This is an archive of the discontinued LLVM Phabricator instance.

Differential D114349

[ARM] CSINC/CSINV patterns from CMOV
ClosedPublic

Authored by dmgreen on Nov 22 2021, 12:10 AM.

Details

Reviewers
SjoerdMeijer
samtebbs
ostannard
simon_tatham
sjarus
Commits
rG5c64d8ef8cc0: [ARM] CSINC/CSINV patterns from CMOV
Summary

We sometimes end up generating CMOV with constant operands that can be simplified to CSINC or CSINV under Arm-8.1m. This adds some simple patterns for them.

Diff Detail

Event Timeline

dmgreen created this revision.Nov 22 2021, 12:10 AM
Herald added a reviewer: sjarus. · View Herald TranscriptNov 22 2021, 12:10 AM
Herald added subscribers: armkevincheng, jsmolens, eric-k256 and 2 others. · View Herald Transcript
dmgreen requested review of this revision.Nov 22 2021, 12:10 AM
Herald added a project: Restricted Project. · View Herald TranscriptNov 22 2021, 12:10 AM
Harbormaster completed remote builds in B135345: Diff 388806.Nov 22 2021, 12:11 AM
dmgreen added a parent revision: D114013: [ARM] Fold away unnecessary CSET/CMPZ.Nov 23 2021, 3:04 AM
dmgreen added a child revision: D114449: [ARM] Make v2i1 legal.Nov 23 2021, 2:13 PM
samtebbs accepted this revision.Nov 26 2021, 7:12 AM

👍

This revision is now accepted and ready to land.Nov 26 2021, 7:12 AM
This revision was landed with ongoing or failed builds.Nov 27 2021, 12:21 PM
Closed by commit rG5c64d8ef8cc0: [ARM] CSINC/CSINV patterns from CMOV (authored by dmgreen). · Explain Why
This revision was automatically updated to reflect the committed changes.
dmgreen added a commit: rG5c64d8ef8cc0: [ARM] CSINC/CSINV patterns from CMOV.

Revision Contents

PathSize
llvm/
lib/
Target/
ARM/
9 lines
test/
CodeGen/
Thumb2/
84 lines
552 lines
1371 lines
464 lines
116 lines
368 lines
64 lines
40 lines
128 lines
50 lines
1236 lines
2366 lines
2304 lines
56 lines
72 lines
21 lines
152 lines
152 lines

Diff 388806

llvm/lib/Target/ARM/ARMInstrThumb2.td

Show First 20 Lines • Show All 5,601 Lines • ▼ Show 20 Lines multiclass CSPats<SDNode Node, Instruction Insn> {
def : T2Pat<(Node (i32 0), (i32 0), imm0_31:$imm), def : T2Pat<(Node (i32 0), (i32 0), imm0_31:$imm),
(Insn ZR, ZR, imm0_31:$imm)>; (Insn ZR, ZR, imm0_31:$imm)>;
} }
defm : CSPats<ARMcsinc, t2CSINC>; defm : CSPats<ARMcsinc, t2CSINC>;
defm : CSPats<ARMcsinv, t2CSINV>; defm : CSPats<ARMcsinv, t2CSINV>;
defm : CSPats<ARMcsneg, t2CSNEG>; defm : CSPats<ARMcsneg, t2CSNEG>;
def : T2Pat<(ARMcmov (i32 1), (i32 0), cmovpred:$imm),
(t2CSINC ZR, ZR, imm0_31:$imm)>;
def : T2Pat<(ARMcmov (i32 -1), (i32 0), cmovpred:$imm),
(t2CSINV ZR, ZR, imm0_31:$imm)>;
def : T2Pat<(ARMcmov (i32 0), (i32 1), cmovpred:$imm),
(t2CSINC ZR, ZR, (inv_cond_XFORM imm:$imm))>;
def : T2Pat<(ARMcmov (i32 0), (i32 -1), cmovpred:$imm),
(t2CSINV ZR, ZR, (inv_cond_XFORM imm:$imm))>;
multiclass ModifiedV8_1CSEL<Instruction Insn, dag modvalue> { multiclass ModifiedV8_1CSEL<Instruction Insn, dag modvalue> {
def : T2Pat<(ARMcmov modvalue, GPRwithZR:$tval, cmovpred:$imm), def : T2Pat<(ARMcmov modvalue, GPRwithZR:$tval, cmovpred:$imm),
(Insn GPRwithZR:$tval, GPRwithZR:$fval, imm0_31:$imm)>; (Insn GPRwithZR:$tval, GPRwithZR:$fval, imm0_31:$imm)>;
def : T2Pat<(ARMcmov GPRwithZR:$tval, modvalue, cmovpred:$imm), def : T2Pat<(ARMcmov GPRwithZR:$tval, modvalue, cmovpred:$imm),
(Insn GPRwithZR:$tval, GPRwithZR:$fval, (Insn GPRwithZR:$tval, GPRwithZR:$fval,
(i32 (inv_cond_XFORM imm:$imm)))>; (i32 (inv_cond_XFORM imm:$imm)))>;
} }
defm : ModifiedV8_1CSEL<t2CSINC, (add rGPR:$fval, 1)>; defm : ModifiedV8_1CSEL<t2CSINC, (add rGPR:$fval, 1)>;
Show All 21 Lines

llvm/test/CodeGen/Thumb2/active_lane_mask.ll

; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -mtriple=thumbv8.1m.main -mattr=+mve %s -o - | FileCheck %s ; RUN: llc -mtriple=thumbv8.1m.main -mattr=+mve %s -o - | FileCheck %s
define <2 x i64> @v2i64(i32 %index, i32 %TC, <2 x i64> %V1, <2 x i64> %V2) { define <2 x i64> @v2i64(i32 %index, i32 %TC, <2 x i64> %V1, <2 x i64> %V2) {
; CHECK-LABEL: v2i64: ; CHECK-LABEL: v2i64:
; CHECK: @ %bb.0: ; CHECK: @ %bb.0:
; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, r9, lr} ; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, lr}
; CHECK-NEXT: sub sp, #4
; CHECK-NEXT: vmov q0[2], q0[0], r0, r0 ; CHECK-NEXT: vmov q0[2], q0[0], r0, r0
; CHECK-NEXT: vmov.i64 q1, #0xffffffff ; CHECK-NEXT: vmov.i64 q1, #0xffffffff
; CHECK-NEXT: vand q0, q0, q1 ; CHECK-NEXT: vand q0, q0, q1
; CHECK-NEXT: vmov q2[2], q2[0], r1, r1 ; CHECK-NEXT: vmov q2[2], q2[0], r1, r1
; CHECK-NEXT: vmov r0, r12, d1 ; CHECK-NEXT: vmov r0, r12, d1
; CHECK-NEXT: movs r7, #0 ; CHECK-NEXT: vmov lr, s0
; CHECK-NEXT: adds.w r8, r0, #1 ; CHECK-NEXT: adds r0, #1
; CHECK-NEXT: adc lr, r12, #0 ; CHECK-NEXT: vmov q0[2], q0[0], lr, r0
; CHECK-NEXT: vmov r12, s0 ; CHECK-NEXT: adc r12, r12, #0
; CHECK-NEXT: vmov q0[2], q0[0], r12, r8
; CHECK-NEXT: vand q0, q0, q1 ; CHECK-NEXT: vand q0, q0, q1
; CHECK-NEXT: vand q1, q2, q1 ; CHECK-NEXT: vand q1, q2, q1
; CHECK-NEXT: vmov r4, r5, d1 ; CHECK-NEXT: vmov r4, r5, d1
; CHECK-NEXT: vmov.i32 q2, #0x1 ; CHECK-NEXT: vmov.i32 q2, #0x1
; CHECK-NEXT: vmov r1, r6, d3 ; CHECK-NEXT: vmov r1, r6, d3
; CHECK-NEXT: eors r0, r4
; CHECK-NEXT: subs r1, r4, r1 ; CHECK-NEXT: subs r1, r4, r1
; CHECK-NEXT: sbcs.w r1, r5, r6 ; CHECK-NEXT: sbcs.w r1, r5, r6
; CHECK-NEXT: vmov r5, r6, d0 ; CHECK-NEXT: vmov r5, r6, d0
; CHECK-NEXT: mov.w r1, #0 ; CHECK-NEXT: cset r1, lo
; CHECK-NEXT: it lo
; CHECK-NEXT: movlo r1, #1
; CHECK-NEXT: cmp r1, #0 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: vmov r0, r1, d2 ; CHECK-NEXT: vmov r7, r1, d2
; CHECK-NEXT: csetm r9, ne ; CHECK-NEXT: csetm r8, ne
; CHECK-NEXT: subs r0, r5, r0 ; CHECK-NEXT: subs r7, r5, r7
; CHECK-NEXT: sbcs.w r0, r6, r1 ; CHECK-NEXT: sbcs.w r1, r6, r1
; CHECK-NEXT: it lo ; CHECK-NEXT: cset r1, lo
; CHECK-NEXT: movlo r7, #1 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: cmp r7, #0 ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: orrs.w r0, r0, r12
; CHECK-NEXT: vmov q0[2], q0[0], r0, r9
; CHECK-NEXT: eor.w r0, r4, r8
; CHECK-NEXT: orrs.w r0, r0, lr
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: teq.w r5, r12 ; CHECK-NEXT: teq.w r5, lr
; CHECK-NEXT: vmov q0[2], q0[0], r1, r8
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: vmov q1[2], q1[0], r1, r0 ; CHECK-NEXT: vmov q1[2], q1[0], r1, r0
; CHECK-NEXT: veor q1, q1, q2 ; CHECK-NEXT: veor q1, q1, q2
; CHECK-NEXT: vldr d5, [sp, #32] ; CHECK-NEXT: vldr d5, [sp, #24]
; CHECK-NEXT: vand q0, q1, q0 ; CHECK-NEXT: vand q0, q1, q0
; CHECK-NEXT: vmov d4, r2, r3 ; CHECK-NEXT: vmov d4, r2, r3
; CHECK-NEXT: vmov r0, s2 ; CHECK-NEXT: vmov r0, s2
; CHECK-NEXT: vmov r1, s0 ; CHECK-NEXT: vmov r1, s0
; CHECK-NEXT: and r0, r0, #1 ; CHECK-NEXT: and r0, r0, #1
; CHECK-NEXT: and r1, r1, #1 ; CHECK-NEXT: and r1, r1, #1
; CHECK-NEXT: rsbs r0, r0, #0 ; CHECK-NEXT: rsbs r0, r0, #0
; CHECK-NEXT: rsbs r1, r1, #0 ; CHECK-NEXT: rsbs r1, r1, #0
; CHECK-NEXT: vmov q0[2], q0[0], r1, r0 ; CHECK-NEXT: vmov q0[2], q0[0], r1, r0
; CHECK-NEXT: vmov q0[3], q0[1], r1, r0 ; CHECK-NEXT: vmov q0[3], q0[1], r1, r0
; CHECK-NEXT: add r0, sp, #40 ; CHECK-NEXT: add r0, sp, #32
; CHECK-NEXT: vldrw.u32 q1, [r0] ; CHECK-NEXT: vldrw.u32 q1, [r0]
; CHECK-NEXT: vbic q1, q1, q0 ; CHECK-NEXT: vbic q1, q1, q0
; CHECK-NEXT: vand q0, q2, q0 ; CHECK-NEXT: vand q0, q2, q0
; CHECK-NEXT: vorr q0, q0, q1 ; CHECK-NEXT: vorr q0, q0, q1
; CHECK-NEXT: vmov r0, r1, d0 ; CHECK-NEXT: vmov r0, r1, d0
; CHECK-NEXT: vmov r2, r3, d1 ; CHECK-NEXT: vmov r2, r3, d1
; CHECK-NEXT: add sp, #4 ; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, pc}
; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, pc}
%active.lane.mask = call <2 x i1> @llvm.get.active.lane.mask.v2i1.i32(i32 %index, i32 %TC) %active.lane.mask = call <2 x i1> @llvm.get.active.lane.mask.v2i1.i32(i32 %index, i32 %TC)
%select = select <2 x i1> %active.lane.mask, <2 x i64> %V1, <2 x i64> %V2 %select = select <2 x i1> %active.lane.mask, <2 x i64> %V1, <2 x i64> %V2
ret <2 x i64> %select ret <2 x i64> %select
} }
define <4 x i32> @v4i32(i32 %index, i32 %TC, <4 x i32> %V1, <4 x i32> %V2) { define <4 x i32> @v4i32(i32 %index, i32 %TC, <4 x i32> %V1, <4 x i32> %V2) {
; CHECK-LABEL: v4i32: ; CHECK-LABEL: v4i32:
; CHECK: @ %bb.0: ; CHECK: @ %bb.0:
▲ Show 20 Lines • Show All 387 Lines • ▼ Show 20 Lines
; CHECK-NEXT: vmov q1[2], q1[0], r2, r2 ; CHECK-NEXT: vmov q1[2], q1[0], r2, r2
; CHECK-NEXT: bic r0, r0, #1 ; CHECK-NEXT: bic r0, r0, #1
; CHECK-NEXT: adr r2, .LCPI5_0 ; CHECK-NEXT: adr r2, .LCPI5_0
; CHECK-NEXT: subs r0, #2 ; CHECK-NEXT: subs r0, #2
; CHECK-NEXT: movs r3, #1 ; CHECK-NEXT: movs r3, #1
; CHECK-NEXT: vmov.i64 q0, #0xffffffff ; CHECK-NEXT: vmov.i64 q0, #0xffffffff
; CHECK-NEXT: vldrw.u32 q2, [r2] ; CHECK-NEXT: vldrw.u32 q2, [r2]
; CHECK-NEXT: add.w lr, r3, r0, lsr #1 ; CHECK-NEXT: add.w lr, r3, r0, lsr #1
; CHECK-NEXT: mov.w r12, #0 ; CHECK-NEXT: mov.w r8, #0
; CHECK-NEXT: vand q1, q1, q0 ; CHECK-NEXT: vand q1, q1, q0
; CHECK-NEXT: .LBB5_2: @ %vector.body ; CHECK-NEXT: .LBB5_2: @ %vector.body
; CHECK-NEXT: @ =>This Inner Loop Header: Depth=1 ; CHECK-NEXT: @ =>This Inner Loop Header: Depth=1
; CHECK-NEXT: vmov q3[2], q3[0], r12, r12 ; CHECK-NEXT: vmov q3[2], q3[0], r8, r8
; CHECK-NEXT: vmov r6, r7, d3 ; CHECK-NEXT: vmov r6, r7, d3
; CHECK-NEXT: vand q3, q3, q0 ; CHECK-NEXT: vand q3, q3, q0
; CHECK-NEXT: add.w r12, r12, #2 ; CHECK-NEXT: add.w r8, r8, #2
; CHECK-NEXT: vmov r2, r3, d7 ; CHECK-NEXT: vmov r2, r3, d7
; CHECK-NEXT: vmov r9, s12 ; CHECK-NEXT: vmov r9, s12
; CHECK-NEXT: adds r0, r2, #1 ; CHECK-NEXT: adds r2, #1
; CHECK-NEXT: vmov q3[2], q3[0], r9, r0 ; CHECK-NEXT: vmov q3[2], q3[0], r9, r2
; CHECK-NEXT: adc r8, r3, #0 ; CHECK-NEXT: adc r12, r3, #0
; CHECK-NEXT: vand q3, q3, q0 ; CHECK-NEXT: vand q3, q3, q0
; CHECK-NEXT: vmov r3, r2, d2 ; CHECK-NEXT: vmov r0, r3, d2
; CHECK-NEXT: vmov r4, r5, d7 ; CHECK-NEXT: vmov r4, r5, d7
; CHECK-NEXT: subs r6, r4, r6 ; CHECK-NEXT: subs r6, r4, r6
; CHECK-NEXT: eor.w r0, r0, r4
; CHECK-NEXT: sbcs r5, r7 ; CHECK-NEXT: sbcs r5, r7
; CHECK-NEXT: vmov r6, r7, d6 ; CHECK-NEXT: vmov r6, r7, d6
; CHECK-NEXT: mov.w r5, #0 ; CHECK-NEXT: cset r5, lo
; CHECK-NEXT: it lo
; CHECK-NEXT: movlo r5, #1
; CHECK-NEXT: cmp r5, #0 ; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: csetm r5, ne ; CHECK-NEXT: csetm r5, ne
; CHECK-NEXT: subs r3, r6, r3 ; CHECK-NEXT: subs r0, r6, r0
; CHECK-NEXT: sbcs.w r2, r7, r2 ; CHECK-NEXT: sbcs.w r0, r7, r3
; CHECK-NEXT: mov.w r2, #0 ; CHECK-NEXT: cset r0, lo
; CHECK-NEXT: it lo ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: movlo r2, #1 ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: cmp r2, #0 ; CHECK-NEXT: vmov q3[2], q3[0], r0, r5
; CHECK-NEXT: csetm r2, ne ; CHECK-NEXT: vmov q3[3], q3[1], r0, r5
; CHECK-NEXT: orrs.w r0, r0, r8 ; CHECK-NEXT: eor.w r0, r4, r2
; CHECK-NEXT: vmov q3[2], q3[0], r2, r5 ; CHECK-NEXT: orrs.w r0, r0, r12
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: teq.w r6, r9 ; CHECK-NEXT: teq.w r6, r9
; CHECK-NEXT: vmov q3[3], q3[1], r2, r5
; CHECK-NEXT: csetm r2, ne ; CHECK-NEXT: csetm r2, ne
; CHECK-NEXT: vmov q4[2], q4[0], r2, r0 ; CHECK-NEXT: vmov q4[2], q4[0], r2, r0
; CHECK-NEXT: vmov q4[3], q4[1], r2, r0 ; CHECK-NEXT: vmov q4[3], q4[1], r2, r0
; CHECK-NEXT: veor q4, q4, q2 ; CHECK-NEXT: veor q4, q4, q2
; CHECK-NEXT: vand q4, q4, q3 ; CHECK-NEXT: vand q4, q4, q3
; CHECK-NEXT: @ implicit-def: $q3 ; CHECK-NEXT: @ implicit-def: $q3
; CHECK-NEXT: vmov r2, s16 ; CHECK-NEXT: vmov r2, s16
; CHECK-NEXT: vmov r0, s18 ; CHECK-NEXT: vmov r0, s18
▲ Show 20 Lines • Show All 74 Lines • Show Last 20 Lines

llvm/test/CodeGen/Thumb2/mve-fmas.ll

Show First 20 Lines • Show All 398 Lines • ▼ Show 20 Lines
; CHECK-MVE-VMLA-NEXT: bx lr ; CHECK-MVE-VMLA-NEXT: bx lr
; ;
; CHECK-MVE-LABEL: vfma16_v1_pred: ; CHECK-MVE-LABEL: vfma16_v1_pred:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s14, s4 ; CHECK-MVE-NEXT: vmovx.f16 s14, s4
; CHECK-MVE-NEXT: vmovx.f16 s13, s0 ; CHECK-MVE-NEXT: vmovx.f16 s13, s0
; CHECK-MVE-NEXT: vcmp.f16 s14, #0 ; CHECK-MVE-NEXT: vcmp.f16 s14, #0
; CHECK-MVE-NEXT: vmovx.f16 s12, s8 ; CHECK-MVE-NEXT: vmovx.f16 s12, s8
; CHECK-MVE-NEXT: vmov.f32 s15, s13
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmov.f32 s15, s13
; CHECK-MVE-NEXT: vmla.f16 s15, s14, s12 ; CHECK-MVE-NEXT: vmla.f16 s15, s14, s12
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmov.f32 s14, s0 ; CHECK-MVE-NEXT: vmov.f32 s14, s0
; CHECK-MVE-NEXT: vmla.f16 s14, s4, s8
; CHECK-MVE-NEXT: vmovx.f16 s8, s5
; CHECK-MVE-NEXT: vmovx.f16 s4, s9
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s12, s13, s15 ; CHECK-MVE-NEXT: vseleq.f16 s12, s13, s15
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vcmp.f16 s8, #0
; CHECK-MVE-NEXT: vmla.f16 s14, s4, s8 ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s0, s14 ; CHECK-MVE-NEXT: vseleq.f16 s0, s0, s14
; CHECK-MVE-NEXT: vmovx.f16 s8, s5 ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vins.f16 s0, s12 ; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: vmovx.f16 s12, s1 ; CHECK-MVE-NEXT: vmovx.f16 s12, s1
; CHECK-MVE-NEXT: vcmp.f16 s8, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s9
; CHECK-MVE-NEXT: vmov.f32 s14, s12 ; CHECK-MVE-NEXT: vmov.f32 s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmla.f16 s14, s8, s4
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s5, #0 ; CHECK-MVE-NEXT: vcmp.f16 s5, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s12, s14 ; CHECK-MVE-NEXT: vmla.f16 s14, s8, s4
; CHECK-MVE-NEXT: vmov.f32 s8, s1 ; CHECK-MVE-NEXT: vmov.f32 s8, s1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmla.f16 s8, s5, s9 ; CHECK-MVE-NEXT: vmla.f16 s8, s5, s9
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: vseleq.f16 s4, s12, s14
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s12, s2
; CHECK-MVE-NEXT: vmov.f32 s14, s12
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s1, s8 ; CHECK-MVE-NEXT: vseleq.f16 s1, s1, s8
; CHECK-MVE-NEXT: vmovx.f16 s8, s6 ; CHECK-MVE-NEXT: vmovx.f16 s8, s6
; CHECK-MVE-NEXT: vmovx.f16 s12, s2
; CHECK-MVE-NEXT: vcmp.f16 s8, #0 ; CHECK-MVE-NEXT: vcmp.f16 s8, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s10 ; CHECK-MVE-NEXT: vmovx.f16 s4, s10
; CHECK-MVE-NEXT: vmov.f32 s14, s12
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: vmla.f16 s14, s8, s4 ; CHECK-MVE-NEXT: vmla.f16 s14, s8, s4
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s6, #0 ; CHECK-MVE-NEXT: vcmp.f16 s6, #0
; CHECK-MVE-NEXT: vmov.f32 s8, s2 ; CHECK-MVE-NEXT: vmov.f32 s8, s2
; CHECK-MVE-NEXT: vseleq.f16 s4, s12, s14
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmla.f16 s8, s6, s10 ; CHECK-MVE-NEXT: vmla.f16 s8, s6, s10
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s6, s7 ; CHECK-MVE-NEXT: vmovx.f16 s6, s7
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s12, s14
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f16 s6, #0
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s2, s8 ; CHECK-MVE-NEXT: vseleq.f16 s2, s2, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s3 ; CHECK-MVE-NEXT: vmovx.f16 s8, s3
; CHECK-MVE-NEXT: vcmp.f16 s6, #0
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s11 ; CHECK-MVE-NEXT: vmovx.f16 s4, s11
; CHECK-MVE-NEXT: vmov.f32 s10, s8 ; CHECK-MVE-NEXT: vmov.f32 s10, s8
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: vmla.f16 s10, s6, s4 ; CHECK-MVE-NEXT: vmla.f16 s10, s6, s4
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s7, #0 ; CHECK-MVE-NEXT: vcmp.f16 s7, #0
; CHECK-MVE-NEXT: vmov.f32 s6, s3 ; CHECK-MVE-NEXT: vmov.f32 s6, s3
; CHECK-MVE-NEXT: vmla.f16 s6, s7, s11
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s10 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s10
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r0, #1
; CHECK-MVE-NEXT: vmla.f16 s6, s7, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s3, s6 ; CHECK-MVE-NEXT: vseleq.f16 s3, s3, s6
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
entry: entry:
%0 = fmul <8 x half> %src2, %src3 %0 = fmul <8 x half> %src2, %src3
%1 = fadd <8 x half> %src1, %0 %1 = fadd <8 x half> %src1, %0
%c = fcmp olt <8 x half> %src2, zeroinitializer %c = fcmp olt <8 x half> %src2, zeroinitializer
Show All 16 Lines
; CHECK-MVE-VMLA-NEXT: bx lr ; CHECK-MVE-VMLA-NEXT: bx lr
; ;
; CHECK-MVE-LABEL: vfma16_v2_pred: ; CHECK-MVE-LABEL: vfma16_v2_pred:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s14, s4 ; CHECK-MVE-NEXT: vmovx.f16 s14, s4
; CHECK-MVE-NEXT: vmovx.f16 s13, s0 ; CHECK-MVE-NEXT: vmovx.f16 s13, s0
; CHECK-MVE-NEXT: vcmp.f16 s14, #0 ; CHECK-MVE-NEXT: vcmp.f16 s14, #0
; CHECK-MVE-NEXT: vmovx.f16 s12, s8 ; CHECK-MVE-NEXT: vmovx.f16 s12, s8
; CHECK-MVE-NEXT: vmov.f32 s15, s13
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmov.f32 s15, s13
; CHECK-MVE-NEXT: vmla.f16 s15, s14, s12 ; CHECK-MVE-NEXT: vmla.f16 s15, s14, s12
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmov.f32 s14, s0 ; CHECK-MVE-NEXT: vmov.f32 s14, s0
; CHECK-MVE-NEXT: vmla.f16 s14, s4, s8
; CHECK-MVE-NEXT: vmovx.f16 s8, s5
; CHECK-MVE-NEXT: vmovx.f16 s4, s9
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s12, s13, s15 ; CHECK-MVE-NEXT: vseleq.f16 s12, s13, s15
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vcmp.f16 s8, #0
; CHECK-MVE-NEXT: vmla.f16 s14, s4, s8 ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s0, s14 ; CHECK-MVE-NEXT: vseleq.f16 s0, s0, s14
; CHECK-MVE-NEXT: vmovx.f16 s8, s5 ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vins.f16 s0, s12 ; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: vmovx.f16 s12, s1 ; CHECK-MVE-NEXT: vmovx.f16 s12, s1
; CHECK-MVE-NEXT: vcmp.f16 s8, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s9
; CHECK-MVE-NEXT: vmov.f32 s14, s12 ; CHECK-MVE-NEXT: vmov.f32 s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmla.f16 s14, s8, s4
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s5, #0 ; CHECK-MVE-NEXT: vcmp.f16 s5, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s12, s14 ; CHECK-MVE-NEXT: vmla.f16 s14, s8, s4
; CHECK-MVE-NEXT: vmov.f32 s8, s1 ; CHECK-MVE-NEXT: vmov.f32 s8, s1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmla.f16 s8, s5, s9 ; CHECK-MVE-NEXT: vmla.f16 s8, s5, s9
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: vseleq.f16 s4, s12, s14
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s12, s2
; CHECK-MVE-NEXT: vmov.f32 s14, s12
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s1, s8 ; CHECK-MVE-NEXT: vseleq.f16 s1, s1, s8
; CHECK-MVE-NEXT: vmovx.f16 s8, s6 ; CHECK-MVE-NEXT: vmovx.f16 s8, s6
; CHECK-MVE-NEXT: vmovx.f16 s12, s2
; CHECK-MVE-NEXT: vcmp.f16 s8, #0 ; CHECK-MVE-NEXT: vcmp.f16 s8, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s10 ; CHECK-MVE-NEXT: vmovx.f16 s4, s10
; CHECK-MVE-NEXT: vmov.f32 s14, s12
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: vmla.f16 s14, s8, s4 ; CHECK-MVE-NEXT: vmla.f16 s14, s8, s4
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s6, #0 ; CHECK-MVE-NEXT: vcmp.f16 s6, #0
; CHECK-MVE-NEXT: vmov.f32 s8, s2 ; CHECK-MVE-NEXT: vmov.f32 s8, s2
; CHECK-MVE-NEXT: vseleq.f16 s4, s12, s14
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmla.f16 s8, s6, s10 ; CHECK-MVE-NEXT: vmla.f16 s8, s6, s10
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s6, s7 ; CHECK-MVE-NEXT: vmovx.f16 s6, s7
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s12, s14
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f16 s6, #0
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s2, s8 ; CHECK-MVE-NEXT: vseleq.f16 s2, s2, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s3 ; CHECK-MVE-NEXT: vmovx.f16 s8, s3
; CHECK-MVE-NEXT: vcmp.f16 s6, #0
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s11 ; CHECK-MVE-NEXT: vmovx.f16 s4, s11
; CHECK-MVE-NEXT: vmov.f32 s10, s8 ; CHECK-MVE-NEXT: vmov.f32 s10, s8
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: vmla.f16 s10, s6, s4 ; CHECK-MVE-NEXT: vmla.f16 s10, s6, s4
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s7, #0 ; CHECK-MVE-NEXT: vcmp.f16 s7, #0
; CHECK-MVE-NEXT: vmov.f32 s6, s3 ; CHECK-MVE-NEXT: vmov.f32 s6, s3
; CHECK-MVE-NEXT: vmla.f16 s6, s7, s11
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s10 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s10
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r0, #1
; CHECK-MVE-NEXT: vmla.f16 s6, s7, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s3, s6 ; CHECK-MVE-NEXT: vseleq.f16 s3, s3, s6
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
entry: entry:
%0 = fmul <8 x half> %src2, %src3 %0 = fmul <8 x half> %src2, %src3
%1 = fadd <8 x half> %0, %src1 %1 = fadd <8 x half> %0, %src1
%c = fcmp olt <8 x half> %src2, zeroinitializer %c = fcmp olt <8 x half> %src2, zeroinitializer
Show All 16 Lines
; CHECK-MVE-VMLA-NEXT: bx lr ; CHECK-MVE-VMLA-NEXT: bx lr
; ;
; CHECK-MVE-LABEL: vfms16_pred: ; CHECK-MVE-LABEL: vfms16_pred:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s14, s4 ; CHECK-MVE-NEXT: vmovx.f16 s14, s4
; CHECK-MVE-NEXT: vmovx.f16 s13, s0 ; CHECK-MVE-NEXT: vmovx.f16 s13, s0
; CHECK-MVE-NEXT: vcmp.f16 s14, #0 ; CHECK-MVE-NEXT: vcmp.f16 s14, #0
; CHECK-MVE-NEXT: vmovx.f16 s12, s8 ; CHECK-MVE-NEXT: vmovx.f16 s12, s8
; CHECK-MVE-NEXT: vmov.f32 s15, s13
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmov.f32 s15, s13
; CHECK-MVE-NEXT: vmls.f16 s15, s14, s12 ; CHECK-MVE-NEXT: vmls.f16 s15, s14, s12
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmov.f32 s14, s0 ; CHECK-MVE-NEXT: vmov.f32 s14, s0
; CHECK-MVE-NEXT: vmls.f16 s14, s4, s8
; CHECK-MVE-NEXT: vmovx.f16 s8, s5
; CHECK-MVE-NEXT: vmovx.f16 s4, s9
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s12, s13, s15 ; CHECK-MVE-NEXT: vseleq.f16 s12, s13, s15
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vcmp.f16 s8, #0
; CHECK-MVE-NEXT: vmls.f16 s14, s4, s8 ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s0, s14 ; CHECK-MVE-NEXT: vseleq.f16 s0, s0, s14
; CHECK-MVE-NEXT: vmovx.f16 s8, s5 ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vins.f16 s0, s12 ; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: vmovx.f16 s12, s1 ; CHECK-MVE-NEXT: vmovx.f16 s12, s1
; CHECK-MVE-NEXT: vcmp.f16 s8, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s9
; CHECK-MVE-NEXT: vmov.f32 s14, s12 ; CHECK-MVE-NEXT: vmov.f32 s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmls.f16 s14, s8, s4
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s5, #0 ; CHECK-MVE-NEXT: vcmp.f16 s5, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s12, s14 ; CHECK-MVE-NEXT: vmls.f16 s14, s8, s4
; CHECK-MVE-NEXT: vmov.f32 s8, s1 ; CHECK-MVE-NEXT: vmov.f32 s8, s1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmls.f16 s8, s5, s9 ; CHECK-MVE-NEXT: vmls.f16 s8, s5, s9
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: vseleq.f16 s4, s12, s14
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s12, s2
; CHECK-MVE-NEXT: vmov.f32 s14, s12
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s1, s8 ; CHECK-MVE-NEXT: vseleq.f16 s1, s1, s8
; CHECK-MVE-NEXT: vmovx.f16 s8, s6 ; CHECK-MVE-NEXT: vmovx.f16 s8, s6
; CHECK-MVE-NEXT: vmovx.f16 s12, s2
; CHECK-MVE-NEXT: vcmp.f16 s8, #0 ; CHECK-MVE-NEXT: vcmp.f16 s8, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s10 ; CHECK-MVE-NEXT: vmovx.f16 s4, s10
; CHECK-MVE-NEXT: vmov.f32 s14, s12
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: vmls.f16 s14, s8, s4 ; CHECK-MVE-NEXT: vmls.f16 s14, s8, s4
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s6, #0 ; CHECK-MVE-NEXT: vcmp.f16 s6, #0
; CHECK-MVE-NEXT: vmov.f32 s8, s2 ; CHECK-MVE-NEXT: vmov.f32 s8, s2
; CHECK-MVE-NEXT: vseleq.f16 s4, s12, s14
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmls.f16 s8, s6, s10 ; CHECK-MVE-NEXT: vmls.f16 s8, s6, s10
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s6, s7 ; CHECK-MVE-NEXT: vmovx.f16 s6, s7
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s12, s14
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f16 s6, #0
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s2, s8 ; CHECK-MVE-NEXT: vseleq.f16 s2, s2, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s3 ; CHECK-MVE-NEXT: vmovx.f16 s8, s3
; CHECK-MVE-NEXT: vcmp.f16 s6, #0
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s11 ; CHECK-MVE-NEXT: vmovx.f16 s4, s11
; CHECK-MVE-NEXT: vmov.f32 s10, s8 ; CHECK-MVE-NEXT: vmov.f32 s10, s8
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: vmls.f16 s10, s6, s4 ; CHECK-MVE-NEXT: vmls.f16 s10, s6, s4
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s7, #0 ; CHECK-MVE-NEXT: vcmp.f16 s7, #0
; CHECK-MVE-NEXT: vmov.f32 s6, s3 ; CHECK-MVE-NEXT: vmov.f32 s6, s3
; CHECK-MVE-NEXT: vmls.f16 s6, s7, s11
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s10 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s10
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r0, #1
; CHECK-MVE-NEXT: vmls.f16 s6, s7, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s3, s6 ; CHECK-MVE-NEXT: vseleq.f16 s3, s3, s6
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
entry: entry:
%0 = fmul <8 x half> %src2, %src3 %0 = fmul <8 x half> %src2, %src3
%1 = fsub <8 x half> %src1, %0 %1 = fsub <8 x half> %src1, %0
%c = fcmp olt <8 x half> %src2, zeroinitializer %c = fcmp olt <8 x half> %src2, zeroinitializer
Show All 21 Lines
; CHECK-MVE-VMLA-NEXT: bx lr ; CHECK-MVE-VMLA-NEXT: bx lr
; ;
; CHECK-MVE-LABEL: vfmar16_pred: ; CHECK-MVE-LABEL: vfmar16_pred:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s10, s4 ; CHECK-MVE-NEXT: vmovx.f16 s10, s4
; CHECK-MVE-NEXT: vmovx.f16 s12, s0 ; CHECK-MVE-NEXT: vmovx.f16 s12, s0
; CHECK-MVE-NEXT: vcmp.f16 s10, #0 ; CHECK-MVE-NEXT: vcmp.f16 s10, #0
; CHECK-MVE-NEXT: vcvtb.f16.f32 s8, s8 ; CHECK-MVE-NEXT: vcvtb.f16.f32 s8, s8
; CHECK-MVE-NEXT: vmov.f32 s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmov.f32 s14, s12
; CHECK-MVE-NEXT: vmla.f16 s14, s10, s8 ; CHECK-MVE-NEXT: vmla.f16 s14, s10, s8
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s10, s12, s14 ; CHECK-MVE-NEXT: vseleq.f16 s10, s12, s14
; CHECK-MVE-NEXT: vmov.f32 s12, s0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmov.f32 s12, s0
; CHECK-MVE-NEXT: vmla.f16 s12, s4, s8 ; CHECK-MVE-NEXT: vmla.f16 s12, s4, s8
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s0, s12
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s0, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vins.f16 s0, s10 ; CHECK-MVE-NEXT: vins.f16 s0, s10
; CHECK-MVE-NEXT: vmovx.f16 s10, s1 ; CHECK-MVE-NEXT: vmovx.f16 s10, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmov.f32 s12, s10 ; CHECK-MVE-NEXT: vmov.f32 s12, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmla.f16 s12, s4, s8
; CHECK-MVE-NEXT: vcmp.f16 s5, #0 ; CHECK-MVE-NEXT: vcmp.f16 s5, #0
; CHECK-MVE-NEXT: vmla.f16 s12, s4, s8
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s10, s12 ; CHECK-MVE-NEXT: vseleq.f16 s4, s10, s12
; CHECK-MVE-NEXT: vmov.f32 s10, s1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmov.f32 s10, s1
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmla.f16 s10, s5, s8 ; CHECK-MVE-NEXT: vmla.f16 s10, s5, s8
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s1, s10 ; CHECK-MVE-NEXT: vseleq.f16 s1, s1, s10
; CHECK-MVE-NEXT: vmovx.f16 s10, s2 ; CHECK-MVE-NEXT: vmovx.f16 s10, s2
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmov.f32 s12, s10 ; CHECK-MVE-NEXT: vmov.f32 s12, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmla.f16 s12, s4, s8 ; CHECK-MVE-NEXT: vmla.f16 s12, s4, s8
; CHECK-MVE-NEXT: vcmp.f16 s6, #0 ; CHECK-MVE-NEXT: vcmp.f16 s6, #0
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s10, s12 ; CHECK-MVE-NEXT: vseleq.f16 s4, s10, s12
; CHECK-MVE-NEXT: vmov.f32 s10, s2
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmov.f32 s10, s2
; CHECK-MVE-NEXT: vmla.f16 s10, s6, s8 ; CHECK-MVE-NEXT: vmla.f16 s10, s6, s8
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s2, s10
; CHECK-MVE-NEXT: vmovx.f16 s6, s3 ; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s2, s10
; CHECK-MVE-NEXT: vmov.f32 s10, s6
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmov.f32 s10, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: vmla.f16 s10, s4, s8 ; CHECK-MVE-NEXT: vmla.f16 s10, s4, s8
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f16 s7, #0 ; CHECK-MVE-NEXT: vcmp.f16 s7, #0
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s10 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s10
; CHECK-MVE-NEXT: vmov.f32 s6, s3
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: vmov.f32 s6, s3
; CHECK-MVE-NEXT: movmi r0, #1
; CHECK-MVE-NEXT: vmla.f16 s6, s7, s8 ; CHECK-MVE-NEXT: vmla.f16 s6, s7, s8
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s3, s6 ; CHECK-MVE-NEXT: vseleq.f16 s3, s3, s6
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
entry: entry:
%src3 = fptrunc float %src3o to half %src3 = fptrunc float %src3o to half
%i = insertelement <8 x half> undef, half %src3, i32 0 %i = insertelement <8 x half> undef, half %src3, i32 0
%sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer %sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer
Show All 23 Lines
; CHECK-MVE-VMLA-NEXT: bx lr ; CHECK-MVE-VMLA-NEXT: bx lr
; ;
; CHECK-MVE-LABEL: vfma16_pred: ; CHECK-MVE-LABEL: vfma16_pred:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s10, s4 ; CHECK-MVE-NEXT: vmovx.f16 s10, s4
; CHECK-MVE-NEXT: vcvtb.f16.f32 s8, s8 ; CHECK-MVE-NEXT: vcvtb.f16.f32 s8, s8
; CHECK-MVE-NEXT: vcmp.f16 s10, #0 ; CHECK-MVE-NEXT: vcmp.f16 s10, #0
; CHECK-MVE-NEXT: vmovx.f16 s12, s0 ; CHECK-MVE-NEXT: vmovx.f16 s12, s0
; CHECK-MVE-NEXT: vmov.f32 s14, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmov.f32 s14, s8
; CHECK-MVE-NEXT: vmla.f16 s14, s12, s10 ; CHECK-MVE-NEXT: vmla.f16 s14, s12, s10
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s10, s12, s14 ; CHECK-MVE-NEXT: vseleq.f16 s10, s12, s14
; CHECK-MVE-NEXT: vmov.f32 s12, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmov.f32 s12, s8
; CHECK-MVE-NEXT: vmla.f16 s12, s0, s4 ; CHECK-MVE-NEXT: vmla.f16 s12, s0, s4
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: vseleq.f16 s0, s0, s12
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vins.f16 s0, s10 ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s0, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vins.f16 s0, s10
; CHECK-MVE-NEXT: vmovx.f16 s10, s1 ; CHECK-MVE-NEXT: vmovx.f16 s10, s1
; CHECK-MVE-NEXT: vmov.f32 s12, s8 ; CHECK-MVE-NEXT: vmov.f32 s12, s8
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmla.f16 s12, s10, s4
; CHECK-MVE-NEXT: vcmp.f16 s5, #0 ; CHECK-MVE-NEXT: vcmp.f16 s5, #0
; CHECK-MVE-NEXT: vmla.f16 s12, s10, s4
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s10, s12 ; CHECK-MVE-NEXT: vseleq.f16 s4, s10, s12
; CHECK-MVE-NEXT: vmov.f32 s10, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmov.f32 s10, s8
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmla.f16 s10, s1, s5 ; CHECK-MVE-NEXT: vmla.f16 s10, s1, s5
; CHECK-MVE-NEXT: vmov.f32 s12, s8 ; CHECK-MVE-NEXT: vmov.f32 s12, s8
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s1, s10 ; CHECK-MVE-NEXT: vseleq.f16 s1, s1, s10
; CHECK-MVE-NEXT: vmovx.f16 s10, s2 ; CHECK-MVE-NEXT: vmovx.f16 s10, s2
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmla.f16 s12, s10, s4 ; CHECK-MVE-NEXT: vmla.f16 s12, s10, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f16 s6, #0 ; CHECK-MVE-NEXT: vcmp.f16 s6, #0
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s10, s12 ; CHECK-MVE-NEXT: vseleq.f16 s4, s10, s12
; CHECK-MVE-NEXT: vmov.f32 s10, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmov.f32 s10, s8
; CHECK-MVE-NEXT: vmla.f16 s10, s2, s6 ; CHECK-MVE-NEXT: vmla.f16 s10, s2, s6
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: movmi r1, #1 ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s2, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s2, s10
; CHECK-MVE-NEXT: vmov.f32 s10, s8 ; CHECK-MVE-NEXT: vmov.f32 s10, s8
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: vmla.f16 s10, s6, s4 ; CHECK-MVE-NEXT: vmla.f16 s10, s6, s4
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f16 s7, #0 ; CHECK-MVE-NEXT: vcmp.f16 s7, #0
; CHECK-MVE-NEXT: vmla.f16 s8, s3, s7
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s10 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s10
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r0, #1
; CHECK-MVE-NEXT: vmla.f16 s8, s3, s7
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s3, s8 ; CHECK-MVE-NEXT: vseleq.f16 s3, s3, s8
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
entry: entry:
%src3 = fptrunc float %src3o to half %src3 = fptrunc float %src3o to half
%i = insertelement <8 x half> undef, half %src3, i32 0 %i = insertelement <8 x half> undef, half %src3, i32 0
%sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer %sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer
Show All 15 Lines
; CHECK-MVE-VMLA-LABEL: vfma32_v1_pred: ; CHECK-MVE-VMLA-LABEL: vfma32_v1_pred:
; CHECK-MVE-VMLA: @ %bb.0: @ %entry ; CHECK-MVE-VMLA: @ %bb.0: @ %entry
; CHECK-MVE-VMLA-NEXT: vpt.f32 lt, q1, zr ; CHECK-MVE-VMLA-NEXT: vpt.f32 lt, q1, zr
; CHECK-MVE-VMLA-NEXT: vfmat.f32 q0, q1, q2 ; CHECK-MVE-VMLA-NEXT: vfmat.f32 q0, q1, q2
; CHECK-MVE-VMLA-NEXT: bx lr ; CHECK-MVE-VMLA-NEXT: bx lr
; ;
; CHECK-MVE-LABEL: vfma32_v1_pred: ; CHECK-MVE-LABEL: vfma32_v1_pred:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmov.f32 s14, s0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vcmp.f32 s5, #0 ; CHECK-MVE-NEXT: vcmp.f32 s5, #0
; CHECK-MVE-NEXT: movs r2, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s4, #0 ; CHECK-MVE-NEXT: vcmp.f32 s4, #0
; CHECK-MVE-NEXT: vmov.f32 s14, s0
; CHECK-MVE-NEXT: vmov.f32 s12, s1 ; CHECK-MVE-NEXT: vmov.f32 s12, s1
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s7, #0
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmla.f32 s14, s4, s8 ; CHECK-MVE-NEXT: vmla.f32 s14, s4, s8
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vmov.f32 s4, s3 ; CHECK-MVE-NEXT: vmov.f32 s4, s3
; CHECK-MVE-NEXT: mov.w r0, #0
; CHECK-MVE-NEXT: vmov.f32 s8, s2 ; CHECK-MVE-NEXT: vmov.f32 s8, s2
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r3, #1
; CHECK-MVE-NEXT: vcmp.f32 s6, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r0, #1
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vmla.f32 s12, s5, s9 ; CHECK-MVE-NEXT: vmla.f32 s12, s5, s9
; CHECK-MVE-NEXT: vmla.f32 s4, s7, s11 ; CHECK-MVE-NEXT: vmla.f32 s4, s7, s11
; CHECK-MVE-NEXT: vmla.f32 s8, s6, s10 ; CHECK-MVE-NEXT: vmla.f32 s8, s6, s10
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s7, #0
; CHECK-MVE-NEXT: cset r1, mi
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s6, #0
; CHECK-MVE-NEXT: cset r2, mi
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: cset r3, mi
; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s3, s4 ; CHECK-MVE-NEXT: vseleq.f32 s3, s3, s4
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s2, s8 ; CHECK-MVE-NEXT: vseleq.f32 s2, s2, s8
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s1, s12 ; CHECK-MVE-NEXT: vseleq.f32 s1, s1, s12
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s0, s14 ; CHECK-MVE-NEXT: vseleq.f32 s0, s0, s14
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
entry: entry:
%0 = fmul <4 x float> %src2, %src3 %0 = fmul <4 x float> %src2, %src3
%1 = fadd <4 x float> %src1, %0 %1 = fadd <4 x float> %src1, %0
%c = fcmp olt <4 x float> %src2, zeroinitializer %c = fcmp olt <4 x float> %src2, zeroinitializer
%s = select <4 x i1> %c, <4 x float> %1, <4 x float> %src1 %s = select <4 x i1> %c, <4 x float> %1, <4 x float> %src1
ret <4 x float> %s ret <4 x float> %s
Show All 10 Lines
; CHECK-MVE-VMLA-LABEL: vfma32_v2_pred: ; CHECK-MVE-VMLA-LABEL: vfma32_v2_pred:
; CHECK-MVE-VMLA: @ %bb.0: @ %entry ; CHECK-MVE-VMLA: @ %bb.0: @ %entry
; CHECK-MVE-VMLA-NEXT: vpt.f32 lt, q1, zr ; CHECK-MVE-VMLA-NEXT: vpt.f32 lt, q1, zr
; CHECK-MVE-VMLA-NEXT: vfmat.f32 q0, q1, q2 ; CHECK-MVE-VMLA-NEXT: vfmat.f32 q0, q1, q2
; CHECK-MVE-VMLA-NEXT: bx lr ; CHECK-MVE-VMLA-NEXT: bx lr
; ;
; CHECK-MVE-LABEL: vfma32_v2_pred: ; CHECK-MVE-LABEL: vfma32_v2_pred:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmov.f32 s14, s0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vcmp.f32 s5, #0 ; CHECK-MVE-NEXT: vcmp.f32 s5, #0
; CHECK-MVE-NEXT: movs r2, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s4, #0 ; CHECK-MVE-NEXT: vcmp.f32 s4, #0
; CHECK-MVE-NEXT: vmov.f32 s14, s0
; CHECK-MVE-NEXT: vmov.f32 s12, s1 ; CHECK-MVE-NEXT: vmov.f32 s12, s1
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s7, #0
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmla.f32 s14, s4, s8 ; CHECK-MVE-NEXT: vmla.f32 s14, s4, s8
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vmov.f32 s4, s3 ; CHECK-MVE-NEXT: vmov.f32 s4, s3
; CHECK-MVE-NEXT: mov.w r0, #0
; CHECK-MVE-NEXT: vmov.f32 s8, s2 ; CHECK-MVE-NEXT: vmov.f32 s8, s2
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r3, #1
; CHECK-MVE-NEXT: vcmp.f32 s6, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r0, #1
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vmla.f32 s12, s5, s9 ; CHECK-MVE-NEXT: vmla.f32 s12, s5, s9
; CHECK-MVE-NEXT: vmla.f32 s4, s7, s11 ; CHECK-MVE-NEXT: vmla.f32 s4, s7, s11
; CHECK-MVE-NEXT: vmla.f32 s8, s6, s10 ; CHECK-MVE-NEXT: vmla.f32 s8, s6, s10
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s7, #0
; CHECK-MVE-NEXT: cset r1, mi
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s6, #0
; CHECK-MVE-NEXT: cset r2, mi
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: cset r3, mi
; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s3, s4 ; CHECK-MVE-NEXT: vseleq.f32 s3, s3, s4
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s2, s8 ; CHECK-MVE-NEXT: vseleq.f32 s2, s2, s8
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s1, s12 ; CHECK-MVE-NEXT: vseleq.f32 s1, s1, s12
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s0, s14 ; CHECK-MVE-NEXT: vseleq.f32 s0, s0, s14
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
entry: entry:
%0 = fmul <4 x float> %src2, %src3 %0 = fmul <4 x float> %src2, %src3
%1 = fadd <4 x float> %0, %src1 %1 = fadd <4 x float> %0, %src1
%c = fcmp olt <4 x float> %src2, zeroinitializer %c = fcmp olt <4 x float> %src2, zeroinitializer
%s = select <4 x i1> %c, <4 x float> %1, <4 x float> %src1 %s = select <4 x i1> %c, <4 x float> %1, <4 x float> %src1
ret <4 x float> %s ret <4 x float> %s
Show All 10 Lines
; CHECK-MVE-VMLA-LABEL: vfms32_pred: ; CHECK-MVE-VMLA-LABEL: vfms32_pred:
; CHECK-MVE-VMLA: @ %bb.0: @ %entry ; CHECK-MVE-VMLA: @ %bb.0: @ %entry
; CHECK-MVE-VMLA-NEXT: vpt.f32 lt, q1, zr ; CHECK-MVE-VMLA-NEXT: vpt.f32 lt, q1, zr
; CHECK-MVE-VMLA-NEXT: vfmst.f32 q0, q1, q2 ; CHECK-MVE-VMLA-NEXT: vfmst.f32 q0, q1, q2
; CHECK-MVE-VMLA-NEXT: bx lr ; CHECK-MVE-VMLA-NEXT: bx lr
; ;
; CHECK-MVE-LABEL: vfms32_pred: ; CHECK-MVE-LABEL: vfms32_pred:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmov.f32 s14, s0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vcmp.f32 s5, #0 ; CHECK-MVE-NEXT: vcmp.f32 s5, #0
; CHECK-MVE-NEXT: movs r2, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s4, #0 ; CHECK-MVE-NEXT: vcmp.f32 s4, #0
; CHECK-MVE-NEXT: vmov.f32 s14, s0
; CHECK-MVE-NEXT: vmov.f32 s12, s1 ; CHECK-MVE-NEXT: vmov.f32 s12, s1
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s7, #0
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmls.f32 s14, s4, s8 ; CHECK-MVE-NEXT: vmls.f32 s14, s4, s8
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vmov.f32 s4, s3 ; CHECK-MVE-NEXT: vmov.f32 s4, s3
; CHECK-MVE-NEXT: mov.w r0, #0
; CHECK-MVE-NEXT: vmov.f32 s8, s2 ; CHECK-MVE-NEXT: vmov.f32 s8, s2
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r3, #1
; CHECK-MVE-NEXT: vcmp.f32 s6, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r0, #1
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vmls.f32 s12, s5, s9 ; CHECK-MVE-NEXT: vmls.f32 s12, s5, s9
; CHECK-MVE-NEXT: vmls.f32 s4, s7, s11 ; CHECK-MVE-NEXT: vmls.f32 s4, s7, s11
; CHECK-MVE-NEXT: vmls.f32 s8, s6, s10 ; CHECK-MVE-NEXT: vmls.f32 s8, s6, s10
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s7, #0
; CHECK-MVE-NEXT: cset r1, mi
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s6, #0
; CHECK-MVE-NEXT: cset r2, mi
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: cset r3, mi
; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s3, s4 ; CHECK-MVE-NEXT: vseleq.f32 s3, s3, s4
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s2, s8 ; CHECK-MVE-NEXT: vseleq.f32 s2, s2, s8
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s1, s12 ; CHECK-MVE-NEXT: vseleq.f32 s1, s1, s12
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s0, s14 ; CHECK-MVE-NEXT: vseleq.f32 s0, s0, s14
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
entry: entry:
%0 = fmul <4 x float> %src2, %src3 %0 = fmul <4 x float> %src2, %src3
%1 = fsub <4 x float> %src1, %0 %1 = fsub <4 x float> %src1, %0
%c = fcmp olt <4 x float> %src2, zeroinitializer %c = fcmp olt <4 x float> %src2, zeroinitializer
%s = select <4 x i1> %c, <4 x float> %1, <4 x float> %src1 %s = select <4 x i1> %c, <4 x float> %1, <4 x float> %src1
ret <4 x float> %s ret <4 x float> %s
Show All 13 Lines
; CHECK-MVE-VMLA: @ %bb.0: @ %entry ; CHECK-MVE-VMLA: @ %bb.0: @ %entry
; CHECK-MVE-VMLA-NEXT: vmov r0, s8 ; CHECK-MVE-VMLA-NEXT: vmov r0, s8
; CHECK-MVE-VMLA-NEXT: vpt.f32 lt, q1, zr ; CHECK-MVE-VMLA-NEXT: vpt.f32 lt, q1, zr
; CHECK-MVE-VMLA-NEXT: vfmat.f32 q0, q1, r0 ; CHECK-MVE-VMLA-NEXT: vfmat.f32 q0, q1, r0
; CHECK-MVE-VMLA-NEXT: bx lr ; CHECK-MVE-VMLA-NEXT: bx lr
; ;
; CHECK-MVE-LABEL: vfmar32_pred: ; CHECK-MVE-LABEL: vfmar32_pred:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmov.f32 s12, s0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vcmp.f32 s5, #0 ; CHECK-MVE-NEXT: vcmp.f32 s5, #0
; CHECK-MVE-NEXT: movs r2, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s4, #0 ; CHECK-MVE-NEXT: vcmp.f32 s4, #0
; CHECK-MVE-NEXT: vmov.f32 s12, s0
; CHECK-MVE-NEXT: vmov.f32 s14, s2 ; CHECK-MVE-NEXT: vmov.f32 s14, s2
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: vmov.f32 s10, s1 ; CHECK-MVE-NEXT: vmov.f32 s10, s1
; CHECK-MVE-NEXT: vmla.f32 s12, s4, s8
; CHECK-MVE-NEXT: vmov.f32 s4, s3
; CHECK-MVE-NEXT: vmla.f32 s14, s6, s8
; CHECK-MVE-NEXT: vmla.f32 s10, s5, s8
; CHECK-MVE-NEXT: vmla.f32 s4, s7, s8
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s7, #0 ; CHECK-MVE-NEXT: vcmp.f32 s7, #0
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r1, mi
; CHECK-MVE-NEXT: movmi r2, #1
; CHECK-MVE-NEXT: vmla.f32 s12, s4, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmov.f32 s4, s3
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s6, #0 ; CHECK-MVE-NEXT: vcmp.f32 s6, #0
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r2, mi
; CHECK-MVE-NEXT: movmi r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r3, mi
; CHECK-MVE-NEXT: movmi r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vmla.f32 s10, s5, s8
; CHECK-MVE-NEXT: vmla.f32 s14, s6, s8
; CHECK-MVE-NEXT: vmla.f32 s4, s7, s8
; CHECK-MVE-NEXT: vseleq.f32 s3, s3, s4 ; CHECK-MVE-NEXT: vseleq.f32 s3, s3, s4
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s2, s14 ; CHECK-MVE-NEXT: vseleq.f32 s2, s2, s14
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s1, s10 ; CHECK-MVE-NEXT: vseleq.f32 s1, s1, s10
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s0, s12 ; CHECK-MVE-NEXT: vseleq.f32 s0, s0, s12
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
entry: entry:
%i = insertelement <4 x float> undef, float %src3, i32 0 %i = insertelement <4 x float> undef, float %src3, i32 0
%sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer %sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer
%0 = fmul <4 x float> %src2, %sp %0 = fmul <4 x float> %src2, %sp
%1 = fadd <4 x float> %src1, %0 %1 = fadd <4 x float> %src1, %0
%c = fcmp olt <4 x float> %src2, zeroinitializer %c = fcmp olt <4 x float> %src2, zeroinitializer
Show All 14 Lines
; CHECK-MVE-VMLA: @ %bb.0: @ %entry ; CHECK-MVE-VMLA: @ %bb.0: @ %entry
; CHECK-MVE-VMLA-NEXT: vmov r0, s8 ; CHECK-MVE-VMLA-NEXT: vmov r0, s8
; CHECK-MVE-VMLA-NEXT: vpt.f32 lt, q1, zr ; CHECK-MVE-VMLA-NEXT: vpt.f32 lt, q1, zr
; CHECK-MVE-VMLA-NEXT: vfmast.f32 q0, q1, r0 ; CHECK-MVE-VMLA-NEXT: vfmast.f32 q0, q1, r0
; CHECK-MVE-VMLA-NEXT: bx lr ; CHECK-MVE-VMLA-NEXT: bx lr
; ;
; CHECK-MVE-LABEL: vfmas32_pred: ; CHECK-MVE-LABEL: vfmas32_pred:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmov.f32 s12, s8
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vcmp.f32 s5, #0 ; CHECK-MVE-NEXT: vcmp.f32 s5, #0
; CHECK-MVE-NEXT: movs r2, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s4, #0 ; CHECK-MVE-NEXT: vcmp.f32 s4, #0
; CHECK-MVE-NEXT: vmov.f32 s12, s8
; CHECK-MVE-NEXT: vmov.f32 s10, s8 ; CHECK-MVE-NEXT: vmov.f32 s10, s8
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: vmla.f32 s12, s0, s4
; CHECK-MVE-NEXT: movmi r1, #1 ; CHECK-MVE-NEXT: vmov.f32 s4, s8
; CHECK-MVE-NEXT: vmla.f32 s8, s2, s6
; CHECK-MVE-NEXT: vmla.f32 s10, s1, s5
; CHECK-MVE-NEXT: vmla.f32 s4, s3, s7
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s7, #0 ; CHECK-MVE-NEXT: vcmp.f32 s7, #0
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r1, mi
; CHECK-MVE-NEXT: movmi r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmla.f32 s12, s0, s4
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vmov.f32 s4, s8
; CHECK-MVE-NEXT: mov.w r0, #0
; CHECK-MVE-NEXT: vcmp.f32 s6, #0 ; CHECK-MVE-NEXT: vcmp.f32 s6, #0
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r2, mi
; CHECK-MVE-NEXT: movmi r3, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r3, mi
; CHECK-MVE-NEXT: movmi r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vmla.f32 s10, s1, s5
; CHECK-MVE-NEXT: vmla.f32 s8, s2, s6
; CHECK-MVE-NEXT: vmla.f32 s4, s3, s7
; CHECK-MVE-NEXT: vseleq.f32 s3, s3, s4 ; CHECK-MVE-NEXT: vseleq.f32 s3, s3, s4
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s2, s8 ; CHECK-MVE-NEXT: vseleq.f32 s2, s2, s8
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s1, s10 ; CHECK-MVE-NEXT: vseleq.f32 s1, s1, s10
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s0, s12 ; CHECK-MVE-NEXT: vseleq.f32 s0, s0, s12
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
entry: entry:
%i = insertelement <4 x float> undef, float %src3, i32 0 %i = insertelement <4 x float> undef, float %src3, i32 0
%sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer %sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer
%0 = fmul <4 x float> %src1, %src2 %0 = fmul <4 x float> %src1, %src2
%1 = fadd <4 x float> %sp, %0 %1 = fadd <4 x float> %sp, %0
%c = fcmp olt <4 x float> %src2, zeroinitializer %c = fcmp olt <4 x float> %src2, zeroinitializer
%s = select <4 x i1> %c, <4 x float> %1, <4 x float> %src1 %s = select <4 x i1> %c, <4 x float> %1, <4 x float> %src1
ret <4 x float> %s ret <4 x float> %s
} }

llvm/test/CodeGen/Thumb2/mve-fpclamptosat_vec.ll

Show All 12 Lines
; CHECK-NEXT: vmov q4, q0 ; CHECK-NEXT: vmov q4, q0
; CHECK-NEXT: vmov r0, r1, d9 ; CHECK-NEXT: vmov r0, r1, d9
; CHECK-NEXT: bl __aeabi_d2lz ; CHECK-NEXT: bl __aeabi_d2lz
; CHECK-NEXT: mov r4, r0 ; CHECK-NEXT: mov r4, r0
; CHECK-NEXT: mov r5, r1 ; CHECK-NEXT: mov r5, r1
; CHECK-NEXT: vmov r0, r1, d8 ; CHECK-NEXT: vmov r0, r1, d8
; CHECK-NEXT: bl __aeabi_d2lz ; CHECK-NEXT: bl __aeabi_d2lz
; CHECK-NEXT: adr r3, .LCPI0_0 ; CHECK-NEXT: adr r3, .LCPI0_0
; CHECK-NEXT: mvn r12, #-2147483648 ; CHECK-NEXT: mvn r2, #-2147483648
; CHECK-NEXT: vldrw.u32 q0, [r3] ; CHECK-NEXT: vldrw.u32 q0, [r3]
; CHECK-NEXT: subs.w r3, r4, r12 ; CHECK-NEXT: subs r3, r4, r2
; CHECK-NEXT: sbcs r3, r5, #0 ; CHECK-NEXT: sbcs r3, r5, #0
; CHECK-NEXT: vmov q1[2], q1[0], r0, r4 ; CHECK-NEXT: vmov q1[2], q1[0], r0, r4
; CHECK-NEXT: mov.w r3, #0 ; CHECK-NEXT: cset r3, lt
; CHECK-NEXT: vmov q1[3], q1[1], r1, r5 ; CHECK-NEXT: vmov q1[3], q1[1], r1, r5
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r3, #1
; CHECK-NEXT: cmp r3, #0 ; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: mov.w r5, #-1
; CHECK-NEXT: csetm r3, ne ; CHECK-NEXT: csetm r3, ne
; CHECK-NEXT: subs.w r0, r0, r12 ; CHECK-NEXT: subs r0, r0, r2
; CHECK-NEXT: sbcs r0, r1, #0 ; CHECK-NEXT: sbcs r0, r1, #0
; CHECK-NEXT: mov.w r12, #-1 ; CHECK-NEXT: adr r4, .LCPI0_1
; CHECK-NEXT: mov.w r0, #0 ; CHECK-NEXT: cset r0, lt
; CHECK-NEXT: mov.w r2, #0
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r0, #1
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: adr r4, .LCPI0_1
; CHECK-NEXT: vmov q2[2], q2[0], r0, r3 ; CHECK-NEXT: vmov q2[2], q2[0], r0, r3
; CHECK-NEXT: vmov q2[3], q2[1], r0, r3 ; CHECK-NEXT: vmov q2[3], q2[1], r0, r3
; CHECK-NEXT: vand q1, q1, q2 ; CHECK-NEXT: vand q1, q1, q2
; CHECK-NEXT: vbic q0, q0, q2 ; CHECK-NEXT: vbic q0, q0, q2
; CHECK-NEXT: vorr q0, q1, q0 ; CHECK-NEXT: vorr q0, q1, q0
; CHECK-NEXT: vldrw.u32 q1, [r4] ; CHECK-NEXT: vldrw.u32 q1, [r4]
; CHECK-NEXT: vmov r0, r1, d1 ; CHECK-NEXT: vmov r0, r1, d1
; CHECK-NEXT: vmov r3, r5, d0 ; CHECK-NEXT: vmov r2, r3, d0
; CHECK-NEXT: rsbs.w r0, r0, #-2147483648 ; CHECK-NEXT: rsbs.w r0, r0, #-2147483648
; CHECK-NEXT: sbcs.w r0, r12, r1 ; CHECK-NEXT: sbcs.w r0, r5, r1
; CHECK-NEXT: mov.w r0, #0 ; CHECK-NEXT: cset r0, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r0, #1
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: rsbs.w r1, r3, #-2147483648 ; CHECK-NEXT: rsbs.w r1, r2, #-2147483648
; CHECK-NEXT: sbcs.w r1, r12, r5 ; CHECK-NEXT: sbcs.w r1, r5, r3
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r1, lt
; CHECK-NEXT: movlt r2, #1 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: vmov q2[2], q2[0], r1, r0 ; CHECK-NEXT: vmov q2[2], q2[0], r1, r0
; CHECK-NEXT: vmov q2[3], q2[1], r1, r0 ; CHECK-NEXT: vmov q2[3], q2[1], r1, r0
; CHECK-NEXT: vbic q1, q1, q2 ; CHECK-NEXT: vbic q1, q1, q2
; CHECK-NEXT: vand q0, q0, q2 ; CHECK-NEXT: vand q0, q0, q2
; CHECK-NEXT: vorr q0, q0, q1 ; CHECK-NEXT: vorr q0, q0, q1
; CHECK-NEXT: vpop {d8, d9} ; CHECK-NEXT: vpop {d8, d9}
; CHECK-NEXT: pop {r4, r5, r7, pc} ; CHECK-NEXT: pop {r4, r5, r7, pc}
Show All 28 Lines
; CHECK-NEXT: vpush {d8, d9} ; CHECK-NEXT: vpush {d8, d9}
; CHECK-NEXT: vmov q4, q0 ; CHECK-NEXT: vmov q4, q0
; CHECK-NEXT: vmov r0, r1, d9 ; CHECK-NEXT: vmov r0, r1, d9
; CHECK-NEXT: bl __aeabi_d2ulz ; CHECK-NEXT: bl __aeabi_d2ulz
; CHECK-NEXT: mov r4, r0 ; CHECK-NEXT: mov r4, r0
; CHECK-NEXT: mov r5, r1 ; CHECK-NEXT: mov r5, r1
; CHECK-NEXT: vmov r0, r1, d8 ; CHECK-NEXT: vmov r0, r1, d8
; CHECK-NEXT: bl __aeabi_d2ulz ; CHECK-NEXT: bl __aeabi_d2ulz
; CHECK-NEXT: subs.w r3, r4, #-1 ; CHECK-NEXT: subs.w r2, r4, #-1
; CHECK-NEXT: vmov q1[2], q1[0], r0, r4 ; CHECK-NEXT: vmov q1[2], q1[0], r0, r4
; CHECK-NEXT: sbcs r3, r5, #0 ; CHECK-NEXT: sbcs r2, r5, #0
; CHECK-NEXT: mov.w r2, #0
; CHECK-NEXT: mov.w r3, #0
; CHECK-NEXT: vmov.i64 q0, #0xffffffff ; CHECK-NEXT: vmov.i64 q0, #0xffffffff
; CHECK-NEXT: it lo ; CHECK-NEXT: cset r2, lo
; CHECK-NEXT: movlo r3, #1 ; CHECK-NEXT: vmov q1[3], q1[1], r1, r5
; CHECK-NEXT: cmp r3, #0 ; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: csetm r3, ne ; CHECK-NEXT: csetm r2, ne
; CHECK-NEXT: subs.w r0, r0, #-1 ; CHECK-NEXT: subs.w r0, r0, #-1
; CHECK-NEXT: sbcs r0, r1, #0 ; CHECK-NEXT: sbcs r0, r1, #0
; CHECK-NEXT: it lo ; CHECK-NEXT: cset r0, lo
; CHECK-NEXT: movlo r2, #1 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: vmov q1[3], q1[1], r1, r5
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: vmov q2[2], q2[0], r0, r3 ; CHECK-NEXT: vmov q2[2], q2[0], r0, r2
; CHECK-NEXT: vmov q2[3], q2[1], r0, r3 ; CHECK-NEXT: vmov q2[3], q2[1], r0, r2
; CHECK-NEXT: vand q1, q1, q2 ; CHECK-NEXT: vand q1, q1, q2
; CHECK-NEXT: vbic q0, q0, q2 ; CHECK-NEXT: vbic q0, q0, q2
; CHECK-NEXT: vorr q0, q1, q0 ; CHECK-NEXT: vorr q0, q1, q0
; CHECK-NEXT: vpop {d8, d9} ; CHECK-NEXT: vpop {d8, d9}
; CHECK-NEXT: pop {r4, r5, r7, pc} ; CHECK-NEXT: pop {r4, r5, r7, pc}
entry: entry:
%conv = fptoui <2 x double> %x to <2 x i64> %conv = fptoui <2 x double> %x to <2 x i64>
%0 = icmp ult <2 x i64> %conv, <i64 4294967295, i64 4294967295> %0 = icmp ult <2 x i64> %conv, <i64 4294967295, i64 4294967295>
Show All 11 Lines
; CHECK-NEXT: vpush {d8, d9} ; CHECK-NEXT: vpush {d8, d9}
; CHECK-NEXT: vmov q4, q0 ; CHECK-NEXT: vmov q4, q0
; CHECK-NEXT: vmov r0, r1, d9 ; CHECK-NEXT: vmov r0, r1, d9
; CHECK-NEXT: bl __aeabi_d2lz ; CHECK-NEXT: bl __aeabi_d2lz
; CHECK-NEXT: mov r4, r0 ; CHECK-NEXT: mov r4, r0
; CHECK-NEXT: mov r5, r1 ; CHECK-NEXT: mov r5, r1
; CHECK-NEXT: vmov r0, r1, d8 ; CHECK-NEXT: vmov r0, r1, d8
; CHECK-NEXT: bl __aeabi_d2lz ; CHECK-NEXT: bl __aeabi_d2lz
; CHECK-NEXT: subs.w r3, r4, #-1 ; CHECK-NEXT: subs.w r2, r4, #-1
; CHECK-NEXT: vmov q1[2], q1[0], r0, r4 ; CHECK-NEXT: vmov q1[2], q1[0], r0, r4
; CHECK-NEXT: sbcs r3, r5, #0 ; CHECK-NEXT: sbcs r2, r5, #0
; CHECK-NEXT: vmov.i64 q0, #0xffffffff
; CHECK-NEXT: mov.w r3, #0
; CHECK-NEXT: vmov q1[3], q1[1], r1, r5 ; CHECK-NEXT: vmov q1[3], q1[1], r1, r5
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r2, lt
; CHECK-NEXT: movlt r3, #1 ; CHECK-NEXT: vmov.i64 q0, #0xffffffff
; CHECK-NEXT: cmp r3, #0 ; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: csetm r3, ne ; CHECK-NEXT: mov.w r5, #0
; CHECK-NEXT: csetm r2, ne
; CHECK-NEXT: subs.w r0, r0, #-1 ; CHECK-NEXT: subs.w r0, r0, #-1
; CHECK-NEXT: sbcs r0, r1, #0 ; CHECK-NEXT: sbcs r0, r1, #0
; CHECK-NEXT: mov.w r2, #0 ; CHECK-NEXT: cset r0, lt
; CHECK-NEXT: mov.w r0, #0
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r0, #1
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: vmov q2[2], q2[0], r0, r3 ; CHECK-NEXT: vmov q2[2], q2[0], r0, r2
; CHECK-NEXT: vmov q2[3], q2[1], r0, r3 ; CHECK-NEXT: vmov q2[3], q2[1], r0, r2
; CHECK-NEXT: vand q1, q1, q2 ; CHECK-NEXT: vand q1, q1, q2
; CHECK-NEXT: vbic q0, q0, q2 ; CHECK-NEXT: vbic q0, q0, q2
; CHECK-NEXT: vorr q0, q1, q0 ; CHECK-NEXT: vorr q0, q1, q0
; CHECK-NEXT: vmov r0, r1, d1 ; CHECK-NEXT: vmov r0, r1, d1
; CHECK-NEXT: vmov r3, r5, d0 ; CHECK-NEXT: vmov r2, r3, d0
; CHECK-NEXT: rsbs r0, r0, #0 ; CHECK-NEXT: rsbs r0, r0, #0
; CHECK-NEXT: sbcs.w r0, r2, r1 ; CHECK-NEXT: sbcs.w r0, r5, r1
; CHECK-NEXT: mov.w r0, #0 ; CHECK-NEXT: cset r0, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r0, #1
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: rsbs r1, r3, #0 ; CHECK-NEXT: rsbs r1, r2, #0
; CHECK-NEXT: sbcs.w r1, r2, r5 ; CHECK-NEXT: sbcs.w r1, r5, r3
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r1, lt
; CHECK-NEXT: movlt r2, #1 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: vmov q1[2], q1[0], r1, r0 ; CHECK-NEXT: vmov q1[2], q1[0], r1, r0
; CHECK-NEXT: vmov q1[3], q1[1], r1, r0 ; CHECK-NEXT: vmov q1[3], q1[1], r1, r0
; CHECK-NEXT: vand q0, q0, q1 ; CHECK-NEXT: vand q0, q0, q1
; CHECK-NEXT: vpop {d8, d9} ; CHECK-NEXT: vpop {d8, d9}
; CHECK-NEXT: pop {r4, r5, r7, pc} ; CHECK-NEXT: pop {r4, r5, r7, pc}
entry: entry:
%conv = fptosi <2 x double> %x to <2 x i64> %conv = fptosi <2 x double> %x to <2 x i64>
%0 = icmp slt <2 x i64> %conv, <i64 4294967295, i64 4294967295> %0 = icmp slt <2 x i64> %conv, <i64 4294967295, i64 4294967295>
%spec.store.select = select <2 x i1> %0, <2 x i64> %conv, <2 x i64> <i64 4294967295, i64 4294967295> %spec.store.select = select <2 x i1> %0, <2 x i64> %conv, <2 x i64> <i64 4294967295, i64 4294967295>
%1 = icmp sgt <2 x i64> %spec.store.select, zeroinitializer %1 = icmp sgt <2 x i64> %spec.store.select, zeroinitializer
%spec.store.select7 = select <2 x i1> %1, <2 x i64> %spec.store.select, <2 x i64> zeroinitializer %spec.store.select7 = select <2 x i1> %1, <2 x i64> %spec.store.select, <2 x i64> zeroinitializer
%conv6 = trunc <2 x i64> %spec.store.select7 to <2 x i32> %conv6 = trunc <2 x i64> %spec.store.select7 to <2 x i32>
ret <2 x i32> %conv6 ret <2 x i32> %conv6
} }
define arm_aapcs_vfpcc <4 x i32> @stest_f32i32(<4 x float> %x) { define arm_aapcs_vfpcc <4 x i32> @stest_f32i32(<4 x float> %x) {
; CHECK-LABEL: stest_f32i32: ; CHECK-LABEL: stest_f32i32:
; CHECK: @ %bb.0: @ %entry ; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: .save {r4, r5, r6, r7, r8, r9, r10, r11, lr} ; CHECK-NEXT: .save {r4, r5, r6, r7, r8, r9, lr}
; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, r9, r10, r11, lr} ; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, r9, lr}
; CHECK-NEXT: .pad #4 ; CHECK-NEXT: .pad #4
; CHECK-NEXT: sub sp, #4 ; CHECK-NEXT: sub sp, #4
; CHECK-NEXT: .vsave {d8, d9} ; CHECK-NEXT: .vsave {d8, d9, d10, d11}
; CHECK-NEXT: vpush {d8, d9} ; CHECK-NEXT: vpush {d8, d9, d10, d11}
; CHECK-NEXT: .pad #8
; CHECK-NEXT: sub sp, #8
; CHECK-NEXT: vmov q4, q0 ; CHECK-NEXT: vmov q4, q0
; CHECK-NEXT: vmov r0, r6, d8 ; CHECK-NEXT: vmov r4, r0, d8
; CHECK-NEXT: bl __aeabi_f2lz ; CHECK-NEXT: bl __aeabi_f2lz
; CHECK-NEXT: mov r9, r0 ; CHECK-NEXT: mov r5, r0
; CHECK-NEXT: mvn r0, #-2147483648 ; CHECK-NEXT: mov r0, r4
; CHECK-NEXT: subs.w r0, r9, r0 ; CHECK-NEXT: mov r6, r1
; CHECK-NEXT: str r1, [sp, #4] @ 4-byte Spill
; CHECK-NEXT: sbcs r0, r1, #0
; CHECK-NEXT: mvn r4, #-2147483648
; CHECK-NEXT: mov.w r0, #0
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r0, #1
; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: mov r0, r6
; CHECK-NEXT: csetm r11, ne
; CHECK-NEXT: bl __aeabi_f2lz ; CHECK-NEXT: bl __aeabi_f2lz
; CHECK-NEXT: mov r6, r0 ; CHECK-NEXT: vmov r4, r2, d9
; CHECK-NEXT: subs r0, r0, r4 ; CHECK-NEXT: adr r3, .LCPI3_0
; CHECK-NEXT: mvn r7, #-2147483648
; CHECK-NEXT: vldrw.u32 q5, [r3]
; CHECK-NEXT: subs r3, r5, r7
; CHECK-NEXT: vmov q0[2], q0[0], r0, r5
; CHECK-NEXT: sbcs r3, r6, #0
; CHECK-NEXT: vmov q0[3], q0[1], r1, r6
; CHECK-NEXT: cset r3, lt
; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: csetm r3, ne
; CHECK-NEXT: subs r0, r0, r7
; CHECK-NEXT: sbcs r0, r1, #0 ; CHECK-NEXT: sbcs r0, r1, #0
; CHECK-NEXT: mov r10, r1 ; CHECK-NEXT: cset r0, lt
; CHECK-NEXT: mov.w r0, #0
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r0, #1
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: vmov r4, r0, d9 ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: csetm r8, ne ; CHECK-NEXT: vmov q1[2], q1[0], r0, r3
; CHECK-NEXT: vmov q1[3], q1[1], r0, r3
; CHECK-NEXT: vand q0, q0, q1
; CHECK-NEXT: vbic q1, q5, q1
; CHECK-NEXT: vorr q4, q0, q1
; CHECK-NEXT: vmov r9, r8, d8
; CHECK-NEXT: mov r0, r2
; CHECK-NEXT: bl __aeabi_f2lz ; CHECK-NEXT: bl __aeabi_f2lz
; CHECK-NEXT: mov r5, r0 ; CHECK-NEXT: mov r5, r0
; CHECK-NEXT: mov r0, r4 ; CHECK-NEXT: mov r0, r4
; CHECK-NEXT: mov r7, r1 ; CHECK-NEXT: mov r6, r1
; CHECK-NEXT: bl __aeabi_f2lz ; CHECK-NEXT: bl __aeabi_f2lz
; CHECK-NEXT: adr r2, .LCPI3_0 ; CHECK-NEXT: adr r3, .LCPI3_1
; CHECK-NEXT: mvn r4, #-2147483648 ; CHECK-NEXT: vmov q1[2], q1[0], r0, r5
; CHECK-NEXT: vldrw.u32 q0, [r2] ; CHECK-NEXT: vldrw.u32 q0, [r3]
; CHECK-NEXT: adr r2, .LCPI3_1 ; CHECK-NEXT: subs r3, r5, r7
; CHECK-NEXT: vldrw.u32 q2, [r2] ; CHECK-NEXT: sbcs r3, r6, #0
; CHECK-NEXT: subs r2, r5, r4 ; CHECK-NEXT: vmov q1[3], q1[1], r1, r6
; CHECK-NEXT: ldr r2, [sp, #4] @ 4-byte Reload ; CHECK-NEXT: cset r3, lt
; CHECK-NEXT: vmov q1[2], q1[0], r9, r6 ; CHECK-NEXT: mov.w r2, #-1
; CHECK-NEXT: vmov q4[2], q4[0], r11, r8
; CHECK-NEXT: vmov q3[2], q3[0], r0, r5
; CHECK-NEXT: vmov q1[3], q1[1], r2, r10
; CHECK-NEXT: vmov q4[3], q4[1], r11, r8
; CHECK-NEXT: vand q1, q1, q4
; CHECK-NEXT: vbic q4, q2, q4
; CHECK-NEXT: vorr q1, q1, q4
; CHECK-NEXT: vmov q3[3], q3[1], r1, r7
; CHECK-NEXT: vmov r2, r3, d2
; CHECK-NEXT: sbcs r7, r7, #0
; CHECK-NEXT: vmov r6, r5, d3
; CHECK-NEXT: mov.w r7, #0
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r7, #1
; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csetm r7, ne
; CHECK-NEXT: subs r0, r0, r4
; CHECK-NEXT: sbcs r0, r1, #0
; CHECK-NEXT: mov.w r0, #0
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r0, #1
; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: mov.w r0, #-1
; CHECK-NEXT: vmov q4[2], q4[0], r1, r7
; CHECK-NEXT: vmov q4[3], q4[1], r1, r7
; CHECK-NEXT: vand q3, q3, q4
; CHECK-NEXT: vbic q2, q2, q4
; CHECK-NEXT: vorr q2, q3, q2
; CHECK-NEXT: vmov r1, r7, d5
; CHECK-NEXT: rsbs.w r2, r2, #-2147483648
; CHECK-NEXT: sbcs.w r2, r0, r3
; CHECK-NEXT: mov.w r2, #0
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r2, #1
; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: csetm r2, ne
; CHECK-NEXT: rsbs.w r3, r6, #-2147483648
; CHECK-NEXT: sbcs.w r3, r0, r5
; CHECK-NEXT: vmov r6, r5, d4
; CHECK-NEXT: mov.w r3, #0
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r3, #1
; CHECK-NEXT: cmp r3, #0 ; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: csetm r3, ne ; CHECK-NEXT: csetm r3, ne
; CHECK-NEXT: subs r0, r0, r7
; CHECK-NEXT: sbcs r0, r1, #0
; CHECK-NEXT: vmov r1, r7, d9
; CHECK-NEXT: cset r0, lt
; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: rsbs.w r6, r9, #-2147483648
; CHECK-NEXT: vmov q2[2], q2[0], r0, r3
; CHECK-NEXT: sbcs.w r6, r2, r8
; CHECK-NEXT: vmov q2[3], q2[1], r0, r3
; CHECK-NEXT: cset r6, lt
; CHECK-NEXT: vand q1, q1, q2
; CHECK-NEXT: vbic q2, q5, q2
; CHECK-NEXT: vorr q1, q1, q2
; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: vmov r0, r3, d3
; CHECK-NEXT: csetm r6, ne
; CHECK-NEXT: vmov r5, r4, d2
; CHECK-NEXT: rsbs.w r1, r1, #-2147483648 ; CHECK-NEXT: rsbs.w r1, r1, #-2147483648
; CHECK-NEXT: sbcs.w r1, r0, r7 ; CHECK-NEXT: sbcs.w r1, r2, r7
; CHECK-NEXT: mov.w r1, #0 ; CHECK-NEXT: cset r1, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r1, #1
; CHECK-NEXT: cmp r1, #0 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: rsbs.w r7, r6, #-2147483648 ; CHECK-NEXT: rsbs.w r0, r0, #-2147483648
; CHECK-NEXT: sbcs r0, r5 ; CHECK-NEXT: sbcs.w r0, r2, r3
; CHECK-NEXT: mov.w r0, #0 ; CHECK-NEXT: cset r0, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r0, #1
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: vmov.32 q3[1], r0 ; CHECK-NEXT: rsbs.w r3, r5, #-2147483648
; CHECK-NEXT: vmov q3[2], q3[0], r0, r1 ; CHECK-NEXT: sbcs r2, r4
; CHECK-NEXT: vbic q4, q0, q3 ; CHECK-NEXT: cset r2, lt
; CHECK-NEXT: vand q2, q2, q3 ; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: vmov.32 q3[1], r2 ; CHECK-NEXT: csetm r2, ne
; CHECK-NEXT: vorr q2, q2, q4 ; CHECK-NEXT: vmov.32 q2[1], r2
; CHECK-NEXT: vmov q3[2], q3[0], r2, r3 ; CHECK-NEXT: vmov q2[2], q2[0], r2, r0
; CHECK-NEXT: vbic q0, q0, q3 ; CHECK-NEXT: vbic q3, q0, q2
; CHECK-NEXT: vand q1, q1, q3 ; CHECK-NEXT: vand q1, q1, q2
; CHECK-NEXT: vorr q0, q1, q0 ; CHECK-NEXT: vmov.32 q2[1], r6
; CHECK-NEXT: vorr q1, q1, q3
; CHECK-NEXT: vmov q2[2], q2[0], r6, r1
; CHECK-NEXT: vbic q0, q0, q2
; CHECK-NEXT: vand q2, q4, q2
; CHECK-NEXT: vorr q0, q2, q0
; CHECK-NEXT: vmov.f32 s1, s2 ; CHECK-NEXT: vmov.f32 s1, s2
; CHECK-NEXT: vmov.f32 s2, s8 ; CHECK-NEXT: vmov.f32 s2, s4
; CHECK-NEXT: vmov.f32 s3, s10 ; CHECK-NEXT: vmov.f32 s3, s6
; CHECK-NEXT: add sp, #8 ; CHECK-NEXT: vpop {d8, d9, d10, d11}
; CHECK-NEXT: vpop {d8, d9}
; CHECK-NEXT: add sp, #4 ; CHECK-NEXT: add sp, #4
; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, r10, r11, pc} ; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, pc}
; CHECK-NEXT: .p2align 4 ; CHECK-NEXT: .p2align 4
; CHECK-NEXT: @ %bb.1: ; CHECK-NEXT: @ %bb.1:
; CHECK-NEXT: .LCPI3_0: ; CHECK-NEXT: .LCPI3_0:
; CHECK-NEXT: .long 2147483648 @ 0x80000000
; CHECK-NEXT: .long 4294967295 @ 0xffffffff
; CHECK-NEXT: .long 2147483648 @ 0x80000000
; CHECK-NEXT: .long 4294967295 @ 0xffffffff
; CHECK-NEXT: .LCPI3_1:
; CHECK-NEXT: .long 2147483647 @ 0x7fffffff ; CHECK-NEXT: .long 2147483647 @ 0x7fffffff
; CHECK-NEXT: .long 0 @ 0x0 ; CHECK-NEXT: .long 0 @ 0x0
; CHECK-NEXT: .long 2147483647 @ 0x7fffffff ; CHECK-NEXT: .long 2147483647 @ 0x7fffffff
; CHECK-NEXT: .long 0 @ 0x0 ; CHECK-NEXT: .long 0 @ 0x0
; CHECK-NEXT: .LCPI3_1:
; CHECK-NEXT: .long 2147483648 @ 0x80000000
; CHECK-NEXT: .long 4294967295 @ 0xffffffff
; CHECK-NEXT: .long 2147483648 @ 0x80000000
; CHECK-NEXT: .long 4294967295 @ 0xffffffff
entry: entry:
%conv = fptosi <4 x float> %x to <4 x i64> %conv = fptosi <4 x float> %x to <4 x i64>
%0 = icmp slt <4 x i64> %conv, <i64 2147483647, i64 2147483647, i64 2147483647, i64 2147483647> %0 = icmp slt <4 x i64> %conv, <i64 2147483647, i64 2147483647, i64 2147483647, i64 2147483647>
%spec.store.select = select <4 x i1> %0, <4 x i64> %conv, <4 x i64> <i64 2147483647, i64 2147483647, i64 2147483647, i64 2147483647> %spec.store.select = select <4 x i1> %0, <4 x i64> %conv, <4 x i64> <i64 2147483647, i64 2147483647, i64 2147483647, i64 2147483647>
%1 = icmp sgt <4 x i64> %spec.store.select, <i64 -2147483648, i64 -2147483648, i64 -2147483648, i64 -2147483648> %1 = icmp sgt <4 x i64> %spec.store.select, <i64 -2147483648, i64 -2147483648, i64 -2147483648, i64 -2147483648>
%spec.store.select7 = select <4 x i1> %1, <4 x i64> %spec.store.select, <4 x i64> <i64 -2147483648, i64 -2147483648, i64 -2147483648, i64 -2147483648> %spec.store.select7 = select <4 x i1> %1, <4 x i64> %spec.store.select, <4 x i64> <i64 -2147483648, i64 -2147483648, i64 -2147483648, i64 -2147483648>
%conv6 = trunc <4 x i64> %spec.store.select7 to <4 x i32> %conv6 = trunc <4 x i64> %spec.store.select7 to <4 x i32>
ret <4 x i32> %conv6 ret <4 x i32> %conv6
Show All 16 Lines
; CHECK-NEXT: mov r6, r0 ; CHECK-NEXT: mov r6, r0
; CHECK-NEXT: vmov r5, r0, d8 ; CHECK-NEXT: vmov r5, r0, d8
; CHECK-NEXT: mov r9, r1 ; CHECK-NEXT: mov r9, r1
; CHECK-NEXT: bl __aeabi_f2ulz ; CHECK-NEXT: bl __aeabi_f2ulz
; CHECK-NEXT: mov r7, r0 ; CHECK-NEXT: mov r7, r0
; CHECK-NEXT: mov r0, r5 ; CHECK-NEXT: mov r0, r5
; CHECK-NEXT: mov r4, r1 ; CHECK-NEXT: mov r4, r1
; CHECK-NEXT: bl __aeabi_f2ulz ; CHECK-NEXT: bl __aeabi_f2ulz
; CHECK-NEXT: subs.w r3, r7, #-1 ; CHECK-NEXT: subs.w r2, r7, #-1
; CHECK-NEXT: vmov q0[2], q0[0], r0, r7 ; CHECK-NEXT: vmov q0[2], q0[0], r0, r7
; CHECK-NEXT: sbcs r3, r4, #0 ; CHECK-NEXT: sbcs r2, r4, #0
; CHECK-NEXT: mov.w r2, #0
; CHECK-NEXT: mov.w r3, #0
; CHECK-NEXT: vmov q1[2], q1[0], r10, r6 ; CHECK-NEXT: vmov q1[2], q1[0], r10, r6
; CHECK-NEXT: it lo ; CHECK-NEXT: cset r2, lo
; CHECK-NEXT: movlo r3, #1 ; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: cmp r3, #0 ; CHECK-NEXT: csetm r2, ne
; CHECK-NEXT: csetm r3, ne
; CHECK-NEXT: subs.w r0, r0, #-1 ; CHECK-NEXT: subs.w r0, r0, #-1
; CHECK-NEXT: sbcs r0, r1, #0 ; CHECK-NEXT: sbcs r0, r1, #0
; CHECK-NEXT: mov.w r0, #0 ; CHECK-NEXT: cset r0, lo
; CHECK-NEXT: it lo
; CHECK-NEXT: movlo r0, #1
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: subs.w r1, r6, #-1 ; CHECK-NEXT: subs.w r1, r6, #-1
; CHECK-NEXT: sbcs r1, r9, #0 ; CHECK-NEXT: sbcs r1, r9, #0
; CHECK-NEXT: mov.w r1, #0 ; CHECK-NEXT: cset r1, lo
; CHECK-NEXT: it lo
; CHECK-NEXT: movlo r1, #1
; CHECK-NEXT: cmp r1, #0 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: subs.w r7, r10, #-1 ; CHECK-NEXT: subs.w r3, r10, #-1
; CHECK-NEXT: sbcs r7, r8, #0 ; CHECK-NEXT: sbcs r3, r8, #0
; CHECK-NEXT: it lo ; CHECK-NEXT: cset r3, lo
; CHECK-NEXT: movlo r2, #1 ; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: cmp r2, #0 ; CHECK-NEXT: csetm r3, ne
; CHECK-NEXT: csetm r2, ne ; CHECK-NEXT: vmov.32 q2[1], r3
; CHECK-NEXT: vmov.32 q2[1], r2 ; CHECK-NEXT: vmov q2[2], q2[0], r3, r1
; CHECK-NEXT: vmov q2[2], q2[0], r2, r1
; CHECK-NEXT: vand q1, q1, q2 ; CHECK-NEXT: vand q1, q1, q2
; CHECK-NEXT: vorn q1, q1, q2 ; CHECK-NEXT: vorn q1, q1, q2
; CHECK-NEXT: vmov.32 q2[1], r0 ; CHECK-NEXT: vmov.32 q2[1], r0
; CHECK-NEXT: vmov q2[2], q2[0], r0, r3 ; CHECK-NEXT: vmov q2[2], q2[0], r0, r2
; CHECK-NEXT: vand q0, q0, q2 ; CHECK-NEXT: vand q0, q0, q2
; CHECK-NEXT: vorn q0, q0, q2 ; CHECK-NEXT: vorn q0, q0, q2
; CHECK-NEXT: vmov.f32 s1, s2 ; CHECK-NEXT: vmov.f32 s1, s2
; CHECK-NEXT: vmov.f32 s2, s4 ; CHECK-NEXT: vmov.f32 s2, s4
; CHECK-NEXT: vmov.f32 s3, s6 ; CHECK-NEXT: vmov.f32 s3, s6
; CHECK-NEXT: vpop {d8, d9} ; CHECK-NEXT: vpop {d8, d9}
; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, r10, pc} ; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, r10, pc}
entry: entry:
Show All 19 Lines
; CHECK-NEXT: mov r5, r0 ; CHECK-NEXT: mov r5, r0
; CHECK-NEXT: mov r0, r4 ; CHECK-NEXT: mov r0, r4
; CHECK-NEXT: mov r6, r1 ; CHECK-NEXT: mov r6, r1
; CHECK-NEXT: bl __aeabi_f2lz ; CHECK-NEXT: bl __aeabi_f2lz
; CHECK-NEXT: vmov r4, r2, d8 ; CHECK-NEXT: vmov r4, r2, d8
; CHECK-NEXT: subs.w r3, r5, #-1 ; CHECK-NEXT: subs.w r3, r5, #-1
; CHECK-NEXT: sbcs r3, r6, #0 ; CHECK-NEXT: sbcs r3, r6, #0
; CHECK-NEXT: vmov q0[2], q0[0], r0, r5 ; CHECK-NEXT: vmov q0[2], q0[0], r0, r5
; CHECK-NEXT: mov.w r3, #0 ; CHECK-NEXT: cset r3, lt
; CHECK-NEXT: vmov.i64 q5, #0xffffffff ; CHECK-NEXT: vmov.i64 q5, #0xffffffff
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r3, #1
; CHECK-NEXT: cmp r3, #0 ; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: vmov q0[3], q0[1], r1, r6
; CHECK-NEXT: csetm r3, ne ; CHECK-NEXT: csetm r3, ne
; CHECK-NEXT: subs.w r0, r0, #-1 ; CHECK-NEXT: subs.w r0, r0, #-1
; CHECK-NEXT: sbcs r0, r1, #0 ; CHECK-NEXT: sbcs r0, r1, #0
; CHECK-NEXT: vmov q0[3], q0[1], r1, r6 ; CHECK-NEXT: cset r0, lt
; CHECK-NEXT: mov.w r0, #0
; CHECK-NEXT: mov.w r7, #0
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r0, #1
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: vmov q1[2], q1[0], r0, r3 ; CHECK-NEXT: vmov q1[2], q1[0], r0, r3
; CHECK-NEXT: vmov q1[3], q1[1], r0, r3 ; CHECK-NEXT: vmov q1[3], q1[1], r0, r3
; CHECK-NEXT: vand q0, q0, q1 ; CHECK-NEXT: vand q0, q0, q1
; CHECK-NEXT: vbic q1, q5, q1 ; CHECK-NEXT: vbic q1, q5, q1
; CHECK-NEXT: vorr q4, q0, q1 ; CHECK-NEXT: vorr q4, q0, q1
; CHECK-NEXT: vmov r9, r8, d9 ; CHECK-NEXT: vmov r9, r8, d9
; CHECK-NEXT: mov r0, r2 ; CHECK-NEXT: mov r0, r2
; CHECK-NEXT: bl __aeabi_f2lz ; CHECK-NEXT: bl __aeabi_f2lz
; CHECK-NEXT: mov r5, r0 ; CHECK-NEXT: mov r5, r0
; CHECK-NEXT: mov r0, r4 ; CHECK-NEXT: subs.w r0, r0, #-1
; CHECK-NEXT: sbcs r0, r1, #0
; CHECK-NEXT: mov r6, r1 ; CHECK-NEXT: mov r6, r1
; CHECK-NEXT: cset r0, lt
; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: mov r0, r4
; CHECK-NEXT: csetm r7, ne
; CHECK-NEXT: bl __aeabi_f2lz ; CHECK-NEXT: bl __aeabi_f2lz
; CHECK-NEXT: subs.w r2, r5, #-1 ; CHECK-NEXT: subs.w r3, r0, #-1
; CHECK-NEXT: vmov q0[2], q0[0], r0, r5 ; CHECK-NEXT: vmov q0[2], q0[0], r0, r5
; CHECK-NEXT: sbcs r2, r6, #0
; CHECK-NEXT: vmov q0[3], q0[1], r1, r6
; CHECK-NEXT: mov.w r2, #0
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r2, #1
; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: csetm r2, ne
; CHECK-NEXT: subs.w r0, r0, #-1
; CHECK-NEXT: sbcs r0, r1, #0 ; CHECK-NEXT: sbcs r0, r1, #0
; CHECK-NEXT: vmov q0[3], q0[1], r1, r6
; CHECK-NEXT: cset r0, lt
; CHECK-NEXT: vmov r1, r3, d8 ; CHECK-NEXT: vmov r1, r3, d8
; CHECK-NEXT: mov.w r0, #0
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r0, #1
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: mov.w r2, #0
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: rsbs.w r6, r9, #0 ; CHECK-NEXT: rsbs.w r6, r9, #0
; CHECK-NEXT: vmov q1[2], q1[0], r0, r2 ; CHECK-NEXT: vmov q1[2], q1[0], r0, r7
; CHECK-NEXT: sbcs.w r6, r7, r8 ; CHECK-NEXT: sbcs.w r6, r2, r8
; CHECK-NEXT: vmov q1[3], q1[1], r0, r2 ; CHECK-NEXT: vmov q1[3], q1[1], r0, r7
; CHECK-NEXT: mov.w r6, #0 ; CHECK-NEXT: cset r6, lt
; CHECK-NEXT: vand q0, q0, q1 ; CHECK-NEXT: vand q0, q0, q1
; CHECK-NEXT: vbic q1, q5, q1 ; CHECK-NEXT: vbic q1, q5, q1
; CHECK-NEXT: vorr q0, q0, q1 ; CHECK-NEXT: vorr q0, q0, q1
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r6, #1
; CHECK-NEXT: vmov r0, r2, d1
; CHECK-NEXT: cmp r6, #0 ; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: vmov r5, r4, d0 ; CHECK-NEXT: vmov r0, r7, d1
; CHECK-NEXT: csetm r6, ne ; CHECK-NEXT: csetm r6, ne
; CHECK-NEXT: vmov r5, r4, d0
; CHECK-NEXT: rsbs r1, r1, #0 ; CHECK-NEXT: rsbs r1, r1, #0
; CHECK-NEXT: sbcs.w r1, r7, r3 ; CHECK-NEXT: sbcs.w r1, r2, r3
; CHECK-NEXT: mov.w r1, #0 ; CHECK-NEXT: cset r1, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r1, #1
; CHECK-NEXT: cmp r1, #0 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: vmov q2[2], q2[0], r1, r6 ; CHECK-NEXT: vmov q2[2], q2[0], r1, r6
; CHECK-NEXT: vand q2, q4, q2 ; CHECK-NEXT: vand q2, q4, q2
; CHECK-NEXT: rsbs r0, r0, #0 ; CHECK-NEXT: rsbs r0, r0, #0
; CHECK-NEXT: sbcs.w r0, r7, r2 ; CHECK-NEXT: sbcs.w r0, r2, r7
; CHECK-NEXT: mov.w r0, #0 ; CHECK-NEXT: cset r0, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r0, #1
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: rsbs r2, r5, #0 ; CHECK-NEXT: rsbs r3, r5, #0
; CHECK-NEXT: sbcs.w r2, r7, r4 ; CHECK-NEXT: sbcs r2, r4
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r2, lt
; CHECK-NEXT: movlt r7, #1 ; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csetm r2, ne ; CHECK-NEXT: csetm r2, ne
; CHECK-NEXT: vmov q1[2], q1[0], r2, r0 ; CHECK-NEXT: vmov q1[2], q1[0], r2, r0
; CHECK-NEXT: vand q0, q0, q1 ; CHECK-NEXT: vand q0, q0, q1
; CHECK-NEXT: vmov.f32 s1, s2 ; CHECK-NEXT: vmov.f32 s1, s2
; CHECK-NEXT: vmov.f32 s2, s8 ; CHECK-NEXT: vmov.f32 s2, s8
; CHECK-NEXT: vmov.f32 s3, s10 ; CHECK-NEXT: vmov.f32 s3, s10
; CHECK-NEXT: vpop {d8, d9, d10, d11} ; CHECK-NEXT: vpop {d8, d9, d10, d11}
; CHECK-NEXT: add sp, #4 ; CHECK-NEXT: add sp, #4
▲ Show 20 Lines • Show All 95 Lines • ▼ Show 20 Lines
; CHECK-NEXT: mov r5, r0 ; CHECK-NEXT: mov r5, r0
; CHECK-NEXT: vmov.u16 r0, q4[0] ; CHECK-NEXT: vmov.u16 r0, q4[0]
; CHECK-NEXT: mov r7, r1 ; CHECK-NEXT: mov r7, r1
; CHECK-NEXT: bl __fixhfdi ; CHECK-NEXT: bl __fixhfdi
; CHECK-NEXT: rsbs r3, r5, #0 ; CHECK-NEXT: rsbs r3, r5, #0
; CHECK-NEXT: mov.w r2, #0 ; CHECK-NEXT: mov.w r2, #0
; CHECK-NEXT: sbcs.w r3, r2, r7 ; CHECK-NEXT: sbcs.w r3, r2, r7
; CHECK-NEXT: vmov q0[2], q0[0], r0, r6 ; CHECK-NEXT: vmov q0[2], q0[0], r0, r6
; CHECK-NEXT: mov.w r3, #0 ; CHECK-NEXT: cset r3, lt
; CHECK-NEXT: vmov q1[2], q1[0], r4, r5 ; CHECK-NEXT: vmov q1[2], q1[0], r4, r5
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r3, #1
; CHECK-NEXT: cmp r3, #0 ; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: csetm r3, ne ; CHECK-NEXT: csetm r3, ne
; CHECK-NEXT: rsbs r7, r4, #0 ; CHECK-NEXT: rsbs r7, r4, #0
; CHECK-NEXT: sbcs.w r7, r2, r8 ; CHECK-NEXT: sbcs.w r7, r2, r8
; CHECK-NEXT: mov.w r7, #0 ; CHECK-NEXT: cset r7, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r7, #1
; CHECK-NEXT: cmp r7, #0 ; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csetm r7, ne ; CHECK-NEXT: csetm r7, ne
; CHECK-NEXT: rsbs r6, r6, #0 ; CHECK-NEXT: rsbs r6, r6, #0
; CHECK-NEXT: sbcs.w r6, r2, r9 ; CHECK-NEXT: sbcs.w r6, r2, r9
; CHECK-NEXT: vmov q3[2], q3[0], r7, r3 ; CHECK-NEXT: vmov q3[2], q3[0], r7, r3
; CHECK-NEXT: mov.w r6, #0 ; CHECK-NEXT: cset r6, lt
; CHECK-NEXT: vand q1, q1, q3 ; CHECK-NEXT: vand q1, q1, q3
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r6, #1
; CHECK-NEXT: cmp r6, #0 ; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: csetm r6, ne ; CHECK-NEXT: csetm r6, ne
; CHECK-NEXT: rsbs r0, r0, #0 ; CHECK-NEXT: rsbs r0, r0, #0
; CHECK-NEXT: sbcs.w r0, r2, r1 ; CHECK-NEXT: sbcs.w r0, r2, r1
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r0, lt
; CHECK-NEXT: movlt r2, #1 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: vmov q2[2], q2[0], r0, r6 ; CHECK-NEXT: vmov q2[2], q2[0], r0, r6
; CHECK-NEXT: vand q0, q0, q2 ; CHECK-NEXT: vand q0, q0, q2
; CHECK-NEXT: vmov.f32 s1, s2 ; CHECK-NEXT: vmov.f32 s1, s2
; CHECK-NEXT: vmov.f32 s2, s4 ; CHECK-NEXT: vmov.f32 s2, s4
; CHECK-NEXT: vmov.f32 s3, s6 ; CHECK-NEXT: vmov.f32 s3, s6
; CHECK-NEXT: vpop {d8, d9} ; CHECK-NEXT: vpop {d8, d9}
; CHECK-NEXT: add sp, #4 ; CHECK-NEXT: add sp, #4
Show All 22 Lines
; CHECK-NEXT: bl __aeabi_d2lz ; CHECK-NEXT: bl __aeabi_d2lz
; CHECK-NEXT: mov r4, r0 ; CHECK-NEXT: mov r4, r0
; CHECK-NEXT: mov r5, r1 ; CHECK-NEXT: mov r5, r1
; CHECK-NEXT: vmov r0, r1, d8 ; CHECK-NEXT: vmov r0, r1, d8
; CHECK-NEXT: bl __aeabi_d2lz ; CHECK-NEXT: bl __aeabi_d2lz
; CHECK-NEXT: vmov q0[2], q0[0], r0, r4 ; CHECK-NEXT: vmov q0[2], q0[0], r0, r4
; CHECK-NEXT: movw r4, #32767 ; CHECK-NEXT: movw r4, #32767
; CHECK-NEXT: vmov q0[3], q0[1], r1, r5 ; CHECK-NEXT: vmov q0[3], q0[1], r1, r5
; CHECK-NEXT: adr.w r12, .LCPI9_0 ; CHECK-NEXT: adr r5, .LCPI9_0
; CHECK-NEXT: vmov r1, r2, d1 ; CHECK-NEXT: vmov r0, r1, d1
; CHECK-NEXT: vldrw.u32 q1, [r12] ; CHECK-NEXT: vldrw.u32 q1, [r5]
; CHECK-NEXT: vmov r3, r5, d0 ; CHECK-NEXT: vmov r2, r3, d0
; CHECK-NEXT: movw lr, #32768
; CHECK-NEXT: movt lr, #65535
; CHECK-NEXT: mov.w r12, #-1 ; CHECK-NEXT: mov.w r12, #-1
; CHECK-NEXT: movs r0, #0 ; CHECK-NEXT: adr r5, .LCPI9_1
; CHECK-NEXT: subs r1, r1, r4 ; CHECK-NEXT: subs r0, r0, r4
; CHECK-NEXT: sbcs r1, r2, #0 ; CHECK-NEXT: sbcs r0, r1, #0
; CHECK-NEXT: mov.w r1, #0 ; CHECK-NEXT: cset r0, lt
; CHECK-NEXT: it lt ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: movlt r1, #1 ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: subs r1, r2, r4
; CHECK-NEXT: sbcs r1, r3, #0
; CHECK-NEXT: movw r4, #32768
; CHECK-NEXT: cset r1, lt
; CHECK-NEXT: movt r4, #65535
; CHECK-NEXT: cmp r1, #0 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: subs r2, r3, r4 ; CHECK-NEXT: vmov q2[2], q2[0], r1, r0
; CHECK-NEXT: sbcs r2, r5, #0 ; CHECK-NEXT: vmov q2[3], q2[1], r1, r0
; CHECK-NEXT: adr r4, .LCPI9_1
; CHECK-NEXT: mov.w r2, #0
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r2, #1
; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: csetm r2, ne
; CHECK-NEXT: vmov q2[2], q2[0], r2, r1
; CHECK-NEXT: vmov q2[3], q2[1], r2, r1
; CHECK-NEXT: vbic q1, q1, q2 ; CHECK-NEXT: vbic q1, q1, q2
; CHECK-NEXT: vand q0, q0, q2 ; CHECK-NEXT: vand q0, q0, q2
; CHECK-NEXT: vorr q0, q0, q1 ; CHECK-NEXT: vorr q0, q0, q1
; CHECK-NEXT: vldrw.u32 q1, [r4] ; CHECK-NEXT: vldrw.u32 q1, [r5]
; CHECK-NEXT: vmov r1, r2, d1 ; CHECK-NEXT: vmov r0, r1, d1
; CHECK-NEXT: vmov r3, r5, d0 ; CHECK-NEXT: vmov r2, r3, d0
; CHECK-NEXT: subs.w r1, lr, r1 ; CHECK-NEXT: subs r0, r4, r0
; CHECK-NEXT: sbcs.w r1, r12, r2 ; CHECK-NEXT: sbcs.w r0, r12, r1
; CHECK-NEXT: mov.w r1, #0 ; CHECK-NEXT: cset r0, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r1, #1
; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: subs.w r2, lr, r3
; CHECK-NEXT: sbcs.w r2, r12, r5
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r0, #1
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: vmov q2[2], q2[0], r0, r1 ; CHECK-NEXT: subs r1, r4, r2
; CHECK-NEXT: vmov q2[3], q2[1], r0, r1 ; CHECK-NEXT: sbcs.w r1, r12, r3
; CHECK-NEXT: cset r1, lt
; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: vmov q2[2], q2[0], r1, r0
; CHECK-NEXT: vmov q2[3], q2[1], r1, r0
; CHECK-NEXT: vbic q1, q1, q2 ; CHECK-NEXT: vbic q1, q1, q2
; CHECK-NEXT: vand q0, q0, q2 ; CHECK-NEXT: vand q0, q0, q2
; CHECK-NEXT: vorr q0, q0, q1 ; CHECK-NEXT: vorr q0, q0, q1
; CHECK-NEXT: vpop {d8, d9} ; CHECK-NEXT: vpop {d8, d9}
; CHECK-NEXT: pop {r4, r5, r7, pc} ; CHECK-NEXT: pop {r4, r5, r7, pc}
; CHECK-NEXT: .p2align 4 ; CHECK-NEXT: .p2align 4
; CHECK-NEXT: @ %bb.1: ; CHECK-NEXT: @ %bb.1:
; CHECK-NEXT: .LCPI9_0: ; CHECK-NEXT: .LCPI9_0:
Show All 26 Lines
; CHECK-NEXT: vmov q4, q0 ; CHECK-NEXT: vmov q4, q0
; CHECK-NEXT: vmov r0, r1, d9 ; CHECK-NEXT: vmov r0, r1, d9
; CHECK-NEXT: bl __aeabi_d2ulz ; CHECK-NEXT: bl __aeabi_d2ulz
; CHECK-NEXT: mov r4, r0 ; CHECK-NEXT: mov r4, r0
; CHECK-NEXT: mov r5, r1 ; CHECK-NEXT: mov r5, r1
; CHECK-NEXT: vmov r0, r1, d8 ; CHECK-NEXT: vmov r0, r1, d8
; CHECK-NEXT: bl __aeabi_d2ulz ; CHECK-NEXT: bl __aeabi_d2ulz
; CHECK-NEXT: vmov q0[2], q0[0], r0, r4 ; CHECK-NEXT: vmov q0[2], q0[0], r0, r4
; CHECK-NEXT: movw r4, #65535 ; CHECK-NEXT: vmov.i64 q1, #0xffff
; CHECK-NEXT: vmov q0[3], q0[1], r1, r5 ; CHECK-NEXT: vmov q0[3], q0[1], r1, r5
; CHECK-NEXT: movs r5, #0 ; CHECK-NEXT: movw r5, #65535
; CHECK-NEXT: vmov r0, r1, d1 ; CHECK-NEXT: vmov r0, r1, d1
; CHECK-NEXT: vmov.i64 q1, #0xffff
; CHECK-NEXT: vmov r2, r3, d0 ; CHECK-NEXT: vmov r2, r3, d0
; CHECK-NEXT: subs r0, r0, r4 ; CHECK-NEXT: subs r0, r0, r5
; CHECK-NEXT: sbcs r0, r1, #0 ; CHECK-NEXT: sbcs r0, r1, #0
; CHECK-NEXT: mov.w r0, #0 ; CHECK-NEXT: cset r0, lo
; CHECK-NEXT: it lo
; CHECK-NEXT: movlo r0, #1
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: subs r1, r2, r4 ; CHECK-NEXT: subs r1, r2, r5
; CHECK-NEXT: sbcs r1, r3, #0 ; CHECK-NEXT: sbcs r1, r3, #0
; CHECK-NEXT: it lo ; CHECK-NEXT: cset r1, lo
; CHECK-NEXT: movlo r5, #1 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: vmov q2[2], q2[0], r1, r0 ; CHECK-NEXT: vmov q2[2], q2[0], r1, r0
; CHECK-NEXT: vmov q2[3], q2[1], r1, r0 ; CHECK-NEXT: vmov q2[3], q2[1], r1, r0
; CHECK-NEXT: vbic q1, q1, q2 ; CHECK-NEXT: vbic q1, q1, q2
; CHECK-NEXT: vand q0, q0, q2 ; CHECK-NEXT: vand q0, q0, q2
; CHECK-NEXT: vorr q0, q0, q1 ; CHECK-NEXT: vorr q0, q0, q1
; CHECK-NEXT: vpop {d8, d9} ; CHECK-NEXT: vpop {d8, d9}
; CHECK-NEXT: pop {r4, r5, r7, pc} ; CHECK-NEXT: pop {r4, r5, r7, pc}
Show All 15 Lines
; CHECK-NEXT: vmov q4, q0 ; CHECK-NEXT: vmov q4, q0
; CHECK-NEXT: vmov r0, r1, d9 ; CHECK-NEXT: vmov r0, r1, d9
; CHECK-NEXT: bl __aeabi_d2lz ; CHECK-NEXT: bl __aeabi_d2lz
; CHECK-NEXT: mov r4, r0 ; CHECK-NEXT: mov r4, r0
; CHECK-NEXT: mov r5, r1 ; CHECK-NEXT: mov r5, r1
; CHECK-NEXT: vmov r0, r1, d8 ; CHECK-NEXT: vmov r0, r1, d8
; CHECK-NEXT: bl __aeabi_d2lz ; CHECK-NEXT: bl __aeabi_d2lz
; CHECK-NEXT: vmov q0[2], q0[0], r0, r4 ; CHECK-NEXT: vmov q0[2], q0[0], r0, r4
; CHECK-NEXT: movw r4, #65535
; CHECK-NEXT: vmov q0[3], q0[1], r1, r5
; CHECK-NEXT: vmov.i64 q1, #0xffff ; CHECK-NEXT: vmov.i64 q1, #0xffff
; CHECK-NEXT: vmov r1, r2, d1 ; CHECK-NEXT: vmov q0[3], q0[1], r1, r5
; CHECK-NEXT: movs r0, #0 ; CHECK-NEXT: movw r5, #65535
; CHECK-NEXT: vmov r3, r5, d0 ; CHECK-NEXT: vmov r0, r1, d1
; CHECK-NEXT: subs r1, r1, r4 ; CHECK-NEXT: vmov r2, r3, d0
; CHECK-NEXT: sbcs r1, r2, #0 ; CHECK-NEXT: subs r0, r0, r5
; CHECK-NEXT: mov.w r1, #0 ; CHECK-NEXT: sbcs r0, r1, #0
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r0, lt
; CHECK-NEXT: movlt r1, #1 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: subs r1, r2, r5
; CHECK-NEXT: sbcs r1, r3, #0
; CHECK-NEXT: mov.w r5, #0
; CHECK-NEXT: cset r1, lt
; CHECK-NEXT: cmp r1, #0 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: subs r2, r3, r4 ; CHECK-NEXT: vmov q2[2], q2[0], r1, r0
; CHECK-NEXT: sbcs r2, r5, #0 ; CHECK-NEXT: vmov q2[3], q2[1], r1, r0
; CHECK-NEXT: mov.w r2, #0
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r2, #1
; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: csetm r2, ne
; CHECK-NEXT: vmov q2[2], q2[0], r2, r1
; CHECK-NEXT: vmov q2[3], q2[1], r2, r1
; CHECK-NEXT: vbic q1, q1, q2 ; CHECK-NEXT: vbic q1, q1, q2
; CHECK-NEXT: vand q0, q0, q2 ; CHECK-NEXT: vand q0, q0, q2
; CHECK-NEXT: vorr q0, q0, q1 ; CHECK-NEXT: vorr q0, q0, q1
; CHECK-NEXT: vmov r1, r2, d1 ; CHECK-NEXT: vmov r0, r1, d1
; CHECK-NEXT: vmov r3, r5, d0 ; CHECK-NEXT: vmov r2, r3, d0
; CHECK-NEXT: rsbs r1, r1, #0 ; CHECK-NEXT: rsbs r0, r0, #0
; CHECK-NEXT: sbcs.w r1, r0, r2 ; CHECK-NEXT: sbcs.w r0, r5, r1
; CHECK-NEXT: mov.w r1, #0 ; CHECK-NEXT: cset r0, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r1, #1
; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: rsbs r2, r3, #0
; CHECK-NEXT: sbcs.w r2, r0, r5
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r0, #1
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: vmov q1[2], q1[0], r0, r1 ; CHECK-NEXT: rsbs r1, r2, #0
; CHECK-NEXT: vmov q1[3], q1[1], r0, r1 ; CHECK-NEXT: sbcs.w r1, r5, r3
; CHECK-NEXT: cset r1, lt
; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: vmov q1[2], q1[0], r1, r0
; CHECK-NEXT: vmov q1[3], q1[1], r1, r0
; CHECK-NEXT: vand q0, q0, q1 ; CHECK-NEXT: vand q0, q0, q1
; CHECK-NEXT: vpop {d8, d9} ; CHECK-NEXT: vpop {d8, d9}
; CHECK-NEXT: pop {r4, r5, r7, pc} ; CHECK-NEXT: pop {r4, r5, r7, pc}
entry: entry:
%conv = fptosi <2 x double> %x to <2 x i32> %conv = fptosi <2 x double> %x to <2 x i32>
%0 = icmp slt <2 x i32> %conv, <i32 65535, i32 65535> %0 = icmp slt <2 x i32> %conv, <i32 65535, i32 65535>
%spec.store.select = select <2 x i1> %0, <2 x i32> %conv, <2 x i32> <i32 65535, i32 65535> %spec.store.select = select <2 x i1> %0, <2 x i32> %conv, <2 x i32> <i32 65535, i32 65535>
%1 = icmp sgt <2 x i32> %spec.store.select, zeroinitializer %1 = icmp sgt <2 x i32> %spec.store.select, zeroinitializer
▲ Show 20 Lines • Show All 194 Lines • ▼ Show 20 Linesentry:
ret <8 x i16> %conv6 ret <8 x i16> %conv6
} }
; i64 saturate ; i64 saturate
define arm_aapcs_vfpcc <2 x i64> @stest_f64i64(<2 x double> %x) { define arm_aapcs_vfpcc <2 x i64> @stest_f64i64(<2 x double> %x) {
; CHECK-LABEL: stest_f64i64: ; CHECK-LABEL: stest_f64i64:
; CHECK: @ %bb.0: @ %entry ; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: .save {r4, r5, r6, r7, r8, r9, r10, r11, lr} ; CHECK-NEXT: .save {r4, r5, r6, r7, r8, r9, r10, lr}
; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, r9, r10, r11, lr} ; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, r9, r10, lr}
; CHECK-NEXT: .pad #4
; CHECK-NEXT: sub sp, #4
; CHECK-NEXT: .vsave {d8, d9} ; CHECK-NEXT: .vsave {d8, d9}
; CHECK-NEXT: vpush {d8, d9} ; CHECK-NEXT: vpush {d8, d9}
; CHECK-NEXT: vmov q4, q0 ; CHECK-NEXT: vmov q4, q0
; CHECK-NEXT: movs r4, #0
; CHECK-NEXT: vmov r0, r1, d9 ; CHECK-NEXT: vmov r0, r1, d9
; CHECK-NEXT: mvn r9, #-2147483648
; CHECK-NEXT: bl __fixdfti ; CHECK-NEXT: bl __fixdfti
; CHECK-NEXT: subs.w r7, r0, #-1 ; CHECK-NEXT: vmov r12, lr, d8
; CHECK-NEXT: mov.w r5, #-1 ; CHECK-NEXT: subs.w r4, r0, #-1
; CHECK-NEXT: sbcs.w r7, r1, r9 ; CHECK-NEXT: mvn r9, #-2147483648
; CHECK-NEXT: sbcs.w r4, r1, r9
; CHECK-NEXT: sbcs r4, r2, #0
; CHECK-NEXT: mov.w r7, #-1
; CHECK-NEXT: sbcs r4, r3, #0
; CHECK-NEXT: mov.w r10, #-2147483648 ; CHECK-NEXT: mov.w r10, #-2147483648
; CHECK-NEXT: sbcs r7, r2, #0 ; CHECK-NEXT: cset r4, lt
; CHECK-NEXT: sbcs r7, r3, #0 ; CHECK-NEXT: cmp r4, #0
; CHECK-NEXT: mov.w r7, #0 ; CHECK-NEXT: csel r3, r3, r4, ne
; CHECK-NEXT: it lt ; CHECK-NEXT: csel r2, r2, r4, ne
; CHECK-NEXT: movlt r7, #1 ; CHECK-NEXT: csel r4, r0, r7, ne
; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csel r11, r0, r5, ne
; CHECK-NEXT: csel r3, r3, r7, ne
; CHECK-NEXT: csel r2, r2, r7, ne
; CHECK-NEXT: csel r1, r1, r9, ne ; CHECK-NEXT: csel r1, r1, r9, ne
; CHECK-NEXT: rsbs.w r0, r11, #0 ; CHECK-NEXT: rsbs r0, r4, #0
; CHECK-NEXT: mov.w r7, #0
; CHECK-NEXT: sbcs.w r0, r10, r1 ; CHECK-NEXT: sbcs.w r0, r10, r1
; CHECK-NEXT: sbcs.w r0, r5, r2 ; CHECK-NEXT: sbcs.w r0, r7, r2
; CHECK-NEXT: sbcs.w r0, r5, r3 ; CHECK-NEXT: sbcs.w r0, r7, r3
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r5, lt
; CHECK-NEXT: movlt r7, #1 ; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csel r8, r1, r10, ne ; CHECK-NEXT: csel r8, r1, r10, ne
; CHECK-NEXT: vmov r0, r1, d8 ; CHECK-NEXT: mov r0, r12
; CHECK-NEXT: mov r1, lr
; CHECK-NEXT: bl __fixdfti ; CHECK-NEXT: bl __fixdfti
; CHECK-NEXT: subs.w r6, r0, #-1 ; CHECK-NEXT: subs.w r6, r0, #-1
; CHECK-NEXT: sbcs.w r6, r1, r9 ; CHECK-NEXT: sbcs.w r6, r1, r9
; CHECK-NEXT: sbcs r6, r2, #0 ; CHECK-NEXT: sbcs r6, r2, #0
; CHECK-NEXT: sbcs r6, r3, #0 ; CHECK-NEXT: sbcs r6, r3, #0
; CHECK-NEXT: mov.w r6, #0 ; CHECK-NEXT: cset r6, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r6, #1
; CHECK-NEXT: cmp r6, #0 ; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: csel r0, r0, r5, ne ; CHECK-NEXT: csel r0, r0, r7, ne
; CHECK-NEXT: csel r1, r1, r9, ne ; CHECK-NEXT: csel r1, r1, r9, ne
; CHECK-NEXT: csel r3, r3, r6, ne ; CHECK-NEXT: csel r3, r3, r6, ne
; CHECK-NEXT: csel r2, r2, r6, ne ; CHECK-NEXT: csel r2, r2, r6, ne
; CHECK-NEXT: rsbs r6, r0, #0 ; CHECK-NEXT: rsbs r6, r0, #0
; CHECK-NEXT: sbcs.w r6, r10, r1 ; CHECK-NEXT: sbcs.w r6, r10, r1
; CHECK-NEXT: sbcs.w r2, r5, r2 ; CHECK-NEXT: sbcs.w r2, r7, r2
; CHECK-NEXT: sbcs.w r2, r5, r3 ; CHECK-NEXT: sbcs.w r2, r7, r3
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r2, lt
; CHECK-NEXT: movlt r4, #1 ; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: cmp r4, #0
; CHECK-NEXT: csel r1, r1, r10, ne ; CHECK-NEXT: csel r1, r1, r10, ne
; CHECK-NEXT: cmp r7, #0 ; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: csel r2, r11, r7, ne ; CHECK-NEXT: csel r3, r4, r5, ne
; CHECK-NEXT: cmp r4, #0 ; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: csel r0, r0, r4, ne ; CHECK-NEXT: csel r0, r0, r2, ne
; CHECK-NEXT: vmov q0[2], q0[0], r0, r2 ; CHECK-NEXT: vmov q0[2], q0[0], r0, r3
; CHECK-NEXT: vmov q0[3], q0[1], r1, r8 ; CHECK-NEXT: vmov q0[3], q0[1], r1, r8
; CHECK-NEXT: vpop {d8, d9} ; CHECK-NEXT: vpop {d8, d9}
; CHECK-NEXT: add sp, #4 ; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, r10, pc}
; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, r10, r11, pc}
entry: entry:
%conv = fptosi <2 x double> %x to <2 x i128> %conv = fptosi <2 x double> %x to <2 x i128>
%0 = icmp slt <2 x i128> %conv, <i128 9223372036854775807, i128 9223372036854775807> %0 = icmp slt <2 x i128> %conv, <i128 9223372036854775807, i128 9223372036854775807>
%spec.store.select = select <2 x i1> %0, <2 x i128> %conv, <2 x i128> <i128 9223372036854775807, i128 9223372036854775807> %spec.store.select = select <2 x i1> %0, <2 x i128> %conv, <2 x i128> <i128 9223372036854775807, i128 9223372036854775807>
%1 = icmp sgt <2 x i128> %spec.store.select, <i128 -9223372036854775808, i128 -9223372036854775808> %1 = icmp sgt <2 x i128> %spec.store.select, <i128 -9223372036854775808, i128 -9223372036854775808>
%spec.store.select7 = select <2 x i1> %1, <2 x i128> %spec.store.select, <2 x i128> <i128 -9223372036854775808, i128 -9223372036854775808> %spec.store.select7 = select <2 x i1> %1, <2 x i128> %spec.store.select, <2 x i128> <i128 -9223372036854775808, i128 -9223372036854775808>
%conv6 = trunc <2 x i128> %spec.store.select7 to <2 x i64> %conv6 = trunc <2 x i128> %spec.store.select7 to <2 x i64>
ret <2 x i64> %conv6 ret <2 x i64> %conv6
} }
define arm_aapcs_vfpcc <2 x i64> @utest_f64i64(<2 x double> %x) { define arm_aapcs_vfpcc <2 x i64> @utest_f64i64(<2 x double> %x) {
; CHECK-LABEL: utest_f64i64: ; CHECK-LABEL: utest_f64i64:
; CHECK: @ %bb.0: @ %entry ; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: .save {r4, r5, r6, r7, r8, lr} ; CHECK-NEXT: .save {r4, r5, r6, r7, lr}
; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, lr} ; CHECK-NEXT: push {r4, r5, r6, r7, lr}
; CHECK-NEXT: .pad #4
; CHECK-NEXT: sub sp, #4
; CHECK-NEXT: .vsave {d8, d9} ; CHECK-NEXT: .vsave {d8, d9}
; CHECK-NEXT: vpush {d8, d9} ; CHECK-NEXT: vpush {d8, d9}
; CHECK-NEXT: vmov q4, q0 ; CHECK-NEXT: vmov q4, q0
; CHECK-NEXT: vmov r0, r1, d9 ; CHECK-NEXT: vmov r0, r1, d9
; CHECK-NEXT: bl __fixunsdfti ; CHECK-NEXT: bl __fixunsdfti
; CHECK-NEXT: mov r8, r1 ; CHECK-NEXT: mov r5, r1
; CHECK-NEXT: vmov r4, r1, d8 ; CHECK-NEXT: vmov r4, r1, d8
; CHECK-NEXT: subs r2, #1 ; CHECK-NEXT: subs r2, #1
; CHECK-NEXT: mov.w r7, #0
; CHECK-NEXT: sbcs r2, r3, #0 ; CHECK-NEXT: sbcs r2, r3, #0
; CHECK-NEXT: it lo ; CHECK-NEXT: cset r6, lo
; CHECK-NEXT: movlo r7, #1 ; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: cmp r7, #0 ; CHECK-NEXT: csel r7, r0, r6, ne
; CHECK-NEXT: mov.w r6, #0
; CHECK-NEXT: csel r5, r0, r7, ne
; CHECK-NEXT: mov r0, r4 ; CHECK-NEXT: mov r0, r4
; CHECK-NEXT: bl __fixunsdfti ; CHECK-NEXT: bl __fixunsdfti
; CHECK-NEXT: subs r2, #1 ; CHECK-NEXT: subs r2, #1
; CHECK-NEXT: sbcs r2, r3, #0 ; CHECK-NEXT: sbcs r2, r3, #0
; CHECK-NEXT: it lo ; CHECK-NEXT: cset r2, lo
; CHECK-NEXT: movlo r6, #1 ; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: cmp r6, #0 ; CHECK-NEXT: csel r0, r0, r2, ne
; CHECK-NEXT: csel r0, r0, r6, ne
; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csel r2, r8, r7, ne
; CHECK-NEXT: cmp r6, #0 ; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: csel r1, r1, r6, ne ; CHECK-NEXT: csel r3, r5, r6, ne
; CHECK-NEXT: vmov q0[2], q0[0], r0, r5 ; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: vmov q0[3], q0[1], r1, r2 ; CHECK-NEXT: csel r1, r1, r2, ne
; CHECK-NEXT: vmov q0[2], q0[0], r0, r7
; CHECK-NEXT: vmov q0[3], q0[1], r1, r3
; CHECK-NEXT: vpop {d8, d9} ; CHECK-NEXT: vpop {d8, d9}
; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, pc} ; CHECK-NEXT: add sp, #4
; CHECK-NEXT: pop {r4, r5, r6, r7, pc}
entry: entry:
%conv = fptoui <2 x double> %x to <2 x i128> %conv = fptoui <2 x double> %x to <2 x i128>
%0 = icmp ult <2 x i128> %conv, <i128 18446744073709551616, i128 18446744073709551616> %0 = icmp ult <2 x i128> %conv, <i128 18446744073709551616, i128 18446744073709551616>
%spec.store.select = select <2 x i1> %0, <2 x i128> %conv, <2 x i128> <i128 18446744073709551616, i128 18446744073709551616> %spec.store.select = select <2 x i1> %0, <2 x i128> %conv, <2 x i128> <i128 18446744073709551616, i128 18446744073709551616>
%conv6 = trunc <2 x i128> %spec.store.select to <2 x i64> %conv6 = trunc <2 x i128> %spec.store.select to <2 x i64>
ret <2 x i64> %conv6 ret <2 x i64> %conv6
} }
define arm_aapcs_vfpcc <2 x i64> @ustest_f64i64(<2 x double> %x) { define arm_aapcs_vfpcc <2 x i64> @ustest_f64i64(<2 x double> %x) {
; CHECK-LABEL: ustest_f64i64: ; CHECK-LABEL: ustest_f64i64:
; CHECK: @ %bb.0: @ %entry ; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: .save {r4, r5, r6, r7, r8, r9, lr} ; CHECK-NEXT: .save {r4, r5, r6, r7, r8, r9, lr}
; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, r9, lr} ; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, r9, lr}
; CHECK-NEXT: .pad #4 ; CHECK-NEXT: .pad #4
; CHECK-NEXT: sub sp, #4 ; CHECK-NEXT: sub sp, #4
; CHECK-NEXT: .vsave {d8, d9} ; CHECK-NEXT: .vsave {d8, d9}
; CHECK-NEXT: vpush {d8, d9} ; CHECK-NEXT: vpush {d8, d9}
; CHECK-NEXT: vmov q4, q0 ; CHECK-NEXT: vmov q4, q0
; CHECK-NEXT: vmov r0, r1, d9 ; CHECK-NEXT: vmov r0, r1, d9
; CHECK-NEXT: bl __fixdfti ; CHECK-NEXT: bl __fixdfti
; CHECK-NEXT: vmov r12, lr, d8 ; CHECK-NEXT: vmov r12, lr, d8
; CHECK-NEXT: subs r5, r2, #1 ; CHECK-NEXT: subs r4, r2, #1
; CHECK-NEXT: sbcs r5, r3, #0 ; CHECK-NEXT: sbcs r4, r3, #0
; CHECK-NEXT: mov.w r6, #0
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r6, #1
; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: mov.w r8, #1 ; CHECK-NEXT: mov.w r8, #1
; CHECK-NEXT: csel r0, r0, r6, ne ; CHECK-NEXT: cset r5, lt
; CHECK-NEXT: csel r3, r3, r6, ne ; CHECK-NEXT: movs r7, #0
; CHECK-NEXT: csel r5, r1, r6, ne ; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: csel r0, r0, r5, ne
; CHECK-NEXT: csel r3, r3, r5, ne
; CHECK-NEXT: csel r2, r2, r8, ne ; CHECK-NEXT: csel r2, r2, r8, ne
; CHECK-NEXT: csel r4, r1, r5, ne
; CHECK-NEXT: rsbs r1, r0, #0 ; CHECK-NEXT: rsbs r1, r0, #0
; CHECK-NEXT: mov.w r4, #0 ; CHECK-NEXT: sbcs.w r1, r7, r4
; CHECK-NEXT: sbcs.w r1, r4, r5 ; CHECK-NEXT: sbcs.w r1, r7, r2
; CHECK-NEXT: sbcs.w r1, r4, r2 ; CHECK-NEXT: sbcs.w r1, r7, r3
; CHECK-NEXT: mov.w r6, #0 ; CHECK-NEXT: cset r6, lt
; CHECK-NEXT: sbcs.w r1, r4, r3
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r6, #1
; CHECK-NEXT: cmp r6, #0 ; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: csel r9, r0, r6, ne ; CHECK-NEXT: csel r9, r0, r6, ne
; CHECK-NEXT: mov r0, r12 ; CHECK-NEXT: mov r0, r12
; CHECK-NEXT: mov r1, lr ; CHECK-NEXT: mov r1, lr
; CHECK-NEXT: bl __fixdfti ; CHECK-NEXT: bl __fixdfti
; CHECK-NEXT: subs r7, r2, #1 ; CHECK-NEXT: subs r5, r2, #1
; CHECK-NEXT: sbcs r7, r3, #0 ; CHECK-NEXT: sbcs r5, r3, #0
; CHECK-NEXT: mov.w r7, #0 ; CHECK-NEXT: cset r5, lt
; CHECK-NEXT: it lt ; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: movlt r7, #1 ; CHECK-NEXT: csel r0, r0, r5, ne
; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csel r0, r0, r7, ne
; CHECK-NEXT: csel r2, r2, r8, ne ; CHECK-NEXT: csel r2, r2, r8, ne
; CHECK-NEXT: csel r3, r3, r7, ne ; CHECK-NEXT: csel r3, r3, r5, ne
; CHECK-NEXT: csel r1, r1, r7, ne ; CHECK-NEXT: csel r1, r1, r5, ne
; CHECK-NEXT: rsbs r7, r0, #0 ; CHECK-NEXT: rsbs r5, r0, #0
; CHECK-NEXT: sbcs.w r7, r4, r1 ; CHECK-NEXT: sbcs.w r5, r7, r1
; CHECK-NEXT: sbcs.w r2, r4, r2 ; CHECK-NEXT: sbcs.w r2, r7, r2
; CHECK-NEXT: sbcs.w r2, r4, r3 ; CHECK-NEXT: sbcs.w r2, r7, r3
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r2, lt
; CHECK-NEXT: movlt r4, #1 ; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: cmp r4, #0 ; CHECK-NEXT: csel r0, r0, r2, ne
; CHECK-NEXT: csel r0, r0, r4, ne
; CHECK-NEXT: cmp r6, #0 ; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: csel r2, r5, r6, ne ; CHECK-NEXT: csel r3, r4, r6, ne
; CHECK-NEXT: cmp r4, #0 ; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: csel r1, r1, r4, ne ; CHECK-NEXT: csel r1, r1, r2, ne
; CHECK-NEXT: vmov q0[2], q0[0], r0, r9 ; CHECK-NEXT: vmov q0[2], q0[0], r0, r9
; CHECK-NEXT: vmov q0[3], q0[1], r1, r2 ; CHECK-NEXT: vmov q0[3], q0[1], r1, r3
; CHECK-NEXT: vpop {d8, d9} ; CHECK-NEXT: vpop {d8, d9}
; CHECK-NEXT: add sp, #4 ; CHECK-NEXT: add sp, #4
; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, pc} ; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, pc}
entry: entry:
%conv = fptosi <2 x double> %x to <2 x i128> %conv = fptosi <2 x double> %x to <2 x i128>
%0 = icmp slt <2 x i128> %conv, <i128 18446744073709551616, i128 18446744073709551616> %0 = icmp slt <2 x i128> %conv, <i128 18446744073709551616, i128 18446744073709551616>
%spec.store.select = select <2 x i1> %0, <2 x i128> %conv, <2 x i128> <i128 18446744073709551616, i128 18446744073709551616> %spec.store.select = select <2 x i1> %0, <2 x i128> %conv, <2 x i128> <i128 18446744073709551616, i128 18446744073709551616>
%1 = icmp sgt <2 x i128> %spec.store.select, zeroinitializer %1 = icmp sgt <2 x i128> %spec.store.select, zeroinitializer
%spec.store.select7 = select <2 x i1> %1, <2 x i128> %spec.store.select, <2 x i128> zeroinitializer %spec.store.select7 = select <2 x i1> %1, <2 x i128> %spec.store.select, <2 x i128> zeroinitializer
%conv6 = trunc <2 x i128> %spec.store.select7 to <2 x i64> %conv6 = trunc <2 x i128> %spec.store.select7 to <2 x i64>
ret <2 x i64> %conv6 ret <2 x i64> %conv6
} }
define arm_aapcs_vfpcc <2 x i64> @stest_f32i64(<2 x float> %x) { define arm_aapcs_vfpcc <2 x i64> @stest_f32i64(<2 x float> %x) {
; CHECK-LABEL: stest_f32i64: ; CHECK-LABEL: stest_f32i64:
; CHECK: @ %bb.0: @ %entry ; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: .save {r4, r5, r6, r7, r8, r9, r10, r11, lr} ; CHECK-NEXT: .save {r4, r5, r6, r7, r8, r9, r10, r11, lr}
; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, r9, r10, r11, lr} ; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, r9, r10, r11, lr}
; CHECK-NEXT: .pad #4 ; CHECK-NEXT: .pad #4
; CHECK-NEXT: sub sp, #4 ; CHECK-NEXT: sub sp, #4
; CHECK-NEXT: vmov r8, r0, d0 ; CHECK-NEXT: vmov r9, r0, d0
; CHECK-NEXT: bl __fixsfti ; CHECK-NEXT: bl __fixsfti
; CHECK-NEXT: subs.w r7, r0, #-1 ; CHECK-NEXT: subs.w r7, r0, #-1
; CHECK-NEXT: mvn r10, #-2147483648 ; CHECK-NEXT: mvn r10, #-2147483648
; CHECK-NEXT: sbcs.w r7, r1, r10 ; CHECK-NEXT: sbcs.w r7, r1, r10
; CHECK-NEXT: mov.w r11, #-2147483648 ; CHECK-NEXT: mov.w r4, #-1
; CHECK-NEXT: sbcs r7, r2, #0 ; CHECK-NEXT: sbcs r7, r2, #0
; CHECK-NEXT: mov.w r4, #0 ; CHECK-NEXT: mov.w r11, #-2147483648
; CHECK-NEXT: sbcs r7, r3, #0 ; CHECK-NEXT: sbcs r7, r3, #0
; CHECK-NEXT: mov.w r5, #0 ; CHECK-NEXT: cset r7, lt
; CHECK-NEXT: mov.w r7, #0
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r7, #1
; CHECK-NEXT: cmp r7, #0 ; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csel r5, r0, r4, ne
; CHECK-NEXT: csel r3, r3, r7, ne ; CHECK-NEXT: csel r3, r3, r7, ne
; CHECK-NEXT: csel r2, r2, r7, ne ; CHECK-NEXT: csel r2, r2, r7, ne
; CHECK-NEXT: mov.w r7, #-1
; CHECK-NEXT: csel r1, r1, r10, ne ; CHECK-NEXT: csel r1, r1, r10, ne
; CHECK-NEXT: csel r9, r0, r7, ne ; CHECK-NEXT: rsbs r0, r5, #0
; CHECK-NEXT: rsbs.w r0, r9, #0
; CHECK-NEXT: sbcs.w r0, r11, r1 ; CHECK-NEXT: sbcs.w r0, r11, r1
; CHECK-NEXT: sbcs.w r0, r7, r2 ; CHECK-NEXT: sbcs.w r0, r4, r2
; CHECK-NEXT: sbcs.w r0, r7, r3 ; CHECK-NEXT: sbcs.w r0, r4, r3
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r6, lt
; CHECK-NEXT: movlt r4, #1 ; CHECK-NEXT: mov r0, r9
; CHECK-NEXT: cmp r4, #0
; CHECK-NEXT: csel r0, r1, r11, ne
; CHECK-NEXT: str r0, [sp] @ 4-byte Spill
; CHECK-NEXT: mov r0, r8
; CHECK-NEXT: bl __fixsfti
; CHECK-NEXT: subs.w r6, r0, #-1
; CHECK-NEXT: sbcs.w r6, r1, r10
; CHECK-NEXT: sbcs r6, r2, #0
; CHECK-NEXT: sbcs r6, r3, #0
; CHECK-NEXT: mov.w r6, #0
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r6, #1
; CHECK-NEXT: cmp r6, #0 ; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: csel r0, r0, r7, ne ; CHECK-NEXT: csel r8, r1, r11, ne
; CHECK-NEXT: bl __fixsfti
; CHECK-NEXT: subs.w r7, r0, #-1
; CHECK-NEXT: sbcs.w r7, r1, r10
; CHECK-NEXT: sbcs r7, r2, #0
; CHECK-NEXT: sbcs r7, r3, #0
; CHECK-NEXT: cset r7, lt
; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csel r0, r0, r4, ne
; CHECK-NEXT: csel r1, r1, r10, ne ; CHECK-NEXT: csel r1, r1, r10, ne
; CHECK-NEXT: csel r3, r3, r6, ne ; CHECK-NEXT: csel r3, r3, r7, ne
; CHECK-NEXT: csel r2, r2, r6, ne ; CHECK-NEXT: csel r2, r2, r7, ne
; CHECK-NEXT: rsbs r6, r0, #0 ; CHECK-NEXT: rsbs r7, r0, #0
; CHECK-NEXT: sbcs.w r6, r11, r1 ; CHECK-NEXT: sbcs.w r7, r11, r1
; CHECK-NEXT: sbcs.w r2, r7, r2 ; CHECK-NEXT: sbcs.w r2, r4, r2
; CHECK-NEXT: sbcs.w r2, r7, r3 ; CHECK-NEXT: sbcs.w r2, r4, r3
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r2, lt
; CHECK-NEXT: movlt r5, #1 ; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: csel r1, r1, r11, ne ; CHECK-NEXT: csel r1, r1, r11, ne
; CHECK-NEXT: cmp r4, #0 ; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: csel r2, r9, r4, ne ; CHECK-NEXT: csel r3, r5, r6, ne
; CHECK-NEXT: cmp r5, #0 ; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: csel r0, r0, r5, ne ; CHECK-NEXT: csel r0, r0, r2, ne
; CHECK-NEXT: vmov q0[2], q0[0], r0, r2 ; CHECK-NEXT: vmov q0[2], q0[0], r0, r3
; CHECK-NEXT: ldr r0, [sp] @ 4-byte Reload ; CHECK-NEXT: vmov q0[3], q0[1], r1, r8
; CHECK-NEXT: vmov q0[3], q0[1], r1, r0
; CHECK-NEXT: add sp, #4 ; CHECK-NEXT: add sp, #4
; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, r10, r11, pc} ; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, r10, r11, pc}
entry: entry:
%conv = fptosi <2 x float> %x to <2 x i128> %conv = fptosi <2 x float> %x to <2 x i128>
%0 = icmp slt <2 x i128> %conv, <i128 9223372036854775807, i128 9223372036854775807> %0 = icmp slt <2 x i128> %conv, <i128 9223372036854775807, i128 9223372036854775807>
%spec.store.select = select <2 x i1> %0, <2 x i128> %conv, <2 x i128> <i128 9223372036854775807, i128 9223372036854775807> %spec.store.select = select <2 x i1> %0, <2 x i128> %conv, <2 x i128> <i128 9223372036854775807, i128 9223372036854775807>
%1 = icmp sgt <2 x i128> %spec.store.select, <i128 -9223372036854775808, i128 -9223372036854775808> %1 = icmp sgt <2 x i128> %spec.store.select, <i128 -9223372036854775808, i128 -9223372036854775808>
%spec.store.select7 = select <2 x i1> %1, <2 x i128> %spec.store.select, <2 x i128> <i128 -9223372036854775808, i128 -9223372036854775808> %spec.store.select7 = select <2 x i1> %1, <2 x i128> %spec.store.select, <2 x i128> <i128 -9223372036854775808, i128 -9223372036854775808>
%conv6 = trunc <2 x i128> %spec.store.select7 to <2 x i64> %conv6 = trunc <2 x i128> %spec.store.select7 to <2 x i64>
ret <2 x i64> %conv6 ret <2 x i64> %conv6
} }
define arm_aapcs_vfpcc <2 x i64> @utest_f32i64(<2 x float> %x) { define arm_aapcs_vfpcc <2 x i64> @utest_f32i64(<2 x float> %x) {
; CHECK-LABEL: utest_f32i64: ; CHECK-LABEL: utest_f32i64:
; CHECK: @ %bb.0: @ %entry ; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: .save {r4, r5, r6, r7, r8, lr} ; CHECK-NEXT: .save {r4, r5, r6, r7, lr}
; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, lr} ; CHECK-NEXT: push {r4, r5, r6, r7, lr}
; CHECK-NEXT: .pad #4
; CHECK-NEXT: sub sp, #4
; CHECK-NEXT: vmov r4, r0, d0 ; CHECK-NEXT: vmov r4, r0, d0
; CHECK-NEXT: bl __fixunssfti ; CHECK-NEXT: bl __fixunssfti
; CHECK-NEXT: mov r8, r1 ; CHECK-NEXT: mov r5, r1
; CHECK-NEXT: subs r1, r2, #1 ; CHECK-NEXT: subs r1, r2, #1
; CHECK-NEXT: sbcs r1, r3, #0 ; CHECK-NEXT: sbcs r1, r3, #0
; CHECK-NEXT: mov.w r7, #0 ; CHECK-NEXT: cset r6, lo
; CHECK-NEXT: it lo ; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: movlo r7, #1 ; CHECK-NEXT: csel r7, r0, r6, ne
; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csel r5, r0, r7, ne
; CHECK-NEXT: mov r0, r4 ; CHECK-NEXT: mov r0, r4
; CHECK-NEXT: movs r6, #0
; CHECK-NEXT: bl __fixunssfti ; CHECK-NEXT: bl __fixunssfti
; CHECK-NEXT: subs r2, #1 ; CHECK-NEXT: subs r2, #1
; CHECK-NEXT: sbcs r2, r3, #0 ; CHECK-NEXT: sbcs r2, r3, #0
; CHECK-NEXT: it lo ; CHECK-NEXT: cset r2, lo
; CHECK-NEXT: movlo r6, #1 ; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: cmp r6, #0 ; CHECK-NEXT: csel r0, r0, r2, ne
; CHECK-NEXT: csel r0, r0, r6, ne
; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csel r2, r8, r7, ne
; CHECK-NEXT: cmp r6, #0 ; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: csel r1, r1, r6, ne ; CHECK-NEXT: csel r3, r5, r6, ne
; CHECK-NEXT: vmov q0[2], q0[0], r0, r5 ; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: vmov q0[3], q0[1], r1, r2 ; CHECK-NEXT: csel r1, r1, r2, ne
; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, pc} ; CHECK-NEXT: vmov q0[2], q0[0], r0, r7
; CHECK-NEXT: vmov q0[3], q0[1], r1, r3
; CHECK-NEXT: add sp, #4
; CHECK-NEXT: pop {r4, r5, r6, r7, pc}
entry: entry:
%conv = fptoui <2 x float> %x to <2 x i128> %conv = fptoui <2 x float> %x to <2 x i128>
%0 = icmp ult <2 x i128> %conv, <i128 18446744073709551616, i128 18446744073709551616> %0 = icmp ult <2 x i128> %conv, <i128 18446744073709551616, i128 18446744073709551616>
%spec.store.select = select <2 x i1> %0, <2 x i128> %conv, <2 x i128> <i128 18446744073709551616, i128 18446744073709551616> %spec.store.select = select <2 x i1> %0, <2 x i128> %conv, <2 x i128> <i128 18446744073709551616, i128 18446744073709551616>
%conv6 = trunc <2 x i128> %spec.store.select to <2 x i64> %conv6 = trunc <2 x i128> %spec.store.select to <2 x i64>
ret <2 x i64> %conv6 ret <2 x i64> %conv6
} }
define arm_aapcs_vfpcc <2 x i64> @ustest_f32i64(<2 x float> %x) { define arm_aapcs_vfpcc <2 x i64> @ustest_f32i64(<2 x float> %x) {
; CHECK-LABEL: ustest_f32i64: ; CHECK-LABEL: ustest_f32i64:
; CHECK: @ %bb.0: @ %entry ; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: .save {r4, r5, r6, r7, r8, r9, lr} ; CHECK-NEXT: .save {r4, r5, r6, r7, r8, r9, lr}
; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, r9, lr} ; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, r9, lr}
; CHECK-NEXT: .pad #4 ; CHECK-NEXT: .pad #4
; CHECK-NEXT: sub sp, #4 ; CHECK-NEXT: sub sp, #4
; CHECK-NEXT: vmov r4, r0, d0 ; CHECK-NEXT: vmov r6, r0, d0
; CHECK-NEXT: bl __fixsfti
; CHECK-NEXT: subs r7, r2, #1
; CHECK-NEXT: mov.w r9, #1
; CHECK-NEXT: sbcs r7, r3, #0
; CHECK-NEXT: mov.w r5, #0
; CHECK-NEXT: mov.w r7, #0
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r7, #1
; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csel r0, r0, r7, ne
; CHECK-NEXT: csel r3, r3, r7, ne
; CHECK-NEXT: csel r6, r1, r7, ne
; CHECK-NEXT: csel r2, r2, r9, ne
; CHECK-NEXT: rsbs r1, r0, #0
; CHECK-NEXT: mov.w r7, #0
; CHECK-NEXT: sbcs.w r1, r5, r6
; CHECK-NEXT: sbcs.w r1, r5, r2
; CHECK-NEXT: sbcs.w r1, r5, r3
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r7, #1
; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csel r8, r0, r7, ne
; CHECK-NEXT: mov r0, r4
; CHECK-NEXT: bl __fixsfti ; CHECK-NEXT: bl __fixsfti
; CHECK-NEXT: subs r4, r2, #1 ; CHECK-NEXT: subs r5, r2, #1
; CHECK-NEXT: sbcs r4, r3, #0 ; CHECK-NEXT: mov.w r8, #1
; CHECK-NEXT: mov.w r4, #0 ; CHECK-NEXT: sbcs r5, r3, #0
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r4, lt
; CHECK-NEXT: movlt r4, #1
; CHECK-NEXT: cmp r4, #0 ; CHECK-NEXT: cmp r4, #0
; CHECK-NEXT: csel r0, r0, r4, ne ; CHECK-NEXT: csel r0, r0, r4, ne
; CHECK-NEXT: csel r2, r2, r9, ne
; CHECK-NEXT: csel r3, r3, r4, ne ; CHECK-NEXT: csel r3, r3, r4, ne
; CHECK-NEXT: csel r1, r1, r4, ne ; CHECK-NEXT: csel r5, r1, r4, ne
; CHECK-NEXT: rsbs r4, r0, #0 ; CHECK-NEXT: csel r2, r2, r8, ne
; CHECK-NEXT: sbcs.w r4, r5, r1 ; CHECK-NEXT: rsbs r1, r0, #0
; CHECK-NEXT: sbcs.w r2, r5, r2 ; CHECK-NEXT: mov.w r4, #0
; CHECK-NEXT: sbcs.w r2, r5, r3 ; CHECK-NEXT: sbcs.w r1, r4, r5
; CHECK-NEXT: it lt ; CHECK-NEXT: sbcs.w r1, r4, r2
; CHECK-NEXT: movlt r5, #1 ; CHECK-NEXT: sbcs.w r1, r4, r3
; CHECK-NEXT: cmp r5, #0 ; CHECK-NEXT: cset r7, lt
; CHECK-NEXT: csel r0, r0, r5, ne
; CHECK-NEXT: cmp r7, #0 ; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csel r2, r6, r7, ne ; CHECK-NEXT: csel r9, r0, r7, ne
; CHECK-NEXT: cmp r5, #0 ; CHECK-NEXT: mov r0, r6
; CHECK-NEXT: csel r1, r1, r5, ne ; CHECK-NEXT: bl __fixsfti
; CHECK-NEXT: vmov q0[2], q0[0], r0, r8 ; CHECK-NEXT: subs r6, r2, #1
; CHECK-NEXT: vmov q0[3], q0[1], r1, r2 ; CHECK-NEXT: sbcs r6, r3, #0
; CHECK-NEXT: cset r6, lt
; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: csel r0, r0, r6, ne
; CHECK-NEXT: csel r2, r2, r8, ne
; CHECK-NEXT: csel r3, r3, r6, ne
; CHECK-NEXT: csel r1, r1, r6, ne
; CHECK-NEXT: rsbs r6, r0, #0
; CHECK-NEXT: sbcs.w r6, r4, r1
; CHECK-NEXT: sbcs.w r2, r4, r2
; CHECK-NEXT: sbcs.w r2, r4, r3
; CHECK-NEXT: cset r2, lt
; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: csel r0, r0, r2, ne
; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csel r3, r5, r7, ne
; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: csel r1, r1, r2, ne
; CHECK-NEXT: vmov q0[2], q0[0], r0, r9
; CHECK-NEXT: vmov q0[3], q0[1], r1, r3
; CHECK-NEXT: add sp, #4 ; CHECK-NEXT: add sp, #4
; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, pc} ; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, pc}
entry: entry:
%conv = fptosi <2 x float> %x to <2 x i128> %conv = fptosi <2 x float> %x to <2 x i128>
%0 = icmp slt <2 x i128> %conv, <i128 18446744073709551616, i128 18446744073709551616> %0 = icmp slt <2 x i128> %conv, <i128 18446744073709551616, i128 18446744073709551616>
%spec.store.select = select <2 x i1> %0, <2 x i128> %conv, <2 x i128> <i128 18446744073709551616, i128 18446744073709551616> %spec.store.select = select <2 x i1> %0, <2 x i128> %conv, <2 x i128> <i128 18446744073709551616, i128 18446744073709551616>
%1 = icmp sgt <2 x i128> %spec.store.select, zeroinitializer %1 = icmp sgt <2 x i128> %spec.store.select, zeroinitializer
%spec.store.select7 = select <2 x i1> %1, <2 x i128> %spec.store.select, <2 x i128> zeroinitializer %spec.store.select7 = select <2 x i1> %1, <2 x i128> %spec.store.select, <2 x i128> zeroinitializer
%conv6 = trunc <2 x i128> %spec.store.select7 to <2 x i64> %conv6 = trunc <2 x i128> %spec.store.select7 to <2 x i64>
ret <2 x i64> %conv6 ret <2 x i64> %conv6
} }
define arm_aapcs_vfpcc <2 x i64> @stest_f16i64(<2 x half> %x) { define arm_aapcs_vfpcc <2 x i64> @stest_f16i64(<2 x half> %x) {
; CHECK-LABEL: stest_f16i64: ; CHECK-LABEL: stest_f16i64:
; CHECK: @ %bb.0: @ %entry ; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: .save {r4, r5, r6, r7, r8, r9, r10, r11, lr} ; CHECK-NEXT: .save {r4, r5, r6, r7, r8, r9, r10, lr}
; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, r9, r10, r11, lr} ; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, r9, r10, lr}
; CHECK-NEXT: .pad #4
; CHECK-NEXT: sub sp, #4
; CHECK-NEXT: .vsave {d8, d9} ; CHECK-NEXT: .vsave {d8, d9}
; CHECK-NEXT: vpush {d8, d9} ; CHECK-NEXT: vpush {d8, d9}
; CHECK-NEXT: vmov.u16 r0, q0[1] ; CHECK-NEXT: vmov.u16 r0, q0[1]
; CHECK-NEXT: vmov q4, q0 ; CHECK-NEXT: vmov q4, q0
; CHECK-NEXT: bl __fixhfti ; CHECK-NEXT: bl __fixhfti
; CHECK-NEXT: subs.w r7, r0, #-1 ; CHECK-NEXT: subs.w r7, r0, #-1
; CHECK-NEXT: mvn r9, #-2147483648 ; CHECK-NEXT: mvn r9, #-2147483648
; CHECK-NEXT: sbcs.w r7, r1, r9 ; CHECK-NEXT: sbcs.w r7, r1, r9
; CHECK-NEXT: mov.w r6, #-1
; CHECK-NEXT: sbcs r7, r2, #0
; CHECK-NEXT: mov.w r10, #-2147483648 ; CHECK-NEXT: mov.w r10, #-2147483648
; CHECK-NEXT: sbcs r7, r2, #0
; CHECK-NEXT: sbcs r7, r3, #0 ; CHECK-NEXT: sbcs r7, r3, #0
; CHECK-NEXT: mov.w r4, #0 ; CHECK-NEXT: cset r7, lt
; CHECK-NEXT: mov.w r7, #0
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r7, #1
; CHECK-NEXT: cmp r7, #0 ; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csel r11, r0, r6, ne
; CHECK-NEXT: csel r3, r3, r7, ne ; CHECK-NEXT: csel r3, r3, r7, ne
; CHECK-NEXT: csel r2, r2, r7, ne ; CHECK-NEXT: csel r2, r2, r7, ne
; CHECK-NEXT: mov.w r7, #-1
; CHECK-NEXT: csel r1, r1, r9, ne ; CHECK-NEXT: csel r1, r1, r9, ne
; CHECK-NEXT: rsbs.w r0, r11, #0 ; CHECK-NEXT: csel r4, r0, r7, ne
; CHECK-NEXT: mov.w r7, #0 ; CHECK-NEXT: rsbs r0, r4, #0
; CHECK-NEXT: sbcs.w r0, r10, r1 ; CHECK-NEXT: sbcs.w r0, r10, r1
; CHECK-NEXT: sbcs.w r0, r6, r2 ; CHECK-NEXT: sbcs.w r0, r7, r2
; CHECK-NEXT: sbcs.w r0, r6, r3 ; CHECK-NEXT: sbcs.w r0, r7, r3
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r5, lt
; CHECK-NEXT: movlt r7, #1
; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: vmov.u16 r0, q4[0] ; CHECK-NEXT: vmov.u16 r0, q4[0]
; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: csel r8, r1, r10, ne ; CHECK-NEXT: csel r8, r1, r10, ne
; CHECK-NEXT: bl __fixhfti ; CHECK-NEXT: bl __fixhfti
; CHECK-NEXT: subs.w r5, r0, #-1 ; CHECK-NEXT: subs.w r6, r0, #-1
; CHECK-NEXT: sbcs.w r5, r1, r9 ; CHECK-NEXT: sbcs.w r6, r1, r9
; CHECK-NEXT: sbcs r5, r2, #0 ; CHECK-NEXT: sbcs r6, r2, #0
; CHECK-NEXT: sbcs r5, r3, #0 ; CHECK-NEXT: sbcs r6, r3, #0
; CHECK-NEXT: mov.w r5, #0 ; CHECK-NEXT: cset r6, lt
; CHECK-NEXT: it lt ; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: movlt r5, #1 ; CHECK-NEXT: csel r0, r0, r7, ne
; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: csel r0, r0, r6, ne
; CHECK-NEXT: csel r1, r1, r9, ne ; CHECK-NEXT: csel r1, r1, r9, ne
; CHECK-NEXT: csel r3, r3, r5, ne ; CHECK-NEXT: csel r3, r3, r6, ne
; CHECK-NEXT: csel r2, r2, r5, ne ; CHECK-NEXT: csel r2, r2, r6, ne
; CHECK-NEXT: rsbs r5, r0, #0 ; CHECK-NEXT: rsbs r6, r0, #0
; CHECK-NEXT: sbcs.w r5, r10, r1 ; CHECK-NEXT: sbcs.w r6, r10, r1
; CHECK-NEXT: sbcs.w r2, r6, r2 ; CHECK-NEXT: sbcs.w r2, r7, r2
; CHECK-NEXT: sbcs.w r2, r6, r3 ; CHECK-NEXT: sbcs.w r2, r7, r3
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r2, lt
; CHECK-NEXT: movlt r4, #1 ; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: cmp r4, #0
; CHECK-NEXT: csel r1, r1, r10, ne ; CHECK-NEXT: csel r1, r1, r10, ne
; CHECK-NEXT: cmp r7, #0 ; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: csel r2, r11, r7, ne ; CHECK-NEXT: csel r3, r4, r5, ne
; CHECK-NEXT: cmp r4, #0 ; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: csel r0, r0, r4, ne ; CHECK-NEXT: csel r0, r0, r2, ne
; CHECK-NEXT: vmov q0[2], q0[0], r0, r2 ; CHECK-NEXT: vmov q0[2], q0[0], r0, r3
; CHECK-NEXT: vmov q0[3], q0[1], r1, r8 ; CHECK-NEXT: vmov q0[3], q0[1], r1, r8
; CHECK-NEXT: vpop {d8, d9} ; CHECK-NEXT: vpop {d8, d9}
; CHECK-NEXT: add sp, #4 ; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, r10, pc}
; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, r10, r11, pc}
entry: entry:
%conv = fptosi <2 x half> %x to <2 x i128> %conv = fptosi <2 x half> %x to <2 x i128>
%0 = icmp slt <2 x i128> %conv, <i128 9223372036854775807, i128 9223372036854775807> %0 = icmp slt <2 x i128> %conv, <i128 9223372036854775807, i128 9223372036854775807>
%spec.store.select = select <2 x i1> %0, <2 x i128> %conv, <2 x i128> <i128 9223372036854775807, i128 9223372036854775807> %spec.store.select = select <2 x i1> %0, <2 x i128> %conv, <2 x i128> <i128 9223372036854775807, i128 9223372036854775807>
%1 = icmp sgt <2 x i128> %spec.store.select, <i128 -9223372036854775808, i128 -9223372036854775808> %1 = icmp sgt <2 x i128> %spec.store.select, <i128 -9223372036854775808, i128 -9223372036854775808>
%spec.store.select7 = select <2 x i1> %1, <2 x i128> %spec.store.select, <2 x i128> <i128 -9223372036854775808, i128 -9223372036854775808> %spec.store.select7 = select <2 x i1> %1, <2 x i128> %spec.store.select, <2 x i128> <i128 -9223372036854775808, i128 -9223372036854775808>
%conv6 = trunc <2 x i128> %spec.store.select7 to <2 x i64> %conv6 = trunc <2 x i128> %spec.store.select7 to <2 x i64>
ret <2 x i64> %conv6 ret <2 x i64> %conv6
Show All 32 Lines
; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, r9, lr} ; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, r9, lr}
; CHECK-NEXT: .pad #4 ; CHECK-NEXT: .pad #4
; CHECK-NEXT: sub sp, #4 ; CHECK-NEXT: sub sp, #4
; CHECK-NEXT: .vsave {d8, d9} ; CHECK-NEXT: .vsave {d8, d9}
; CHECK-NEXT: vpush {d8, d9} ; CHECK-NEXT: vpush {d8, d9}
; CHECK-NEXT: vmov.u16 r0, q0[1] ; CHECK-NEXT: vmov.u16 r0, q0[1]
; CHECK-NEXT: vmov q4, q0 ; CHECK-NEXT: vmov q4, q0
; CHECK-NEXT: bl __fixhfti ; CHECK-NEXT: bl __fixhfti
; CHECK-NEXT: subs r7, r2, #1 ; CHECK-NEXT: subs r5, r2, #1
; CHECK-NEXT: mov.w r8, #1 ; CHECK-NEXT: mov.w r8, #1
; CHECK-NEXT: sbcs r7, r3, #0 ; CHECK-NEXT: sbcs r5, r3, #0
; CHECK-NEXT: mov.w r4, #0
; CHECK-NEXT: mov.w r7, #0 ; CHECK-NEXT: mov.w r7, #0
; CHECK-NEXT: mov.w r6, #0 ; CHECK-NEXT: cset r5, lt
; CHECK-NEXT: it lt ; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: movlt r7, #1 ; CHECK-NEXT: csel r0, r0, r5, ne
; CHECK-NEXT: cmp r7, #0 ; CHECK-NEXT: csel r3, r3, r5, ne
; CHECK-NEXT: csel r0, r0, r7, ne
; CHECK-NEXT: csel r3, r3, r7, ne
; CHECK-NEXT: csel r2, r2, r8, ne ; CHECK-NEXT: csel r2, r2, r8, ne
; CHECK-NEXT: csel r5, r1, r7, ne ; CHECK-NEXT: csel r4, r1, r5, ne
; CHECK-NEXT: rsbs r1, r0, #0 ; CHECK-NEXT: rsbs r1, r0, #0
; CHECK-NEXT: sbcs.w r1, r4, r5 ; CHECK-NEXT: sbcs.w r1, r7, r4
; CHECK-NEXT: sbcs.w r1, r4, r2 ; CHECK-NEXT: sbcs.w r1, r7, r2
; CHECK-NEXT: sbcs.w r1, r4, r3 ; CHECK-NEXT: sbcs.w r1, r7, r3
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r6, lt
; CHECK-NEXT: movlt r6, #1
; CHECK-NEXT: cmp r6, #0 ; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: csel r9, r0, r6, ne ; CHECK-NEXT: csel r9, r0, r6, ne
; CHECK-NEXT: vmov.u16 r0, q4[0] ; CHECK-NEXT: vmov.u16 r0, q4[0]
; CHECK-NEXT: bl __fixhfti ; CHECK-NEXT: bl __fixhfti
; CHECK-NEXT: subs r7, r2, #1 ; CHECK-NEXT: subs r5, r2, #1
; CHECK-NEXT: sbcs r7, r3, #0 ; CHECK-NEXT: sbcs r5, r3, #0
; CHECK-NEXT: mov.w r7, #0 ; CHECK-NEXT: cset r5, lt
; CHECK-NEXT: it lt ; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: movlt r7, #1 ; CHECK-NEXT: csel r0, r0, r5, ne
; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csel r0, r0, r7, ne
; CHECK-NEXT: csel r2, r2, r8, ne ; CHECK-NEXT: csel r2, r2, r8, ne
; CHECK-NEXT: csel r3, r3, r7, ne ; CHECK-NEXT: csel r3, r3, r5, ne
; CHECK-NEXT: csel r1, r1, r7, ne ; CHECK-NEXT: csel r1, r1, r5, ne
; CHECK-NEXT: rsbs r7, r0, #0 ; CHECK-NEXT: rsbs r5, r0, #0
; CHECK-NEXT: sbcs.w r7, r4, r1 ; CHECK-NEXT: sbcs.w r5, r7, r1
; CHECK-NEXT: sbcs.w r2, r4, r2 ; CHECK-NEXT: sbcs.w r2, r7, r2
; CHECK-NEXT: sbcs.w r2, r4, r3 ; CHECK-NEXT: sbcs.w r2, r7, r3
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r2, lt
; CHECK-NEXT: movlt r4, #1 ; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: cmp r4, #0 ; CHECK-NEXT: csel r0, r0, r2, ne
; CHECK-NEXT: csel r0, r0, r4, ne
; CHECK-NEXT: cmp r6, #0 ; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: csel r2, r5, r6, ne ; CHECK-NEXT: csel r3, r4, r6, ne
; CHECK-NEXT: cmp r4, #0 ; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: csel r1, r1, r4, ne ; CHECK-NEXT: csel r1, r1, r2, ne
; CHECK-NEXT: vmov q0[2], q0[0], r0, r9 ; CHECK-NEXT: vmov q0[2], q0[0], r0, r9
; CHECK-NEXT: vmov q0[3], q0[1], r1, r2 ; CHECK-NEXT: vmov q0[3], q0[1], r1, r3
; CHECK-NEXT: vpop {d8, d9} ; CHECK-NEXT: vpop {d8, d9}
; CHECK-NEXT: add sp, #4 ; CHECK-NEXT: add sp, #4
; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, pc} ; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, pc}
entry: entry:
%conv = fptosi <2 x half> %x to <2 x i128> %conv = fptosi <2 x half> %x to <2 x i128>
%0 = icmp slt <2 x i128> %conv, <i128 18446744073709551616, i128 18446744073709551616> %0 = icmp slt <2 x i128> %conv, <i128 18446744073709551616, i128 18446744073709551616>
%spec.store.select = select <2 x i1> %0, <2 x i128> %conv, <2 x i128> <i128 18446744073709551616, i128 18446744073709551616> %spec.store.select = select <2 x i1> %0, <2 x i128> %conv, <2 x i128> <i128 18446744073709551616, i128 18446744073709551616>
%1 = icmp sgt <2 x i128> %spec.store.select, zeroinitializer %1 = icmp sgt <2 x i128> %spec.store.select, zeroinitializer
▲ Show 20 Lines • Show All 1,089 Lines • ▼ Show 20 Linesentry:
%spec.store.select7 = call <2 x i128> @llvm.smax.v2i128(<2 x i128> %spec.store.select, <2 x i128> <i128 -9223372036854775808, i128 -9223372036854775808>) %spec.store.select7 = call <2 x i128> @llvm.smax.v2i128(<2 x i128> %spec.store.select, <2 x i128> <i128 -9223372036854775808, i128 -9223372036854775808>)
%conv6 = trunc <2 x i128> %spec.store.select7 to <2 x i64> %conv6 = trunc <2 x i128> %spec.store.select7 to <2 x i64>
ret <2 x i64> %conv6 ret <2 x i64> %conv6
} }
define arm_aapcs_vfpcc <2 x i64> @utest_f64i64_mm(<2 x double> %x) { define arm_aapcs_vfpcc <2 x i64> @utest_f64i64_mm(<2 x double> %x) {
; CHECK-LABEL: utest_f64i64_mm: ; CHECK-LABEL: utest_f64i64_mm:
; CHECK: @ %bb.0: @ %entry ; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: .save {r4, r5, r6, r7, r8, r9, lr} ; CHECK-NEXT: .save {r4, r5, r6, r7, r8, lr}
; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, r9, lr} ; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, lr}
; CHECK-NEXT: .pad #4
; CHECK-NEXT: sub sp, #4
; CHECK-NEXT: .vsave {d8, d9} ; CHECK-NEXT: .vsave {d8, d9}
; CHECK-NEXT: vpush {d8, d9} ; CHECK-NEXT: vpush {d8, d9}
; CHECK-NEXT: vmov q4, q0 ; CHECK-NEXT: vmov q4, q0
; CHECK-NEXT: vmov r0, r1, d9 ; CHECK-NEXT: vmov r0, r1, d9
; CHECK-NEXT: bl __fixunsdfti ; CHECK-NEXT: bl __fixunsdfti
; CHECK-NEXT: mov r8, r1 ; CHECK-NEXT: mov r8, r1
; CHECK-NEXT: vmov r4, r1, d8 ; CHECK-NEXT: vmov r4, r1, d8
; CHECK-NEXT: eor r7, r2, #1 ; CHECK-NEXT: eor r6, r2, #1
; CHECK-NEXT: subs r2, #1 ; CHECK-NEXT: subs r2, #1
; CHECK-NEXT: sbcs r2, r3, #0 ; CHECK-NEXT: sbcs r2, r3, #0
; CHECK-NEXT: mov.w r5, #0 ; CHECK-NEXT: orr.w r6, r6, r3
; CHECK-NEXT: it lo ; CHECK-NEXT: cset r7, lo
; CHECK-NEXT: movlo r5, #1
; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: orr.w r7, r7, r3
; CHECK-NEXT: csel r0, r0, r5, ne
; CHECK-NEXT: cmp r7, #0 ; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: mov.w r6, #0 ; CHECK-NEXT: csel r0, r0, r7, ne
; CHECK-NEXT: csel r9, r0, r7, ne ; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: csel r5, r0, r6, ne
; CHECK-NEXT: mov r0, r4 ; CHECK-NEXT: mov r0, r4
; CHECK-NEXT: bl __fixunsdfti ; CHECK-NEXT: bl __fixunsdfti
; CHECK-NEXT: eor r4, r2, #1 ; CHECK-NEXT: eor r4, r2, #1
; CHECK-NEXT: subs r2, #1 ; CHECK-NEXT: subs r2, #1
; CHECK-NEXT: sbcs r2, r3, #0 ; CHECK-NEXT: sbcs r2, r3, #0
; CHECK-NEXT: it lo
; CHECK-NEXT: movlo r6, #1
; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: orr.w r4, r4, r3 ; CHECK-NEXT: orr.w r4, r4, r3
; CHECK-NEXT: csel r0, r0, r6, ne ; CHECK-NEXT: cset r2, lo
; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: csel r0, r0, r2, ne
; CHECK-NEXT: cmp r4, #0 ; CHECK-NEXT: cmp r4, #0
; CHECK-NEXT: csel r0, r0, r4, ne ; CHECK-NEXT: csel r0, r0, r4, ne
; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: csel r2, r8, r5, ne
; CHECK-NEXT: cmp r7, #0 ; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csel r2, r2, r7, ne ; CHECK-NEXT: csel r3, r8, r7, ne
; CHECK-NEXT: cmp r6, #0 ; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: csel r1, r1, r6, ne ; CHECK-NEXT: csel r3, r3, r6, ne
; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: csel r1, r1, r2, ne
; CHECK-NEXT: cmp r4, #0 ; CHECK-NEXT: cmp r4, #0
; CHECK-NEXT: csel r1, r1, r4, ne ; CHECK-NEXT: csel r1, r1, r4, ne
; CHECK-NEXT: vmov q0[2], q0[0], r0, r9 ; CHECK-NEXT: vmov q0[2], q0[0], r0, r5
; CHECK-NEXT: vmov q0[3], q0[1], r1, r2 ; CHECK-NEXT: vmov q0[3], q0[1], r1, r3
; CHECK-NEXT: vpop {d8, d9} ; CHECK-NEXT: vpop {d8, d9}
; CHECK-NEXT: add sp, #4 ; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, pc}
; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, pc}
entry: entry:
%conv = fptoui <2 x double> %x to <2 x i128> %conv = fptoui <2 x double> %x to <2 x i128>
%spec.store.select = call <2 x i128> @llvm.umin.v2i128(<2 x i128> %conv, <2 x i128> <i128 18446744073709551616, i128 18446744073709551616>) %spec.store.select = call <2 x i128> @llvm.umin.v2i128(<2 x i128> %conv, <2 x i128> <i128 18446744073709551616, i128 18446744073709551616>)
%conv6 = trunc <2 x i128> %spec.store.select to <2 x i64> %conv6 = trunc <2 x i128> %spec.store.select to <2 x i64>
ret <2 x i64> %conv6 ret <2 x i64> %conv6
} }
define arm_aapcs_vfpcc <2 x i64> @ustest_f64i64_mm(<2 x double> %x) { define arm_aapcs_vfpcc <2 x i64> @ustest_f64i64_mm(<2 x double> %x) {
; CHECK-LABEL: ustest_f64i64_mm: ; CHECK-LABEL: ustest_f64i64_mm:
; CHECK: @ %bb.0: @ %entry ; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: .save {r4, r5, r6, r7, r8, r9, r10, r11, lr} ; CHECK-NEXT: .save {r4, r5, r6, r7, r8, r9, r10, r11, lr}
; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, r9, r10, r11, lr} ; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, r9, r10, r11, lr}
; CHECK-NEXT: .pad #4 ; CHECK-NEXT: .pad #4
; CHECK-NEXT: sub sp, #4 ; CHECK-NEXT: sub sp, #4
; CHECK-NEXT: .vsave {d8, d9} ; CHECK-NEXT: .vsave {d8, d9}
; CHECK-NEXT: vpush {d8, d9} ; CHECK-NEXT: vpush {d8, d9}
; CHECK-NEXT: vmov q4, q0 ; CHECK-NEXT: vmov q4, q0
; CHECK-NEXT: vmov r0, r1, d9 ; CHECK-NEXT: vmov r0, r1, d9
; CHECK-NEXT: bl __fixdfti ; CHECK-NEXT: bl __fixdfti
; CHECK-NEXT: subs r6, r2, #1 ; CHECK-NEXT: subs r6, r2, #1
; CHECK-NEXT: eor r7, r2, #1 ; CHECK-NEXT: eor r7, r2, #1
; CHECK-NEXT: sbcs r6, r3, #0 ; CHECK-NEXT: sbcs r6, r3, #0
; CHECK-NEXT: orr.w r7, r7, r3 ; CHECK-NEXT: orr.w r7, r7, r3
; CHECK-NEXT: mov.w r6, #0 ; CHECK-NEXT: cset r6, lt
; CHECK-NEXT: mov.w r10, #1 ; CHECK-NEXT: mov.w r10, #1
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r6, #1
; CHECK-NEXT: cmp r6, #0 ; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: mov.w r11, #0
; CHECK-NEXT: csel r0, r0, r6, ne ; CHECK-NEXT: csel r0, r0, r6, ne
; CHECK-NEXT: cmp r7, #0 ; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csel r0, r0, r7, ne ; CHECK-NEXT: csel r0, r0, r7, ne
; CHECK-NEXT: cmp r6, #0 ; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: csel r1, r1, r6, ne ; CHECK-NEXT: csel r1, r1, r6, ne
; CHECK-NEXT: cmp r7, #0 ; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csel r11, r1, r7, ne ; CHECK-NEXT: csel r4, r1, r7, ne
; CHECK-NEXT: movs r4, #0 ; CHECK-NEXT: cmp r4, #0
; CHECK-NEXT: cmp.w r11, #0 ; CHECK-NEXT: csel r1, r0, r4, ne
; CHECK-NEXT: mov.w r7, #0 ; CHECK-NEXT: csel r7, r0, r1, eq
; CHECK-NEXT: csel r1, r0, r11, ne
; CHECK-NEXT: csel r6, r0, r1, eq
; CHECK-NEXT: cmp r2, #1 ; CHECK-NEXT: cmp r2, #1
; CHECK-NEXT: csel r1, r2, r10, lo ; CHECK-NEXT: csel r1, r2, r10, lo
; CHECK-NEXT: cmp r3, #0 ; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: csel r2, r2, r10, mi ; CHECK-NEXT: csel r2, r2, r10, mi
; CHECK-NEXT: csel r1, r1, r2, eq ; CHECK-NEXT: csel r1, r1, r2, eq
; CHECK-NEXT: csel r2, r3, r4, mi ; CHECK-NEXT: csel r2, r3, r11, mi
; CHECK-NEXT: rsbs r3, r1, #0 ; CHECK-NEXT: rsbs r3, r1, #0
; CHECK-NEXT: sbcs.w r3, r4, r2 ; CHECK-NEXT: sbcs.w r3, r11, r2
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r6, lt
; CHECK-NEXT: movlt r7, #1 ; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: cmp r7, #0 ; CHECK-NEXT: csel r3, r0, r6, ne
; CHECK-NEXT: csel r3, r0, r7, ne
; CHECK-NEXT: orrs.w r9, r1, r2 ; CHECK-NEXT: orrs.w r9, r1, r2
; CHECK-NEXT: vmov r0, r1, d8 ; CHECK-NEXT: vmov r0, r1, d8
; CHECK-NEXT: csel r8, r6, r3, eq ; CHECK-NEXT: csel r8, r7, r3, eq
; CHECK-NEXT: bl __fixdfti ; CHECK-NEXT: bl __fixdfti
; CHECK-NEXT: subs r5, r2, #1 ; CHECK-NEXT: subs r5, r2, #1
; CHECK-NEXT: eor r6, r2, #1 ; CHECK-NEXT: eor r7, r2, #1
; CHECK-NEXT: sbcs r5, r3, #0 ; CHECK-NEXT: sbcs r5, r3, #0
; CHECK-NEXT: orr.w r6, r6, r3 ; CHECK-NEXT: orr.w r7, r7, r3
; CHECK-NEXT: mov.w r5, #0 ; CHECK-NEXT: cset r5, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r5, #1
; CHECK-NEXT: cmp r5, #0 ; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: csel r0, r0, r5, ne ; CHECK-NEXT: csel r0, r0, r5, ne
; CHECK-NEXT: cmp r6, #0 ; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csel r0, r0, r6, ne ; CHECK-NEXT: csel r0, r0, r7, ne
; CHECK-NEXT: cmp r5, #0 ; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: csel r1, r1, r5, ne ; CHECK-NEXT: csel r1, r1, r5, ne
; CHECK-NEXT: cmp r6, #0 ; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csel r1, r1, r6, ne ; CHECK-NEXT: csel r1, r1, r7, ne
; CHECK-NEXT: cmp r1, #0 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: csel r6, r0, r1, ne ; CHECK-NEXT: csel r7, r0, r1, ne
; CHECK-NEXT: csel r6, r0, r6, eq ; CHECK-NEXT: csel r7, r0, r7, eq
; CHECK-NEXT: cmp r2, #1 ; CHECK-NEXT: cmp r2, #1
; CHECK-NEXT: csel r5, r2, r10, lo ; CHECK-NEXT: csel r5, r2, r10, lo
; CHECK-NEXT: cmp r3, #0 ; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: csel r2, r2, r10, mi ; CHECK-NEXT: csel r2, r2, r10, mi
; CHECK-NEXT: csel r3, r3, r4, mi ; CHECK-NEXT: csel r3, r3, r11, mi
; CHECK-NEXT: csel r2, r5, r2, eq ; CHECK-NEXT: csel r2, r5, r2, eq
; CHECK-NEXT: rsbs r5, r2, #0 ; CHECK-NEXT: rsbs r5, r2, #0
; CHECK-NEXT: sbcs.w r5, r4, r3 ; CHECK-NEXT: sbcs.w r5, r11, r3
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r5, lt
; CHECK-NEXT: movlt r4, #1 ; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: cmp r4, #0 ; CHECK-NEXT: csel r0, r0, r5, ne
; CHECK-NEXT: csel r0, r0, r4, ne
; CHECK-NEXT: orrs r2, r3 ; CHECK-NEXT: orrs r2, r3
; CHECK-NEXT: csel r0, r6, r0, eq ; CHECK-NEXT: csel r0, r7, r0, eq
; CHECK-NEXT: cmp r7, #0 ; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: csel r3, r11, r7, ne ; CHECK-NEXT: csel r3, r4, r6, ne
; CHECK-NEXT: cmp.w r9, #0 ; CHECK-NEXT: cmp.w r9, #0
; CHECK-NEXT: csel r3, r11, r3, eq ; CHECK-NEXT: csel r3, r4, r3, eq
; CHECK-NEXT: cmp r4, #0 ; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: csel r7, r1, r4, ne ; CHECK-NEXT: csel r7, r1, r5, ne
; CHECK-NEXT: cmp r2, #0 ; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: csel r1, r1, r7, eq ; CHECK-NEXT: csel r1, r1, r7, eq
; CHECK-NEXT: vmov q0[2], q0[0], r0, r8 ; CHECK-NEXT: vmov q0[2], q0[0], r0, r8
; CHECK-NEXT: vmov q0[3], q0[1], r1, r3 ; CHECK-NEXT: vmov q0[3], q0[1], r1, r3
; CHECK-NEXT: vpop {d8, d9} ; CHECK-NEXT: vpop {d8, d9}
; CHECK-NEXT: add sp, #4 ; CHECK-NEXT: add sp, #4
; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, r10, r11, pc} ; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, r10, r11, pc}
entry: entry:
▲ Show 20 Lines • Show All 116 Lines • ▼ Show 20 Linesentry:
%spec.store.select7 = call <2 x i128> @llvm.smax.v2i128(<2 x i128> %spec.store.select, <2 x i128> <i128 -9223372036854775808, i128 -9223372036854775808>) %spec.store.select7 = call <2 x i128> @llvm.smax.v2i128(<2 x i128> %spec.store.select, <2 x i128> <i128 -9223372036854775808, i128 -9223372036854775808>)
%conv6 = trunc <2 x i128> %spec.store.select7 to <2 x i64> %conv6 = trunc <2 x i128> %spec.store.select7 to <2 x i64>
ret <2 x i64> %conv6 ret <2 x i64> %conv6
} }
define arm_aapcs_vfpcc <2 x i64> @utest_f32i64_mm(<2 x float> %x) { define arm_aapcs_vfpcc <2 x i64> @utest_f32i64_mm(<2 x float> %x) {
; CHECK-LABEL: utest_f32i64_mm: ; CHECK-LABEL: utest_f32i64_mm:
; CHECK: @ %bb.0: @ %entry ; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: .save {r4, r5, r6, r7, r8, r9, lr} ; CHECK-NEXT: .save {r4, r5, r6, r7, r8, lr}
; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, r9, lr} ; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, lr}
; CHECK-NEXT: .pad #4
; CHECK-NEXT: sub sp, #4
; CHECK-NEXT: vmov r5, r0, d0 ; CHECK-NEXT: vmov r5, r0, d0
; CHECK-NEXT: bl __fixunssfti ; CHECK-NEXT: bl __fixunssfti
; CHECK-NEXT: mov r8, r1 ; CHECK-NEXT: mov r8, r1
; CHECK-NEXT: eor r1, r2, #1 ; CHECK-NEXT: eor r1, r2, #1
; CHECK-NEXT: orr.w r7, r1, r3 ; CHECK-NEXT: orr.w r6, r1, r3
; CHECK-NEXT: subs r1, r2, #1 ; CHECK-NEXT: subs r1, r2, #1
; CHECK-NEXT: sbcs r1, r3, #0 ; CHECK-NEXT: sbcs r1, r3, #0
; CHECK-NEXT: mov.w r4, #0 ; CHECK-NEXT: cset r7, lo
; CHECK-NEXT: it lo
; CHECK-NEXT: movlo r4, #1
; CHECK-NEXT: cmp r4, #0
; CHECK-NEXT: csel r0, r0, r4, ne
; CHECK-NEXT: cmp r7, #0 ; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csel r9, r0, r7, ne ; CHECK-NEXT: csel r0, r0, r7, ne
; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: csel r4, r0, r6, ne
; CHECK-NEXT: mov r0, r5 ; CHECK-NEXT: mov r0, r5
; CHECK-NEXT: movs r6, #0
; CHECK-NEXT: bl __fixunssfti ; CHECK-NEXT: bl __fixunssfti
; CHECK-NEXT: eor r5, r2, #1 ; CHECK-NEXT: eor r5, r2, #1
; CHECK-NEXT: subs r2, #1 ; CHECK-NEXT: subs r2, #1
; CHECK-NEXT: sbcs r2, r3, #0 ; CHECK-NEXT: sbcs r2, r3, #0
; CHECK-NEXT: it lo
; CHECK-NEXT: movlo r6, #1
; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: orr.w r5, r5, r3 ; CHECK-NEXT: orr.w r5, r5, r3
; CHECK-NEXT: csel r0, r0, r6, ne ; CHECK-NEXT: cset r2, lo
; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: csel r0, r0, r2, ne
; CHECK-NEXT: cmp r5, #0 ; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: csel r0, r0, r5, ne ; CHECK-NEXT: csel r0, r0, r5, ne
; CHECK-NEXT: cmp r4, #0
; CHECK-NEXT: csel r2, r8, r4, ne
; CHECK-NEXT: cmp r7, #0 ; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csel r2, r2, r7, ne ; CHECK-NEXT: csel r3, r8, r7, ne
; CHECK-NEXT: cmp r6, #0 ; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: csel r1, r1, r6, ne ; CHECK-NEXT: csel r3, r3, r6, ne
; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: csel r1, r1, r2, ne
; CHECK-NEXT: cmp r5, #0 ; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: csel r1, r1, r5, ne ; CHECK-NEXT: csel r1, r1, r5, ne
; CHECK-NEXT: vmov q0[2], q0[0], r0, r9 ; CHECK-NEXT: vmov q0[2], q0[0], r0, r4
; CHECK-NEXT: vmov q0[3], q0[1], r1, r2 ; CHECK-NEXT: vmov q0[3], q0[1], r1, r3
; CHECK-NEXT: add sp, #4 ; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, pc}
; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, pc}
entry: entry:
%conv = fptoui <2 x float> %x to <2 x i128> %conv = fptoui <2 x float> %x to <2 x i128>
%spec.store.select = call <2 x i128> @llvm.umin.v2i128(<2 x i128> %conv, <2 x i128> <i128 18446744073709551616, i128 18446744073709551616>) %spec.store.select = call <2 x i128> @llvm.umin.v2i128(<2 x i128> %conv, <2 x i128> <i128 18446744073709551616, i128 18446744073709551616>)
%conv6 = trunc <2 x i128> %spec.store.select to <2 x i64> %conv6 = trunc <2 x i128> %spec.store.select to <2 x i64>
ret <2 x i64> %conv6 ret <2 x i64> %conv6
} }
define arm_aapcs_vfpcc <2 x i64> @ustest_f32i64_mm(<2 x float> %x) { define arm_aapcs_vfpcc <2 x i64> @ustest_f32i64_mm(<2 x float> %x) {
; CHECK-LABEL: ustest_f32i64_mm: ; CHECK-LABEL: ustest_f32i64_mm:
; CHECK: @ %bb.0: @ %entry ; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: .save {r4, r5, r6, r7, r8, r9, r10, r11, lr} ; CHECK-NEXT: .save {r4, r5, r6, r7, r8, r9, r10, r11, lr}
; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, r9, r10, r11, lr} ; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, r9, r10, r11, lr}
; CHECK-NEXT: .pad #4 ; CHECK-NEXT: .pad #4
; CHECK-NEXT: sub sp, #4 ; CHECK-NEXT: sub sp, #4
; CHECK-NEXT: vmov r9, r0, d0 ; CHECK-NEXT: vmov r10, r0, d0
; CHECK-NEXT: bl __fixsfti ; CHECK-NEXT: bl __fixsfti
; CHECK-NEXT: subs r6, r2, #1 ; CHECK-NEXT: subs r6, r2, #1
; CHECK-NEXT: eor r7, r2, #1 ; CHECK-NEXT: eor r7, r2, #1
; CHECK-NEXT: sbcs r6, r3, #0 ; CHECK-NEXT: sbcs r6, r3, #0
; CHECK-NEXT: orr.w r7, r7, r3 ; CHECK-NEXT: orr.w r7, r7, r3
; CHECK-NEXT: mov.w r6, #0 ; CHECK-NEXT: cset r6, lt
; CHECK-NEXT: mov.w r11, #1 ; CHECK-NEXT: mov.w r11, #1
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r6, #1
; CHECK-NEXT: cmp r6, #0 ; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: mov.w r8, #0
; CHECK-NEXT: csel r0, r0, r6, ne ; CHECK-NEXT: csel r0, r0, r6, ne
; CHECK-NEXT: cmp r7, #0 ; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csel r0, r0, r7, ne ; CHECK-NEXT: csel r0, r0, r7, ne
; CHECK-NEXT: cmp r6, #0 ; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: csel r1, r1, r6, ne ; CHECK-NEXT: csel r1, r1, r6, ne
; CHECK-NEXT: cmp r7, #0 ; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csel r8, r1, r7, ne ; CHECK-NEXT: csel r5, r1, r7, ne
; CHECK-NEXT: movs r5, #0 ; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: cmp.w r8, #0 ; CHECK-NEXT: csel r1, r0, r5, ne
; CHECK-NEXT: csel r1, r0, r8, ne
; CHECK-NEXT: csel r1, r0, r1, eq ; CHECK-NEXT: csel r1, r0, r1, eq
; CHECK-NEXT: cmp r2, #1 ; CHECK-NEXT: cmp r2, #1
; CHECK-NEXT: csel r7, r2, r11, lo ; CHECK-NEXT: csel r7, r2, r11, lo
; CHECK-NEXT: cmp r3, #0 ; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: csel r2, r2, r11, mi ; CHECK-NEXT: csel r2, r2, r11, mi
; CHECK-NEXT: csel r3, r3, r5, mi ; CHECK-NEXT: csel r3, r3, r8, mi
; CHECK-NEXT: csel r2, r7, r2, eq ; CHECK-NEXT: csel r2, r7, r2, eq
; CHECK-NEXT: rsbs r7, r2, #0 ; CHECK-NEXT: rsbs r7, r2, #0
; CHECK-NEXT: sbcs.w r7, r5, r3 ; CHECK-NEXT: sbcs.w r7, r8, r3
; CHECK-NEXT: mov.w r7, #0 ; CHECK-NEXT: cset r7, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r7, #1
; CHECK-NEXT: cmp r7, #0 ; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csel r0, r0, r7, ne ; CHECK-NEXT: csel r0, r0, r7, ne
; CHECK-NEXT: orrs.w r10, r2, r3 ; CHECK-NEXT: orrs.w r9, r2, r3
; CHECK-NEXT: csel r0, r1, r0, eq ; CHECK-NEXT: csel r0, r1, r0, eq
; CHECK-NEXT: str r0, [sp] @ 4-byte Spill ; CHECK-NEXT: str r0, [sp] @ 4-byte Spill
; CHECK-NEXT: mov r0, r9 ; CHECK-NEXT: mov r0, r10
; CHECK-NEXT: bl __fixsfti ; CHECK-NEXT: bl __fixsfti
; CHECK-NEXT: subs r6, r2, #1 ; CHECK-NEXT: subs r4, r2, #1
; CHECK-NEXT: eor r4, r2, #1 ; CHECK-NEXT: eor r6, r2, #1
; CHECK-NEXT: sbcs r6, r3, #0 ; CHECK-NEXT: sbcs r4, r3, #0
; CHECK-NEXT: orr.w r4, r4, r3 ; CHECK-NEXT: orr.w r6, r6, r3
; CHECK-NEXT: mov.w r6, #0 ; CHECK-NEXT: cset r4, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r6, #1
; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: csel r0, r0, r6, ne
; CHECK-NEXT: cmp r4, #0 ; CHECK-NEXT: cmp r4, #0
; CHECK-NEXT: csel r0, r0, r4, ne ; CHECK-NEXT: csel r0, r0, r4, ne
; CHECK-NEXT: cmp r6, #0 ; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: csel r1, r1, r6, ne ; CHECK-NEXT: csel r0, r0, r6, ne
; CHECK-NEXT: cmp r4, #0 ; CHECK-NEXT: cmp r4, #0
; CHECK-NEXT: csel r1, r1, r4, ne ; CHECK-NEXT: csel r1, r1, r4, ne
; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: csel r1, r1, r6, ne
; CHECK-NEXT: cmp r1, #0 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: csel r6, r0, r1, ne ; CHECK-NEXT: csel r6, r0, r1, ne
; CHECK-NEXT: csel r6, r0, r6, eq ; CHECK-NEXT: csel r6, r0, r6, eq
; CHECK-NEXT: cmp r2, #1 ; CHECK-NEXT: cmp r2, #1
; CHECK-NEXT: csel r4, r2, r11, lo ; CHECK-NEXT: csel r4, r2, r11, lo
; CHECK-NEXT: cmp r3, #0 ; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: csel r2, r2, r11, mi ; CHECK-NEXT: csel r2, r2, r11, mi
; CHECK-NEXT: csel r3, r3, r5, mi ; CHECK-NEXT: csel r3, r3, r8, mi
; CHECK-NEXT: csel r2, r4, r2, eq ; CHECK-NEXT: csel r2, r4, r2, eq
; CHECK-NEXT: rsbs r4, r2, #0 ; CHECK-NEXT: rsbs r4, r2, #0
; CHECK-NEXT: sbcs.w r4, r5, r3 ; CHECK-NEXT: sbcs.w r4, r8, r3
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r4, lt
; CHECK-NEXT: movlt r5, #1 ; CHECK-NEXT: cmp r4, #0
; CHECK-NEXT: cmp r5, #0 ; CHECK-NEXT: csel r0, r0, r4, ne
; CHECK-NEXT: csel r0, r0, r5, ne
; CHECK-NEXT: orrs r2, r3 ; CHECK-NEXT: orrs r2, r3
; CHECK-NEXT: csel r0, r6, r0, eq ; CHECK-NEXT: csel r0, r6, r0, eq
; CHECK-NEXT: cmp r7, #0 ; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csel r3, r8, r7, ne ; CHECK-NEXT: csel r3, r5, r7, ne
; CHECK-NEXT: cmp.w r10, #0 ; CHECK-NEXT: cmp.w r9, #0
; CHECK-NEXT: csel r3, r8, r3, eq ; CHECK-NEXT: csel r3, r5, r3, eq
; CHECK-NEXT: cmp r5, #0 ; CHECK-NEXT: cmp r4, #0
; CHECK-NEXT: csel r7, r1, r5, ne ; CHECK-NEXT: csel r7, r1, r4, ne
; CHECK-NEXT: cmp r2, #0 ; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: ldr r2, [sp] @ 4-byte Reload ; CHECK-NEXT: ldr r2, [sp] @ 4-byte Reload
; CHECK-NEXT: csel r1, r1, r7, eq ; CHECK-NEXT: csel r1, r1, r7, eq
; CHECK-NEXT: vmov q0[2], q0[0], r0, r2 ; CHECK-NEXT: vmov q0[2], q0[0], r0, r2
; CHECK-NEXT: vmov q0[3], q0[1], r1, r3 ; CHECK-NEXT: vmov q0[3], q0[1], r1, r3
; CHECK-NEXT: add sp, #4 ; CHECK-NEXT: add sp, #4
; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, r10, r11, pc} ; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, r10, r11, pc}
entry: entry:
▲ Show 20 Lines • Show All 161 Lines • ▼ Show 20 Lines
; CHECK-NEXT: vpush {d8, d9} ; CHECK-NEXT: vpush {d8, d9}
; CHECK-NEXT: vmov.u16 r0, q0[1] ; CHECK-NEXT: vmov.u16 r0, q0[1]
; CHECK-NEXT: vmov q4, q0 ; CHECK-NEXT: vmov q4, q0
; CHECK-NEXT: bl __fixhfti ; CHECK-NEXT: bl __fixhfti
; CHECK-NEXT: subs r6, r2, #1 ; CHECK-NEXT: subs r6, r2, #1
; CHECK-NEXT: eor r7, r2, #1 ; CHECK-NEXT: eor r7, r2, #1
; CHECK-NEXT: sbcs r6, r3, #0 ; CHECK-NEXT: sbcs r6, r3, #0
; CHECK-NEXT: orr.w r7, r7, r3 ; CHECK-NEXT: orr.w r7, r7, r3
; CHECK-NEXT: mov.w r6, #0 ; CHECK-NEXT: cset r6, lt
; CHECK-NEXT: mov.w r10, #1 ; CHECK-NEXT: mov.w r10, #1
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r6, #1
; CHECK-NEXT: cmp r6, #0 ; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: mov.w r11, #0
; CHECK-NEXT: csel r0, r0, r6, ne ; CHECK-NEXT: csel r0, r0, r6, ne
; CHECK-NEXT: cmp r7, #0 ; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csel r0, r0, r7, ne ; CHECK-NEXT: csel r0, r0, r7, ne
; CHECK-NEXT: cmp r6, #0 ; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: csel r1, r1, r6, ne ; CHECK-NEXT: csel r1, r1, r6, ne
; CHECK-NEXT: cmp r7, #0 ; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csel r11, r1, r7, ne ; CHECK-NEXT: csel r4, r1, r7, ne
; CHECK-NEXT: movs r4, #0 ; CHECK-NEXT: cmp r4, #0
; CHECK-NEXT: cmp.w r11, #0 ; CHECK-NEXT: csel r1, r0, r4, ne
; CHECK-NEXT: csel r1, r0, r11, ne
; CHECK-NEXT: csel r1, r0, r1, eq ; CHECK-NEXT: csel r1, r0, r1, eq
; CHECK-NEXT: cmp r2, #1 ; CHECK-NEXT: cmp r2, #1
; CHECK-NEXT: csel r7, r2, r10, lo ; CHECK-NEXT: csel r6, r2, r10, lo
; CHECK-NEXT: cmp r3, #0 ; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: csel r2, r2, r10, mi ; CHECK-NEXT: csel r2, r2, r10, mi
; CHECK-NEXT: csel r3, r3, r4, mi ; CHECK-NEXT: csel r3, r3, r11, mi
; CHECK-NEXT: csel r2, r7, r2, eq ; CHECK-NEXT: csel r2, r6, r2, eq
; CHECK-NEXT: rsbs r7, r2, #0 ; CHECK-NEXT: rsbs r6, r2, #0
; CHECK-NEXT: sbcs.w r7, r4, r3 ; CHECK-NEXT: sbcs.w r6, r11, r3
; CHECK-NEXT: mov.w r7, #0 ; CHECK-NEXT: cset r6, lt
; CHECK-NEXT: it lt ; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: movlt r7, #1 ; CHECK-NEXT: csel r0, r0, r6, ne
; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csel r0, r0, r7, ne
; CHECK-NEXT: orrs.w r9, r2, r3 ; CHECK-NEXT: orrs.w r9, r2, r3
; CHECK-NEXT: csel r8, r1, r0, eq ; CHECK-NEXT: csel r8, r1, r0, eq
; CHECK-NEXT: vmov.u16 r0, q4[0] ; CHECK-NEXT: vmov.u16 r0, q4[0]
; CHECK-NEXT: bl __fixhfti ; CHECK-NEXT: bl __fixhfti
; CHECK-NEXT: subs r5, r2, #1 ; CHECK-NEXT: subs r5, r2, #1
; CHECK-NEXT: eor r6, r2, #1 ; CHECK-NEXT: eor r7, r2, #1
; CHECK-NEXT: sbcs r5, r3, #0 ; CHECK-NEXT: sbcs r5, r3, #0
; CHECK-NEXT: orr.w r6, r6, r3 ; CHECK-NEXT: orr.w r7, r7, r3
; CHECK-NEXT: mov.w r5, #0 ; CHECK-NEXT: cset r5, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r5, #1
; CHECK-NEXT: cmp r5, #0 ; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: csel r0, r0, r5, ne ; CHECK-NEXT: csel r0, r0, r5, ne
; CHECK-NEXT: cmp r6, #0 ; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csel r0, r0, r6, ne ; CHECK-NEXT: csel r0, r0, r7, ne
; CHECK-NEXT: cmp r5, #0 ; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: csel r1, r1, r5, ne ; CHECK-NEXT: csel r1, r1, r5, ne
; CHECK-NEXT: cmp r6, #0 ; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csel r1, r1, r6, ne ; CHECK-NEXT: csel r1, r1, r7, ne
; CHECK-NEXT: cmp r1, #0 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: csel r6, r0, r1, ne ; CHECK-NEXT: csel r7, r0, r1, ne
; CHECK-NEXT: csel r6, r0, r6, eq ; CHECK-NEXT: csel r7, r0, r7, eq
; CHECK-NEXT: cmp r2, #1 ; CHECK-NEXT: cmp r2, #1
; CHECK-NEXT: csel r5, r2, r10, lo ; CHECK-NEXT: csel r5, r2, r10, lo
; CHECK-NEXT: cmp r3, #0 ; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: csel r2, r2, r10, mi ; CHECK-NEXT: csel r2, r2, r10, mi
; CHECK-NEXT: csel r3, r3, r4, mi ; CHECK-NEXT: csel r3, r3, r11, mi
; CHECK-NEXT: csel r2, r5, r2, eq ; CHECK-NEXT: csel r2, r5, r2, eq
; CHECK-NEXT: rsbs r5, r2, #0 ; CHECK-NEXT: rsbs r5, r2, #0
; CHECK-NEXT: sbcs.w r5, r4, r3 ; CHECK-NEXT: sbcs.w r5, r11, r3
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r5, lt
; CHECK-NEXT: movlt r4, #1 ; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: cmp r4, #0 ; CHECK-NEXT: csel r0, r0, r5, ne
; CHECK-NEXT: csel r0, r0, r4, ne
; CHECK-NEXT: orrs r2, r3 ; CHECK-NEXT: orrs r2, r3
; CHECK-NEXT: csel r0, r6, r0, eq ; CHECK-NEXT: csel r0, r7, r0, eq
; CHECK-NEXT: cmp r7, #0 ; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: csel r3, r11, r7, ne ; CHECK-NEXT: csel r3, r4, r6, ne
; CHECK-NEXT: cmp.w r9, #0 ; CHECK-NEXT: cmp.w r9, #0
; CHECK-NEXT: csel r3, r11, r3, eq ; CHECK-NEXT: csel r3, r4, r3, eq
; CHECK-NEXT: cmp r4, #0 ; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: csel r7, r1, r4, ne ; CHECK-NEXT: csel r7, r1, r5, ne
; CHECK-NEXT: cmp r2, #0 ; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: csel r1, r1, r7, eq ; CHECK-NEXT: csel r1, r1, r7, eq
; CHECK-NEXT: vmov q0[2], q0[0], r0, r8 ; CHECK-NEXT: vmov q0[2], q0[0], r0, r8
; CHECK-NEXT: vmov q0[3], q0[1], r1, r3 ; CHECK-NEXT: vmov q0[3], q0[1], r1, r3
; CHECK-NEXT: vpop {d8, d9} ; CHECK-NEXT: vpop {d8, d9}
; CHECK-NEXT: add sp, #4 ; CHECK-NEXT: add sp, #4
; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, r10, r11, pc} ; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, r10, r11, pc}
entry: entry:
Show All 25 Lines

llvm/test/CodeGen/Thumb2/mve-masked-ldst.ll

Show First 20 Lines • Show All 87 Lines • ▼ Show 20 Lines
define void @foo_sext_v2i64_v2i32(<2 x i64> *%dest, <2 x i32> *%mask, <2 x i32> *%src) { define void @foo_sext_v2i64_v2i32(<2 x i64> *%dest, <2 x i32> *%mask, <2 x i32> *%src) {
; CHECK-LE-LABEL: foo_sext_v2i64_v2i32: ; CHECK-LE-LABEL: foo_sext_v2i64_v2i32:
; CHECK-LE: @ %bb.0: @ %entry ; CHECK-LE: @ %bb.0: @ %entry
; CHECK-LE-NEXT: .save {r4, r5, r7, lr} ; CHECK-LE-NEXT: .save {r4, r5, r7, lr}
; CHECK-LE-NEXT: push {r4, r5, r7, lr} ; CHECK-LE-NEXT: push {r4, r5, r7, lr}
; CHECK-LE-NEXT: .pad #4 ; CHECK-LE-NEXT: .pad #4
; CHECK-LE-NEXT: sub sp, #4 ; CHECK-LE-NEXT: sub sp, #4
; CHECK-LE-NEXT: ldrd lr, r5, [r1] ; CHECK-LE-NEXT: ldrd lr, r1, [r1]
; CHECK-LE-NEXT: movs r3, #0 ; CHECK-LE-NEXT: rsbs r3, r1, #0
; CHECK-LE-NEXT: @ implicit-def: $q0 ; CHECK-LE-NEXT: mov.w r12, #0
; CHECK-LE-NEXT: rsbs.w r1, lr, #0 ; CHECK-LE-NEXT: sbcs.w r3, r12, r1, asr #31
; CHECK-LE-NEXT: vmov q1[2], q1[0], lr, r5 ; CHECK-LE-NEXT: cset r3, lt
; CHECK-LE-NEXT: sbcs.w r1, r3, lr, asr #31 ; CHECK-LE-NEXT: @ implicit-def: $q1
; CHECK-LE-NEXT: mov.w r1, #0 ; CHECK-LE-NEXT: vmov q0[2], q0[0], lr, r1
; CHECK-LE-NEXT: it lt
; CHECK-LE-NEXT: movlt r1, #1
; CHECK-LE-NEXT: rsbs r4, r5, #0
; CHECK-LE-NEXT: sbcs.w r4, r3, r5, asr #31
; CHECK-LE-NEXT: it lt
; CHECK-LE-NEXT: movlt r3, #1
; CHECK-LE-NEXT: cmp r3, #0 ; CHECK-LE-NEXT: cmp r3, #0
; CHECK-LE-NEXT: it ne ; CHECK-LE-NEXT: it ne
; CHECK-LE-NEXT: mvnne r3, #1 ; CHECK-LE-NEXT: mvnne r3, #1
; CHECK-LE-NEXT: bfi r3, r1, #0, #1 ; CHECK-LE-NEXT: rsbs.w r4, lr, #0
; CHECK-LE-NEXT: vmov r4, s4 ; CHECK-LE-NEXT: sbcs.w r4, r12, lr, asr #31
; CHECK-LE-NEXT: mov.w lr, #0
; CHECK-LE-NEXT: cset r4, lt
; CHECK-LE-NEXT: bfi r3, r4, #0, #1
; CHECK-LE-NEXT: and r12, r3, #3 ; CHECK-LE-NEXT: and r12, r3, #3
; CHECK-LE-NEXT: lsls r1, r3, #31 ; CHECK-LE-NEXT: lsls r3, r3, #31
; CHECK-LE-NEXT: itt ne ; CHECK-LE-NEXT: itt ne
; CHECK-LE-NEXT: ldrne r1, [r2] ; CHECK-LE-NEXT: ldrne r3, [r2]
; CHECK-LE-NEXT: vmovne.32 q0[0], r1 ; CHECK-LE-NEXT: vmovne.32 q1[0], r3
; CHECK-LE-NEXT: lsls.w r1, r12, #30 ; CHECK-LE-NEXT: lsls.w r1, r12, #30
; CHECK-LE-NEXT: itt mi ; CHECK-LE-NEXT: itt mi
; CHECK-LE-NEXT: ldrmi r1, [r2, #4] ; CHECK-LE-NEXT: ldrmi r1, [r2, #4]
; CHECK-LE-NEXT: vmovmi.32 q0[2], r1 ; CHECK-LE-NEXT: vmovmi.32 q1[2], r1
; CHECK-LE-NEXT: vmov r3, s0 ; CHECK-LE-NEXT: vmov r1, s6
; CHECK-LE-NEXT: movs r2, #0 ; CHECK-LE-NEXT: vmov r2, s2
; CHECK-LE-NEXT: vmov r1, s2 ; CHECK-LE-NEXT: vmov r3, s4
; CHECK-LE-NEXT: vmov q0[2], q0[0], r3, r1 ; CHECK-LE-NEXT: vmov q1[2], q1[0], r3, r1
; CHECK-LE-NEXT: rsbs r5, r4, #0 ; CHECK-LE-NEXT: rsbs r5, r2, #0
; CHECK-LE-NEXT: asr.w lr, r3, #31
; CHECK-LE-NEXT: vmov r3, s6
; CHECK-LE-NEXT: asr.w r12, r1, #31 ; CHECK-LE-NEXT: asr.w r12, r1, #31
; CHECK-LE-NEXT: sbcs.w r1, r2, r4, asr #31 ; CHECK-LE-NEXT: sbcs.w r1, lr, r2, asr #31
; CHECK-LE-NEXT: mov.w r1, #0 ; CHECK-LE-NEXT: vmov r1, s0
; CHECK-LE-NEXT: vmov q0[3], q0[1], lr, r12 ; CHECK-LE-NEXT: cset r2, lt
; CHECK-LE-NEXT: it lt
; CHECK-LE-NEXT: movlt r1, #1
; CHECK-LE-NEXT: rsbs r5, r3, #0
; CHECK-LE-NEXT: sbcs.w r3, r2, r3, asr #31
; CHECK-LE-NEXT: it lt
; CHECK-LE-NEXT: movlt r2, #1
; CHECK-LE-NEXT: cmp r2, #0 ; CHECK-LE-NEXT: cmp r2, #0
; CHECK-LE-NEXT: asr.w r4, r3, #31
; CHECK-LE-NEXT: it ne ; CHECK-LE-NEXT: it ne
; CHECK-LE-NEXT: mvnne r2, #1 ; CHECK-LE-NEXT: mvnne r2, #1
; CHECK-LE-NEXT: vmov q1[3], q1[1], r4, r12
; CHECK-LE-NEXT: rsbs r3, r1, #0
; CHECK-LE-NEXT: sbcs.w r1, lr, r1, asr #31
; CHECK-LE-NEXT: cset r1, lt
; CHECK-LE-NEXT: bfi r2, r1, #0, #1 ; CHECK-LE-NEXT: bfi r2, r1, #0, #1
; CHECK-LE-NEXT: and r1, r2, #3 ; CHECK-LE-NEXT: and r1, r2, #3
; CHECK-LE-NEXT: lsls r2, r2, #31 ; CHECK-LE-NEXT: lsls r2, r2, #31
; CHECK-LE-NEXT: it ne ; CHECK-LE-NEXT: it ne
; CHECK-LE-NEXT: vstrne d0, [r0] ; CHECK-LE-NEXT: vstrne d2, [r0]
; CHECK-LE-NEXT: lsls r1, r1, #30 ; CHECK-LE-NEXT: lsls r1, r1, #30
; CHECK-LE-NEXT: it mi ; CHECK-LE-NEXT: it mi
; CHECK-LE-NEXT: vstrmi d1, [r0, #8] ; CHECK-LE-NEXT: vstrmi d3, [r0, #8]
; CHECK-LE-NEXT: add sp, #4 ; CHECK-LE-NEXT: add sp, #4
; CHECK-LE-NEXT: pop {r4, r5, r7, pc} ; CHECK-LE-NEXT: pop {r4, r5, r7, pc}
; ;
; CHECK-BE-LABEL: foo_sext_v2i64_v2i32: ; CHECK-BE-LABEL: foo_sext_v2i64_v2i32:
; CHECK-BE: @ %bb.0: @ %entry ; CHECK-BE: @ %bb.0: @ %entry
; CHECK-BE-NEXT: .save {r4, r5, r7, lr} ; CHECK-BE-NEXT: .save {r4, r5, r7, lr}
; CHECK-BE-NEXT: push {r4, r5, r7, lr} ; CHECK-BE-NEXT: push {r4, r5, r7, lr}
; CHECK-BE-NEXT: .pad #4 ; CHECK-BE-NEXT: .pad #4
; CHECK-BE-NEXT: sub sp, #4 ; CHECK-BE-NEXT: sub sp, #4
; CHECK-BE-NEXT: ldrd r12, lr, [r1] ; CHECK-BE-NEXT: ldrd r3, lr, [r1]
; CHECK-BE-NEXT: rsbs.w r1, lr, #0 ; CHECK-BE-NEXT: mov.w r12, #0
; CHECK-BE-NEXT: mov.w r3, #0 ; CHECK-BE-NEXT: @ implicit-def: $q2
; CHECK-BE-NEXT: sbcs.w r1, r3, lr, asr #31 ; CHECK-BE-NEXT: rsbs r1, r3, #0
; CHECK-BE-NEXT: vmov q0[3], q0[1], r12, lr ; CHECK-BE-NEXT: vmov q0[3], q0[1], r3, lr
; CHECK-BE-NEXT: mov.w lr, #0 ; CHECK-BE-NEXT: sbcs.w r1, r12, r3, asr #31
; CHECK-BE-NEXT: it lt ; CHECK-BE-NEXT: cset r3, lt
; CHECK-BE-NEXT: movlt.w lr, #1
; CHECK-BE-NEXT: rsbs.w r1, r12, #0
; CHECK-BE-NEXT: sbcs.w r1, r3, r12, asr #31
; CHECK-BE-NEXT: it lt
; CHECK-BE-NEXT: movlt r3, #1
; CHECK-BE-NEXT: cmp r3, #0 ; CHECK-BE-NEXT: cmp r3, #0
; CHECK-BE-NEXT: it ne ; CHECK-BE-NEXT: it ne
; CHECK-BE-NEXT: mvnne r3, #1 ; CHECK-BE-NEXT: mvnne r3, #1
; CHECK-BE-NEXT: bfi r3, lr, #0, #1 ; CHECK-BE-NEXT: rsbs.w r1, lr, #0
; CHECK-BE-NEXT: @ implicit-def: $q2 ; CHECK-BE-NEXT: sbcs.w r1, r12, lr, asr #31
; CHECK-BE-NEXT: cset r1, lt
; CHECK-BE-NEXT: bfi r3, r1, #0, #1
; CHECK-BE-NEXT: and r1, r3, #3 ; CHECK-BE-NEXT: and r1, r3, #3
; CHECK-BE-NEXT: lsls r3, r3, #30 ; CHECK-BE-NEXT: lsls r3, r3, #30
; CHECK-BE-NEXT: bpl .LBB5_2 ; CHECK-BE-NEXT: bpl .LBB5_2
; CHECK-BE-NEXT: @ %bb.1: @ %cond.load ; CHECK-BE-NEXT: @ %bb.1: @ %cond.load
; CHECK-BE-NEXT: ldr r3, [r2] ; CHECK-BE-NEXT: ldr r3, [r2]
; CHECK-BE-NEXT: vmov.32 q1[1], r3 ; CHECK-BE-NEXT: vmov.32 q1[1], r3
; CHECK-BE-NEXT: vrev64.32 q2, q1 ; CHECK-BE-NEXT: vrev64.32 q2, q1
; CHECK-BE-NEXT: .LBB5_2: @ %else ; CHECK-BE-NEXT: .LBB5_2: @ %else
; CHECK-BE-NEXT: vrev64.32 q1, q0 ; CHECK-BE-NEXT: vrev64.32 q1, q0
; CHECK-BE-NEXT: lsls r1, r1, #31 ; CHECK-BE-NEXT: lsls r1, r1, #31
; CHECK-BE-NEXT: beq .LBB5_4 ; CHECK-BE-NEXT: beq .LBB5_4
; CHECK-BE-NEXT: @ %bb.3: @ %cond.load1 ; CHECK-BE-NEXT: @ %bb.3: @ %cond.load1
; CHECK-BE-NEXT: ldr r1, [r2, #4] ; CHECK-BE-NEXT: ldr r1, [r2, #4]
; CHECK-BE-NEXT: vrev64.32 q0, q2 ; CHECK-BE-NEXT: vrev64.32 q0, q2
; CHECK-BE-NEXT: vmov.32 q0[3], r1 ; CHECK-BE-NEXT: vmov.32 q0[3], r1
; CHECK-BE-NEXT: vrev64.32 q2, q0 ; CHECK-BE-NEXT: vrev64.32 q2, q0
; CHECK-BE-NEXT: .LBB5_4: @ %else2 ; CHECK-BE-NEXT: .LBB5_4: @ %else2
; CHECK-BE-NEXT: vrev64.32 q0, q2 ; CHECK-BE-NEXT: vrev64.32 q0, q2
; CHECK-BE-NEXT: vrev64.32 q2, q1 ; CHECK-BE-NEXT: vrev64.32 q2, q1
; CHECK-BE-NEXT: vmov r2, s11
; CHECK-BE-NEXT: movs r4, #0
; CHECK-BE-NEXT: vmov r1, s3 ; CHECK-BE-NEXT: vmov r1, s3
; CHECK-BE-NEXT: movs r4, #0
; CHECK-BE-NEXT: vmov r3, s1 ; CHECK-BE-NEXT: vmov r3, s1
; CHECK-BE-NEXT: rsbs r5, r2, #0
; CHECK-BE-NEXT: sbcs.w r2, r4, r2, asr #31
; CHECK-BE-NEXT: vmov r2, s9 ; CHECK-BE-NEXT: vmov r2, s9
; CHECK-BE-NEXT: asr.w r12, r1, #31 ; CHECK-BE-NEXT: asr.w r12, r1, #31
; CHECK-BE-NEXT: asr.w lr, r3, #31 ; CHECK-BE-NEXT: asr.w lr, r3, #31
; CHECK-BE-NEXT: rsbs r5, r2, #0
; CHECK-BE-NEXT: vmov q1[2], q1[0], lr, r12 ; CHECK-BE-NEXT: vmov q1[2], q1[0], lr, r12
; CHECK-BE-NEXT: sbcs.w r2, r4, r2, asr #31
; CHECK-BE-NEXT: vmov q1[3], q1[1], r3, r1 ; CHECK-BE-NEXT: vmov q1[3], q1[1], r3, r1
; CHECK-BE-NEXT: mov.w r1, #0 ; CHECK-BE-NEXT: vmov r1, s11
; CHECK-BE-NEXT: it lt ; CHECK-BE-NEXT: cset r2, lt
; CHECK-BE-NEXT: movlt r1, #1
; CHECK-BE-NEXT: vrev64.32 q0, q1 ; CHECK-BE-NEXT: vrev64.32 q0, q1
; CHECK-BE-NEXT: rsbs r3, r2, #0 ; CHECK-BE-NEXT: cmp r2, #0
; CHECK-BE-NEXT: sbcs.w r2, r4, r2, asr #31
; CHECK-BE-NEXT: it lt
; CHECK-BE-NEXT: movlt r4, #1
; CHECK-BE-NEXT: cmp r4, #0
; CHECK-BE-NEXT: it ne ; CHECK-BE-NEXT: it ne
; CHECK-BE-NEXT: mvnne r4, #1 ; CHECK-BE-NEXT: mvnne r2, #1
; CHECK-BE-NEXT: bfi r4, r1, #0, #1 ; CHECK-BE-NEXT: rsbs r3, r1, #0
; CHECK-BE-NEXT: and r1, r4, #3 ; CHECK-BE-NEXT: sbcs.w r1, r4, r1, asr #31
; CHECK-BE-NEXT: lsls r2, r4, #30 ; CHECK-BE-NEXT: cset r1, lt
; CHECK-BE-NEXT: bfi r2, r1, #0, #1
; CHECK-BE-NEXT: and r1, r2, #3
; CHECK-BE-NEXT: lsls r2, r2, #30
; CHECK-BE-NEXT: it mi ; CHECK-BE-NEXT: it mi
; CHECK-BE-NEXT: vstrmi d0, [r0] ; CHECK-BE-NEXT: vstrmi d0, [r0]
; CHECK-BE-NEXT: lsls r1, r1, #31 ; CHECK-BE-NEXT: lsls r1, r1, #31
; CHECK-BE-NEXT: it ne ; CHECK-BE-NEXT: it ne
; CHECK-BE-NEXT: vstrne d1, [r0, #8] ; CHECK-BE-NEXT: vstrne d1, [r0, #8]
; CHECK-BE-NEXT: add sp, #4 ; CHECK-BE-NEXT: add sp, #4
; CHECK-BE-NEXT: pop {r4, r5, r7, pc} ; CHECK-BE-NEXT: pop {r4, r5, r7, pc}
entry: entry:
%0 = load <2 x i32>, <2 x i32>* %mask, align 4 %0 = load <2 x i32>, <2 x i32>* %mask, align 4
%1 = icmp sgt <2 x i32> %0, zeroinitializer %1 = icmp sgt <2 x i32> %0, zeroinitializer
%2 = call <2 x i32> @llvm.masked.load.v2i32.p0v2i32(<2 x i32>* %src, i32 4, <2 x i1> %1, <2 x i32> undef) %2 = call <2 x i32> @llvm.masked.load.v2i32.p0v2i32(<2 x i32>* %src, i32 4, <2 x i1> %1, <2 x i32> undef)
%3 = sext <2 x i32> %2 to <2 x i64> %3 = sext <2 x i32> %2 to <2 x i64>
call void @llvm.masked.store.v2i64.p0v2i64(<2 x i64> %3, <2 x i64>* %dest, i32 8, <2 x i1> %1) call void @llvm.masked.store.v2i64.p0v2i64(<2 x i64> %3, <2 x i64>* %dest, i32 8, <2 x i1> %1)
ret void ret void
} }
define void @foo_sext_v2i64_v2i32_unaligned(<2 x i64> *%dest, <2 x i32> *%mask, <2 x i32> *%src) { define void @foo_sext_v2i64_v2i32_unaligned(<2 x i64> *%dest, <2 x i32> *%mask, <2 x i32> *%src) {
; CHECK-LE-LABEL: foo_sext_v2i64_v2i32_unaligned: ; CHECK-LE-LABEL: foo_sext_v2i64_v2i32_unaligned:
; CHECK-LE: @ %bb.0: @ %entry ; CHECK-LE: @ %bb.0: @ %entry
; CHECK-LE-NEXT: .save {r4, r5, r7, lr} ; CHECK-LE-NEXT: .save {r4, r5, r7, lr}
; CHECK-LE-NEXT: push {r4, r5, r7, lr} ; CHECK-LE-NEXT: push {r4, r5, r7, lr}
; CHECK-LE-NEXT: .pad #4 ; CHECK-LE-NEXT: .pad #4
; CHECK-LE-NEXT: sub sp, #4 ; CHECK-LE-NEXT: sub sp, #4
; CHECK-LE-NEXT: ldrd lr, r5, [r1] ; CHECK-LE-NEXT: ldrd lr, r1, [r1]
; CHECK-LE-NEXT: movs r3, #0 ; CHECK-LE-NEXT: rsbs r3, r1, #0
; CHECK-LE-NEXT: mov.w r12, #0
; CHECK-LE-NEXT: sbcs.w r3, r12, r1, asr #31
; CHECK-LE-NEXT: cset r3, lt
; CHECK-LE-NEXT: @ implicit-def: $q0 ; CHECK-LE-NEXT: @ implicit-def: $q0
; CHECK-LE-NEXT: rsbs.w r1, lr, #0 ; CHECK-LE-NEXT: vmov q1[2], q1[0], lr, r1
; CHECK-LE-NEXT: vmov q1[2], q1[0], lr, r5
; CHECK-LE-NEXT: sbcs.w r1, r3, lr, asr #31
; CHECK-LE-NEXT: mov.w r1, #0
; CHECK-LE-NEXT: it lt
; CHECK-LE-NEXT: movlt r1, #1
; CHECK-LE-NEXT: rsbs r4, r5, #0
; CHECK-LE-NEXT: sbcs.w r4, r3, r5, asr #31
; CHECK-LE-NEXT: it lt
; CHECK-LE-NEXT: movlt r3, #1
; CHECK-LE-NEXT: cmp r3, #0 ; CHECK-LE-NEXT: cmp r3, #0
; CHECK-LE-NEXT: it ne ; CHECK-LE-NEXT: it ne
; CHECK-LE-NEXT: mvnne r3, #1 ; CHECK-LE-NEXT: mvnne r3, #1
; CHECK-LE-NEXT: bfi r3, r1, #0, #1 ; CHECK-LE-NEXT: rsbs.w r4, lr, #0
; CHECK-LE-NEXT: vmov r4, s4 ; CHECK-LE-NEXT: sbcs.w r4, r12, lr, asr #31
; CHECK-LE-NEXT: mov.w lr, #0
; CHECK-LE-NEXT: cset r4, lt
; CHECK-LE-NEXT: bfi r3, r4, #0, #1
; CHECK-LE-NEXT: and r12, r3, #3 ; CHECK-LE-NEXT: and r12, r3, #3
; CHECK-LE-NEXT: lsls r1, r3, #31 ; CHECK-LE-NEXT: lsls r3, r3, #31
; CHECK-LE-NEXT: itt ne ; CHECK-LE-NEXT: itt ne
; CHECK-LE-NEXT: ldrne r1, [r2] ; CHECK-LE-NEXT: ldrne r3, [r2]
; CHECK-LE-NEXT: vmovne.32 q0[0], r1 ; CHECK-LE-NEXT: vmovne.32 q0[0], r3
; CHECK-LE-NEXT: lsls.w r1, r12, #30 ; CHECK-LE-NEXT: lsls.w r1, r12, #30
; CHECK-LE-NEXT: itt mi ; CHECK-LE-NEXT: itt mi
; CHECK-LE-NEXT: ldrmi r1, [r2, #4] ; CHECK-LE-NEXT: ldrmi r1, [r2, #4]
; CHECK-LE-NEXT: vmovmi.32 q0[2], r1 ; CHECK-LE-NEXT: vmovmi.32 q0[2], r1
; CHECK-LE-NEXT: vmov r3, s0
; CHECK-LE-NEXT: movs r2, #0
; CHECK-LE-NEXT: vmov r1, s2 ; CHECK-LE-NEXT: vmov r1, s2
; CHECK-LE-NEXT: vmov r2, s6
; CHECK-LE-NEXT: vmov r3, s0
; CHECK-LE-NEXT: vmov q0[2], q0[0], r3, r1 ; CHECK-LE-NEXT: vmov q0[2], q0[0], r3, r1
; CHECK-LE-NEXT: rsbs r5, r4, #0 ; CHECK-LE-NEXT: rsbs r5, r2, #0
; CHECK-LE-NEXT: asr.w lr, r3, #31
; CHECK-LE-NEXT: vmov r3, s6
; CHECK-LE-NEXT: asr.w r12, r1, #31 ; CHECK-LE-NEXT: asr.w r12, r1, #31
; CHECK-LE-NEXT: sbcs.w r1, r2, r4, asr #31 ; CHECK-LE-NEXT: sbcs.w r1, lr, r2, asr #31
; CHECK-LE-NEXT: mov.w r1, #0 ; CHECK-LE-NEXT: vmov r1, s4
; CHECK-LE-NEXT: vmov q0[3], q0[1], lr, r12 ; CHECK-LE-NEXT: cset r2, lt
; CHECK-LE-NEXT: it lt
; CHECK-LE-NEXT: movlt r1, #1
; CHECK-LE-NEXT: rsbs r5, r3, #0
; CHECK-LE-NEXT: sbcs.w r3, r2, r3, asr #31
; CHECK-LE-NEXT: it lt
; CHECK-LE-NEXT: movlt r2, #1
; CHECK-LE-NEXT: cmp r2, #0 ; CHECK-LE-NEXT: cmp r2, #0
; CHECK-LE-NEXT: asr.w r4, r3, #31
; CHECK-LE-NEXT: it ne ; CHECK-LE-NEXT: it ne
; CHECK-LE-NEXT: mvnne r2, #1 ; CHECK-LE-NEXT: mvnne r2, #1
; CHECK-LE-NEXT: vmov q0[3], q0[1], r4, r12
; CHECK-LE-NEXT: rsbs r3, r1, #0
; CHECK-LE-NEXT: sbcs.w r1, lr, r1, asr #31
; CHECK-LE-NEXT: cset r1, lt
; CHECK-LE-NEXT: bfi r2, r1, #0, #1 ; CHECK-LE-NEXT: bfi r2, r1, #0, #1
; CHECK-LE-NEXT: and r1, r2, #3 ; CHECK-LE-NEXT: and r1, r2, #3
; CHECK-LE-NEXT: lsls r2, r2, #31 ; CHECK-LE-NEXT: lsls r2, r2, #31
; CHECK-LE-NEXT: itt ne ; CHECK-LE-NEXT: itt ne
; CHECK-LE-NEXT: vmovne r2, r3, d0 ; CHECK-LE-NEXT: vmovne r2, r3, d0
; CHECK-LE-NEXT: strdne r2, r3, [r0] ; CHECK-LE-NEXT: strdne r2, r3, [r0]
; CHECK-LE-NEXT: lsls r1, r1, #30 ; CHECK-LE-NEXT: lsls r1, r1, #30
; CHECK-LE-NEXT: itt mi ; CHECK-LE-NEXT: itt mi
; CHECK-LE-NEXT: vmovmi r1, r2, d1 ; CHECK-LE-NEXT: vmovmi r1, r2, d1
; CHECK-LE-NEXT: strdmi r1, r2, [r0, #8] ; CHECK-LE-NEXT: strdmi r1, r2, [r0, #8]
; CHECK-LE-NEXT: add sp, #4 ; CHECK-LE-NEXT: add sp, #4
; CHECK-LE-NEXT: pop {r4, r5, r7, pc} ; CHECK-LE-NEXT: pop {r4, r5, r7, pc}
; ;
; CHECK-BE-LABEL: foo_sext_v2i64_v2i32_unaligned: ; CHECK-BE-LABEL: foo_sext_v2i64_v2i32_unaligned:
; CHECK-BE: @ %bb.0: @ %entry ; CHECK-BE: @ %bb.0: @ %entry
; CHECK-BE-NEXT: .save {r4, r5, r7, lr} ; CHECK-BE-NEXT: .save {r4, r5, r7, lr}
; CHECK-BE-NEXT: push {r4, r5, r7, lr} ; CHECK-BE-NEXT: push {r4, r5, r7, lr}
; CHECK-BE-NEXT: .pad #4 ; CHECK-BE-NEXT: .pad #4
; CHECK-BE-NEXT: sub sp, #4 ; CHECK-BE-NEXT: sub sp, #4
; CHECK-BE-NEXT: ldrd r12, lr, [r1] ; CHECK-BE-NEXT: ldrd r3, lr, [r1]
; CHECK-BE-NEXT: rsbs.w r1, lr, #0 ; CHECK-BE-NEXT: mov.w r12, #0
; CHECK-BE-NEXT: mov.w r3, #0 ; CHECK-BE-NEXT: @ implicit-def: $q2
; CHECK-BE-NEXT: sbcs.w r1, r3, lr, asr #31 ; CHECK-BE-NEXT: rsbs r1, r3, #0
; CHECK-BE-NEXT: vmov q0[3], q0[1], r12, lr ; CHECK-BE-NEXT: vmov q0[3], q0[1], r3, lr
; CHECK-BE-NEXT: mov.w lr, #0 ; CHECK-BE-NEXT: sbcs.w r1, r12, r3, asr #31
; CHECK-BE-NEXT: it lt ; CHECK-BE-NEXT: cset r3, lt
; CHECK-BE-NEXT: movlt.w lr, #1
; CHECK-BE-NEXT: rsbs.w r1, r12, #0
; CHECK-BE-NEXT: sbcs.w r1, r3, r12, asr #31
; CHECK-BE-NEXT: it lt
; CHECK-BE-NEXT: movlt r3, #1
; CHECK-BE-NEXT: cmp r3, #0 ; CHECK-BE-NEXT: cmp r3, #0
; CHECK-BE-NEXT: it ne ; CHECK-BE-NEXT: it ne
; CHECK-BE-NEXT: mvnne r3, #1 ; CHECK-BE-NEXT: mvnne r3, #1
; CHECK-BE-NEXT: bfi r3, lr, #0, #1 ; CHECK-BE-NEXT: rsbs.w r1, lr, #0
; CHECK-BE-NEXT: @ implicit-def: $q2 ; CHECK-BE-NEXT: sbcs.w r1, r12, lr, asr #31
; CHECK-BE-NEXT: cset r1, lt
; CHECK-BE-NEXT: bfi r3, r1, #0, #1
; CHECK-BE-NEXT: and r1, r3, #3 ; CHECK-BE-NEXT: and r1, r3, #3
; CHECK-BE-NEXT: lsls r3, r3, #30 ; CHECK-BE-NEXT: lsls r3, r3, #30
; CHECK-BE-NEXT: bpl .LBB6_2 ; CHECK-BE-NEXT: bpl .LBB6_2
; CHECK-BE-NEXT: @ %bb.1: @ %cond.load ; CHECK-BE-NEXT: @ %bb.1: @ %cond.load
; CHECK-BE-NEXT: ldr r3, [r2] ; CHECK-BE-NEXT: ldr r3, [r2]
; CHECK-BE-NEXT: vmov.32 q1[1], r3 ; CHECK-BE-NEXT: vmov.32 q1[1], r3
; CHECK-BE-NEXT: vrev64.32 q2, q1 ; CHECK-BE-NEXT: vrev64.32 q2, q1
; CHECK-BE-NEXT: .LBB6_2: @ %else ; CHECK-BE-NEXT: .LBB6_2: @ %else
; CHECK-BE-NEXT: vrev64.32 q1, q0 ; CHECK-BE-NEXT: vrev64.32 q1, q0
; CHECK-BE-NEXT: lsls r1, r1, #31 ; CHECK-BE-NEXT: lsls r1, r1, #31
; CHECK-BE-NEXT: beq .LBB6_4 ; CHECK-BE-NEXT: beq .LBB6_4
; CHECK-BE-NEXT: @ %bb.3: @ %cond.load1 ; CHECK-BE-NEXT: @ %bb.3: @ %cond.load1
; CHECK-BE-NEXT: ldr r1, [r2, #4] ; CHECK-BE-NEXT: ldr r1, [r2, #4]
; CHECK-BE-NEXT: vrev64.32 q0, q2 ; CHECK-BE-NEXT: vrev64.32 q0, q2
; CHECK-BE-NEXT: vmov.32 q0[3], r1 ; CHECK-BE-NEXT: vmov.32 q0[3], r1
; CHECK-BE-NEXT: vrev64.32 q2, q0 ; CHECK-BE-NEXT: vrev64.32 q2, q0
; CHECK-BE-NEXT: .LBB6_4: @ %else2 ; CHECK-BE-NEXT: .LBB6_4: @ %else2
; CHECK-BE-NEXT: vrev64.32 q0, q2 ; CHECK-BE-NEXT: vrev64.32 q0, q2
; CHECK-BE-NEXT: vrev64.32 q2, q1 ; CHECK-BE-NEXT: vrev64.32 q2, q1
; CHECK-BE-NEXT: vmov r2, s11
; CHECK-BE-NEXT: movs r4, #0
; CHECK-BE-NEXT: vmov r1, s3 ; CHECK-BE-NEXT: vmov r1, s3
; CHECK-BE-NEXT: movs r4, #0
; CHECK-BE-NEXT: vmov r3, s1 ; CHECK-BE-NEXT: vmov r3, s1
; CHECK-BE-NEXT: rsbs r5, r2, #0
; CHECK-BE-NEXT: sbcs.w r2, r4, r2, asr #31
; CHECK-BE-NEXT: vmov r2, s9 ; CHECK-BE-NEXT: vmov r2, s9
; CHECK-BE-NEXT: asr.w r12, r1, #31 ; CHECK-BE-NEXT: asr.w r12, r1, #31
; CHECK-BE-NEXT: asr.w lr, r3, #31 ; CHECK-BE-NEXT: asr.w lr, r3, #31
; CHECK-BE-NEXT: rsbs r5, r2, #0
; CHECK-BE-NEXT: vmov q1[2], q1[0], lr, r12 ; CHECK-BE-NEXT: vmov q1[2], q1[0], lr, r12
; CHECK-BE-NEXT: sbcs.w r2, r4, r2, asr #31
; CHECK-BE-NEXT: vmov q1[3], q1[1], r3, r1 ; CHECK-BE-NEXT: vmov q1[3], q1[1], r3, r1
; CHECK-BE-NEXT: mov.w r1, #0 ; CHECK-BE-NEXT: vmov r1, s11
; CHECK-BE-NEXT: it lt ; CHECK-BE-NEXT: cset r2, lt
; CHECK-BE-NEXT: movlt r1, #1
; CHECK-BE-NEXT: vrev64.32 q0, q1 ; CHECK-BE-NEXT: vrev64.32 q0, q1
; CHECK-BE-NEXT: rsbs r3, r2, #0 ; CHECK-BE-NEXT: cmp r2, #0
; CHECK-BE-NEXT: sbcs.w r2, r4, r2, asr #31
; CHECK-BE-NEXT: it lt
; CHECK-BE-NEXT: movlt r4, #1
; CHECK-BE-NEXT: cmp r4, #0
; CHECK-BE-NEXT: it ne ; CHECK-BE-NEXT: it ne
; CHECK-BE-NEXT: mvnne r4, #1 ; CHECK-BE-NEXT: mvnne r2, #1
; CHECK-BE-NEXT: bfi r4, r1, #0, #1 ; CHECK-BE-NEXT: rsbs r3, r1, #0
; CHECK-BE-NEXT: and r1, r4, #3 ; CHECK-BE-NEXT: sbcs.w r1, r4, r1, asr #31
; CHECK-BE-NEXT: lsls r2, r4, #30 ; CHECK-BE-NEXT: cset r1, lt
; CHECK-BE-NEXT: bfi r2, r1, #0, #1
; CHECK-BE-NEXT: and r1, r2, #3
; CHECK-BE-NEXT: lsls r2, r2, #30
; CHECK-BE-NEXT: itt mi ; CHECK-BE-NEXT: itt mi
; CHECK-BE-NEXT: vmovmi r2, r3, d0 ; CHECK-BE-NEXT: vmovmi r2, r3, d0
; CHECK-BE-NEXT: strdmi r3, r2, [r0] ; CHECK-BE-NEXT: strdmi r3, r2, [r0]
; CHECK-BE-NEXT: lsls r1, r1, #31 ; CHECK-BE-NEXT: lsls r1, r1, #31
; CHECK-BE-NEXT: itt ne ; CHECK-BE-NEXT: itt ne
; CHECK-BE-NEXT: vmovne r1, r2, d1 ; CHECK-BE-NEXT: vmovne r1, r2, d1
; CHECK-BE-NEXT: strdne r2, r1, [r0, #8] ; CHECK-BE-NEXT: strdne r2, r1, [r0, #8]
; CHECK-BE-NEXT: add sp, #4 ; CHECK-BE-NEXT: add sp, #4
; CHECK-BE-NEXT: pop {r4, r5, r7, pc} ; CHECK-BE-NEXT: pop {r4, r5, r7, pc}
entry: entry:
%0 = load <2 x i32>, <2 x i32>* %mask, align 4 %0 = load <2 x i32>, <2 x i32>* %mask, align 4
%1 = icmp sgt <2 x i32> %0, zeroinitializer %1 = icmp sgt <2 x i32> %0, zeroinitializer
%2 = call <2 x i32> @llvm.masked.load.v2i32.p0v2i32(<2 x i32>* %src, i32 2, <2 x i1> %1, <2 x i32> undef) %2 = call <2 x i32> @llvm.masked.load.v2i32.p0v2i32(<2 x i32>* %src, i32 2, <2 x i1> %1, <2 x i32> undef)
%3 = sext <2 x i32> %2 to <2 x i64> %3 = sext <2 x i32> %2 to <2 x i64>
call void @llvm.masked.store.v2i64.p0v2i64(<2 x i64> %3, <2 x i64>* %dest, i32 4, <2 x i1> %1) call void @llvm.masked.store.v2i64.p0v2i64(<2 x i64> %3, <2 x i64>* %dest, i32 4, <2 x i1> %1)
ret void ret void
} }
define void @foo_zext_v2i64_v2i32(<2 x i64> *%dest, <2 x i32> *%mask, <2 x i32> *%src) { define void @foo_zext_v2i64_v2i32(<2 x i64> *%dest, <2 x i32> *%mask, <2 x i32> *%src) {
; CHECK-LE-LABEL: foo_zext_v2i64_v2i32: ; CHECK-LE-LABEL: foo_zext_v2i64_v2i32:
; CHECK-LE: @ %bb.0: @ %entry ; CHECK-LE: @ %bb.0: @ %entry
; CHECK-LE-NEXT: .save {r4, r5, r7, lr} ; CHECK-LE-NEXT: .save {r4, lr}
; CHECK-LE-NEXT: push {r4, r5, r7, lr} ; CHECK-LE-NEXT: push {r4, lr}
; CHECK-LE-NEXT: .pad #4 ; CHECK-LE-NEXT: .pad #4
; CHECK-LE-NEXT: sub sp, #4 ; CHECK-LE-NEXT: sub sp, #4
; CHECK-LE-NEXT: ldrd lr, r5, [r1] ; CHECK-LE-NEXT: ldrd lr, r1, [r1]
; CHECK-LE-NEXT: movs r3, #0 ; CHECK-LE-NEXT: rsbs r3, r1, #0
; CHECK-LE-NEXT: mov.w r12, #0
; CHECK-LE-NEXT: sbcs.w r3, r12, r1, asr #31
; CHECK-LE-NEXT: cset r3, lt
; CHECK-LE-NEXT: @ implicit-def: $q0 ; CHECK-LE-NEXT: @ implicit-def: $q0
; CHECK-LE-NEXT: vmov.i64 q2, #0xffffffff ; CHECK-LE-NEXT: vmov q1[2], q1[0], lr, r1
; CHECK-LE-NEXT: rsbs.w r1, lr, #0
; CHECK-LE-NEXT: vmov q1[2], q1[0], lr, r5
; CHECK-LE-NEXT: sbcs.w r1, r3, lr, asr #31
; CHECK-LE-NEXT: mov.w r1, #0
; CHECK-LE-NEXT: it lt
; CHECK-LE-NEXT: movlt r1, #1
; CHECK-LE-NEXT: rsbs r4, r5, #0
; CHECK-LE-NEXT: sbcs.w r4, r3, r5, asr #31
; CHECK-LE-NEXT: it lt
; CHECK-LE-NEXT: movlt r3, #1
; CHECK-LE-NEXT: cmp r3, #0 ; CHECK-LE-NEXT: cmp r3, #0
; CHECK-LE-NEXT: it ne ; CHECK-LE-NEXT: it ne
; CHECK-LE-NEXT: mvnne r3, #1 ; CHECK-LE-NEXT: mvnne r3, #1
; CHECK-LE-NEXT: bfi r3, r1, #0, #1 ; CHECK-LE-NEXT: rsbs.w r4, lr, #0
; CHECK-LE-NEXT: vmov.i64 q2, #0xffffffff
; CHECK-LE-NEXT: sbcs.w r4, r12, lr, asr #31
; CHECK-LE-NEXT: cset r4, lt
; CHECK-LE-NEXT: bfi r3, r4, #0, #1
; CHECK-LE-NEXT: and r12, r3, #3 ; CHECK-LE-NEXT: and r12, r3, #3
; CHECK-LE-NEXT: lsls r1, r3, #31 ; CHECK-LE-NEXT: lsls r3, r3, #31
; CHECK-LE-NEXT: itt ne ; CHECK-LE-NEXT: itt ne
; CHECK-LE-NEXT: ldrne r1, [r2] ; CHECK-LE-NEXT: ldrne r3, [r2]
; CHECK-LE-NEXT: vmovne.32 q0[0], r1 ; CHECK-LE-NEXT: vmovne.32 q0[0], r3
; CHECK-LE-NEXT: lsls.w r1, r12, #30 ; CHECK-LE-NEXT: lsls.w r1, r12, #30
; CHECK-LE-NEXT: itt mi ; CHECK-LE-NEXT: itt mi
; CHECK-LE-NEXT: ldrmi r1, [r2, #4] ; CHECK-LE-NEXT: ldrmi r1, [r2, #4]
; CHECK-LE-NEXT: vmovmi.32 q0[2], r1 ; CHECK-LE-NEXT: vmovmi.32 q0[2], r1
; CHECK-LE-NEXT: vmov r1, s4 ; CHECK-LE-NEXT: vmov r2, s6
; CHECK-LE-NEXT: movs r2, #0 ; CHECK-LE-NEXT: movs r1, #0
; CHECK-LE-NEXT: vand q0, q0, q2 ; CHECK-LE-NEXT: vand q0, q0, q2
; CHECK-LE-NEXT: rsbs r3, r1, #0 ; CHECK-LE-NEXT: rsbs r3, r2, #0
; CHECK-LE-NEXT: vmov r3, s6 ; CHECK-LE-NEXT: vmov r3, s4
; CHECK-LE-NEXT: sbcs.w r1, r2, r1, asr #31 ; CHECK-LE-NEXT: sbcs.w r2, r1, r2, asr #31
; CHECK-LE-NEXT: mov.w r1, #0 ; CHECK-LE-NEXT: cset r2, lt
; CHECK-LE-NEXT: it lt
; CHECK-LE-NEXT: movlt r1, #1
; CHECK-LE-NEXT: rsbs r5, r3, #0
; CHECK-LE-NEXT: sbcs.w r3, r2, r3, asr #31
; CHECK-LE-NEXT: it lt
; CHECK-LE-NEXT: movlt r2, #1
; CHECK-LE-NEXT: cmp r2, #0 ; CHECK-LE-NEXT: cmp r2, #0
; CHECK-LE-NEXT: it ne ; CHECK-LE-NEXT: it ne
; CHECK-LE-NEXT: mvnne r2, #1 ; CHECK-LE-NEXT: mvnne r2, #1
; CHECK-LE-NEXT: rsbs r4, r3, #0
; CHECK-LE-NEXT: sbcs.w r1, r1, r3, asr #31
; CHECK-LE-NEXT: cset r1, lt
; CHECK-LE-NEXT: bfi r2, r1, #0, #1 ; CHECK-LE-NEXT: bfi r2, r1, #0, #1
; CHECK-LE-NEXT: and r1, r2, #3 ; CHECK-LE-NEXT: and r1, r2, #3
; CHECK-LE-NEXT: lsls r2, r2, #31 ; CHECK-LE-NEXT: lsls r2, r2, #31
; CHECK-LE-NEXT: it ne ; CHECK-LE-NEXT: it ne
; CHECK-LE-NEXT: vstrne d0, [r0] ; CHECK-LE-NEXT: vstrne d0, [r0]
; CHECK-LE-NEXT: lsls r1, r1, #30 ; CHECK-LE-NEXT: lsls r1, r1, #30
; CHECK-LE-NEXT: it mi ; CHECK-LE-NEXT: it mi
; CHECK-LE-NEXT: vstrmi d1, [r0, #8] ; CHECK-LE-NEXT: vstrmi d1, [r0, #8]
; CHECK-LE-NEXT: add sp, #4 ; CHECK-LE-NEXT: add sp, #4
; CHECK-LE-NEXT: pop {r4, r5, r7, pc} ; CHECK-LE-NEXT: pop {r4, pc}
; ;
; CHECK-BE-LABEL: foo_zext_v2i64_v2i32: ; CHECK-BE-LABEL: foo_zext_v2i64_v2i32:
; CHECK-BE: @ %bb.0: @ %entry ; CHECK-BE: @ %bb.0: @ %entry
; CHECK-BE-NEXT: .save {r7, lr} ; CHECK-BE-NEXT: .save {r7, lr}
; CHECK-BE-NEXT: push {r7, lr} ; CHECK-BE-NEXT: push {r7, lr}
; CHECK-BE-NEXT: .pad #4 ; CHECK-BE-NEXT: .pad #4
; CHECK-BE-NEXT: sub sp, #4 ; CHECK-BE-NEXT: sub sp, #4
; CHECK-BE-NEXT: ldrd r12, lr, [r1] ; CHECK-BE-NEXT: ldrd r3, lr, [r1]
; CHECK-BE-NEXT: rsbs.w r1, lr, #0 ; CHECK-BE-NEXT: mov.w r12, #0
; CHECK-BE-NEXT: mov.w r3, #0 ; CHECK-BE-NEXT: @ implicit-def: $q0
; CHECK-BE-NEXT: sbcs.w r1, r3, lr, asr #31 ; CHECK-BE-NEXT: rsbs r1, r3, #0
; CHECK-BE-NEXT: vmov q0[3], q0[1], r12, lr ; CHECK-BE-NEXT: vmov q1[3], q1[1], r3, lr
; CHECK-BE-NEXT: mov.w lr, #0 ; CHECK-BE-NEXT: sbcs.w r1, r12, r3, asr #31
; CHECK-BE-NEXT: it lt ; CHECK-BE-NEXT: cset r3, lt
; CHECK-BE-NEXT: movlt.w lr, #1
; CHECK-BE-NEXT: rsbs.w r1, r12, #0
; CHECK-BE-NEXT: sbcs.w r1, r3, r12, asr #31
; CHECK-BE-NEXT: it lt
; CHECK-BE-NEXT: movlt r3, #1
; CHECK-BE-NEXT: cmp r3, #0 ; CHECK-BE-NEXT: cmp r3, #0
; CHECK-BE-NEXT: it ne ; CHECK-BE-NEXT: it ne
; CHECK-BE-NEXT: mvnne r3, #1 ; CHECK-BE-NEXT: mvnne r3, #1
; CHECK-BE-NEXT: bfi r3, lr, #0, #1 ; CHECK-BE-NEXT: rsbs.w r1, lr, #0
; CHECK-BE-NEXT: @ implicit-def: $q1 ; CHECK-BE-NEXT: sbcs.w r1, r12, lr, asr #31
; CHECK-BE-NEXT: cset r1, lt
; CHECK-BE-NEXT: bfi r3, r1, #0, #1
; CHECK-BE-NEXT: and r1, r3, #3 ; CHECK-BE-NEXT: and r1, r3, #3
; CHECK-BE-NEXT: lsls r3, r3, #30 ; CHECK-BE-NEXT: lsls r3, r3, #30
; CHECK-BE-NEXT: bpl .LBB7_2 ; CHECK-BE-NEXT: bpl .LBB7_2
; CHECK-BE-NEXT: @ %bb.1: @ %cond.load ; CHECK-BE-NEXT: @ %bb.1: @ %cond.load
; CHECK-BE-NEXT: ldr r3, [r2] ; CHECK-BE-NEXT: ldr r3, [r2]
; CHECK-BE-NEXT: vmov.32 q2[1], r3 ; CHECK-BE-NEXT: vmov.32 q2[1], r3
; CHECK-BE-NEXT: vrev64.32 q1, q2 ; CHECK-BE-NEXT: vrev64.32 q0, q2
; CHECK-BE-NEXT: .LBB7_2: @ %else ; CHECK-BE-NEXT: .LBB7_2: @ %else
; CHECK-BE-NEXT: vrev64.32 q2, q0 ; CHECK-BE-NEXT: vrev64.32 q2, q1
; CHECK-BE-NEXT: lsls r1, r1, #31 ; CHECK-BE-NEXT: lsls r1, r1, #31
; CHECK-BE-NEXT: beq .LBB7_4 ; CHECK-BE-NEXT: beq .LBB7_4
; CHECK-BE-NEXT: @ %bb.3: @ %cond.load1 ; CHECK-BE-NEXT: @ %bb.3: @ %cond.load1
; CHECK-BE-NEXT: ldr r1, [r2, #4] ; CHECK-BE-NEXT: ldr r1, [r2, #4]
; CHECK-BE-NEXT: vrev64.32 q0, q1
; CHECK-BE-NEXT: vmov.32 q0[3], r1
; CHECK-BE-NEXT: vrev64.32 q1, q0 ; CHECK-BE-NEXT: vrev64.32 q1, q0
; CHECK-BE-NEXT: vmov.32 q1[3], r1
; CHECK-BE-NEXT: vrev64.32 q0, q1
; CHECK-BE-NEXT: .LBB7_4: @ %else2 ; CHECK-BE-NEXT: .LBB7_4: @ %else2
; CHECK-BE-NEXT: vrev64.32 q3, q2 ; CHECK-BE-NEXT: vrev64.32 q3, q2
; CHECK-BE-NEXT: movs r2, #0
; CHECK-BE-NEXT: vmov r1, s15
; CHECK-BE-NEXT: mov.w r12, #0 ; CHECK-BE-NEXT: mov.w r12, #0
; CHECK-BE-NEXT: vmov.i64 q0, #0xffffffff ; CHECK-BE-NEXT: vmov r2, s13
; CHECK-BE-NEXT: vand q0, q1, q0 ; CHECK-BE-NEXT: vmov.i64 q1, #0xffffffff
; CHECK-BE-NEXT: rsbs r3, r1, #0 ; CHECK-BE-NEXT: vand q0, q0, q1
; CHECK-BE-NEXT: vmov r3, s13 ; CHECK-BE-NEXT: rsbs r3, r2, #0
; CHECK-BE-NEXT: sbcs.w r1, r2, r1, asr #31 ; CHECK-BE-NEXT: vmov r3, s15
; CHECK-BE-NEXT: it lt ; CHECK-BE-NEXT: sbcs.w r2, r12, r2, asr #31
; CHECK-BE-NEXT: movlt.w r12, #1 ; CHECK-BE-NEXT: cset r2, lt
; CHECK-BE-NEXT: rsbs r1, r3, #0
; CHECK-BE-NEXT: sbcs.w r1, r2, r3, asr #31
; CHECK-BE-NEXT: it lt
; CHECK-BE-NEXT: movlt r2, #1
; CHECK-BE-NEXT: cmp r2, #0 ; CHECK-BE-NEXT: cmp r2, #0
; CHECK-BE-NEXT: it ne ; CHECK-BE-NEXT: it ne
; CHECK-BE-NEXT: mvnne r2, #1 ; CHECK-BE-NEXT: mvnne r2, #1
; CHECK-BE-NEXT: bfi r2, r12, #0, #1 ; CHECK-BE-NEXT: rsbs r1, r3, #0
; CHECK-BE-NEXT: sbcs.w r1, r12, r3, asr #31
; CHECK-BE-NEXT: cset r1, lt
; CHECK-BE-NEXT: bfi r2, r1, #0, #1
; CHECK-BE-NEXT: and r1, r2, #3 ; CHECK-BE-NEXT: and r1, r2, #3
; CHECK-BE-NEXT: lsls r2, r2, #30 ; CHECK-BE-NEXT: lsls r2, r2, #30
; CHECK-BE-NEXT: it mi ; CHECK-BE-NEXT: it mi
; CHECK-BE-NEXT: vstrmi d0, [r0] ; CHECK-BE-NEXT: vstrmi d0, [r0]
; CHECK-BE-NEXT: lsls r1, r1, #31 ; CHECK-BE-NEXT: lsls r1, r1, #31
; CHECK-BE-NEXT: it ne ; CHECK-BE-NEXT: it ne
; CHECK-BE-NEXT: vstrne d1, [r0, #8] ; CHECK-BE-NEXT: vstrne d1, [r0, #8]
; CHECK-BE-NEXT: add sp, #4 ; CHECK-BE-NEXT: add sp, #4
; CHECK-BE-NEXT: pop {r7, pc} ; CHECK-BE-NEXT: pop {r7, pc}
entry: entry:
%0 = load <2 x i32>, <2 x i32>* %mask, align 4 %0 = load <2 x i32>, <2 x i32>* %mask, align 4
%1 = icmp sgt <2 x i32> %0, zeroinitializer %1 = icmp sgt <2 x i32> %0, zeroinitializer
%2 = call <2 x i32> @llvm.masked.load.v2i32.p0v2i32(<2 x i32>* %src, i32 4, <2 x i1> %1, <2 x i32> undef) %2 = call <2 x i32> @llvm.masked.load.v2i32.p0v2i32(<2 x i32>* %src, i32 4, <2 x i1> %1, <2 x i32> undef)
%3 = zext <2 x i32> %2 to <2 x i64> %3 = zext <2 x i32> %2 to <2 x i64>
call void @llvm.masked.store.v2i64.p0v2i64(<2 x i64> %3, <2 x i64>* %dest, i32 8, <2 x i1> %1) call void @llvm.masked.store.v2i64.p0v2i64(<2 x i64> %3, <2 x i64>* %dest, i32 8, <2 x i1> %1)
ret void ret void
} }
define void @foo_zext_v2i64_v2i32_unaligned(<2 x i64> *%dest, <2 x i32> *%mask, <2 x i32> *%src) { define void @foo_zext_v2i64_v2i32_unaligned(<2 x i64> *%dest, <2 x i32> *%mask, <2 x i32> *%src) {
; CHECK-LE-LABEL: foo_zext_v2i64_v2i32_unaligned: ; CHECK-LE-LABEL: foo_zext_v2i64_v2i32_unaligned:
; CHECK-LE: @ %bb.0: @ %entry ; CHECK-LE: @ %bb.0: @ %entry
; CHECK-LE-NEXT: .save {r4, r5, r7, lr} ; CHECK-LE-NEXT: .save {r4, lr}
; CHECK-LE-NEXT: push {r4, r5, r7, lr} ; CHECK-LE-NEXT: push {r4, lr}
; CHECK-LE-NEXT: .pad #4 ; CHECK-LE-NEXT: .pad #4
; CHECK-LE-NEXT: sub sp, #4 ; CHECK-LE-NEXT: sub sp, #4
; CHECK-LE-NEXT: ldrd lr, r5, [r1] ; CHECK-LE-NEXT: ldrd lr, r1, [r1]
; CHECK-LE-NEXT: movs r3, #0 ; CHECK-LE-NEXT: rsbs r3, r1, #0
; CHECK-LE-NEXT: mov.w r12, #0
; CHECK-LE-NEXT: sbcs.w r3, r12, r1, asr #31
; CHECK-LE-NEXT: cset r3, lt
; CHECK-LE-NEXT: @ implicit-def: $q0 ; CHECK-LE-NEXT: @ implicit-def: $q0
; CHECK-LE-NEXT: vmov.i64 q2, #0xffffffff ; CHECK-LE-NEXT: vmov q1[2], q1[0], lr, r1
; CHECK-LE-NEXT: rsbs.w r1, lr, #0
; CHECK-LE-NEXT: vmov q1[2], q1[0], lr, r5
; CHECK-LE-NEXT: sbcs.w r1, r3, lr, asr #31
; CHECK-LE-NEXT: mov.w r1, #0
; CHECK-LE-NEXT: it lt
; CHECK-LE-NEXT: movlt r1, #1
; CHECK-LE-NEXT: rsbs r4, r5, #0
; CHECK-LE-NEXT: sbcs.w r4, r3, r5, asr #31
; CHECK-LE-NEXT: it lt
; CHECK-LE-NEXT: movlt r3, #1
; CHECK-LE-NEXT: cmp r3, #0 ; CHECK-LE-NEXT: cmp r3, #0
; CHECK-LE-NEXT: it ne ; CHECK-LE-NEXT: it ne
; CHECK-LE-NEXT: mvnne r3, #1 ; CHECK-LE-NEXT: mvnne r3, #1
; CHECK-LE-NEXT: bfi r3, r1, #0, #1 ; CHECK-LE-NEXT: rsbs.w r4, lr, #0
; CHECK-LE-NEXT: vmov.i64 q2, #0xffffffff
; CHECK-LE-NEXT: sbcs.w r4, r12, lr, asr #31
; CHECK-LE-NEXT: cset r4, lt
; CHECK-LE-NEXT: bfi r3, r4, #0, #1
; CHECK-LE-NEXT: and r12, r3, #3 ; CHECK-LE-NEXT: and r12, r3, #3
; CHECK-LE-NEXT: lsls r1, r3, #31 ; CHECK-LE-NEXT: lsls r3, r3, #31
; CHECK-LE-NEXT: itt ne ; CHECK-LE-NEXT: itt ne
; CHECK-LE-NEXT: ldrne r1, [r2] ; CHECK-LE-NEXT: ldrne r3, [r2]
; CHECK-LE-NEXT: vmovne.32 q0[0], r1 ; CHECK-LE-NEXT: vmovne.32 q0[0], r3
; CHECK-LE-NEXT: lsls.w r1, r12, #30 ; CHECK-LE-NEXT: lsls.w r1, r12, #30
; CHECK-LE-NEXT: itt mi ; CHECK-LE-NEXT: itt mi
; CHECK-LE-NEXT: ldrmi r1, [r2, #4] ; CHECK-LE-NEXT: ldrmi r1, [r2, #4]
; CHECK-LE-NEXT: vmovmi.32 q0[2], r1 ; CHECK-LE-NEXT: vmovmi.32 q0[2], r1
; CHECK-LE-NEXT: vmov r1, s4 ; CHECK-LE-NEXT: vmov r2, s6
; CHECK-LE-NEXT: movs r2, #0 ; CHECK-LE-NEXT: movs r1, #0
; CHECK-LE-NEXT: vand q0, q0, q2 ; CHECK-LE-NEXT: vand q0, q0, q2
; CHECK-LE-NEXT: rsbs r3, r1, #0 ; CHECK-LE-NEXT: rsbs r3, r2, #0
; CHECK-LE-NEXT: vmov r3, s6 ; CHECK-LE-NEXT: vmov r3, s4
; CHECK-LE-NEXT: sbcs.w r1, r2, r1, asr #31 ; CHECK-LE-NEXT: sbcs.w r2, r1, r2, asr #31
; CHECK-LE-NEXT: mov.w r1, #0 ; CHECK-LE-NEXT: cset r2, lt
; CHECK-LE-NEXT: it lt
; CHECK-LE-NEXT: movlt r1, #1
; CHECK-LE-NEXT: rsbs r5, r3, #0
; CHECK-LE-NEXT: sbcs.w r3, r2, r3, asr #31
; CHECK-LE-NEXT: it lt
; CHECK-LE-NEXT: movlt r2, #1
; CHECK-LE-NEXT: cmp r2, #0 ; CHECK-LE-NEXT: cmp r2, #0
; CHECK-LE-NEXT: it ne ; CHECK-LE-NEXT: it ne
; CHECK-LE-NEXT: mvnne r2, #1 ; CHECK-LE-NEXT: mvnne r2, #1
; CHECK-LE-NEXT: rsbs r4, r3, #0
; CHECK-LE-NEXT: sbcs.w r1, r1, r3, asr #31
; CHECK-LE-NEXT: cset r1, lt
; CHECK-LE-NEXT: bfi r2, r1, #0, #1 ; CHECK-LE-NEXT: bfi r2, r1, #0, #1
; CHECK-LE-NEXT: and r1, r2, #3 ; CHECK-LE-NEXT: and r1, r2, #3
; CHECK-LE-NEXT: lsls r2, r2, #31 ; CHECK-LE-NEXT: lsls r2, r2, #31
; CHECK-LE-NEXT: itt ne ; CHECK-LE-NEXT: itt ne
; CHECK-LE-NEXT: vmovne r2, r3, d0 ; CHECK-LE-NEXT: vmovne r2, r3, d0
; CHECK-LE-NEXT: strdne r2, r3, [r0] ; CHECK-LE-NEXT: strdne r2, r3, [r0]
; CHECK-LE-NEXT: lsls r1, r1, #30 ; CHECK-LE-NEXT: lsls r1, r1, #30
; CHECK-LE-NEXT: itt mi ; CHECK-LE-NEXT: itt mi
; CHECK-LE-NEXT: vmovmi r1, r2, d1 ; CHECK-LE-NEXT: vmovmi r1, r2, d1
; CHECK-LE-NEXT: strdmi r1, r2, [r0, #8] ; CHECK-LE-NEXT: strdmi r1, r2, [r0, #8]
; CHECK-LE-NEXT: add sp, #4 ; CHECK-LE-NEXT: add sp, #4
; CHECK-LE-NEXT: pop {r4, r5, r7, pc} ; CHECK-LE-NEXT: pop {r4, pc}
; ;
; CHECK-BE-LABEL: foo_zext_v2i64_v2i32_unaligned: ; CHECK-BE-LABEL: foo_zext_v2i64_v2i32_unaligned:
; CHECK-BE: @ %bb.0: @ %entry ; CHECK-BE: @ %bb.0: @ %entry
; CHECK-BE-NEXT: .save {r7, lr} ; CHECK-BE-NEXT: .save {r7, lr}
; CHECK-BE-NEXT: push {r7, lr} ; CHECK-BE-NEXT: push {r7, lr}
; CHECK-BE-NEXT: .pad #4 ; CHECK-BE-NEXT: .pad #4
; CHECK-BE-NEXT: sub sp, #4 ; CHECK-BE-NEXT: sub sp, #4
; CHECK-BE-NEXT: ldrd r12, lr, [r1] ; CHECK-BE-NEXT: ldrd r3, lr, [r1]
; CHECK-BE-NEXT: rsbs.w r1, lr, #0 ; CHECK-BE-NEXT: mov.w r12, #0
; CHECK-BE-NEXT: mov.w r3, #0 ; CHECK-BE-NEXT: @ implicit-def: $q0
; CHECK-BE-NEXT: sbcs.w r1, r3, lr, asr #31 ; CHECK-BE-NEXT: rsbs r1, r3, #0
; CHECK-BE-NEXT: vmov q0[3], q0[1], r12, lr ; CHECK-BE-NEXT: vmov q1[3], q1[1], r3, lr
; CHECK-BE-NEXT: mov.w lr, #0 ; CHECK-BE-NEXT: sbcs.w r1, r12, r3, asr #31
; CHECK-BE-NEXT: it lt ; CHECK-BE-NEXT: cset r3, lt
; CHECK-BE-NEXT: movlt.w lr, #1
; CHECK-BE-NEXT: rsbs.w r1, r12, #0
; CHECK-BE-NEXT: sbcs.w r1, r3, r12, asr #31
; CHECK-BE-NEXT: it lt
; CHECK-BE-NEXT: movlt r3, #1
; CHECK-BE-NEXT: cmp r3, #0 ; CHECK-BE-NEXT: cmp r3, #0
; CHECK-BE-NEXT: it ne ; CHECK-BE-NEXT: it ne
; CHECK-BE-NEXT: mvnne r3, #1 ; CHECK-BE-NEXT: mvnne r3, #1
; CHECK-BE-NEXT: bfi r3, lr, #0, #1 ; CHECK-BE-NEXT: rsbs.w r1, lr, #0
; CHECK-BE-NEXT: @ implicit-def: $q1 ; CHECK-BE-NEXT: sbcs.w r1, r12, lr, asr #31
; CHECK-BE-NEXT: cset r1, lt
; CHECK-BE-NEXT: bfi r3, r1, #0, #1
; CHECK-BE-NEXT: and r1, r3, #3 ; CHECK-BE-NEXT: and r1, r3, #3
; CHECK-BE-NEXT: lsls r3, r3, #30 ; CHECK-BE-NEXT: lsls r3, r3, #30
; CHECK-BE-NEXT: bpl .LBB8_2 ; CHECK-BE-NEXT: bpl .LBB8_2
; CHECK-BE-NEXT: @ %bb.1: @ %cond.load ; CHECK-BE-NEXT: @ %bb.1: @ %cond.load
; CHECK-BE-NEXT: ldr r3, [r2] ; CHECK-BE-NEXT: ldr r3, [r2]
; CHECK-BE-NEXT: vmov.32 q2[1], r3 ; CHECK-BE-NEXT: vmov.32 q2[1], r3
; CHECK-BE-NEXT: vrev64.32 q1, q2 ; CHECK-BE-NEXT: vrev64.32 q0, q2
; CHECK-BE-NEXT: .LBB8_2: @ %else ; CHECK-BE-NEXT: .LBB8_2: @ %else
; CHECK-BE-NEXT: vrev64.32 q2, q0 ; CHECK-BE-NEXT: vrev64.32 q2, q1
; CHECK-BE-NEXT: lsls r1, r1, #31 ; CHECK-BE-NEXT: lsls r1, r1, #31
; CHECK-BE-NEXT: beq .LBB8_4 ; CHECK-BE-NEXT: beq .LBB8_4
; CHECK-BE-NEXT: @ %bb.3: @ %cond.load1 ; CHECK-BE-NEXT: @ %bb.3: @ %cond.load1
; CHECK-BE-NEXT: ldr r1, [r2, #4] ; CHECK-BE-NEXT: ldr r1, [r2, #4]
; CHECK-BE-NEXT: vrev64.32 q0, q1
; CHECK-BE-NEXT: vmov.32 q0[3], r1
; CHECK-BE-NEXT: vrev64.32 q1, q0 ; CHECK-BE-NEXT: vrev64.32 q1, q0
; CHECK-BE-NEXT: vmov.32 q1[3], r1
; CHECK-BE-NEXT: vrev64.32 q0, q1
; CHECK-BE-NEXT: .LBB8_4: @ %else2 ; CHECK-BE-NEXT: .LBB8_4: @ %else2
; CHECK-BE-NEXT: vrev64.32 q3, q2 ; CHECK-BE-NEXT: vrev64.32 q3, q2
; CHECK-BE-NEXT: movs r2, #0
; CHECK-BE-NEXT: vmov r1, s15
; CHECK-BE-NEXT: mov.w r12, #0 ; CHECK-BE-NEXT: mov.w r12, #0
; CHECK-BE-NEXT: vmov.i64 q0, #0xffffffff ; CHECK-BE-NEXT: vmov r2, s13
; CHECK-BE-NEXT: vand q0, q1, q0 ; CHECK-BE-NEXT: vmov.i64 q1, #0xffffffff
; CHECK-BE-NEXT: rsbs r3, r1, #0 ; CHECK-BE-NEXT: vand q0, q0, q1
; CHECK-BE-NEXT: vmov r3, s13 ; CHECK-BE-NEXT: rsbs r3, r2, #0
; CHECK-BE-NEXT: sbcs.w r1, r2, r1, asr #31 ; CHECK-BE-NEXT: vmov r3, s15
; CHECK-BE-NEXT: it lt ; CHECK-BE-NEXT: sbcs.w r2, r12, r2, asr #31
; CHECK-BE-NEXT: movlt.w r12, #1 ; CHECK-BE-NEXT: cset r2, lt
; CHECK-BE-NEXT: rsbs r1, r3, #0
; CHECK-BE-NEXT: sbcs.w r1, r2, r3, asr #31
; CHECK-BE-NEXT: it lt
; CHECK-BE-NEXT: movlt r2, #1
; CHECK-BE-NEXT: cmp r2, #0 ; CHECK-BE-NEXT: cmp r2, #0
; CHECK-BE-NEXT: it ne ; CHECK-BE-NEXT: it ne
; CHECK-BE-NEXT: mvnne r2, #1 ; CHECK-BE-NEXT: mvnne r2, #1
; CHECK-BE-NEXT: bfi r2, r12, #0, #1 ; CHECK-BE-NEXT: rsbs r1, r3, #0
; CHECK-BE-NEXT: sbcs.w r1, r12, r3, asr #31
; CHECK-BE-NEXT: cset r1, lt
; CHECK-BE-NEXT: bfi r2, r1, #0, #1
; CHECK-BE-NEXT: and r1, r2, #3 ; CHECK-BE-NEXT: and r1, r2, #3
; CHECK-BE-NEXT: lsls r2, r2, #30 ; CHECK-BE-NEXT: lsls r2, r2, #30
; CHECK-BE-NEXT: itt mi ; CHECK-BE-NEXT: itt mi
; CHECK-BE-NEXT: vmovmi r2, r3, d0 ; CHECK-BE-NEXT: vmovmi r2, r3, d0
; CHECK-BE-NEXT: strdmi r3, r2, [r0] ; CHECK-BE-NEXT: strdmi r3, r2, [r0]
; CHECK-BE-NEXT: lsls r1, r1, #31 ; CHECK-BE-NEXT: lsls r1, r1, #31
; CHECK-BE-NEXT: itt ne ; CHECK-BE-NEXT: itt ne
; CHECK-BE-NEXT: vmovne r1, r2, d1 ; CHECK-BE-NEXT: vmovne r1, r2, d1
▲ Show 20 Lines • Show All 563 Lines • Show Last 20 Lines

llvm/test/CodeGen/Thumb2/mve-masked-load.ll

Show First 20 Lines • Show All 1,748 Lines • ▼ Show 20 Lines
define arm_aapcs_vfpcc <2 x i64> @masked_v2i64_align4_zero(<2 x i64> *%dest, <2 x i64> %a) { define arm_aapcs_vfpcc <2 x i64> @masked_v2i64_align4_zero(<2 x i64> *%dest, <2 x i64> %a) {
; CHECK-LE-LABEL: masked_v2i64_align4_zero: ; CHECK-LE-LABEL: masked_v2i64_align4_zero:
; CHECK-LE: @ %bb.0: @ %entry ; CHECK-LE: @ %bb.0: @ %entry
; CHECK-LE-NEXT: .save {r7, lr} ; CHECK-LE-NEXT: .save {r7, lr}
; CHECK-LE-NEXT: push {r7, lr} ; CHECK-LE-NEXT: push {r7, lr}
; CHECK-LE-NEXT: .pad #4 ; CHECK-LE-NEXT: .pad #4
; CHECK-LE-NEXT: sub sp, #4 ; CHECK-LE-NEXT: sub sp, #4
; CHECK-LE-NEXT: vmov r1, r2, d0 ; CHECK-LE-NEXT: vmov r1, r2, d1
; CHECK-LE-NEXT: movs r3, #0 ; CHECK-LE-NEXT: mov.w lr, #0
; CHECK-LE-NEXT: vmov lr, r12, d1 ; CHECK-LE-NEXT: vmov r3, r12, d0
; CHECK-LE-NEXT: rsbs r1, r1, #0 ; CHECK-LE-NEXT: rsbs r1, r1, #0
; CHECK-LE-NEXT: sbcs.w r1, r3, r2 ; CHECK-LE-NEXT: sbcs.w r1, lr, r2
; CHECK-LE-NEXT: mov.w r1, #0 ; CHECK-LE-NEXT: cset r2, lt
; CHECK-LE-NEXT: it lt ; CHECK-LE-NEXT: cmp r2, #0
; CHECK-LE-NEXT: movlt r1, #1
; CHECK-LE-NEXT: rsbs.w r2, lr, #0
; CHECK-LE-NEXT: sbcs.w r2, r3, r12
; CHECK-LE-NEXT: it lt
; CHECK-LE-NEXT: movlt r3, #1
; CHECK-LE-NEXT: cmp r3, #0
; CHECK-LE-NEXT: it ne ; CHECK-LE-NEXT: it ne
; CHECK-LE-NEXT: mvnne r3, #1 ; CHECK-LE-NEXT: mvnne r2, #1
; CHECK-LE-NEXT: bfi r3, r1, #0, #1 ; CHECK-LE-NEXT: rsbs r1, r3, #0
; CHECK-LE-NEXT: and r1, r3, #3 ; CHECK-LE-NEXT: sbcs.w r1, lr, r12
; CHECK-LE-NEXT: lsls r2, r3, #31 ; CHECK-LE-NEXT: cset r1, lt
; CHECK-LE-NEXT: bfi r2, r1, #0, #1
; CHECK-LE-NEXT: and r1, r2, #3
; CHECK-LE-NEXT: lsls r2, r2, #31
; CHECK-LE-NEXT: beq .LBB49_2 ; CHECK-LE-NEXT: beq .LBB49_2
; CHECK-LE-NEXT: @ %bb.1: @ %cond.load ; CHECK-LE-NEXT: @ %bb.1: @ %cond.load
; CHECK-LE-NEXT: vldr d1, .LCPI49_0 ; CHECK-LE-NEXT: vldr d1, .LCPI49_0
; CHECK-LE-NEXT: vldr d0, [r0] ; CHECK-LE-NEXT: vldr d0, [r0]
; CHECK-LE-NEXT: b .LBB49_3 ; CHECK-LE-NEXT: b .LBB49_3
; CHECK-LE-NEXT: .LBB49_2: ; CHECK-LE-NEXT: .LBB49_2:
; CHECK-LE-NEXT: vmov.i32 q0, #0x0 ; CHECK-LE-NEXT: vmov.i32 q0, #0x0
; CHECK-LE-NEXT: .LBB49_3: @ %else ; CHECK-LE-NEXT: .LBB49_3: @ %else
Show All 10 Lines
; ;
; CHECK-BE-LABEL: masked_v2i64_align4_zero: ; CHECK-BE-LABEL: masked_v2i64_align4_zero:
; CHECK-BE: @ %bb.0: @ %entry ; CHECK-BE: @ %bb.0: @ %entry
; CHECK-BE-NEXT: .save {r7, lr} ; CHECK-BE-NEXT: .save {r7, lr}
; CHECK-BE-NEXT: push {r7, lr} ; CHECK-BE-NEXT: push {r7, lr}
; CHECK-BE-NEXT: .pad #4 ; CHECK-BE-NEXT: .pad #4
; CHECK-BE-NEXT: sub sp, #4 ; CHECK-BE-NEXT: sub sp, #4
; CHECK-BE-NEXT: vrev64.32 q1, q0 ; CHECK-BE-NEXT: vrev64.32 q1, q0
; CHECK-BE-NEXT: movs r3, #0 ; CHECK-BE-NEXT: mov.w lr, #0
; CHECK-BE-NEXT: vmov r1, r2, d3 ; CHECK-BE-NEXT: vmov r1, r2, d2
; CHECK-BE-NEXT: vmov r12, lr, d2 ; CHECK-BE-NEXT: vmov r12, r3, d3
; CHECK-BE-NEXT: rsbs r2, r2, #0 ; CHECK-BE-NEXT: rsbs r2, r2, #0
; CHECK-BE-NEXT: sbcs.w r1, r3, r1 ; CHECK-BE-NEXT: sbcs.w r1, lr, r1
; CHECK-BE-NEXT: mov.w r1, #0 ; CHECK-BE-NEXT: cset r2, lt
; CHECK-BE-NEXT: it lt ; CHECK-BE-NEXT: cmp r2, #0
; CHECK-BE-NEXT: movlt r1, #1
; CHECK-BE-NEXT: rsbs.w r2, lr, #0
; CHECK-BE-NEXT: sbcs.w r2, r3, r12
; CHECK-BE-NEXT: it lt
; CHECK-BE-NEXT: movlt r3, #1
; CHECK-BE-NEXT: cmp r3, #0
; CHECK-BE-NEXT: it ne ; CHECK-BE-NEXT: it ne
; CHECK-BE-NEXT: mvnne r3, #1 ; CHECK-BE-NEXT: mvnne r2, #1
; CHECK-BE-NEXT: bfi r3, r1, #0, #1 ; CHECK-BE-NEXT: rsbs r1, r3, #0
; CHECK-BE-NEXT: and r1, r3, #3 ; CHECK-BE-NEXT: sbcs.w r1, lr, r12
; CHECK-BE-NEXT: lsls r2, r3, #30 ; CHECK-BE-NEXT: cset r1, lt
; CHECK-BE-NEXT: bfi r2, r1, #0, #1
; CHECK-BE-NEXT: and r1, r2, #3
; CHECK-BE-NEXT: lsls r2, r2, #30
; CHECK-BE-NEXT: bpl .LBB49_2 ; CHECK-BE-NEXT: bpl .LBB49_2
; CHECK-BE-NEXT: @ %bb.1: @ %cond.load ; CHECK-BE-NEXT: @ %bb.1: @ %cond.load
; CHECK-BE-NEXT: vldr d1, .LCPI49_0 ; CHECK-BE-NEXT: vldr d1, .LCPI49_0
; CHECK-BE-NEXT: vldr d0, [r0] ; CHECK-BE-NEXT: vldr d0, [r0]
; CHECK-BE-NEXT: b .LBB49_3 ; CHECK-BE-NEXT: b .LBB49_3
; CHECK-BE-NEXT: .LBB49_2: ; CHECK-BE-NEXT: .LBB49_2:
; CHECK-BE-NEXT: vmov.i32 q0, #0x0 ; CHECK-BE-NEXT: vmov.i32 q0, #0x0
; CHECK-BE-NEXT: .LBB49_3: @ %else ; CHECK-BE-NEXT: .LBB49_3: @ %else
Show All 15 Lines
define arm_aapcs_vfpcc <2 x double> @masked_v2f64_align4_zero(<2 x double> *%dest, <2 x double> %a, <2 x i64> %b) { define arm_aapcs_vfpcc <2 x double> @masked_v2f64_align4_zero(<2 x double> *%dest, <2 x double> %a, <2 x i64> %b) {
; CHECK-LE-LABEL: masked_v2f64_align4_zero: ; CHECK-LE-LABEL: masked_v2f64_align4_zero:
; CHECK-LE: @ %bb.0: @ %entry ; CHECK-LE: @ %bb.0: @ %entry
; CHECK-LE-NEXT: .save {r7, lr} ; CHECK-LE-NEXT: .save {r7, lr}
; CHECK-LE-NEXT: push {r7, lr} ; CHECK-LE-NEXT: push {r7, lr}
; CHECK-LE-NEXT: .pad #4 ; CHECK-LE-NEXT: .pad #4
; CHECK-LE-NEXT: sub sp, #4 ; CHECK-LE-NEXT: sub sp, #4
; CHECK-LE-NEXT: vmov r1, r2, d2 ; CHECK-LE-NEXT: vmov r1, r2, d3
; CHECK-LE-NEXT: movs r3, #0 ; CHECK-LE-NEXT: mov.w lr, #0
; CHECK-LE-NEXT: vmov lr, r12, d3 ; CHECK-LE-NEXT: vmov r3, r12, d2
; CHECK-LE-NEXT: rsbs r1, r1, #0 ; CHECK-LE-NEXT: rsbs r1, r1, #0
; CHECK-LE-NEXT: sbcs.w r1, r3, r2 ; CHECK-LE-NEXT: sbcs.w r1, lr, r2
; CHECK-LE-NEXT: mov.w r1, #0 ; CHECK-LE-NEXT: cset r2, lt
; CHECK-LE-NEXT: it lt ; CHECK-LE-NEXT: cmp r2, #0
; CHECK-LE-NEXT: movlt r1, #1
; CHECK-LE-NEXT: rsbs.w r2, lr, #0
; CHECK-LE-NEXT: sbcs.w r2, r3, r12
; CHECK-LE-NEXT: it lt
; CHECK-LE-NEXT: movlt r3, #1
; CHECK-LE-NEXT: cmp r3, #0
; CHECK-LE-NEXT: it ne ; CHECK-LE-NEXT: it ne
; CHECK-LE-NEXT: mvnne r3, #1 ; CHECK-LE-NEXT: mvnne r2, #1
; CHECK-LE-NEXT: bfi r3, r1, #0, #1 ; CHECK-LE-NEXT: rsbs r1, r3, #0
; CHECK-LE-NEXT: and r1, r3, #3 ; CHECK-LE-NEXT: sbcs.w r1, lr, r12
; CHECK-LE-NEXT: lsls r2, r3, #31 ; CHECK-LE-NEXT: cset r1, lt
; CHECK-LE-NEXT: bfi r2, r1, #0, #1
; CHECK-LE-NEXT: and r1, r2, #3
; CHECK-LE-NEXT: lsls r2, r2, #31
; CHECK-LE-NEXT: beq .LBB50_2 ; CHECK-LE-NEXT: beq .LBB50_2
; CHECK-LE-NEXT: @ %bb.1: @ %cond.load ; CHECK-LE-NEXT: @ %bb.1: @ %cond.load
; CHECK-LE-NEXT: vldr d1, .LCPI50_0 ; CHECK-LE-NEXT: vldr d1, .LCPI50_0
; CHECK-LE-NEXT: vldr d0, [r0] ; CHECK-LE-NEXT: vldr d0, [r0]
; CHECK-LE-NEXT: b .LBB50_3 ; CHECK-LE-NEXT: b .LBB50_3
; CHECK-LE-NEXT: .LBB50_2: ; CHECK-LE-NEXT: .LBB50_2:
; CHECK-LE-NEXT: vmov.i32 q0, #0x0 ; CHECK-LE-NEXT: vmov.i32 q0, #0x0
; CHECK-LE-NEXT: .LBB50_3: @ %else ; CHECK-LE-NEXT: .LBB50_3: @ %else
Show All 10 Lines
; ;
; CHECK-BE-LABEL: masked_v2f64_align4_zero: ; CHECK-BE-LABEL: masked_v2f64_align4_zero:
; CHECK-BE: @ %bb.0: @ %entry ; CHECK-BE: @ %bb.0: @ %entry
; CHECK-BE-NEXT: .save {r7, lr} ; CHECK-BE-NEXT: .save {r7, lr}
; CHECK-BE-NEXT: push {r7, lr} ; CHECK-BE-NEXT: push {r7, lr}
; CHECK-BE-NEXT: .pad #4 ; CHECK-BE-NEXT: .pad #4
; CHECK-BE-NEXT: sub sp, #4 ; CHECK-BE-NEXT: sub sp, #4
; CHECK-BE-NEXT: vrev64.32 q0, q1 ; CHECK-BE-NEXT: vrev64.32 q0, q1
; CHECK-BE-NEXT: movs r3, #0 ; CHECK-BE-NEXT: mov.w lr, #0
; CHECK-BE-NEXT: vmov r1, r2, d1 ; CHECK-BE-NEXT: vmov r1, r2, d0
; CHECK-BE-NEXT: vmov r12, lr, d0 ; CHECK-BE-NEXT: vmov r12, r3, d1
; CHECK-BE-NEXT: rsbs r2, r2, #0 ; CHECK-BE-NEXT: rsbs r2, r2, #0
; CHECK-BE-NEXT: sbcs.w r1, r3, r1 ; CHECK-BE-NEXT: sbcs.w r1, lr, r1
; CHECK-BE-NEXT: mov.w r1, #0 ; CHECK-BE-NEXT: cset r2, lt
; CHECK-BE-NEXT: it lt ; CHECK-BE-NEXT: cmp r2, #0
; CHECK-BE-NEXT: movlt r1, #1
; CHECK-BE-NEXT: rsbs.w r2, lr, #0
; CHECK-BE-NEXT: sbcs.w r2, r3, r12
; CHECK-BE-NEXT: it lt
; CHECK-BE-NEXT: movlt r3, #1
; CHECK-BE-NEXT: cmp r3, #0
; CHECK-BE-NEXT: it ne ; CHECK-BE-NEXT: it ne
; CHECK-BE-NEXT: mvnne r3, #1 ; CHECK-BE-NEXT: mvnne r2, #1
; CHECK-BE-NEXT: bfi r3, r1, #0, #1 ; CHECK-BE-NEXT: rsbs r1, r3, #0
; CHECK-BE-NEXT: and r1, r3, #3 ; CHECK-BE-NEXT: sbcs.w r1, lr, r12
; CHECK-BE-NEXT: lsls r2, r3, #30 ; CHECK-BE-NEXT: cset r1, lt
; CHECK-BE-NEXT: bfi r2, r1, #0, #1
; CHECK-BE-NEXT: and r1, r2, #3
; CHECK-BE-NEXT: lsls r2, r2, #30
; CHECK-BE-NEXT: bpl .LBB50_2 ; CHECK-BE-NEXT: bpl .LBB50_2
; CHECK-BE-NEXT: @ %bb.1: @ %cond.load ; CHECK-BE-NEXT: @ %bb.1: @ %cond.load
; CHECK-BE-NEXT: vldr d1, .LCPI50_0 ; CHECK-BE-NEXT: vldr d1, .LCPI50_0
; CHECK-BE-NEXT: vldr d0, [r0] ; CHECK-BE-NEXT: vldr d0, [r0]
; CHECK-BE-NEXT: b .LBB50_3 ; CHECK-BE-NEXT: b .LBB50_3
; CHECK-BE-NEXT: .LBB50_2: ; CHECK-BE-NEXT: .LBB50_2:
; CHECK-BE-NEXT: vmov.i32 q0, #0x0 ; CHECK-BE-NEXT: vmov.i32 q0, #0x0
; CHECK-BE-NEXT: .LBB50_3: @ %else ; CHECK-BE-NEXT: .LBB50_3: @ %else
▲ Show 20 Lines • Show All 267 Lines • Show Last 20 Lines

llvm/test/CodeGen/Thumb2/mve-masked-store.ll

Show First 20 Lines • Show All 933 Lines • ▼ Show 20 Lines
define arm_aapcs_vfpcc void @masked_v2i64(<2 x i64> *%dest, <2 x i64> %a) { define arm_aapcs_vfpcc void @masked_v2i64(<2 x i64> *%dest, <2 x i64> %a) {
; CHECK-LE-LABEL: masked_v2i64: ; CHECK-LE-LABEL: masked_v2i64:
; CHECK-LE: @ %bb.0: @ %entry ; CHECK-LE: @ %bb.0: @ %entry
; CHECK-LE-NEXT: .save {r7, lr} ; CHECK-LE-NEXT: .save {r7, lr}
; CHECK-LE-NEXT: push {r7, lr} ; CHECK-LE-NEXT: push {r7, lr}
; CHECK-LE-NEXT: .pad #4 ; CHECK-LE-NEXT: .pad #4
; CHECK-LE-NEXT: sub sp, #4 ; CHECK-LE-NEXT: sub sp, #4
; CHECK-LE-NEXT: vmov r1, r2, d0 ; CHECK-LE-NEXT: vmov r1, r2, d1
; CHECK-LE-NEXT: movs r3, #0 ; CHECK-LE-NEXT: mov.w lr, #0
; CHECK-LE-NEXT: vmov lr, r12, d1 ; CHECK-LE-NEXT: vmov r3, r12, d0
; CHECK-LE-NEXT: rsbs r1, r1, #0 ; CHECK-LE-NEXT: rsbs r1, r1, #0
; CHECK-LE-NEXT: sbcs.w r1, r3, r2 ; CHECK-LE-NEXT: sbcs.w r1, lr, r2
; CHECK-LE-NEXT: mov.w r1, #0 ; CHECK-LE-NEXT: cset r2, lt
; CHECK-LE-NEXT: it lt ; CHECK-LE-NEXT: cmp r2, #0
; CHECK-LE-NEXT: movlt r1, #1
; CHECK-LE-NEXT: rsbs.w r2, lr, #0
; CHECK-LE-NEXT: sbcs.w r2, r3, r12
; CHECK-LE-NEXT: it lt
; CHECK-LE-NEXT: movlt r3, #1
; CHECK-LE-NEXT: cmp r3, #0
; CHECK-LE-NEXT: it ne ; CHECK-LE-NEXT: it ne
; CHECK-LE-NEXT: mvnne r3, #1 ; CHECK-LE-NEXT: mvnne r2, #1
; CHECK-LE-NEXT: bfi r3, r1, #0, #1 ; CHECK-LE-NEXT: rsbs r1, r3, #0
; CHECK-LE-NEXT: and r1, r3, #3 ; CHECK-LE-NEXT: sbcs.w r1, lr, r12
; CHECK-LE-NEXT: lsls r2, r3, #31 ; CHECK-LE-NEXT: cset r1, lt
; CHECK-LE-NEXT: bfi r2, r1, #0, #1
; CHECK-LE-NEXT: and r1, r2, #3
; CHECK-LE-NEXT: lsls r2, r2, #31
; CHECK-LE-NEXT: it ne ; CHECK-LE-NEXT: it ne
; CHECK-LE-NEXT: vstrne d0, [r0] ; CHECK-LE-NEXT: vstrne d0, [r0]
; CHECK-LE-NEXT: lsls r1, r1, #30 ; CHECK-LE-NEXT: lsls r1, r1, #30
; CHECK-LE-NEXT: it mi ; CHECK-LE-NEXT: it mi
; CHECK-LE-NEXT: vstrmi d1, [r0, #8] ; CHECK-LE-NEXT: vstrmi d1, [r0, #8]
; CHECK-LE-NEXT: add sp, #4 ; CHECK-LE-NEXT: add sp, #4
; CHECK-LE-NEXT: pop {r7, pc} ; CHECK-LE-NEXT: pop {r7, pc}
; ;
; CHECK-BE-LABEL: masked_v2i64: ; CHECK-BE-LABEL: masked_v2i64:
; CHECK-BE: @ %bb.0: @ %entry ; CHECK-BE: @ %bb.0: @ %entry
; CHECK-BE-NEXT: .save {r7, lr} ; CHECK-BE-NEXT: .save {r7, lr}
; CHECK-BE-NEXT: push {r7, lr} ; CHECK-BE-NEXT: push {r7, lr}
; CHECK-BE-NEXT: .pad #4 ; CHECK-BE-NEXT: .pad #4
; CHECK-BE-NEXT: sub sp, #4 ; CHECK-BE-NEXT: sub sp, #4
; CHECK-BE-NEXT: vrev64.32 q1, q0 ; CHECK-BE-NEXT: vrev64.32 q1, q0
; CHECK-BE-NEXT: movs r3, #0 ; CHECK-BE-NEXT: mov.w lr, #0
; CHECK-BE-NEXT: vmov r1, r2, d3 ; CHECK-BE-NEXT: vmov r1, r2, d2
; CHECK-BE-NEXT: vmov r12, lr, d2 ; CHECK-BE-NEXT: vmov r12, r3, d3
; CHECK-BE-NEXT: rsbs r2, r2, #0 ; CHECK-BE-NEXT: rsbs r2, r2, #0
; CHECK-BE-NEXT: sbcs.w r1, r3, r1 ; CHECK-BE-NEXT: sbcs.w r1, lr, r1
; CHECK-BE-NEXT: mov.w r1, #0 ; CHECK-BE-NEXT: cset r2, lt
; CHECK-BE-NEXT: it lt ; CHECK-BE-NEXT: cmp r2, #0
; CHECK-BE-NEXT: movlt r1, #1
; CHECK-BE-NEXT: rsbs.w r2, lr, #0
; CHECK-BE-NEXT: sbcs.w r2, r3, r12
; CHECK-BE-NEXT: it lt
; CHECK-BE-NEXT: movlt r3, #1
; CHECK-BE-NEXT: cmp r3, #0
; CHECK-BE-NEXT: it ne ; CHECK-BE-NEXT: it ne
; CHECK-BE-NEXT: mvnne r3, #1 ; CHECK-BE-NEXT: mvnne r2, #1
; CHECK-BE-NEXT: bfi r3, r1, #0, #1 ; CHECK-BE-NEXT: rsbs r1, r3, #0
; CHECK-BE-NEXT: and r1, r3, #3 ; CHECK-BE-NEXT: sbcs.w r1, lr, r12
; CHECK-BE-NEXT: lsls r2, r3, #30 ; CHECK-BE-NEXT: cset r1, lt
; CHECK-BE-NEXT: bfi r2, r1, #0, #1
; CHECK-BE-NEXT: and r1, r2, #3
; CHECK-BE-NEXT: lsls r2, r2, #30
; CHECK-BE-NEXT: it mi ; CHECK-BE-NEXT: it mi
; CHECK-BE-NEXT: vstrmi d0, [r0] ; CHECK-BE-NEXT: vstrmi d0, [r0]
; CHECK-BE-NEXT: lsls r1, r1, #31 ; CHECK-BE-NEXT: lsls r1, r1, #31
; CHECK-BE-NEXT: it ne ; CHECK-BE-NEXT: it ne
; CHECK-BE-NEXT: vstrne d1, [r0, #8] ; CHECK-BE-NEXT: vstrne d1, [r0, #8]
; CHECK-BE-NEXT: add sp, #4 ; CHECK-BE-NEXT: add sp, #4
; CHECK-BE-NEXT: pop {r7, pc} ; CHECK-BE-NEXT: pop {r7, pc}
entry: entry:
%c = icmp sgt <2 x i64> %a, zeroinitializer %c = icmp sgt <2 x i64> %a, zeroinitializer
call void @llvm.masked.store.v2i64.p0v2i64(<2 x i64> %a, <2 x i64>* %dest, i32 8, <2 x i1> %c) call void @llvm.masked.store.v2i64.p0v2i64(<2 x i64> %a, <2 x i64>* %dest, i32 8, <2 x i1> %c)
ret void ret void
} }
define arm_aapcs_vfpcc void @masked_v2f64(<2 x double> *%dest, <2 x double> %a, <2 x i64> %b) { define arm_aapcs_vfpcc void @masked_v2f64(<2 x double> *%dest, <2 x double> %a, <2 x i64> %b) {
; CHECK-LE-LABEL: masked_v2f64: ; CHECK-LE-LABEL: masked_v2f64:
; CHECK-LE: @ %bb.0: @ %entry ; CHECK-LE: @ %bb.0: @ %entry
; CHECK-LE-NEXT: .save {r7, lr} ; CHECK-LE-NEXT: .save {r7, lr}
; CHECK-LE-NEXT: push {r7, lr} ; CHECK-LE-NEXT: push {r7, lr}
; CHECK-LE-NEXT: .pad #4 ; CHECK-LE-NEXT: .pad #4
; CHECK-LE-NEXT: sub sp, #4 ; CHECK-LE-NEXT: sub sp, #4
; CHECK-LE-NEXT: vmov r1, r2, d2 ; CHECK-LE-NEXT: vmov r1, r2, d3
; CHECK-LE-NEXT: movs r3, #0 ; CHECK-LE-NEXT: mov.w lr, #0
; CHECK-LE-NEXT: vmov lr, r12, d3 ; CHECK-LE-NEXT: vmov r3, r12, d2
; CHECK-LE-NEXT: rsbs r1, r1, #0 ; CHECK-LE-NEXT: rsbs r1, r1, #0
; CHECK-LE-NEXT: sbcs.w r1, r3, r2 ; CHECK-LE-NEXT: sbcs.w r1, lr, r2
; CHECK-LE-NEXT: mov.w r1, #0 ; CHECK-LE-NEXT: cset r2, lt
; CHECK-LE-NEXT: it lt ; CHECK-LE-NEXT: cmp r2, #0
; CHECK-LE-NEXT: movlt r1, #1
; CHECK-LE-NEXT: rsbs.w r2, lr, #0
; CHECK-LE-NEXT: sbcs.w r2, r3, r12
; CHECK-LE-NEXT: it lt
; CHECK-LE-NEXT: movlt r3, #1
; CHECK-LE-NEXT: cmp r3, #0
; CHECK-LE-NEXT: it ne ; CHECK-LE-NEXT: it ne
; CHECK-LE-NEXT: mvnne r3, #1 ; CHECK-LE-NEXT: mvnne r2, #1
; CHECK-LE-NEXT: bfi r3, r1, #0, #1 ; CHECK-LE-NEXT: rsbs r1, r3, #0
; CHECK-LE-NEXT: and r1, r3, #3 ; CHECK-LE-NEXT: sbcs.w r1, lr, r12
; CHECK-LE-NEXT: lsls r2, r3, #31 ; CHECK-LE-NEXT: cset r1, lt
; CHECK-LE-NEXT: bfi r2, r1, #0, #1
; CHECK-LE-NEXT: and r1, r2, #3
; CHECK-LE-NEXT: lsls r2, r2, #31
; CHECK-LE-NEXT: it ne ; CHECK-LE-NEXT: it ne
; CHECK-LE-NEXT: vstrne d0, [r0] ; CHECK-LE-NEXT: vstrne d0, [r0]
; CHECK-LE-NEXT: lsls r1, r1, #30 ; CHECK-LE-NEXT: lsls r1, r1, #30
; CHECK-LE-NEXT: it mi ; CHECK-LE-NEXT: it mi
; CHECK-LE-NEXT: vstrmi d1, [r0, #8] ; CHECK-LE-NEXT: vstrmi d1, [r0, #8]
; CHECK-LE-NEXT: add sp, #4 ; CHECK-LE-NEXT: add sp, #4
; CHECK-LE-NEXT: pop {r7, pc} ; CHECK-LE-NEXT: pop {r7, pc}
; ;
; CHECK-BE-LABEL: masked_v2f64: ; CHECK-BE-LABEL: masked_v2f64:
; CHECK-BE: @ %bb.0: @ %entry ; CHECK-BE: @ %bb.0: @ %entry
; CHECK-BE-NEXT: .save {r7, lr} ; CHECK-BE-NEXT: .save {r7, lr}
; CHECK-BE-NEXT: push {r7, lr} ; CHECK-BE-NEXT: push {r7, lr}
; CHECK-BE-NEXT: .pad #4 ; CHECK-BE-NEXT: .pad #4
; CHECK-BE-NEXT: sub sp, #4 ; CHECK-BE-NEXT: sub sp, #4
; CHECK-BE-NEXT: vrev64.32 q2, q1 ; CHECK-BE-NEXT: vrev64.32 q2, q1
; CHECK-BE-NEXT: movs r3, #0 ; CHECK-BE-NEXT: mov.w lr, #0
; CHECK-BE-NEXT: vmov r1, r2, d5 ; CHECK-BE-NEXT: vmov r1, r2, d4
; CHECK-BE-NEXT: vmov r12, lr, d4 ; CHECK-BE-NEXT: vmov r12, r3, d5
; CHECK-BE-NEXT: rsbs r2, r2, #0 ; CHECK-BE-NEXT: rsbs r2, r2, #0
; CHECK-BE-NEXT: sbcs.w r1, r3, r1 ; CHECK-BE-NEXT: sbcs.w r1, lr, r1
; CHECK-BE-NEXT: mov.w r1, #0 ; CHECK-BE-NEXT: cset r2, lt
; CHECK-BE-NEXT: it lt ; CHECK-BE-NEXT: cmp r2, #0
; CHECK-BE-NEXT: movlt r1, #1
; CHECK-BE-NEXT: rsbs.w r2, lr, #0
; CHECK-BE-NEXT: sbcs.w r2, r3, r12
; CHECK-BE-NEXT: it lt
; CHECK-BE-NEXT: movlt r3, #1
; CHECK-BE-NEXT: cmp r3, #0
; CHECK-BE-NEXT: it ne ; CHECK-BE-NEXT: it ne
; CHECK-BE-NEXT: mvnne r3, #1 ; CHECK-BE-NEXT: mvnne r2, #1
; CHECK-BE-NEXT: bfi r3, r1, #0, #1 ; CHECK-BE-NEXT: rsbs r1, r3, #0
; CHECK-BE-NEXT: and r1, r3, #3 ; CHECK-BE-NEXT: sbcs.w r1, lr, r12
; CHECK-BE-NEXT: lsls r2, r3, #30 ; CHECK-BE-NEXT: cset r1, lt
; CHECK-BE-NEXT: bfi r2, r1, #0, #1
; CHECK-BE-NEXT: and r1, r2, #3
; CHECK-BE-NEXT: lsls r2, r2, #30
; CHECK-BE-NEXT: it mi ; CHECK-BE-NEXT: it mi
; CHECK-BE-NEXT: vstrmi d0, [r0] ; CHECK-BE-NEXT: vstrmi d0, [r0]
; CHECK-BE-NEXT: lsls r1, r1, #31 ; CHECK-BE-NEXT: lsls r1, r1, #31
; CHECK-BE-NEXT: it ne ; CHECK-BE-NEXT: it ne
; CHECK-BE-NEXT: vstrne d1, [r0, #8] ; CHECK-BE-NEXT: vstrne d1, [r0, #8]
; CHECK-BE-NEXT: add sp, #4 ; CHECK-BE-NEXT: add sp, #4
; CHECK-BE-NEXT: pop {r7, pc} ; CHECK-BE-NEXT: pop {r7, pc}
entry: entry:
▲ Show 20 Lines • Show All 147 Lines • ▼ Show 20 Lines
} }
define arm_aapcs_vfpcc void @masked_v4f16_align4(<4 x half> *%dest, <4 x float> %a) { define arm_aapcs_vfpcc void @masked_v4f16_align4(<4 x half> *%dest, <4 x float> %a) {
; CHECK-LE-LABEL: masked_v4f16_align4: ; CHECK-LE-LABEL: masked_v4f16_align4:
; CHECK-LE: @ %bb.0: @ %entry ; CHECK-LE: @ %bb.0: @ %entry
; CHECK-LE-NEXT: .pad #4 ; CHECK-LE-NEXT: .pad #4
; CHECK-LE-NEXT: sub sp, #4 ; CHECK-LE-NEXT: sub sp, #4
; CHECK-LE-NEXT: vcmp.f32 s0, #0 ; CHECK-LE-NEXT: vcmp.f32 s0, #0
; CHECK-LE-NEXT: movs r1, #0 ; CHECK-LE-NEXT: vcvtb.f16.f32 s4, s0
; CHECK-LE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-LE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-LE-NEXT: it gt ; CHECK-LE-NEXT: vcmp.f32 s1, #0
; CHECK-LE-NEXT: movgt r1, #1 ; CHECK-LE-NEXT: vcvtt.f16.f32 s4, s1
; CHECK-LE-NEXT: vcvtb.f16.f32 s6, s2
; CHECK-LE-NEXT: vcvtt.f16.f32 s5, s3
; CHECK-LE-NEXT: cset r1, gt
; CHECK-LE-NEXT: cmp r1, #0 ; CHECK-LE-NEXT: cmp r1, #0
; CHECK-LE-NEXT: mov.w r1, #0 ; CHECK-LE-NEXT: mov.w r1, #0
; CHECK-LE-NEXT: csetm r3, ne ; CHECK-LE-NEXT: csetm r2, ne
; CHECK-LE-NEXT: vcmp.f32 s1, #0
; CHECK-LE-NEXT: bfi r1, r3, #0, #1
; CHECK-LE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-LE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-LE-NEXT: mov.w r3, #0 ; CHECK-LE-NEXT: bfi r1, r2, #0, #1
; CHECK-LE-NEXT: vcmp.f32 s2, #0 ; CHECK-LE-NEXT: vcmp.f32 s2, #0
; CHECK-LE-NEXT: it gt ; CHECK-LE-NEXT: cset r2, gt
; CHECK-LE-NEXT: movgt r3, #1 ; CHECK-LE-NEXT: cmp r2, #0
; CHECK-LE-NEXT: cmp r3, #0 ; CHECK-LE-NEXT: csetm r2, ne
; CHECK-LE-NEXT: csetm r3, ne
; CHECK-LE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-LE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-LE-NEXT: bfi r1, r3, #1, #1 ; CHECK-LE-NEXT: bfi r1, r2, #1, #1
; CHECK-LE-NEXT: mov.w r3, #0
; CHECK-LE-NEXT: it gt
; CHECK-LE-NEXT: movgt r3, #1
; CHECK-LE-NEXT: cmp r3, #0
; CHECK-LE-NEXT: vcmp.f32 s3, #0 ; CHECK-LE-NEXT: vcmp.f32 s3, #0
; CHECK-LE-NEXT: csetm r3, ne ; CHECK-LE-NEXT: cset r2, gt
; CHECK-LE-NEXT: movs r2, #0 ; CHECK-LE-NEXT: cmp r2, #0
; CHECK-LE-NEXT: csetm r2, ne
; CHECK-LE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-LE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-LE-NEXT: it gt ; CHECK-LE-NEXT: bfi r1, r2, #2, #1
; CHECK-LE-NEXT: movgt r2, #1 ; CHECK-LE-NEXT: cset r2, gt
; CHECK-LE-NEXT: cmp r2, #0 ; CHECK-LE-NEXT: cmp r2, #0
; CHECK-LE-NEXT: bfi r1, r3, #2, #1
; CHECK-LE-NEXT: csetm r2, ne ; CHECK-LE-NEXT: csetm r2, ne
; CHECK-LE-NEXT: bfi r1, r2, #3, #1 ; CHECK-LE-NEXT: bfi r1, r2, #3, #1
; CHECK-LE-NEXT: vcvtb.f16.f32 s4, s0
; CHECK-LE-NEXT: vcvtt.f16.f32 s4, s1
; CHECK-LE-NEXT: vcvtb.f16.f32 s6, s2
; CHECK-LE-NEXT: vcvtt.f16.f32 s5, s3
; CHECK-LE-NEXT: lsls r2, r1, #31 ; CHECK-LE-NEXT: lsls r2, r1, #31
; CHECK-LE-NEXT: bne .LBB25_5 ; CHECK-LE-NEXT: bne .LBB25_5
; CHECK-LE-NEXT: @ %bb.1: @ %else ; CHECK-LE-NEXT: @ %bb.1: @ %else
; CHECK-LE-NEXT: lsls r2, r1, #30 ; CHECK-LE-NEXT: lsls r2, r1, #30
; CHECK-LE-NEXT: bmi .LBB25_6 ; CHECK-LE-NEXT: bmi .LBB25_6
; CHECK-LE-NEXT: .LBB25_2: @ %else2 ; CHECK-LE-NEXT: .LBB25_2: @ %else2
; CHECK-LE-NEXT: lsls r2, r1, #29 ; CHECK-LE-NEXT: lsls r2, r1, #29
; CHECK-LE-NEXT: bmi .LBB25_7 ; CHECK-LE-NEXT: bmi .LBB25_7
Show All 22 Lines
; CHECK-LE-NEXT: add sp, #4 ; CHECK-LE-NEXT: add sp, #4
; CHECK-LE-NEXT: bx lr ; CHECK-LE-NEXT: bx lr
; ;
; CHECK-BE-LABEL: masked_v4f16_align4: ; CHECK-BE-LABEL: masked_v4f16_align4:
; CHECK-BE: @ %bb.0: @ %entry ; CHECK-BE: @ %bb.0: @ %entry
; CHECK-BE-NEXT: .pad #4 ; CHECK-BE-NEXT: .pad #4
; CHECK-BE-NEXT: sub sp, #4 ; CHECK-BE-NEXT: sub sp, #4
; CHECK-BE-NEXT: vrev64.32 q1, q0 ; CHECK-BE-NEXT: vrev64.32 q1, q0
; CHECK-BE-NEXT: movs r1, #0
; CHECK-BE-NEXT: vcmp.f32 s7, #0 ; CHECK-BE-NEXT: vcmp.f32 s7, #0
; CHECK-BE-NEXT: movs r2, #0 ; CHECK-BE-NEXT: vcvtb.f16.f32 s0, s4
; CHECK-BE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-BE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-BE-NEXT: it gt ; CHECK-BE-NEXT: vcmp.f32 s6, #0
; CHECK-BE-NEXT: movgt r1, #1 ; CHECK-BE-NEXT: vcvtt.f16.f32 s0, s5
; CHECK-BE-NEXT: vcvtb.f16.f32 s2, s6
; CHECK-BE-NEXT: vcvtt.f16.f32 s1, s7
; CHECK-BE-NEXT: cset r1, gt
; CHECK-BE-NEXT: cmp r1, #0 ; CHECK-BE-NEXT: cmp r1, #0
; CHECK-BE-NEXT: mov.w r1, #0 ; CHECK-BE-NEXT: mov.w r1, #0
; CHECK-BE-NEXT: csetm r3, ne ; CHECK-BE-NEXT: csetm r2, ne
; CHECK-BE-NEXT: vcmp.f32 s6, #0
; CHECK-BE-NEXT: bfi r1, r3, #0, #1
; CHECK-BE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-BE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-BE-NEXT: mov.w r3, #0 ; CHECK-BE-NEXT: bfi r1, r2, #0, #1
; CHECK-BE-NEXT: vcmp.f32 s5, #0 ; CHECK-BE-NEXT: vcmp.f32 s5, #0
; CHECK-BE-NEXT: it gt ; CHECK-BE-NEXT: cset r2, gt
; CHECK-BE-NEXT: movgt r3, #1 ; CHECK-BE-NEXT: cmp r2, #0
; CHECK-BE-NEXT: cmp r3, #0 ; CHECK-BE-NEXT: csetm r2, ne
; CHECK-BE-NEXT: csetm r3, ne
; CHECK-BE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-BE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-BE-NEXT: bfi r1, r3, #1, #1 ; CHECK-BE-NEXT: bfi r1, r2, #1, #1
; CHECK-BE-NEXT: mov.w r3, #0
; CHECK-BE-NEXT: it gt
; CHECK-BE-NEXT: movgt r3, #1
; CHECK-BE-NEXT: cmp r3, #0
; CHECK-BE-NEXT: vcmp.f32 s4, #0 ; CHECK-BE-NEXT: vcmp.f32 s4, #0
; CHECK-BE-NEXT: csetm r3, ne ; CHECK-BE-NEXT: cset r2, gt
; CHECK-BE-NEXT: cmp r2, #0
; CHECK-BE-NEXT: csetm r2, ne
; CHECK-BE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-BE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-BE-NEXT: it gt ; CHECK-BE-NEXT: bfi r1, r2, #2, #1
; CHECK-BE-NEXT: movgt r2, #1 ; CHECK-BE-NEXT: cset r2, gt
; CHECK-BE-NEXT: cmp r2, #0 ; CHECK-BE-NEXT: cmp r2, #0
; CHECK-BE-NEXT: bfi r1, r3, #2, #1
; CHECK-BE-NEXT: csetm r2, ne ; CHECK-BE-NEXT: csetm r2, ne
; CHECK-BE-NEXT: vcvtb.f16.f32 s0, s4
; CHECK-BE-NEXT: bfi r1, r2, #3, #1 ; CHECK-BE-NEXT: bfi r1, r2, #3, #1
; CHECK-BE-NEXT: vcvtt.f16.f32 s0, s5
; CHECK-BE-NEXT: vcvtb.f16.f32 s2, s6
; CHECK-BE-NEXT: vcvtt.f16.f32 s1, s7
; CHECK-BE-NEXT: lsls r2, r1, #28 ; CHECK-BE-NEXT: lsls r2, r1, #28
; CHECK-BE-NEXT: bmi .LBB25_5 ; CHECK-BE-NEXT: bmi .LBB25_5
; CHECK-BE-NEXT: @ %bb.1: @ %else ; CHECK-BE-NEXT: @ %bb.1: @ %else
; CHECK-BE-NEXT: lsls r2, r1, #29 ; CHECK-BE-NEXT: lsls r2, r1, #29
; CHECK-BE-NEXT: bmi .LBB25_6 ; CHECK-BE-NEXT: bmi .LBB25_6
; CHECK-BE-NEXT: .LBB25_2: @ %else2 ; CHECK-BE-NEXT: .LBB25_2: @ %else2
; CHECK-BE-NEXT: lsls r2, r1, #30 ; CHECK-BE-NEXT: lsls r2, r1, #30
; CHECK-BE-NEXT: bmi .LBB25_7 ; CHECK-BE-NEXT: bmi .LBB25_7
Show All 29 Lines
} }
define arm_aapcs_vfpcc void @masked_v4f16_align2(<4 x half> *%dest, <4 x float> %a) { define arm_aapcs_vfpcc void @masked_v4f16_align2(<4 x half> *%dest, <4 x float> %a) {
; CHECK-LE-LABEL: masked_v4f16_align2: ; CHECK-LE-LABEL: masked_v4f16_align2:
; CHECK-LE: @ %bb.0: @ %entry ; CHECK-LE: @ %bb.0: @ %entry
; CHECK-LE-NEXT: .pad #4 ; CHECK-LE-NEXT: .pad #4
; CHECK-LE-NEXT: sub sp, #4 ; CHECK-LE-NEXT: sub sp, #4
; CHECK-LE-NEXT: vcmp.f32 s0, #0 ; CHECK-LE-NEXT: vcmp.f32 s0, #0
; CHECK-LE-NEXT: movs r1, #0 ; CHECK-LE-NEXT: vcvtb.f16.f32 s4, s0
; CHECK-LE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-LE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-LE-NEXT: it gt ; CHECK-LE-NEXT: vcmp.f32 s1, #0
; CHECK-LE-NEXT: movgt r1, #1 ; CHECK-LE-NEXT: vcvtt.f16.f32 s4, s1
; CHECK-LE-NEXT: vcvtb.f16.f32 s6, s2
; CHECK-LE-NEXT: vcvtt.f16.f32 s5, s3
; CHECK-LE-NEXT: cset r1, gt
; CHECK-LE-NEXT: cmp r1, #0 ; CHECK-LE-NEXT: cmp r1, #0
; CHECK-LE-NEXT: mov.w r1, #0 ; CHECK-LE-NEXT: mov.w r1, #0
; CHECK-LE-NEXT: csetm r3, ne ; CHECK-LE-NEXT: csetm r2, ne
; CHECK-LE-NEXT: vcmp.f32 s1, #0
; CHECK-LE-NEXT: bfi r1, r3, #0, #1
; CHECK-LE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-LE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-LE-NEXT: mov.w r3, #0 ; CHECK-LE-NEXT: bfi r1, r2, #0, #1
; CHECK-LE-NEXT: vcmp.f32 s2, #0 ; CHECK-LE-NEXT: vcmp.f32 s2, #0
; CHECK-LE-NEXT: it gt ; CHECK-LE-NEXT: cset r2, gt
; CHECK-LE-NEXT: movgt r3, #1 ; CHECK-LE-NEXT: cmp r2, #0
; CHECK-LE-NEXT: cmp r3, #0 ; CHECK-LE-NEXT: csetm r2, ne
; CHECK-LE-NEXT: csetm r3, ne
; CHECK-LE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-LE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-LE-NEXT: bfi r1, r3, #1, #1 ; CHECK-LE-NEXT: bfi r1, r2, #1, #1
; CHECK-LE-NEXT: mov.w r3, #0
; CHECK-LE-NEXT: it gt
; CHECK-LE-NEXT: movgt r3, #1
; CHECK-LE-NEXT: cmp r3, #0
; CHECK-LE-NEXT: vcmp.f32 s3, #0 ; CHECK-LE-NEXT: vcmp.f32 s3, #0
; CHECK-LE-NEXT: csetm r3, ne ; CHECK-LE-NEXT: cset r2, gt
; CHECK-LE-NEXT: movs r2, #0 ; CHECK-LE-NEXT: cmp r2, #0
; CHECK-LE-NEXT: csetm r2, ne
; CHECK-LE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-LE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-LE-NEXT: it gt ; CHECK-LE-NEXT: bfi r1, r2, #2, #1
; CHECK-LE-NEXT: movgt r2, #1 ; CHECK-LE-NEXT: cset r2, gt
; CHECK-LE-NEXT: cmp r2, #0 ; CHECK-LE-NEXT: cmp r2, #0
; CHECK-LE-NEXT: bfi r1, r3, #2, #1
; CHECK-LE-NEXT: csetm r2, ne ; CHECK-LE-NEXT: csetm r2, ne
; CHECK-LE-NEXT: bfi r1, r2, #3, #1 ; CHECK-LE-NEXT: bfi r1, r2, #3, #1
; CHECK-LE-NEXT: vcvtb.f16.f32 s4, s0
; CHECK-LE-NEXT: vcvtt.f16.f32 s4, s1
; CHECK-LE-NEXT: vcvtb.f16.f32 s6, s2
; CHECK-LE-NEXT: vcvtt.f16.f32 s5, s3
; CHECK-LE-NEXT: lsls r2, r1, #31 ; CHECK-LE-NEXT: lsls r2, r1, #31
; CHECK-LE-NEXT: bne .LBB26_5 ; CHECK-LE-NEXT: bne .LBB26_5
; CHECK-LE-NEXT: @ %bb.1: @ %else ; CHECK-LE-NEXT: @ %bb.1: @ %else
; CHECK-LE-NEXT: lsls r2, r1, #30 ; CHECK-LE-NEXT: lsls r2, r1, #30
; CHECK-LE-NEXT: bmi .LBB26_6 ; CHECK-LE-NEXT: bmi .LBB26_6
; CHECK-LE-NEXT: .LBB26_2: @ %else2 ; CHECK-LE-NEXT: .LBB26_2: @ %else2
; CHECK-LE-NEXT: lsls r2, r1, #29 ; CHECK-LE-NEXT: lsls r2, r1, #29
; CHECK-LE-NEXT: bmi .LBB26_7 ; CHECK-LE-NEXT: bmi .LBB26_7
Show All 22 Lines
; CHECK-LE-NEXT: add sp, #4 ; CHECK-LE-NEXT: add sp, #4
; CHECK-LE-NEXT: bx lr ; CHECK-LE-NEXT: bx lr
; ;
; CHECK-BE-LABEL: masked_v4f16_align2: ; CHECK-BE-LABEL: masked_v4f16_align2:
; CHECK-BE: @ %bb.0: @ %entry ; CHECK-BE: @ %bb.0: @ %entry
; CHECK-BE-NEXT: .pad #4 ; CHECK-BE-NEXT: .pad #4
; CHECK-BE-NEXT: sub sp, #4 ; CHECK-BE-NEXT: sub sp, #4
; CHECK-BE-NEXT: vrev64.32 q1, q0 ; CHECK-BE-NEXT: vrev64.32 q1, q0
; CHECK-BE-NEXT: movs r1, #0
; CHECK-BE-NEXT: vcmp.f32 s7, #0 ; CHECK-BE-NEXT: vcmp.f32 s7, #0
; CHECK-BE-NEXT: movs r2, #0 ; CHECK-BE-NEXT: vcvtb.f16.f32 s0, s4
; CHECK-BE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-BE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-BE-NEXT: it gt ; CHECK-BE-NEXT: vcmp.f32 s6, #0
; CHECK-BE-NEXT: movgt r1, #1 ; CHECK-BE-NEXT: vcvtt.f16.f32 s0, s5
; CHECK-BE-NEXT: vcvtb.f16.f32 s2, s6
; CHECK-BE-NEXT: vcvtt.f16.f32 s1, s7
; CHECK-BE-NEXT: cset r1, gt
; CHECK-BE-NEXT: cmp r1, #0 ; CHECK-BE-NEXT: cmp r1, #0
; CHECK-BE-NEXT: mov.w r1, #0 ; CHECK-BE-NEXT: mov.w r1, #0
; CHECK-BE-NEXT: csetm r3, ne ; CHECK-BE-NEXT: csetm r2, ne
; CHECK-BE-NEXT: vcmp.f32 s6, #0
; CHECK-BE-NEXT: bfi r1, r3, #0, #1
; CHECK-BE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-BE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-BE-NEXT: mov.w r3, #0 ; CHECK-BE-NEXT: bfi r1, r2, #0, #1
; CHECK-BE-NEXT: vcmp.f32 s5, #0 ; CHECK-BE-NEXT: vcmp.f32 s5, #0
; CHECK-BE-NEXT: it gt ; CHECK-BE-NEXT: cset r2, gt
; CHECK-BE-NEXT: movgt r3, #1 ; CHECK-BE-NEXT: cmp r2, #0
; CHECK-BE-NEXT: cmp r3, #0 ; CHECK-BE-NEXT: csetm r2, ne
; CHECK-BE-NEXT: csetm r3, ne
; CHECK-BE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-BE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-BE-NEXT: bfi r1, r3, #1, #1 ; CHECK-BE-NEXT: bfi r1, r2, #1, #1
; CHECK-BE-NEXT: mov.w r3, #0
; CHECK-BE-NEXT: it gt
; CHECK-BE-NEXT: movgt r3, #1
; CHECK-BE-NEXT: cmp r3, #0
; CHECK-BE-NEXT: vcmp.f32 s4, #0 ; CHECK-BE-NEXT: vcmp.f32 s4, #0
; CHECK-BE-NEXT: csetm r3, ne ; CHECK-BE-NEXT: cset r2, gt
; CHECK-BE-NEXT: cmp r2, #0
; CHECK-BE-NEXT: csetm r2, ne
; CHECK-BE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-BE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-BE-NEXT: it gt ; CHECK-BE-NEXT: bfi r1, r2, #2, #1
; CHECK-BE-NEXT: movgt r2, #1 ; CHECK-BE-NEXT: cset r2, gt
; CHECK-BE-NEXT: cmp r2, #0 ; CHECK-BE-NEXT: cmp r2, #0
; CHECK-BE-NEXT: bfi r1, r3, #2, #1
; CHECK-BE-NEXT: csetm r2, ne ; CHECK-BE-NEXT: csetm r2, ne
; CHECK-BE-NEXT: vcvtb.f16.f32 s0, s4
; CHECK-BE-NEXT: bfi r1, r2, #3, #1 ; CHECK-BE-NEXT: bfi r1, r2, #3, #1
; CHECK-BE-NEXT: vcvtt.f16.f32 s0, s5
; CHECK-BE-NEXT: vcvtb.f16.f32 s2, s6
; CHECK-BE-NEXT: vcvtt.f16.f32 s1, s7
; CHECK-BE-NEXT: lsls r2, r1, #28 ; CHECK-BE-NEXT: lsls r2, r1, #28
; CHECK-BE-NEXT: bmi .LBB26_5 ; CHECK-BE-NEXT: bmi .LBB26_5
; CHECK-BE-NEXT: @ %bb.1: @ %else ; CHECK-BE-NEXT: @ %bb.1: @ %else
; CHECK-BE-NEXT: lsls r2, r1, #29 ; CHECK-BE-NEXT: lsls r2, r1, #29
; CHECK-BE-NEXT: bmi .LBB26_6 ; CHECK-BE-NEXT: bmi .LBB26_6
; CHECK-BE-NEXT: .LBB26_2: @ %else2 ; CHECK-BE-NEXT: .LBB26_2: @ %else2
; CHECK-BE-NEXT: lsls r2, r1, #30 ; CHECK-BE-NEXT: lsls r2, r1, #30
; CHECK-BE-NEXT: bmi .LBB26_7 ; CHECK-BE-NEXT: bmi .LBB26_7
Show All 29 Lines
} }
define arm_aapcs_vfpcc void @masked_v4f16_align1(<4 x half> *%dest, <4 x float> %a) { define arm_aapcs_vfpcc void @masked_v4f16_align1(<4 x half> *%dest, <4 x float> %a) {
; CHECK-LE-LABEL: masked_v4f16_align1: ; CHECK-LE-LABEL: masked_v4f16_align1:
; CHECK-LE: @ %bb.0: @ %entry ; CHECK-LE: @ %bb.0: @ %entry
; CHECK-LE-NEXT: .pad #20 ; CHECK-LE-NEXT: .pad #20
; CHECK-LE-NEXT: sub sp, #20 ; CHECK-LE-NEXT: sub sp, #20
; CHECK-LE-NEXT: vcmp.f32 s0, #0 ; CHECK-LE-NEXT: vcmp.f32 s0, #0
; CHECK-LE-NEXT: movs r1, #0 ; CHECK-LE-NEXT: vcvtb.f16.f32 s4, s0
; CHECK-LE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-LE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-LE-NEXT: it gt ; CHECK-LE-NEXT: vcmp.f32 s1, #0
; CHECK-LE-NEXT: movgt r1, #1 ; CHECK-LE-NEXT: vcvtt.f16.f32 s4, s1
; CHECK-LE-NEXT: vcvtb.f16.f32 s6, s2
; CHECK-LE-NEXT: vcvtt.f16.f32 s5, s3
; CHECK-LE-NEXT: cset r1, gt
; CHECK-LE-NEXT: cmp r1, #0 ; CHECK-LE-NEXT: cmp r1, #0
; CHECK-LE-NEXT: mov.w r1, #0 ; CHECK-LE-NEXT: mov.w r1, #0
; CHECK-LE-NEXT: csetm r3, ne ; CHECK-LE-NEXT: csetm r2, ne
; CHECK-LE-NEXT: vcmp.f32 s1, #0
; CHECK-LE-NEXT: bfi r1, r3, #0, #1
; CHECK-LE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-LE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-LE-NEXT: mov.w r3, #0 ; CHECK-LE-NEXT: bfi r1, r2, #0, #1
; CHECK-LE-NEXT: vcmp.f32 s2, #0 ; CHECK-LE-NEXT: vcmp.f32 s2, #0
; CHECK-LE-NEXT: it gt ; CHECK-LE-NEXT: cset r2, gt
; CHECK-LE-NEXT: movgt r3, #1 ; CHECK-LE-NEXT: cmp r2, #0
; CHECK-LE-NEXT: cmp r3, #0 ; CHECK-LE-NEXT: csetm r2, ne
; CHECK-LE-NEXT: csetm r3, ne
; CHECK-LE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-LE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-LE-NEXT: bfi r1, r3, #1, #1 ; CHECK-LE-NEXT: bfi r1, r2, #1, #1
; CHECK-LE-NEXT: mov.w r3, #0
; CHECK-LE-NEXT: it gt
; CHECK-LE-NEXT: movgt r3, #1
; CHECK-LE-NEXT: cmp r3, #0
; CHECK-LE-NEXT: vcmp.f32 s3, #0 ; CHECK-LE-NEXT: vcmp.f32 s3, #0
; CHECK-LE-NEXT: csetm r3, ne ; CHECK-LE-NEXT: cset r2, gt
; CHECK-LE-NEXT: movs r2, #0 ; CHECK-LE-NEXT: cmp r2, #0
; CHECK-LE-NEXT: csetm r2, ne
; CHECK-LE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-LE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-LE-NEXT: it gt ; CHECK-LE-NEXT: bfi r1, r2, #2, #1
; CHECK-LE-NEXT: movgt r2, #1 ; CHECK-LE-NEXT: cset r2, gt
; CHECK-LE-NEXT: cmp r2, #0 ; CHECK-LE-NEXT: cmp r2, #0
; CHECK-LE-NEXT: bfi r1, r3, #2, #1
; CHECK-LE-NEXT: csetm r2, ne ; CHECK-LE-NEXT: csetm r2, ne
; CHECK-LE-NEXT: bfi r1, r2, #3, #1 ; CHECK-LE-NEXT: bfi r1, r2, #3, #1
; CHECK-LE-NEXT: vcvtb.f16.f32 s4, s0
; CHECK-LE-NEXT: vcvtt.f16.f32 s4, s1
; CHECK-LE-NEXT: vcvtb.f16.f32 s6, s2
; CHECK-LE-NEXT: vcvtt.f16.f32 s5, s3
; CHECK-LE-NEXT: lsls r2, r1, #31 ; CHECK-LE-NEXT: lsls r2, r1, #31
; CHECK-LE-NEXT: bne .LBB27_5 ; CHECK-LE-NEXT: bne .LBB27_5
; CHECK-LE-NEXT: @ %bb.1: @ %else ; CHECK-LE-NEXT: @ %bb.1: @ %else
; CHECK-LE-NEXT: lsls r2, r1, #30 ; CHECK-LE-NEXT: lsls r2, r1, #30
; CHECK-LE-NEXT: bmi .LBB27_6 ; CHECK-LE-NEXT: bmi .LBB27_6
; CHECK-LE-NEXT: .LBB27_2: @ %else2 ; CHECK-LE-NEXT: .LBB27_2: @ %else2
; CHECK-LE-NEXT: lsls r2, r1, #29 ; CHECK-LE-NEXT: lsls r2, r1, #29
; CHECK-LE-NEXT: bmi .LBB27_7 ; CHECK-LE-NEXT: bmi .LBB27_7
Show All 30 Lines
; CHECK-LE-NEXT: add sp, #20 ; CHECK-LE-NEXT: add sp, #20
; CHECK-LE-NEXT: bx lr ; CHECK-LE-NEXT: bx lr
; ;
; CHECK-BE-LABEL: masked_v4f16_align1: ; CHECK-BE-LABEL: masked_v4f16_align1:
; CHECK-BE: @ %bb.0: @ %entry ; CHECK-BE: @ %bb.0: @ %entry
; CHECK-BE-NEXT: .pad #20 ; CHECK-BE-NEXT: .pad #20
; CHECK-BE-NEXT: sub sp, #20 ; CHECK-BE-NEXT: sub sp, #20
; CHECK-BE-NEXT: vrev64.32 q1, q0 ; CHECK-BE-NEXT: vrev64.32 q1, q0
; CHECK-BE-NEXT: movs r1, #0
; CHECK-BE-NEXT: vcmp.f32 s7, #0 ; CHECK-BE-NEXT: vcmp.f32 s7, #0
; CHECK-BE-NEXT: movs r2, #0 ; CHECK-BE-NEXT: vcvtb.f16.f32 s0, s4
; CHECK-BE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-BE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-BE-NEXT: it gt ; CHECK-BE-NEXT: vcmp.f32 s6, #0
; CHECK-BE-NEXT: movgt r1, #1 ; CHECK-BE-NEXT: vcvtt.f16.f32 s0, s5
; CHECK-BE-NEXT: vcvtb.f16.f32 s2, s6
; CHECK-BE-NEXT: vcvtt.f16.f32 s1, s7
; CHECK-BE-NEXT: cset r1, gt
; CHECK-BE-NEXT: cmp r1, #0 ; CHECK-BE-NEXT: cmp r1, #0
; CHECK-BE-NEXT: mov.w r1, #0 ; CHECK-BE-NEXT: mov.w r1, #0
; CHECK-BE-NEXT: csetm r3, ne ; CHECK-BE-NEXT: csetm r2, ne
; CHECK-BE-NEXT: vcmp.f32 s6, #0
; CHECK-BE-NEXT: bfi r1, r3, #0, #1
; CHECK-BE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-BE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-BE-NEXT: mov.w r3, #0 ; CHECK-BE-NEXT: bfi r1, r2, #0, #1
; CHECK-BE-NEXT: vcmp.f32 s5, #0 ; CHECK-BE-NEXT: vcmp.f32 s5, #0
; CHECK-BE-NEXT: it gt ; CHECK-BE-NEXT: cset r2, gt
; CHECK-BE-NEXT: movgt r3, #1 ; CHECK-BE-NEXT: cmp r2, #0
; CHECK-BE-NEXT: cmp r3, #0 ; CHECK-BE-NEXT: csetm r2, ne
; CHECK-BE-NEXT: csetm r3, ne
; CHECK-BE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-BE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-BE-NEXT: bfi r1, r3, #1, #1 ; CHECK-BE-NEXT: bfi r1, r2, #1, #1
; CHECK-BE-NEXT: mov.w r3, #0
; CHECK-BE-NEXT: it gt
; CHECK-BE-NEXT: movgt r3, #1
; CHECK-BE-NEXT: cmp r3, #0
; CHECK-BE-NEXT: vcmp.f32 s4, #0 ; CHECK-BE-NEXT: vcmp.f32 s4, #0
; CHECK-BE-NEXT: csetm r3, ne ; CHECK-BE-NEXT: cset r2, gt
; CHECK-BE-NEXT: cmp r2, #0
; CHECK-BE-NEXT: csetm r2, ne
; CHECK-BE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-BE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-BE-NEXT: it gt ; CHECK-BE-NEXT: bfi r1, r2, #2, #1
; CHECK-BE-NEXT: movgt r2, #1 ; CHECK-BE-NEXT: cset r2, gt
; CHECK-BE-NEXT: cmp r2, #0 ; CHECK-BE-NEXT: cmp r2, #0
; CHECK-BE-NEXT: bfi r1, r3, #2, #1
; CHECK-BE-NEXT: csetm r2, ne ; CHECK-BE-NEXT: csetm r2, ne
; CHECK-BE-NEXT: vcvtb.f16.f32 s0, s4
; CHECK-BE-NEXT: bfi r1, r2, #3, #1 ; CHECK-BE-NEXT: bfi r1, r2, #3, #1
; CHECK-BE-NEXT: vcvtt.f16.f32 s0, s5
; CHECK-BE-NEXT: vcvtb.f16.f32 s2, s6
; CHECK-BE-NEXT: vcvtt.f16.f32 s1, s7
; CHECK-BE-NEXT: lsls r2, r1, #28 ; CHECK-BE-NEXT: lsls r2, r1, #28
; CHECK-BE-NEXT: bmi .LBB27_5 ; CHECK-BE-NEXT: bmi .LBB27_5
; CHECK-BE-NEXT: @ %bb.1: @ %else ; CHECK-BE-NEXT: @ %bb.1: @ %else
; CHECK-BE-NEXT: lsls r2, r1, #29 ; CHECK-BE-NEXT: lsls r2, r1, #29
; CHECK-BE-NEXT: bmi .LBB27_6 ; CHECK-BE-NEXT: bmi .LBB27_6
; CHECK-BE-NEXT: .LBB27_2: @ %else2 ; CHECK-BE-NEXT: .LBB27_2: @ %else2
; CHECK-BE-NEXT: lsls r2, r1, #30 ; CHECK-BE-NEXT: lsls r2, r1, #30
; CHECK-BE-NEXT: bmi .LBB27_7 ; CHECK-BE-NEXT: bmi .LBB27_7
▲ Show 20 Lines • Show All 50 Lines • Show Last 20 Lines

llvm/test/CodeGen/Thumb2/mve-minmax.ll

Show All 32 Linesentry:
%0 = icmp slt <4 x i32> %s1, %s2 %0 = icmp slt <4 x i32> %s1, %s2
%1 = select <4 x i1> %0, <4 x i32> %s1, <4 x i32> %s2 %1 = select <4 x i1> %0, <4 x i32> %s1, <4 x i32> %s2
ret <4 x i32> %1 ret <4 x i32> %1
} }
define arm_aapcs_vfpcc <2 x i64> @smin_v2i64(<2 x i64> %s1, <2 x i64> %s2) { define arm_aapcs_vfpcc <2 x i64> @smin_v2i64(<2 x i64> %s1, <2 x i64> %s2) {
; CHECK-LABEL: smin_v2i64: ; CHECK-LABEL: smin_v2i64:
; CHECK: @ %bb.0: @ %entry ; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: .save {r4, r5, r6, lr} ; CHECK-NEXT: .save {r4, r5, r7, lr}
; CHECK-NEXT: push {r4, r5, r6, lr} ; CHECK-NEXT: push {r4, r5, r7, lr}
; CHECK-NEXT: vmov r0, r1, d3 ; CHECK-NEXT: vmov r0, r1, d3
; CHECK-NEXT: movs r6, #0
; CHECK-NEXT: vmov r2, r3, d1 ; CHECK-NEXT: vmov r2, r3, d1
; CHECK-NEXT: vmov r12, lr, d2 ; CHECK-NEXT: vmov r12, lr, d2
; CHECK-NEXT: vmov r4, r5, d0 ; CHECK-NEXT: vmov r4, r5, d0
; CHECK-NEXT: subs r0, r2, r0 ; CHECK-NEXT: subs r0, r2, r0
; CHECK-NEXT: sbcs.w r0, r3, r1 ; CHECK-NEXT: sbcs.w r0, r3, r1
; CHECK-NEXT: mov.w r0, #0 ; CHECK-NEXT: cset r0, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r0, #1
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: subs.w r1, r4, r12 ; CHECK-NEXT: subs.w r1, r4, r12
; CHECK-NEXT: sbcs.w r1, r5, lr ; CHECK-NEXT: sbcs.w r1, r5, lr
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r1, lt
; CHECK-NEXT: movlt r6, #1 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: vmov q2[2], q2[0], r1, r0 ; CHECK-NEXT: vmov q2[2], q2[0], r1, r0
; CHECK-NEXT: vmov q2[3], q2[1], r1, r0 ; CHECK-NEXT: vmov q2[3], q2[1], r1, r0
; CHECK-NEXT: vbic q1, q1, q2 ; CHECK-NEXT: vbic q1, q1, q2
; CHECK-NEXT: vand q0, q0, q2 ; CHECK-NEXT: vand q0, q0, q2
; CHECK-NEXT: vorr q0, q0, q1 ; CHECK-NEXT: vorr q0, q0, q1
; CHECK-NEXT: pop {r4, r5, r6, pc} ; CHECK-NEXT: pop {r4, r5, r7, pc}
entry: entry:
%0 = icmp slt <2 x i64> %s1, %s2 %0 = icmp slt <2 x i64> %s1, %s2
%1 = select <2 x i1> %0, <2 x i64> %s1, <2 x i64> %s2 %1 = select <2 x i1> %0, <2 x i64> %s1, <2 x i64> %s2
ret <2 x i64> %1 ret <2 x i64> %1
} }
define arm_aapcs_vfpcc <16 x i8> @umin_v16i8(<16 x i8> %s1, <16 x i8> %s2) { define arm_aapcs_vfpcc <16 x i8> @umin_v16i8(<16 x i8> %s1, <16 x i8> %s2) {
; CHECK-LABEL: umin_v16i8: ; CHECK-LABEL: umin_v16i8:
Show All 26 Linesentry:
%0 = icmp ult <4 x i32> %s1, %s2 %0 = icmp ult <4 x i32> %s1, %s2
%1 = select <4 x i1> %0, <4 x i32> %s1, <4 x i32> %s2 %1 = select <4 x i1> %0, <4 x i32> %s1, <4 x i32> %s2
ret <4 x i32> %1 ret <4 x i32> %1
} }
define arm_aapcs_vfpcc <2 x i64> @umin_v2i64(<2 x i64> %s1, <2 x i64> %s2) { define arm_aapcs_vfpcc <2 x i64> @umin_v2i64(<2 x i64> %s1, <2 x i64> %s2) {
; CHECK-LABEL: umin_v2i64: ; CHECK-LABEL: umin_v2i64:
; CHECK: @ %bb.0: @ %entry ; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: .save {r4, r5, r6, lr} ; CHECK-NEXT: .save {r4, r5, r7, lr}
; CHECK-NEXT: push {r4, r5, r6, lr} ; CHECK-NEXT: push {r4, r5, r7, lr}
; CHECK-NEXT: vmov r0, r1, d3 ; CHECK-NEXT: vmov r0, r1, d3
; CHECK-NEXT: movs r6, #0
; CHECK-NEXT: vmov r2, r3, d1 ; CHECK-NEXT: vmov r2, r3, d1
; CHECK-NEXT: vmov r12, lr, d2 ; CHECK-NEXT: vmov r12, lr, d2
; CHECK-NEXT: vmov r4, r5, d0 ; CHECK-NEXT: vmov r4, r5, d0
; CHECK-NEXT: subs r0, r2, r0 ; CHECK-NEXT: subs r0, r2, r0
; CHECK-NEXT: sbcs.w r0, r3, r1 ; CHECK-NEXT: sbcs.w r0, r3, r1
; CHECK-NEXT: mov.w r0, #0 ; CHECK-NEXT: cset r0, lo
; CHECK-NEXT: it lo
; CHECK-NEXT: movlo r0, #1
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: subs.w r1, r4, r12 ; CHECK-NEXT: subs.w r1, r4, r12
; CHECK-NEXT: sbcs.w r1, r5, lr ; CHECK-NEXT: sbcs.w r1, r5, lr
; CHECK-NEXT: it lo ; CHECK-NEXT: cset r1, lo
; CHECK-NEXT: movlo r6, #1 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: vmov q2[2], q2[0], r1, r0 ; CHECK-NEXT: vmov q2[2], q2[0], r1, r0
; CHECK-NEXT: vmov q2[3], q2[1], r1, r0 ; CHECK-NEXT: vmov q2[3], q2[1], r1, r0
; CHECK-NEXT: vbic q1, q1, q2 ; CHECK-NEXT: vbic q1, q1, q2
; CHECK-NEXT: vand q0, q0, q2 ; CHECK-NEXT: vand q0, q0, q2
; CHECK-NEXT: vorr q0, q0, q1 ; CHECK-NEXT: vorr q0, q0, q1
; CHECK-NEXT: pop {r4, r5, r6, pc} ; CHECK-NEXT: pop {r4, r5, r7, pc}
entry: entry:
%0 = icmp ult <2 x i64> %s1, %s2 %0 = icmp ult <2 x i64> %s1, %s2
%1 = select <2 x i1> %0, <2 x i64> %s1, <2 x i64> %s2 %1 = select <2 x i1> %0, <2 x i64> %s1, <2 x i64> %s2
ret <2 x i64> %1 ret <2 x i64> %1
} }
define arm_aapcs_vfpcc <16 x i8> @smax_v16i8(<16 x i8> %s1, <16 x i8> %s2) { define arm_aapcs_vfpcc <16 x i8> @smax_v16i8(<16 x i8> %s1, <16 x i8> %s2) {
Show All 27 Linesentry:
%0 = icmp sgt <4 x i32> %s1, %s2 %0 = icmp sgt <4 x i32> %s1, %s2
%1 = select <4 x i1> %0, <4 x i32> %s1, <4 x i32> %s2 %1 = select <4 x i1> %0, <4 x i32> %s1, <4 x i32> %s2
ret <4 x i32> %1 ret <4 x i32> %1
} }
define arm_aapcs_vfpcc <2 x i64> @smax_v2i64(<2 x i64> %s1, <2 x i64> %s2) { define arm_aapcs_vfpcc <2 x i64> @smax_v2i64(<2 x i64> %s1, <2 x i64> %s2) {
; CHECK-LABEL: smax_v2i64: ; CHECK-LABEL: smax_v2i64:
; CHECK: @ %bb.0: @ %entry ; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: .save {r4, r5, r6, lr} ; CHECK-NEXT: .save {r4, r5, r7, lr}
; CHECK-NEXT: push {r4, r5, r6, lr} ; CHECK-NEXT: push {r4, r5, r7, lr}
; CHECK-NEXT: vmov r0, r1, d1 ; CHECK-NEXT: vmov r0, r1, d1
; CHECK-NEXT: movs r6, #0
; CHECK-NEXT: vmov r2, r3, d3 ; CHECK-NEXT: vmov r2, r3, d3
; CHECK-NEXT: vmov r12, lr, d0 ; CHECK-NEXT: vmov r12, lr, d0
; CHECK-NEXT: vmov r4, r5, d2 ; CHECK-NEXT: vmov r4, r5, d2
; CHECK-NEXT: subs r0, r2, r0 ; CHECK-NEXT: subs r0, r2, r0
; CHECK-NEXT: sbcs.w r0, r3, r1 ; CHECK-NEXT: sbcs.w r0, r3, r1
; CHECK-NEXT: mov.w r0, #0 ; CHECK-NEXT: cset r0, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r0, #1
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: subs.w r1, r4, r12 ; CHECK-NEXT: subs.w r1, r4, r12
; CHECK-NEXT: sbcs.w r1, r5, lr ; CHECK-NEXT: sbcs.w r1, r5, lr
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r1, lt
; CHECK-NEXT: movlt r6, #1 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: vmov q2[2], q2[0], r1, r0 ; CHECK-NEXT: vmov q2[2], q2[0], r1, r0
; CHECK-NEXT: vmov q2[3], q2[1], r1, r0 ; CHECK-NEXT: vmov q2[3], q2[1], r1, r0
; CHECK-NEXT: vbic q1, q1, q2 ; CHECK-NEXT: vbic q1, q1, q2
; CHECK-NEXT: vand q0, q0, q2 ; CHECK-NEXT: vand q0, q0, q2
; CHECK-NEXT: vorr q0, q0, q1 ; CHECK-NEXT: vorr q0, q0, q1
; CHECK-NEXT: pop {r4, r5, r6, pc} ; CHECK-NEXT: pop {r4, r5, r7, pc}
entry: entry:
%0 = icmp sgt <2 x i64> %s1, %s2 %0 = icmp sgt <2 x i64> %s1, %s2
%1 = select <2 x i1> %0, <2 x i64> %s1, <2 x i64> %s2 %1 = select <2 x i1> %0, <2 x i64> %s1, <2 x i64> %s2
ret <2 x i64> %1 ret <2 x i64> %1
} }
define arm_aapcs_vfpcc <16 x i8> @umax_v16i8(<16 x i8> %s1, <16 x i8> %s2) { define arm_aapcs_vfpcc <16 x i8> @umax_v16i8(<16 x i8> %s1, <16 x i8> %s2) {
; CHECK-LABEL: umax_v16i8: ; CHECK-LABEL: umax_v16i8:
Show All 26 Linesentry:
%0 = icmp ugt <4 x i32> %s1, %s2 %0 = icmp ugt <4 x i32> %s1, %s2
%1 = select <4 x i1> %0, <4 x i32> %s1, <4 x i32> %s2 %1 = select <4 x i1> %0, <4 x i32> %s1, <4 x i32> %s2
ret <4 x i32> %1 ret <4 x i32> %1
} }
define arm_aapcs_vfpcc <2 x i64> @umax_v2i64(<2 x i64> %s1, <2 x i64> %s2) { define arm_aapcs_vfpcc <2 x i64> @umax_v2i64(<2 x i64> %s1, <2 x i64> %s2) {
; CHECK-LABEL: umax_v2i64: ; CHECK-LABEL: umax_v2i64:
; CHECK: @ %bb.0: @ %entry ; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: .save {r4, r5, r6, lr} ; CHECK-NEXT: .save {r4, r5, r7, lr}
; CHECK-NEXT: push {r4, r5, r6, lr} ; CHECK-NEXT: push {r4, r5, r7, lr}
; CHECK-NEXT: vmov r0, r1, d1 ; CHECK-NEXT: vmov r0, r1, d1
; CHECK-NEXT: movs r6, #0
; CHECK-NEXT: vmov r2, r3, d3 ; CHECK-NEXT: vmov r2, r3, d3
; CHECK-NEXT: vmov r12, lr, d0 ; CHECK-NEXT: vmov r12, lr, d0
; CHECK-NEXT: vmov r4, r5, d2 ; CHECK-NEXT: vmov r4, r5, d2
; CHECK-NEXT: subs r0, r2, r0 ; CHECK-NEXT: subs r0, r2, r0
; CHECK-NEXT: sbcs.w r0, r3, r1 ; CHECK-NEXT: sbcs.w r0, r3, r1
; CHECK-NEXT: mov.w r0, #0 ; CHECK-NEXT: cset r0, lo
; CHECK-NEXT: it lo
; CHECK-NEXT: movlo r0, #1
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: subs.w r1, r4, r12 ; CHECK-NEXT: subs.w r1, r4, r12
; CHECK-NEXT: sbcs.w r1, r5, lr ; CHECK-NEXT: sbcs.w r1, r5, lr
; CHECK-NEXT: it lo ; CHECK-NEXT: cset r1, lo
; CHECK-NEXT: movlo r6, #1 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: cmp r6, #0
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: vmov q2[2], q2[0], r1, r0 ; CHECK-NEXT: vmov q2[2], q2[0], r1, r0
; CHECK-NEXT: vmov q2[3], q2[1], r1, r0 ; CHECK-NEXT: vmov q2[3], q2[1], r1, r0
; CHECK-NEXT: vbic q1, q1, q2 ; CHECK-NEXT: vbic q1, q1, q2
; CHECK-NEXT: vand q0, q0, q2 ; CHECK-NEXT: vand q0, q0, q2
; CHECK-NEXT: vorr q0, q0, q1 ; CHECK-NEXT: vorr q0, q0, q1
; CHECK-NEXT: pop {r4, r5, r6, pc} ; CHECK-NEXT: pop {r4, r5, r7, pc}
entry: entry:
%0 = icmp ugt <2 x i64> %s1, %s2 %0 = icmp ugt <2 x i64> %s1, %s2
%1 = select <2 x i1> %0, <2 x i64> %s1, <2 x i64> %s2 %1 = select <2 x i1> %0, <2 x i64> %s1, <2 x i64> %s2
ret <2 x i64> %1 ret <2 x i64> %1
} }
define arm_aapcs_vfpcc <4 x float> @maxnm_float32_t(<4 x float> %src1, <4 x float> %src2) { define arm_aapcs_vfpcc <4 x float> @maxnm_float32_t(<4 x float> %src1, <4 x float> %src2) {
▲ Show 20 Lines • Show All 91 Lines • Show Last 20 Lines

llvm/test/CodeGen/Thumb2/mve-pred-ext.ll

Show First 20 Lines • Show All 69 Lines • ▼ Show 20 Linesentry:
%0 = sext <16 x i1> %c to <16 x i8> %0 = sext <16 x i1> %c to <16 x i8>
ret <16 x i8> %0 ret <16 x i8> %0
} }
define arm_aapcs_vfpcc <2 x i64> @sext_v2i1_v2i64(<2 x i64> %src) { define arm_aapcs_vfpcc <2 x i64> @sext_v2i1_v2i64(<2 x i64> %src) {
; CHECK-LABEL: sext_v2i1_v2i64: ; CHECK-LABEL: sext_v2i1_v2i64:
; CHECK: @ %bb.0: @ %entry ; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: vmov r0, r1, d1 ; CHECK-NEXT: vmov r0, r1, d1
; CHECK-NEXT: movs r3, #0 ; CHECK-NEXT: mov.w r12, #0
; CHECK-NEXT: vmov r2, r12, d0 ; CHECK-NEXT: vmov r2, r3, d0
; CHECK-NEXT: rsbs r0, r0, #0 ; CHECK-NEXT: rsbs r0, r0, #0
; CHECK-NEXT: sbcs.w r0, r3, r1 ; CHECK-NEXT: sbcs.w r0, r12, r1
; CHECK-NEXT: mov.w r0, #0 ; CHECK-NEXT: cset r0, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r0, #1
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: rsbs r1, r2, #0 ; CHECK-NEXT: rsbs r1, r2, #0
; CHECK-NEXT: sbcs.w r1, r3, r12 ; CHECK-NEXT: sbcs.w r1, r12, r3
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r1, lt
; CHECK-NEXT: movlt r3, #1 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: vmov q0[2], q0[0], r1, r0 ; CHECK-NEXT: vmov q0[2], q0[0], r1, r0
; CHECK-NEXT: vmov q0[3], q0[1], r1, r0 ; CHECK-NEXT: vmov q0[3], q0[1], r1, r0
; CHECK-NEXT: bx lr ; CHECK-NEXT: bx lr
entry: entry:
%c = icmp sgt <2 x i64> %src, zeroinitializer %c = icmp sgt <2 x i64> %src, zeroinitializer
%0 = sext <2 x i1> %c to <2 x i64> %0 = sext <2 x i1> %c to <2 x i64>
ret <2 x i64> %0 ret <2 x i64> %0
▲ Show 20 Lines • Show All 111 Lines • ▼ Show 20 Lines
} }
define arm_aapcs_vfpcc <2 x i64> @zext_v2i1_v2i64(<2 x i64> %src) { define arm_aapcs_vfpcc <2 x i64> @zext_v2i1_v2i64(<2 x i64> %src) {
; CHECK-LABEL: zext_v2i1_v2i64: ; CHECK-LABEL: zext_v2i1_v2i64:
; CHECK: @ %bb.0: @ %entry ; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: .save {r7, lr} ; CHECK-NEXT: .save {r7, lr}
; CHECK-NEXT: push {r7, lr} ; CHECK-NEXT: push {r7, lr}
; CHECK-NEXT: vmov r0, r1, d1 ; CHECK-NEXT: vmov r0, r1, d1
; CHECK-NEXT: movs r3, #0 ; CHECK-NEXT: mov.w r12, #0
; CHECK-NEXT: vmov lr, r12, d0 ; CHECK-NEXT: vmov r2, lr, d0
; CHECK-NEXT: adr r2, .LCPI12_0 ; CHECK-NEXT: adr r3, .LCPI12_0
; CHECK-NEXT: vldrw.u32 q0, [r2] ; CHECK-NEXT: vldrw.u32 q0, [r3]
; CHECK-NEXT: rsbs r0, r0, #0 ; CHECK-NEXT: rsbs r0, r0, #0
; CHECK-NEXT: sbcs.w r0, r3, r1 ; CHECK-NEXT: sbcs.w r0, r12, r1
; CHECK-NEXT: mov.w r0, #0 ; CHECK-NEXT: cset r0, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r0, #1
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: rsbs.w r1, lr, #0 ; CHECK-NEXT: rsbs r1, r2, #0
; CHECK-NEXT: sbcs.w r1, r3, r12 ; CHECK-NEXT: sbcs.w r1, r12, lr
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r1, lt
; CHECK-NEXT: movlt r3, #1 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: vmov q1[2], q1[0], r1, r0 ; CHECK-NEXT: vmov q1[2], q1[0], r1, r0
; CHECK-NEXT: vand q0, q1, q0 ; CHECK-NEXT: vand q0, q1, q0
; CHECK-NEXT: pop {r7, pc} ; CHECK-NEXT: pop {r7, pc}
; CHECK-NEXT: .p2align 4 ; CHECK-NEXT: .p2align 4
; CHECK-NEXT: @ %bb.1: ; CHECK-NEXT: @ %bb.1:
; CHECK-NEXT: .LCPI12_0: ; CHECK-NEXT: .LCPI12_0:
; CHECK-NEXT: .long 1 @ 0x1 ; CHECK-NEXT: .long 1 @ 0x1
▲ Show 20 Lines • Show All 235 Lines • Show Last 20 Lines

llvm/test/CodeGen/Thumb2/mve-satmul-loops.ll

Show All 40 Lines
; CHECK-NEXT: ldrd r5, r4, [r0], #8 ; CHECK-NEXT: ldrd r5, r4, [r0], #8
; CHECK-NEXT: mov.w r3, #-1 ; CHECK-NEXT: mov.w r3, #-1
; CHECK-NEXT: ldrd r8, r7, [r1], #8 ; CHECK-NEXT: ldrd r8, r7, [r1], #8
; CHECK-NEXT: smull r4, r7, r7, r4 ; CHECK-NEXT: smull r4, r7, r7, r4
; CHECK-NEXT: asrl r4, r7, #31 ; CHECK-NEXT: asrl r4, r7, #31
; CHECK-NEXT: smull r6, r5, r8, r5 ; CHECK-NEXT: smull r6, r5, r8, r5
; CHECK-NEXT: rsbs.w r9, r4, #-2147483648 ; CHECK-NEXT: rsbs.w r9, r4, #-2147483648
; CHECK-NEXT: sbcs r3, r7 ; CHECK-NEXT: sbcs r3, r7
; CHECK-NEXT: mov.w r3, #0 ; CHECK-NEXT: cset r3, lt
; CHECK-NEXT: asrl r6, r5, #31 ; CHECK-NEXT: asrl r6, r5, #31
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r3, #1
; CHECK-NEXT: cmp r3, #0 ; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: vmov q4[2], q4[0], r6, r4
; CHECK-NEXT: csetm r9, ne ; CHECK-NEXT: csetm r9, ne
; CHECK-NEXT: rsbs.w r3, r6, #-2147483648 ; CHECK-NEXT: rsbs.w r3, r6, #-2147483648
; CHECK-NEXT: mov.w r3, #-1 ; CHECK-NEXT: mov.w r3, #-1
; CHECK-NEXT: vmov q4[2], q4[0], r6, r4
; CHECK-NEXT: sbcs r3, r5
; CHECK-NEXT: vmov q4[3], q4[1], r5, r7 ; CHECK-NEXT: vmov q4[3], q4[1], r5, r7
; CHECK-NEXT: mov.w r3, #0 ; CHECK-NEXT: sbcs r3, r5
; CHECK-NEXT: mvn r6, #-2147483648 ; CHECK-NEXT: mvn r6, #-2147483648
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r3, lt
; CHECK-NEXT: movlt r3, #1
; CHECK-NEXT: cmp r3, #0 ; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: csetm r3, ne ; CHECK-NEXT: csetm r3, ne
; CHECK-NEXT: vmov q2[2], q2[0], r3, r9 ; CHECK-NEXT: vmov q2[2], q2[0], r3, r9
; CHECK-NEXT: vmov q2[3], q2[1], r3, r9 ; CHECK-NEXT: vmov q2[3], q2[1], r3, r9
; CHECK-NEXT: vbic q3, q0, q2 ; CHECK-NEXT: vbic q3, q0, q2
; CHECK-NEXT: vand q2, q4, q2 ; CHECK-NEXT: vand q2, q4, q2
; CHECK-NEXT: vorr q2, q2, q3 ; CHECK-NEXT: vorr q2, q2, q3
; CHECK-NEXT: vmov r3, r4, d4 ; CHECK-NEXT: vmov r3, r4, d4
; CHECK-NEXT: subs r3, r3, r6 ; CHECK-NEXT: subs r3, r3, r6
; CHECK-NEXT: sbcs r3, r4, #0 ; CHECK-NEXT: sbcs r3, r4, #0
; CHECK-NEXT: vmov r4, r5, d5 ; CHECK-NEXT: vmov r4, r5, d5
; CHECK-NEXT: mov.w r3, #0 ; CHECK-NEXT: cset r3, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r3, #1
; CHECK-NEXT: cmp r3, #0 ; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: csetm r3, ne ; CHECK-NEXT: csetm r3, ne
; CHECK-NEXT: vmov.32 q3[1], r3 ; CHECK-NEXT: vmov.32 q3[1], r3
; CHECK-NEXT: subs r4, r4, r6 ; CHECK-NEXT: subs r4, r4, r6
; CHECK-NEXT: sbcs r4, r5, #0 ; CHECK-NEXT: sbcs r4, r5, #0
; CHECK-NEXT: mov.w r4, #0 ; CHECK-NEXT: cset r4, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r4, #1
; CHECK-NEXT: cmp r4, #0 ; CHECK-NEXT: cmp r4, #0
; CHECK-NEXT: csetm r4, ne ; CHECK-NEXT: csetm r4, ne
; CHECK-NEXT: vmov q3[2], q3[0], r3, r4 ; CHECK-NEXT: vmov q3[2], q3[0], r3, r4
; CHECK-NEXT: vbic q4, q1, q3 ; CHECK-NEXT: vbic q4, q1, q3
; CHECK-NEXT: vand q2, q2, q3 ; CHECK-NEXT: vand q2, q2, q3
; CHECK-NEXT: vorr q2, q2, q4 ; CHECK-NEXT: vorr q2, q2, q4
; CHECK-NEXT: vmov r3, s10 ; CHECK-NEXT: vmov r3, s10
; CHECK-NEXT: vmov r4, s8 ; CHECK-NEXT: vmov r4, s8
Show All 11 Lines
; CHECK-NEXT: .LBB0_7: @ %for.body ; CHECK-NEXT: .LBB0_7: @ %for.body
; CHECK-NEXT: @ =>This Inner Loop Header: Depth=1 ; CHECK-NEXT: @ =>This Inner Loop Header: Depth=1
; CHECK-NEXT: ldr r3, [r12], #4 ; CHECK-NEXT: ldr r3, [r12], #4
; CHECK-NEXT: ldr r4, [r10], #4 ; CHECK-NEXT: ldr r4, [r10], #4
; CHECK-NEXT: smull r4, r3, r4, r3 ; CHECK-NEXT: smull r4, r3, r4, r3
; CHECK-NEXT: asrl r4, r3, #31 ; CHECK-NEXT: asrl r4, r3, #31
; CHECK-NEXT: subs r5, r1, r4 ; CHECK-NEXT: subs r5, r1, r4
; CHECK-NEXT: sbcs.w r5, r0, r3 ; CHECK-NEXT: sbcs.w r5, r0, r3
; CHECK-NEXT: mov.w r5, #0 ; CHECK-NEXT: cset r5, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r5, #1
; CHECK-NEXT: cmp r5, #0 ; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: csel r4, r4, r1, ne ; CHECK-NEXT: csel r4, r4, r1, ne
; CHECK-NEXT: csel r3, r3, r0, ne ; CHECK-NEXT: csel r3, r3, r0, ne
; CHECK-NEXT: subs r5, r4, r2 ; CHECK-NEXT: subs r5, r4, r2
; CHECK-NEXT: sbcs r3, r3, #0 ; CHECK-NEXT: sbcs r3, r3, #0
; CHECK-NEXT: csel r3, r4, r2, lt ; CHECK-NEXT: csel r3, r4, r2, lt
; CHECK-NEXT: str r3, [r11], #4 ; CHECK-NEXT: str r3, [r11], #4
; CHECK-NEXT: le lr, .LBB0_7 ; CHECK-NEXT: le lr, .LBB0_7
▲ Show 20 Lines • Show All 94 Lines • ▼ Show 20 Lines
; CHECK-NEXT: sub sp, #4 ; CHECK-NEXT: sub sp, #4
; CHECK-NEXT: .vsave {d8, d9, d10, d11, d12, d13} ; CHECK-NEXT: .vsave {d8, d9, d10, d11, d12, d13}
; CHECK-NEXT: vpush {d8, d9, d10, d11, d12, d13} ; CHECK-NEXT: vpush {d8, d9, d10, d11, d12, d13}
; CHECK-NEXT: .pad #16 ; CHECK-NEXT: .pad #16
; CHECK-NEXT: sub sp, #16 ; CHECK-NEXT: sub sp, #16
; CHECK-NEXT: cmp r3, #0 ; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: beq.w .LBB1_8 ; CHECK-NEXT: beq.w .LBB1_8
; CHECK-NEXT: @ %bb.1: @ %for.body.preheader ; CHECK-NEXT: @ %bb.1: @ %for.body.preheader
; CHECK-NEXT: movs r7, #0
; CHECK-NEXT: cmp r3, #3 ; CHECK-NEXT: cmp r3, #3
; CHECK-NEXT: bhi .LBB1_3 ; CHECK-NEXT: bhi .LBB1_3
; CHECK-NEXT: @ %bb.2: ; CHECK-NEXT: @ %bb.2:
; CHECK-NEXT: movs r7, #0
; CHECK-NEXT: mov r12, r0 ; CHECK-NEXT: mov r12, r0
; CHECK-NEXT: mov r9, r1 ; CHECK-NEXT: mov r9, r1
; CHECK-NEXT: mov r11, r2 ; CHECK-NEXT: mov r11, r2
; CHECK-NEXT: b .LBB1_6 ; CHECK-NEXT: b .LBB1_6
; CHECK-NEXT: .LBB1_3: @ %vector.ph ; CHECK-NEXT: .LBB1_3: @ %vector.ph
; CHECK-NEXT: str r3, [sp, #4] @ 4-byte Spill ; CHECK-NEXT: str r3, [sp, #4] @ 4-byte Spill
; CHECK-NEXT: bic r3, r3, #3 ; CHECK-NEXT: bic r3, r3, #3
; CHECK-NEXT: subs r7, r3, #4 ; CHECK-NEXT: subs r7, r3, #4
Show All 20 Lines
; CHECK-NEXT: vmov.f32 s22, s15 ; CHECK-NEXT: vmov.f32 s22, s15
; CHECK-NEXT: vmullb.s32 q6, q5, q4 ; CHECK-NEXT: vmullb.s32 q6, q5, q4
; CHECK-NEXT: vmov.f32 s10, s9 ; CHECK-NEXT: vmov.f32 s10, s9
; CHECK-NEXT: vmov r4, r7, d13 ; CHECK-NEXT: vmov r4, r7, d13
; CHECK-NEXT: asrl r4, r7, #31 ; CHECK-NEXT: asrl r4, r7, #31
; CHECK-NEXT: vmov r6, s12 ; CHECK-NEXT: vmov r6, s12
; CHECK-NEXT: rsbs.w r5, r4, #-2147483648 ; CHECK-NEXT: rsbs.w r5, r4, #-2147483648
; CHECK-NEXT: sbcs.w r5, r2, r7 ; CHECK-NEXT: sbcs.w r5, r2, r7
; CHECK-NEXT: mov.w r5, #0 ; CHECK-NEXT: cset r5, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r5, #1
; CHECK-NEXT: cmp r5, #0 ; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: vmov r10, r5, d12 ; CHECK-NEXT: vmov r10, r5, d12
; CHECK-NEXT: csetm r8, ne
; CHECK-NEXT: asrl r10, r5, #31 ; CHECK-NEXT: asrl r10, r5, #31
; CHECK-NEXT: csetm r8, ne
; CHECK-NEXT: rsbs.w r3, r10, #-2147483648 ; CHECK-NEXT: rsbs.w r3, r10, #-2147483648
; CHECK-NEXT: vmov q6[2], q6[0], r10, r4 ; CHECK-NEXT: vmov q6[2], q6[0], r10, r4
; CHECK-NEXT: sbcs.w r3, r2, r5 ; CHECK-NEXT: sbcs.w r3, r2, r5
; CHECK-NEXT: vmov q6[3], q6[1], r5, r7 ; CHECK-NEXT: vmov q6[3], q6[1], r5, r7
; CHECK-NEXT: mov.w r3, #0 ; CHECK-NEXT: cset r3, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r3, #1
; CHECK-NEXT: cmp r3, #0 ; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: csetm r3, ne ; CHECK-NEXT: csetm r3, ne
; CHECK-NEXT: vmov q4[2], q4[0], r3, r8 ; CHECK-NEXT: vmov q4[2], q4[0], r3, r8
; CHECK-NEXT: vmov q4[3], q4[1], r3, r8 ; CHECK-NEXT: vmov q4[3], q4[1], r3, r8
; CHECK-NEXT: mvn r8, #-2147483648 ; CHECK-NEXT: mvn r8, #-2147483648
; CHECK-NEXT: vbic q5, q0, q4 ; CHECK-NEXT: vbic q5, q0, q4
; CHECK-NEXT: vand q4, q6, q4 ; CHECK-NEXT: vand q4, q6, q4
; CHECK-NEXT: vorr q4, q4, q5 ; CHECK-NEXT: vorr q4, q4, q5
; CHECK-NEXT: vmov r3, r4, d8 ; CHECK-NEXT: vmov r3, r4, d8
; CHECK-NEXT: subs.w r3, r3, r8 ; CHECK-NEXT: subs.w r3, r3, r8
; CHECK-NEXT: sbcs r3, r4, #0 ; CHECK-NEXT: sbcs r3, r4, #0
; CHECK-NEXT: vmov r4, r5, d9 ; CHECK-NEXT: vmov r4, r5, d9
; CHECK-NEXT: mov.w r3, #0 ; CHECK-NEXT: cset r3, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r3, #1
; CHECK-NEXT: cmp r3, #0 ; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: csetm r3, ne ; CHECK-NEXT: csetm r3, ne
; CHECK-NEXT: vmov.32 q5[1], r3 ; CHECK-NEXT: vmov.32 q5[1], r3
; CHECK-NEXT: subs.w r4, r4, r8 ; CHECK-NEXT: subs.w r4, r4, r8
; CHECK-NEXT: sbcs r4, r5, #0 ; CHECK-NEXT: sbcs r4, r5, #0
; CHECK-NEXT: vmov r5, s8 ; CHECK-NEXT: vmov r5, s8
; CHECK-NEXT: mov.w r4, #0 ; CHECK-NEXT: cset r4, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r4, #1
; CHECK-NEXT: cmp r4, #0 ; CHECK-NEXT: cmp r4, #0
; CHECK-NEXT: csetm r4, ne ; CHECK-NEXT: csetm r4, ne
; CHECK-NEXT: vmov q5[2], q5[0], r3, r4 ; CHECK-NEXT: vmov q5[2], q5[0], r3, r4
; CHECK-NEXT: vmov r3, s10 ; CHECK-NEXT: vmov r3, s10
; CHECK-NEXT: vmov.f32 s10, s13 ; CHECK-NEXT: vmov.f32 s10, s13
; CHECK-NEXT: vbic q6, q1, q5 ; CHECK-NEXT: vbic q6, q1, q5
; CHECK-NEXT: vand q4, q4, q5 ; CHECK-NEXT: vand q4, q4, q5
; CHECK-NEXT: vorr q4, q4, q6 ; CHECK-NEXT: vorr q4, q4, q6
; CHECK-NEXT: vmov r4, s10 ; CHECK-NEXT: vmov r4, s10
; CHECK-NEXT: smull r6, r5, r6, r5 ; CHECK-NEXT: smull r6, r5, r6, r5
; CHECK-NEXT: asrl r6, r5, #31 ; CHECK-NEXT: asrl r6, r5, #31
; CHECK-NEXT: smull r4, r7, r4, r3 ; CHECK-NEXT: smull r4, r7, r4, r3
; CHECK-NEXT: asrl r4, r7, #31 ; CHECK-NEXT: asrl r4, r7, #31
; CHECK-NEXT: rsbs.w r3, r4, #-2147483648 ; CHECK-NEXT: rsbs.w r3, r4, #-2147483648
; CHECK-NEXT: vmov q5[2], q5[0], r6, r4 ; CHECK-NEXT: vmov q5[2], q5[0], r6, r4
; CHECK-NEXT: sbcs.w r3, r2, r7 ; CHECK-NEXT: sbcs.w r3, r2, r7
; CHECK-NEXT: vmov q5[3], q5[1], r5, r7 ; CHECK-NEXT: vmov q5[3], q5[1], r5, r7
; CHECK-NEXT: mov.w r3, #0 ; CHECK-NEXT: cset r3, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r3, #1
; CHECK-NEXT: cmp r3, #0 ; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: csetm r3, ne ; CHECK-NEXT: csetm r3, ne
; CHECK-NEXT: rsbs.w r1, r6, #-2147483648 ; CHECK-NEXT: rsbs.w r1, r6, #-2147483648
; CHECK-NEXT: sbcs.w r1, r2, r5 ; CHECK-NEXT: sbcs.w r1, r2, r5
; CHECK-NEXT: mov.w r1, #0 ; CHECK-NEXT: cset r1, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r1, #1
; CHECK-NEXT: cmp r1, #0 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: vmov q2[2], q2[0], r1, r3 ; CHECK-NEXT: vmov q2[2], q2[0], r1, r3
; CHECK-NEXT: vmov q2[3], q2[1], r1, r3 ; CHECK-NEXT: vmov q2[3], q2[1], r1, r3
; CHECK-NEXT: ldrd r1, r2, [sp, #8] @ 8-byte Folded Reload ; CHECK-NEXT: ldrd r1, r2, [sp, #8] @ 8-byte Folded Reload
; CHECK-NEXT: vbic q3, q0, q2 ; CHECK-NEXT: vbic q3, q0, q2
; CHECK-NEXT: vand q2, q5, q2 ; CHECK-NEXT: vand q2, q5, q2
; CHECK-NEXT: vorr q2, q2, q3 ; CHECK-NEXT: vorr q2, q2, q3
; CHECK-NEXT: vmov r4, r3, d4 ; CHECK-NEXT: vmov r4, r3, d4
; CHECK-NEXT: subs.w r4, r4, r8 ; CHECK-NEXT: subs.w r4, r4, r8
; CHECK-NEXT: sbcs r3, r3, #0 ; CHECK-NEXT: sbcs r3, r3, #0
; CHECK-NEXT: mov.w r3, #0 ; CHECK-NEXT: cset r3, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r3, #1
; CHECK-NEXT: cmp r3, #0 ; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: vmov r3, r4, d5 ; CHECK-NEXT: vmov r3, r4, d5
; CHECK-NEXT: csetm r5, ne ; CHECK-NEXT: csetm r5, ne
; CHECK-NEXT: vmov.32 q3[1], r5 ; CHECK-NEXT: vmov.32 q3[1], r5
; CHECK-NEXT: subs.w r3, r3, r8 ; CHECK-NEXT: subs.w r3, r3, r8
; CHECK-NEXT: sbcs r3, r4, #0 ; CHECK-NEXT: sbcs r3, r4, #0
; CHECK-NEXT: mov.w r3, #0 ; CHECK-NEXT: cset r3, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r3, #1
; CHECK-NEXT: cmp r3, #0 ; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: csetm r3, ne ; CHECK-NEXT: csetm r3, ne
; CHECK-NEXT: vmov q3[2], q3[0], r5, r3 ; CHECK-NEXT: vmov q3[2], q3[0], r5, r3
; CHECK-NEXT: vbic q5, q1, q3 ; CHECK-NEXT: vbic q5, q1, q3
; CHECK-NEXT: vand q2, q2, q3 ; CHECK-NEXT: vand q2, q2, q3
; CHECK-NEXT: vorr q2, q2, q5 ; CHECK-NEXT: vorr q2, q2, q5
; CHECK-NEXT: vmov.f32 s9, s10 ; CHECK-NEXT: vmov.f32 s9, s10
; CHECK-NEXT: vmov.f32 s10, s16 ; CHECK-NEXT: vmov.f32 s10, s16
Show All 12 Lines
; CHECK-NEXT: .LBB1_7: @ %for.body ; CHECK-NEXT: .LBB1_7: @ %for.body
; CHECK-NEXT: @ =>This Inner Loop Header: Depth=1 ; CHECK-NEXT: @ =>This Inner Loop Header: Depth=1
; CHECK-NEXT: ldr r1, [r12], #4 ; CHECK-NEXT: ldr r1, [r12], #4
; CHECK-NEXT: ldr r4, [r9], #4 ; CHECK-NEXT: ldr r4, [r9], #4
; CHECK-NEXT: smull r4, r1, r4, r1 ; CHECK-NEXT: smull r4, r1, r4, r1
; CHECK-NEXT: asrl r4, r1, #31 ; CHECK-NEXT: asrl r4, r1, #31
; CHECK-NEXT: subs r5, r3, r4 ; CHECK-NEXT: subs r5, r3, r4
; CHECK-NEXT: sbcs.w r5, r0, r1 ; CHECK-NEXT: sbcs.w r5, r0, r1
; CHECK-NEXT: mov.w r5, #0 ; CHECK-NEXT: cset r5, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r5, #1
; CHECK-NEXT: cmp r5, #0 ; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: csel r4, r4, r3, ne ; CHECK-NEXT: csel r4, r4, r3, ne
; CHECK-NEXT: csel r1, r1, r0, ne ; CHECK-NEXT: csel r1, r1, r0, ne
; CHECK-NEXT: subs r5, r4, r2 ; CHECK-NEXT: subs r5, r4, r2
; CHECK-NEXT: sbcs r1, r1, #0 ; CHECK-NEXT: sbcs r1, r1, #0
; CHECK-NEXT: csel r1, r4, r2, lt ; CHECK-NEXT: csel r1, r4, r2, lt
; CHECK-NEXT: str r1, [r11], #4 ; CHECK-NEXT: str r1, [r11], #4
; CHECK-NEXT: le lr, .LBB1_7 ; CHECK-NEXT: le lr, .LBB1_7
▲ Show 20 Lines • Show All 134 Lines • ▼ Show 20 Lines
; CHECK-NEXT: vmov.f32 s26, s19 ; CHECK-NEXT: vmov.f32 s26, s19
; CHECK-NEXT: vmov.f32 s28, s22 ; CHECK-NEXT: vmov.f32 s28, s22
; CHECK-NEXT: vmov.f32 s30, s23 ; CHECK-NEXT: vmov.f32 s30, s23
; CHECK-NEXT: vmullb.s32 q0, q7, q6 ; CHECK-NEXT: vmullb.s32 q0, q7, q6
; CHECK-NEXT: vmov r6, r5, d1 ; CHECK-NEXT: vmov r6, r5, d1
; CHECK-NEXT: asrl r6, r5, #31 ; CHECK-NEXT: asrl r6, r5, #31
; CHECK-NEXT: rsbs.w r7, r6, #-2147483648 ; CHECK-NEXT: rsbs.w r7, r6, #-2147483648
; CHECK-NEXT: sbcs.w r7, r12, r5 ; CHECK-NEXT: sbcs.w r7, r12, r5
; CHECK-NEXT: mov.w r7, #0 ; CHECK-NEXT: cset r7, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r7, #1
; CHECK-NEXT: cmp r7, #0 ; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: vmov r4, r7, d0 ; CHECK-NEXT: vmov r4, r7, d0
; CHECK-NEXT: csetm r10, ne
; CHECK-NEXT: asrl r4, r7, #31 ; CHECK-NEXT: asrl r4, r7, #31
; CHECK-NEXT: csetm r10, ne
; CHECK-NEXT: rsbs.w r3, r4, #-2147483648 ; CHECK-NEXT: rsbs.w r3, r4, #-2147483648
; CHECK-NEXT: vmov q7[2], q7[0], r4, r6 ; CHECK-NEXT: vmov q7[2], q7[0], r4, r6
; CHECK-NEXT: sbcs.w r3, r12, r7 ; CHECK-NEXT: sbcs.w r3, r12, r7
; CHECK-NEXT: vmov q7[3], q7[1], r7, r5 ; CHECK-NEXT: vmov q7[3], q7[1], r7, r5
; CHECK-NEXT: mov.w r3, #0 ; CHECK-NEXT: cset r3, lt
; CHECK-NEXT: vmov r7, s20 ; CHECK-NEXT: vmov r7, s20
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r3, #1
; CHECK-NEXT: cmp r3, #0 ; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: csetm r3, ne ; CHECK-NEXT: csetm r3, ne
; CHECK-NEXT: vmov q0[2], q0[0], r3, r10 ; CHECK-NEXT: vmov q0[2], q0[0], r3, r10
; CHECK-NEXT: vmov q0[3], q0[1], r3, r10 ; CHECK-NEXT: vmov q0[3], q0[1], r3, r10
; CHECK-NEXT: vbic q6, q2, q0 ; CHECK-NEXT: vbic q6, q2, q0
; CHECK-NEXT: vand q0, q7, q0 ; CHECK-NEXT: vand q0, q7, q0
; CHECK-NEXT: vorr q6, q0, q6 ; CHECK-NEXT: vorr q6, q0, q6
; CHECK-NEXT: vmov r3, r4, d12 ; CHECK-NEXT: vmov r3, r4, d12
; CHECK-NEXT: subs.w r3, r3, r8 ; CHECK-NEXT: subs.w r3, r3, r8
; CHECK-NEXT: sbcs r3, r4, #0 ; CHECK-NEXT: sbcs r3, r4, #0
; CHECK-NEXT: vmov r4, r5, d13 ; CHECK-NEXT: vmov r4, r5, d13
; CHECK-NEXT: mov.w r3, #0 ; CHECK-NEXT: cset r3, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r3, #1
; CHECK-NEXT: cmp r3, #0 ; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: csetm r3, ne ; CHECK-NEXT: csetm r3, ne
; CHECK-NEXT: vmov.32 q0[1], r3 ; CHECK-NEXT: vmov.32 q0[1], r3
; CHECK-NEXT: subs.w r4, r4, r8 ; CHECK-NEXT: subs.w r4, r4, r8
; CHECK-NEXT: sbcs r4, r5, #0 ; CHECK-NEXT: sbcs r4, r5, #0
; CHECK-NEXT: mov.w r4, #0 ; CHECK-NEXT: cset r4, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r4, #1
; CHECK-NEXT: cmp r4, #0 ; CHECK-NEXT: cmp r4, #0
; CHECK-NEXT: csetm r4, ne ; CHECK-NEXT: csetm r4, ne
; CHECK-NEXT: vmov q0[2], q0[0], r3, r4 ; CHECK-NEXT: vmov q0[2], q0[0], r3, r4
; CHECK-NEXT: vbic q7, q3, q0 ; CHECK-NEXT: vbic q7, q3, q0
; CHECK-NEXT: vand q0, q6, q0 ; CHECK-NEXT: vand q0, q6, q0
; CHECK-NEXT: vorr q6, q0, q7 ; CHECK-NEXT: vorr q6, q0, q7
; CHECK-NEXT: vmov.f32 s2, s17 ; CHECK-NEXT: vmov.f32 s2, s17
; CHECK-NEXT: vmov r3, s2 ; CHECK-NEXT: vmov r3, s2
; CHECK-NEXT: vmov.f32 s2, s21 ; CHECK-NEXT: vmov.f32 s2, s21
; CHECK-NEXT: vmov r4, s2 ; CHECK-NEXT: vmov r4, s2
; CHECK-NEXT: smull r6, r5, r4, r3 ; CHECK-NEXT: smull r6, r5, r4, r3
; CHECK-NEXT: vmov r4, s16 ; CHECK-NEXT: vmov r4, s16
; CHECK-NEXT: asrl r6, r5, #31 ; CHECK-NEXT: asrl r6, r5, #31
; CHECK-NEXT: rsbs.w r3, r6, #-2147483648 ; CHECK-NEXT: rsbs.w r3, r6, #-2147483648
; CHECK-NEXT: sbcs.w r3, r12, r5 ; CHECK-NEXT: sbcs.w r3, r12, r5
; CHECK-NEXT: mov.w r3, #0 ; CHECK-NEXT: cset r3, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r3, #1
; CHECK-NEXT: cmp r3, #0 ; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: csetm r10, ne ; CHECK-NEXT: csetm r10, ne
; CHECK-NEXT: smull r4, r7, r7, r4 ; CHECK-NEXT: smull r4, r7, r7, r4
; CHECK-NEXT: asrl r4, r7, #31 ; CHECK-NEXT: asrl r4, r7, #31
; CHECK-NEXT: rsbs.w r3, r4, #-2147483648 ; CHECK-NEXT: rsbs.w r3, r4, #-2147483648
; CHECK-NEXT: vmov q5[2], q5[0], r4, r6 ; CHECK-NEXT: vmov q5[2], q5[0], r4, r6
; CHECK-NEXT: sbcs.w r3, r12, r7 ; CHECK-NEXT: sbcs.w r3, r12, r7
; CHECK-NEXT: vmov q5[3], q5[1], r7, r5 ; CHECK-NEXT: vmov q5[3], q5[1], r7, r5
; CHECK-NEXT: mov.w r3, #0 ; CHECK-NEXT: cset r3, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r3, #1
; CHECK-NEXT: cmp r3, #0 ; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: csetm r3, ne ; CHECK-NEXT: csetm r3, ne
; CHECK-NEXT: vmov q0[2], q0[0], r3, r10 ; CHECK-NEXT: vmov q0[2], q0[0], r3, r10
; CHECK-NEXT: vmov q0[3], q0[1], r3, r10 ; CHECK-NEXT: vmov q0[3], q0[1], r3, r10
; CHECK-NEXT: vbic q4, q2, q0 ; CHECK-NEXT: vbic q4, q2, q0
; CHECK-NEXT: vand q0, q5, q0 ; CHECK-NEXT: vand q0, q5, q0
; CHECK-NEXT: vorr q4, q0, q4 ; CHECK-NEXT: vorr q4, q0, q4
; CHECK-NEXT: vmov r3, r4, d8 ; CHECK-NEXT: vmov r3, r4, d8
; CHECK-NEXT: subs.w r3, r3, r8 ; CHECK-NEXT: subs.w r3, r3, r8
; CHECK-NEXT: sbcs r3, r4, #0 ; CHECK-NEXT: sbcs r3, r4, #0
; CHECK-NEXT: vmov r4, r5, d9 ; CHECK-NEXT: vmov r4, r5, d9
; CHECK-NEXT: mov.w r3, #0 ; CHECK-NEXT: cset r3, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r3, #1
; CHECK-NEXT: cmp r3, #0 ; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: csetm r3, ne ; CHECK-NEXT: csetm r3, ne
; CHECK-NEXT: vmov.32 q0[1], r3 ; CHECK-NEXT: vmov.32 q0[1], r3
; CHECK-NEXT: subs.w r4, r4, r8 ; CHECK-NEXT: subs.w r4, r4, r8
; CHECK-NEXT: sbcs r4, r5, #0 ; CHECK-NEXT: sbcs r4, r5, #0
; CHECK-NEXT: mov.w r4, #0 ; CHECK-NEXT: cset r4, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r4, #1
; CHECK-NEXT: cmp r4, #0 ; CHECK-NEXT: cmp r4, #0
; CHECK-NEXT: csetm r4, ne ; CHECK-NEXT: csetm r4, ne
; CHECK-NEXT: vmov q0[2], q0[0], r3, r4 ; CHECK-NEXT: vmov q0[2], q0[0], r3, r4
; CHECK-NEXT: vbic q5, q3, q0 ; CHECK-NEXT: vbic q5, q3, q0
; CHECK-NEXT: vand q0, q4, q0 ; CHECK-NEXT: vand q0, q4, q0
; CHECK-NEXT: vorr q0, q0, q5 ; CHECK-NEXT: vorr q0, q0, q5
; CHECK-NEXT: vmov.f32 s1, s2 ; CHECK-NEXT: vmov.f32 s1, s2
; CHECK-NEXT: vmov.f32 s2, s24 ; CHECK-NEXT: vmov.f32 s2, s24
▲ Show 20 Lines • Show All 97 Lines • ▼ Show 20 Lines
; CHECK-NEXT: @ =>This Inner Loop Header: Depth=1 ; CHECK-NEXT: @ =>This Inner Loop Header: Depth=1
; CHECK-NEXT: ldrd r4, r7, [r0], #8 ; CHECK-NEXT: ldrd r4, r7, [r0], #8
; CHECK-NEXT: ldrd r5, r10, [r1], #8 ; CHECK-NEXT: ldrd r5, r10, [r1], #8
; CHECK-NEXT: umull r4, r5, r5, r4 ; CHECK-NEXT: umull r4, r5, r5, r4
; CHECK-NEXT: lsrl r4, r5, #31 ; CHECK-NEXT: lsrl r4, r5, #31
; CHECK-NEXT: subs.w r6, r4, #-1 ; CHECK-NEXT: subs.w r6, r4, #-1
; CHECK-NEXT: umull r6, r7, r10, r7 ; CHECK-NEXT: umull r6, r7, r10, r7
; CHECK-NEXT: sbcs r5, r5, #0 ; CHECK-NEXT: sbcs r5, r5, #0
; CHECK-NEXT: mov.w r5, #0 ; CHECK-NEXT: cset r5, lo
; CHECK-NEXT: it lo
; CHECK-NEXT: movlo r5, #1
; CHECK-NEXT: cmp r5, #0 ; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: lsrl r6, r7, #31 ; CHECK-NEXT: lsrl r6, r7, #31
; CHECK-NEXT: csetm r9, ne ; CHECK-NEXT: csetm r9, ne
; CHECK-NEXT: subs.w r5, r6, #-1 ; CHECK-NEXT: subs.w r5, r6, #-1
; CHECK-NEXT: vmov.32 q0[1], r9 ; CHECK-NEXT: vmov.32 q0[1], r9
; CHECK-NEXT: sbcs r5, r7, #0 ; CHECK-NEXT: sbcs r5, r7, #0
; CHECK-NEXT: vmov q1[2], q1[0], r4, r6 ; CHECK-NEXT: vmov q1[2], q1[0], r4, r6
; CHECK-NEXT: mov.w r5, #0 ; CHECK-NEXT: cset r5, lo
; CHECK-NEXT: it lo
; CHECK-NEXT: movlo r5, #1
; CHECK-NEXT: cmp r5, #0 ; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: csetm r5, ne ; CHECK-NEXT: csetm r5, ne
; CHECK-NEXT: vmov q0[2], q0[0], r9, r5 ; CHECK-NEXT: vmov q0[2], q0[0], r9, r5
; CHECK-NEXT: vand q1, q1, q0 ; CHECK-NEXT: vand q1, q1, q0
; CHECK-NEXT: vorn q0, q1, q0 ; CHECK-NEXT: vorn q0, q1, q0
; CHECK-NEXT: vmov r4, s2 ; CHECK-NEXT: vmov r4, s2
; CHECK-NEXT: vmov r5, s0 ; CHECK-NEXT: vmov r5, s0
; CHECK-NEXT: strd r5, r4, [r2], #8 ; CHECK-NEXT: strd r5, r4, [r2], #8
▲ Show 20 Lines • Show All 98 Lines • ▼ Show 20 Lines
; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, r9, r10, r11, lr} ; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, r9, r10, r11, lr}
; CHECK-NEXT: .pad #4 ; CHECK-NEXT: .pad #4
; CHECK-NEXT: sub sp, #4 ; CHECK-NEXT: sub sp, #4
; CHECK-NEXT: .vsave {d8, d9} ; CHECK-NEXT: .vsave {d8, d9}
; CHECK-NEXT: vpush {d8, d9} ; CHECK-NEXT: vpush {d8, d9}
; CHECK-NEXT: cmp r3, #0 ; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: beq.w .LBB4_8 ; CHECK-NEXT: beq.w .LBB4_8
; CHECK-NEXT: @ %bb.1: @ %for.body.preheader ; CHECK-NEXT: @ %bb.1: @ %for.body.preheader
; CHECK-NEXT: mov.w r8, #0
; CHECK-NEXT: cmp r3, #3 ; CHECK-NEXT: cmp r3, #3
; CHECK-NEXT: bhi .LBB4_3 ; CHECK-NEXT: bhi .LBB4_3
; CHECK-NEXT: @ %bb.2: ; CHECK-NEXT: @ %bb.2:
; CHECK-NEXT: mov.w r8, #0
; CHECK-NEXT: mov r12, r0 ; CHECK-NEXT: mov r12, r0
; CHECK-NEXT: mov r9, r1 ; CHECK-NEXT: mov r9, r1
; CHECK-NEXT: mov r10, r2 ; CHECK-NEXT: mov r10, r2
; CHECK-NEXT: b .LBB4_6 ; CHECK-NEXT: b .LBB4_6
; CHECK-NEXT: .LBB4_3: @ %vector.ph ; CHECK-NEXT: .LBB4_3: @ %vector.ph
; CHECK-NEXT: bic r8, r3, #3 ; CHECK-NEXT: bic r8, r3, #3
; CHECK-NEXT: movs r6, #1 ; CHECK-NEXT: movs r6, #1
; CHECK-NEXT: sub.w r7, r8, #4 ; CHECK-NEXT: sub.w r7, r8, #4
Show All 12 Lines
; CHECK-NEXT: vmullb.u32 q4, q3, q1 ; CHECK-NEXT: vmullb.u32 q4, q3, q1
; CHECK-NEXT: vmov.f32 s2, s1 ; CHECK-NEXT: vmov.f32 s2, s1
; CHECK-NEXT: vmov r4, r5, d8 ; CHECK-NEXT: vmov r4, r5, d8
; CHECK-NEXT: lsrl r4, r5, #31 ; CHECK-NEXT: lsrl r4, r5, #31
; CHECK-NEXT: vmov.f32 s10, s9 ; CHECK-NEXT: vmov.f32 s10, s9
; CHECK-NEXT: subs.w r6, r4, #-1 ; CHECK-NEXT: subs.w r6, r4, #-1
; CHECK-NEXT: sbcs r5, r5, #0 ; CHECK-NEXT: sbcs r5, r5, #0
; CHECK-NEXT: vmov r6, r7, d9 ; CHECK-NEXT: vmov r6, r7, d9
; CHECK-NEXT: mov.w r5, #0 ; CHECK-NEXT: cset r5, lo
; CHECK-NEXT: lsrl r6, r7, #31 ; CHECK-NEXT: lsrl r6, r7, #31
; CHECK-NEXT: it lo
; CHECK-NEXT: movlo r5, #1
; CHECK-NEXT: cmp r5, #0 ; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: vmov q3[2], q3[0], r4, r6
; CHECK-NEXT: csetm r11, ne ; CHECK-NEXT: csetm r11, ne
; CHECK-NEXT: subs.w r5, r6, #-1 ; CHECK-NEXT: subs.w r5, r6, #-1
; CHECK-NEXT: sbcs r5, r7, #0 ; CHECK-NEXT: sbcs r5, r7, #0
; CHECK-NEXT: vmov.32 q1[1], r11 ; CHECK-NEXT: vmov.32 q1[1], r11
; CHECK-NEXT: mov.w r5, #0 ; CHECK-NEXT: cset r5, lo
; CHECK-NEXT: vmov q3[2], q3[0], r4, r6
; CHECK-NEXT: it lo
; CHECK-NEXT: movlo r5, #1
; CHECK-NEXT: cmp r5, #0 ; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: csetm r5, ne ; CHECK-NEXT: csetm r5, ne
; CHECK-NEXT: vmov q1[2], q1[0], r11, r5 ; CHECK-NEXT: vmov q1[2], q1[0], r11, r5
; CHECK-NEXT: vand q3, q3, q1 ; CHECK-NEXT: vand q3, q3, q1
; CHECK-NEXT: vorn q1, q3, q1 ; CHECK-NEXT: vorn q1, q3, q1
; CHECK-NEXT: vmullb.u32 q3, q2, q0 ; CHECK-NEXT: vmullb.u32 q3, q2, q0
; CHECK-NEXT: vmov r4, r5, d6 ; CHECK-NEXT: vmov r4, r5, d6
; CHECK-NEXT: lsrl r4, r5, #31 ; CHECK-NEXT: lsrl r4, r5, #31
; CHECK-NEXT: subs.w r6, r4, #-1 ; CHECK-NEXT: subs.w r6, r4, #-1
; CHECK-NEXT: sbcs r5, r5, #0 ; CHECK-NEXT: sbcs r5, r5, #0
; CHECK-NEXT: vmov r6, r7, d7 ; CHECK-NEXT: vmov r6, r7, d7
; CHECK-NEXT: mov.w r5, #0 ; CHECK-NEXT: cset r5, lo
; CHECK-NEXT: lsrl r6, r7, #31 ; CHECK-NEXT: lsrl r6, r7, #31
; CHECK-NEXT: it lo
; CHECK-NEXT: movlo r5, #1
; CHECK-NEXT: cmp r5, #0 ; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: vmov q2[2], q2[0], r4, r6
; CHECK-NEXT: csetm r11, ne ; CHECK-NEXT: csetm r11, ne
; CHECK-NEXT: subs.w r5, r6, #-1 ; CHECK-NEXT: subs.w r5, r6, #-1
; CHECK-NEXT: sbcs r5, r7, #0 ; CHECK-NEXT: sbcs r5, r7, #0
; CHECK-NEXT: vmov.32 q0[1], r11 ; CHECK-NEXT: vmov.32 q0[1], r11
; CHECK-NEXT: mov.w r5, #0 ; CHECK-NEXT: cset r5, lo
; CHECK-NEXT: vmov q2[2], q2[0], r4, r6
; CHECK-NEXT: it lo
; CHECK-NEXT: movlo r5, #1
; CHECK-NEXT: cmp r5, #0 ; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: csetm r5, ne ; CHECK-NEXT: csetm r5, ne
; CHECK-NEXT: vmov q0[2], q0[0], r11, r5 ; CHECK-NEXT: vmov q0[2], q0[0], r11, r5
; CHECK-NEXT: vand q2, q2, q0 ; CHECK-NEXT: vand q2, q2, q0
; CHECK-NEXT: vorn q0, q2, q0 ; CHECK-NEXT: vorn q0, q2, q0
; CHECK-NEXT: vmov.f32 s1, s2 ; CHECK-NEXT: vmov.f32 s1, s2
; CHECK-NEXT: vmov.f32 s2, s4 ; CHECK-NEXT: vmov.f32 s2, s4
; CHECK-NEXT: vmov.f32 s3, s6 ; CHECK-NEXT: vmov.f32 s3, s6
▲ Show 20 Lines • Show All 2,377 Lines • Show Last 20 Lines

llvm/test/CodeGen/Thumb2/mve-vcmp.ll

Show First 20 Lines • Show All 388 Lines • ▼ Show 20 Linesentry:
%c = icmp eq <2 x i64> %src, %srcb %c = icmp eq <2 x i64> %src, %srcb
%s = select <2 x i1> %c, <2 x i64> %a, <2 x i64> %b %s = select <2 x i1> %c, <2 x i64> %a, <2 x i64> %b
ret <2 x i64> %s ret <2 x i64> %s
} }
define arm_aapcs_vfpcc <2 x i64> @vcmp_slt_v2i64(<2 x i64> %src, <2 x i64> %srcb, <2 x i64> %a, <2 x i64> %b) { define arm_aapcs_vfpcc <2 x i64> @vcmp_slt_v2i64(<2 x i64> %src, <2 x i64> %srcb, <2 x i64> %a, <2 x i64> %b) {
; CHECK-LABEL: vcmp_slt_v2i64: ; CHECK-LABEL: vcmp_slt_v2i64:
; CHECK: @ %bb.0: @ %entry ; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: .save {r7, lr} ; CHECK-NEXT: vmov r0, r1, d3
; CHECK-NEXT: push {r7, lr}
; CHECK-NEXT: vmov r0, r12, d3
; CHECK-NEXT: movs r1, #0
; CHECK-NEXT: vmov r2, r3, d1 ; CHECK-NEXT: vmov r2, r3, d1
; CHECK-NEXT: subs r0, r2, r0 ; CHECK-NEXT: subs r0, r2, r0
; CHECK-NEXT: sbcs.w r0, r3, r12 ; CHECK-NEXT: sbcs.w r0, r3, r1
; CHECK-NEXT: vmov lr, r12, d2 ; CHECK-NEXT: vmov r1, r12, d2
; CHECK-NEXT: vmov r3, r2, d0 ; CHECK-NEXT: vmov r3, r2, d0
; CHECK-NEXT: mov.w r0, #0 ; CHECK-NEXT: cset r0, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r0, #1
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: subs.w r3, r3, lr ; CHECK-NEXT: subs r1, r3, r1
; CHECK-NEXT: sbcs.w r2, r2, r12 ; CHECK-NEXT: sbcs.w r1, r2, r12
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r1, lt
; CHECK-NEXT: movlt r1, #1
; CHECK-NEXT: cmp r1, #0 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: vmov q0[2], q0[0], r1, r0 ; CHECK-NEXT: vmov q0[2], q0[0], r1, r0
; CHECK-NEXT: vmov q0[3], q0[1], r1, r0 ; CHECK-NEXT: vmov q0[3], q0[1], r1, r0
; CHECK-NEXT: vbic q1, q3, q0 ; CHECK-NEXT: vbic q1, q3, q0
; CHECK-NEXT: vand q0, q2, q0 ; CHECK-NEXT: vand q0, q2, q0
; CHECK-NEXT: vorr q0, q0, q1 ; CHECK-NEXT: vorr q0, q0, q1
; CHECK-NEXT: pop {r7, pc} ; CHECK-NEXT: bx lr
entry: entry:
%c = icmp slt <2 x i64> %src, %srcb %c = icmp slt <2 x i64> %src, %srcb
%s = select <2 x i1> %c, <2 x i64> %a, <2 x i64> %b %s = select <2 x i1> %c, <2 x i64> %a, <2 x i64> %b
ret <2 x i64> %s ret <2 x i64> %s
} }
define arm_aapcs_vfpcc <2 x i32> @vcmp_eq_v2i32(<2 x i64> %src, <2 x i64> %srcb, <2 x i32> %a, <2 x i32> %b) { define arm_aapcs_vfpcc <2 x i32> @vcmp_eq_v2i32(<2 x i64> %src, <2 x i64> %srcb, <2 x i32> %a, <2 x i32> %b) {
; CHECK-LABEL: vcmp_eq_v2i32: ; CHECK-LABEL: vcmp_eq_v2i32:
Show All 20 Linesentry:
%c = icmp eq <2 x i64> %src, %srcb %c = icmp eq <2 x i64> %src, %srcb
%s = select <2 x i1> %c, <2 x i32> %a, <2 x i32> %b %s = select <2 x i1> %c, <2 x i32> %a, <2 x i32> %b
ret <2 x i32> %s ret <2 x i32> %s
} }
define arm_aapcs_vfpcc <2 x i32> @vcmp_multi_v2i32(<2 x i64> %a, <2 x i32> %b, <2 x i32> %c) { define arm_aapcs_vfpcc <2 x i32> @vcmp_multi_v2i32(<2 x i64> %a, <2 x i32> %b, <2 x i32> %c) {
; CHECK-LABEL: vcmp_multi_v2i32: ; CHECK-LABEL: vcmp_multi_v2i32:
; CHECK: @ %bb.0: ; CHECK: @ %bb.0:
; CHECK-NEXT: .save {r4, lr} ; CHECK-NEXT: .save {r7, lr}
; CHECK-NEXT: push {r4, lr} ; CHECK-NEXT: push {r7, lr}
; CHECK-NEXT: .vsave {d8, d9} ; CHECK-NEXT: .vsave {d8, d9}
; CHECK-NEXT: vpush {d8, d9} ; CHECK-NEXT: vpush {d8, d9}
; CHECK-NEXT: vmov r0, r1, d1 ; CHECK-NEXT: vmov r0, r1, d1
; CHECK-NEXT: movs r3, #0
; CHECK-NEXT: orrs r0, r1 ; CHECK-NEXT: orrs r0, r1
; CHECK-NEXT: vmov r1, r2, d0 ; CHECK-NEXT: vmov r1, r2, d0
; CHECK-NEXT: csetm r0, eq ; CHECK-NEXT: csetm r0, eq
; CHECK-NEXT: orrs r1, r2 ; CHECK-NEXT: orrs r1, r2
; CHECK-NEXT: vmov r2, s10 ; CHECK-NEXT: vmov r2, s10
; CHECK-NEXT: csetm r1, eq ; CHECK-NEXT: csetm r1, eq
; CHECK-NEXT: vmov q0[2], q0[0], r1, r0 ; CHECK-NEXT: vmov q0[2], q0[0], r1, r0
; CHECK-NEXT: vmov q0[3], q0[1], r1, r0 ; CHECK-NEXT: vmov q0[3], q0[1], r1, r0
; CHECK-NEXT: vbic q0, q2, q0 ; CHECK-NEXT: vbic q0, q2, q0
; CHECK-NEXT: vmov r0, s2 ; CHECK-NEXT: vmov r0, s2
; CHECK-NEXT: subs r1, r0, r2 ; CHECK-NEXT: subs r3, r0, r2
; CHECK-NEXT: asr.w r12, r0, #31 ; CHECK-NEXT: asr.w r1, r0, #31
; CHECK-NEXT: sbcs.w r1, r12, r2, asr #31 ; CHECK-NEXT: vmov r3, s8
; CHECK-NEXT: mov.w r1, #0 ; CHECK-NEXT: sbcs.w r1, r1, r2, asr #31
; CHECK-NEXT: vmov r2, s0 ; CHECK-NEXT: vmov r2, s0
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r1, lt
; CHECK-NEXT: movlt r1, #1
; CHECK-NEXT: cmp r1, #0 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: vmov r1, s8
; CHECK-NEXT: csetm lr, ne ; CHECK-NEXT: csetm lr, ne
; CHECK-NEXT: subs r1, r2, r3
; CHECK-NEXT: asr.w r12, r2, #31 ; CHECK-NEXT: asr.w r12, r2, #31
; CHECK-NEXT: subs r4, r2, r1 ; CHECK-NEXT: sbcs.w r1, r12, r3, asr #31
; CHECK-NEXT: sbcs.w r1, r12, r1, asr #31 ; CHECK-NEXT: cset r1, lt
; CHECK-NEXT: it lt ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: movlt r3, #1
; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: vmov q3[2], q3[0], r1, lr ; CHECK-NEXT: vmov q3[2], q3[0], r1, lr
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: cmp r2, #0 ; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: vmov q3[3], q3[1], r1, lr ; CHECK-NEXT: vmov q3[3], q3[1], r1, lr
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: vmov q4[2], q4[0], r1, r0 ; CHECK-NEXT: vmov q4[2], q4[0], r1, r0
; CHECK-NEXT: vmov q4[3], q4[1], r1, r0 ; CHECK-NEXT: vmov q4[3], q4[1], r1, r0
; CHECK-NEXT: vmov r0, s6 ; CHECK-NEXT: vmov r0, s6
; CHECK-NEXT: vmov r1, s4 ; CHECK-NEXT: vmov r1, s4
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: cmp r1, #0 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: vmov q1[2], q1[0], r1, r0 ; CHECK-NEXT: vmov q1[2], q1[0], r1, r0
; CHECK-NEXT: vmov q1[3], q1[1], r1, r0 ; CHECK-NEXT: vmov q1[3], q1[1], r1, r0
; CHECK-NEXT: vand q1, q1, q4 ; CHECK-NEXT: vand q1, q1, q4
; CHECK-NEXT: vand q1, q3, q1 ; CHECK-NEXT: vand q1, q3, q1
; CHECK-NEXT: vbic q0, q0, q1 ; CHECK-NEXT: vbic q0, q0, q1
; CHECK-NEXT: vand q1, q2, q1 ; CHECK-NEXT: vand q1, q2, q1
; CHECK-NEXT: vorr q0, q1, q0 ; CHECK-NEXT: vorr q0, q1, q0
; CHECK-NEXT: vpop {d8, d9} ; CHECK-NEXT: vpop {d8, d9}
; CHECK-NEXT: pop {r4, pc} ; CHECK-NEXT: pop {r7, pc}
%a4 = icmp eq <2 x i64> %a, zeroinitializer %a4 = icmp eq <2 x i64> %a, zeroinitializer
%a5 = select <2 x i1> %a4, <2 x i32> zeroinitializer, <2 x i32> %c %a5 = select <2 x i1> %a4, <2 x i32> zeroinitializer, <2 x i32> %c
%a6 = icmp ne <2 x i32> %b, zeroinitializer %a6 = icmp ne <2 x i32> %b, zeroinitializer
%a7 = icmp slt <2 x i32> %a5, %c %a7 = icmp slt <2 x i32> %a5, %c
%a8 = icmp ne <2 x i32> %a5, zeroinitializer %a8 = icmp ne <2 x i32> %a5, zeroinitializer
%a9 = and <2 x i1> %a6, %a8 %a9 = and <2 x i1> %a6, %a8
%a10 = and <2 x i1> %a7, %a9 %a10 = and <2 x i1> %a7, %a9
%a11 = select <2 x i1> %a10, <2 x i32> %c, <2 x i32> %a5 %a11 = select <2 x i1> %a10, <2 x i32> %c, <2 x i32> %a5
ret <2 x i32> %a11 ret <2 x i32> %a11
} }

llvm/test/CodeGen/Thumb2/mve-vcmpf.ll

; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -mtriple=thumbv8.1m.main-none-none-eabi -mattr=+mve,+fullfp16 -verify-machineinstrs %s -o - | FileCheck %s --check-prefix=CHECK-MVE ; RUN: llc -mtriple=thumbv8.1m.main-none-none-eabi -mattr=+mve,+fullfp16 -verify-machineinstrs %s -o - | FileCheck %s --check-prefix=CHECK-MVE
; RUN: llc -mtriple=thumbv8.1m.main-none-none-eabi -mattr=+mve.fp -verify-machineinstrs %s -o - | FileCheck %s --check-prefix=CHECK-MVEFP ; RUN: llc -mtriple=thumbv8.1m.main-none-none-eabi -mattr=+mve.fp -verify-machineinstrs %s -o - | FileCheck %s --check-prefix=CHECK-MVEFP
define arm_aapcs_vfpcc <4 x float> @vcmp_oeq_v4f32(<4 x float> %src, <4 x float> %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_oeq_v4f32(<4 x float> %src, <4 x float> %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_oeq_v4f32: ; CHECK-MVE-LABEL: vcmp_oeq_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s5 ; CHECK-MVE-NEXT: vcmp.f32 s1, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s4 ; CHECK-MVE-NEXT: vcmp.f32 s0, s4
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, s7 ; CHECK-MVE-NEXT: vcmp.f32 s3, s7
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r1, eq
; CHECK-MVE-NEXT: moveq r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s6 ; CHECK-MVE-NEXT: vcmp.f32 s2, s6
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r2, eq
; CHECK-MVE-NEXT: moveq r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r3, eq
; CHECK-MVE-NEXT: moveq r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_oeq_v4f32: ; CHECK-MVEFP-LABEL: vcmp_oeq_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f32 eq, q0, q1 ; CHECK-MVEFP-NEXT: vcmp.f32 eq, q0, q1
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp oeq <4 x float> %src, %src2 %c = fcmp oeq <4 x float> %src, %src2
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_one_v4f32(<4 x float> %src, <4 x float> %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_one_v4f32(<4 x float> %src, <4 x float> %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_one_v4f32: ; CHECK-MVE-LABEL: vcmp_one_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s5 ; CHECK-MVE-NEXT: vcmp.f32 s1, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s1, s5 ; CHECK-MVE-NEXT: vcmp.f32 s1, s5
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s4 ; CHECK-MVE-NEXT: vcmp.f32 s0, s4
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1 ; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r2, #1
; CHECK-MVE-NEXT: vcmp.f32 s3, s7 ; CHECK-MVE-NEXT: vcmp.f32 s3, s7
; CHECK-MVE-NEXT: mov.w r3, #0 ; CHECK-MVE-NEXT: cset r1, mi
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r2, #1 ; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r3, #1
; CHECK-MVE-NEXT: mov.w r0, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s6 ; CHECK-MVE-NEXT: vcmp.f32 s2, s6
; CHECK-MVE-NEXT: cset r2, mi
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r3, #1 ; CHECK-MVE-NEXT: movgt r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r3, mi
; CHECK-MVE-NEXT: movmi r0, #1
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r0, #1 ; CHECK-MVE-NEXT: movgt r3, #1
; CHECK-MVE-NEXT: cmp r3, #0 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_one_v4f32: ; CHECK-MVEFP-LABEL: vcmp_one_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vpt.f32 le, q1, q0 ; CHECK-MVEFP-NEXT: vpt.f32 le, q1, q0
; CHECK-MVEFP-NEXT: vcmpt.f32 le, q0, q1 ; CHECK-MVEFP-NEXT: vcmpt.f32 le, q0, q1
; CHECK-MVEFP-NEXT: vpsel q0, q3, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q3, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp one <4 x float> %src, %src2 %c = fcmp one <4 x float> %src, %src2
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_ogt_v4f32(<4 x float> %src, <4 x float> %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_ogt_v4f32(<4 x float> %src, <4 x float> %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_ogt_v4f32: ; CHECK-MVE-LABEL: vcmp_ogt_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s5 ; CHECK-MVE-NEXT: vcmp.f32 s1, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s4 ; CHECK-MVE-NEXT: vcmp.f32 s0, s4
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, s7 ; CHECK-MVE-NEXT: vcmp.f32 s3, s7
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r1, gt
; CHECK-MVE-NEXT: movgt r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s6 ; CHECK-MVE-NEXT: vcmp.f32 s2, s6
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r2, gt
; CHECK-MVE-NEXT: movgt r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r3, gt
; CHECK-MVE-NEXT: movgt r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ogt_v4f32: ; CHECK-MVEFP-LABEL: vcmp_ogt_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f32 gt, q0, q1 ; CHECK-MVEFP-NEXT: vcmp.f32 gt, q0, q1
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ogt <4 x float> %src, %src2 %c = fcmp ogt <4 x float> %src, %src2
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_oge_v4f32(<4 x float> %src, <4 x float> %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_oge_v4f32(<4 x float> %src, <4 x float> %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_oge_v4f32: ; CHECK-MVE-LABEL: vcmp_oge_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s5 ; CHECK-MVE-NEXT: vcmp.f32 s1, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s4 ; CHECK-MVE-NEXT: vcmp.f32 s0, s4
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, s7 ; CHECK-MVE-NEXT: vcmp.f32 s3, s7
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r1, ge
; CHECK-MVE-NEXT: movge r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s6 ; CHECK-MVE-NEXT: vcmp.f32 s2, s6
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r2, ge
; CHECK-MVE-NEXT: movge r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r3, ge
; CHECK-MVE-NEXT: movge r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_oge_v4f32: ; CHECK-MVEFP-LABEL: vcmp_oge_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f32 ge, q0, q1 ; CHECK-MVEFP-NEXT: vcmp.f32 ge, q0, q1
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp oge <4 x float> %src, %src2 %c = fcmp oge <4 x float> %src, %src2
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_olt_v4f32(<4 x float> %src, <4 x float> %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_olt_v4f32(<4 x float> %src, <4 x float> %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_olt_v4f32: ; CHECK-MVE-LABEL: vcmp_olt_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s5 ; CHECK-MVE-NEXT: vcmp.f32 s1, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s4 ; CHECK-MVE-NEXT: vcmp.f32 s0, s4
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, s7 ; CHECK-MVE-NEXT: vcmp.f32 s3, s7
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r1, mi
; CHECK-MVE-NEXT: movmi r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s6 ; CHECK-MVE-NEXT: vcmp.f32 s2, s6
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r2, mi
; CHECK-MVE-NEXT: movmi r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r3, mi
; CHECK-MVE-NEXT: movmi r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_olt_v4f32: ; CHECK-MVEFP-LABEL: vcmp_olt_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f32 gt, q1, q0 ; CHECK-MVEFP-NEXT: vcmp.f32 gt, q1, q0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp olt <4 x float> %src, %src2 %c = fcmp olt <4 x float> %src, %src2
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_ole_v4f32(<4 x float> %src, <4 x float> %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_ole_v4f32(<4 x float> %src, <4 x float> %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_ole_v4f32: ; CHECK-MVE-LABEL: vcmp_ole_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s5 ; CHECK-MVE-NEXT: vcmp.f32 s1, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s4 ; CHECK-MVE-NEXT: vcmp.f32 s0, s4
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, s7 ; CHECK-MVE-NEXT: vcmp.f32 s3, s7
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r1, ls
; CHECK-MVE-NEXT: movls r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s6 ; CHECK-MVE-NEXT: vcmp.f32 s2, s6
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r2, ls
; CHECK-MVE-NEXT: movls r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r3, ls
; CHECK-MVE-NEXT: movls r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ole_v4f32: ; CHECK-MVEFP-LABEL: vcmp_ole_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f32 ge, q1, q0 ; CHECK-MVEFP-NEXT: vcmp.f32 ge, q1, q0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ole <4 x float> %src, %src2 %c = fcmp ole <4 x float> %src, %src2
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_ueq_v4f32(<4 x float> %src, <4 x float> %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_ueq_v4f32(<4 x float> %src, <4 x float> %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_ueq_v4f32: ; CHECK-MVE-LABEL: vcmp_ueq_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s5 ; CHECK-MVE-NEXT: vcmp.f32 s1, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s1, s5 ; CHECK-MVE-NEXT: vcmp.f32 s1, s5
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s4 ; CHECK-MVE-NEXT: vcmp.f32 s0, s4
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1 ; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r2, #1
; CHECK-MVE-NEXT: vcmp.f32 s3, s7 ; CHECK-MVE-NEXT: vcmp.f32 s3, s7
; CHECK-MVE-NEXT: mov.w r3, #0 ; CHECK-MVE-NEXT: cset r1, eq
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r2, #1 ; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r3, #1
; CHECK-MVE-NEXT: mov.w r0, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s6 ; CHECK-MVE-NEXT: vcmp.f32 s2, s6
; CHECK-MVE-NEXT: cset r2, eq
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r3, #1 ; CHECK-MVE-NEXT: movvs r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r3, eq
; CHECK-MVE-NEXT: moveq r0, #1
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r0, #1 ; CHECK-MVE-NEXT: movvs r3, #1
; CHECK-MVE-NEXT: cmp r3, #0 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ueq_v4f32: ; CHECK-MVEFP-LABEL: vcmp_ueq_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vpt.f32 le, q1, q0 ; CHECK-MVEFP-NEXT: vpt.f32 le, q1, q0
; CHECK-MVEFP-NEXT: vcmpt.f32 le, q0, q1 ; CHECK-MVEFP-NEXT: vcmpt.f32 le, q0, q1
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ueq <4 x float> %src, %src2 %c = fcmp ueq <4 x float> %src, %src2
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_une_v4f32(<4 x float> %src, <4 x float> %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_une_v4f32(<4 x float> %src, <4 x float> %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_une_v4f32: ; CHECK-MVE-LABEL: vcmp_une_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s5 ; CHECK-MVE-NEXT: vcmp.f32 s1, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s4 ; CHECK-MVE-NEXT: vcmp.f32 s0, s4
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, s7 ; CHECK-MVE-NEXT: vcmp.f32 s3, s7
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r1, ne
; CHECK-MVE-NEXT: movne r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s6 ; CHECK-MVE-NEXT: vcmp.f32 s2, s6
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r2, ne
; CHECK-MVE-NEXT: movne r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r3, ne
; CHECK-MVE-NEXT: movne r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_une_v4f32: ; CHECK-MVEFP-LABEL: vcmp_une_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f32 ne, q0, q1 ; CHECK-MVEFP-NEXT: vcmp.f32 ne, q0, q1
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp une <4 x float> %src, %src2 %c = fcmp une <4 x float> %src, %src2
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_ugt_v4f32(<4 x float> %src, <4 x float> %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_ugt_v4f32(<4 x float> %src, <4 x float> %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_ugt_v4f32: ; CHECK-MVE-LABEL: vcmp_ugt_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s5 ; CHECK-MVE-NEXT: vcmp.f32 s1, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s4 ; CHECK-MVE-NEXT: vcmp.f32 s0, s4
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, s7 ; CHECK-MVE-NEXT: vcmp.f32 s3, s7
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r1, hi
; CHECK-MVE-NEXT: movhi r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s6 ; CHECK-MVE-NEXT: vcmp.f32 s2, s6
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r2, hi
; CHECK-MVE-NEXT: movhi r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r3, hi
; CHECK-MVE-NEXT: movhi r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ugt_v4f32: ; CHECK-MVEFP-LABEL: vcmp_ugt_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f32 lt, q1, q0 ; CHECK-MVEFP-NEXT: vcmp.f32 lt, q1, q0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ugt <4 x float> %src, %src2 %c = fcmp ugt <4 x float> %src, %src2
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_uge_v4f32(<4 x float> %src, <4 x float> %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_uge_v4f32(<4 x float> %src, <4 x float> %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_uge_v4f32: ; CHECK-MVE-LABEL: vcmp_uge_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s5 ; CHECK-MVE-NEXT: vcmp.f32 s1, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s4 ; CHECK-MVE-NEXT: vcmp.f32 s0, s4
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, s7 ; CHECK-MVE-NEXT: vcmp.f32 s3, s7
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r1, pl
; CHECK-MVE-NEXT: movpl r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s6 ; CHECK-MVE-NEXT: vcmp.f32 s2, s6
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r2, pl
; CHECK-MVE-NEXT: movpl r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r3, pl
; CHECK-MVE-NEXT: movpl r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_uge_v4f32: ; CHECK-MVEFP-LABEL: vcmp_uge_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f32 le, q1, q0 ; CHECK-MVEFP-NEXT: vcmp.f32 le, q1, q0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp uge <4 x float> %src, %src2 %c = fcmp uge <4 x float> %src, %src2
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_ult_v4f32(<4 x float> %src, <4 x float> %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_ult_v4f32(<4 x float> %src, <4 x float> %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_ult_v4f32: ; CHECK-MVE-LABEL: vcmp_ult_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s5 ; CHECK-MVE-NEXT: vcmp.f32 s1, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s4 ; CHECK-MVE-NEXT: vcmp.f32 s0, s4
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, s7 ; CHECK-MVE-NEXT: vcmp.f32 s3, s7
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r1, lt
; CHECK-MVE-NEXT: movlt r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s6 ; CHECK-MVE-NEXT: vcmp.f32 s2, s6
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r2, lt
; CHECK-MVE-NEXT: movlt r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r3, lt
; CHECK-MVE-NEXT: movlt r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ult_v4f32: ; CHECK-MVEFP-LABEL: vcmp_ult_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f32 lt, q0, q1 ; CHECK-MVEFP-NEXT: vcmp.f32 lt, q0, q1
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ult <4 x float> %src, %src2 %c = fcmp ult <4 x float> %src, %src2
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_ule_v4f32(<4 x float> %src, <4 x float> %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_ule_v4f32(<4 x float> %src, <4 x float> %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_ule_v4f32: ; CHECK-MVE-LABEL: vcmp_ule_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s5 ; CHECK-MVE-NEXT: vcmp.f32 s1, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s4 ; CHECK-MVE-NEXT: vcmp.f32 s0, s4
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, s7 ; CHECK-MVE-NEXT: vcmp.f32 s3, s7
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r1, le
; CHECK-MVE-NEXT: movle r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s6 ; CHECK-MVE-NEXT: vcmp.f32 s2, s6
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r2, le
; CHECK-MVE-NEXT: movle r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r3, le
; CHECK-MVE-NEXT: movle r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ule_v4f32: ; CHECK-MVEFP-LABEL: vcmp_ule_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f32 le, q0, q1 ; CHECK-MVEFP-NEXT: vcmp.f32 le, q0, q1
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ule <4 x float> %src, %src2 %c = fcmp ule <4 x float> %src, %src2
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_ord_v4f32(<4 x float> %src, <4 x float> %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_ord_v4f32(<4 x float> %src, <4 x float> %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_ord_v4f32: ; CHECK-MVE-LABEL: vcmp_ord_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s5 ; CHECK-MVE-NEXT: vcmp.f32 s1, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s4 ; CHECK-MVE-NEXT: vcmp.f32 s0, s4
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, s7 ; CHECK-MVE-NEXT: vcmp.f32 s3, s7
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r1, vc
; CHECK-MVE-NEXT: movvc r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s6 ; CHECK-MVE-NEXT: vcmp.f32 s2, s6
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r2, vc
; CHECK-MVE-NEXT: movvc r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r3, vc
; CHECK-MVE-NEXT: movvc r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ord_v4f32: ; CHECK-MVEFP-LABEL: vcmp_ord_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vpt.f32 le, q1, q0 ; CHECK-MVEFP-NEXT: vpt.f32 le, q1, q0
; CHECK-MVEFP-NEXT: vcmpt.f32 lt, q0, q1 ; CHECK-MVEFP-NEXT: vcmpt.f32 lt, q0, q1
; CHECK-MVEFP-NEXT: vpsel q0, q3, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q3, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ord <4 x float> %src, %src2 %c = fcmp ord <4 x float> %src, %src2
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_uno_v4f32(<4 x float> %src, <4 x float> %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_uno_v4f32(<4 x float> %src, <4 x float> %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_uno_v4f32: ; CHECK-MVE-LABEL: vcmp_uno_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s5 ; CHECK-MVE-NEXT: vcmp.f32 s1, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s4 ; CHECK-MVE-NEXT: vcmp.f32 s0, s4
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, s7 ; CHECK-MVE-NEXT: vcmp.f32 s3, s7
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r1, vs
; CHECK-MVE-NEXT: movvs r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s6 ; CHECK-MVE-NEXT: vcmp.f32 s2, s6
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r2, vs
; CHECK-MVE-NEXT: movvs r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r3, vs
; CHECK-MVE-NEXT: movvs r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_uno_v4f32: ; CHECK-MVEFP-LABEL: vcmp_uno_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vpt.f32 le, q1, q0 ; CHECK-MVEFP-NEXT: vpt.f32 le, q1, q0
; CHECK-MVEFP-NEXT: vcmpt.f32 lt, q0, q1 ; CHECK-MVEFP-NEXT: vcmpt.f32 lt, q0, q1
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
Show All 9 Lines
define arm_aapcs_vfpcc <8 x half> @vcmp_oeq_v8f16(<8 x half> %src, <8 x half> %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_oeq_v8f16(<8 x half> %src, <8 x half> %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_oeq_v8f16: ; CHECK-MVE-LABEL: vcmp_oeq_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: .vsave {d8, d9} ; CHECK-MVE-NEXT: .vsave {d8, d9}
; CHECK-MVE-NEXT: vpush {d8, d9} ; CHECK-MVE-NEXT: vpush {d8, d9}
; CHECK-MVE-NEXT: vmovx.f16 s16, s4 ; CHECK-MVE-NEXT: vmovx.f16 s16, s4
; CHECK-MVE-NEXT: vmovx.f16 s18, s0 ; CHECK-MVE-NEXT: vmovx.f16 s18, s0
; CHECK-MVE-NEXT: vcmp.f16 s18, s16 ; CHECK-MVE-NEXT: vcmp.f16 s18, s16
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s16, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s0, s4 ; CHECK-MVE-NEXT: vcmp.f16 s0, s4
; CHECK-MVE-NEXT: vmovx.f16 s16, s8
; CHECK-MVE-NEXT: vmovx.f16 s18, s12 ; CHECK-MVE-NEXT: vmovx.f16 s18, s12
; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s16, s18, s16 ; CHECK-MVE-NEXT: vseleq.f16 s16, s18, s16
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: vmovx.f16 s8, s1 ; CHECK-MVE-NEXT: vmovx.f16 s8, s1
; CHECK-MVE-NEXT: vcmp.f16 s8, s4 ; CHECK-MVE-NEXT: vcmp.f16 s8, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s9 ; CHECK-MVE-NEXT: vmovx.f16 s4, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s13 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s1, s5 ; CHECK-MVE-NEXT: vcmp.f16 s1, s5
; CHECK-MVE-NEXT: vins.f16 s0, s16
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s2 ; CHECK-MVE-NEXT: vmovx.f16 s8, s2
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vcmp.f16 s8, s4 ; CHECK-MVE-NEXT: vcmp.f16 s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s10 ; CHECK-MVE-NEXT: vmovx.f16 s4, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s2, s6 ; CHECK-MVE-NEXT: vcmp.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s4, s11 ; CHECK-MVE-NEXT: vmovx.f16 s4, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s6, s15 ; CHECK-MVE-NEXT: vmovx.f16 s6, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s7 ; CHECK-MVE-NEXT: vcmp.f16 s3, s7
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s16
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: vpop {d8, d9} ; CHECK-MVE-NEXT: vpop {d8, d9}
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_oeq_v8f16: ; CHECK-MVEFP-LABEL: vcmp_oeq_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f16 eq, q0, q1 ; CHECK-MVEFP-NEXT: vcmp.f16 eq, q0, q1
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp oeq <8 x half> %src, %src2 %c = fcmp oeq <8 x half> %src, %src2
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_one_v8f16(<8 x half> %src, <8 x half> %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_one_v8f16(<8 x half> %src, <8 x half> %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_one_v8f16: ; CHECK-MVE-LABEL: vcmp_one_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: .vsave {d8, d9} ; CHECK-MVE-NEXT: .vsave {d8, d9}
; CHECK-MVE-NEXT: vpush {d8, d9} ; CHECK-MVE-NEXT: vpush {d8, d9}
; CHECK-MVE-NEXT: vmovx.f16 s16, s4 ; CHECK-MVE-NEXT: vmovx.f16 s16, s4
; CHECK-MVE-NEXT: vmovx.f16 s18, s0 ; CHECK-MVE-NEXT: vmovx.f16 s18, s0
; CHECK-MVE-NEXT: vcmp.f16 s18, s16 ; CHECK-MVE-NEXT: vcmp.f16 s18, s16
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s16, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s0, s4 ; CHECK-MVE-NEXT: vcmp.f16 s0, s4
; CHECK-MVE-NEXT: vmovx.f16 s16, s8
; CHECK-MVE-NEXT: vmovx.f16 s18, s12 ; CHECK-MVE-NEXT: vmovx.f16 s18, s12
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s16, s18, s16 ; CHECK-MVE-NEXT: vseleq.f16 s16, s18, s16
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1 ; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: vmovx.f16 s8, s1 ; CHECK-MVE-NEXT: vmovx.f16 s8, s1
; CHECK-MVE-NEXT: vcmp.f16 s8, s4 ; CHECK-MVE-NEXT: vcmp.f16 s8, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s9 ; CHECK-MVE-NEXT: vmovx.f16 s4, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s13 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s1, s5 ; CHECK-MVE-NEXT: vcmp.f16 s1, s5
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vins.f16 s0, s16
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s8, s2 ; CHECK-MVE-NEXT: vmovx.f16 s8, s2
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1 ; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vcmp.f16 s8, s4 ; CHECK-MVE-NEXT: vcmp.f16 s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s10 ; CHECK-MVE-NEXT: vmovx.f16 s4, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s2, s6 ; CHECK-MVE-NEXT: vcmp.f16 s2, s6
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1 ; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s11 ; CHECK-MVE-NEXT: vmovx.f16 s4, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s6, s15 ; CHECK-MVE-NEXT: vmovx.f16 s6, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s7 ; CHECK-MVE-NEXT: vcmp.f16 s3, s7
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r0, #1
; CHECK-MVE-NEXT: vins.f16 s0, s16
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r0, #1 ; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: vpop {d8, d9} ; CHECK-MVE-NEXT: vpop {d8, d9}
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
Show All 12 Lines
define arm_aapcs_vfpcc <8 x half> @vcmp_ogt_v8f16(<8 x half> %src, <8 x half> %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_ogt_v8f16(<8 x half> %src, <8 x half> %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_ogt_v8f16: ; CHECK-MVE-LABEL: vcmp_ogt_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: .vsave {d8, d9} ; CHECK-MVE-NEXT: .vsave {d8, d9}
; CHECK-MVE-NEXT: vpush {d8, d9} ; CHECK-MVE-NEXT: vpush {d8, d9}
; CHECK-MVE-NEXT: vmovx.f16 s16, s4 ; CHECK-MVE-NEXT: vmovx.f16 s16, s4
; CHECK-MVE-NEXT: vmovx.f16 s18, s0 ; CHECK-MVE-NEXT: vmovx.f16 s18, s0
; CHECK-MVE-NEXT: vcmp.f16 s18, s16 ; CHECK-MVE-NEXT: vcmp.f16 s18, s16
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s16, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s0, s4 ; CHECK-MVE-NEXT: vcmp.f16 s0, s4
; CHECK-MVE-NEXT: vmovx.f16 s16, s8
; CHECK-MVE-NEXT: vmovx.f16 s18, s12 ; CHECK-MVE-NEXT: vmovx.f16 s18, s12
; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s16, s18, s16 ; CHECK-MVE-NEXT: vseleq.f16 s16, s18, s16
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: vmovx.f16 s8, s1 ; CHECK-MVE-NEXT: vmovx.f16 s8, s1
; CHECK-MVE-NEXT: vcmp.f16 s8, s4 ; CHECK-MVE-NEXT: vcmp.f16 s8, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s9 ; CHECK-MVE-NEXT: vmovx.f16 s4, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s13 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s1, s5 ; CHECK-MVE-NEXT: vcmp.f16 s1, s5
; CHECK-MVE-NEXT: vins.f16 s0, s16
; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s2 ; CHECK-MVE-NEXT: vmovx.f16 s8, s2
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: movgt r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vcmp.f16 s8, s4 ; CHECK-MVE-NEXT: vcmp.f16 s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s10 ; CHECK-MVE-NEXT: vmovx.f16 s4, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s2, s6 ; CHECK-MVE-NEXT: vcmp.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s4, s11 ; CHECK-MVE-NEXT: vmovx.f16 s4, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s6, s15 ; CHECK-MVE-NEXT: vmovx.f16 s6, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s7 ; CHECK-MVE-NEXT: vcmp.f16 s3, s7
; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s16
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: vpop {d8, d9} ; CHECK-MVE-NEXT: vpop {d8, d9}
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ogt_v8f16: ; CHECK-MVEFP-LABEL: vcmp_ogt_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f16 gt, q0, q1 ; CHECK-MVEFP-NEXT: vcmp.f16 gt, q0, q1
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ogt <8 x half> %src, %src2 %c = fcmp ogt <8 x half> %src, %src2
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_oge_v8f16(<8 x half> %src, <8 x half> %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_oge_v8f16(<8 x half> %src, <8 x half> %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_oge_v8f16: ; CHECK-MVE-LABEL: vcmp_oge_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: .vsave {d8, d9} ; CHECK-MVE-NEXT: .vsave {d8, d9}
; CHECK-MVE-NEXT: vpush {d8, d9} ; CHECK-MVE-NEXT: vpush {d8, d9}
; CHECK-MVE-NEXT: vmovx.f16 s16, s4 ; CHECK-MVE-NEXT: vmovx.f16 s16, s4
; CHECK-MVE-NEXT: vmovx.f16 s18, s0 ; CHECK-MVE-NEXT: vmovx.f16 s18, s0
; CHECK-MVE-NEXT: vcmp.f16 s18, s16 ; CHECK-MVE-NEXT: vcmp.f16 s18, s16
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s16, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ge
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s0, s4 ; CHECK-MVE-NEXT: vcmp.f16 s0, s4
; CHECK-MVE-NEXT: vmovx.f16 s16, s8
; CHECK-MVE-NEXT: vmovx.f16 s18, s12 ; CHECK-MVE-NEXT: vmovx.f16 s18, s12
; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s16, s18, s16 ; CHECK-MVE-NEXT: vseleq.f16 s16, s18, s16
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it ge
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: vmovx.f16 s8, s1 ; CHECK-MVE-NEXT: vmovx.f16 s8, s1
; CHECK-MVE-NEXT: vcmp.f16 s8, s4 ; CHECK-MVE-NEXT: vcmp.f16 s8, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ge
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s9 ; CHECK-MVE-NEXT: vmovx.f16 s4, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s13 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s1, s5 ; CHECK-MVE-NEXT: vcmp.f16 s1, s5
; CHECK-MVE-NEXT: vins.f16 s0, s16
; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s2 ; CHECK-MVE-NEXT: vmovx.f16 s8, s2
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: movge r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vcmp.f16 s8, s4 ; CHECK-MVE-NEXT: vcmp.f16 s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s10 ; CHECK-MVE-NEXT: vmovx.f16 s4, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ge
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s2, s6 ; CHECK-MVE-NEXT: vcmp.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ge
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s4, s11 ; CHECK-MVE-NEXT: vmovx.f16 s4, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s6, s15 ; CHECK-MVE-NEXT: vmovx.f16 s6, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s7 ; CHECK-MVE-NEXT: vcmp.f16 s3, s7
; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: movge r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s16
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: vpop {d8, d9} ; CHECK-MVE-NEXT: vpop {d8, d9}
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_oge_v8f16: ; CHECK-MVEFP-LABEL: vcmp_oge_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f16 ge, q0, q1 ; CHECK-MVEFP-NEXT: vcmp.f16 ge, q0, q1
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp oge <8 x half> %src, %src2 %c = fcmp oge <8 x half> %src, %src2
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_olt_v8f16(<8 x half> %src, <8 x half> %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_olt_v8f16(<8 x half> %src, <8 x half> %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_olt_v8f16: ; CHECK-MVE-LABEL: vcmp_olt_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: .vsave {d8, d9} ; CHECK-MVE-NEXT: .vsave {d8, d9}
; CHECK-MVE-NEXT: vpush {d8, d9} ; CHECK-MVE-NEXT: vpush {d8, d9}
; CHECK-MVE-NEXT: vmovx.f16 s16, s4 ; CHECK-MVE-NEXT: vmovx.f16 s16, s4
; CHECK-MVE-NEXT: vmovx.f16 s18, s0 ; CHECK-MVE-NEXT: vmovx.f16 s18, s0
; CHECK-MVE-NEXT: vcmp.f16 s18, s16 ; CHECK-MVE-NEXT: vcmp.f16 s18, s16
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s16, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s0, s4 ; CHECK-MVE-NEXT: vcmp.f16 s0, s4
; CHECK-MVE-NEXT: vmovx.f16 s16, s8
; CHECK-MVE-NEXT: vmovx.f16 s18, s12 ; CHECK-MVE-NEXT: vmovx.f16 s18, s12
; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s16, s18, s16 ; CHECK-MVE-NEXT: vseleq.f16 s16, s18, s16
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: vmovx.f16 s8, s1 ; CHECK-MVE-NEXT: vmovx.f16 s8, s1
; CHECK-MVE-NEXT: vcmp.f16 s8, s4 ; CHECK-MVE-NEXT: vcmp.f16 s8, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s9 ; CHECK-MVE-NEXT: vmovx.f16 s4, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s13 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s1, s5 ; CHECK-MVE-NEXT: vcmp.f16 s1, s5
; CHECK-MVE-NEXT: vins.f16 s0, s16
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s2 ; CHECK-MVE-NEXT: vmovx.f16 s8, s2
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vcmp.f16 s8, s4 ; CHECK-MVE-NEXT: vcmp.f16 s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s10 ; CHECK-MVE-NEXT: vmovx.f16 s4, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s2, s6 ; CHECK-MVE-NEXT: vcmp.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s4, s11 ; CHECK-MVE-NEXT: vmovx.f16 s4, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s6, s15 ; CHECK-MVE-NEXT: vmovx.f16 s6, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s7 ; CHECK-MVE-NEXT: vcmp.f16 s3, s7
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s16
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: vpop {d8, d9} ; CHECK-MVE-NEXT: vpop {d8, d9}
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_olt_v8f16: ; CHECK-MVEFP-LABEL: vcmp_olt_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f16 gt, q1, q0 ; CHECK-MVEFP-NEXT: vcmp.f16 gt, q1, q0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp olt <8 x half> %src, %src2 %c = fcmp olt <8 x half> %src, %src2
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_ole_v8f16(<8 x half> %src, <8 x half> %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_ole_v8f16(<8 x half> %src, <8 x half> %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_ole_v8f16: ; CHECK-MVE-LABEL: vcmp_ole_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: .vsave {d8, d9} ; CHECK-MVE-NEXT: .vsave {d8, d9}
; CHECK-MVE-NEXT: vpush {d8, d9} ; CHECK-MVE-NEXT: vpush {d8, d9}
; CHECK-MVE-NEXT: vmovx.f16 s16, s4 ; CHECK-MVE-NEXT: vmovx.f16 s16, s4
; CHECK-MVE-NEXT: vmovx.f16 s18, s0 ; CHECK-MVE-NEXT: vmovx.f16 s18, s0
; CHECK-MVE-NEXT: vcmp.f16 s18, s16 ; CHECK-MVE-NEXT: vcmp.f16 s18, s16
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s16, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ls
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s0, s4 ; CHECK-MVE-NEXT: vcmp.f16 s0, s4
; CHECK-MVE-NEXT: vmovx.f16 s16, s8
; CHECK-MVE-NEXT: vmovx.f16 s18, s12 ; CHECK-MVE-NEXT: vmovx.f16 s18, s12
; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s16, s18, s16 ; CHECK-MVE-NEXT: vseleq.f16 s16, s18, s16
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it ls
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: vmovx.f16 s8, s1 ; CHECK-MVE-NEXT: vmovx.f16 s8, s1
; CHECK-MVE-NEXT: vcmp.f16 s8, s4 ; CHECK-MVE-NEXT: vcmp.f16 s8, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ls
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s9 ; CHECK-MVE-NEXT: vmovx.f16 s4, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s13 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s1, s5 ; CHECK-MVE-NEXT: vcmp.f16 s1, s5
; CHECK-MVE-NEXT: vins.f16 s0, s16
; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s2 ; CHECK-MVE-NEXT: vmovx.f16 s8, s2
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: movls r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vcmp.f16 s8, s4 ; CHECK-MVE-NEXT: vcmp.f16 s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s10 ; CHECK-MVE-NEXT: vmovx.f16 s4, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ls
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s2, s6 ; CHECK-MVE-NEXT: vcmp.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ls
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s4, s11 ; CHECK-MVE-NEXT: vmovx.f16 s4, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s6, s15 ; CHECK-MVE-NEXT: vmovx.f16 s6, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s7 ; CHECK-MVE-NEXT: vcmp.f16 s3, s7
; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: movls r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s16
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: vpop {d8, d9} ; CHECK-MVE-NEXT: vpop {d8, d9}
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ole_v8f16: ; CHECK-MVEFP-LABEL: vcmp_ole_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f16 ge, q1, q0 ; CHECK-MVEFP-NEXT: vcmp.f16 ge, q1, q0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ole <8 x half> %src, %src2 %c = fcmp ole <8 x half> %src, %src2
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_ueq_v8f16(<8 x half> %src, <8 x half> %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_ueq_v8f16(<8 x half> %src, <8 x half> %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_ueq_v8f16: ; CHECK-MVE-LABEL: vcmp_ueq_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: .vsave {d8, d9} ; CHECK-MVE-NEXT: .vsave {d8, d9}
; CHECK-MVE-NEXT: vpush {d8, d9} ; CHECK-MVE-NEXT: vpush {d8, d9}
; CHECK-MVE-NEXT: vmovx.f16 s16, s4 ; CHECK-MVE-NEXT: vmovx.f16 s16, s4
; CHECK-MVE-NEXT: vmovx.f16 s18, s0 ; CHECK-MVE-NEXT: vmovx.f16 s18, s0
; CHECK-MVE-NEXT: vcmp.f16 s18, s16 ; CHECK-MVE-NEXT: vcmp.f16 s18, s16
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s16, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s0, s4 ; CHECK-MVE-NEXT: vcmp.f16 s0, s4
; CHECK-MVE-NEXT: vmovx.f16 s16, s8
; CHECK-MVE-NEXT: vmovx.f16 s18, s12 ; CHECK-MVE-NEXT: vmovx.f16 s18, s12
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s16, s18, s16 ; CHECK-MVE-NEXT: vseleq.f16 s16, s18, s16
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1 ; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: vmovx.f16 s8, s1 ; CHECK-MVE-NEXT: vmovx.f16 s8, s1
; CHECK-MVE-NEXT: vcmp.f16 s8, s4 ; CHECK-MVE-NEXT: vcmp.f16 s8, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s9 ; CHECK-MVE-NEXT: vmovx.f16 s4, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s13 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s1, s5 ; CHECK-MVE-NEXT: vcmp.f16 s1, s5
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vins.f16 s0, s16
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s8, s2 ; CHECK-MVE-NEXT: vmovx.f16 s8, s2
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1 ; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vcmp.f16 s8, s4 ; CHECK-MVE-NEXT: vcmp.f16 s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s10 ; CHECK-MVE-NEXT: vmovx.f16 s4, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s2, s6 ; CHECK-MVE-NEXT: vcmp.f16 s2, s6
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1 ; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s11 ; CHECK-MVE-NEXT: vmovx.f16 s4, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s6, s15 ; CHECK-MVE-NEXT: vmovx.f16 s6, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s7 ; CHECK-MVE-NEXT: vcmp.f16 s3, s7
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r0, #1
; CHECK-MVE-NEXT: vins.f16 s0, s16
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r0, #1 ; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: vpop {d8, d9} ; CHECK-MVE-NEXT: vpop {d8, d9}
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
Show All 12 Lines
define arm_aapcs_vfpcc <8 x half> @vcmp_une_v8f16(<8 x half> %src, <8 x half> %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_une_v8f16(<8 x half> %src, <8 x half> %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_une_v8f16: ; CHECK-MVE-LABEL: vcmp_une_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: .vsave {d8, d9} ; CHECK-MVE-NEXT: .vsave {d8, d9}
; CHECK-MVE-NEXT: vpush {d8, d9} ; CHECK-MVE-NEXT: vpush {d8, d9}
; CHECK-MVE-NEXT: vmovx.f16 s16, s4 ; CHECK-MVE-NEXT: vmovx.f16 s16, s4
; CHECK-MVE-NEXT: vmovx.f16 s18, s0 ; CHECK-MVE-NEXT: vmovx.f16 s18, s0
; CHECK-MVE-NEXT: vcmp.f16 s18, s16 ; CHECK-MVE-NEXT: vcmp.f16 s18, s16
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s16, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ne
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s0, s4 ; CHECK-MVE-NEXT: vcmp.f16 s0, s4
; CHECK-MVE-NEXT: vmovx.f16 s16, s8
; CHECK-MVE-NEXT: vmovx.f16 s18, s12 ; CHECK-MVE-NEXT: vmovx.f16 s18, s12
; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s16, s18, s16 ; CHECK-MVE-NEXT: vseleq.f16 s16, s18, s16
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it ne
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: vmovx.f16 s8, s1 ; CHECK-MVE-NEXT: vmovx.f16 s8, s1
; CHECK-MVE-NEXT: vcmp.f16 s8, s4 ; CHECK-MVE-NEXT: vcmp.f16 s8, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ne
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s9 ; CHECK-MVE-NEXT: vmovx.f16 s4, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s13 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s1, s5 ; CHECK-MVE-NEXT: vcmp.f16 s1, s5
; CHECK-MVE-NEXT: vins.f16 s0, s16
; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s2 ; CHECK-MVE-NEXT: vmovx.f16 s8, s2
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: movne r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vcmp.f16 s8, s4 ; CHECK-MVE-NEXT: vcmp.f16 s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s10 ; CHECK-MVE-NEXT: vmovx.f16 s4, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ne
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s2, s6 ; CHECK-MVE-NEXT: vcmp.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ne
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s4, s11 ; CHECK-MVE-NEXT: vmovx.f16 s4, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s6, s15 ; CHECK-MVE-NEXT: vmovx.f16 s6, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s7 ; CHECK-MVE-NEXT: vcmp.f16 s3, s7
; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: movne r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s16
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: vpop {d8, d9} ; CHECK-MVE-NEXT: vpop {d8, d9}
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_une_v8f16: ; CHECK-MVEFP-LABEL: vcmp_une_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f16 ne, q0, q1 ; CHECK-MVEFP-NEXT: vcmp.f16 ne, q0, q1
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp une <8 x half> %src, %src2 %c = fcmp une <8 x half> %src, %src2
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_ugt_v8f16(<8 x half> %src, <8 x half> %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_ugt_v8f16(<8 x half> %src, <8 x half> %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_ugt_v8f16: ; CHECK-MVE-LABEL: vcmp_ugt_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: .vsave {d8, d9} ; CHECK-MVE-NEXT: .vsave {d8, d9}
; CHECK-MVE-NEXT: vpush {d8, d9} ; CHECK-MVE-NEXT: vpush {d8, d9}
; CHECK-MVE-NEXT: vmovx.f16 s16, s4 ; CHECK-MVE-NEXT: vmovx.f16 s16, s4
; CHECK-MVE-NEXT: vmovx.f16 s18, s0 ; CHECK-MVE-NEXT: vmovx.f16 s18, s0
; CHECK-MVE-NEXT: vcmp.f16 s18, s16 ; CHECK-MVE-NEXT: vcmp.f16 s18, s16
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s16, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it hi
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s0, s4 ; CHECK-MVE-NEXT: vcmp.f16 s0, s4
; CHECK-MVE-NEXT: vmovx.f16 s16, s8
; CHECK-MVE-NEXT: vmovx.f16 s18, s12 ; CHECK-MVE-NEXT: vmovx.f16 s18, s12
; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s16, s18, s16 ; CHECK-MVE-NEXT: vseleq.f16 s16, s18, s16
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it hi
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: vmovx.f16 s8, s1 ; CHECK-MVE-NEXT: vmovx.f16 s8, s1
; CHECK-MVE-NEXT: vcmp.f16 s8, s4 ; CHECK-MVE-NEXT: vcmp.f16 s8, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it hi
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s9 ; CHECK-MVE-NEXT: vmovx.f16 s4, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s13 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s1, s5 ; CHECK-MVE-NEXT: vcmp.f16 s1, s5
; CHECK-MVE-NEXT: vins.f16 s0, s16
; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s2 ; CHECK-MVE-NEXT: vmovx.f16 s8, s2
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: movhi r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vcmp.f16 s8, s4 ; CHECK-MVE-NEXT: vcmp.f16 s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s10 ; CHECK-MVE-NEXT: vmovx.f16 s4, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it hi
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s2, s6 ; CHECK-MVE-NEXT: vcmp.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it hi
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s4, s11 ; CHECK-MVE-NEXT: vmovx.f16 s4, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s6, s15 ; CHECK-MVE-NEXT: vmovx.f16 s6, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s7 ; CHECK-MVE-NEXT: vcmp.f16 s3, s7
; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: movhi r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s16
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: vpop {d8, d9} ; CHECK-MVE-NEXT: vpop {d8, d9}
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ugt_v8f16: ; CHECK-MVEFP-LABEL: vcmp_ugt_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f16 lt, q1, q0 ; CHECK-MVEFP-NEXT: vcmp.f16 lt, q1, q0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ugt <8 x half> %src, %src2 %c = fcmp ugt <8 x half> %src, %src2
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_uge_v8f16(<8 x half> %src, <8 x half> %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_uge_v8f16(<8 x half> %src, <8 x half> %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_uge_v8f16: ; CHECK-MVE-LABEL: vcmp_uge_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: .vsave {d8, d9} ; CHECK-MVE-NEXT: .vsave {d8, d9}
; CHECK-MVE-NEXT: vpush {d8, d9} ; CHECK-MVE-NEXT: vpush {d8, d9}
; CHECK-MVE-NEXT: vmovx.f16 s16, s4 ; CHECK-MVE-NEXT: vmovx.f16 s16, s4
; CHECK-MVE-NEXT: vmovx.f16 s18, s0 ; CHECK-MVE-NEXT: vmovx.f16 s18, s0
; CHECK-MVE-NEXT: vcmp.f16 s18, s16 ; CHECK-MVE-NEXT: vcmp.f16 s18, s16
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s16, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it pl
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s0, s4 ; CHECK-MVE-NEXT: vcmp.f16 s0, s4
; CHECK-MVE-NEXT: vmovx.f16 s16, s8
; CHECK-MVE-NEXT: vmovx.f16 s18, s12 ; CHECK-MVE-NEXT: vmovx.f16 s18, s12
; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s16, s18, s16 ; CHECK-MVE-NEXT: vseleq.f16 s16, s18, s16
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it pl
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: vmovx.f16 s8, s1 ; CHECK-MVE-NEXT: vmovx.f16 s8, s1
; CHECK-MVE-NEXT: vcmp.f16 s8, s4 ; CHECK-MVE-NEXT: vcmp.f16 s8, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it pl
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s9 ; CHECK-MVE-NEXT: vmovx.f16 s4, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s13 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s1, s5 ; CHECK-MVE-NEXT: vcmp.f16 s1, s5
; CHECK-MVE-NEXT: vins.f16 s0, s16
; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s2 ; CHECK-MVE-NEXT: vmovx.f16 s8, s2
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: movpl r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vcmp.f16 s8, s4 ; CHECK-MVE-NEXT: vcmp.f16 s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s10 ; CHECK-MVE-NEXT: vmovx.f16 s4, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it pl
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s2, s6 ; CHECK-MVE-NEXT: vcmp.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it pl
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s4, s11 ; CHECK-MVE-NEXT: vmovx.f16 s4, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s6, s15 ; CHECK-MVE-NEXT: vmovx.f16 s6, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s7 ; CHECK-MVE-NEXT: vcmp.f16 s3, s7
; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: movpl r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s16
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: vpop {d8, d9} ; CHECK-MVE-NEXT: vpop {d8, d9}
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_uge_v8f16: ; CHECK-MVEFP-LABEL: vcmp_uge_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f16 le, q1, q0 ; CHECK-MVEFP-NEXT: vcmp.f16 le, q1, q0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp uge <8 x half> %src, %src2 %c = fcmp uge <8 x half> %src, %src2
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_ult_v8f16(<8 x half> %src, <8 x half> %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_ult_v8f16(<8 x half> %src, <8 x half> %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_ult_v8f16: ; CHECK-MVE-LABEL: vcmp_ult_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: .vsave {d8, d9} ; CHECK-MVE-NEXT: .vsave {d8, d9}
; CHECK-MVE-NEXT: vpush {d8, d9} ; CHECK-MVE-NEXT: vpush {d8, d9}
; CHECK-MVE-NEXT: vmovx.f16 s16, s4 ; CHECK-MVE-NEXT: vmovx.f16 s16, s4
; CHECK-MVE-NEXT: vmovx.f16 s18, s0 ; CHECK-MVE-NEXT: vmovx.f16 s18, s0
; CHECK-MVE-NEXT: vcmp.f16 s18, s16 ; CHECK-MVE-NEXT: vcmp.f16 s18, s16
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s16, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it lt
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s0, s4 ; CHECK-MVE-NEXT: vcmp.f16 s0, s4
; CHECK-MVE-NEXT: vmovx.f16 s16, s8
; CHECK-MVE-NEXT: vmovx.f16 s18, s12 ; CHECK-MVE-NEXT: vmovx.f16 s18, s12
; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s16, s18, s16 ; CHECK-MVE-NEXT: vseleq.f16 s16, s18, s16
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it lt
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: vmovx.f16 s8, s1 ; CHECK-MVE-NEXT: vmovx.f16 s8, s1
; CHECK-MVE-NEXT: vcmp.f16 s8, s4 ; CHECK-MVE-NEXT: vcmp.f16 s8, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it lt
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s9 ; CHECK-MVE-NEXT: vmovx.f16 s4, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s13 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s1, s5 ; CHECK-MVE-NEXT: vcmp.f16 s1, s5
; CHECK-MVE-NEXT: vins.f16 s0, s16
; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s2 ; CHECK-MVE-NEXT: vmovx.f16 s8, s2
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: movlt r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vcmp.f16 s8, s4 ; CHECK-MVE-NEXT: vcmp.f16 s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s10 ; CHECK-MVE-NEXT: vmovx.f16 s4, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it lt
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s2, s6 ; CHECK-MVE-NEXT: vcmp.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it lt
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s4, s11 ; CHECK-MVE-NEXT: vmovx.f16 s4, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s6, s15 ; CHECK-MVE-NEXT: vmovx.f16 s6, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s7 ; CHECK-MVE-NEXT: vcmp.f16 s3, s7
; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: movlt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s16
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: vpop {d8, d9} ; CHECK-MVE-NEXT: vpop {d8, d9}
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ult_v8f16: ; CHECK-MVEFP-LABEL: vcmp_ult_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f16 lt, q0, q1 ; CHECK-MVEFP-NEXT: vcmp.f16 lt, q0, q1
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ult <8 x half> %src, %src2 %c = fcmp ult <8 x half> %src, %src2
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_ule_v8f16(<8 x half> %src, <8 x half> %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_ule_v8f16(<8 x half> %src, <8 x half> %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_ule_v8f16: ; CHECK-MVE-LABEL: vcmp_ule_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: .vsave {d8, d9} ; CHECK-MVE-NEXT: .vsave {d8, d9}
; CHECK-MVE-NEXT: vpush {d8, d9} ; CHECK-MVE-NEXT: vpush {d8, d9}
; CHECK-MVE-NEXT: vmovx.f16 s16, s4 ; CHECK-MVE-NEXT: vmovx.f16 s16, s4
; CHECK-MVE-NEXT: vmovx.f16 s18, s0 ; CHECK-MVE-NEXT: vmovx.f16 s18, s0
; CHECK-MVE-NEXT: vcmp.f16 s18, s16 ; CHECK-MVE-NEXT: vcmp.f16 s18, s16
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s16, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it le
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s0, s4 ; CHECK-MVE-NEXT: vcmp.f16 s0, s4
; CHECK-MVE-NEXT: vmovx.f16 s16, s8
; CHECK-MVE-NEXT: vmovx.f16 s18, s12 ; CHECK-MVE-NEXT: vmovx.f16 s18, s12
; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s16, s18, s16 ; CHECK-MVE-NEXT: vseleq.f16 s16, s18, s16
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it le
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: vmovx.f16 s8, s1 ; CHECK-MVE-NEXT: vmovx.f16 s8, s1
; CHECK-MVE-NEXT: vcmp.f16 s8, s4 ; CHECK-MVE-NEXT: vcmp.f16 s8, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it le
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s9 ; CHECK-MVE-NEXT: vmovx.f16 s4, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s13 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s1, s5 ; CHECK-MVE-NEXT: vcmp.f16 s1, s5
; CHECK-MVE-NEXT: vins.f16 s0, s16
; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s2 ; CHECK-MVE-NEXT: vmovx.f16 s8, s2
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: movle r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vcmp.f16 s8, s4 ; CHECK-MVE-NEXT: vcmp.f16 s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s10 ; CHECK-MVE-NEXT: vmovx.f16 s4, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it le
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s2, s6 ; CHECK-MVE-NEXT: vcmp.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it le
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s4, s11 ; CHECK-MVE-NEXT: vmovx.f16 s4, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s6, s15 ; CHECK-MVE-NEXT: vmovx.f16 s6, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s7 ; CHECK-MVE-NEXT: vcmp.f16 s3, s7
; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: movle r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s16
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: vpop {d8, d9} ; CHECK-MVE-NEXT: vpop {d8, d9}
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ule_v8f16: ; CHECK-MVEFP-LABEL: vcmp_ule_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f16 le, q0, q1 ; CHECK-MVEFP-NEXT: vcmp.f16 le, q0, q1
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ule <8 x half> %src, %src2 %c = fcmp ule <8 x half> %src, %src2
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_ord_v8f16(<8 x half> %src, <8 x half> %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_ord_v8f16(<8 x half> %src, <8 x half> %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_ord_v8f16: ; CHECK-MVE-LABEL: vcmp_ord_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: .vsave {d8, d9} ; CHECK-MVE-NEXT: .vsave {d8, d9}
; CHECK-MVE-NEXT: vpush {d8, d9} ; CHECK-MVE-NEXT: vpush {d8, d9}
; CHECK-MVE-NEXT: vmovx.f16 s16, s4 ; CHECK-MVE-NEXT: vmovx.f16 s16, s4
; CHECK-MVE-NEXT: vmovx.f16 s18, s0 ; CHECK-MVE-NEXT: vmovx.f16 s18, s0
; CHECK-MVE-NEXT: vcmp.f16 s18, s16 ; CHECK-MVE-NEXT: vcmp.f16 s18, s16
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s16, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vc
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s0, s4 ; CHECK-MVE-NEXT: vcmp.f16 s0, s4
; CHECK-MVE-NEXT: vmovx.f16 s16, s8
; CHECK-MVE-NEXT: vmovx.f16 s18, s12 ; CHECK-MVE-NEXT: vmovx.f16 s18, s12
; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s16, s18, s16 ; CHECK-MVE-NEXT: vseleq.f16 s16, s18, s16
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it vc
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: vmovx.f16 s8, s1 ; CHECK-MVE-NEXT: vmovx.f16 s8, s1
; CHECK-MVE-NEXT: vcmp.f16 s8, s4 ; CHECK-MVE-NEXT: vcmp.f16 s8, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vc
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s9 ; CHECK-MVE-NEXT: vmovx.f16 s4, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s13 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s1, s5 ; CHECK-MVE-NEXT: vcmp.f16 s1, s5
; CHECK-MVE-NEXT: vins.f16 s0, s16
; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s2 ; CHECK-MVE-NEXT: vmovx.f16 s8, s2
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: movvc r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vcmp.f16 s8, s4 ; CHECK-MVE-NEXT: vcmp.f16 s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s10 ; CHECK-MVE-NEXT: vmovx.f16 s4, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vc
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s2, s6 ; CHECK-MVE-NEXT: vcmp.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vc
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s4, s11 ; CHECK-MVE-NEXT: vmovx.f16 s4, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s6, s15 ; CHECK-MVE-NEXT: vmovx.f16 s6, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s7 ; CHECK-MVE-NEXT: vcmp.f16 s3, s7
; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: movvc r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s16
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: vpop {d8, d9} ; CHECK-MVE-NEXT: vpop {d8, d9}
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ord_v8f16: ; CHECK-MVEFP-LABEL: vcmp_ord_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vpt.f16 le, q1, q0 ; CHECK-MVEFP-NEXT: vpt.f16 le, q1, q0
Show All 9 Lines
define arm_aapcs_vfpcc <8 x half> @vcmp_uno_v8f16(<8 x half> %src, <8 x half> %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_uno_v8f16(<8 x half> %src, <8 x half> %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_uno_v8f16: ; CHECK-MVE-LABEL: vcmp_uno_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: .vsave {d8, d9} ; CHECK-MVE-NEXT: .vsave {d8, d9}
; CHECK-MVE-NEXT: vpush {d8, d9} ; CHECK-MVE-NEXT: vpush {d8, d9}
; CHECK-MVE-NEXT: vmovx.f16 s16, s4 ; CHECK-MVE-NEXT: vmovx.f16 s16, s4
; CHECK-MVE-NEXT: vmovx.f16 s18, s0 ; CHECK-MVE-NEXT: vmovx.f16 s18, s0
; CHECK-MVE-NEXT: vcmp.f16 s18, s16 ; CHECK-MVE-NEXT: vcmp.f16 s18, s16
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s16, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s0, s4 ; CHECK-MVE-NEXT: vcmp.f16 s0, s4
; CHECK-MVE-NEXT: vmovx.f16 s16, s8
; CHECK-MVE-NEXT: vmovx.f16 s18, s12 ; CHECK-MVE-NEXT: vmovx.f16 s18, s12
; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s16, s18, s16 ; CHECK-MVE-NEXT: vseleq.f16 s16, s18, s16
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: vmovx.f16 s8, s1 ; CHECK-MVE-NEXT: vmovx.f16 s8, s1
; CHECK-MVE-NEXT: vcmp.f16 s8, s4 ; CHECK-MVE-NEXT: vcmp.f16 s8, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s9 ; CHECK-MVE-NEXT: vmovx.f16 s4, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s13 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s1, s5 ; CHECK-MVE-NEXT: vcmp.f16 s1, s5
; CHECK-MVE-NEXT: vins.f16 s0, s16
; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s2 ; CHECK-MVE-NEXT: vmovx.f16 s8, s2
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: movvs r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vcmp.f16 s8, s4 ; CHECK-MVE-NEXT: vcmp.f16 s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s10 ; CHECK-MVE-NEXT: vmovx.f16 s4, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s2, s6 ; CHECK-MVE-NEXT: vcmp.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s4, s11 ; CHECK-MVE-NEXT: vmovx.f16 s4, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s6, s15 ; CHECK-MVE-NEXT: vmovx.f16 s6, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s7 ; CHECK-MVE-NEXT: vcmp.f16 s3, s7
; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s16
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: vpop {d8, d9} ; CHECK-MVE-NEXT: vpop {d8, d9}
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_uno_v8f16: ; CHECK-MVEFP-LABEL: vcmp_uno_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vpt.f16 le, q1, q0 ; CHECK-MVEFP-NEXT: vpt.f16 le, q1, q0
; CHECK-MVEFP-NEXT: vcmpt.f16 lt, q0, q1 ; CHECK-MVEFP-NEXT: vcmpt.f16 lt, q0, q1
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp uno <8 x half> %src, %src2 %c = fcmp uno <8 x half> %src, %src2
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }

llvm/test/CodeGen/Thumb2/mve-vcmpfr.ll

; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -mtriple=thumbv8.1m.main-none-none-eabi -mattr=+mve,+fullfp16 -verify-machineinstrs %s -o - | FileCheck %s --check-prefix=CHECK-MVE ; RUN: llc -mtriple=thumbv8.1m.main-none-none-eabi -mattr=+mve,+fullfp16 -verify-machineinstrs %s -o - | FileCheck %s --check-prefix=CHECK-MVE
; RUN: llc -mtriple=thumbv8.1m.main-none-none-eabi -mattr=+mve.fp -verify-machineinstrs %s -o - | FileCheck %s --check-prefix=CHECK-MVEFP ; RUN: llc -mtriple=thumbv8.1m.main-none-none-eabi -mattr=+mve.fp -verify-machineinstrs %s -o - | FileCheck %s --check-prefix=CHECK-MVEFP
define arm_aapcs_vfpcc <4 x float> @vcmp_oeq_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_oeq_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_oeq_v4f32: ; CHECK-MVE-LABEL: vcmp_oeq_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s4 ; CHECK-MVE-NEXT: vcmp.f32 s1, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s4 ; CHECK-MVE-NEXT: vcmp.f32 s0, s4
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, s4 ; CHECK-MVE-NEXT: vcmp.f32 s3, s4
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r1, eq
; CHECK-MVE-NEXT: moveq r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s4 ; CHECK-MVE-NEXT: vcmp.f32 s2, s4
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r2, eq
; CHECK-MVE-NEXT: moveq r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r3, eq
; CHECK-MVE-NEXT: moveq r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_oeq_v4f32: ; CHECK-MVEFP-LABEL: vcmp_oeq_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov r0, s4 ; CHECK-MVEFP-NEXT: vmov r0, s4
; CHECK-MVEFP-NEXT: vcmp.f32 eq, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f32 eq, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <4 x float> undef, float %src2, i32 0 %i = insertelement <4 x float> undef, float %src2, i32 0
%sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer %sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer
%c = fcmp oeq <4 x float> %src, %sp %c = fcmp oeq <4 x float> %src, %sp
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_one_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_one_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_one_v4f32: ; CHECK-MVE-LABEL: vcmp_one_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s4 ; CHECK-MVE-NEXT: vcmp.f32 s1, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s1, s4 ; CHECK-MVE-NEXT: vcmp.f32 s1, s4
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s4 ; CHECK-MVE-NEXT: vcmp.f32 s0, s4
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1 ; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r2, #1
; CHECK-MVE-NEXT: vcmp.f32 s3, s4 ; CHECK-MVE-NEXT: vcmp.f32 s3, s4
; CHECK-MVE-NEXT: mov.w r3, #0 ; CHECK-MVE-NEXT: cset r1, mi
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r2, #1 ; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r3, #1
; CHECK-MVE-NEXT: mov.w r0, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s4 ; CHECK-MVE-NEXT: vcmp.f32 s2, s4
; CHECK-MVE-NEXT: cset r2, mi
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r3, #1 ; CHECK-MVE-NEXT: movgt r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r3, mi
; CHECK-MVE-NEXT: movmi r0, #1
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r0, #1 ; CHECK-MVE-NEXT: movgt r3, #1
; CHECK-MVE-NEXT: cmp r3, #0 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_one_v4f32: ; CHECK-MVEFP-LABEL: vcmp_one_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov r0, s4 ; CHECK-MVEFP-NEXT: vmov r0, s4
; CHECK-MVEFP-NEXT: vpt.f32 ge, q0, r0 ; CHECK-MVEFP-NEXT: vpt.f32 ge, q0, r0
; CHECK-MVEFP-NEXT: vcmpt.f32 le, q0, r0 ; CHECK-MVEFP-NEXT: vcmpt.f32 le, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q3, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q3, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <4 x float> undef, float %src2, i32 0 %i = insertelement <4 x float> undef, float %src2, i32 0
%sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer %sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer
%c = fcmp one <4 x float> %src, %sp %c = fcmp one <4 x float> %src, %sp
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_ogt_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_ogt_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_ogt_v4f32: ; CHECK-MVE-LABEL: vcmp_ogt_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s4 ; CHECK-MVE-NEXT: vcmp.f32 s1, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s4 ; CHECK-MVE-NEXT: vcmp.f32 s0, s4
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, s4 ; CHECK-MVE-NEXT: vcmp.f32 s3, s4
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r1, gt
; CHECK-MVE-NEXT: movgt r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s4 ; CHECK-MVE-NEXT: vcmp.f32 s2, s4
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r2, gt
; CHECK-MVE-NEXT: movgt r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r3, gt
; CHECK-MVE-NEXT: movgt r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ogt_v4f32: ; CHECK-MVEFP-LABEL: vcmp_ogt_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov r0, s4 ; CHECK-MVEFP-NEXT: vmov r0, s4
; CHECK-MVEFP-NEXT: vcmp.f32 gt, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f32 gt, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <4 x float> undef, float %src2, i32 0 %i = insertelement <4 x float> undef, float %src2, i32 0
%sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer %sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer
%c = fcmp ogt <4 x float> %src, %sp %c = fcmp ogt <4 x float> %src, %sp
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_oge_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_oge_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_oge_v4f32: ; CHECK-MVE-LABEL: vcmp_oge_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s4 ; CHECK-MVE-NEXT: vcmp.f32 s1, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s4 ; CHECK-MVE-NEXT: vcmp.f32 s0, s4
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, s4 ; CHECK-MVE-NEXT: vcmp.f32 s3, s4
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r1, ge
; CHECK-MVE-NEXT: movge r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s4 ; CHECK-MVE-NEXT: vcmp.f32 s2, s4
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r2, ge
; CHECK-MVE-NEXT: movge r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r3, ge
; CHECK-MVE-NEXT: movge r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_oge_v4f32: ; CHECK-MVEFP-LABEL: vcmp_oge_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov r0, s4 ; CHECK-MVEFP-NEXT: vmov r0, s4
; CHECK-MVEFP-NEXT: vcmp.f32 ge, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f32 ge, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <4 x float> undef, float %src2, i32 0 %i = insertelement <4 x float> undef, float %src2, i32 0
%sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer %sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer
%c = fcmp oge <4 x float> %src, %sp %c = fcmp oge <4 x float> %src, %sp
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_olt_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_olt_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_olt_v4f32: ; CHECK-MVE-LABEL: vcmp_olt_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s4 ; CHECK-MVE-NEXT: vcmp.f32 s1, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s4 ; CHECK-MVE-NEXT: vcmp.f32 s0, s4
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, s4 ; CHECK-MVE-NEXT: vcmp.f32 s3, s4
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r1, mi
; CHECK-MVE-NEXT: movmi r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s4 ; CHECK-MVE-NEXT: vcmp.f32 s2, s4
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r2, mi
; CHECK-MVE-NEXT: movmi r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r3, mi
; CHECK-MVE-NEXT: movmi r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_olt_v4f32: ; CHECK-MVEFP-LABEL: vcmp_olt_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov r0, s4 ; CHECK-MVEFP-NEXT: vmov r0, s4
; CHECK-MVEFP-NEXT: vcmp.f32 lt, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f32 lt, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <4 x float> undef, float %src2, i32 0 %i = insertelement <4 x float> undef, float %src2, i32 0
%sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer %sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer
%c = fcmp olt <4 x float> %src, %sp %c = fcmp olt <4 x float> %src, %sp
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_ole_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_ole_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_ole_v4f32: ; CHECK-MVE-LABEL: vcmp_ole_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s4 ; CHECK-MVE-NEXT: vcmp.f32 s1, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s4 ; CHECK-MVE-NEXT: vcmp.f32 s0, s4
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, s4 ; CHECK-MVE-NEXT: vcmp.f32 s3, s4
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r1, ls
; CHECK-MVE-NEXT: movls r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s4 ; CHECK-MVE-NEXT: vcmp.f32 s2, s4
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r2, ls
; CHECK-MVE-NEXT: movls r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r3, ls
; CHECK-MVE-NEXT: movls r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ole_v4f32: ; CHECK-MVEFP-LABEL: vcmp_ole_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov r0, s4 ; CHECK-MVEFP-NEXT: vmov r0, s4
; CHECK-MVEFP-NEXT: vcmp.f32 le, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f32 le, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <4 x float> undef, float %src2, i32 0 %i = insertelement <4 x float> undef, float %src2, i32 0
%sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer %sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer
%c = fcmp ole <4 x float> %src, %sp %c = fcmp ole <4 x float> %src, %sp
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_ueq_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_ueq_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_ueq_v4f32: ; CHECK-MVE-LABEL: vcmp_ueq_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s4 ; CHECK-MVE-NEXT: vcmp.f32 s1, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s1, s4 ; CHECK-MVE-NEXT: vcmp.f32 s1, s4
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s4 ; CHECK-MVE-NEXT: vcmp.f32 s0, s4
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1 ; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r2, #1
; CHECK-MVE-NEXT: vcmp.f32 s3, s4 ; CHECK-MVE-NEXT: vcmp.f32 s3, s4
; CHECK-MVE-NEXT: mov.w r3, #0 ; CHECK-MVE-NEXT: cset r1, eq
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r2, #1 ; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r3, #1
; CHECK-MVE-NEXT: mov.w r0, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s4 ; CHECK-MVE-NEXT: vcmp.f32 s2, s4
; CHECK-MVE-NEXT: cset r2, eq
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r3, #1 ; CHECK-MVE-NEXT: movvs r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r3, eq
; CHECK-MVE-NEXT: moveq r0, #1
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r0, #1 ; CHECK-MVE-NEXT: movvs r3, #1
; CHECK-MVE-NEXT: cmp r3, #0 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ueq_v4f32: ; CHECK-MVEFP-LABEL: vcmp_ueq_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov r0, s4 ; CHECK-MVEFP-NEXT: vmov r0, s4
; CHECK-MVEFP-NEXT: vpt.f32 ge, q0, r0 ; CHECK-MVEFP-NEXT: vpt.f32 ge, q0, r0
; CHECK-MVEFP-NEXT: vcmpt.f32 le, q0, r0 ; CHECK-MVEFP-NEXT: vcmpt.f32 le, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <4 x float> undef, float %src2, i32 0 %i = insertelement <4 x float> undef, float %src2, i32 0
%sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer %sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer
%c = fcmp ueq <4 x float> %src, %sp %c = fcmp ueq <4 x float> %src, %sp
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_une_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_une_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_une_v4f32: ; CHECK-MVE-LABEL: vcmp_une_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s4 ; CHECK-MVE-NEXT: vcmp.f32 s1, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s4 ; CHECK-MVE-NEXT: vcmp.f32 s0, s4
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, s4 ; CHECK-MVE-NEXT: vcmp.f32 s3, s4
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r1, ne
; CHECK-MVE-NEXT: movne r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s4 ; CHECK-MVE-NEXT: vcmp.f32 s2, s4
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r2, ne
; CHECK-MVE-NEXT: movne r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r3, ne
; CHECK-MVE-NEXT: movne r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_une_v4f32: ; CHECK-MVEFP-LABEL: vcmp_une_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov r0, s4 ; CHECK-MVEFP-NEXT: vmov r0, s4
; CHECK-MVEFP-NEXT: vcmp.f32 ne, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f32 ne, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <4 x float> undef, float %src2, i32 0 %i = insertelement <4 x float> undef, float %src2, i32 0
%sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer %sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer
%c = fcmp une <4 x float> %src, %sp %c = fcmp une <4 x float> %src, %sp
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_ugt_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_ugt_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_ugt_v4f32: ; CHECK-MVE-LABEL: vcmp_ugt_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s4 ; CHECK-MVE-NEXT: vcmp.f32 s1, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s4 ; CHECK-MVE-NEXT: vcmp.f32 s0, s4
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, s4 ; CHECK-MVE-NEXT: vcmp.f32 s3, s4
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r1, hi
; CHECK-MVE-NEXT: movhi r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s4 ; CHECK-MVE-NEXT: vcmp.f32 s2, s4
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r2, hi
; CHECK-MVE-NEXT: movhi r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r3, hi
; CHECK-MVE-NEXT: movhi r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ugt_v4f32: ; CHECK-MVEFP-LABEL: vcmp_ugt_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov r0, s4 ; CHECK-MVEFP-NEXT: vmov r0, s4
; CHECK-MVEFP-NEXT: vcmp.f32 gt, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f32 gt, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <4 x float> undef, float %src2, i32 0 %i = insertelement <4 x float> undef, float %src2, i32 0
%sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer %sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer
%c = fcmp ugt <4 x float> %src, %sp %c = fcmp ugt <4 x float> %src, %sp
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_uge_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_uge_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_uge_v4f32: ; CHECK-MVE-LABEL: vcmp_uge_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s4 ; CHECK-MVE-NEXT: vcmp.f32 s1, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s4 ; CHECK-MVE-NEXT: vcmp.f32 s0, s4
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, s4 ; CHECK-MVE-NEXT: vcmp.f32 s3, s4
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r1, pl
; CHECK-MVE-NEXT: movpl r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s4 ; CHECK-MVE-NEXT: vcmp.f32 s2, s4
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r2, pl
; CHECK-MVE-NEXT: movpl r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r3, pl
; CHECK-MVE-NEXT: movpl r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_uge_v4f32: ; CHECK-MVEFP-LABEL: vcmp_uge_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov r0, s4 ; CHECK-MVEFP-NEXT: vmov r0, s4
; CHECK-MVEFP-NEXT: vcmp.f32 ge, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f32 ge, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <4 x float> undef, float %src2, i32 0 %i = insertelement <4 x float> undef, float %src2, i32 0
%sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer %sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer
%c = fcmp uge <4 x float> %src, %sp %c = fcmp uge <4 x float> %src, %sp
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_ult_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_ult_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_ult_v4f32: ; CHECK-MVE-LABEL: vcmp_ult_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s4 ; CHECK-MVE-NEXT: vcmp.f32 s1, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s4 ; CHECK-MVE-NEXT: vcmp.f32 s0, s4
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, s4 ; CHECK-MVE-NEXT: vcmp.f32 s3, s4
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r1, lt
; CHECK-MVE-NEXT: movlt r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s4 ; CHECK-MVE-NEXT: vcmp.f32 s2, s4
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r2, lt
; CHECK-MVE-NEXT: movlt r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r3, lt
; CHECK-MVE-NEXT: movlt r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ult_v4f32: ; CHECK-MVEFP-LABEL: vcmp_ult_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov r0, s4 ; CHECK-MVEFP-NEXT: vmov r0, s4
; CHECK-MVEFP-NEXT: vcmp.f32 lt, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f32 lt, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <4 x float> undef, float %src2, i32 0 %i = insertelement <4 x float> undef, float %src2, i32 0
%sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer %sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer
%c = fcmp ult <4 x float> %src, %sp %c = fcmp ult <4 x float> %src, %sp
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_ule_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_ule_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_ule_v4f32: ; CHECK-MVE-LABEL: vcmp_ule_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s4 ; CHECK-MVE-NEXT: vcmp.f32 s1, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s4 ; CHECK-MVE-NEXT: vcmp.f32 s0, s4
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, s4 ; CHECK-MVE-NEXT: vcmp.f32 s3, s4
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r1, le
; CHECK-MVE-NEXT: movle r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s4 ; CHECK-MVE-NEXT: vcmp.f32 s2, s4
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r2, le
; CHECK-MVE-NEXT: movle r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r3, le
; CHECK-MVE-NEXT: movle r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ule_v4f32: ; CHECK-MVEFP-LABEL: vcmp_ule_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov r0, s4 ; CHECK-MVEFP-NEXT: vmov r0, s4
; CHECK-MVEFP-NEXT: vcmp.f32 le, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f32 le, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <4 x float> undef, float %src2, i32 0 %i = insertelement <4 x float> undef, float %src2, i32 0
%sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer %sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer
%c = fcmp ule <4 x float> %src, %sp %c = fcmp ule <4 x float> %src, %sp
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_ord_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_ord_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_ord_v4f32: ; CHECK-MVE-LABEL: vcmp_ord_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s4 ; CHECK-MVE-NEXT: vcmp.f32 s1, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s4 ; CHECK-MVE-NEXT: vcmp.f32 s0, s4
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, s4 ; CHECK-MVE-NEXT: vcmp.f32 s3, s4
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r1, vc
; CHECK-MVE-NEXT: movvc r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s4 ; CHECK-MVE-NEXT: vcmp.f32 s2, s4
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r2, vc
; CHECK-MVE-NEXT: movvc r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r3, vc
; CHECK-MVE-NEXT: movvc r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ord_v4f32: ; CHECK-MVEFP-LABEL: vcmp_ord_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov r0, s4 ; CHECK-MVEFP-NEXT: vmov r0, s4
; CHECK-MVEFP-NEXT: vpt.f32 ge, q0, r0 ; CHECK-MVEFP-NEXT: vpt.f32 ge, q0, r0
; CHECK-MVEFP-NEXT: vcmpt.f32 lt, q0, r0 ; CHECK-MVEFP-NEXT: vcmpt.f32 lt, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q3, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q3, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <4 x float> undef, float %src2, i32 0 %i = insertelement <4 x float> undef, float %src2, i32 0
%sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer %sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer
%c = fcmp ord <4 x float> %src, %sp %c = fcmp ord <4 x float> %src, %sp
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_uno_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_uno_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_uno_v4f32: ; CHECK-MVE-LABEL: vcmp_uno_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s4 ; CHECK-MVE-NEXT: vcmp.f32 s1, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s4 ; CHECK-MVE-NEXT: vcmp.f32 s0, s4
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, s4 ; CHECK-MVE-NEXT: vcmp.f32 s3, s4
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r1, vs
; CHECK-MVE-NEXT: movvs r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s4 ; CHECK-MVE-NEXT: vcmp.f32 s2, s4
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r2, vs
; CHECK-MVE-NEXT: movvs r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r3, vs
; CHECK-MVE-NEXT: movvs r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_uno_v4f32: ; CHECK-MVEFP-LABEL: vcmp_uno_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov r0, s4 ; CHECK-MVEFP-NEXT: vmov r0, s4
; CHECK-MVEFP-NEXT: vpt.f32 ge, q0, r0 ; CHECK-MVEFP-NEXT: vpt.f32 ge, q0, r0
; CHECK-MVEFP-NEXT: vcmpt.f32 lt, q0, r0 ; CHECK-MVEFP-NEXT: vcmpt.f32 lt, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <4 x float> undef, float %src2, i32 0 %i = insertelement <4 x float> undef, float %src2, i32 0
%sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer %sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer
%c = fcmp uno <4 x float> %src, %sp %c = fcmp uno <4 x float> %src, %sp
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_oeq_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_oeq_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_oeq_v8f16: ; CHECK-MVE-LABEL: vcmp_oeq_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s6, s0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s8 ; CHECK-MVE-NEXT: vmovx.f16 s6, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s0, s4 ; CHECK-MVE-NEXT: vcmp.f16 s0, s4
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vins.f16 s0, s6 ; CHECK-MVE-NEXT: vins.f16 s0, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s1 ; CHECK-MVE-NEXT: vmovx.f16 s6, s1
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s9 ; CHECK-MVE-NEXT: vmovx.f16 s6, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s1, s4 ; CHECK-MVE-NEXT: vcmp.f16 s1, s4
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s6 ; CHECK-MVE-NEXT: vins.f16 s1, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s2 ; CHECK-MVE-NEXT: vmovx.f16 s6, s2
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s10 ; CHECK-MVE-NEXT: vmovx.f16 s6, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s2, s4 ; CHECK-MVE-NEXT: vcmp.f16 s2, s4
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s2, s6 ; CHECK-MVE-NEXT: vins.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s3 ; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s11 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s4 ; CHECK-MVE-NEXT: vcmp.f16 s3, s4
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s6 ; CHECK-MVE-NEXT: vins.f16 s3, s6
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_oeq_v8f16: ; CHECK-MVEFP-LABEL: vcmp_oeq_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov.f16 r0, s4 ; CHECK-MVEFP-NEXT: vmov.f16 r0, s4
; CHECK-MVEFP-NEXT: vcmp.f16 eq, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f16 eq, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <8 x half> undef, half %src2, i32 0 %i = insertelement <8 x half> undef, half %src2, i32 0
%sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer %sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer
%c = fcmp oeq <8 x half> %src, %sp %c = fcmp oeq <8 x half> %src, %sp
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_one_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_one_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_one_v8f16: ; CHECK-MVE-LABEL: vcmp_one_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s6, s0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s8 ; CHECK-MVE-NEXT: vmovx.f16 s6, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s0, s4 ; CHECK-MVE-NEXT: vcmp.f16 s0, s4
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1 ; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vins.f16 s0, s6 ; CHECK-MVE-NEXT: vins.f16 s0, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s1 ; CHECK-MVE-NEXT: vmovx.f16 s6, s1
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s9 ; CHECK-MVE-NEXT: vmovx.f16 s6, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s1, s4 ; CHECK-MVE-NEXT: vcmp.f16 s1, s4
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: movmi r1, #1 ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1 ; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s6 ; CHECK-MVE-NEXT: vins.f16 s1, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s2 ; CHECK-MVE-NEXT: vmovx.f16 s6, s2
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s10 ; CHECK-MVE-NEXT: vmovx.f16 s6, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s2, s4 ; CHECK-MVE-NEXT: vcmp.f16 s2, s4
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: movmi r1, #1 ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1 ; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s2, s6 ; CHECK-MVE-NEXT: vins.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s3 ; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vmovx.f16 s8, s15 ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s4 ; CHECK-MVE-NEXT: vcmp.f16 s3, s4
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r0, #1
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r0, #1 ; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s6 ; CHECK-MVE-NEXT: vins.f16 s3, s6
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_one_v8f16: ; CHECK-MVEFP-LABEL: vcmp_one_v8f16:
Show All 10 Linesentry:
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_ogt_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_ogt_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_ogt_v8f16: ; CHECK-MVE-LABEL: vcmp_ogt_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s6, s0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s8 ; CHECK-MVE-NEXT: vmovx.f16 s6, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s0, s4 ; CHECK-MVE-NEXT: vcmp.f16 s0, s4
; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vins.f16 s0, s6 ; CHECK-MVE-NEXT: vins.f16 s0, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s1 ; CHECK-MVE-NEXT: vmovx.f16 s6, s1
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s9 ; CHECK-MVE-NEXT: vmovx.f16 s6, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s1, s4 ; CHECK-MVE-NEXT: vcmp.f16 s1, s4
; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: movgt r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s6 ; CHECK-MVE-NEXT: vins.f16 s1, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s2 ; CHECK-MVE-NEXT: vmovx.f16 s6, s2
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s10 ; CHECK-MVE-NEXT: vmovx.f16 s6, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s2, s4 ; CHECK-MVE-NEXT: vcmp.f16 s2, s4
; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: movgt r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s2, s6 ; CHECK-MVE-NEXT: vins.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s3 ; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s11 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s4 ; CHECK-MVE-NEXT: vcmp.f16 s3, s4
; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s6 ; CHECK-MVE-NEXT: vins.f16 s3, s6
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ogt_v8f16: ; CHECK-MVEFP-LABEL: vcmp_ogt_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov.f16 r0, s4 ; CHECK-MVEFP-NEXT: vmov.f16 r0, s4
; CHECK-MVEFP-NEXT: vcmp.f16 gt, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f16 gt, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <8 x half> undef, half %src2, i32 0 %i = insertelement <8 x half> undef, half %src2, i32 0
%sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer %sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer
%c = fcmp ogt <8 x half> %src, %sp %c = fcmp ogt <8 x half> %src, %sp
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_oge_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_oge_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_oge_v8f16: ; CHECK-MVE-LABEL: vcmp_oge_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s6, s0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s8 ; CHECK-MVE-NEXT: vmovx.f16 s6, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ge
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s0, s4 ; CHECK-MVE-NEXT: vcmp.f16 s0, s4
; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vins.f16 s0, s6 ; CHECK-MVE-NEXT: vins.f16 s0, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s1 ; CHECK-MVE-NEXT: vmovx.f16 s6, s1
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s9 ; CHECK-MVE-NEXT: vmovx.f16 s6, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ge
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s1, s4 ; CHECK-MVE-NEXT: vcmp.f16 s1, s4
; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: movge r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s6 ; CHECK-MVE-NEXT: vins.f16 s1, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s2 ; CHECK-MVE-NEXT: vmovx.f16 s6, s2
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s10 ; CHECK-MVE-NEXT: vmovx.f16 s6, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ge
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s2, s4 ; CHECK-MVE-NEXT: vcmp.f16 s2, s4
; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: movge r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s2, s6 ; CHECK-MVE-NEXT: vins.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s3 ; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s11 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ge
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s4 ; CHECK-MVE-NEXT: vcmp.f16 s3, s4
; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: movge r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s6 ; CHECK-MVE-NEXT: vins.f16 s3, s6
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_oge_v8f16: ; CHECK-MVEFP-LABEL: vcmp_oge_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov.f16 r0, s4 ; CHECK-MVEFP-NEXT: vmov.f16 r0, s4
; CHECK-MVEFP-NEXT: vcmp.f16 ge, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f16 ge, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <8 x half> undef, half %src2, i32 0 %i = insertelement <8 x half> undef, half %src2, i32 0
%sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer %sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer
%c = fcmp oge <8 x half> %src, %sp %c = fcmp oge <8 x half> %src, %sp
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_olt_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_olt_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_olt_v8f16: ; CHECK-MVE-LABEL: vcmp_olt_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s6, s0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s8 ; CHECK-MVE-NEXT: vmovx.f16 s6, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s0, s4 ; CHECK-MVE-NEXT: vcmp.f16 s0, s4
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vins.f16 s0, s6 ; CHECK-MVE-NEXT: vins.f16 s0, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s1 ; CHECK-MVE-NEXT: vmovx.f16 s6, s1
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s9 ; CHECK-MVE-NEXT: vmovx.f16 s6, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s1, s4 ; CHECK-MVE-NEXT: vcmp.f16 s1, s4
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s6 ; CHECK-MVE-NEXT: vins.f16 s1, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s2 ; CHECK-MVE-NEXT: vmovx.f16 s6, s2
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s10 ; CHECK-MVE-NEXT: vmovx.f16 s6, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s2, s4 ; CHECK-MVE-NEXT: vcmp.f16 s2, s4
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s2, s6 ; CHECK-MVE-NEXT: vins.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s3 ; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s11 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s4 ; CHECK-MVE-NEXT: vcmp.f16 s3, s4
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s6 ; CHECK-MVE-NEXT: vins.f16 s3, s6
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_olt_v8f16: ; CHECK-MVEFP-LABEL: vcmp_olt_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov.f16 r0, s4 ; CHECK-MVEFP-NEXT: vmov.f16 r0, s4
; CHECK-MVEFP-NEXT: vcmp.f16 lt, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f16 lt, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <8 x half> undef, half %src2, i32 0 %i = insertelement <8 x half> undef, half %src2, i32 0
%sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer %sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer
%c = fcmp olt <8 x half> %src, %sp %c = fcmp olt <8 x half> %src, %sp
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_ole_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_ole_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_ole_v8f16: ; CHECK-MVE-LABEL: vcmp_ole_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s6, s0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s8 ; CHECK-MVE-NEXT: vmovx.f16 s6, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ls
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s0, s4 ; CHECK-MVE-NEXT: vcmp.f16 s0, s4
; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vins.f16 s0, s6 ; CHECK-MVE-NEXT: vins.f16 s0, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s1 ; CHECK-MVE-NEXT: vmovx.f16 s6, s1
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s9 ; CHECK-MVE-NEXT: vmovx.f16 s6, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ls
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s1, s4 ; CHECK-MVE-NEXT: vcmp.f16 s1, s4
; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: movls r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s6 ; CHECK-MVE-NEXT: vins.f16 s1, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s2 ; CHECK-MVE-NEXT: vmovx.f16 s6, s2
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s10 ; CHECK-MVE-NEXT: vmovx.f16 s6, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ls
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s2, s4 ; CHECK-MVE-NEXT: vcmp.f16 s2, s4
; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: movls r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s2, s6 ; CHECK-MVE-NEXT: vins.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s3 ; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s11 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ls
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s4 ; CHECK-MVE-NEXT: vcmp.f16 s3, s4
; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: movls r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s6 ; CHECK-MVE-NEXT: vins.f16 s3, s6
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ole_v8f16: ; CHECK-MVEFP-LABEL: vcmp_ole_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov.f16 r0, s4 ; CHECK-MVEFP-NEXT: vmov.f16 r0, s4
; CHECK-MVEFP-NEXT: vcmp.f16 le, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f16 le, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <8 x half> undef, half %src2, i32 0 %i = insertelement <8 x half> undef, half %src2, i32 0
%sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer %sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer
%c = fcmp ole <8 x half> %src, %sp %c = fcmp ole <8 x half> %src, %sp
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_ueq_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_ueq_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_ueq_v8f16: ; CHECK-MVE-LABEL: vcmp_ueq_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s6, s0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s8 ; CHECK-MVE-NEXT: vmovx.f16 s6, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s0, s4 ; CHECK-MVE-NEXT: vcmp.f16 s0, s4
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1 ; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vins.f16 s0, s6 ; CHECK-MVE-NEXT: vins.f16 s0, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s1 ; CHECK-MVE-NEXT: vmovx.f16 s6, s1
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s9 ; CHECK-MVE-NEXT: vmovx.f16 s6, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s1, s4 ; CHECK-MVE-NEXT: vcmp.f16 s1, s4
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: moveq r1, #1 ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1 ; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s6 ; CHECK-MVE-NEXT: vins.f16 s1, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s2 ; CHECK-MVE-NEXT: vmovx.f16 s6, s2
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s10 ; CHECK-MVE-NEXT: vmovx.f16 s6, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s2, s4 ; CHECK-MVE-NEXT: vcmp.f16 s2, s4
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: moveq r1, #1 ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1 ; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s2, s6 ; CHECK-MVE-NEXT: vins.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s3 ; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vmovx.f16 s8, s15 ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s4 ; CHECK-MVE-NEXT: vcmp.f16 s3, s4
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r0, #1
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r0, #1 ; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s6 ; CHECK-MVE-NEXT: vins.f16 s3, s6
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ueq_v8f16: ; CHECK-MVEFP-LABEL: vcmp_ueq_v8f16:
Show All 10 Linesentry:
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_une_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_une_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_une_v8f16: ; CHECK-MVE-LABEL: vcmp_une_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s6, s0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s8 ; CHECK-MVE-NEXT: vmovx.f16 s6, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ne
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s0, s4 ; CHECK-MVE-NEXT: vcmp.f16 s0, s4
; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vins.f16 s0, s6 ; CHECK-MVE-NEXT: vins.f16 s0, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s1 ; CHECK-MVE-NEXT: vmovx.f16 s6, s1
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s9 ; CHECK-MVE-NEXT: vmovx.f16 s6, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ne
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s1, s4 ; CHECK-MVE-NEXT: vcmp.f16 s1, s4
; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: movne r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s6 ; CHECK-MVE-NEXT: vins.f16 s1, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s2 ; CHECK-MVE-NEXT: vmovx.f16 s6, s2
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s10 ; CHECK-MVE-NEXT: vmovx.f16 s6, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ne
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s2, s4 ; CHECK-MVE-NEXT: vcmp.f16 s2, s4
; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: movne r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s2, s6 ; CHECK-MVE-NEXT: vins.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s3 ; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s11 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ne
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s4 ; CHECK-MVE-NEXT: vcmp.f16 s3, s4
; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: movne r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s6 ; CHECK-MVE-NEXT: vins.f16 s3, s6
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_une_v8f16: ; CHECK-MVEFP-LABEL: vcmp_une_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov.f16 r0, s4 ; CHECK-MVEFP-NEXT: vmov.f16 r0, s4
; CHECK-MVEFP-NEXT: vcmp.f16 ne, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f16 ne, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <8 x half> undef, half %src2, i32 0 %i = insertelement <8 x half> undef, half %src2, i32 0
%sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer %sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer
%c = fcmp une <8 x half> %src, %sp %c = fcmp une <8 x half> %src, %sp
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_ugt_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_ugt_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_ugt_v8f16: ; CHECK-MVE-LABEL: vcmp_ugt_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s6, s0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s8 ; CHECK-MVE-NEXT: vmovx.f16 s6, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it hi
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s0, s4 ; CHECK-MVE-NEXT: vcmp.f16 s0, s4
; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vins.f16 s0, s6 ; CHECK-MVE-NEXT: vins.f16 s0, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s1 ; CHECK-MVE-NEXT: vmovx.f16 s6, s1
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s9 ; CHECK-MVE-NEXT: vmovx.f16 s6, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it hi
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s1, s4 ; CHECK-MVE-NEXT: vcmp.f16 s1, s4
; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: movhi r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s6 ; CHECK-MVE-NEXT: vins.f16 s1, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s2 ; CHECK-MVE-NEXT: vmovx.f16 s6, s2
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s10 ; CHECK-MVE-NEXT: vmovx.f16 s6, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it hi
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s2, s4 ; CHECK-MVE-NEXT: vcmp.f16 s2, s4
; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: movhi r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s2, s6 ; CHECK-MVE-NEXT: vins.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s3 ; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s11 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it hi
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s4 ; CHECK-MVE-NEXT: vcmp.f16 s3, s4
; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: movhi r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s6 ; CHECK-MVE-NEXT: vins.f16 s3, s6
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ugt_v8f16: ; CHECK-MVEFP-LABEL: vcmp_ugt_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov.f16 r0, s4 ; CHECK-MVEFP-NEXT: vmov.f16 r0, s4
; CHECK-MVEFP-NEXT: vcmp.f16 gt, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f16 gt, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <8 x half> undef, half %src2, i32 0 %i = insertelement <8 x half> undef, half %src2, i32 0
%sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer %sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer
%c = fcmp ugt <8 x half> %src, %sp %c = fcmp ugt <8 x half> %src, %sp
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_uge_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_uge_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_uge_v8f16: ; CHECK-MVE-LABEL: vcmp_uge_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s6, s0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s8 ; CHECK-MVE-NEXT: vmovx.f16 s6, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it pl
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s0, s4 ; CHECK-MVE-NEXT: vcmp.f16 s0, s4
; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vins.f16 s0, s6 ; CHECK-MVE-NEXT: vins.f16 s0, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s1 ; CHECK-MVE-NEXT: vmovx.f16 s6, s1
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s9 ; CHECK-MVE-NEXT: vmovx.f16 s6, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it pl
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s1, s4 ; CHECK-MVE-NEXT: vcmp.f16 s1, s4
; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: movpl r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s6 ; CHECK-MVE-NEXT: vins.f16 s1, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s2 ; CHECK-MVE-NEXT: vmovx.f16 s6, s2
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s10 ; CHECK-MVE-NEXT: vmovx.f16 s6, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it pl
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s2, s4 ; CHECK-MVE-NEXT: vcmp.f16 s2, s4
; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: movpl r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s2, s6 ; CHECK-MVE-NEXT: vins.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s3 ; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s11 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it pl
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s4 ; CHECK-MVE-NEXT: vcmp.f16 s3, s4
; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: movpl r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s6 ; CHECK-MVE-NEXT: vins.f16 s3, s6
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_uge_v8f16: ; CHECK-MVEFP-LABEL: vcmp_uge_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov.f16 r0, s4 ; CHECK-MVEFP-NEXT: vmov.f16 r0, s4
; CHECK-MVEFP-NEXT: vcmp.f16 ge, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f16 ge, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <8 x half> undef, half %src2, i32 0 %i = insertelement <8 x half> undef, half %src2, i32 0
%sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer %sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer
%c = fcmp uge <8 x half> %src, %sp %c = fcmp uge <8 x half> %src, %sp
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_ult_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_ult_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_ult_v8f16: ; CHECK-MVE-LABEL: vcmp_ult_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s6, s0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s8 ; CHECK-MVE-NEXT: vmovx.f16 s6, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it lt
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s0, s4 ; CHECK-MVE-NEXT: vcmp.f16 s0, s4
; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vins.f16 s0, s6 ; CHECK-MVE-NEXT: vins.f16 s0, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s1 ; CHECK-MVE-NEXT: vmovx.f16 s6, s1
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s9 ; CHECK-MVE-NEXT: vmovx.f16 s6, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it lt
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s1, s4 ; CHECK-MVE-NEXT: vcmp.f16 s1, s4
; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: movlt r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s6 ; CHECK-MVE-NEXT: vins.f16 s1, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s2 ; CHECK-MVE-NEXT: vmovx.f16 s6, s2
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s10 ; CHECK-MVE-NEXT: vmovx.f16 s6, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it lt
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s2, s4 ; CHECK-MVE-NEXT: vcmp.f16 s2, s4
; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: movlt r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s2, s6 ; CHECK-MVE-NEXT: vins.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s3 ; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s11 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it lt
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s4 ; CHECK-MVE-NEXT: vcmp.f16 s3, s4
; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: movlt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s6 ; CHECK-MVE-NEXT: vins.f16 s3, s6
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ult_v8f16: ; CHECK-MVEFP-LABEL: vcmp_ult_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov.f16 r0, s4 ; CHECK-MVEFP-NEXT: vmov.f16 r0, s4
; CHECK-MVEFP-NEXT: vcmp.f16 lt, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f16 lt, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <8 x half> undef, half %src2, i32 0 %i = insertelement <8 x half> undef, half %src2, i32 0
%sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer %sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer
%c = fcmp ult <8 x half> %src, %sp %c = fcmp ult <8 x half> %src, %sp
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_ule_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_ule_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_ule_v8f16: ; CHECK-MVE-LABEL: vcmp_ule_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s6, s0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s8 ; CHECK-MVE-NEXT: vmovx.f16 s6, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it le
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s0, s4 ; CHECK-MVE-NEXT: vcmp.f16 s0, s4
; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vins.f16 s0, s6 ; CHECK-MVE-NEXT: vins.f16 s0, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s1 ; CHECK-MVE-NEXT: vmovx.f16 s6, s1
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s9 ; CHECK-MVE-NEXT: vmovx.f16 s6, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it le
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s1, s4 ; CHECK-MVE-NEXT: vcmp.f16 s1, s4
; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: movle r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s6 ; CHECK-MVE-NEXT: vins.f16 s1, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s2 ; CHECK-MVE-NEXT: vmovx.f16 s6, s2
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s10 ; CHECK-MVE-NEXT: vmovx.f16 s6, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it le
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s2, s4 ; CHECK-MVE-NEXT: vcmp.f16 s2, s4
; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: movle r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s2, s6 ; CHECK-MVE-NEXT: vins.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s3 ; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s11 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it le
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s4 ; CHECK-MVE-NEXT: vcmp.f16 s3, s4
; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: movle r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s6 ; CHECK-MVE-NEXT: vins.f16 s3, s6
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ule_v8f16: ; CHECK-MVEFP-LABEL: vcmp_ule_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov.f16 r0, s4 ; CHECK-MVEFP-NEXT: vmov.f16 r0, s4
; CHECK-MVEFP-NEXT: vcmp.f16 le, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f16 le, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <8 x half> undef, half %src2, i32 0 %i = insertelement <8 x half> undef, half %src2, i32 0
%sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer %sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer
%c = fcmp ule <8 x half> %src, %sp %c = fcmp ule <8 x half> %src, %sp
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_ord_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_ord_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_ord_v8f16: ; CHECK-MVE-LABEL: vcmp_ord_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s6, s0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s8 ; CHECK-MVE-NEXT: vmovx.f16 s6, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vc
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s0, s4 ; CHECK-MVE-NEXT: vcmp.f16 s0, s4
; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vins.f16 s0, s6 ; CHECK-MVE-NEXT: vins.f16 s0, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s1 ; CHECK-MVE-NEXT: vmovx.f16 s6, s1
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s9 ; CHECK-MVE-NEXT: vmovx.f16 s6, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vc
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s1, s4 ; CHECK-MVE-NEXT: vcmp.f16 s1, s4
; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: movvc r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s6 ; CHECK-MVE-NEXT: vins.f16 s1, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s2 ; CHECK-MVE-NEXT: vmovx.f16 s6, s2
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s10 ; CHECK-MVE-NEXT: vmovx.f16 s6, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vc
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s2, s4 ; CHECK-MVE-NEXT: vcmp.f16 s2, s4
; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: movvc r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s2, s6 ; CHECK-MVE-NEXT: vins.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s3 ; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s11 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vc
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s4 ; CHECK-MVE-NEXT: vcmp.f16 s3, s4
; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: movvc r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s6 ; CHECK-MVE-NEXT: vins.f16 s3, s6
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ord_v8f16: ; CHECK-MVEFP-LABEL: vcmp_ord_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov.f16 r0, s4 ; CHECK-MVEFP-NEXT: vmov.f16 r0, s4
; CHECK-MVEFP-NEXT: vpt.f16 ge, q0, r0 ; CHECK-MVEFP-NEXT: vpt.f16 ge, q0, r0
; CHECK-MVEFP-NEXT: vcmpt.f16 lt, q0, r0 ; CHECK-MVEFP-NEXT: vcmpt.f16 lt, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q3, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q3, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <8 x half> undef, half %src2, i32 0 %i = insertelement <8 x half> undef, half %src2, i32 0
%sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer %sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer
%c = fcmp ord <8 x half> %src, %sp %c = fcmp ord <8 x half> %src, %sp
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_uno_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_uno_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_uno_v8f16: ; CHECK-MVE-LABEL: vcmp_uno_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s6, s0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s8 ; CHECK-MVE-NEXT: vmovx.f16 s6, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s0, s4 ; CHECK-MVE-NEXT: vcmp.f16 s0, s4
; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vins.f16 s0, s6 ; CHECK-MVE-NEXT: vins.f16 s0, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s1 ; CHECK-MVE-NEXT: vmovx.f16 s6, s1
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s9 ; CHECK-MVE-NEXT: vmovx.f16 s6, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s1, s4 ; CHECK-MVE-NEXT: vcmp.f16 s1, s4
; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: movvs r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s6 ; CHECK-MVE-NEXT: vins.f16 s1, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s2 ; CHECK-MVE-NEXT: vmovx.f16 s6, s2
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s10 ; CHECK-MVE-NEXT: vmovx.f16 s6, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s2, s4 ; CHECK-MVE-NEXT: vcmp.f16 s2, s4
; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: movvs r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s2, s6 ; CHECK-MVE-NEXT: vins.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s3 ; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s11 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s4 ; CHECK-MVE-NEXT: vcmp.f16 s3, s4
; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s6 ; CHECK-MVE-NEXT: vins.f16 s3, s6
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_uno_v8f16: ; CHECK-MVEFP-LABEL: vcmp_uno_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov.f16 r0, s4 ; CHECK-MVEFP-NEXT: vmov.f16 r0, s4
Show All 11 Lines
; Reversed ; Reversed
define arm_aapcs_vfpcc <4 x float> @vcmp_r_oeq_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_r_oeq_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_r_oeq_v4f32: ; CHECK-MVE-LABEL: vcmp_r_oeq_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s4, s1 ; CHECK-MVE-NEXT: vcmp.f32 s4, s1
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s4, s0 ; CHECK-MVE-NEXT: vcmp.f32 s4, s0
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s4, s3 ; CHECK-MVE-NEXT: vcmp.f32 s4, s3
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r1, eq
; CHECK-MVE-NEXT: moveq r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s4, s2 ; CHECK-MVE-NEXT: vcmp.f32 s4, s2
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r2, eq
; CHECK-MVE-NEXT: moveq r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r3, eq
; CHECK-MVE-NEXT: moveq r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_oeq_v4f32: ; CHECK-MVEFP-LABEL: vcmp_r_oeq_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov r0, s4 ; CHECK-MVEFP-NEXT: vmov r0, s4
; CHECK-MVEFP-NEXT: vcmp.f32 eq, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f32 eq, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <4 x float> undef, float %src2, i32 0 %i = insertelement <4 x float> undef, float %src2, i32 0
%sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer %sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer
%c = fcmp oeq <4 x float> %sp, %src %c = fcmp oeq <4 x float> %sp, %src
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_r_one_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_r_one_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_r_one_v4f32: ; CHECK-MVE-LABEL: vcmp_r_one_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s4, s1 ; CHECK-MVE-NEXT: vcmp.f32 s4, s1
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s4, s1 ; CHECK-MVE-NEXT: vcmp.f32 s4, s1
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s4, s0 ; CHECK-MVE-NEXT: vcmp.f32 s4, s0
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1 ; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r2, #1
; CHECK-MVE-NEXT: vcmp.f32 s4, s3 ; CHECK-MVE-NEXT: vcmp.f32 s4, s3
; CHECK-MVE-NEXT: mov.w r3, #0 ; CHECK-MVE-NEXT: cset r1, mi
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r2, #1 ; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r3, #1
; CHECK-MVE-NEXT: mov.w r0, #0
; CHECK-MVE-NEXT: vcmp.f32 s4, s2 ; CHECK-MVE-NEXT: vcmp.f32 s4, s2
; CHECK-MVE-NEXT: cset r2, mi
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r3, #1 ; CHECK-MVE-NEXT: movgt r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r3, mi
; CHECK-MVE-NEXT: movmi r0, #1
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r0, #1 ; CHECK-MVE-NEXT: movgt r3, #1
; CHECK-MVE-NEXT: cmp r3, #0 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_one_v4f32: ; CHECK-MVEFP-LABEL: vcmp_r_one_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov r0, s4 ; CHECK-MVEFP-NEXT: vmov r0, s4
; CHECK-MVEFP-NEXT: vpt.f32 le, q0, r0 ; CHECK-MVEFP-NEXT: vpt.f32 le, q0, r0
; CHECK-MVEFP-NEXT: vcmpt.f32 ge, q0, r0 ; CHECK-MVEFP-NEXT: vcmpt.f32 ge, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q3, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q3, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <4 x float> undef, float %src2, i32 0 %i = insertelement <4 x float> undef, float %src2, i32 0
%sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer %sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer
%c = fcmp one <4 x float> %sp, %src %c = fcmp one <4 x float> %sp, %src
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_r_ogt_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_r_ogt_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_r_ogt_v4f32: ; CHECK-MVE-LABEL: vcmp_r_ogt_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s4, s1 ; CHECK-MVE-NEXT: vcmp.f32 s4, s1
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s4, s0 ; CHECK-MVE-NEXT: vcmp.f32 s4, s0
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s4, s3 ; CHECK-MVE-NEXT: vcmp.f32 s4, s3
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r1, gt
; CHECK-MVE-NEXT: movgt r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s4, s2 ; CHECK-MVE-NEXT: vcmp.f32 s4, s2
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r2, gt
; CHECK-MVE-NEXT: movgt r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r3, gt
; CHECK-MVE-NEXT: movgt r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_ogt_v4f32: ; CHECK-MVEFP-LABEL: vcmp_r_ogt_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov r0, s4 ; CHECK-MVEFP-NEXT: vmov r0, s4
; CHECK-MVEFP-NEXT: vcmp.f32 lt, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f32 lt, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <4 x float> undef, float %src2, i32 0 %i = insertelement <4 x float> undef, float %src2, i32 0
%sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer %sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer
%c = fcmp ogt <4 x float> %sp, %src %c = fcmp ogt <4 x float> %sp, %src
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_r_oge_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_r_oge_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_r_oge_v4f32: ; CHECK-MVE-LABEL: vcmp_r_oge_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s4, s1 ; CHECK-MVE-NEXT: vcmp.f32 s4, s1
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s4, s0 ; CHECK-MVE-NEXT: vcmp.f32 s4, s0
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s4, s3 ; CHECK-MVE-NEXT: vcmp.f32 s4, s3
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r1, ge
; CHECK-MVE-NEXT: movge r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s4, s2 ; CHECK-MVE-NEXT: vcmp.f32 s4, s2
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r2, ge
; CHECK-MVE-NEXT: movge r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r3, ge
; CHECK-MVE-NEXT: movge r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_oge_v4f32: ; CHECK-MVEFP-LABEL: vcmp_r_oge_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov r0, s4 ; CHECK-MVEFP-NEXT: vmov r0, s4
; CHECK-MVEFP-NEXT: vcmp.f32 le, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f32 le, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <4 x float> undef, float %src2, i32 0 %i = insertelement <4 x float> undef, float %src2, i32 0
%sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer %sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer
%c = fcmp oge <4 x float> %sp, %src %c = fcmp oge <4 x float> %sp, %src
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_r_olt_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_r_olt_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_r_olt_v4f32: ; CHECK-MVE-LABEL: vcmp_r_olt_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s4, s1 ; CHECK-MVE-NEXT: vcmp.f32 s4, s1
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s4, s0 ; CHECK-MVE-NEXT: vcmp.f32 s4, s0
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s4, s3 ; CHECK-MVE-NEXT: vcmp.f32 s4, s3
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r1, mi
; CHECK-MVE-NEXT: movmi r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s4, s2 ; CHECK-MVE-NEXT: vcmp.f32 s4, s2
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r2, mi
; CHECK-MVE-NEXT: movmi r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r3, mi
; CHECK-MVE-NEXT: movmi r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_olt_v4f32: ; CHECK-MVEFP-LABEL: vcmp_r_olt_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov r0, s4 ; CHECK-MVEFP-NEXT: vmov r0, s4
; CHECK-MVEFP-NEXT: vcmp.f32 gt, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f32 gt, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <4 x float> undef, float %src2, i32 0 %i = insertelement <4 x float> undef, float %src2, i32 0
%sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer %sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer
%c = fcmp olt <4 x float> %sp, %src %c = fcmp olt <4 x float> %sp, %src
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_r_ole_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_r_ole_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_r_ole_v4f32: ; CHECK-MVE-LABEL: vcmp_r_ole_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s4, s1 ; CHECK-MVE-NEXT: vcmp.f32 s4, s1
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s4, s0 ; CHECK-MVE-NEXT: vcmp.f32 s4, s0
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s4, s3 ; CHECK-MVE-NEXT: vcmp.f32 s4, s3
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r1, ls
; CHECK-MVE-NEXT: movls r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s4, s2 ; CHECK-MVE-NEXT: vcmp.f32 s4, s2
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r2, ls
; CHECK-MVE-NEXT: movls r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r3, ls
; CHECK-MVE-NEXT: movls r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_ole_v4f32: ; CHECK-MVEFP-LABEL: vcmp_r_ole_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov r0, s4 ; CHECK-MVEFP-NEXT: vmov r0, s4
; CHECK-MVEFP-NEXT: vcmp.f32 ge, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f32 ge, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <4 x float> undef, float %src2, i32 0 %i = insertelement <4 x float> undef, float %src2, i32 0
%sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer %sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer
%c = fcmp ole <4 x float> %sp, %src %c = fcmp ole <4 x float> %sp, %src
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_r_ueq_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_r_ueq_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_r_ueq_v4f32: ; CHECK-MVE-LABEL: vcmp_r_ueq_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s4, s1 ; CHECK-MVE-NEXT: vcmp.f32 s4, s1
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s4, s1 ; CHECK-MVE-NEXT: vcmp.f32 s4, s1
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s4, s0 ; CHECK-MVE-NEXT: vcmp.f32 s4, s0
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1 ; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r2, #1
; CHECK-MVE-NEXT: vcmp.f32 s4, s3 ; CHECK-MVE-NEXT: vcmp.f32 s4, s3
; CHECK-MVE-NEXT: mov.w r3, #0 ; CHECK-MVE-NEXT: cset r1, eq
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r2, #1 ; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r3, #1
; CHECK-MVE-NEXT: mov.w r0, #0
; CHECK-MVE-NEXT: vcmp.f32 s4, s2 ; CHECK-MVE-NEXT: vcmp.f32 s4, s2
; CHECK-MVE-NEXT: cset r2, eq
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r3, #1 ; CHECK-MVE-NEXT: movvs r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r3, eq
; CHECK-MVE-NEXT: moveq r0, #1
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r0, #1 ; CHECK-MVE-NEXT: movvs r3, #1
; CHECK-MVE-NEXT: cmp r3, #0 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_ueq_v4f32: ; CHECK-MVEFP-LABEL: vcmp_r_ueq_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov r0, s4 ; CHECK-MVEFP-NEXT: vmov r0, s4
; CHECK-MVEFP-NEXT: vpt.f32 le, q0, r0 ; CHECK-MVEFP-NEXT: vpt.f32 le, q0, r0
; CHECK-MVEFP-NEXT: vcmpt.f32 ge, q0, r0 ; CHECK-MVEFP-NEXT: vcmpt.f32 ge, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <4 x float> undef, float %src2, i32 0 %i = insertelement <4 x float> undef, float %src2, i32 0
%sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer %sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer
%c = fcmp ueq <4 x float> %sp, %src %c = fcmp ueq <4 x float> %sp, %src
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_r_une_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_r_une_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_r_une_v4f32: ; CHECK-MVE-LABEL: vcmp_r_une_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s4, s1 ; CHECK-MVE-NEXT: vcmp.f32 s4, s1
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s4, s0 ; CHECK-MVE-NEXT: vcmp.f32 s4, s0
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s4, s3 ; CHECK-MVE-NEXT: vcmp.f32 s4, s3
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r1, ne
; CHECK-MVE-NEXT: movne r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s4, s2 ; CHECK-MVE-NEXT: vcmp.f32 s4, s2
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r2, ne
; CHECK-MVE-NEXT: movne r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r3, ne
; CHECK-MVE-NEXT: movne r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_une_v4f32: ; CHECK-MVEFP-LABEL: vcmp_r_une_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov r0, s4 ; CHECK-MVEFP-NEXT: vmov r0, s4
; CHECK-MVEFP-NEXT: vcmp.f32 ne, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f32 ne, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <4 x float> undef, float %src2, i32 0 %i = insertelement <4 x float> undef, float %src2, i32 0
%sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer %sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer
%c = fcmp une <4 x float> %sp, %src %c = fcmp une <4 x float> %sp, %src
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_r_ugt_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_r_ugt_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_r_ugt_v4f32: ; CHECK-MVE-LABEL: vcmp_r_ugt_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s4, s1 ; CHECK-MVE-NEXT: vcmp.f32 s4, s1
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s4, s0 ; CHECK-MVE-NEXT: vcmp.f32 s4, s0
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s4, s3 ; CHECK-MVE-NEXT: vcmp.f32 s4, s3
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r1, hi
; CHECK-MVE-NEXT: movhi r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s4, s2 ; CHECK-MVE-NEXT: vcmp.f32 s4, s2
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r2, hi
; CHECK-MVE-NEXT: movhi r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r3, hi
; CHECK-MVE-NEXT: movhi r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_ugt_v4f32: ; CHECK-MVEFP-LABEL: vcmp_r_ugt_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov r0, s4 ; CHECK-MVEFP-NEXT: vmov r0, s4
; CHECK-MVEFP-NEXT: vcmp.f32 lt, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f32 lt, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <4 x float> undef, float %src2, i32 0 %i = insertelement <4 x float> undef, float %src2, i32 0
%sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer %sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer
%c = fcmp ugt <4 x float> %sp, %src %c = fcmp ugt <4 x float> %sp, %src
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_r_uge_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_r_uge_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_r_uge_v4f32: ; CHECK-MVE-LABEL: vcmp_r_uge_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s4, s1 ; CHECK-MVE-NEXT: vcmp.f32 s4, s1
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s4, s0 ; CHECK-MVE-NEXT: vcmp.f32 s4, s0
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s4, s3 ; CHECK-MVE-NEXT: vcmp.f32 s4, s3
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r1, pl
; CHECK-MVE-NEXT: movpl r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s4, s2 ; CHECK-MVE-NEXT: vcmp.f32 s4, s2
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r2, pl
; CHECK-MVE-NEXT: movpl r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r3, pl
; CHECK-MVE-NEXT: movpl r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_uge_v4f32: ; CHECK-MVEFP-LABEL: vcmp_r_uge_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov r0, s4 ; CHECK-MVEFP-NEXT: vmov r0, s4
; CHECK-MVEFP-NEXT: vcmp.f32 le, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f32 le, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <4 x float> undef, float %src2, i32 0 %i = insertelement <4 x float> undef, float %src2, i32 0
%sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer %sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer
%c = fcmp uge <4 x float> %sp, %src %c = fcmp uge <4 x float> %sp, %src
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_r_ult_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_r_ult_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_r_ult_v4f32: ; CHECK-MVE-LABEL: vcmp_r_ult_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s4, s1 ; CHECK-MVE-NEXT: vcmp.f32 s4, s1
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s4, s0 ; CHECK-MVE-NEXT: vcmp.f32 s4, s0
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s4, s3 ; CHECK-MVE-NEXT: vcmp.f32 s4, s3
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r1, lt
; CHECK-MVE-NEXT: movlt r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s4, s2 ; CHECK-MVE-NEXT: vcmp.f32 s4, s2
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r2, lt
; CHECK-MVE-NEXT: movlt r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r3, lt
; CHECK-MVE-NEXT: movlt r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_ult_v4f32: ; CHECK-MVEFP-LABEL: vcmp_r_ult_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov r0, s4 ; CHECK-MVEFP-NEXT: vmov r0, s4
; CHECK-MVEFP-NEXT: vcmp.f32 gt, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f32 gt, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <4 x float> undef, float %src2, i32 0 %i = insertelement <4 x float> undef, float %src2, i32 0
%sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer %sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer
%c = fcmp ult <4 x float> %sp, %src %c = fcmp ult <4 x float> %sp, %src
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_r_ule_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_r_ule_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_r_ule_v4f32: ; CHECK-MVE-LABEL: vcmp_r_ule_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s4, s1 ; CHECK-MVE-NEXT: vcmp.f32 s4, s1
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s4, s0 ; CHECK-MVE-NEXT: vcmp.f32 s4, s0
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s4, s3 ; CHECK-MVE-NEXT: vcmp.f32 s4, s3
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r1, le
; CHECK-MVE-NEXT: movle r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s4, s2 ; CHECK-MVE-NEXT: vcmp.f32 s4, s2
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r2, le
; CHECK-MVE-NEXT: movle r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r3, le
; CHECK-MVE-NEXT: movle r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_ule_v4f32: ; CHECK-MVEFP-LABEL: vcmp_r_ule_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov r0, s4 ; CHECK-MVEFP-NEXT: vmov r0, s4
; CHECK-MVEFP-NEXT: vcmp.f32 ge, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f32 ge, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <4 x float> undef, float %src2, i32 0 %i = insertelement <4 x float> undef, float %src2, i32 0
%sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer %sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer
%c = fcmp ule <4 x float> %sp, %src %c = fcmp ule <4 x float> %sp, %src
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_r_ord_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_r_ord_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_r_ord_v4f32: ; CHECK-MVE-LABEL: vcmp_r_ord_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s4, s1 ; CHECK-MVE-NEXT: vcmp.f32 s4, s1
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s4, s0 ; CHECK-MVE-NEXT: vcmp.f32 s4, s0
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s4, s3 ; CHECK-MVE-NEXT: vcmp.f32 s4, s3
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r1, vc
; CHECK-MVE-NEXT: movvc r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s4, s2 ; CHECK-MVE-NEXT: vcmp.f32 s4, s2
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r2, vc
; CHECK-MVE-NEXT: movvc r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r3, vc
; CHECK-MVE-NEXT: movvc r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_ord_v4f32: ; CHECK-MVEFP-LABEL: vcmp_r_ord_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov r0, s4 ; CHECK-MVEFP-NEXT: vmov r0, s4
; CHECK-MVEFP-NEXT: vpt.f32 le, q0, r0 ; CHECK-MVEFP-NEXT: vpt.f32 le, q0, r0
; CHECK-MVEFP-NEXT: vcmpt.f32 gt, q0, r0 ; CHECK-MVEFP-NEXT: vcmpt.f32 gt, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q3, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q3, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <4 x float> undef, float %src2, i32 0 %i = insertelement <4 x float> undef, float %src2, i32 0
%sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer %sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer
%c = fcmp ord <4 x float> %sp, %src %c = fcmp ord <4 x float> %sp, %src
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_r_uno_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_r_uno_v4f32(<4 x float> %src, float %src2, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_r_uno_v4f32: ; CHECK-MVE-LABEL: vcmp_r_uno_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s4, s1 ; CHECK-MVE-NEXT: vcmp.f32 s4, s1
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s4, s0 ; CHECK-MVE-NEXT: vcmp.f32 s4, s0
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s4, s3 ; CHECK-MVE-NEXT: vcmp.f32 s4, s3
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r1, vs
; CHECK-MVE-NEXT: movvs r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s4, s2 ; CHECK-MVE-NEXT: vcmp.f32 s4, s2
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r2, vs
; CHECK-MVE-NEXT: movvs r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r3, vs
; CHECK-MVE-NEXT: movvs r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f32 s3, s15, s11
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f32 s2, s14, s10
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f32 s1, s13, s9
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f32 s0, s12, s8
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_uno_v4f32: ; CHECK-MVEFP-LABEL: vcmp_r_uno_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov r0, s4 ; CHECK-MVEFP-NEXT: vmov r0, s4
; CHECK-MVEFP-NEXT: vpt.f32 le, q0, r0 ; CHECK-MVEFP-NEXT: vpt.f32 le, q0, r0
; CHECK-MVEFP-NEXT: vcmpt.f32 gt, q0, r0 ; CHECK-MVEFP-NEXT: vcmpt.f32 gt, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <4 x float> undef, float %src2, i32 0 %i = insertelement <4 x float> undef, float %src2, i32 0
%sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer %sp = shufflevector <4 x float> %i, <4 x float> undef, <4 x i32> zeroinitializer
%c = fcmp uno <4 x float> %sp, %src %c = fcmp uno <4 x float> %sp, %src
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_r_oeq_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_r_oeq_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_r_oeq_v8f16: ; CHECK-MVE-LABEL: vcmp_r_oeq_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s6, s0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s8 ; CHECK-MVE-NEXT: vmovx.f16 s6, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s4, s0 ; CHECK-MVE-NEXT: vcmp.f16 s4, s0
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vins.f16 s0, s6 ; CHECK-MVE-NEXT: vins.f16 s0, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s1 ; CHECK-MVE-NEXT: vmovx.f16 s6, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s9 ; CHECK-MVE-NEXT: vmovx.f16 s6, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s4, s1 ; CHECK-MVE-NEXT: vcmp.f16 s4, s1
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s6 ; CHECK-MVE-NEXT: vins.f16 s1, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s2 ; CHECK-MVE-NEXT: vmovx.f16 s6, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s10 ; CHECK-MVE-NEXT: vmovx.f16 s6, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s4, s2 ; CHECK-MVE-NEXT: vcmp.f16 s4, s2
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s2, s6 ; CHECK-MVE-NEXT: vins.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s3 ; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s11 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s4, s3 ; CHECK-MVE-NEXT: vcmp.f16 s4, s3
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s6 ; CHECK-MVE-NEXT: vins.f16 s3, s6
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_oeq_v8f16: ; CHECK-MVEFP-LABEL: vcmp_r_oeq_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov.f16 r0, s4 ; CHECK-MVEFP-NEXT: vmov.f16 r0, s4
; CHECK-MVEFP-NEXT: vcmp.f16 eq, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f16 eq, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <8 x half> undef, half %src2, i32 0 %i = insertelement <8 x half> undef, half %src2, i32 0
%sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer %sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer
%c = fcmp oeq <8 x half> %sp, %src %c = fcmp oeq <8 x half> %sp, %src
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_r_one_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_r_one_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_r_one_v8f16: ; CHECK-MVE-LABEL: vcmp_r_one_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s6, s0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s8 ; CHECK-MVE-NEXT: vmovx.f16 s6, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s4, s0 ; CHECK-MVE-NEXT: vcmp.f16 s4, s0
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1 ; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vins.f16 s0, s6 ; CHECK-MVE-NEXT: vins.f16 s0, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s1 ; CHECK-MVE-NEXT: vmovx.f16 s6, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s9 ; CHECK-MVE-NEXT: vmovx.f16 s6, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s4, s1 ; CHECK-MVE-NEXT: vcmp.f16 s4, s1
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: movmi r1, #1 ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1 ; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s6 ; CHECK-MVE-NEXT: vins.f16 s1, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s2 ; CHECK-MVE-NEXT: vmovx.f16 s6, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s10 ; CHECK-MVE-NEXT: vmovx.f16 s6, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s4, s2 ; CHECK-MVE-NEXT: vcmp.f16 s4, s2
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: movmi r1, #1 ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1 ; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s2, s6 ; CHECK-MVE-NEXT: vins.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s3 ; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vmovx.f16 s8, s15 ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s4, s3 ; CHECK-MVE-NEXT: vcmp.f16 s4, s3
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r0, #1
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r0, #1 ; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s6 ; CHECK-MVE-NEXT: vins.f16 s3, s6
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_one_v8f16: ; CHECK-MVEFP-LABEL: vcmp_r_one_v8f16:
Show All 10 Linesentry:
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_r_ogt_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_r_ogt_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_r_ogt_v8f16: ; CHECK-MVE-LABEL: vcmp_r_ogt_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s6, s0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s8 ; CHECK-MVE-NEXT: vmovx.f16 s6, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s4, s0 ; CHECK-MVE-NEXT: vcmp.f16 s4, s0
; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vins.f16 s0, s6 ; CHECK-MVE-NEXT: vins.f16 s0, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s1 ; CHECK-MVE-NEXT: vmovx.f16 s6, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s9 ; CHECK-MVE-NEXT: vmovx.f16 s6, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s4, s1 ; CHECK-MVE-NEXT: vcmp.f16 s4, s1
; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: movgt r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s6 ; CHECK-MVE-NEXT: vins.f16 s1, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s2 ; CHECK-MVE-NEXT: vmovx.f16 s6, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s10 ; CHECK-MVE-NEXT: vmovx.f16 s6, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s4, s2 ; CHECK-MVE-NEXT: vcmp.f16 s4, s2
; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: movgt r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s2, s6 ; CHECK-MVE-NEXT: vins.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s3 ; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s11 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s4, s3 ; CHECK-MVE-NEXT: vcmp.f16 s4, s3
; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s6 ; CHECK-MVE-NEXT: vins.f16 s3, s6
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_ogt_v8f16: ; CHECK-MVEFP-LABEL: vcmp_r_ogt_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov.f16 r0, s4 ; CHECK-MVEFP-NEXT: vmov.f16 r0, s4
; CHECK-MVEFP-NEXT: vcmp.f16 lt, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f16 lt, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <8 x half> undef, half %src2, i32 0 %i = insertelement <8 x half> undef, half %src2, i32 0
%sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer %sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer
%c = fcmp ogt <8 x half> %sp, %src %c = fcmp ogt <8 x half> %sp, %src
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_r_oge_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_r_oge_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_r_oge_v8f16: ; CHECK-MVE-LABEL: vcmp_r_oge_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s6, s0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s8 ; CHECK-MVE-NEXT: vmovx.f16 s6, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ge
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s4, s0 ; CHECK-MVE-NEXT: vcmp.f16 s4, s0
; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vins.f16 s0, s6 ; CHECK-MVE-NEXT: vins.f16 s0, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s1 ; CHECK-MVE-NEXT: vmovx.f16 s6, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s9 ; CHECK-MVE-NEXT: vmovx.f16 s6, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ge
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s4, s1 ; CHECK-MVE-NEXT: vcmp.f16 s4, s1
; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: movge r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s6 ; CHECK-MVE-NEXT: vins.f16 s1, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s2 ; CHECK-MVE-NEXT: vmovx.f16 s6, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s10 ; CHECK-MVE-NEXT: vmovx.f16 s6, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ge
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s4, s2 ; CHECK-MVE-NEXT: vcmp.f16 s4, s2
; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: movge r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s2, s6 ; CHECK-MVE-NEXT: vins.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s3 ; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s11 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ge
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s4, s3 ; CHECK-MVE-NEXT: vcmp.f16 s4, s3
; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: movge r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s6 ; CHECK-MVE-NEXT: vins.f16 s3, s6
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_oge_v8f16: ; CHECK-MVEFP-LABEL: vcmp_r_oge_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov.f16 r0, s4 ; CHECK-MVEFP-NEXT: vmov.f16 r0, s4
; CHECK-MVEFP-NEXT: vcmp.f16 le, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f16 le, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <8 x half> undef, half %src2, i32 0 %i = insertelement <8 x half> undef, half %src2, i32 0
%sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer %sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer
%c = fcmp oge <8 x half> %sp, %src %c = fcmp oge <8 x half> %sp, %src
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_r_olt_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_r_olt_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_r_olt_v8f16: ; CHECK-MVE-LABEL: vcmp_r_olt_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s6, s0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s8 ; CHECK-MVE-NEXT: vmovx.f16 s6, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s4, s0 ; CHECK-MVE-NEXT: vcmp.f16 s4, s0
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vins.f16 s0, s6 ; CHECK-MVE-NEXT: vins.f16 s0, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s1 ; CHECK-MVE-NEXT: vmovx.f16 s6, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s9 ; CHECK-MVE-NEXT: vmovx.f16 s6, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s4, s1 ; CHECK-MVE-NEXT: vcmp.f16 s4, s1
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s6 ; CHECK-MVE-NEXT: vins.f16 s1, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s2 ; CHECK-MVE-NEXT: vmovx.f16 s6, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s10 ; CHECK-MVE-NEXT: vmovx.f16 s6, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s4, s2 ; CHECK-MVE-NEXT: vcmp.f16 s4, s2
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s2, s6 ; CHECK-MVE-NEXT: vins.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s3 ; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s11 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s4, s3 ; CHECK-MVE-NEXT: vcmp.f16 s4, s3
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s6 ; CHECK-MVE-NEXT: vins.f16 s3, s6
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_olt_v8f16: ; CHECK-MVEFP-LABEL: vcmp_r_olt_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov.f16 r0, s4 ; CHECK-MVEFP-NEXT: vmov.f16 r0, s4
; CHECK-MVEFP-NEXT: vcmp.f16 gt, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f16 gt, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <8 x half> undef, half %src2, i32 0 %i = insertelement <8 x half> undef, half %src2, i32 0
%sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer %sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer
%c = fcmp olt <8 x half> %sp, %src %c = fcmp olt <8 x half> %sp, %src
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_r_ole_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_r_ole_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_r_ole_v8f16: ; CHECK-MVE-LABEL: vcmp_r_ole_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s6, s0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s8 ; CHECK-MVE-NEXT: vmovx.f16 s6, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ls
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s4, s0 ; CHECK-MVE-NEXT: vcmp.f16 s4, s0
; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vins.f16 s0, s6 ; CHECK-MVE-NEXT: vins.f16 s0, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s1 ; CHECK-MVE-NEXT: vmovx.f16 s6, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s9 ; CHECK-MVE-NEXT: vmovx.f16 s6, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ls
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s4, s1 ; CHECK-MVE-NEXT: vcmp.f16 s4, s1
; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: movls r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s6 ; CHECK-MVE-NEXT: vins.f16 s1, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s2 ; CHECK-MVE-NEXT: vmovx.f16 s6, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s10 ; CHECK-MVE-NEXT: vmovx.f16 s6, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ls
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s4, s2 ; CHECK-MVE-NEXT: vcmp.f16 s4, s2
; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: movls r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s2, s6 ; CHECK-MVE-NEXT: vins.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s3 ; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s11 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ls
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s4, s3 ; CHECK-MVE-NEXT: vcmp.f16 s4, s3
; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: movls r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s6 ; CHECK-MVE-NEXT: vins.f16 s3, s6
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_ole_v8f16: ; CHECK-MVEFP-LABEL: vcmp_r_ole_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov.f16 r0, s4 ; CHECK-MVEFP-NEXT: vmov.f16 r0, s4
; CHECK-MVEFP-NEXT: vcmp.f16 ge, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f16 ge, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <8 x half> undef, half %src2, i32 0 %i = insertelement <8 x half> undef, half %src2, i32 0
%sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer %sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer
%c = fcmp ole <8 x half> %sp, %src %c = fcmp ole <8 x half> %sp, %src
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_r_ueq_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_r_ueq_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_r_ueq_v8f16: ; CHECK-MVE-LABEL: vcmp_r_ueq_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s6, s0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s8 ; CHECK-MVE-NEXT: vmovx.f16 s6, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s4, s0 ; CHECK-MVE-NEXT: vcmp.f16 s4, s0
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1 ; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vins.f16 s0, s6 ; CHECK-MVE-NEXT: vins.f16 s0, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s1 ; CHECK-MVE-NEXT: vmovx.f16 s6, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s9 ; CHECK-MVE-NEXT: vmovx.f16 s6, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s4, s1 ; CHECK-MVE-NEXT: vcmp.f16 s4, s1
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: moveq r1, #1 ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1 ; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s6 ; CHECK-MVE-NEXT: vins.f16 s1, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s2 ; CHECK-MVE-NEXT: vmovx.f16 s6, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s10 ; CHECK-MVE-NEXT: vmovx.f16 s6, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s4, s2 ; CHECK-MVE-NEXT: vcmp.f16 s4, s2
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: moveq r1, #1 ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1 ; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s2, s6 ; CHECK-MVE-NEXT: vins.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s3 ; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vmovx.f16 s8, s15 ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s4, s3 ; CHECK-MVE-NEXT: vcmp.f16 s4, s3
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r0, #1
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r0, #1 ; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s6 ; CHECK-MVE-NEXT: vins.f16 s3, s6
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_ueq_v8f16: ; CHECK-MVEFP-LABEL: vcmp_r_ueq_v8f16:
Show All 10 Linesentry:
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_r_une_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_r_une_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_r_une_v8f16: ; CHECK-MVE-LABEL: vcmp_r_une_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s6, s0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s8 ; CHECK-MVE-NEXT: vmovx.f16 s6, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ne
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s4, s0 ; CHECK-MVE-NEXT: vcmp.f16 s4, s0
; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vins.f16 s0, s6 ; CHECK-MVE-NEXT: vins.f16 s0, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s1 ; CHECK-MVE-NEXT: vmovx.f16 s6, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s9 ; CHECK-MVE-NEXT: vmovx.f16 s6, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ne
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s4, s1 ; CHECK-MVE-NEXT: vcmp.f16 s4, s1
; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: movne r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s6 ; CHECK-MVE-NEXT: vins.f16 s1, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s2 ; CHECK-MVE-NEXT: vmovx.f16 s6, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s10 ; CHECK-MVE-NEXT: vmovx.f16 s6, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ne
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s4, s2 ; CHECK-MVE-NEXT: vcmp.f16 s4, s2
; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: movne r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s2, s6 ; CHECK-MVE-NEXT: vins.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s3 ; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s11 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ne
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s4, s3 ; CHECK-MVE-NEXT: vcmp.f16 s4, s3
; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: movne r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s6 ; CHECK-MVE-NEXT: vins.f16 s3, s6
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_une_v8f16: ; CHECK-MVEFP-LABEL: vcmp_r_une_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov.f16 r0, s4 ; CHECK-MVEFP-NEXT: vmov.f16 r0, s4
; CHECK-MVEFP-NEXT: vcmp.f16 ne, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f16 ne, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <8 x half> undef, half %src2, i32 0 %i = insertelement <8 x half> undef, half %src2, i32 0
%sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer %sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer
%c = fcmp une <8 x half> %sp, %src %c = fcmp une <8 x half> %sp, %src
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_r_ugt_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_r_ugt_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_r_ugt_v8f16: ; CHECK-MVE-LABEL: vcmp_r_ugt_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s6, s0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s8 ; CHECK-MVE-NEXT: vmovx.f16 s6, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it hi
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s4, s0 ; CHECK-MVE-NEXT: vcmp.f16 s4, s0
; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vins.f16 s0, s6 ; CHECK-MVE-NEXT: vins.f16 s0, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s1 ; CHECK-MVE-NEXT: vmovx.f16 s6, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s9 ; CHECK-MVE-NEXT: vmovx.f16 s6, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it hi
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s4, s1 ; CHECK-MVE-NEXT: vcmp.f16 s4, s1
; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: movhi r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s6 ; CHECK-MVE-NEXT: vins.f16 s1, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s2 ; CHECK-MVE-NEXT: vmovx.f16 s6, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s10 ; CHECK-MVE-NEXT: vmovx.f16 s6, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it hi
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s4, s2 ; CHECK-MVE-NEXT: vcmp.f16 s4, s2
; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: movhi r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s2, s6 ; CHECK-MVE-NEXT: vins.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s3 ; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s11 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it hi
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s4, s3 ; CHECK-MVE-NEXT: vcmp.f16 s4, s3
; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: movhi r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s6 ; CHECK-MVE-NEXT: vins.f16 s3, s6
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_ugt_v8f16: ; CHECK-MVEFP-LABEL: vcmp_r_ugt_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov.f16 r0, s4 ; CHECK-MVEFP-NEXT: vmov.f16 r0, s4
; CHECK-MVEFP-NEXT: vcmp.f16 lt, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f16 lt, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <8 x half> undef, half %src2, i32 0 %i = insertelement <8 x half> undef, half %src2, i32 0
%sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer %sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer
%c = fcmp ugt <8 x half> %sp, %src %c = fcmp ugt <8 x half> %sp, %src
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_r_uge_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_r_uge_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_r_uge_v8f16: ; CHECK-MVE-LABEL: vcmp_r_uge_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s6, s0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s8 ; CHECK-MVE-NEXT: vmovx.f16 s6, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it pl
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s4, s0 ; CHECK-MVE-NEXT: vcmp.f16 s4, s0
; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vins.f16 s0, s6 ; CHECK-MVE-NEXT: vins.f16 s0, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s1 ; CHECK-MVE-NEXT: vmovx.f16 s6, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s9 ; CHECK-MVE-NEXT: vmovx.f16 s6, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it pl
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s4, s1 ; CHECK-MVE-NEXT: vcmp.f16 s4, s1
; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: movpl r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s6 ; CHECK-MVE-NEXT: vins.f16 s1, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s2 ; CHECK-MVE-NEXT: vmovx.f16 s6, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s10 ; CHECK-MVE-NEXT: vmovx.f16 s6, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it pl
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s4, s2 ; CHECK-MVE-NEXT: vcmp.f16 s4, s2
; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: movpl r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s2, s6 ; CHECK-MVE-NEXT: vins.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s3 ; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s11 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it pl
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s4, s3 ; CHECK-MVE-NEXT: vcmp.f16 s4, s3
; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: movpl r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s6 ; CHECK-MVE-NEXT: vins.f16 s3, s6
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_uge_v8f16: ; CHECK-MVEFP-LABEL: vcmp_r_uge_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov.f16 r0, s4 ; CHECK-MVEFP-NEXT: vmov.f16 r0, s4
; CHECK-MVEFP-NEXT: vcmp.f16 le, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f16 le, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <8 x half> undef, half %src2, i32 0 %i = insertelement <8 x half> undef, half %src2, i32 0
%sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer %sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer
%c = fcmp uge <8 x half> %sp, %src %c = fcmp uge <8 x half> %sp, %src
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_r_ult_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_r_ult_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_r_ult_v8f16: ; CHECK-MVE-LABEL: vcmp_r_ult_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s6, s0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s8 ; CHECK-MVE-NEXT: vmovx.f16 s6, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it lt
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s4, s0 ; CHECK-MVE-NEXT: vcmp.f16 s4, s0
; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vins.f16 s0, s6 ; CHECK-MVE-NEXT: vins.f16 s0, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s1 ; CHECK-MVE-NEXT: vmovx.f16 s6, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s9 ; CHECK-MVE-NEXT: vmovx.f16 s6, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it lt
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s4, s1 ; CHECK-MVE-NEXT: vcmp.f16 s4, s1
; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: movlt r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s6 ; CHECK-MVE-NEXT: vins.f16 s1, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s2 ; CHECK-MVE-NEXT: vmovx.f16 s6, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s10 ; CHECK-MVE-NEXT: vmovx.f16 s6, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it lt
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s4, s2 ; CHECK-MVE-NEXT: vcmp.f16 s4, s2
; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: movlt r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s2, s6 ; CHECK-MVE-NEXT: vins.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s3 ; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s11 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it lt
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s4, s3 ; CHECK-MVE-NEXT: vcmp.f16 s4, s3
; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: movlt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s6 ; CHECK-MVE-NEXT: vins.f16 s3, s6
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_ult_v8f16: ; CHECK-MVEFP-LABEL: vcmp_r_ult_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov.f16 r0, s4 ; CHECK-MVEFP-NEXT: vmov.f16 r0, s4
; CHECK-MVEFP-NEXT: vcmp.f16 gt, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f16 gt, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <8 x half> undef, half %src2, i32 0 %i = insertelement <8 x half> undef, half %src2, i32 0
%sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer %sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer
%c = fcmp ult <8 x half> %sp, %src %c = fcmp ult <8 x half> %sp, %src
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_r_ule_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_r_ule_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_r_ule_v8f16: ; CHECK-MVE-LABEL: vcmp_r_ule_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s6, s0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s8 ; CHECK-MVE-NEXT: vmovx.f16 s6, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it le
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s4, s0 ; CHECK-MVE-NEXT: vcmp.f16 s4, s0
; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vins.f16 s0, s6 ; CHECK-MVE-NEXT: vins.f16 s0, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s1 ; CHECK-MVE-NEXT: vmovx.f16 s6, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s9 ; CHECK-MVE-NEXT: vmovx.f16 s6, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it le
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s4, s1 ; CHECK-MVE-NEXT: vcmp.f16 s4, s1
; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: movle r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s6 ; CHECK-MVE-NEXT: vins.f16 s1, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s2 ; CHECK-MVE-NEXT: vmovx.f16 s6, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s10 ; CHECK-MVE-NEXT: vmovx.f16 s6, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it le
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s4, s2 ; CHECK-MVE-NEXT: vcmp.f16 s4, s2
; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: movle r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s2, s6 ; CHECK-MVE-NEXT: vins.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s3 ; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s11 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it le
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s4, s3 ; CHECK-MVE-NEXT: vcmp.f16 s4, s3
; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: movle r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s6 ; CHECK-MVE-NEXT: vins.f16 s3, s6
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_ule_v8f16: ; CHECK-MVEFP-LABEL: vcmp_r_ule_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov.f16 r0, s4 ; CHECK-MVEFP-NEXT: vmov.f16 r0, s4
; CHECK-MVEFP-NEXT: vcmp.f16 ge, q0, r0 ; CHECK-MVEFP-NEXT: vcmp.f16 ge, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q2, q3 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q3
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <8 x half> undef, half %src2, i32 0 %i = insertelement <8 x half> undef, half %src2, i32 0
%sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer %sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer
%c = fcmp ule <8 x half> %sp, %src %c = fcmp ule <8 x half> %sp, %src
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_r_ord_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_r_ord_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_r_ord_v8f16: ; CHECK-MVE-LABEL: vcmp_r_ord_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s6, s0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s8 ; CHECK-MVE-NEXT: vmovx.f16 s6, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vc
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s4, s0 ; CHECK-MVE-NEXT: vcmp.f16 s4, s0
; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vins.f16 s0, s6 ; CHECK-MVE-NEXT: vins.f16 s0, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s1 ; CHECK-MVE-NEXT: vmovx.f16 s6, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s9 ; CHECK-MVE-NEXT: vmovx.f16 s6, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vc
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s4, s1 ; CHECK-MVE-NEXT: vcmp.f16 s4, s1
; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: movvc r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s6 ; CHECK-MVE-NEXT: vins.f16 s1, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s2 ; CHECK-MVE-NEXT: vmovx.f16 s6, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s10 ; CHECK-MVE-NEXT: vmovx.f16 s6, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vc
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s4, s2 ; CHECK-MVE-NEXT: vcmp.f16 s4, s2
; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: movvc r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s2, s6 ; CHECK-MVE-NEXT: vins.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s3 ; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s11 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vc
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s4, s3 ; CHECK-MVE-NEXT: vcmp.f16 s4, s3
; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: movvc r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s6 ; CHECK-MVE-NEXT: vins.f16 s3, s6
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_ord_v8f16: ; CHECK-MVEFP-LABEL: vcmp_r_ord_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov.f16 r0, s4 ; CHECK-MVEFP-NEXT: vmov.f16 r0, s4
; CHECK-MVEFP-NEXT: vpt.f16 le, q0, r0 ; CHECK-MVEFP-NEXT: vpt.f16 le, q0, r0
; CHECK-MVEFP-NEXT: vcmpt.f16 gt, q0, r0 ; CHECK-MVEFP-NEXT: vcmpt.f16 gt, q0, r0
; CHECK-MVEFP-NEXT: vpsel q0, q3, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q3, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%i = insertelement <8 x half> undef, half %src2, i32 0 %i = insertelement <8 x half> undef, half %src2, i32 0
%sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer %sp = shufflevector <8 x half> %i, <8 x half> undef, <8 x i32> zeroinitializer
%c = fcmp ord <8 x half> %sp, %src %c = fcmp ord <8 x half> %sp, %src
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_r_uno_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_r_uno_v8f16(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_r_uno_v8f16: ; CHECK-MVE-LABEL: vcmp_r_uno_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s6, s0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s8 ; CHECK-MVE-NEXT: vmovx.f16 s6, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s4, s0 ; CHECK-MVE-NEXT: vcmp.f16 s4, s0
; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vins.f16 s0, s6 ; CHECK-MVE-NEXT: vins.f16 s0, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s1 ; CHECK-MVE-NEXT: vmovx.f16 s6, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s9 ; CHECK-MVE-NEXT: vmovx.f16 s6, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s4, s1 ; CHECK-MVE-NEXT: vcmp.f16 s4, s1
; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: movvs r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s6 ; CHECK-MVE-NEXT: vins.f16 s1, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s2 ; CHECK-MVE-NEXT: vmovx.f16 s6, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s10 ; CHECK-MVE-NEXT: vmovx.f16 s6, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s4, s2 ; CHECK-MVE-NEXT: vcmp.f16 s4, s2
; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: movvs r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s2, s6 ; CHECK-MVE-NEXT: vins.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s3 ; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, s6 ; CHECK-MVE-NEXT: vcmp.f16 s4, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s11 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s4, s3 ; CHECK-MVE-NEXT: vcmp.f16 s4, s3
; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s6 ; CHECK-MVE-NEXT: vins.f16 s3, s6
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_uno_v8f16: ; CHECK-MVEFP-LABEL: vcmp_r_uno_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov.f16 r0, s4 ; CHECK-MVEFP-NEXT: vmov.f16 r0, s4
Show All 10 Lines
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_oeq_v8f16_bc(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_oeq_v8f16_bc(<8 x half> %src, half %src2, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_oeq_v8f16_bc: ; CHECK-MVE-LABEL: vcmp_oeq_v8f16_bc:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s6, s0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s8 ; CHECK-MVE-NEXT: vmovx.f16 s6, s8
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s5, s12
; CHECK-MVE-NEXT: vcmp.f16 s0, s4 ; CHECK-MVE-NEXT: vcmp.f16 s0, s4
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s5, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8 ; CHECK-MVE-NEXT: vseleq.f16 s0, s12, s8
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vins.f16 s0, s6 ; CHECK-MVE-NEXT: vins.f16 s0, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s1 ; CHECK-MVE-NEXT: vmovx.f16 s6, s1
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s9 ; CHECK-MVE-NEXT: vmovx.f16 s6, s9
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s13
; CHECK-MVE-NEXT: vcmp.f16 s1, s4 ; CHECK-MVE-NEXT: vcmp.f16 s1, s4
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9 ; CHECK-MVE-NEXT: vseleq.f16 s1, s13, s9
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s6 ; CHECK-MVE-NEXT: vins.f16 s1, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s2 ; CHECK-MVE-NEXT: vmovx.f16 s6, s2
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s10 ; CHECK-MVE-NEXT: vmovx.f16 s6, s10
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s14
; CHECK-MVE-NEXT: vcmp.f16 s2, s4 ; CHECK-MVE-NEXT: vcmp.f16 s2, s4
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10 ; CHECK-MVE-NEXT: vseleq.f16 s2, s14, s10
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s2, s6 ; CHECK-MVE-NEXT: vins.f16 s2, s6
; CHECK-MVE-NEXT: vmovx.f16 s6, s3 ; CHECK-MVE-NEXT: vmovx.f16 s6, s3
; CHECK-MVE-NEXT: vcmp.f16 s6, s4 ; CHECK-MVE-NEXT: vcmp.f16 s6, s4
; CHECK-MVE-NEXT: vmovx.f16 s6, s11 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s8, s15
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s4 ; CHECK-MVE-NEXT: vcmp.f16 s3, s4
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6 ; CHECK-MVE-NEXT: vseleq.f16 s6, s8, s6
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11 ; CHECK-MVE-NEXT: vseleq.f16 s3, s15, s11
; CHECK-MVE-NEXT: vins.f16 s3, s6 ; CHECK-MVE-NEXT: vins.f16 s3, s6
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_oeq_v8f16_bc: ; CHECK-MVEFP-LABEL: vcmp_oeq_v8f16_bc:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vmov.f16 r0, s4 ; CHECK-MVEFP-NEXT: vmov.f16 r0, s4
Show All 12 Lines

llvm/test/CodeGen/Thumb2/mve-vcmpfz.ll

; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -mtriple=thumbv8.1m.main-none-none-eabi -mattr=+mve,+fullfp16 -verify-machineinstrs %s -o - | FileCheck %s --check-prefix=CHECK-MVE ; RUN: llc -mtriple=thumbv8.1m.main-none-none-eabi -mattr=+mve,+fullfp16 -verify-machineinstrs %s -o - | FileCheck %s --check-prefix=CHECK-MVE
; RUN: llc -mtriple=thumbv8.1m.main-none-none-eabi -mattr=+mve.fp -verify-machineinstrs %s -o - | FileCheck %s --check-prefix=CHECK-MVEFP ; RUN: llc -mtriple=thumbv8.1m.main-none-none-eabi -mattr=+mve.fp -verify-machineinstrs %s -o - | FileCheck %s --check-prefix=CHECK-MVEFP
define arm_aapcs_vfpcc <4 x float> @vcmp_oeq_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_oeq_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_oeq_v4f32: ; CHECK-MVE-LABEL: vcmp_oeq_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, #0 ; CHECK-MVE-NEXT: vcmp.f32 s1, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, #0 ; CHECK-MVE-NEXT: vcmp.f32 s0, #0
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, #0 ; CHECK-MVE-NEXT: vcmp.f32 s3, #0
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r1, eq
; CHECK-MVE-NEXT: moveq r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, #0 ; CHECK-MVE-NEXT: vcmp.f32 s2, #0
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r2, eq
; CHECK-MVE-NEXT: moveq r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r3, eq
; CHECK-MVE-NEXT: moveq r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_oeq_v4f32: ; CHECK-MVEFP-LABEL: vcmp_oeq_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f32 eq, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f32 eq, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp oeq <4 x float> %src, zeroinitializer %c = fcmp oeq <4 x float> %src, zeroinitializer
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_one_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_one_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_one_v4f32: ; CHECK-MVE-LABEL: vcmp_one_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, #0 ; CHECK-MVE-NEXT: vcmp.f32 s1, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s1, #0 ; CHECK-MVE-NEXT: vcmp.f32 s1, #0
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, #0 ; CHECK-MVE-NEXT: vcmp.f32 s0, #0
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1 ; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r2, #1
; CHECK-MVE-NEXT: vcmp.f32 s3, #0 ; CHECK-MVE-NEXT: vcmp.f32 s3, #0
; CHECK-MVE-NEXT: mov.w r3, #0 ; CHECK-MVE-NEXT: cset r1, mi
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r2, #1 ; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r3, #1
; CHECK-MVE-NEXT: mov.w r0, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, #0 ; CHECK-MVE-NEXT: vcmp.f32 s2, #0
; CHECK-MVE-NEXT: cset r2, mi
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r3, #1 ; CHECK-MVE-NEXT: movgt r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r3, mi
; CHECK-MVE-NEXT: movmi r0, #1
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r0, #1 ; CHECK-MVE-NEXT: movgt r3, #1
; CHECK-MVE-NEXT: cmp r3, #0 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_one_v4f32: ; CHECK-MVEFP-LABEL: vcmp_one_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vpt.f32 ge, q0, zr ; CHECK-MVEFP-NEXT: vpt.f32 ge, q0, zr
; CHECK-MVEFP-NEXT: vcmpt.f32 le, q0, zr ; CHECK-MVEFP-NEXT: vcmpt.f32 le, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q2, q1 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q1
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp one <4 x float> %src, zeroinitializer %c = fcmp one <4 x float> %src, zeroinitializer
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_ogt_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_ogt_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_ogt_v4f32: ; CHECK-MVE-LABEL: vcmp_ogt_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, #0 ; CHECK-MVE-NEXT: vcmp.f32 s1, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, #0 ; CHECK-MVE-NEXT: vcmp.f32 s0, #0
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, #0 ; CHECK-MVE-NEXT: vcmp.f32 s3, #0
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r1, gt
; CHECK-MVE-NEXT: movgt r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, #0 ; CHECK-MVE-NEXT: vcmp.f32 s2, #0
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r2, gt
; CHECK-MVE-NEXT: movgt r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r3, gt
; CHECK-MVE-NEXT: movgt r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ogt_v4f32: ; CHECK-MVEFP-LABEL: vcmp_ogt_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f32 gt, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f32 gt, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ogt <4 x float> %src, zeroinitializer %c = fcmp ogt <4 x float> %src, zeroinitializer
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_oge_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_oge_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_oge_v4f32: ; CHECK-MVE-LABEL: vcmp_oge_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, #0 ; CHECK-MVE-NEXT: vcmp.f32 s1, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, #0 ; CHECK-MVE-NEXT: vcmp.f32 s0, #0
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, #0 ; CHECK-MVE-NEXT: vcmp.f32 s3, #0
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r1, ge
; CHECK-MVE-NEXT: movge r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, #0 ; CHECK-MVE-NEXT: vcmp.f32 s2, #0
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r2, ge
; CHECK-MVE-NEXT: movge r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r3, ge
; CHECK-MVE-NEXT: movge r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_oge_v4f32: ; CHECK-MVEFP-LABEL: vcmp_oge_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f32 ge, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f32 ge, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp oge <4 x float> %src, zeroinitializer %c = fcmp oge <4 x float> %src, zeroinitializer
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_olt_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_olt_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_olt_v4f32: ; CHECK-MVE-LABEL: vcmp_olt_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, #0 ; CHECK-MVE-NEXT: vcmp.f32 s1, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, #0 ; CHECK-MVE-NEXT: vcmp.f32 s0, #0
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, #0 ; CHECK-MVE-NEXT: vcmp.f32 s3, #0
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r1, mi
; CHECK-MVE-NEXT: movmi r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, #0 ; CHECK-MVE-NEXT: vcmp.f32 s2, #0
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r2, mi
; CHECK-MVE-NEXT: movmi r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r3, mi
; CHECK-MVE-NEXT: movmi r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_olt_v4f32: ; CHECK-MVEFP-LABEL: vcmp_olt_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f32 lt, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f32 lt, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp olt <4 x float> %src, zeroinitializer %c = fcmp olt <4 x float> %src, zeroinitializer
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_ole_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_ole_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_ole_v4f32: ; CHECK-MVE-LABEL: vcmp_ole_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, #0 ; CHECK-MVE-NEXT: vcmp.f32 s1, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, #0 ; CHECK-MVE-NEXT: vcmp.f32 s0, #0
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, #0 ; CHECK-MVE-NEXT: vcmp.f32 s3, #0
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r1, ls
; CHECK-MVE-NEXT: movls r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, #0 ; CHECK-MVE-NEXT: vcmp.f32 s2, #0
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r2, ls
; CHECK-MVE-NEXT: movls r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r3, ls
; CHECK-MVE-NEXT: movls r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ole_v4f32: ; CHECK-MVEFP-LABEL: vcmp_ole_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f32 le, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f32 le, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ole <4 x float> %src, zeroinitializer %c = fcmp ole <4 x float> %src, zeroinitializer
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_ueq_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_ueq_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_ueq_v4f32: ; CHECK-MVE-LABEL: vcmp_ueq_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, #0 ; CHECK-MVE-NEXT: vcmp.f32 s1, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s1, #0 ; CHECK-MVE-NEXT: vcmp.f32 s1, #0
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, #0 ; CHECK-MVE-NEXT: vcmp.f32 s0, #0
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1 ; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r2, #1
; CHECK-MVE-NEXT: vcmp.f32 s3, #0 ; CHECK-MVE-NEXT: vcmp.f32 s3, #0
; CHECK-MVE-NEXT: mov.w r3, #0 ; CHECK-MVE-NEXT: cset r1, eq
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r2, #1 ; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r3, #1
; CHECK-MVE-NEXT: mov.w r0, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, #0 ; CHECK-MVE-NEXT: vcmp.f32 s2, #0
; CHECK-MVE-NEXT: cset r2, eq
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r3, #1 ; CHECK-MVE-NEXT: movvs r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r3, eq
; CHECK-MVE-NEXT: moveq r0, #1
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r0, #1 ; CHECK-MVE-NEXT: movvs r3, #1
; CHECK-MVE-NEXT: cmp r3, #0 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ueq_v4f32: ; CHECK-MVEFP-LABEL: vcmp_ueq_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vpt.f32 ge, q0, zr ; CHECK-MVEFP-NEXT: vpt.f32 ge, q0, zr
; CHECK-MVEFP-NEXT: vcmpt.f32 le, q0, zr ; CHECK-MVEFP-NEXT: vcmpt.f32 le, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ueq <4 x float> %src, zeroinitializer %c = fcmp ueq <4 x float> %src, zeroinitializer
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_une_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_une_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_une_v4f32: ; CHECK-MVE-LABEL: vcmp_une_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, #0 ; CHECK-MVE-NEXT: vcmp.f32 s1, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, #0 ; CHECK-MVE-NEXT: vcmp.f32 s0, #0
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, #0 ; CHECK-MVE-NEXT: vcmp.f32 s3, #0
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r1, ne
; CHECK-MVE-NEXT: movne r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, #0 ; CHECK-MVE-NEXT: vcmp.f32 s2, #0
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r2, ne
; CHECK-MVE-NEXT: movne r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r3, ne
; CHECK-MVE-NEXT: movne r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_une_v4f32: ; CHECK-MVEFP-LABEL: vcmp_une_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f32 ne, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f32 ne, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp une <4 x float> %src, zeroinitializer %c = fcmp une <4 x float> %src, zeroinitializer
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_ugt_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_ugt_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_ugt_v4f32: ; CHECK-MVE-LABEL: vcmp_ugt_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, #0 ; CHECK-MVE-NEXT: vcmp.f32 s1, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, #0 ; CHECK-MVE-NEXT: vcmp.f32 s0, #0
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, #0 ; CHECK-MVE-NEXT: vcmp.f32 s3, #0
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r1, hi
; CHECK-MVE-NEXT: movhi r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, #0 ; CHECK-MVE-NEXT: vcmp.f32 s2, #0
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r2, hi
; CHECK-MVE-NEXT: movhi r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r3, hi
; CHECK-MVE-NEXT: movhi r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ugt_v4f32: ; CHECK-MVEFP-LABEL: vcmp_ugt_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f32 gt, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f32 gt, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ugt <4 x float> %src, zeroinitializer %c = fcmp ugt <4 x float> %src, zeroinitializer
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_uge_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_uge_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_uge_v4f32: ; CHECK-MVE-LABEL: vcmp_uge_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, #0 ; CHECK-MVE-NEXT: vcmp.f32 s1, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, #0 ; CHECK-MVE-NEXT: vcmp.f32 s0, #0
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, #0 ; CHECK-MVE-NEXT: vcmp.f32 s3, #0
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r1, pl
; CHECK-MVE-NEXT: movpl r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, #0 ; CHECK-MVE-NEXT: vcmp.f32 s2, #0
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r2, pl
; CHECK-MVE-NEXT: movpl r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r3, pl
; CHECK-MVE-NEXT: movpl r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_uge_v4f32: ; CHECK-MVEFP-LABEL: vcmp_uge_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f32 ge, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f32 ge, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp uge <4 x float> %src, zeroinitializer %c = fcmp uge <4 x float> %src, zeroinitializer
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_ult_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_ult_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_ult_v4f32: ; CHECK-MVE-LABEL: vcmp_ult_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, #0 ; CHECK-MVE-NEXT: vcmp.f32 s1, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, #0 ; CHECK-MVE-NEXT: vcmp.f32 s0, #0
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, #0 ; CHECK-MVE-NEXT: vcmp.f32 s3, #0
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r1, lt
; CHECK-MVE-NEXT: movlt r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, #0 ; CHECK-MVE-NEXT: vcmp.f32 s2, #0
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r2, lt
; CHECK-MVE-NEXT: movlt r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r3, lt
; CHECK-MVE-NEXT: movlt r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ult_v4f32: ; CHECK-MVEFP-LABEL: vcmp_ult_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f32 lt, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f32 lt, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ult <4 x float> %src, zeroinitializer %c = fcmp ult <4 x float> %src, zeroinitializer
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_ule_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_ule_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_ule_v4f32: ; CHECK-MVE-LABEL: vcmp_ule_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, #0 ; CHECK-MVE-NEXT: vcmp.f32 s1, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, #0 ; CHECK-MVE-NEXT: vcmp.f32 s0, #0
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, #0 ; CHECK-MVE-NEXT: vcmp.f32 s3, #0
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r1, le
; CHECK-MVE-NEXT: movle r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, #0 ; CHECK-MVE-NEXT: vcmp.f32 s2, #0
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r2, le
; CHECK-MVE-NEXT: movle r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r3, le
; CHECK-MVE-NEXT: movle r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ule_v4f32: ; CHECK-MVEFP-LABEL: vcmp_ule_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f32 le, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f32 le, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ule <4 x float> %src, zeroinitializer %c = fcmp ule <4 x float> %src, zeroinitializer
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_ord_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_ord_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_ord_v4f32: ; CHECK-MVE-LABEL: vcmp_ord_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s1 ; CHECK-MVE-NEXT: vcmp.f32 s1, s1
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s0 ; CHECK-MVE-NEXT: vcmp.f32 s0, s0
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, s3 ; CHECK-MVE-NEXT: vcmp.f32 s3, s3
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r1, vc
; CHECK-MVE-NEXT: movvc r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s2 ; CHECK-MVE-NEXT: vcmp.f32 s2, s2
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r2, vc
; CHECK-MVE-NEXT: movvc r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r3, vc
; CHECK-MVE-NEXT: movvc r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ord_v4f32: ; CHECK-MVEFP-LABEL: vcmp_ord_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vpt.f32 ge, q0, zr ; CHECK-MVEFP-NEXT: vpt.f32 ge, q0, zr
; CHECK-MVEFP-NEXT: vcmpt.f32 lt, q0, zr ; CHECK-MVEFP-NEXT: vcmpt.f32 lt, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q2, q1 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q1
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ord <4 x float> %src, zeroinitializer %c = fcmp ord <4 x float> %src, zeroinitializer
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_uno_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_uno_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_uno_v4f32: ; CHECK-MVE-LABEL: vcmp_uno_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s1 ; CHECK-MVE-NEXT: vcmp.f32 s1, s1
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s0 ; CHECK-MVE-NEXT: vcmp.f32 s0, s0
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, s3 ; CHECK-MVE-NEXT: vcmp.f32 s3, s3
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r1, vs
; CHECK-MVE-NEXT: movvs r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s2 ; CHECK-MVE-NEXT: vcmp.f32 s2, s2
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r2, vs
; CHECK-MVE-NEXT: movvs r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r3, vs
; CHECK-MVE-NEXT: movvs r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_uno_v4f32: ; CHECK-MVEFP-LABEL: vcmp_uno_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vpt.f32 ge, q0, zr ; CHECK-MVEFP-NEXT: vpt.f32 ge, q0, zr
; CHECK-MVEFP-NEXT: vcmpt.f32 lt, q0, zr ; CHECK-MVEFP-NEXT: vcmpt.f32 lt, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp uno <4 x float> %src, zeroinitializer %c = fcmp uno <4 x float> %src, zeroinitializer
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_oeq_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_oeq_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_oeq_v8f16: ; CHECK-MVE-LABEL: vcmp_oeq_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s12, s0 ; CHECK-MVE-NEXT: vmovx.f16 s12, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s12, #0 ; CHECK-MVE-NEXT: vcmp.f16 s12, #0
; CHECK-MVE-NEXT: vmovx.f16 s12, s4 ; CHECK-MVE-NEXT: vmovx.f16 s12, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s0, #0 ; CHECK-MVE-NEXT: vcmp.f16 s0, #0
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12 ; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s1 ; CHECK-MVE-NEXT: vmovx.f16 s4, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s9 ; CHECK-MVE-NEXT: vmovx.f16 s8, s9
; CHECK-MVE-NEXT: vcmp.f16 s1, #0 ; CHECK-MVE-NEXT: vcmp.f16 s1, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s10 ; CHECK-MVE-NEXT: vmovx.f16 s8, s10
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s2 ; CHECK-MVE-NEXT: vmovx.f16 s4, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s2, #0 ; CHECK-MVE-NEXT: vcmp.f16 s2, #0
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: mov.w r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s3 ; CHECK-MVE-NEXT: vmovx.f16 s4, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, #0 ; CHECK-MVE-NEXT: vcmp.f16 s3, #0
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_oeq_v8f16: ; CHECK-MVEFP-LABEL: vcmp_oeq_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f16 eq, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f16 eq, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp oeq <8 x half> %src, zeroinitializer %c = fcmp oeq <8 x half> %src, zeroinitializer
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_one_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_one_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_one_v8f16: ; CHECK-MVE-LABEL: vcmp_one_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s12, s0 ; CHECK-MVE-NEXT: vmovx.f16 s12, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s12, #0 ; CHECK-MVE-NEXT: vcmp.f16 s12, #0
; CHECK-MVE-NEXT: vmovx.f16 s12, s4 ; CHECK-MVE-NEXT: vmovx.f16 s12, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s0, #0 ; CHECK-MVE-NEXT: vcmp.f16 s0, #0
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12 ; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1 ; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s1 ; CHECK-MVE-NEXT: vmovx.f16 s4, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s9 ; CHECK-MVE-NEXT: vmovx.f16 s8, s9
; CHECK-MVE-NEXT: vcmp.f16 s1, #0 ; CHECK-MVE-NEXT: vcmp.f16 s1, #0
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s8, s10 ; CHECK-MVE-NEXT: vmovx.f16 s8, s10
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1 ; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s2 ; CHECK-MVE-NEXT: vmovx.f16 s4, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s2, #0 ; CHECK-MVE-NEXT: vcmp.f16 s2, #0
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: mov.w r0, #0
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1 ; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s3 ; CHECK-MVE-NEXT: vmovx.f16 s4, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, #0 ; CHECK-MVE-NEXT: vcmp.f16 s3, #0
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r0, #1
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r0, #1 ; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_one_v8f16: ; CHECK-MVEFP-LABEL: vcmp_one_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vpt.f16 ge, q0, zr ; CHECK-MVEFP-NEXT: vpt.f16 ge, q0, zr
; CHECK-MVEFP-NEXT: vcmpt.f16 le, q0, zr ; CHECK-MVEFP-NEXT: vcmpt.f16 le, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q2, q1 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q1
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp one <8 x half> %src, zeroinitializer %c = fcmp one <8 x half> %src, zeroinitializer
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_ogt_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_ogt_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_ogt_v8f16: ; CHECK-MVE-LABEL: vcmp_ogt_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s12, s0 ; CHECK-MVE-NEXT: vmovx.f16 s12, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s12, #0 ; CHECK-MVE-NEXT: vcmp.f16 s12, #0
; CHECK-MVE-NEXT: vmovx.f16 s12, s4 ; CHECK-MVE-NEXT: vmovx.f16 s12, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s0, #0 ; CHECK-MVE-NEXT: vcmp.f16 s0, #0
; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12 ; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s1 ; CHECK-MVE-NEXT: vmovx.f16 s4, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s9 ; CHECK-MVE-NEXT: vmovx.f16 s8, s9
; CHECK-MVE-NEXT: vcmp.f16 s1, #0 ; CHECK-MVE-NEXT: vcmp.f16 s1, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s10 ; CHECK-MVE-NEXT: vmovx.f16 s8, s10
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: movgt r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s2 ; CHECK-MVE-NEXT: vmovx.f16 s4, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s2, #0 ; CHECK-MVE-NEXT: vcmp.f16 s2, #0
; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: mov.w r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s3 ; CHECK-MVE-NEXT: vmovx.f16 s4, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, #0 ; CHECK-MVE-NEXT: vcmp.f16 s3, #0
; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ogt_v8f16: ; CHECK-MVEFP-LABEL: vcmp_ogt_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f16 gt, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f16 gt, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ogt <8 x half> %src, zeroinitializer %c = fcmp ogt <8 x half> %src, zeroinitializer
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_oge_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_oge_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_oge_v8f16: ; CHECK-MVE-LABEL: vcmp_oge_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s12, s0 ; CHECK-MVE-NEXT: vmovx.f16 s12, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s12, #0 ; CHECK-MVE-NEXT: vcmp.f16 s12, #0
; CHECK-MVE-NEXT: vmovx.f16 s12, s4 ; CHECK-MVE-NEXT: vmovx.f16 s12, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ge
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s0, #0 ; CHECK-MVE-NEXT: vcmp.f16 s0, #0
; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12 ; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s1 ; CHECK-MVE-NEXT: vmovx.f16 s4, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ge
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s9 ; CHECK-MVE-NEXT: vmovx.f16 s8, s9
; CHECK-MVE-NEXT: vcmp.f16 s1, #0 ; CHECK-MVE-NEXT: vcmp.f16 s1, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s10 ; CHECK-MVE-NEXT: vmovx.f16 s8, s10
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: movge r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s2 ; CHECK-MVE-NEXT: vmovx.f16 s4, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ge
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s2, #0 ; CHECK-MVE-NEXT: vcmp.f16 s2, #0
; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: mov.w r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it ge
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s3 ; CHECK-MVE-NEXT: vmovx.f16 s4, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ge
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, #0 ; CHECK-MVE-NEXT: vcmp.f16 s3, #0
; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: movge r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_oge_v8f16: ; CHECK-MVEFP-LABEL: vcmp_oge_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f16 ge, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f16 ge, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp oge <8 x half> %src, zeroinitializer %c = fcmp oge <8 x half> %src, zeroinitializer
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_olt_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_olt_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_olt_v8f16: ; CHECK-MVE-LABEL: vcmp_olt_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s12, s0 ; CHECK-MVE-NEXT: vmovx.f16 s12, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s12, #0 ; CHECK-MVE-NEXT: vcmp.f16 s12, #0
; CHECK-MVE-NEXT: vmovx.f16 s12, s4 ; CHECK-MVE-NEXT: vmovx.f16 s12, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s0, #0 ; CHECK-MVE-NEXT: vcmp.f16 s0, #0
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12 ; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s1 ; CHECK-MVE-NEXT: vmovx.f16 s4, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s9 ; CHECK-MVE-NEXT: vmovx.f16 s8, s9
; CHECK-MVE-NEXT: vcmp.f16 s1, #0 ; CHECK-MVE-NEXT: vcmp.f16 s1, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s10 ; CHECK-MVE-NEXT: vmovx.f16 s8, s10
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s2 ; CHECK-MVE-NEXT: vmovx.f16 s4, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s2, #0 ; CHECK-MVE-NEXT: vcmp.f16 s2, #0
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: mov.w r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s3 ; CHECK-MVE-NEXT: vmovx.f16 s4, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, #0 ; CHECK-MVE-NEXT: vcmp.f16 s3, #0
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_olt_v8f16: ; CHECK-MVEFP-LABEL: vcmp_olt_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f16 lt, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f16 lt, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp olt <8 x half> %src, zeroinitializer %c = fcmp olt <8 x half> %src, zeroinitializer
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_ole_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_ole_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_ole_v8f16: ; CHECK-MVE-LABEL: vcmp_ole_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s12, s0 ; CHECK-MVE-NEXT: vmovx.f16 s12, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s12, #0 ; CHECK-MVE-NEXT: vcmp.f16 s12, #0
; CHECK-MVE-NEXT: vmovx.f16 s12, s4 ; CHECK-MVE-NEXT: vmovx.f16 s12, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ls
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s0, #0 ; CHECK-MVE-NEXT: vcmp.f16 s0, #0
; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12 ; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s1 ; CHECK-MVE-NEXT: vmovx.f16 s4, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ls
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s9 ; CHECK-MVE-NEXT: vmovx.f16 s8, s9
; CHECK-MVE-NEXT: vcmp.f16 s1, #0 ; CHECK-MVE-NEXT: vcmp.f16 s1, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s10 ; CHECK-MVE-NEXT: vmovx.f16 s8, s10
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: movls r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s2 ; CHECK-MVE-NEXT: vmovx.f16 s4, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ls
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s2, #0 ; CHECK-MVE-NEXT: vcmp.f16 s2, #0
; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: mov.w r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it ls
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s3 ; CHECK-MVE-NEXT: vmovx.f16 s4, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ls
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, #0 ; CHECK-MVE-NEXT: vcmp.f16 s3, #0
; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: movls r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ole_v8f16: ; CHECK-MVEFP-LABEL: vcmp_ole_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f16 le, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f16 le, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ole <8 x half> %src, zeroinitializer %c = fcmp ole <8 x half> %src, zeroinitializer
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_ueq_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_ueq_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_ueq_v8f16: ; CHECK-MVE-LABEL: vcmp_ueq_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s12, s0 ; CHECK-MVE-NEXT: vmovx.f16 s12, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s12, #0 ; CHECK-MVE-NEXT: vcmp.f16 s12, #0
; CHECK-MVE-NEXT: vmovx.f16 s12, s4 ; CHECK-MVE-NEXT: vmovx.f16 s12, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s0, #0 ; CHECK-MVE-NEXT: vcmp.f16 s0, #0
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12 ; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1 ; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s1 ; CHECK-MVE-NEXT: vmovx.f16 s4, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s9 ; CHECK-MVE-NEXT: vmovx.f16 s8, s9
; CHECK-MVE-NEXT: vcmp.f16 s1, #0 ; CHECK-MVE-NEXT: vcmp.f16 s1, #0
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s8, s10 ; CHECK-MVE-NEXT: vmovx.f16 s8, s10
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1 ; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s2 ; CHECK-MVE-NEXT: vmovx.f16 s4, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s2, #0 ; CHECK-MVE-NEXT: vcmp.f16 s2, #0
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: mov.w r0, #0
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1 ; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s3 ; CHECK-MVE-NEXT: vmovx.f16 s4, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, #0 ; CHECK-MVE-NEXT: vcmp.f16 s3, #0
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r0, #1
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r0, #1 ; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ueq_v8f16: ; CHECK-MVEFP-LABEL: vcmp_ueq_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vpt.f16 ge, q0, zr ; CHECK-MVEFP-NEXT: vpt.f16 ge, q0, zr
; CHECK-MVEFP-NEXT: vcmpt.f16 le, q0, zr ; CHECK-MVEFP-NEXT: vcmpt.f16 le, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ueq <8 x half> %src, zeroinitializer %c = fcmp ueq <8 x half> %src, zeroinitializer
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_une_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_une_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_une_v8f16: ; CHECK-MVE-LABEL: vcmp_une_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s12, s0 ; CHECK-MVE-NEXT: vmovx.f16 s12, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s12, #0 ; CHECK-MVE-NEXT: vcmp.f16 s12, #0
; CHECK-MVE-NEXT: vmovx.f16 s12, s4 ; CHECK-MVE-NEXT: vmovx.f16 s12, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ne
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s0, #0 ; CHECK-MVE-NEXT: vcmp.f16 s0, #0
; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12 ; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s1 ; CHECK-MVE-NEXT: vmovx.f16 s4, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ne
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s9 ; CHECK-MVE-NEXT: vmovx.f16 s8, s9
; CHECK-MVE-NEXT: vcmp.f16 s1, #0 ; CHECK-MVE-NEXT: vcmp.f16 s1, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s10 ; CHECK-MVE-NEXT: vmovx.f16 s8, s10
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: movne r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s2 ; CHECK-MVE-NEXT: vmovx.f16 s4, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ne
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s2, #0 ; CHECK-MVE-NEXT: vcmp.f16 s2, #0
; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: mov.w r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it ne
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s3 ; CHECK-MVE-NEXT: vmovx.f16 s4, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ne
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, #0 ; CHECK-MVE-NEXT: vcmp.f16 s3, #0
; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: movne r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_une_v8f16: ; CHECK-MVEFP-LABEL: vcmp_une_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f16 ne, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f16 ne, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp une <8 x half> %src, zeroinitializer %c = fcmp une <8 x half> %src, zeroinitializer
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_ugt_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_ugt_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_ugt_v8f16: ; CHECK-MVE-LABEL: vcmp_ugt_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s12, s0 ; CHECK-MVE-NEXT: vmovx.f16 s12, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s12, #0 ; CHECK-MVE-NEXT: vcmp.f16 s12, #0
; CHECK-MVE-NEXT: vmovx.f16 s12, s4 ; CHECK-MVE-NEXT: vmovx.f16 s12, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it hi
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s0, #0 ; CHECK-MVE-NEXT: vcmp.f16 s0, #0
; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12 ; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s1 ; CHECK-MVE-NEXT: vmovx.f16 s4, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it hi
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s9 ; CHECK-MVE-NEXT: vmovx.f16 s8, s9
; CHECK-MVE-NEXT: vcmp.f16 s1, #0 ; CHECK-MVE-NEXT: vcmp.f16 s1, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s10 ; CHECK-MVE-NEXT: vmovx.f16 s8, s10
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: movhi r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s2 ; CHECK-MVE-NEXT: vmovx.f16 s4, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it hi
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s2, #0 ; CHECK-MVE-NEXT: vcmp.f16 s2, #0
; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: mov.w r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it hi
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s3 ; CHECK-MVE-NEXT: vmovx.f16 s4, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it hi
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, #0 ; CHECK-MVE-NEXT: vcmp.f16 s3, #0
; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: movhi r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ugt_v8f16: ; CHECK-MVEFP-LABEL: vcmp_ugt_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f16 gt, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f16 gt, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ugt <8 x half> %src, zeroinitializer %c = fcmp ugt <8 x half> %src, zeroinitializer
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_uge_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_uge_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_uge_v8f16: ; CHECK-MVE-LABEL: vcmp_uge_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s12, s0 ; CHECK-MVE-NEXT: vmovx.f16 s12, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s12, #0 ; CHECK-MVE-NEXT: vcmp.f16 s12, #0
; CHECK-MVE-NEXT: vmovx.f16 s12, s4 ; CHECK-MVE-NEXT: vmovx.f16 s12, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it pl
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s0, #0 ; CHECK-MVE-NEXT: vcmp.f16 s0, #0
; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12 ; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s1 ; CHECK-MVE-NEXT: vmovx.f16 s4, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it pl
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s9 ; CHECK-MVE-NEXT: vmovx.f16 s8, s9
; CHECK-MVE-NEXT: vcmp.f16 s1, #0 ; CHECK-MVE-NEXT: vcmp.f16 s1, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s10 ; CHECK-MVE-NEXT: vmovx.f16 s8, s10
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: movpl r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s2 ; CHECK-MVE-NEXT: vmovx.f16 s4, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it pl
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s2, #0 ; CHECK-MVE-NEXT: vcmp.f16 s2, #0
; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: mov.w r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it pl
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s3 ; CHECK-MVE-NEXT: vmovx.f16 s4, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it pl
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, #0 ; CHECK-MVE-NEXT: vcmp.f16 s3, #0
; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: movpl r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_uge_v8f16: ; CHECK-MVEFP-LABEL: vcmp_uge_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f16 ge, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f16 ge, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp uge <8 x half> %src, zeroinitializer %c = fcmp uge <8 x half> %src, zeroinitializer
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_ult_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_ult_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_ult_v8f16: ; CHECK-MVE-LABEL: vcmp_ult_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s12, s0 ; CHECK-MVE-NEXT: vmovx.f16 s12, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s12, #0 ; CHECK-MVE-NEXT: vcmp.f16 s12, #0
; CHECK-MVE-NEXT: vmovx.f16 s12, s4 ; CHECK-MVE-NEXT: vmovx.f16 s12, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it lt
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s0, #0 ; CHECK-MVE-NEXT: vcmp.f16 s0, #0
; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12 ; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s1 ; CHECK-MVE-NEXT: vmovx.f16 s4, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it lt
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s9 ; CHECK-MVE-NEXT: vmovx.f16 s8, s9
; CHECK-MVE-NEXT: vcmp.f16 s1, #0 ; CHECK-MVE-NEXT: vcmp.f16 s1, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s10 ; CHECK-MVE-NEXT: vmovx.f16 s8, s10
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: movlt r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s2 ; CHECK-MVE-NEXT: vmovx.f16 s4, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it lt
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s2, #0 ; CHECK-MVE-NEXT: vcmp.f16 s2, #0
; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: mov.w r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it lt
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s3 ; CHECK-MVE-NEXT: vmovx.f16 s4, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it lt
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, #0 ; CHECK-MVE-NEXT: vcmp.f16 s3, #0
; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: movlt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ult_v8f16: ; CHECK-MVEFP-LABEL: vcmp_ult_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f16 lt, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f16 lt, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ult <8 x half> %src, zeroinitializer %c = fcmp ult <8 x half> %src, zeroinitializer
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_ule_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_ule_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_ule_v8f16: ; CHECK-MVE-LABEL: vcmp_ule_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s12, s0 ; CHECK-MVE-NEXT: vmovx.f16 s12, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s12, #0 ; CHECK-MVE-NEXT: vcmp.f16 s12, #0
; CHECK-MVE-NEXT: vmovx.f16 s12, s4 ; CHECK-MVE-NEXT: vmovx.f16 s12, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it le
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s0, #0 ; CHECK-MVE-NEXT: vcmp.f16 s0, #0
; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12 ; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s1 ; CHECK-MVE-NEXT: vmovx.f16 s4, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it le
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s9 ; CHECK-MVE-NEXT: vmovx.f16 s8, s9
; CHECK-MVE-NEXT: vcmp.f16 s1, #0 ; CHECK-MVE-NEXT: vcmp.f16 s1, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s10 ; CHECK-MVE-NEXT: vmovx.f16 s8, s10
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: movle r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s2 ; CHECK-MVE-NEXT: vmovx.f16 s4, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it le
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s2, #0 ; CHECK-MVE-NEXT: vcmp.f16 s2, #0
; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: mov.w r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it le
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s3 ; CHECK-MVE-NEXT: vmovx.f16 s4, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it le
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, #0 ; CHECK-MVE-NEXT: vcmp.f16 s3, #0
; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: movle r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ule_v8f16: ; CHECK-MVEFP-LABEL: vcmp_ule_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f16 le, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f16 le, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ule <8 x half> %src, zeroinitializer %c = fcmp ule <8 x half> %src, zeroinitializer
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_ord_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_ord_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_ord_v8f16: ; CHECK-MVE-LABEL: vcmp_ord_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s12, s0 ; CHECK-MVE-NEXT: vmovx.f16 s12, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s12, s12 ; CHECK-MVE-NEXT: vcmp.f16 s12, s12
; CHECK-MVE-NEXT: vmovx.f16 s12, s4 ; CHECK-MVE-NEXT: vmovx.f16 s12, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vc
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s0, s0 ; CHECK-MVE-NEXT: vcmp.f16 s0, s0
; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12 ; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s1 ; CHECK-MVE-NEXT: vmovx.f16 s4, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, s4 ; CHECK-MVE-NEXT: vcmp.f16 s4, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vc
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s9 ; CHECK-MVE-NEXT: vmovx.f16 s8, s9
; CHECK-MVE-NEXT: vcmp.f16 s1, s1 ; CHECK-MVE-NEXT: vcmp.f16 s1, s1
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s10 ; CHECK-MVE-NEXT: vmovx.f16 s8, s10
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: movvc r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s2 ; CHECK-MVE-NEXT: vmovx.f16 s4, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, s4 ; CHECK-MVE-NEXT: vcmp.f16 s4, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vc
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s2, s2 ; CHECK-MVE-NEXT: vcmp.f16 s2, s2
; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: mov.w r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it vc
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s3 ; CHECK-MVE-NEXT: vmovx.f16 s4, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, s4 ; CHECK-MVE-NEXT: vcmp.f16 s4, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vc
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s3 ; CHECK-MVE-NEXT: vcmp.f16 s3, s3
; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: movvc r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_ord_v8f16: ; CHECK-MVEFP-LABEL: vcmp_ord_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vpt.f16 ge, q0, zr ; CHECK-MVEFP-NEXT: vpt.f16 ge, q0, zr
; CHECK-MVEFP-NEXT: vcmpt.f16 lt, q0, zr ; CHECK-MVEFP-NEXT: vcmpt.f16 lt, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q2, q1 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q1
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ord <8 x half> %src, zeroinitializer %c = fcmp ord <8 x half> %src, zeroinitializer
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_uno_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_uno_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_uno_v8f16: ; CHECK-MVE-LABEL: vcmp_uno_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s12, s0 ; CHECK-MVE-NEXT: vmovx.f16 s12, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s12, s12 ; CHECK-MVE-NEXT: vcmp.f16 s12, s12
; CHECK-MVE-NEXT: vmovx.f16 s12, s4 ; CHECK-MVE-NEXT: vmovx.f16 s12, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s0, s0 ; CHECK-MVE-NEXT: vcmp.f16 s0, s0
; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12 ; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s1 ; CHECK-MVE-NEXT: vmovx.f16 s4, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, s4 ; CHECK-MVE-NEXT: vcmp.f16 s4, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s9 ; CHECK-MVE-NEXT: vmovx.f16 s8, s9
; CHECK-MVE-NEXT: vcmp.f16 s1, s1 ; CHECK-MVE-NEXT: vcmp.f16 s1, s1
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s10 ; CHECK-MVE-NEXT: vmovx.f16 s8, s10
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: movvs r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s2 ; CHECK-MVE-NEXT: vmovx.f16 s4, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, s4 ; CHECK-MVE-NEXT: vcmp.f16 s4, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s2, s2 ; CHECK-MVE-NEXT: vcmp.f16 s2, s2
; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: mov.w r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s3 ; CHECK-MVE-NEXT: vmovx.f16 s4, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, s4 ; CHECK-MVE-NEXT: vcmp.f16 s4, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s3 ; CHECK-MVE-NEXT: vcmp.f16 s3, s3
; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_uno_v8f16: ; CHECK-MVEFP-LABEL: vcmp_uno_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vpt.f16 ge, q0, zr ; CHECK-MVEFP-NEXT: vpt.f16 ge, q0, zr
; CHECK-MVEFP-NEXT: vcmpt.f16 lt, q0, zr ; CHECK-MVEFP-NEXT: vcmpt.f16 lt, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp uno <8 x half> %src, zeroinitializer %c = fcmp uno <8 x half> %src, zeroinitializer
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
; Reversed ; Reversed
define arm_aapcs_vfpcc <4 x float> @vcmp_r_oeq_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_r_oeq_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_r_oeq_v4f32: ; CHECK-MVE-LABEL: vcmp_r_oeq_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, #0 ; CHECK-MVE-NEXT: vcmp.f32 s1, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, #0 ; CHECK-MVE-NEXT: vcmp.f32 s0, #0
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, #0 ; CHECK-MVE-NEXT: vcmp.f32 s3, #0
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r1, eq
; CHECK-MVE-NEXT: moveq r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, #0 ; CHECK-MVE-NEXT: vcmp.f32 s2, #0
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r2, eq
; CHECK-MVE-NEXT: moveq r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r3, eq
; CHECK-MVE-NEXT: moveq r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_oeq_v4f32: ; CHECK-MVEFP-LABEL: vcmp_r_oeq_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f32 eq, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f32 eq, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp oeq <4 x float> zeroinitializer, %src %c = fcmp oeq <4 x float> zeroinitializer, %src
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_r_one_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_r_one_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_r_one_v4f32: ; CHECK-MVE-LABEL: vcmp_r_one_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, #0 ; CHECK-MVE-NEXT: vcmp.f32 s1, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s1, #0 ; CHECK-MVE-NEXT: vcmp.f32 s1, #0
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, #0 ; CHECK-MVE-NEXT: vcmp.f32 s0, #0
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1 ; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r2, #1
; CHECK-MVE-NEXT: vcmp.f32 s3, #0 ; CHECK-MVE-NEXT: vcmp.f32 s3, #0
; CHECK-MVE-NEXT: mov.w r3, #0 ; CHECK-MVE-NEXT: cset r1, mi
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r2, #1 ; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r3, #1
; CHECK-MVE-NEXT: mov.w r0, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, #0 ; CHECK-MVE-NEXT: vcmp.f32 s2, #0
; CHECK-MVE-NEXT: cset r2, mi
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r3, #1 ; CHECK-MVE-NEXT: movgt r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r3, mi
; CHECK-MVE-NEXT: movmi r0, #1
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r0, #1 ; CHECK-MVE-NEXT: movgt r3, #1
; CHECK-MVE-NEXT: cmp r3, #0 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_one_v4f32: ; CHECK-MVEFP-LABEL: vcmp_r_one_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vpt.f32 le, q0, zr ; CHECK-MVEFP-NEXT: vpt.f32 le, q0, zr
; CHECK-MVEFP-NEXT: vcmpt.f32 ge, q0, zr ; CHECK-MVEFP-NEXT: vcmpt.f32 ge, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q2, q1 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q1
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp one <4 x float> zeroinitializer, %src %c = fcmp one <4 x float> zeroinitializer, %src
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_r_ogt_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_r_ogt_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_r_ogt_v4f32: ; CHECK-MVE-LABEL: vcmp_r_ogt_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, #0 ; CHECK-MVE-NEXT: vcmp.f32 s1, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, #0 ; CHECK-MVE-NEXT: vcmp.f32 s0, #0
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, #0 ; CHECK-MVE-NEXT: vcmp.f32 s3, #0
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r1, mi
; CHECK-MVE-NEXT: movmi r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, #0 ; CHECK-MVE-NEXT: vcmp.f32 s2, #0
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r2, mi
; CHECK-MVE-NEXT: movmi r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r3, mi
; CHECK-MVE-NEXT: movmi r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_ogt_v4f32: ; CHECK-MVEFP-LABEL: vcmp_r_ogt_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f32 lt, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f32 lt, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ogt <4 x float> zeroinitializer, %src %c = fcmp ogt <4 x float> zeroinitializer, %src
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_r_oge_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_r_oge_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_r_oge_v4f32: ; CHECK-MVE-LABEL: vcmp_r_oge_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, #0 ; CHECK-MVE-NEXT: vcmp.f32 s1, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, #0 ; CHECK-MVE-NEXT: vcmp.f32 s0, #0
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, #0 ; CHECK-MVE-NEXT: vcmp.f32 s3, #0
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r1, ls
; CHECK-MVE-NEXT: movls r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, #0 ; CHECK-MVE-NEXT: vcmp.f32 s2, #0
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r2, ls
; CHECK-MVE-NEXT: movls r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r3, ls
; CHECK-MVE-NEXT: movls r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_oge_v4f32: ; CHECK-MVEFP-LABEL: vcmp_r_oge_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f32 le, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f32 le, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp oge <4 x float> zeroinitializer, %src %c = fcmp oge <4 x float> zeroinitializer, %src
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_r_olt_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_r_olt_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_r_olt_v4f32: ; CHECK-MVE-LABEL: vcmp_r_olt_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, #0 ; CHECK-MVE-NEXT: vcmp.f32 s1, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, #0 ; CHECK-MVE-NEXT: vcmp.f32 s0, #0
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, #0 ; CHECK-MVE-NEXT: vcmp.f32 s3, #0
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r1, gt
; CHECK-MVE-NEXT: movgt r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, #0 ; CHECK-MVE-NEXT: vcmp.f32 s2, #0
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r2, gt
; CHECK-MVE-NEXT: movgt r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r3, gt
; CHECK-MVE-NEXT: movgt r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_olt_v4f32: ; CHECK-MVEFP-LABEL: vcmp_r_olt_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f32 gt, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f32 gt, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp olt <4 x float> zeroinitializer, %src %c = fcmp olt <4 x float> zeroinitializer, %src
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_r_ole_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_r_ole_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_r_ole_v4f32: ; CHECK-MVE-LABEL: vcmp_r_ole_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, #0 ; CHECK-MVE-NEXT: vcmp.f32 s1, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, #0 ; CHECK-MVE-NEXT: vcmp.f32 s0, #0
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, #0 ; CHECK-MVE-NEXT: vcmp.f32 s3, #0
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r1, ge
; CHECK-MVE-NEXT: movge r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, #0 ; CHECK-MVE-NEXT: vcmp.f32 s2, #0
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r2, ge
; CHECK-MVE-NEXT: movge r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r3, ge
; CHECK-MVE-NEXT: movge r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_ole_v4f32: ; CHECK-MVEFP-LABEL: vcmp_r_ole_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f32 ge, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f32 ge, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ole <4 x float> zeroinitializer, %src %c = fcmp ole <4 x float> zeroinitializer, %src
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_r_ueq_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_r_ueq_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_r_ueq_v4f32: ; CHECK-MVE-LABEL: vcmp_r_ueq_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, #0 ; CHECK-MVE-NEXT: vcmp.f32 s1, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s1, #0 ; CHECK-MVE-NEXT: vcmp.f32 s1, #0
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, #0 ; CHECK-MVE-NEXT: vcmp.f32 s0, #0
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1 ; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r2, #1
; CHECK-MVE-NEXT: vcmp.f32 s3, #0 ; CHECK-MVE-NEXT: vcmp.f32 s3, #0
; CHECK-MVE-NEXT: mov.w r3, #0 ; CHECK-MVE-NEXT: cset r1, eq
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r2, #1 ; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r3, #1
; CHECK-MVE-NEXT: mov.w r0, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, #0 ; CHECK-MVE-NEXT: vcmp.f32 s2, #0
; CHECK-MVE-NEXT: cset r2, eq
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r3, #1 ; CHECK-MVE-NEXT: movvs r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r3, eq
; CHECK-MVE-NEXT: moveq r0, #1
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r0, #1 ; CHECK-MVE-NEXT: movvs r3, #1
; CHECK-MVE-NEXT: cmp r3, #0 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_ueq_v4f32: ; CHECK-MVEFP-LABEL: vcmp_r_ueq_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vpt.f32 le, q0, zr ; CHECK-MVEFP-NEXT: vpt.f32 le, q0, zr
; CHECK-MVEFP-NEXT: vcmpt.f32 ge, q0, zr ; CHECK-MVEFP-NEXT: vcmpt.f32 ge, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ueq <4 x float> zeroinitializer, %src %c = fcmp ueq <4 x float> zeroinitializer, %src
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_r_une_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_r_une_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_r_une_v4f32: ; CHECK-MVE-LABEL: vcmp_r_une_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, #0 ; CHECK-MVE-NEXT: vcmp.f32 s1, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, #0 ; CHECK-MVE-NEXT: vcmp.f32 s0, #0
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, #0 ; CHECK-MVE-NEXT: vcmp.f32 s3, #0
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r1, ne
; CHECK-MVE-NEXT: movne r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, #0 ; CHECK-MVE-NEXT: vcmp.f32 s2, #0
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r2, ne
; CHECK-MVE-NEXT: movne r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r3, ne
; CHECK-MVE-NEXT: movne r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_une_v4f32: ; CHECK-MVEFP-LABEL: vcmp_r_une_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f32 ne, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f32 ne, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp une <4 x float> zeroinitializer, %src %c = fcmp une <4 x float> zeroinitializer, %src
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_r_ugt_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_r_ugt_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_r_ugt_v4f32: ; CHECK-MVE-LABEL: vcmp_r_ugt_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, #0 ; CHECK-MVE-NEXT: vcmp.f32 s1, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, #0 ; CHECK-MVE-NEXT: vcmp.f32 s0, #0
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, #0 ; CHECK-MVE-NEXT: vcmp.f32 s3, #0
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r1, lt
; CHECK-MVE-NEXT: movlt r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, #0 ; CHECK-MVE-NEXT: vcmp.f32 s2, #0
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r2, lt
; CHECK-MVE-NEXT: movlt r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r3, lt
; CHECK-MVE-NEXT: movlt r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_ugt_v4f32: ; CHECK-MVEFP-LABEL: vcmp_r_ugt_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f32 lt, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f32 lt, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ugt <4 x float> zeroinitializer, %src %c = fcmp ugt <4 x float> zeroinitializer, %src
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_r_uge_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_r_uge_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_r_uge_v4f32: ; CHECK-MVE-LABEL: vcmp_r_uge_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, #0 ; CHECK-MVE-NEXT: vcmp.f32 s1, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, #0 ; CHECK-MVE-NEXT: vcmp.f32 s0, #0
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, #0 ; CHECK-MVE-NEXT: vcmp.f32 s3, #0
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r1, le
; CHECK-MVE-NEXT: movle r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, #0 ; CHECK-MVE-NEXT: vcmp.f32 s2, #0
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r2, le
; CHECK-MVE-NEXT: movle r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r3, le
; CHECK-MVE-NEXT: movle r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_uge_v4f32: ; CHECK-MVEFP-LABEL: vcmp_r_uge_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f32 le, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f32 le, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp uge <4 x float> zeroinitializer, %src %c = fcmp uge <4 x float> zeroinitializer, %src
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_r_ult_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_r_ult_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_r_ult_v4f32: ; CHECK-MVE-LABEL: vcmp_r_ult_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, #0 ; CHECK-MVE-NEXT: vcmp.f32 s1, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, #0 ; CHECK-MVE-NEXT: vcmp.f32 s0, #0
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, #0 ; CHECK-MVE-NEXT: vcmp.f32 s3, #0
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r1, hi
; CHECK-MVE-NEXT: movhi r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, #0 ; CHECK-MVE-NEXT: vcmp.f32 s2, #0
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r2, hi
; CHECK-MVE-NEXT: movhi r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r3, hi
; CHECK-MVE-NEXT: movhi r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_ult_v4f32: ; CHECK-MVEFP-LABEL: vcmp_r_ult_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f32 gt, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f32 gt, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ult <4 x float> zeroinitializer, %src %c = fcmp ult <4 x float> zeroinitializer, %src
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_r_ule_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_r_ule_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_r_ule_v4f32: ; CHECK-MVE-LABEL: vcmp_r_ule_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, #0 ; CHECK-MVE-NEXT: vcmp.f32 s1, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, #0 ; CHECK-MVE-NEXT: vcmp.f32 s0, #0
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, #0 ; CHECK-MVE-NEXT: vcmp.f32 s3, #0
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r1, pl
; CHECK-MVE-NEXT: movpl r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, #0 ; CHECK-MVE-NEXT: vcmp.f32 s2, #0
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r2, pl
; CHECK-MVE-NEXT: movpl r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r3, pl
; CHECK-MVE-NEXT: movpl r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_ule_v4f32: ; CHECK-MVEFP-LABEL: vcmp_r_ule_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f32 ge, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f32 ge, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ule <4 x float> zeroinitializer, %src %c = fcmp ule <4 x float> zeroinitializer, %src
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_r_ord_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_r_ord_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_r_ord_v4f32: ; CHECK-MVE-LABEL: vcmp_r_ord_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s1 ; CHECK-MVE-NEXT: vcmp.f32 s1, s1
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s0 ; CHECK-MVE-NEXT: vcmp.f32 s0, s0
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, s3 ; CHECK-MVE-NEXT: vcmp.f32 s3, s3
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r1, vc
; CHECK-MVE-NEXT: movvc r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s2 ; CHECK-MVE-NEXT: vcmp.f32 s2, s2
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r2, vc
; CHECK-MVE-NEXT: movvc r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r3, vc
; CHECK-MVE-NEXT: movvc r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_ord_v4f32: ; CHECK-MVEFP-LABEL: vcmp_r_ord_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vpt.f32 le, q0, zr ; CHECK-MVEFP-NEXT: vpt.f32 le, q0, zr
; CHECK-MVEFP-NEXT: vcmpt.f32 gt, q0, zr ; CHECK-MVEFP-NEXT: vcmpt.f32 gt, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q2, q1 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q1
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ord <4 x float> zeroinitializer, %src %c = fcmp ord <4 x float> zeroinitializer, %src
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <4 x float> @vcmp_r_uno_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) { define arm_aapcs_vfpcc <4 x float> @vcmp_r_uno_v4f32(<4 x float> %src, <4 x float> %a, <4 x float> %b) {
; CHECK-MVE-LABEL: vcmp_r_uno_v4f32: ; CHECK-MVE-LABEL: vcmp_r_uno_v4f32:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vcmp.f32 s1, s1 ; CHECK-MVE-NEXT: vcmp.f32 s1, s1
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vcmp.f32 s0, s0 ; CHECK-MVE-NEXT: vcmp.f32 s0, s0
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r2, #0
; CHECK-MVE-NEXT: vcmp.f32 s3, s3 ; CHECK-MVE-NEXT: vcmp.f32 s3, s3
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r1, vs
; CHECK-MVE-NEXT: movvs r2, #1
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r3, #0
; CHECK-MVE-NEXT: vcmp.f32 s2, s2 ; CHECK-MVE-NEXT: vcmp.f32 s2, s2
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r2, vs
; CHECK-MVE-NEXT: movvs r3, #1
; CHECK-MVE-NEXT: movs r0, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r3, vs
; CHECK-MVE-NEXT: movvs r0, #1 ; CHECK-MVE-NEXT: cmp r2, #0
; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f32 s3, s11, s7
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r3, #0
; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f32 s2, s10, s6
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f32 s1, s9, s5
; CHECK-MVE-NEXT: cmp r2, #0 ; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f32 s0, s8, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_uno_v4f32: ; CHECK-MVEFP-LABEL: vcmp_r_uno_v4f32:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vpt.f32 le, q0, zr ; CHECK-MVEFP-NEXT: vpt.f32 le, q0, zr
; CHECK-MVEFP-NEXT: vcmpt.f32 gt, q0, zr ; CHECK-MVEFP-NEXT: vcmpt.f32 gt, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp uno <4 x float> zeroinitializer, %src %c = fcmp uno <4 x float> zeroinitializer, %src
%s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b %s = select <4 x i1> %c, <4 x float> %a, <4 x float> %b
ret <4 x float> %s ret <4 x float> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_r_oeq_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_r_oeq_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_r_oeq_v8f16: ; CHECK-MVE-LABEL: vcmp_r_oeq_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s12, s0 ; CHECK-MVE-NEXT: vmovx.f16 s12, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s12, #0 ; CHECK-MVE-NEXT: vcmp.f16 s12, #0
; CHECK-MVE-NEXT: vmovx.f16 s12, s4 ; CHECK-MVE-NEXT: vmovx.f16 s12, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s0, #0 ; CHECK-MVE-NEXT: vcmp.f16 s0, #0
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12 ; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s1 ; CHECK-MVE-NEXT: vmovx.f16 s4, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s9 ; CHECK-MVE-NEXT: vmovx.f16 s8, s9
; CHECK-MVE-NEXT: vcmp.f16 s1, #0 ; CHECK-MVE-NEXT: vcmp.f16 s1, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s10 ; CHECK-MVE-NEXT: vmovx.f16 s8, s10
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s2 ; CHECK-MVE-NEXT: vmovx.f16 s4, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s2, #0 ; CHECK-MVE-NEXT: vcmp.f16 s2, #0
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: mov.w r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s3 ; CHECK-MVE-NEXT: vmovx.f16 s4, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, #0 ; CHECK-MVE-NEXT: vcmp.f16 s3, #0
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_oeq_v8f16: ; CHECK-MVEFP-LABEL: vcmp_r_oeq_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f16 eq, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f16 eq, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp oeq <8 x half> zeroinitializer, %src %c = fcmp oeq <8 x half> zeroinitializer, %src
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_r_one_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_r_one_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_r_one_v8f16: ; CHECK-MVE-LABEL: vcmp_r_one_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s12, s0 ; CHECK-MVE-NEXT: vmovx.f16 s12, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s12, #0 ; CHECK-MVE-NEXT: vcmp.f16 s12, #0
; CHECK-MVE-NEXT: vmovx.f16 s12, s4 ; CHECK-MVE-NEXT: vmovx.f16 s12, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s0, #0 ; CHECK-MVE-NEXT: vcmp.f16 s0, #0
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12 ; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1 ; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s1 ; CHECK-MVE-NEXT: vmovx.f16 s4, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s9 ; CHECK-MVE-NEXT: vmovx.f16 s8, s9
; CHECK-MVE-NEXT: vcmp.f16 s1, #0 ; CHECK-MVE-NEXT: vcmp.f16 s1, #0
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s8, s10 ; CHECK-MVE-NEXT: vmovx.f16 s8, s10
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1 ; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s2 ; CHECK-MVE-NEXT: vmovx.f16 s4, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s2, #0 ; CHECK-MVE-NEXT: vcmp.f16 s2, #0
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: mov.w r0, #0
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1 ; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s3 ; CHECK-MVE-NEXT: vmovx.f16 s4, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, #0 ; CHECK-MVE-NEXT: vcmp.f16 s3, #0
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r0, #1
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r0, #1 ; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_one_v8f16: ; CHECK-MVEFP-LABEL: vcmp_r_one_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vpt.f16 le, q0, zr ; CHECK-MVEFP-NEXT: vpt.f16 le, q0, zr
; CHECK-MVEFP-NEXT: vcmpt.f16 ge, q0, zr ; CHECK-MVEFP-NEXT: vcmpt.f16 ge, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q2, q1 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q1
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp one <8 x half> zeroinitializer, %src %c = fcmp one <8 x half> zeroinitializer, %src
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_r_ogt_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_r_ogt_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_r_ogt_v8f16: ; CHECK-MVE-LABEL: vcmp_r_ogt_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s12, s0 ; CHECK-MVE-NEXT: vmovx.f16 s12, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s12, #0 ; CHECK-MVE-NEXT: vcmp.f16 s12, #0
; CHECK-MVE-NEXT: vmovx.f16 s12, s4 ; CHECK-MVE-NEXT: vmovx.f16 s12, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s0, #0 ; CHECK-MVE-NEXT: vcmp.f16 s0, #0
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12 ; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s1 ; CHECK-MVE-NEXT: vmovx.f16 s4, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s9 ; CHECK-MVE-NEXT: vmovx.f16 s8, s9
; CHECK-MVE-NEXT: vcmp.f16 s1, #0 ; CHECK-MVE-NEXT: vcmp.f16 s1, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s10 ; CHECK-MVE-NEXT: vmovx.f16 s8, s10
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s2 ; CHECK-MVE-NEXT: vmovx.f16 s4, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s2, #0 ; CHECK-MVE-NEXT: vcmp.f16 s2, #0
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: mov.w r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s3 ; CHECK-MVE-NEXT: vmovx.f16 s4, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi
; CHECK-MVE-NEXT: movmi r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, #0 ; CHECK-MVE-NEXT: vcmp.f16 s3, #0
; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it mi ; CHECK-MVE-NEXT: cset r0, mi
; CHECK-MVE-NEXT: movmi r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_ogt_v8f16: ; CHECK-MVEFP-LABEL: vcmp_r_ogt_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f16 lt, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f16 lt, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ogt <8 x half> zeroinitializer, %src %c = fcmp ogt <8 x half> zeroinitializer, %src
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_r_oge_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_r_oge_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_r_oge_v8f16: ; CHECK-MVE-LABEL: vcmp_r_oge_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s12, s0 ; CHECK-MVE-NEXT: vmovx.f16 s12, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s12, #0 ; CHECK-MVE-NEXT: vcmp.f16 s12, #0
; CHECK-MVE-NEXT: vmovx.f16 s12, s4 ; CHECK-MVE-NEXT: vmovx.f16 s12, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ls
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s0, #0 ; CHECK-MVE-NEXT: vcmp.f16 s0, #0
; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12 ; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s1 ; CHECK-MVE-NEXT: vmovx.f16 s4, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ls
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s9 ; CHECK-MVE-NEXT: vmovx.f16 s8, s9
; CHECK-MVE-NEXT: vcmp.f16 s1, #0 ; CHECK-MVE-NEXT: vcmp.f16 s1, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s10 ; CHECK-MVE-NEXT: vmovx.f16 s8, s10
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: movls r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s2 ; CHECK-MVE-NEXT: vmovx.f16 s4, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ls
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s2, #0 ; CHECK-MVE-NEXT: vcmp.f16 s2, #0
; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: mov.w r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it ls
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s3 ; CHECK-MVE-NEXT: vmovx.f16 s4, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ls
; CHECK-MVE-NEXT: movls r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, #0 ; CHECK-MVE-NEXT: vcmp.f16 s3, #0
; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ls ; CHECK-MVE-NEXT: cset r0, ls
; CHECK-MVE-NEXT: movls r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_oge_v8f16: ; CHECK-MVEFP-LABEL: vcmp_r_oge_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f16 le, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f16 le, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp oge <8 x half> zeroinitializer, %src %c = fcmp oge <8 x half> zeroinitializer, %src
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_r_olt_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_r_olt_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_r_olt_v8f16: ; CHECK-MVE-LABEL: vcmp_r_olt_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s12, s0 ; CHECK-MVE-NEXT: vmovx.f16 s12, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s12, #0 ; CHECK-MVE-NEXT: vcmp.f16 s12, #0
; CHECK-MVE-NEXT: vmovx.f16 s12, s4 ; CHECK-MVE-NEXT: vmovx.f16 s12, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s0, #0 ; CHECK-MVE-NEXT: vcmp.f16 s0, #0
; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12 ; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s1 ; CHECK-MVE-NEXT: vmovx.f16 s4, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s9 ; CHECK-MVE-NEXT: vmovx.f16 s8, s9
; CHECK-MVE-NEXT: vcmp.f16 s1, #0 ; CHECK-MVE-NEXT: vcmp.f16 s1, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s10 ; CHECK-MVE-NEXT: vmovx.f16 s8, s10
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: movgt r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s2 ; CHECK-MVE-NEXT: vmovx.f16 s4, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s2, #0 ; CHECK-MVE-NEXT: vcmp.f16 s2, #0
; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: mov.w r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s3 ; CHECK-MVE-NEXT: vmovx.f16 s4, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it gt
; CHECK-MVE-NEXT: movgt r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, #0 ; CHECK-MVE-NEXT: vcmp.f16 s3, #0
; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it gt ; CHECK-MVE-NEXT: cset r0, gt
; CHECK-MVE-NEXT: movgt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_olt_v8f16: ; CHECK-MVEFP-LABEL: vcmp_r_olt_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f16 gt, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f16 gt, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp olt <8 x half> zeroinitializer, %src %c = fcmp olt <8 x half> zeroinitializer, %src
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_r_ole_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_r_ole_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_r_ole_v8f16: ; CHECK-MVE-LABEL: vcmp_r_ole_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s12, s0 ; CHECK-MVE-NEXT: vmovx.f16 s12, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s12, #0 ; CHECK-MVE-NEXT: vcmp.f16 s12, #0
; CHECK-MVE-NEXT: vmovx.f16 s12, s4 ; CHECK-MVE-NEXT: vmovx.f16 s12, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ge
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s0, #0 ; CHECK-MVE-NEXT: vcmp.f16 s0, #0
; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12 ; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s1 ; CHECK-MVE-NEXT: vmovx.f16 s4, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ge
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s9 ; CHECK-MVE-NEXT: vmovx.f16 s8, s9
; CHECK-MVE-NEXT: vcmp.f16 s1, #0 ; CHECK-MVE-NEXT: vcmp.f16 s1, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s10 ; CHECK-MVE-NEXT: vmovx.f16 s8, s10
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: movge r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s2 ; CHECK-MVE-NEXT: vmovx.f16 s4, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ge
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s2, #0 ; CHECK-MVE-NEXT: vcmp.f16 s2, #0
; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: mov.w r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it ge
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s3 ; CHECK-MVE-NEXT: vmovx.f16 s4, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ge
; CHECK-MVE-NEXT: movge r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, #0 ; CHECK-MVE-NEXT: vcmp.f16 s3, #0
; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ge ; CHECK-MVE-NEXT: cset r0, ge
; CHECK-MVE-NEXT: movge r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_ole_v8f16: ; CHECK-MVEFP-LABEL: vcmp_r_ole_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f16 ge, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f16 ge, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ole <8 x half> zeroinitializer, %src %c = fcmp ole <8 x half> zeroinitializer, %src
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_r_ueq_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_r_ueq_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_r_ueq_v8f16: ; CHECK-MVE-LABEL: vcmp_r_ueq_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s12, s0 ; CHECK-MVE-NEXT: vmovx.f16 s12, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s12, #0 ; CHECK-MVE-NEXT: vcmp.f16 s12, #0
; CHECK-MVE-NEXT: vmovx.f16 s12, s4 ; CHECK-MVE-NEXT: vmovx.f16 s12, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s0, #0 ; CHECK-MVE-NEXT: vcmp.f16 s0, #0
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12 ; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1 ; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s1 ; CHECK-MVE-NEXT: vmovx.f16 s4, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s9 ; CHECK-MVE-NEXT: vmovx.f16 s8, s9
; CHECK-MVE-NEXT: vcmp.f16 s1, #0 ; CHECK-MVE-NEXT: vcmp.f16 s1, #0
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s8, s10 ; CHECK-MVE-NEXT: vmovx.f16 s8, s10
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1 ; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s2 ; CHECK-MVE-NEXT: vmovx.f16 s4, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s2, #0 ; CHECK-MVE-NEXT: vcmp.f16 s2, #0
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: mov.w r0, #0
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1 ; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r1, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s3 ; CHECK-MVE-NEXT: vmovx.f16 s4, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq
; CHECK-MVE-NEXT: moveq r1, #1
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, #0 ; CHECK-MVE-NEXT: vcmp.f16 s3, #0
; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it eq ; CHECK-MVE-NEXT: cset r0, eq
; CHECK-MVE-NEXT: moveq r0, #1
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r0, #1 ; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_ueq_v8f16: ; CHECK-MVEFP-LABEL: vcmp_r_ueq_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vpt.f16 le, q0, zr ; CHECK-MVEFP-NEXT: vpt.f16 le, q0, zr
; CHECK-MVEFP-NEXT: vcmpt.f16 ge, q0, zr ; CHECK-MVEFP-NEXT: vcmpt.f16 ge, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ueq <8 x half> zeroinitializer, %src %c = fcmp ueq <8 x half> zeroinitializer, %src
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_r_une_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_r_une_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_r_une_v8f16: ; CHECK-MVE-LABEL: vcmp_r_une_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s12, s0 ; CHECK-MVE-NEXT: vmovx.f16 s12, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s12, #0 ; CHECK-MVE-NEXT: vcmp.f16 s12, #0
; CHECK-MVE-NEXT: vmovx.f16 s12, s4 ; CHECK-MVE-NEXT: vmovx.f16 s12, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ne
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s0, #0 ; CHECK-MVE-NEXT: vcmp.f16 s0, #0
; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12 ; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s1 ; CHECK-MVE-NEXT: vmovx.f16 s4, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ne
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s9 ; CHECK-MVE-NEXT: vmovx.f16 s8, s9
; CHECK-MVE-NEXT: vcmp.f16 s1, #0 ; CHECK-MVE-NEXT: vcmp.f16 s1, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s10 ; CHECK-MVE-NEXT: vmovx.f16 s8, s10
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: movne r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s2 ; CHECK-MVE-NEXT: vmovx.f16 s4, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ne
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s2, #0 ; CHECK-MVE-NEXT: vcmp.f16 s2, #0
; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: mov.w r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it ne
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s3 ; CHECK-MVE-NEXT: vmovx.f16 s4, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ne
; CHECK-MVE-NEXT: movne r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, #0 ; CHECK-MVE-NEXT: vcmp.f16 s3, #0
; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it ne ; CHECK-MVE-NEXT: cset r0, ne
; CHECK-MVE-NEXT: movne r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_une_v8f16: ; CHECK-MVEFP-LABEL: vcmp_r_une_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f16 ne, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f16 ne, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp une <8 x half> zeroinitializer, %src %c = fcmp une <8 x half> zeroinitializer, %src
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_r_ugt_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_r_ugt_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_r_ugt_v8f16: ; CHECK-MVE-LABEL: vcmp_r_ugt_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s12, s0 ; CHECK-MVE-NEXT: vmovx.f16 s12, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s12, #0 ; CHECK-MVE-NEXT: vcmp.f16 s12, #0
; CHECK-MVE-NEXT: vmovx.f16 s12, s4 ; CHECK-MVE-NEXT: vmovx.f16 s12, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it lt
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s0, #0 ; CHECK-MVE-NEXT: vcmp.f16 s0, #0
; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12 ; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s1 ; CHECK-MVE-NEXT: vmovx.f16 s4, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it lt
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s9 ; CHECK-MVE-NEXT: vmovx.f16 s8, s9
; CHECK-MVE-NEXT: vcmp.f16 s1, #0 ; CHECK-MVE-NEXT: vcmp.f16 s1, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s10 ; CHECK-MVE-NEXT: vmovx.f16 s8, s10
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: movlt r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s2 ; CHECK-MVE-NEXT: vmovx.f16 s4, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it lt
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s2, #0 ; CHECK-MVE-NEXT: vcmp.f16 s2, #0
; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: mov.w r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it lt
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s3 ; CHECK-MVE-NEXT: vmovx.f16 s4, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it lt
; CHECK-MVE-NEXT: movlt r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, #0 ; CHECK-MVE-NEXT: vcmp.f16 s3, #0
; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it lt ; CHECK-MVE-NEXT: cset r0, lt
; CHECK-MVE-NEXT: movlt r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_ugt_v8f16: ; CHECK-MVEFP-LABEL: vcmp_r_ugt_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f16 lt, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f16 lt, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ugt <8 x half> zeroinitializer, %src %c = fcmp ugt <8 x half> zeroinitializer, %src
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_r_uge_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_r_uge_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_r_uge_v8f16: ; CHECK-MVE-LABEL: vcmp_r_uge_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s12, s0 ; CHECK-MVE-NEXT: vmovx.f16 s12, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s12, #0 ; CHECK-MVE-NEXT: vcmp.f16 s12, #0
; CHECK-MVE-NEXT: vmovx.f16 s12, s4 ; CHECK-MVE-NEXT: vmovx.f16 s12, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it le
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s0, #0 ; CHECK-MVE-NEXT: vcmp.f16 s0, #0
; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12 ; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s1 ; CHECK-MVE-NEXT: vmovx.f16 s4, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it le
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s9 ; CHECK-MVE-NEXT: vmovx.f16 s8, s9
; CHECK-MVE-NEXT: vcmp.f16 s1, #0 ; CHECK-MVE-NEXT: vcmp.f16 s1, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s10 ; CHECK-MVE-NEXT: vmovx.f16 s8, s10
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: movle r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s2 ; CHECK-MVE-NEXT: vmovx.f16 s4, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it le
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s2, #0 ; CHECK-MVE-NEXT: vcmp.f16 s2, #0
; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: mov.w r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it le
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s3 ; CHECK-MVE-NEXT: vmovx.f16 s4, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it le
; CHECK-MVE-NEXT: movle r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, #0 ; CHECK-MVE-NEXT: vcmp.f16 s3, #0
; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it le ; CHECK-MVE-NEXT: cset r0, le
; CHECK-MVE-NEXT: movle r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_uge_v8f16: ; CHECK-MVEFP-LABEL: vcmp_r_uge_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f16 le, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f16 le, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp uge <8 x half> zeroinitializer, %src %c = fcmp uge <8 x half> zeroinitializer, %src
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_r_ult_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_r_ult_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_r_ult_v8f16: ; CHECK-MVE-LABEL: vcmp_r_ult_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s12, s0 ; CHECK-MVE-NEXT: vmovx.f16 s12, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s12, #0 ; CHECK-MVE-NEXT: vcmp.f16 s12, #0
; CHECK-MVE-NEXT: vmovx.f16 s12, s4 ; CHECK-MVE-NEXT: vmovx.f16 s12, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it hi
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s0, #0 ; CHECK-MVE-NEXT: vcmp.f16 s0, #0
; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12 ; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s1 ; CHECK-MVE-NEXT: vmovx.f16 s4, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it hi
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s9 ; CHECK-MVE-NEXT: vmovx.f16 s8, s9
; CHECK-MVE-NEXT: vcmp.f16 s1, #0 ; CHECK-MVE-NEXT: vcmp.f16 s1, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s10 ; CHECK-MVE-NEXT: vmovx.f16 s8, s10
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: movhi r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s2 ; CHECK-MVE-NEXT: vmovx.f16 s4, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it hi
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s2, #0 ; CHECK-MVE-NEXT: vcmp.f16 s2, #0
; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: mov.w r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it hi
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s3 ; CHECK-MVE-NEXT: vmovx.f16 s4, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it hi
; CHECK-MVE-NEXT: movhi r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, #0 ; CHECK-MVE-NEXT: vcmp.f16 s3, #0
; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it hi ; CHECK-MVE-NEXT: cset r0, hi
; CHECK-MVE-NEXT: movhi r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_ult_v8f16: ; CHECK-MVEFP-LABEL: vcmp_r_ult_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f16 gt, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f16 gt, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ult <8 x half> zeroinitializer, %src %c = fcmp ult <8 x half> zeroinitializer, %src
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_r_ule_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_r_ule_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_r_ule_v8f16: ; CHECK-MVE-LABEL: vcmp_r_ule_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s12, s0 ; CHECK-MVE-NEXT: vmovx.f16 s12, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s12, #0 ; CHECK-MVE-NEXT: vcmp.f16 s12, #0
; CHECK-MVE-NEXT: vmovx.f16 s12, s4 ; CHECK-MVE-NEXT: vmovx.f16 s12, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it pl
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s0, #0 ; CHECK-MVE-NEXT: vcmp.f16 s0, #0
; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12 ; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s1 ; CHECK-MVE-NEXT: vmovx.f16 s4, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it pl
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s9 ; CHECK-MVE-NEXT: vmovx.f16 s8, s9
; CHECK-MVE-NEXT: vcmp.f16 s1, #0 ; CHECK-MVE-NEXT: vcmp.f16 s1, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s10 ; CHECK-MVE-NEXT: vmovx.f16 s8, s10
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: movpl r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s2 ; CHECK-MVE-NEXT: vmovx.f16 s4, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it pl
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s2, #0 ; CHECK-MVE-NEXT: vcmp.f16 s2, #0
; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: mov.w r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it pl
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s3 ; CHECK-MVE-NEXT: vmovx.f16 s4, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, #0 ; CHECK-MVE-NEXT: vcmp.f16 s4, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it pl
; CHECK-MVE-NEXT: movpl r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, #0 ; CHECK-MVE-NEXT: vcmp.f16 s3, #0
; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it pl ; CHECK-MVE-NEXT: cset r0, pl
; CHECK-MVE-NEXT: movpl r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_ule_v8f16: ; CHECK-MVEFP-LABEL: vcmp_r_ule_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vcmp.f16 ge, q0, zr ; CHECK-MVEFP-NEXT: vcmp.f16 ge, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ule <8 x half> zeroinitializer, %src %c = fcmp ule <8 x half> zeroinitializer, %src
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_r_ord_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_r_ord_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_r_ord_v8f16: ; CHECK-MVE-LABEL: vcmp_r_ord_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s12, s0 ; CHECK-MVE-NEXT: vmovx.f16 s12, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s12, s12 ; CHECK-MVE-NEXT: vcmp.f16 s12, s12
; CHECK-MVE-NEXT: vmovx.f16 s12, s4 ; CHECK-MVE-NEXT: vmovx.f16 s12, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vc
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s0, s0 ; CHECK-MVE-NEXT: vcmp.f16 s0, s0
; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12 ; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s1 ; CHECK-MVE-NEXT: vmovx.f16 s4, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, s4 ; CHECK-MVE-NEXT: vcmp.f16 s4, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vc
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s9 ; CHECK-MVE-NEXT: vmovx.f16 s8, s9
; CHECK-MVE-NEXT: vcmp.f16 s1, s1 ; CHECK-MVE-NEXT: vcmp.f16 s1, s1
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s10 ; CHECK-MVE-NEXT: vmovx.f16 s8, s10
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: movvc r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s2 ; CHECK-MVE-NEXT: vmovx.f16 s4, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, s4 ; CHECK-MVE-NEXT: vcmp.f16 s4, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vc
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s2, s2 ; CHECK-MVE-NEXT: vcmp.f16 s2, s2
; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: mov.w r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it vc
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s3 ; CHECK-MVE-NEXT: vmovx.f16 s4, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, s4 ; CHECK-MVE-NEXT: vcmp.f16 s4, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vc
; CHECK-MVE-NEXT: movvc r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s3 ; CHECK-MVE-NEXT: vcmp.f16 s3, s3
; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vc ; CHECK-MVE-NEXT: cset r0, vc
; CHECK-MVE-NEXT: movvc r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_ord_v8f16: ; CHECK-MVEFP-LABEL: vcmp_r_ord_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vpt.f16 le, q0, zr ; CHECK-MVEFP-NEXT: vpt.f16 le, q0, zr
; CHECK-MVEFP-NEXT: vcmpt.f16 gt, q0, zr ; CHECK-MVEFP-NEXT: vcmpt.f16 gt, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q2, q1 ; CHECK-MVEFP-NEXT: vpsel q0, q2, q1
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp ord <8 x half> zeroinitializer, %src %c = fcmp ord <8 x half> zeroinitializer, %src
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }
define arm_aapcs_vfpcc <8 x half> @vcmp_r_uno_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) { define arm_aapcs_vfpcc <8 x half> @vcmp_r_uno_v8f16(<8 x half> %src, <8 x half> %a, <8 x half> %b) {
; CHECK-MVE-LABEL: vcmp_r_uno_v8f16: ; CHECK-MVE-LABEL: vcmp_r_uno_v8f16:
; CHECK-MVE: @ %bb.0: @ %entry ; CHECK-MVE: @ %bb.0: @ %entry
; CHECK-MVE-NEXT: vmovx.f16 s12, s0 ; CHECK-MVE-NEXT: vmovx.f16 s12, s0
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s12, s12 ; CHECK-MVE-NEXT: vcmp.f16 s12, s12
; CHECK-MVE-NEXT: vmovx.f16 s12, s4 ; CHECK-MVE-NEXT: vmovx.f16 s12, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s14, s8
; CHECK-MVE-NEXT: vcmp.f16 s0, s0 ; CHECK-MVE-NEXT: vcmp.f16 s0, s0
; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12 ; CHECK-MVE-NEXT: vseleq.f16 s12, s14, s12
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s0, s8, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s1 ; CHECK-MVE-NEXT: vmovx.f16 s4, s1
; CHECK-MVE-NEXT: vcmp.f16 s4, s4 ; CHECK-MVE-NEXT: vcmp.f16 s4, s4
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s4, s5 ; CHECK-MVE-NEXT: vmovx.f16 s4, s5
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: vmovx.f16 s8, s9 ; CHECK-MVE-NEXT: vmovx.f16 s8, s9
; CHECK-MVE-NEXT: vcmp.f16 s1, s1 ; CHECK-MVE-NEXT: vcmp.f16 s1, s1
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0
; CHECK-MVE-NEXT: vmovx.f16 s8, s10 ; CHECK-MVE-NEXT: vmovx.f16 s8, s10
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: movvs r1, #1 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5 ; CHECK-MVE-NEXT: vseleq.f16 s1, s9, s5
; CHECK-MVE-NEXT: movs r1, #0
; CHECK-MVE-NEXT: vins.f16 s1, s4 ; CHECK-MVE-NEXT: vins.f16 s1, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s2 ; CHECK-MVE-NEXT: vmovx.f16 s4, s2
; CHECK-MVE-NEXT: vcmp.f16 s4, s4 ; CHECK-MVE-NEXT: vcmp.f16 s4, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s6 ; CHECK-MVE-NEXT: vmovx.f16 s4, s6
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s2, s2 ; CHECK-MVE-NEXT: vcmp.f16 s2, s2
; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s8, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: mov.w r1, #0 ; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: mov.w r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6 ; CHECK-MVE-NEXT: vseleq.f16 s2, s10, s6
; CHECK-MVE-NEXT: movs r1, #0 ; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: vins.f16 s2, s4 ; CHECK-MVE-NEXT: vins.f16 s2, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s3 ; CHECK-MVE-NEXT: vmovx.f16 s4, s3
; CHECK-MVE-NEXT: vcmp.f16 s4, s4 ; CHECK-MVE-NEXT: vcmp.f16 s4, s4
; CHECK-MVE-NEXT: vmovx.f16 s4, s7 ; CHECK-MVE-NEXT: vmovx.f16 s4, s7
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vs
; CHECK-MVE-NEXT: movvs r1, #1
; CHECK-MVE-NEXT: vmovx.f16 s6, s11
; CHECK-MVE-NEXT: cmp r1, #0
; CHECK-MVE-NEXT: vcmp.f16 s3, s3 ; CHECK-MVE-NEXT: vcmp.f16 s3, s3
; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4 ; CHECK-MVE-NEXT: vseleq.f16 s4, s6, s4
; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr ; CHECK-MVE-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-MVE-NEXT: it vs ; CHECK-MVE-NEXT: cset r0, vs
; CHECK-MVE-NEXT: movvs r0, #1
; CHECK-MVE-NEXT: cmp r0, #0 ; CHECK-MVE-NEXT: cmp r0, #0
; CHECK-MVE-NEXT: vins.f16 s0, s12
; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7 ; CHECK-MVE-NEXT: vseleq.f16 s3, s11, s7
; CHECK-MVE-NEXT: vins.f16 s3, s4 ; CHECK-MVE-NEXT: vins.f16 s3, s4
; CHECK-MVE-NEXT: bx lr ; CHECK-MVE-NEXT: bx lr
; ;
; CHECK-MVEFP-LABEL: vcmp_r_uno_v8f16: ; CHECK-MVEFP-LABEL: vcmp_r_uno_v8f16:
; CHECK-MVEFP: @ %bb.0: @ %entry ; CHECK-MVEFP: @ %bb.0: @ %entry
; CHECK-MVEFP-NEXT: vpt.f16 le, q0, zr ; CHECK-MVEFP-NEXT: vpt.f16 le, q0, zr
; CHECK-MVEFP-NEXT: vcmpt.f16 gt, q0, zr ; CHECK-MVEFP-NEXT: vcmpt.f16 gt, q0, zr
; CHECK-MVEFP-NEXT: vpsel q0, q1, q2 ; CHECK-MVEFP-NEXT: vpsel q0, q1, q2
; CHECK-MVEFP-NEXT: bx lr ; CHECK-MVEFP-NEXT: bx lr
entry: entry:
%c = fcmp uno <8 x half> zeroinitializer, %src %c = fcmp uno <8 x half> zeroinitializer, %src
%s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b %s = select <8 x i1> %c, <8 x half> %a, <8 x half> %b
ret <8 x half> %s ret <8 x half> %s
} }

llvm/test/CodeGen/Thumb2/mve-vcmpr.ll

Show First 20 Lines • Show All 481 Lines • ▼ Show 20 Linesentry:
%c = icmp eq <2 x i64> %src, %sp %c = icmp eq <2 x i64> %src, %sp
%s = select <2 x i1> %c, <2 x i32> %a, <2 x i32> %b %s = select <2 x i1> %c, <2 x i32> %a, <2 x i32> %b
ret <2 x i32> %s ret <2 x i32> %s
} }
define arm_aapcs_vfpcc <2 x i32> @vcmp_multi_v2i32(<2 x i64> %a, <2 x i32> %b, <2 x i32> %c) { define arm_aapcs_vfpcc <2 x i32> @vcmp_multi_v2i32(<2 x i64> %a, <2 x i32> %b, <2 x i32> %c) {
; CHECK-LABEL: vcmp_multi_v2i32: ; CHECK-LABEL: vcmp_multi_v2i32:
; CHECK: @ %bb.0: ; CHECK: @ %bb.0:
; CHECK-NEXT: .save {r4, lr} ; CHECK-NEXT: .save {r7, lr}
; CHECK-NEXT: push {r4, lr} ; CHECK-NEXT: push {r7, lr}
; CHECK-NEXT: .vsave {d8, d9} ; CHECK-NEXT: .vsave {d8, d9}
; CHECK-NEXT: vpush {d8, d9} ; CHECK-NEXT: vpush {d8, d9}
; CHECK-NEXT: vmov r0, r1, d1 ; CHECK-NEXT: vmov r0, r1, d1
; CHECK-NEXT: movs r3, #0
; CHECK-NEXT: orrs r0, r1 ; CHECK-NEXT: orrs r0, r1
; CHECK-NEXT: vmov r1, r2, d0 ; CHECK-NEXT: vmov r1, r2, d0
; CHECK-NEXT: csetm r0, eq ; CHECK-NEXT: csetm r0, eq
; CHECK-NEXT: orrs r1, r2 ; CHECK-NEXT: orrs r1, r2
; CHECK-NEXT: vmov r2, s10 ; CHECK-NEXT: vmov r2, s10
; CHECK-NEXT: csetm r1, eq ; CHECK-NEXT: csetm r1, eq
; CHECK-NEXT: vmov q0[2], q0[0], r1, r0 ; CHECK-NEXT: vmov q0[2], q0[0], r1, r0
; CHECK-NEXT: vmov q0[3], q0[1], r1, r0 ; CHECK-NEXT: vmov q0[3], q0[1], r1, r0
; CHECK-NEXT: vbic q0, q2, q0 ; CHECK-NEXT: vbic q0, q2, q0
; CHECK-NEXT: vmov r0, s2 ; CHECK-NEXT: vmov r0, s2
; CHECK-NEXT: subs r1, r0, r2 ; CHECK-NEXT: subs r3, r0, r2
; CHECK-NEXT: asr.w r12, r0, #31 ; CHECK-NEXT: asr.w r1, r0, #31
; CHECK-NEXT: sbcs.w r1, r12, r2, asr #31 ; CHECK-NEXT: vmov r3, s8
; CHECK-NEXT: mov.w r1, #0 ; CHECK-NEXT: sbcs.w r1, r1, r2, asr #31
; CHECK-NEXT: vmov r2, s0 ; CHECK-NEXT: vmov r2, s0
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r1, lt
; CHECK-NEXT: movlt r1, #1
; CHECK-NEXT: cmp r1, #0 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: vmov r1, s8
; CHECK-NEXT: csetm lr, ne ; CHECK-NEXT: csetm lr, ne
; CHECK-NEXT: subs r1, r2, r3
; CHECK-NEXT: asr.w r12, r2, #31 ; CHECK-NEXT: asr.w r12, r2, #31
; CHECK-NEXT: subs r4, r2, r1 ; CHECK-NEXT: sbcs.w r1, r12, r3, asr #31
; CHECK-NEXT: sbcs.w r1, r12, r1, asr #31 ; CHECK-NEXT: cset r1, lt
; CHECK-NEXT: it lt ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: movlt r3, #1
; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: vmov q3[2], q3[0], r1, lr ; CHECK-NEXT: vmov q3[2], q3[0], r1, lr
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: cmp r2, #0 ; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: vmov q3[3], q3[1], r1, lr ; CHECK-NEXT: vmov q3[3], q3[1], r1, lr
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: vmov q4[2], q4[0], r1, r0 ; CHECK-NEXT: vmov q4[2], q4[0], r1, r0
; CHECK-NEXT: vmov q4[3], q4[1], r1, r0 ; CHECK-NEXT: vmov q4[3], q4[1], r1, r0
; CHECK-NEXT: vmov r0, s6 ; CHECK-NEXT: vmov r0, s6
; CHECK-NEXT: vmov r1, s4 ; CHECK-NEXT: vmov r1, s4
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: cmp r1, #0 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: vmov q1[2], q1[0], r1, r0 ; CHECK-NEXT: vmov q1[2], q1[0], r1, r0
; CHECK-NEXT: vmov q1[3], q1[1], r1, r0 ; CHECK-NEXT: vmov q1[3], q1[1], r1, r0
; CHECK-NEXT: vand q1, q1, q4 ; CHECK-NEXT: vand q1, q1, q4
; CHECK-NEXT: vand q1, q3, q1 ; CHECK-NEXT: vand q1, q3, q1
; CHECK-NEXT: vbic q0, q0, q1 ; CHECK-NEXT: vbic q0, q0, q1
; CHECK-NEXT: vand q1, q2, q1 ; CHECK-NEXT: vand q1, q2, q1
; CHECK-NEXT: vorr q0, q1, q0 ; CHECK-NEXT: vorr q0, q1, q0
; CHECK-NEXT: vpop {d8, d9} ; CHECK-NEXT: vpop {d8, d9}
; CHECK-NEXT: pop {r4, pc} ; CHECK-NEXT: pop {r7, pc}
%a4 = icmp eq <2 x i64> %a, zeroinitializer %a4 = icmp eq <2 x i64> %a, zeroinitializer
%a5 = select <2 x i1> %a4, <2 x i32> zeroinitializer, <2 x i32> %c %a5 = select <2 x i1> %a4, <2 x i32> zeroinitializer, <2 x i32> %c
%a6 = icmp ne <2 x i32> %b, zeroinitializer %a6 = icmp ne <2 x i32> %b, zeroinitializer
%a7 = icmp slt <2 x i32> %a5, %c %a7 = icmp slt <2 x i32> %a5, %c
%a8 = icmp ne <2 x i32> %a5, zeroinitializer %a8 = icmp ne <2 x i32> %a5, zeroinitializer
%a9 = and <2 x i1> %a6, %a8 %a9 = and <2 x i1> %a6, %a8
%a10 = and <2 x i1> %a7, %a9 %a10 = and <2 x i1> %a7, %a9
%a11 = select <2 x i1> %a10, <2 x i32> %c, <2 x i32> %a5 %a11 = select <2 x i1> %a10, <2 x i32> %c, <2 x i32> %a5
▲ Show 20 Lines • Show All 483 Lines • ▼ Show 20 Linesentry:
%c = icmp eq <2 x i64> %sp, %src %c = icmp eq <2 x i64> %sp, %src
%s = select <2 x i1> %c, <2 x i32> %a, <2 x i32> %b %s = select <2 x i1> %c, <2 x i32> %a, <2 x i32> %b
ret <2 x i32> %s ret <2 x i32> %s
} }
define arm_aapcs_vfpcc <2 x i32> @vcmp_r_multi_v2i32(<2 x i64> %a, <2 x i32> %b, <2 x i32> %c) { define arm_aapcs_vfpcc <2 x i32> @vcmp_r_multi_v2i32(<2 x i64> %a, <2 x i32> %b, <2 x i32> %c) {
; CHECK-LABEL: vcmp_r_multi_v2i32: ; CHECK-LABEL: vcmp_r_multi_v2i32:
; CHECK: @ %bb.0: ; CHECK: @ %bb.0:
; CHECK-NEXT: .save {r4, lr} ; CHECK-NEXT: .save {r7, lr}
; CHECK-NEXT: push {r4, lr} ; CHECK-NEXT: push {r7, lr}
; CHECK-NEXT: .vsave {d8, d9} ; CHECK-NEXT: .vsave {d8, d9}
; CHECK-NEXT: vpush {d8, d9} ; CHECK-NEXT: vpush {d8, d9}
; CHECK-NEXT: vmov r0, r1, d1 ; CHECK-NEXT: vmov r0, r1, d1
; CHECK-NEXT: movs r3, #0
; CHECK-NEXT: orrs r0, r1 ; CHECK-NEXT: orrs r0, r1
; CHECK-NEXT: vmov r1, r2, d0 ; CHECK-NEXT: vmov r1, r2, d0
; CHECK-NEXT: csetm r0, eq ; CHECK-NEXT: csetm r0, eq
; CHECK-NEXT: orrs r1, r2 ; CHECK-NEXT: orrs r1, r2
; CHECK-NEXT: vmov r2, s10 ; CHECK-NEXT: vmov r2, s10
; CHECK-NEXT: csetm r1, eq ; CHECK-NEXT: csetm r1, eq
; CHECK-NEXT: vmov q0[2], q0[0], r1, r0 ; CHECK-NEXT: vmov q0[2], q0[0], r1, r0
; CHECK-NEXT: vmov q0[3], q0[1], r1, r0 ; CHECK-NEXT: vmov q0[3], q0[1], r1, r0
; CHECK-NEXT: vbic q0, q2, q0 ; CHECK-NEXT: vbic q0, q2, q0
; CHECK-NEXT: vmov r0, s2 ; CHECK-NEXT: vmov r0, s2
; CHECK-NEXT: subs r1, r0, r2 ; CHECK-NEXT: subs r3, r0, r2
; CHECK-NEXT: asr.w r12, r0, #31 ; CHECK-NEXT: asr.w r1, r0, #31
; CHECK-NEXT: sbcs.w r1, r12, r2, asr #31 ; CHECK-NEXT: vmov r3, s8
; CHECK-NEXT: mov.w r1, #0 ; CHECK-NEXT: sbcs.w r1, r1, r2, asr #31
; CHECK-NEXT: vmov r2, s0 ; CHECK-NEXT: vmov r2, s0
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r1, lt
; CHECK-NEXT: movlt r1, #1
; CHECK-NEXT: cmp r1, #0 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: vmov r1, s8
; CHECK-NEXT: csetm lr, ne ; CHECK-NEXT: csetm lr, ne
; CHECK-NEXT: subs r1, r2, r3
; CHECK-NEXT: asr.w r12, r2, #31 ; CHECK-NEXT: asr.w r12, r2, #31
; CHECK-NEXT: subs r4, r2, r1 ; CHECK-NEXT: sbcs.w r1, r12, r3, asr #31
; CHECK-NEXT: sbcs.w r1, r12, r1, asr #31 ; CHECK-NEXT: cset r1, lt
; CHECK-NEXT: it lt ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: movlt r3, #1
; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: vmov q3[2], q3[0], r1, lr ; CHECK-NEXT: vmov q3[2], q3[0], r1, lr
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: cmp r2, #0 ; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: vmov q3[3], q3[1], r1, lr ; CHECK-NEXT: vmov q3[3], q3[1], r1, lr
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: vmov q4[2], q4[0], r1, r0 ; CHECK-NEXT: vmov q4[2], q4[0], r1, r0
; CHECK-NEXT: vmov q4[3], q4[1], r1, r0 ; CHECK-NEXT: vmov q4[3], q4[1], r1, r0
; CHECK-NEXT: vmov r0, s6 ; CHECK-NEXT: vmov r0, s6
; CHECK-NEXT: vmov r1, s4 ; CHECK-NEXT: vmov r1, s4
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: cmp r1, #0 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: vmov q1[2], q1[0], r1, r0 ; CHECK-NEXT: vmov q1[2], q1[0], r1, r0
; CHECK-NEXT: vmov q1[3], q1[1], r1, r0 ; CHECK-NEXT: vmov q1[3], q1[1], r1, r0
; CHECK-NEXT: vand q1, q1, q4 ; CHECK-NEXT: vand q1, q1, q4
; CHECK-NEXT: vand q1, q3, q1 ; CHECK-NEXT: vand q1, q3, q1
; CHECK-NEXT: vbic q0, q0, q1 ; CHECK-NEXT: vbic q0, q0, q1
; CHECK-NEXT: vand q1, q2, q1 ; CHECK-NEXT: vand q1, q2, q1
; CHECK-NEXT: vorr q0, q1, q0 ; CHECK-NEXT: vorr q0, q1, q0
; CHECK-NEXT: vpop {d8, d9} ; CHECK-NEXT: vpop {d8, d9}
; CHECK-NEXT: pop {r4, pc} ; CHECK-NEXT: pop {r7, pc}
%a4 = icmp eq <2 x i64> %a, zeroinitializer %a4 = icmp eq <2 x i64> %a, zeroinitializer
%a5 = select <2 x i1> %a4, <2 x i32> zeroinitializer, <2 x i32> %c %a5 = select <2 x i1> %a4, <2 x i32> zeroinitializer, <2 x i32> %c
%a6 = icmp ne <2 x i32> %b, zeroinitializer %a6 = icmp ne <2 x i32> %b, zeroinitializer
%a7 = icmp slt <2 x i32> %a5, %c %a7 = icmp slt <2 x i32> %a5, %c
%a8 = icmp ne <2 x i32> %a5, zeroinitializer %a8 = icmp ne <2 x i32> %a5, zeroinitializer
%a9 = and <2 x i1> %a6, %a8 %a9 = and <2 x i1> %a6, %a8
%a10 = and <2 x i1> %a7, %a9 %a10 = and <2 x i1> %a7, %a9
%a11 = select <2 x i1> %a10, <2 x i32> %c, <2 x i32> %a5 %a11 = select <2 x i1> %a10, <2 x i32> %c, <2 x i32> %a5
ret <2 x i32> %a11 ret <2 x i32> %a11
} }

llvm/test/CodeGen/Thumb2/mve-vmaxv-vminv-scalar.ll

Show First 20 Lines • Show All 487 Lines • ▼ Show 20 Lines; CHECK-NEXT: bx lr
%mt = trunc i32 %max to i16 %mt = trunc i32 %max to i16
%1 = select i1 %cmp, i16 %mt, i16 %x %1 = select i1 %cmp, i16 %mt, i16 %x
ret i16 %1 ret i16 %1
} }
define arm_aapcs_vfpcc i64 @uminv2i64(<2 x i64> %vec, i64 %min) { define arm_aapcs_vfpcc i64 @uminv2i64(<2 x i64> %vec, i64 %min) {
; CHECK-LABEL: uminv2i64: ; CHECK-LABEL: uminv2i64:
; CHECK: @ %bb.0: ; CHECK: @ %bb.0:
; CHECK-NEXT: .save {r4, r5, r7, lr} ; CHECK-NEXT: .save {r4, lr}
; CHECK-NEXT: push {r4, r5, r7, lr} ; CHECK-NEXT: push {r4, lr}
; CHECK-NEXT: vmov r2, r12, d1 ; CHECK-NEXT: vmov r2, r12, d1
; CHECK-NEXT: vmov r3, lr, d0 ; CHECK-NEXT: vmov r3, lr, d0
; CHECK-NEXT: cmp r3, r2 ; CHECK-NEXT: cmp r3, r2
; CHECK-NEXT: csel r4, r3, r2, lo ; CHECK-NEXT: csel r4, r3, r2, lo
; CHECK-NEXT: cmp lr, r12 ; CHECK-NEXT: cmp lr, r12
; CHECK-NEXT: csel r2, r3, r2, lo ; CHECK-NEXT: csel r2, r3, r2, lo
; CHECK-NEXT: csel r3, lr, r12, lo ; CHECK-NEXT: csel r3, lr, r12, lo
; CHECK-NEXT: csel r5, r4, r2, eq ; CHECK-NEXT: csel r2, r4, r2, eq
; CHECK-NEXT: movs r4, #0 ; CHECK-NEXT: subs r4, r2, r0
; CHECK-NEXT: subs r2, r5, r0 ; CHECK-NEXT: sbcs.w r4, r3, r1
; CHECK-NEXT: sbcs.w r2, r3, r1 ; CHECK-NEXT: cset r4, lo
; CHECK-NEXT: it lo
; CHECK-NEXT: movlo r4, #1
; CHECK-NEXT: cmp r4, #0 ; CHECK-NEXT: cmp r4, #0
; CHECK-NEXT: csel r0, r5, r0, ne ; CHECK-NEXT: csel r0, r2, r0, ne
; CHECK-NEXT: csel r1, r3, r1, ne ; CHECK-NEXT: csel r1, r3, r1, ne
; CHECK-NEXT: pop {r4, r5, r7, pc} ; CHECK-NEXT: pop {r4, pc}
%x = call i64 @llvm.vector.reduce.umin.v2i64(<2 x i64> %vec) %x = call i64 @llvm.vector.reduce.umin.v2i64(<2 x i64> %vec)
%cmp = icmp ult i64 %x, %min %cmp = icmp ult i64 %x, %min
%1 = select i1 %cmp, i64 %x, i64 %min %1 = select i1 %cmp, i64 %x, i64 %min
ret i64 %1 ret i64 %1
} }
define arm_aapcs_vfpcc i64 @sminv2i64(<2 x i64> %vec, i64 %min) { define arm_aapcs_vfpcc i64 @sminv2i64(<2 x i64> %vec, i64 %min) {
; CHECK-LABEL: sminv2i64: ; CHECK-LABEL: sminv2i64:
; CHECK: @ %bb.0: ; CHECK: @ %bb.0:
; CHECK-NEXT: .save {r4, r5, r7, lr} ; CHECK-NEXT: .save {r4, lr}
; CHECK-NEXT: push {r4, r5, r7, lr} ; CHECK-NEXT: push {r4, lr}
; CHECK-NEXT: vmov r2, r12, d1 ; CHECK-NEXT: vmov r2, r12, d1
; CHECK-NEXT: vmov r3, lr, d0 ; CHECK-NEXT: vmov r3, lr, d0
; CHECK-NEXT: cmp r3, r2 ; CHECK-NEXT: cmp r3, r2
; CHECK-NEXT: csel r4, r3, r2, lo ; CHECK-NEXT: csel r4, r3, r2, lo
; CHECK-NEXT: cmp lr, r12 ; CHECK-NEXT: cmp lr, r12
; CHECK-NEXT: csel r2, r3, r2, lt ; CHECK-NEXT: csel r2, r3, r2, lt
; CHECK-NEXT: csel r3, lr, r12, lt ; CHECK-NEXT: csel r3, lr, r12, lt
; CHECK-NEXT: csel r5, r4, r2, eq ; CHECK-NEXT: csel r2, r4, r2, eq
; CHECK-NEXT: movs r4, #0 ; CHECK-NEXT: subs r4, r2, r0
; CHECK-NEXT: subs r2, r5, r0 ; CHECK-NEXT: sbcs.w r4, r3, r1
; CHECK-NEXT: sbcs.w r2, r3, r1 ; CHECK-NEXT: cset r4, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r4, #1
; CHECK-NEXT: cmp r4, #0 ; CHECK-NEXT: cmp r4, #0
; CHECK-NEXT: csel r0, r5, r0, ne ; CHECK-NEXT: csel r0, r2, r0, ne
; CHECK-NEXT: csel r1, r3, r1, ne ; CHECK-NEXT: csel r1, r3, r1, ne
; CHECK-NEXT: pop {r4, r5, r7, pc} ; CHECK-NEXT: pop {r4, pc}
%x = call i64 @llvm.vector.reduce.smin.v2i64(<2 x i64> %vec) %x = call i64 @llvm.vector.reduce.smin.v2i64(<2 x i64> %vec)
%cmp = icmp slt i64 %x, %min %cmp = icmp slt i64 %x, %min
%1 = select i1 %cmp, i64 %x, i64 %min %1 = select i1 %cmp, i64 %x, i64 %min
ret i64 %1 ret i64 %1
} }
define arm_aapcs_vfpcc i64 @umaxv2i64(<2 x i64> %vec, i64 %max) { define arm_aapcs_vfpcc i64 @umaxv2i64(<2 x i64> %vec, i64 %max) {
; CHECK-LABEL: umaxv2i64: ; CHECK-LABEL: umaxv2i64:
; CHECK: @ %bb.0: ; CHECK: @ %bb.0:
; CHECK-NEXT: .save {r4, r5, r7, lr} ; CHECK-NEXT: .save {r4, lr}
; CHECK-NEXT: push {r4, r5, r7, lr} ; CHECK-NEXT: push {r4, lr}
; CHECK-NEXT: vmov r2, r12, d1 ; CHECK-NEXT: vmov r2, r12, d1
; CHECK-NEXT: vmov r3, lr, d0 ; CHECK-NEXT: vmov r3, lr, d0
; CHECK-NEXT: cmp r3, r2 ; CHECK-NEXT: cmp r3, r2
; CHECK-NEXT: csel r4, r3, r2, hi ; CHECK-NEXT: csel r4, r3, r2, hi
; CHECK-NEXT: cmp lr, r12 ; CHECK-NEXT: cmp lr, r12
; CHECK-NEXT: csel r2, r3, r2, hi ; CHECK-NEXT: csel r2, r3, r2, hi
; CHECK-NEXT: csel r3, lr, r12, hi ; CHECK-NEXT: csel r3, lr, r12, hi
; CHECK-NEXT: csel r5, r4, r2, eq ; CHECK-NEXT: csel r2, r4, r2, eq
; CHECK-NEXT: movs r4, #0 ; CHECK-NEXT: subs r4, r0, r2
; CHECK-NEXT: subs r2, r0, r5 ; CHECK-NEXT: sbcs.w r4, r1, r3
; CHECK-NEXT: sbcs.w r2, r1, r3 ; CHECK-NEXT: cset r4, lo
; CHECK-NEXT: it lo
; CHECK-NEXT: movlo r4, #1
; CHECK-NEXT: cmp r4, #0 ; CHECK-NEXT: cmp r4, #0
; CHECK-NEXT: csel r0, r5, r0, ne ; CHECK-NEXT: csel r0, r2, r0, ne
; CHECK-NEXT: csel r1, r3, r1, ne ; CHECK-NEXT: csel r1, r3, r1, ne
; CHECK-NEXT: pop {r4, r5, r7, pc} ; CHECK-NEXT: pop {r4, pc}
%x = call i64 @llvm.vector.reduce.umax.v2i64(<2 x i64> %vec) %x = call i64 @llvm.vector.reduce.umax.v2i64(<2 x i64> %vec)
%cmp = icmp ugt i64 %x, %max %cmp = icmp ugt i64 %x, %max
%1 = select i1 %cmp, i64 %x, i64 %max %1 = select i1 %cmp, i64 %x, i64 %max
ret i64 %1 ret i64 %1
} }
define arm_aapcs_vfpcc i64 @smaxv2i64(<2 x i64> %vec, i64 %max) { define arm_aapcs_vfpcc i64 @smaxv2i64(<2 x i64> %vec, i64 %max) {
; CHECK-LABEL: smaxv2i64: ; CHECK-LABEL: smaxv2i64:
; CHECK: @ %bb.0: ; CHECK: @ %bb.0:
; CHECK-NEXT: .save {r4, r5, r7, lr} ; CHECK-NEXT: .save {r4, lr}
; CHECK-NEXT: push {r4, r5, r7, lr} ; CHECK-NEXT: push {r4, lr}
; CHECK-NEXT: vmov r2, r12, d1 ; CHECK-NEXT: vmov r2, r12, d1
; CHECK-NEXT: vmov r3, lr, d0 ; CHECK-NEXT: vmov r3, lr, d0
; CHECK-NEXT: cmp r3, r2 ; CHECK-NEXT: cmp r3, r2
; CHECK-NEXT: csel r4, r3, r2, hi ; CHECK-NEXT: csel r4, r3, r2, hi
; CHECK-NEXT: cmp lr, r12 ; CHECK-NEXT: cmp lr, r12
; CHECK-NEXT: csel r2, r3, r2, gt ; CHECK-NEXT: csel r2, r3, r2, gt
; CHECK-NEXT: csel r3, lr, r12, gt ; CHECK-NEXT: csel r3, lr, r12, gt
; CHECK-NEXT: csel r5, r4, r2, eq ; CHECK-NEXT: csel r2, r4, r2, eq
; CHECK-NEXT: movs r4, #0 ; CHECK-NEXT: subs r4, r0, r2
; CHECK-NEXT: subs r2, r0, r5 ; CHECK-NEXT: sbcs.w r4, r1, r3
; CHECK-NEXT: sbcs.w r2, r1, r3 ; CHECK-NEXT: cset r4, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r4, #1
; CHECK-NEXT: cmp r4, #0 ; CHECK-NEXT: cmp r4, #0
; CHECK-NEXT: csel r0, r5, r0, ne ; CHECK-NEXT: csel r0, r2, r0, ne
; CHECK-NEXT: csel r1, r3, r1, ne ; CHECK-NEXT: csel r1, r3, r1, ne
; CHECK-NEXT: pop {r4, r5, r7, pc} ; CHECK-NEXT: pop {r4, pc}
%x = call i64 @llvm.vector.reduce.smax.v2i64(<2 x i64> %vec) %x = call i64 @llvm.vector.reduce.smax.v2i64(<2 x i64> %vec)
%cmp = icmp sgt i64 %x, %max %cmp = icmp sgt i64 %x, %max
%1 = select i1 %cmp, i64 %x, i64 %max %1 = select i1 %cmp, i64 %x, i64 %max
ret i64 %1 ret i64 %1
} }
declare i8 @llvm.vector.reduce.umin.v16i8(<16 x i8>) declare i8 @llvm.vector.reduce.umin.v16i8(<16 x i8>)
Show All 29 Lines

llvm/test/CodeGen/Thumb2/mve-vqdmulh.ll

Show First 20 Lines • Show All 507 Lines • ▼ Show 20 Lines
; CHECK-NEXT: sub sp, #4 ; CHECK-NEXT: sub sp, #4
; CHECK-NEXT: mov r8, r3 ; CHECK-NEXT: mov r8, r3
; CHECK-NEXT: mov r5, r2 ; CHECK-NEXT: mov r5, r2
; CHECK-NEXT: bl __fixdfti ; CHECK-NEXT: bl __fixdfti
; CHECK-NEXT: subs r7, r2, #1 ; CHECK-NEXT: subs r7, r2, #1
; CHECK-NEXT: mov.w r9, #1 ; CHECK-NEXT: mov.w r9, #1
; CHECK-NEXT: sbcs r7, r3, #0 ; CHECK-NEXT: sbcs r7, r3, #0
; CHECK-NEXT: mov.w r4, #0 ; CHECK-NEXT: mov.w r4, #0
; CHECK-NEXT: mov.w r7, #0 ; CHECK-NEXT: cset r7, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r7, #1
; CHECK-NEXT: cmp r7, #0 ; CHECK-NEXT: cmp r7, #0
; CHECK-NEXT: csel r0, r0, r7, ne ; CHECK-NEXT: csel r0, r0, r7, ne
; CHECK-NEXT: csel r3, r3, r7, ne ; CHECK-NEXT: csel r3, r3, r7, ne
; CHECK-NEXT: csel r2, r2, r9, ne ; CHECK-NEXT: csel r2, r2, r9, ne
; CHECK-NEXT: csel r1, r1, r7, ne ; CHECK-NEXT: csel r1, r1, r7, ne
; CHECK-NEXT: rsbs r7, r0, #0 ; CHECK-NEXT: rsbs r7, r0, #0
; CHECK-NEXT: sbcs.w r7, r4, r1 ; CHECK-NEXT: sbcs.w r7, r4, r1
; CHECK-NEXT: sbcs.w r2, r4, r2 ; CHECK-NEXT: sbcs.w r2, r4, r2
; CHECK-NEXT: sbcs.w r2, r4, r3 ; CHECK-NEXT: sbcs.w r2, r4, r3
; CHECK-NEXT: mov.w r2, #0 ; CHECK-NEXT: cset r2, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r2, #1
; CHECK-NEXT: cmp r2, #0 ; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: csel r6, r0, r2, ne ; CHECK-NEXT: csel r6, r0, r2, ne
; CHECK-NEXT: csel r7, r1, r2, ne ; CHECK-NEXT: csel r7, r1, r2, ne
; CHECK-NEXT: mov r0, r5 ; CHECK-NEXT: mov r0, r5
; CHECK-NEXT: mov r1, r8 ; CHECK-NEXT: mov r1, r8
; CHECK-NEXT: bl __fixdfti ; CHECK-NEXT: bl __fixdfti
; CHECK-NEXT: subs r5, r2, #1 ; CHECK-NEXT: subs r5, r2, #1
; CHECK-NEXT: sbcs r5, r3, #0 ; CHECK-NEXT: sbcs r5, r3, #0
; CHECK-NEXT: mov.w r5, #0 ; CHECK-NEXT: cset r5, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r5, #1
; CHECK-NEXT: cmp r5, #0 ; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: csel r0, r0, r5, ne ; CHECK-NEXT: csel r0, r0, r5, ne
; CHECK-NEXT: csel r3, r3, r5, ne ; CHECK-NEXT: csel r3, r3, r5, ne
; CHECK-NEXT: csel r2, r2, r9, ne ; CHECK-NEXT: csel r2, r2, r9, ne
; CHECK-NEXT: csel r1, r1, r5, ne ; CHECK-NEXT: csel r1, r1, r5, ne
; CHECK-NEXT: rsbs r5, r0, #0 ; CHECK-NEXT: rsbs r5, r0, #0
; CHECK-NEXT: sbcs.w r5, r4, r1 ; CHECK-NEXT: sbcs.w r5, r4, r1
; CHECK-NEXT: sbcs.w r2, r4, r2 ; CHECK-NEXT: sbcs.w r2, r4, r2
; CHECK-NEXT: sbcs.w r2, r4, r3 ; CHECK-NEXT: sbcs.w r2, r4, r3
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r3, lt
; CHECK-NEXT: movlt r4, #1 ; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: cmp r4, #0 ; CHECK-NEXT: csel r2, r0, r3, ne
; CHECK-NEXT: csel r2, r0, r4, ne ; CHECK-NEXT: csel r3, r1, r3, ne
; CHECK-NEXT: csel r3, r1, r4, ne
; CHECK-NEXT: mov r0, r6 ; CHECK-NEXT: mov r0, r6
; CHECK-NEXT: mov r1, r7 ; CHECK-NEXT: mov r1, r7
; CHECK-NEXT: add sp, #4 ; CHECK-NEXT: add sp, #4
; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, pc} ; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, pc}
entry: entry:
%conv = fptosi <2 x double> %x to <2 x i128> %conv = fptosi <2 x double> %x to <2 x i128>
%0 = icmp slt <2 x i128> %conv, <i128 18446744073709551616, i128 18446744073709551616> %0 = icmp slt <2 x i128> %conv, <i128 18446744073709551616, i128 18446744073709551616>
%spec.store.select = select <2 x i1> %0, <2 x i128> %conv, <2 x i128> <i128 18446744073709551616, i128 18446744073709551616> %spec.store.select = select <2 x i1> %0, <2 x i128> %conv, <2 x i128> <i128 18446744073709551616, i128 18446744073709551616>
Show All 9 Lines

llvm/test/CodeGen/Thumb2/mve-vqmovn.ll

Show First 20 Lines • Show All 158 Lines • ▼ Show 20 Linesentry:
%c2 = icmp ult <16 x i8> %s0, <i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15> %c2 = icmp ult <16 x i8> %s0, <i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15>
%s2 = select <16 x i1> %c2, <16 x i8> %s0, <16 x i8> <i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15> %s2 = select <16 x i1> %c2, <16 x i8> %s0, <16 x i8> <i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15>
ret <16 x i8> %s2 ret <16 x i8> %s2
} }
define arm_aapcs_vfpcc <2 x i64> @vqmovni64_smaxmin(<2 x i64> %s0) { define arm_aapcs_vfpcc <2 x i64> @vqmovni64_smaxmin(<2 x i64> %s0) {
; CHECK-LABEL: vqmovni64_smaxmin: ; CHECK-LABEL: vqmovni64_smaxmin:
; CHECK: @ %bb.0: @ %entry ; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: vmov r1, r2, d1 ; CHECK-NEXT: vmov r0, r1, d1
; CHECK-NEXT: mvn r12, #-2147483648 ; CHECK-NEXT: mvn r2, #-2147483648
; CHECK-NEXT: movs r0, #0 ; CHECK-NEXT: subs r0, r0, r2
; CHECK-NEXT: subs.w r1, r1, r12 ; CHECK-NEXT: sbcs r0, r1, #0
; CHECK-NEXT: sbcs r1, r2, #0 ; CHECK-NEXT: vmov r1, r3, d0
; CHECK-NEXT: vmov r2, r3, d0 ; CHECK-NEXT: cset r0, lt
; CHECK-NEXT: mov.w r1, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: it lt ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: movlt r1, #1 ; CHECK-NEXT: subs r1, r1, r2
; CHECK-NEXT: mov.w r2, #-1
; CHECK-NEXT: sbcs r1, r3, #0
; CHECK-NEXT: cset r1, lt
; CHECK-NEXT: cmp r1, #0 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: subs.w r2, r2, r12 ; CHECK-NEXT: vmov q1[2], q1[0], r1, r0
; CHECK-NEXT: mov.w r12, #-1 ; CHECK-NEXT: vmov q1[3], q1[1], r1, r0
; CHECK-NEXT: sbcs r2, r3, #0 ; CHECK-NEXT: adr r0, .LCPI12_0
; CHECK-NEXT: mov.w r2, #0 ; CHECK-NEXT: vldrw.u32 q2, [r0]
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r2, #1
; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: csetm r2, ne
; CHECK-NEXT: vmov q1[2], q1[0], r2, r1
; CHECK-NEXT: vmov q1[3], q1[1], r2, r1
; CHECK-NEXT: adr r1, .LCPI12_0
; CHECK-NEXT: vldrw.u32 q2, [r1]
; CHECK-NEXT: vand q0, q0, q1 ; CHECK-NEXT: vand q0, q0, q1
; CHECK-NEXT: vbic q2, q2, q1 ; CHECK-NEXT: vbic q2, q2, q1
; CHECK-NEXT: vorr q0, q0, q2 ; CHECK-NEXT: vorr q0, q0, q2
; CHECK-NEXT: vmov r1, r2, d1 ; CHECK-NEXT: vmov r0, r1, d1
; CHECK-NEXT: rsbs.w r0, r0, #-2147483648
; CHECK-NEXT: sbcs.w r0, r2, r1
; CHECK-NEXT: vmov r1, r3, d0
; CHECK-NEXT: cset r0, lt
; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: rsbs.w r1, r1, #-2147483648 ; CHECK-NEXT: rsbs.w r1, r1, #-2147483648
; CHECK-NEXT: sbcs.w r1, r12, r2 ; CHECK-NEXT: sbcs.w r1, r2, r3
; CHECK-NEXT: vmov r2, r3, d0 ; CHECK-NEXT: cset r1, lt
; CHECK-NEXT: mov.w r1, #0
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r1, #1
; CHECK-NEXT: cmp r1, #0 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: rsbs.w r2, r2, #-2147483648 ; CHECK-NEXT: vmov q1[2], q1[0], r1, r0
; CHECK-NEXT: sbcs.w r2, r12, r3 ; CHECK-NEXT: vmov q1[3], q1[1], r1, r0
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r0, #1
; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: vmov q1[2], q1[0], r0, r1
; CHECK-NEXT: vmov q1[3], q1[1], r0, r1
; CHECK-NEXT: adr r0, .LCPI12_1 ; CHECK-NEXT: adr r0, .LCPI12_1
; CHECK-NEXT: vldrw.u32 q2, [r0] ; CHECK-NEXT: vldrw.u32 q2, [r0]
; CHECK-NEXT: vand q0, q0, q1 ; CHECK-NEXT: vand q0, q0, q1
; CHECK-NEXT: vbic q2, q2, q1 ; CHECK-NEXT: vbic q2, q2, q1
; CHECK-NEXT: vorr q0, q0, q2 ; CHECK-NEXT: vorr q0, q0, q2
; CHECK-NEXT: bx lr ; CHECK-NEXT: bx lr
; CHECK-NEXT: .p2align 4 ; CHECK-NEXT: .p2align 4
; CHECK-NEXT: @ %bb.1: ; CHECK-NEXT: @ %bb.1:
Show All 13 Linesentry:
%c2 = icmp sgt <2 x i64> %s1, <i64 -2147483648, i64 -2147483648> %c2 = icmp sgt <2 x i64> %s1, <i64 -2147483648, i64 -2147483648>
%s2 = select <2 x i1> %c2, <2 x i64> %s1, <2 x i64> <i64 -2147483648, i64 -2147483648> %s2 = select <2 x i1> %c2, <2 x i64> %s1, <2 x i64> <i64 -2147483648, i64 -2147483648>
ret <2 x i64> %s2 ret <2 x i64> %s2
} }
define arm_aapcs_vfpcc <2 x i64> @vqmovni64_sminmax(<2 x i64> %s0) { define arm_aapcs_vfpcc <2 x i64> @vqmovni64_sminmax(<2 x i64> %s0) {
; CHECK-LABEL: vqmovni64_sminmax: ; CHECK-LABEL: vqmovni64_sminmax:
; CHECK: @ %bb.0: @ %entry ; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: vmov r1, r2, d1 ; CHECK-NEXT: vmov r0, r1, d1
; CHECK-NEXT: mov.w r12, #-1 ; CHECK-NEXT: mov.w r2, #-1
; CHECK-NEXT: movs r0, #0 ; CHECK-NEXT: rsbs.w r0, r0, #-2147483648
; CHECK-NEXT: sbcs.w r0, r2, r1
; CHECK-NEXT: vmov r1, r3, d0
; CHECK-NEXT: cset r0, lt
; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: rsbs.w r1, r1, #-2147483648 ; CHECK-NEXT: rsbs.w r1, r1, #-2147483648
; CHECK-NEXT: sbcs.w r1, r12, r2 ; CHECK-NEXT: sbcs.w r1, r2, r3
; CHECK-NEXT: vmov r2, r3, d0 ; CHECK-NEXT: mvn r2, #-2147483648
; CHECK-NEXT: mov.w r1, #0 ; CHECK-NEXT: cset r1, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r1, #1
; CHECK-NEXT: cmp r1, #0 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: rsbs.w r2, r2, #-2147483648 ; CHECK-NEXT: vmov q1[2], q1[0], r1, r0
; CHECK-NEXT: sbcs.w r2, r12, r3 ; CHECK-NEXT: vmov q1[3], q1[1], r1, r0
; CHECK-NEXT: mvn r12, #-2147483648 ; CHECK-NEXT: adr r0, .LCPI13_0
; CHECK-NEXT: mov.w r2, #0 ; CHECK-NEXT: vldrw.u32 q2, [r0]
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r2, #1
; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: csetm r2, ne
; CHECK-NEXT: vmov q1[2], q1[0], r2, r1
; CHECK-NEXT: vmov q1[3], q1[1], r2, r1
; CHECK-NEXT: adr r1, .LCPI13_0
; CHECK-NEXT: vldrw.u32 q2, [r1]
; CHECK-NEXT: vand q0, q0, q1 ; CHECK-NEXT: vand q0, q0, q1
; CHECK-NEXT: vbic q2, q2, q1 ; CHECK-NEXT: vbic q2, q2, q1
; CHECK-NEXT: vorr q0, q0, q2 ; CHECK-NEXT: vorr q0, q0, q2
; CHECK-NEXT: vmov r1, r2, d1 ; CHECK-NEXT: vmov r0, r1, d1
; CHECK-NEXT: subs.w r1, r1, r12 ; CHECK-NEXT: subs r0, r0, r2
; CHECK-NEXT: sbcs r1, r2, #0 ; CHECK-NEXT: sbcs r0, r1, #0
; CHECK-NEXT: vmov r2, r3, d0 ; CHECK-NEXT: vmov r1, r3, d0
; CHECK-NEXT: mov.w r1, #0 ; CHECK-NEXT: cset r0, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r1, #1
; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: subs.w r2, r2, r12
; CHECK-NEXT: sbcs r2, r3, #0
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r0, #1
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: vmov q1[2], q1[0], r0, r1 ; CHECK-NEXT: subs r1, r1, r2
; CHECK-NEXT: vmov q1[3], q1[1], r0, r1 ; CHECK-NEXT: sbcs r1, r3, #0
; CHECK-NEXT: cset r1, lt
; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: vmov q1[2], q1[0], r1, r0
; CHECK-NEXT: vmov q1[3], q1[1], r1, r0
; CHECK-NEXT: adr r0, .LCPI13_1 ; CHECK-NEXT: adr r0, .LCPI13_1
; CHECK-NEXT: vldrw.u32 q2, [r0] ; CHECK-NEXT: vldrw.u32 q2, [r0]
; CHECK-NEXT: vand q0, q0, q1 ; CHECK-NEXT: vand q0, q0, q1
; CHECK-NEXT: vbic q2, q2, q1 ; CHECK-NEXT: vbic q2, q2, q1
; CHECK-NEXT: vorr q0, q0, q2 ; CHECK-NEXT: vorr q0, q0, q2
; CHECK-NEXT: bx lr ; CHECK-NEXT: bx lr
; CHECK-NEXT: .p2align 4 ; CHECK-NEXT: .p2align 4
; CHECK-NEXT: @ %bb.1: ; CHECK-NEXT: @ %bb.1:
Show All 14 Linesentry:
%s2 = select <2 x i1> %c2, <2 x i64> %s1, <2 x i64> <i64 2147483647, i64 2147483647> %s2 = select <2 x i1> %c2, <2 x i64> %s1, <2 x i64> <i64 2147483647, i64 2147483647>
ret <2 x i64> %s2 ret <2 x i64> %s2
} }
define arm_aapcs_vfpcc <2 x i64> @vqmovni64_umaxmin(<2 x i64> %s0) { define arm_aapcs_vfpcc <2 x i64> @vqmovni64_umaxmin(<2 x i64> %s0) {
; CHECK-LABEL: vqmovni64_umaxmin: ; CHECK-LABEL: vqmovni64_umaxmin:
; CHECK: @ %bb.0: @ %entry ; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: vmov r0, r1, d1 ; CHECK-NEXT: vmov r0, r1, d1
; CHECK-NEXT: movs r2, #0
; CHECK-NEXT: vmov.i64 q2, #0xffffffff ; CHECK-NEXT: vmov.i64 q2, #0xffffffff
; CHECK-NEXT: subs.w r0, r0, #-1 ; CHECK-NEXT: subs.w r0, r0, #-1
; CHECK-NEXT: sbcs r0, r1, #0 ; CHECK-NEXT: sbcs r0, r1, #0
; CHECK-NEXT: vmov r1, r3, d0 ; CHECK-NEXT: vmov r1, r2, d0
; CHECK-NEXT: mov.w r0, #0 ; CHECK-NEXT: cset r0, lo
; CHECK-NEXT: it lo
; CHECK-NEXT: movlo r0, #1
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: subs.w r1, r1, #-1 ; CHECK-NEXT: subs.w r1, r1, #-1
; CHECK-NEXT: sbcs r1, r3, #0 ; CHECK-NEXT: sbcs r1, r2, #0
; CHECK-NEXT: it lo ; CHECK-NEXT: cset r1, lo
; CHECK-NEXT: movlo r2, #1 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: vmov q1[2], q1[0], r1, r0 ; CHECK-NEXT: vmov q1[2], q1[0], r1, r0
; CHECK-NEXT: vmov q1[3], q1[1], r1, r0 ; CHECK-NEXT: vmov q1[3], q1[1], r1, r0
; CHECK-NEXT: vbic q2, q2, q1 ; CHECK-NEXT: vbic q2, q2, q1
; CHECK-NEXT: vand q0, q0, q1 ; CHECK-NEXT: vand q0, q0, q1
; CHECK-NEXT: vorr q0, q0, q2 ; CHECK-NEXT: vorr q0, q0, q2
; CHECK-NEXT: bx lr ; CHECK-NEXT: bx lr
entry: entry:
%c1 = icmp ult <2 x i64> %s0, <i64 4294967295, i64 4294967295> %c1 = icmp ult <2 x i64> %s0, <i64 4294967295, i64 4294967295>
%s1 = select <2 x i1> %c1, <2 x i64> %s0, <2 x i64> <i64 4294967295, i64 4294967295> %s1 = select <2 x i1> %c1, <2 x i64> %s0, <2 x i64> <i64 4294967295, i64 4294967295>
ret <2 x i64> %s1 ret <2 x i64> %s1
} }
define arm_aapcs_vfpcc <2 x i64> @vqmovni64_uminmax(<2 x i64> %s0) { define arm_aapcs_vfpcc <2 x i64> @vqmovni64_uminmax(<2 x i64> %s0) {
; CHECK-LABEL: vqmovni64_uminmax: ; CHECK-LABEL: vqmovni64_uminmax:
; CHECK: @ %bb.0: @ %entry ; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: vmov r0, r1, d1 ; CHECK-NEXT: vmov r0, r1, d1
; CHECK-NEXT: movs r2, #0
; CHECK-NEXT: vmov.i64 q2, #0xffffffff ; CHECK-NEXT: vmov.i64 q2, #0xffffffff
; CHECK-NEXT: subs.w r0, r0, #-1 ; CHECK-NEXT: subs.w r0, r0, #-1
; CHECK-NEXT: sbcs r0, r1, #0 ; CHECK-NEXT: sbcs r0, r1, #0
; CHECK-NEXT: vmov r1, r3, d0 ; CHECK-NEXT: vmov r1, r2, d0
; CHECK-NEXT: mov.w r0, #0 ; CHECK-NEXT: cset r0, lo
; CHECK-NEXT: it lo
; CHECK-NEXT: movlo r0, #1
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: subs.w r1, r1, #-1 ; CHECK-NEXT: subs.w r1, r1, #-1
; CHECK-NEXT: sbcs r1, r3, #0 ; CHECK-NEXT: sbcs r1, r2, #0
; CHECK-NEXT: it lo ; CHECK-NEXT: cset r1, lo
; CHECK-NEXT: movlo r2, #1 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: vmov q1[2], q1[0], r1, r0 ; CHECK-NEXT: vmov q1[2], q1[0], r1, r0
; CHECK-NEXT: vmov q1[3], q1[1], r1, r0 ; CHECK-NEXT: vmov q1[3], q1[1], r1, r0
; CHECK-NEXT: vbic q2, q2, q1 ; CHECK-NEXT: vbic q2, q2, q1
; CHECK-NEXT: vand q0, q0, q1 ; CHECK-NEXT: vand q0, q0, q1
; CHECK-NEXT: vorr q0, q0, q2 ; CHECK-NEXT: vorr q0, q0, q2
; CHECK-NEXT: bx lr ; CHECK-NEXT: bx lr
entry: entry:
%c2 = icmp ult <2 x i64> %s0, <i64 4294967295, i64 4294967295> %c2 = icmp ult <2 x i64> %s0, <i64 4294967295, i64 4294967295>
%s2 = select <2 x i1> %c2, <2 x i64> %s0, <2 x i64> <i64 4294967295, i64 4294967295> %s2 = select <2 x i1> %c2, <2 x i64> %s0, <2 x i64> <i64 4294967295, i64 4294967295>
ret <2 x i64> %s2 ret <2 x i64> %s2
} }

llvm/test/CodeGen/Thumb2/mve-vqshrn.ll

Show First 20 Lines • Show All 174 Lines • ▼ Show 20 Linesentry:
%c2 = icmp ult <16 x i8> %s0, <i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15> %c2 = icmp ult <16 x i8> %s0, <i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15>
%s2 = select <16 x i1> %c2, <16 x i8> %s0, <16 x i8> <i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15> %s2 = select <16 x i1> %c2, <16 x i8> %s0, <16 x i8> <i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15, i8 15>
ret <16 x i8> %s2 ret <16 x i8> %s2
} }
define arm_aapcs_vfpcc <2 x i64> @vqshrni64_smaxmin(<2 x i64> %so) { define arm_aapcs_vfpcc <2 x i64> @vqshrni64_smaxmin(<2 x i64> %so) {
; CHECK-LABEL: vqshrni64_smaxmin: ; CHECK-LABEL: vqshrni64_smaxmin:
; CHECK: @ %bb.0: @ %entry ; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: .save {r7, lr} ; CHECK-NEXT: vmov r0, r1, d1
; CHECK-NEXT: push {r7, lr} ; CHECK-NEXT: mvn r12, #-2147483648
; CHECK-NEXT: vmov r2, r3, d1 ; CHECK-NEXT: vmov r2, r3, d0
; CHECK-NEXT: mvn lr, #-2147483648
; CHECK-NEXT: vmov r0, r1, d0
; CHECK-NEXT: asrl r2, r3, #3
; CHECK-NEXT: asrl r0, r1, #3 ; CHECK-NEXT: asrl r0, r1, #3
; CHECK-NEXT: mov.w r12, #0 ; CHECK-NEXT: asrl r2, r3, #3
; CHECK-NEXT: vmov q0[2], q0[0], r0, r2 ; CHECK-NEXT: vmov q0[2], q0[0], r2, r0
; CHECK-NEXT: subs.w r2, r2, lr ; CHECK-NEXT: subs.w r0, r0, r12
; CHECK-NEXT: sbcs r2, r3, #0
; CHECK-NEXT: vmov q0[3], q0[1], r1, r3
; CHECK-NEXT: mov.w r2, #0
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r2, #1
; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: csetm r2, ne
; CHECK-NEXT: subs.w r0, r0, lr
; CHECK-NEXT: sbcs r0, r1, #0 ; CHECK-NEXT: sbcs r0, r1, #0
; CHECK-NEXT: mov.w r0, #0 ; CHECK-NEXT: vmov q0[3], q0[1], r3, r1
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r0, lt
; CHECK-NEXT: movlt r0, #1
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: vmov q1[2], q1[0], r0, r2 ; CHECK-NEXT: subs.w r1, r2, r12
; CHECK-NEXT: vmov q1[3], q1[1], r0, r2 ; CHECK-NEXT: sbcs r1, r3, #0
; CHECK-NEXT: mov.w r2, #-1
; CHECK-NEXT: cset r1, lt
; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: vmov q1[2], q1[0], r1, r0
; CHECK-NEXT: vmov q1[3], q1[1], r1, r0
; CHECK-NEXT: adr r0, .LCPI12_0 ; CHECK-NEXT: adr r0, .LCPI12_0
; CHECK-NEXT: vldrw.u32 q2, [r0] ; CHECK-NEXT: vldrw.u32 q2, [r0]
; CHECK-NEXT: vand q0, q0, q1 ; CHECK-NEXT: vand q0, q0, q1
; CHECK-NEXT: mov.w r2, #-1
; CHECK-NEXT: vbic q1, q2, q1 ; CHECK-NEXT: vbic q1, q2, q1
; CHECK-NEXT: vorr q0, q0, q1 ; CHECK-NEXT: vorr q0, q0, q1
; CHECK-NEXT: vmov r0, r1, d1 ; CHECK-NEXT: vmov r0, r1, d1
; CHECK-NEXT: rsbs.w r0, r0, #-2147483648 ; CHECK-NEXT: rsbs.w r0, r0, #-2147483648
; CHECK-NEXT: sbcs.w r0, r2, r1 ; CHECK-NEXT: sbcs.w r0, r2, r1
; CHECK-NEXT: vmov r1, r3, d0 ; CHECK-NEXT: vmov r1, r3, d0
; CHECK-NEXT: mov.w r0, #0 ; CHECK-NEXT: cset r0, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r0, #1
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: rsbs.w r1, r1, #-2147483648 ; CHECK-NEXT: rsbs.w r1, r1, #-2147483648
; CHECK-NEXT: sbcs.w r1, r2, r3 ; CHECK-NEXT: sbcs.w r1, r2, r3
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r1, lt
; CHECK-NEXT: movlt.w r12, #1 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: cmp.w r12, #0
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: vmov q1[2], q1[0], r1, r0 ; CHECK-NEXT: vmov q1[2], q1[0], r1, r0
; CHECK-NEXT: vmov q1[3], q1[1], r1, r0 ; CHECK-NEXT: vmov q1[3], q1[1], r1, r0
; CHECK-NEXT: adr r0, .LCPI12_1 ; CHECK-NEXT: adr r0, .LCPI12_1
; CHECK-NEXT: vldrw.u32 q2, [r0] ; CHECK-NEXT: vldrw.u32 q2, [r0]
; CHECK-NEXT: vand q0, q0, q1 ; CHECK-NEXT: vand q0, q0, q1
; CHECK-NEXT: vbic q2, q2, q1 ; CHECK-NEXT: vbic q2, q2, q1
; CHECK-NEXT: vorr q0, q0, q2 ; CHECK-NEXT: vorr q0, q0, q2
; CHECK-NEXT: pop {r7, pc} ; CHECK-NEXT: bx lr
; CHECK-NEXT: .p2align 4 ; CHECK-NEXT: .p2align 4
; CHECK-NEXT: @ %bb.1: ; CHECK-NEXT: @ %bb.1:
; CHECK-NEXT: .LCPI12_0: ; CHECK-NEXT: .LCPI12_0:
; CHECK-NEXT: .long 2147483647 @ 0x7fffffff ; CHECK-NEXT: .long 2147483647 @ 0x7fffffff
; CHECK-NEXT: .long 0 @ 0x0 ; CHECK-NEXT: .long 0 @ 0x0
; CHECK-NEXT: .long 2147483647 @ 0x7fffffff ; CHECK-NEXT: .long 2147483647 @ 0x7fffffff
; CHECK-NEXT: .long 0 @ 0x0 ; CHECK-NEXT: .long 0 @ 0x0
; CHECK-NEXT: .LCPI12_1: ; CHECK-NEXT: .LCPI12_1:
; CHECK-NEXT: .long 2147483648 @ 0x80000000 ; CHECK-NEXT: .long 2147483648 @ 0x80000000
; CHECK-NEXT: .long 4294967295 @ 0xffffffff ; CHECK-NEXT: .long 4294967295 @ 0xffffffff
; CHECK-NEXT: .long 2147483648 @ 0x80000000 ; CHECK-NEXT: .long 2147483648 @ 0x80000000
; CHECK-NEXT: .long 4294967295 @ 0xffffffff ; CHECK-NEXT: .long 4294967295 @ 0xffffffff
entry: entry:
%s0 = ashr <2 x i64> %so, <i64 3, i64 3> %s0 = ashr <2 x i64> %so, <i64 3, i64 3>
%c1 = icmp slt <2 x i64> %s0, <i64 2147483647, i64 2147483647> %c1 = icmp slt <2 x i64> %s0, <i64 2147483647, i64 2147483647>
%s1 = select <2 x i1> %c1, <2 x i64> %s0, <2 x i64> <i64 2147483647, i64 2147483647> %s1 = select <2 x i1> %c1, <2 x i64> %s0, <2 x i64> <i64 2147483647, i64 2147483647>
%c2 = icmp sgt <2 x i64> %s1, <i64 -2147483648, i64 -2147483648> %c2 = icmp sgt <2 x i64> %s1, <i64 -2147483648, i64 -2147483648>
%s2 = select <2 x i1> %c2, <2 x i64> %s1, <2 x i64> <i64 -2147483648, i64 -2147483648> %s2 = select <2 x i1> %c2, <2 x i64> %s1, <2 x i64> <i64 -2147483648, i64 -2147483648>
ret <2 x i64> %s2 ret <2 x i64> %s2
} }
define arm_aapcs_vfpcc <2 x i64> @vqshrni64_sminmax(<2 x i64> %so) { define arm_aapcs_vfpcc <2 x i64> @vqshrni64_sminmax(<2 x i64> %so) {
; CHECK-LABEL: vqshrni64_sminmax: ; CHECK-LABEL: vqshrni64_sminmax:
; CHECK: @ %bb.0: @ %entry ; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: .save {r4, r5, r7, lr} ; CHECK-NEXT: .save {r4, lr}
; CHECK-NEXT: push {r4, r5, r7, lr} ; CHECK-NEXT: push {r4, lr}
; CHECK-NEXT: vmov r2, r1, d1 ; CHECK-NEXT: vmov r0, r1, d1
; CHECK-NEXT: mov.w r12, #-1 ; CHECK-NEXT: mov.w r12, #-1
; CHECK-NEXT: asrl r2, r1, #3 ; CHECK-NEXT: asrl r0, r1, #3
; CHECK-NEXT: mov.w lr, #0 ; CHECK-NEXT: rsbs.w r3, r0, #-2147483648
; CHECK-NEXT: rsbs.w r3, r2, #-2147483648
; CHECK-NEXT: sbcs.w r3, r12, r1 ; CHECK-NEXT: sbcs.w r3, r12, r1
; CHECK-NEXT: mov.w r3, #0 ; CHECK-NEXT: cset r3, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r3, #1
; CHECK-NEXT: cmp r3, #0 ; CHECK-NEXT: cmp r3, #0
; CHECK-NEXT: vmov r4, r3, d0 ; CHECK-NEXT: vmov r4, r3, d0
; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: asrl r4, r3, #3 ; CHECK-NEXT: asrl r4, r3, #3
; CHECK-NEXT: rsbs.w r5, r4, #-2147483648 ; CHECK-NEXT: csetm lr, ne
; CHECK-NEXT: vmov q2[2], q2[0], r4, r2 ; CHECK-NEXT: rsbs.w r2, r4, #-2147483648
; CHECK-NEXT: sbcs.w r5, r12, r3 ; CHECK-NEXT: vmov q2[2], q2[0], r4, r0
; CHECK-NEXT: sbcs.w r2, r12, r3
; CHECK-NEXT: vmov q2[3], q2[1], r3, r1 ; CHECK-NEXT: vmov q2[3], q2[1], r3, r1
; CHECK-NEXT: mov.w r5, #0 ; CHECK-NEXT: cset r2, lt
; CHECK-NEXT: cmp r2, #0
; CHECK-NEXT: csetm r2, ne
; CHECK-NEXT: vmov q0[2], q0[0], r2, lr
; CHECK-NEXT: vmov q0[3], q0[1], r2, lr
; CHECK-NEXT: adr r2, .LCPI13_0
; CHECK-NEXT: vldrw.u32 q1, [r2]
; CHECK-NEXT: mvn r2, #-2147483648 ; CHECK-NEXT: mvn r2, #-2147483648
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r5, #1
; CHECK-NEXT: cmp r5, #0
; CHECK-NEXT: csetm r5, ne
; CHECK-NEXT: vmov q0[2], q0[0], r5, r0
; CHECK-NEXT: vmov q0[3], q0[1], r5, r0
; CHECK-NEXT: adr r0, .LCPI13_0
; CHECK-NEXT: vldrw.u32 q1, [r0]
; CHECK-NEXT: vbic q1, q1, q0 ; CHECK-NEXT: vbic q1, q1, q0
; CHECK-NEXT: vand q0, q2, q0 ; CHECK-NEXT: vand q0, q2, q0
; CHECK-NEXT: vorr q0, q0, q1 ; CHECK-NEXT: vorr q0, q0, q1
; CHECK-NEXT: vmov r0, r1, d1 ; CHECK-NEXT: vmov r0, r1, d1
; CHECK-NEXT: subs r0, r0, r2 ; CHECK-NEXT: subs r0, r0, r2
; CHECK-NEXT: sbcs r0, r1, #0 ; CHECK-NEXT: sbcs r0, r1, #0
; CHECK-NEXT: vmov r1, r3, d0 ; CHECK-NEXT: vmov r1, r3, d0
; CHECK-NEXT: mov.w r0, #0 ; CHECK-NEXT: cset r0, lt
; CHECK-NEXT: it lt
; CHECK-NEXT: movlt r0, #1
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: subs r1, r1, r2 ; CHECK-NEXT: subs r1, r1, r2
; CHECK-NEXT: sbcs r1, r3, #0 ; CHECK-NEXT: sbcs r1, r3, #0
; CHECK-NEXT: it lt ; CHECK-NEXT: cset r1, lt
; CHECK-NEXT: movlt.w lr, #1 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: cmp.w lr, #0
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: vmov q1[2], q1[0], r1, r0 ; CHECK-NEXT: vmov q1[2], q1[0], r1, r0
; CHECK-NEXT: vmov q1[3], q1[1], r1, r0 ; CHECK-NEXT: vmov q1[3], q1[1], r1, r0
; CHECK-NEXT: adr r0, .LCPI13_1 ; CHECK-NEXT: adr r0, .LCPI13_1
; CHECK-NEXT: vldrw.u32 q2, [r0] ; CHECK-NEXT: vldrw.u32 q2, [r0]
; CHECK-NEXT: vand q0, q0, q1 ; CHECK-NEXT: vand q0, q0, q1
; CHECK-NEXT: vbic q2, q2, q1 ; CHECK-NEXT: vbic q2, q2, q1
; CHECK-NEXT: vorr q0, q0, q2 ; CHECK-NEXT: vorr q0, q0, q2
; CHECK-NEXT: pop {r4, r5, r7, pc} ; CHECK-NEXT: pop {r4, pc}
; CHECK-NEXT: .p2align 4 ; CHECK-NEXT: .p2align 4
; CHECK-NEXT: @ %bb.1: ; CHECK-NEXT: @ %bb.1:
; CHECK-NEXT: .LCPI13_0: ; CHECK-NEXT: .LCPI13_0:
; CHECK-NEXT: .long 2147483648 @ 0x80000000 ; CHECK-NEXT: .long 2147483648 @ 0x80000000
; CHECK-NEXT: .long 4294967295 @ 0xffffffff ; CHECK-NEXT: .long 4294967295 @ 0xffffffff
; CHECK-NEXT: .long 2147483648 @ 0x80000000 ; CHECK-NEXT: .long 2147483648 @ 0x80000000
; CHECK-NEXT: .long 4294967295 @ 0xffffffff ; CHECK-NEXT: .long 4294967295 @ 0xffffffff
; CHECK-NEXT: .LCPI13_1: ; CHECK-NEXT: .LCPI13_1:
; CHECK-NEXT: .long 2147483647 @ 0x7fffffff ; CHECK-NEXT: .long 2147483647 @ 0x7fffffff
; CHECK-NEXT: .long 0 @ 0x0 ; CHECK-NEXT: .long 0 @ 0x0
; CHECK-NEXT: .long 2147483647 @ 0x7fffffff ; CHECK-NEXT: .long 2147483647 @ 0x7fffffff
; CHECK-NEXT: .long 0 @ 0x0 ; CHECK-NEXT: .long 0 @ 0x0
entry: entry:
%s0 = ashr <2 x i64> %so, <i64 3, i64 3> %s0 = ashr <2 x i64> %so, <i64 3, i64 3>
%c1 = icmp sgt <2 x i64> %s0, <i64 -2147483648, i64 -2147483648> %c1 = icmp sgt <2 x i64> %s0, <i64 -2147483648, i64 -2147483648>
%s1 = select <2 x i1> %c1, <2 x i64> %s0, <2 x i64> <i64 -2147483648, i64 -2147483648> %s1 = select <2 x i1> %c1, <2 x i64> %s0, <2 x i64> <i64 -2147483648, i64 -2147483648>
%c2 = icmp slt <2 x i64> %s1, <i64 2147483647, i64 2147483647> %c2 = icmp slt <2 x i64> %s1, <i64 2147483647, i64 2147483647>
%s2 = select <2 x i1> %c2, <2 x i64> %s1, <2 x i64> <i64 2147483647, i64 2147483647> %s2 = select <2 x i1> %c2, <2 x i64> %s1, <2 x i64> <i64 2147483647, i64 2147483647>
ret <2 x i64> %s2 ret <2 x i64> %s2
} }
define arm_aapcs_vfpcc <2 x i64> @vqshrni64_umaxmin(<2 x i64> %so) { define arm_aapcs_vfpcc <2 x i64> @vqshrni64_umaxmin(<2 x i64> %so) {
; CHECK-LABEL: vqshrni64_umaxmin: ; CHECK-LABEL: vqshrni64_umaxmin:
; CHECK: @ %bb.0: @ %entry ; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: vmov r0, r3, d1 ; CHECK-NEXT: vmov r0, r1, d1
; CHECK-NEXT: mov.w r12, #0
; CHECK-NEXT: vmov r2, r1, d0
; CHECK-NEXT: lsrl r0, r3, #3
; CHECK-NEXT: lsrl r2, r1, #3
; CHECK-NEXT: vmov.i64 q2, #0xffffffff ; CHECK-NEXT: vmov.i64 q2, #0xffffffff
; CHECK-NEXT: vmov r2, r3, d0
; CHECK-NEXT: lsrl r0, r1, #3
; CHECK-NEXT: lsrl r2, r3, #3
; CHECK-NEXT: vmov q0[2], q0[0], r2, r0 ; CHECK-NEXT: vmov q0[2], q0[0], r2, r0
; CHECK-NEXT: subs.w r0, r0, #-1 ; CHECK-NEXT: subs.w r0, r0, #-1
; CHECK-NEXT: sbcs r0, r3, #0 ; CHECK-NEXT: sbcs r0, r1, #0
; CHECK-NEXT: vmov q0[3], q0[1], r1, r3 ; CHECK-NEXT: vmov q0[3], q0[1], r3, r1
; CHECK-NEXT: mov.w r0, #0 ; CHECK-NEXT: cset r0, lo
; CHECK-NEXT: it lo
; CHECK-NEXT: movlo r0, #1
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: subs.w r2, r2, #-1 ; CHECK-NEXT: subs.w r1, r2, #-1
; CHECK-NEXT: sbcs r1, r1, #0 ; CHECK-NEXT: sbcs r1, r3, #0
; CHECK-NEXT: it lo ; CHECK-NEXT: cset r1, lo
; CHECK-NEXT: movlo.w r12, #1 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: cmp.w r12, #0
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: vmov q1[2], q1[0], r1, r0 ; CHECK-NEXT: vmov q1[2], q1[0], r1, r0
; CHECK-NEXT: vmov q1[3], q1[1], r1, r0 ; CHECK-NEXT: vmov q1[3], q1[1], r1, r0
; CHECK-NEXT: vand q0, q0, q1 ; CHECK-NEXT: vand q0, q0, q1
; CHECK-NEXT: vbic q1, q2, q1 ; CHECK-NEXT: vbic q1, q2, q1
; CHECK-NEXT: vorr q0, q0, q1 ; CHECK-NEXT: vorr q0, q0, q1
; CHECK-NEXT: bx lr ; CHECK-NEXT: bx lr
entry: entry:
%s0 = lshr <2 x i64> %so, <i64 3, i64 3> %s0 = lshr <2 x i64> %so, <i64 3, i64 3>
%c1 = icmp ult <2 x i64> %s0, <i64 4294967295, i64 4294967295> %c1 = icmp ult <2 x i64> %s0, <i64 4294967295, i64 4294967295>
%s1 = select <2 x i1> %c1, <2 x i64> %s0, <2 x i64> <i64 4294967295, i64 4294967295> %s1 = select <2 x i1> %c1, <2 x i64> %s0, <2 x i64> <i64 4294967295, i64 4294967295>
ret <2 x i64> %s1 ret <2 x i64> %s1
} }
define arm_aapcs_vfpcc <2 x i64> @vqshrni64_uminmax(<2 x i64> %so) { define arm_aapcs_vfpcc <2 x i64> @vqshrni64_uminmax(<2 x i64> %so) {
; CHECK-LABEL: vqshrni64_uminmax: ; CHECK-LABEL: vqshrni64_uminmax:
; CHECK: @ %bb.0: @ %entry ; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: vmov r0, r3, d1 ; CHECK-NEXT: vmov r0, r1, d1
; CHECK-NEXT: mov.w r12, #0
; CHECK-NEXT: vmov r2, r1, d0
; CHECK-NEXT: lsrl r0, r3, #3
; CHECK-NEXT: lsrl r2, r1, #3
; CHECK-NEXT: vmov.i64 q2, #0xffffffff ; CHECK-NEXT: vmov.i64 q2, #0xffffffff
; CHECK-NEXT: vmov r2, r3, d0
; CHECK-NEXT: lsrl r0, r1, #3
; CHECK-NEXT: lsrl r2, r3, #3
; CHECK-NEXT: vmov q0[2], q0[0], r2, r0 ; CHECK-NEXT: vmov q0[2], q0[0], r2, r0
; CHECK-NEXT: subs.w r0, r0, #-1 ; CHECK-NEXT: subs.w r0, r0, #-1
; CHECK-NEXT: sbcs r0, r3, #0 ; CHECK-NEXT: sbcs r0, r1, #0
; CHECK-NEXT: vmov q0[3], q0[1], r1, r3 ; CHECK-NEXT: vmov q0[3], q0[1], r3, r1
; CHECK-NEXT: mov.w r0, #0 ; CHECK-NEXT: cset r0, lo
; CHECK-NEXT: it lo
; CHECK-NEXT: movlo r0, #1
; CHECK-NEXT: cmp r0, #0 ; CHECK-NEXT: cmp r0, #0
; CHECK-NEXT: csetm r0, ne ; CHECK-NEXT: csetm r0, ne
; CHECK-NEXT: subs.w r2, r2, #-1 ; CHECK-NEXT: subs.w r1, r2, #-1
; CHECK-NEXT: sbcs r1, r1, #0 ; CHECK-NEXT: sbcs r1, r3, #0
; CHECK-NEXT: it lo ; CHECK-NEXT: cset r1, lo
; CHECK-NEXT: movlo.w r12, #1 ; CHECK-NEXT: cmp r1, #0
; CHECK-NEXT: cmp.w r12, #0
; CHECK-NEXT: csetm r1, ne ; CHECK-NEXT: csetm r1, ne
; CHECK-NEXT: vmov q1[2], q1[0], r1, r0 ; CHECK-NEXT: vmov q1[2], q1[0], r1, r0
; CHECK-NEXT: vmov q1[3], q1[1], r1, r0 ; CHECK-NEXT: vmov q1[3], q1[1], r1, r0
; CHECK-NEXT: vand q0, q0, q1 ; CHECK-NEXT: vand q0, q0, q1
; CHECK-NEXT: vbic q1, q2, q1 ; CHECK-NEXT: vbic q1, q2, q1
; CHECK-NEXT: vorr q0, q0, q1 ; CHECK-NEXT: vorr q0, q0, q1
; CHECK-NEXT: bx lr ; CHECK-NEXT: bx lr
entry: entry:
%s0 = lshr <2 x i64> %so, <i64 3, i64 3> %s0 = lshr <2 x i64> %so, <i64 3, i64 3>
%c2 = icmp ult <2 x i64> %s0, <i64 4294967295, i64 4294967295> %c2 = icmp ult <2 x i64> %s0, <i64 4294967295, i64 4294967295>
%s2 = select <2 x i1> %c2, <2 x i64> %s0, <2 x i64> <i64 4294967295, i64 4294967295> %s2 = select <2 x i1> %c2, <2 x i64> %s0, <2 x i64> <i64 4294967295, i64 4294967295>
ret <2 x i64> %s2 ret <2 x i64> %s2
} }