AMDGPU: Inline constant when materalizing FI with add on gfx9

[arsenm]

Summary

This was relying on the SGPR usable for the carry out clobber to also
be used for the input. There was no carry out on gfx9. With no carry
out clobber to worry about, so the literal can just be directly used
with a VOP2 add.

[rampitec]

LGTM

[arsenm]

r371791

lib/Target/AMDGPU/SIInstrInfo.cpp

 }
 
 MachineInstrBuilder SIInstrInfo::getAddNoCarry(MachineBasicBlock &MBB,
                                                MachineBasicBlock::iterator I,
                                                const DebugLoc &DL,
                                                Register DestReg,
                                                RegScavenger &RS) const {
   if (ST.hasAddNoCarry())
-    return BuildMI(MBB, I, DL, get(AMDGPU::V_ADD_U32_e64), DestReg);
+    return BuildMI(MBB, I, DL, get(AMDGPU::V_ADD_U32_e32), DestReg);
 
   Register UnusedCarry = RS.scavengeRegister(RI.getBoolRC(), I, 0, false);
   // TODO: Users need to deal with this.
   if (!UnusedCarry.isValid())
     return MachineInstrBuilder();
 
   return BuildMI(MBB, I, DL, get(AMDGPU::V_ADD_I32_e64), DestReg)
            .addReg(UnusedCarry, RegState::Define | RegState::Dead);

lib/Target/AMDGPU/SIRegisterInfo.cpp

default: {
             Register ScaledReg =
               RS->scavengeRegister(&AMDGPU::VGPR_32RegClass, MIB, 0);
 
             BuildMI(*MBB, *MIB, DL, TII->get(AMDGPU::V_LSHRREV_B32_e64),
                     ScaledReg)
               .addImm(ST.getWavefrontSizeLog2())
               .addReg(DiffReg, RegState::Kill);
 
+            const bool IsVOP2 = MIB->getOpcode() == AMDGPU::V_ADD_U32_e32;
+
             // TODO: Fold if use instruction is another add of a constant.
-            if (AMDGPU::isInlinableLiteral32(Offset, ST.hasInv2PiInlineImm())) {
+            if (IsVOP2 || AMDGPU::isInlinableLiteral32(Offset, ST.hasInv2PiInlineImm())) {
               // FIXME: This can fail
               MIB.addImm(Offset);
               MIB.addReg(ScaledReg, RegState::Kill);
+              if (!IsVOP2)
                 MIB.addImm(0); // clamp bit
             } else {
               Register ConstOffsetReg =
                 RS->scavengeRegister(&AMDGPU::SReg_32_XM0RegClass, MIB, 0, false);
 
               // This should always be able to use the unused carry out.
               assert(ConstOffsetReg && "this scavenge should not be able to fail");
 
               BuildMI(*MBB, *MIB, DL, TII->get(AMDGPU::S_MOV_B32), ConstOffsetReg)

test/CodeGen/AMDGPU/frame-index-elimination.ll

 
 ret:
   ret void
 }
 
 ; Added offset can't be used with VOP3 add
 ; GCN-LABEL: {{^}}func_other_fi_user_non_inline_imm_offset_i32:
 ; GCN: s_sub_u32 [[SUB:s[0-9]+|vcc_lo|vcc_hi]], s32, s33
-; GCN-DAG: s_movk_i32 [[K:s[0-9]+|vcc_lo|vcc_hi]], 0x200
+; CI-DAG: s_movk_i32 [[K:s[0-9]+|vcc_lo|vcc_hi]], 0x200
 
 ; CI-DAG: v_lshr_b32_e64 [[SCALED:v[0-9]+]], [[SUB]], 6
 ; CI: v_add_i32_e32 [[VZ:v[0-9]+]], vcc, [[K]], [[SCALED]]
 
 ; GFX9-DAG: v_lshrrev_b32_e64 [[SCALED:v[0-9]+]], 6, [[SUB]]
-; GFX9: v_add_u32_e32 [[VZ:v[0-9]+]], [[K]], [[SCALED]]
+; GFX9: v_add_u32_e32 [[VZ:v[0-9]+]], 0x200, [[SCALED]]
 
 ; GCN: v_mul_u32_u24_e32 [[VZ]], 9, [[VZ]]
 ; GCN: ds_write_b32 v0, [[VZ]]
 define void @func_other_fi_user_non_inline_imm_offset_i32() #0 {
   %alloca0 = alloca [128 x i32], align 4, addrspace(5)
   %alloca1 = alloca [8 x i32], align 4, addrspace(5)
   %gep0 = getelementptr inbounds [128 x i32], [128 x i32] addrspace(5)* %alloca0, i32 0, i32 65
   %gep1 = getelementptr inbounds [8 x i32], [8 x i32] addrspace(5)* %alloca1, i32 0, i32 0
   store volatile i32 7, i32 addrspace(5)* %gep0
   %ptrtoint = ptrtoint i32 addrspace(5)* %gep1 to i32
   %mul = mul i32 %ptrtoint, 9
   store volatile i32 %mul, i32 addrspace(3)* undef
   ret void
 }
 
 ; GCN-LABEL: {{^}}func_other_fi_user_non_inline_imm_offset_i32_vcc_live:
 ; GCN: s_sub_u32 [[DIFF:s[0-9]+]], s32, s33
-; GCN-DAG: s_movk_i32 [[OFFSET:s[0-9]+]], 0x200
+; CI-DAG: s_movk_i32 [[OFFSET:s[0-9]+]], 0x200
 
 ; CI-DAG: v_lshr_b32_e64 [[SCALED:v[0-9]+]], [[DIFF]], 6
 ; CI: v_add_i32_e64 [[VZ:v[0-9]+]], s{{\[[0-9]+:[0-9]+\]}}, [[OFFSET]], [[SCALED]]
 
 ; GFX9-DAG: v_lshrrev_b32_e64 [[SCALED:v[0-9]+]], 6, [[DIFF]]
-; GFX9: v_add_u32_e32 [[VZ:v[0-9]+]], [[OFFSET]], [[SCALED]]
+; GFX9: v_add_u32_e32 [[VZ:v[0-9]+]], 0x200, [[SCALED]]
 
 ; GCN: v_mul_u32_u24_e32 [[VZ]], 9, [[VZ]]
 ; GCN: ds_write_b32 v0, [[VZ]]
 define void @func_other_fi_user_non_inline_imm_offset_i32_vcc_live() #0 {
   %alloca0 = alloca [128 x i32], align 4, addrspace(5)
   %alloca1 = alloca [8 x i32], align 4, addrspace(5)
   %vcc = call i64 asm sideeffect "; def $0", "={vcc}"()
   %gep0 = getelementptr inbounds [128 x i32], [128 x i32] addrspace(5)* %alloca0, i32 0, i32 65

test/CodeGen/AMDGPU/pei-scavenge-sgpr-gfx9.mir

+# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
+# RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx900 -verify-machineinstrs -run-pass=prologepilog %s -o - | FileCheck %s
+
+# Test what happens when an SGPR is unavailable for the unused add. The non-inline constant needs to be folded into the add instruction and not materialized in a register.
+
+---
+name: scavenge_sgpr_pei_no_sgprs
+tracksRegLiveness: true
+
+stack:
+  - { id: 0, type: default, offset: 0, size: 4, alignment: 8192 }
+  - { id: 1, type: default, offset: 0, size: 4, alignment: 8192 }
+
+machineFunctionInfo:
+  isEntryFunction: false
+  scratchRSrcReg:  $sgpr0_sgpr1_sgpr2_sgpr3
+  scratchWaveOffsetReg: $sgpr34
+  frameOffsetReg:  $sgpr33
+  stackPtrOffsetReg:  $sgpr32
+
+body:             |
+  bb.0:
+    liveins: $vgpr1
+
+    ; CHECK-LABEL: name: scavenge_sgpr_pei_no_sgprs
+    ; CHECK: liveins: $vgpr1
+    ; CHECK: $sgpr27 = frame-setup COPY $sgpr33
+    ; CHECK: $sgpr4 = frame-setup S_ADD_U32 $sgpr32, 524224, implicit-def $scc
+    ; CHECK: $sgpr33 = frame-setup S_AND_B32 killed $sgpr4, 4294443008, implicit-def $scc
+    ; CHECK: $sgpr32 = frame-setup S_ADD_U32 $sgpr32, 1572864, implicit-def $scc
+    ; CHECK: S_NOP 0, implicit-def $sgpr4, implicit-def $sgpr5, implicit-def $sgpr6, implicit-def $sgpr7, implicit-def $sgpr8, implicit-def $sgpr9, implicit-def $sgpr10, implicit-def $sgpr11, implicit-def $sgpr12, implicit-def $sgpr13, implicit-def $sgpr14, implicit-def $sgpr15, implicit-def $sgpr16, implicit-def $sgpr17, implicit-def $sgpr18, implicit-def $sgpr19, implicit-def $sgpr20, implicit-def $sgpr21, implicit-def $sgpr22, implicit-def $sgpr23, implicit-def $sgpr24, implicit-def $sgpr25, implicit-def $sgpr26, implicit-def $sgpr17, implicit-def $sgpr28, implicit-def $sgpr29, implicit-def $sgpr30, implicit-def $sgpr31, implicit-def $vcc
+    ; CHECK: $sgpr33 = S_SUB_U32 $sgpr33, $sgpr34, implicit-def $scc
+    ; CHECK: $vgpr3 = V_LSHRREV_B32_e64 6, killed $sgpr33, implicit $exec
+    ; CHECK: $vgpr2 = V_ADD_U32_e32 8192, killed $vgpr3, implicit $exec
+    ; CHECK: $sgpr33 = S_ADD_U32 $sgpr33, $sgpr34, implicit-def $scc
+    ; CHECK: $vgpr0 = V_OR_B32_e32 killed $vgpr2, $vgpr1, implicit $exec, implicit $sgpr4, implicit $sgpr5, implicit $sgpr6, implicit $sgpr7, implicit $sgpr8, implicit $sgpr9, implicit $sgpr10, implicit $sgpr11, implicit $sgpr12, implicit $sgpr13, implicit $sgpr14, implicit $sgpr15, implicit $sgpr16, implicit $sgpr17, implicit $sgpr18, implicit $sgpr19, implicit $sgpr20, implicit $sgpr21, implicit $sgpr22, implicit $sgpr23, implicit $sgpr24, implicit $sgpr25, implicit $sgpr26, implicit $sgpr17, implicit $sgpr28, implicit $sgpr29, implicit $sgpr30, implicit $sgpr31
+    ; CHECK: $sgpr32 = frame-destroy S_SUB_U32 $sgpr32, 1572864, implicit-def $scc
+    ; CHECK: $sgpr33 = frame-setup COPY $sgpr27
+    ; CHECK: S_ENDPGM 0, implicit $vcc
+    S_NOP 0, implicit-def $sgpr4, implicit-def $sgpr5, implicit-def $sgpr6, implicit-def $sgpr7, implicit-def $sgpr8, implicit-def $sgpr9, implicit-def $sgpr10, implicit-def $sgpr11, implicit-def $sgpr12, implicit-def $sgpr13, implicit-def $sgpr14, implicit-def $sgpr15, implicit-def $sgpr16, implicit-def $sgpr17, implicit-def $sgpr18, implicit-def $sgpr19, implicit-def $sgpr20, implicit-def $sgpr21, implicit-def $sgpr22, implicit-def $sgpr23, implicit-def $sgpr24, implicit-def $sgpr25, implicit-def $sgpr26, implicit-def $sgpr17, implicit-def $sgpr28, implicit-def $sgpr29, implicit-def $sgpr30, implicit-def $sgpr31, implicit-def $vcc
+    $vgpr0 = V_OR_B32_e32 %stack.1, $vgpr1, implicit $exec, implicit $sgpr4, implicit $sgpr5, implicit $sgpr6, implicit $sgpr7, implicit $sgpr8, implicit $sgpr9, implicit $sgpr10, implicit $sgpr11, implicit $sgpr12, implicit $sgpr13, implicit $sgpr14, implicit $sgpr15, implicit $sgpr16, implicit $sgpr17, implicit $sgpr18, implicit $sgpr19, implicit $sgpr20, implicit $sgpr21, implicit $sgpr22, implicit $sgpr23, implicit $sgpr24, implicit $sgpr25, implicit $sgpr26, implicit $sgpr17, implicit $sgpr28, implicit $sgpr29, implicit $sgpr30, implicit $sgpr31
+    S_ENDPGM 0, implicit $vcc
+...