[ARM] Identity shuffles are legal

[dmgreen]

Summary

Identity shuffles, of the form (0, 1, 2, 3, ...) are perfectly legal under MVE (they essentially just become bitcasts). We were not catching that in the existing set of what we considered legal though. On NEON, they would be covered by vext's, but that is not generally available in MVE.

This uses ShuffleVectorInst::isIdentityMask which is a little odd to use here but does what we want and prevents us from just rewriting what is the same function.

[ostannard]

LGTM

llvm/lib/Target/ARM/ARMISelLowering.cpp

bool ARMTargetLowering::isShuffleMaskLegal(ArrayRef<int> M, EVT VT) const {
   }
 
   bool ReverseVEXT, isV_UNDEF;
   unsigned Imm, WhichResult;
 
   unsigned EltSize = VT.getScalarSizeInBits();
   if (EltSize >= 32 ||
       ShuffleVectorSDNode::isSplatMask(&M[0], VT) ||
+      ShuffleVectorInst::isIdentityMask(M) ||
       isVREVMask(M, VT, 64) ||
       isVREVMask(M, VT, 32) ||
       isVREVMask(M, VT, 16))
     return true;
   else if (Subtarget->hasNEON() &&
            (isVEXTMask(M, VT, ReverseVEXT, Imm) ||
             isVTBLMask(M, VT) ||
             isNEONTwoResultShuffleMask(M, VT, WhichResult, isV_UNDEF)))

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

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc -mtriple=thumbv8.1m.main-arm-none-eabi -mattr=+mve -verify-machineinstrs %s -o - | FileCheck %s --check-prefix=CHECK
 
 define arm_aapcs_vfpcc <4 x i32> @sext_0246(<8 x i16> %src) {
 ; CHECK-LABEL: sext_0246:
 ; CHECK:       @ %bb.0: @ %entry
-; CHECK-NEXT:    vmov.u16 r0, q0[0]
-; CHECK-NEXT:    vmov.32 q1[0], r0
-; CHECK-NEXT:    vmov.u16 r0, q0[2]
-; CHECK-NEXT:    vmov.32 q1[1], r0
-; CHECK-NEXT:    vmov.u16 r0, q0[4]
-; CHECK-NEXT:    vmov.32 q1[2], r0
-; CHECK-NEXT:    vmov.u16 r0, q0[6]
-; CHECK-NEXT:    vmov.32 q1[3], r0
-; CHECK-NEXT:    vmovlb.s16 q0, q1
+; CHECK-NEXT:    vmovlb.s16 q0, q0
 ; CHECK-NEXT:    bx lr
 entry:
   %strided.vec = shufflevector <8 x i16> %src, <8 x i16> undef, <4 x i32> <i32 0, i32 2, i32 4, i32 6>
   %out = sext <4 x i16> %strided.vec to <4 x i32>
   ret <4 x i32> %out
 }
 
 define arm_aapcs_vfpcc <4 x i32> @sext_1357(<8 x i16> %src) {
 ; CHECK-LABEL: sext_1357:
 ; CHECK:       @ %bb.0: @ %entry
 ; CHECK-NEXT:    vrev32.16 q0, q0
 ; CHECK-NEXT:    vmovlb.s16 q0, q0
 ; CHECK-NEXT:    bx lr
 entry:
   %strided.vec = shufflevector <8 x i16> %src, <8 x i16> undef, <4 x i32> <i32 1, i32 3, i32 5, i32 7>
   %out = sext <4 x i16> %strided.vec to <4 x i32>
   ret <4 x i32> %out
 }
 
 define arm_aapcs_vfpcc <4 x i32> @zext_0246(<8 x i16> %src) {
 ; CHECK-LABEL: zext_0246:
 ; CHECK:       @ %bb.0: @ %entry
-; CHECK-NEXT:    vmov.u16 r0, q0[0]
-; CHECK-NEXT:    vmov.32 q1[0], r0
-; CHECK-NEXT:    vmov.u16 r0, q0[2]
-; CHECK-NEXT:    vmov.32 q1[1], r0
-; CHECK-NEXT:    vmov.u16 r0, q0[4]
-; CHECK-NEXT:    vmov.32 q1[2], r0
-; CHECK-NEXT:    vmov.u16 r0, q0[6]
-; CHECK-NEXT:    vmov.32 q1[3], r0
-; CHECK-NEXT:    vmovlb.u16 q0, q1
+; CHECK-NEXT:    vmovlb.u16 q0, q0
 ; CHECK-NEXT:    bx lr
 entry:
   %strided.vec = shufflevector <8 x i16> %src, <8 x i16> undef, <4 x i32> <i32 0, i32 2, i32 4, i32 6>
   %out = zext <4 x i16> %strided.vec to <4 x i32>
   ret <4 x i32> %out
 }
 
 define arm_aapcs_vfpcc <4 x i32> @zext_1357(<8 x i16> %src) {
 ; CHECK-LABEL: zext_1357:
 ; CHECK:       @ %bb.0: @ %entry
 ; CHECK-NEXT:    vrev32.16 q0, q0
 ; CHECK-NEXT:    vmovlb.u16 q0, q0
 ; CHECK-NEXT:    bx lr
 entry:
   %strided.vec = shufflevector <8 x i16> %src, <8 x i16> undef, <4 x i32> <i32 1, i32 3, i32 5, i32 7>
   %out = zext <4 x i16> %strided.vec to <4 x i32>
   ret <4 x i32> %out
 }
 
 define arm_aapcs_vfpcc <8 x i16> @sext_02468101214(<16 x i8> %src) {
 ; CHECK-LABEL: sext_02468101214:
 ; CHECK:       @ %bb.0: @ %entry
-; CHECK-NEXT:    vmov.u8 r0, q0[0]
-; CHECK-NEXT:    vmov.16 q1[0], r0
-; CHECK-NEXT:    vmov.u8 r0, q0[2]
-; CHECK-NEXT:    vmov.16 q1[1], r0
-; CHECK-NEXT:    vmov.u8 r0, q0[4]
-; CHECK-NEXT:    vmov.16 q1[2], r0
-; CHECK-NEXT:    vmov.u8 r0, q0[6]
-; CHECK-NEXT:    vmov.16 q1[3], r0
-; CHECK-NEXT:    vmov.u8 r0, q0[8]
-; CHECK-NEXT:    vmov.16 q1[4], r0
-; CHECK-NEXT:    vmov.u8 r0, q0[10]
-; CHECK-NEXT:    vmov.16 q1[5], r0
-; CHECK-NEXT:    vmov.u8 r0, q0[12]
-; CHECK-NEXT:    vmov.16 q1[6], r0
-; CHECK-NEXT:    vmov.u8 r0, q0[14]
-; CHECK-NEXT:    vmov.16 q1[7], r0
-; CHECK-NEXT:    vmovlb.s8 q0, q1
+; CHECK-NEXT:    vmovlb.s8 q0, q0
 ; CHECK-NEXT:    bx lr
 entry:
   %strided.vec = shufflevector <16 x i8> %src, <16 x i8> undef, <8 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14>
   %out = sext <8 x i8> %strided.vec to <8 x i16>
   ret <8 x i16> %out
 }
 
 define arm_aapcs_vfpcc <8 x i16> @sext_13579111315(<16 x i8> %src) {
 ; CHECK-LABEL: sext_13579111315:
 ; CHECK:       @ %bb.0: @ %entry
 ; CHECK-NEXT:    vrev16.8 q0, q0
 ; CHECK-NEXT:    vmovlb.s8 q0, q0
 ; CHECK-NEXT:    bx lr
 entry:
   %strided.vec = shufflevector <16 x i8> %src, <16 x i8> undef, <8 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15>
   %out = sext <8 x i8> %strided.vec to <8 x i16>
   ret <8 x i16> %out
 }
 
 define arm_aapcs_vfpcc <8 x i16> @zext_02468101214(<16 x i8> %src) {
 ; CHECK-LABEL: zext_02468101214:
 ; CHECK:       @ %bb.0: @ %entry
-; CHECK-NEXT:    vmov.u8 r0, q0[0]
-; CHECK-NEXT:    vmov.16 q1[0], r0
-; CHECK-NEXT:    vmov.u8 r0, q0[2]
-; CHECK-NEXT:    vmov.16 q1[1], r0
-; CHECK-NEXT:    vmov.u8 r0, q0[4]
-; CHECK-NEXT:    vmov.16 q1[2], r0
-; CHECK-NEXT:    vmov.u8 r0, q0[6]
-; CHECK-NEXT:    vmov.16 q1[3], r0
-; CHECK-NEXT:    vmov.u8 r0, q0[8]
-; CHECK-NEXT:    vmov.16 q1[4], r0
-; CHECK-NEXT:    vmov.u8 r0, q0[10]
-; CHECK-NEXT:    vmov.16 q1[5], r0
-; CHECK-NEXT:    vmov.u8 r0, q0[12]
-; CHECK-NEXT:    vmov.16 q1[6], r0
-; CHECK-NEXT:    vmov.u8 r0, q0[14]
-; CHECK-NEXT:    vmov.16 q1[7], r0
-; CHECK-NEXT:    vmovlb.u8 q0, q1
+; CHECK-NEXT:    vmovlb.u8 q0, q0
 ; CHECK-NEXT:    bx lr
 entry:
   %strided.vec = shufflevector <16 x i8> %src, <16 x i8> undef, <8 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14>
   %out = zext <8 x i8> %strided.vec to <8 x i16>
   ret <8 x i16> %out
 }
 
 define arm_aapcs_vfpcc <8 x i16> @zext_13579111315(<16 x i8> %src) {