1 files changed, 113 insertions, 24 deletions
diff --git a/llvm/test/CodeGen/X86/stack-folding-int-avx512vnni.ll b/llvm/test/CodeGen/X86/stack-folding-int-avx512vnni.ll
index cc927f7c0a5..bdb6925d92e 100644
--- a/llvm/test/CodeGen/X86/stack-folding-int-avx512vnni.ll
+++ b/llvm/test/CodeGen/X86/stack-folding-int-avx512vnni.ll
@@ -1,27 +1,49 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc -O3 -disable-peephole -mtriple=x86_64-unknown-unknown -mattr=+avx512f,+avx512bw,+avx512dq,+avx512vbmi,+avx512cd,+avx512vpopcntdq,+avx512vnni < %s | FileCheck %s
 
 target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
 target triple = "x86_64-unknown-unknown"
 
 define <16 x i32> @stack_fold_vpdpwssd(<16 x i32> %a0, <16 x i32> %a1, <16 x i32> %a2) {
-  ;CHECK-LABEL: stack_fold_vpdpwssd:
-  ;CHECK:       vpdpwssd {{-?[0-9]*}}(%rsp), {{%zmm[0-9][0-9]*}}, {{%zmm[0-9][0-9]*}} {{.*#+}} 64-byte Folded Reload
+; CHECK-LABEL: stack_fold_vpdpwssd:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vmovups %zmm2, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
+; CHECK-NEXT:    #APP
+; CHECK-NEXT:    nop
+; CHECK-NEXT:    #NO_APP
+; CHECK-NEXT:    vpdpwssd {{[-0-9]+}}(%r{{[sb]}}p), %zmm1, %zmm0 # 64-byte Folded Reload
+; CHECK-NEXT:    retq
   %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
   %2 = call <16 x i32> @llvm.x86.avx512.vpdpwssd.512(<16 x i32> %a0, <16 x i32> %a1, <16 x i32> %a2)
   ret <16 x i32> %2
 }
 
 define <16 x i32> @stack_fold_vpdpwssd_commuted(<16 x i32> %a0, <16 x i32> %a1, <16 x i32> %a2) {
-  ;CHECK-LABEL: stack_fold_vpdpwssd_commuted:
-  ;CHECK:       vpdpwssd {{-?[0-9]*}}(%rsp), {{%zmm[0-9][0-9]*}}, {{%zmm[0-9][0-9]*}} {{.*#+}} 64-byte Folded Reload
+; CHECK-LABEL: stack_fold_vpdpwssd_commuted:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vmovups %zmm2, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
+; CHECK-NEXT:    #APP
+; CHECK-NEXT:    nop
+; CHECK-NEXT:    #NO_APP
+; CHECK-NEXT:    vpdpwssd {{[-0-9]+}}(%r{{[sb]}}p), %zmm1, %zmm0 # 64-byte Folded Reload
+; CHECK-NEXT:    retq
   %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
   %2 = call <16 x i32> @llvm.x86.avx512.vpdpwssd.512(<16 x i32> %a0, <16 x i32> %a2, <16 x i32> %a1)
   ret <16 x i32> %2
 }
 
 define <16 x i32> @stack_fold_vpdpwssd_mask(<16 x i32>* %a0, <16 x i32> %a1, <16 x i32> %a2, i16 %mask) {
-  ;CHECK-LABEL: stack_fold_vpdpwssd_mask:
-  ;CHECK:       vpdpwssd {{-?[0-9]*}}(%rsp), {{%zmm[0-9][0-9]*}}, {{%zmm[0-9][0-9]*}} {{{%k[0-7]}}} {{.*#+}} 64-byte Folded Reload
+; CHECK-LABEL: stack_fold_vpdpwssd_mask:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
+; CHECK-NEXT:    #APP
+; CHECK-NEXT:    nop
+; CHECK-NEXT:    #NO_APP
+; CHECK-NEXT:    vmovaps (%rdi), %zmm2
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vpdpwssd {{[-0-9]+}}(%r{{[sb]}}p), %zmm0, %zmm2 {%k1} # 64-byte Folded Reload
+; CHECK-NEXT:    vmovaps %zmm2, %zmm0
+; CHECK-NEXT:    retq
   %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
   %2 = load <16 x i32>, <16 x i32>* %a0
   %3 = call <16 x i32> @llvm.x86.avx512.vpdpwssd.512(<16 x i32> %2, <16 x i32> %a1, <16 x i32> %a2)
@@ -31,8 +53,17 @@ define <16 x i32> @stack_fold_vpdpwssd_mask(<16 x i32>* %a0, <16 x i32> %a1, <16
 }
 
 define <16 x i32> @stack_fold_vpdpwssd_mask_commuted(<16 x i32>* %a0, <16 x i32> %a1, <16 x i32> %a2, i16 %mask) {
-  ;CHECK-LABEL: stack_fold_vpdpwssd_mask_commuted:
-  ;CHECK:       vpdpwssd {{-?[0-9]*}}(%rsp), {{%zmm[0-9][0-9]*}}, {{%zmm[0-9][0-9]*}} {{{%k[0-7]}}} {{.*#+}} 64-byte Folded Reload
+; CHECK-LABEL: stack_fold_vpdpwssd_mask_commuted:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
+; CHECK-NEXT:    #APP
+; CHECK-NEXT:    nop
+; CHECK-NEXT:    #NO_APP
+; CHECK-NEXT:    vmovaps (%rdi), %zmm2
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vpdpwssd {{[-0-9]+}}(%r{{[sb]}}p), %zmm0, %zmm2 {%k1} # 64-byte Folded Reload
+; CHECK-NEXT:    vmovaps %zmm2, %zmm0
+; CHECK-NEXT:    retq
   %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
   %2 = load <16 x i32>, <16 x i32>* %a0
   %3 = call <16 x i32> @llvm.x86.avx512.vpdpwssd.512(<16 x i32> %2, <16 x i32> %a2, <16 x i32> %a1)
@@ -42,8 +73,15 @@ define <16 x i32> @stack_fold_vpdpwssd_mask_commuted(<16 x i32>* %a0, <16 x i32>
 }
 
 define <16 x i32> @stack_fold_vpdpwssd_maskz(<16 x i32> %a0, <16 x i32> %a1, <16 x i32> %a2, i16* %mask) {
-  ;CHECK-LABEL: stack_fold_vpdpwssd_maskz:
-  ;CHECK:       vpdpwssd {{-?[0-9]*}}(%rsp), {{%zmm[0-9][0-9]*}}, {{%zmm[0-9][0-9]*}} {{{%k[0-7]}}} {z} {{.*#+}} 64-byte Folded Reload
+; CHECK-LABEL: stack_fold_vpdpwssd_maskz:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vmovups %zmm2, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
+; CHECK-NEXT:    #APP
+; CHECK-NEXT:    nop
+; CHECK-NEXT:    #NO_APP
+; CHECK-NEXT:    kmovw (%rdi), %k1
+; CHECK-NEXT:    vpdpwssd {{[-0-9]+}}(%r{{[sb]}}p), %zmm1, %zmm0 {%k1} {z} # 64-byte Folded Reload
+; CHECK-NEXT:    retq
   %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
   %2 = call <16 x i32> @llvm.x86.avx512.vpdpwssd.512(<16 x i32> %a0, <16 x i32> %a1, <16 x i32> %a2)
   %3 = load i16, i16* %mask
@@ -53,8 +91,15 @@ define <16 x i32> @stack_fold_vpdpwssd_maskz(<16 x i32> %a0, <16 x i32> %a1, <16
 }
 
 define <16 x i32> @stack_fold_vpdpwssd_maskz_commuted(<16 x i32> %a0, <16 x i32> %a1, <16 x i32> %a2, i16* %mask) {
-  ;CHECK-LABEL: stack_fold_vpdpwssd_maskz_commuted:
-  ;CHECK:       vpdpwssd {{-?[0-9]*}}(%rsp), {{%zmm[0-9][0-9]*}}, {{%zmm[0-9][0-9]*}} {{{%k[0-7]}}} {z} {{.*#+}} 64-byte Folded Reload
+; CHECK-LABEL: stack_fold_vpdpwssd_maskz_commuted:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vmovups %zmm2, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
+; CHECK-NEXT:    #APP
+; CHECK-NEXT:    nop
+; CHECK-NEXT:    #NO_APP
+; CHECK-NEXT:    kmovw (%rdi), %k1
+; CHECK-NEXT:    vpdpwssd {{[-0-9]+}}(%r{{[sb]}}p), %zmm1, %zmm0 {%k1} {z} # 64-byte Folded Reload
+; CHECK-NEXT:    retq
   %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
   %2 = call <16 x i32> @llvm.x86.avx512.vpdpwssd.512(<16 x i32> %a0, <16 x i32> %a2, <16 x i32> %a1)
   %3 = load i16, i16* %mask
@@ -64,24 +109,45 @@ define <16 x i32> @stack_fold_vpdpwssd_maskz_commuted(<16 x i32> %a0, <16 x i32>
 }
 
 define <16 x i32> @stack_fold_vpdpwssds(<16 x i32> %a0, <16 x i32> %a1, <16 x i32> %a2) {
-  ;CHECK-LABEL: stack_fold_vpdpwssds:
-  ;CHECK:       vpdpwssds {{-?[0-9]*}}(%rsp), {{%zmm[0-9][0-9]*}}, {{%zmm[0-9][0-9]*}} {{.*#+}} 64-byte Folded Reload
+; CHECK-LABEL: stack_fold_vpdpwssds:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vmovups %zmm2, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
+; CHECK-NEXT:    #APP
+; CHECK-NEXT:    nop
+; CHECK-NEXT:    #NO_APP
+; CHECK-NEXT:    vpdpwssds {{[-0-9]+}}(%r{{[sb]}}p), %zmm1, %zmm0 # 64-byte Folded Reload
+; CHECK-NEXT:    retq
   %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
   %2 = call <16 x i32> @llvm.x86.avx512.vpdpwssds.512(<16 x i32> %a0, <16 x i32> %a1, <16 x i32> %a2)
   ret <16 x i32> %2
 }
 
 define <16 x i32> @stack_fold_vpdpwssds_commuted(<16 x i32> %a0, <16 x i32> %a1, <16 x i32> %a2) {
-  ;CHECK-LABEL: stack_fold_vpdpwssds_commuted:
-  ;CHECK:       vpdpwssds {{-?[0-9]*}}(%rsp), {{%zmm[0-9][0-9]*}}, {{%zmm[0-9][0-9]*}} {{.*#+}} 64-byte Folded Reload
+; CHECK-LABEL: stack_fold_vpdpwssds_commuted:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vmovups %zmm2, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
+; CHECK-NEXT:    #APP
+; CHECK-NEXT:    nop
+; CHECK-NEXT:    #NO_APP
+; CHECK-NEXT:    vpdpwssds {{[-0-9]+}}(%r{{[sb]}}p), %zmm1, %zmm0 # 64-byte Folded Reload
+; CHECK-NEXT:    retq
   %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
   %2 = call <16 x i32> @llvm.x86.avx512.vpdpwssds.512(<16 x i32> %a0, <16 x i32> %a2, <16 x i32> %a1)
   ret <16 x i32> %2
 }
 
 define <16 x i32> @stack_fold_vpdpwssds_mask(<16 x i32>* %a0, <16 x i32> %a1, <16 x i32> %a2, i16 %mask) {
-  ;CHECK-LABEL: stack_fold_vpdpwssds_mask:
-  ;CHECK:       vpdpwssds {{-?[0-9]*}}(%rsp), {{%zmm[0-9][0-9]*}}, {{%zmm[0-9][0-9]*}} {{{%k[0-7]}}} {{.*#+}} 64-byte Folded Reload
+; CHECK-LABEL: stack_fold_vpdpwssds_mask:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
+; CHECK-NEXT:    #APP
+; CHECK-NEXT:    nop
+; CHECK-NEXT:    #NO_APP
+; CHECK-NEXT:    vmovaps (%rdi), %zmm2
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vpdpwssds {{[-0-9]+}}(%r{{[sb]}}p), %zmm0, %zmm2 {%k1} # 64-byte Folded Reload
+; CHECK-NEXT:    vmovaps %zmm2, %zmm0
+; CHECK-NEXT:    retq
   %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
   %2 = load <16 x i32>, <16 x i32>* %a0
   %3 = call <16 x i32> @llvm.x86.avx512.vpdpwssds.512(<16 x i32> %2, <16 x i32> %a1, <16 x i32> %a2)
@@ -91,8 +157,17 @@ define <16 x i32> @stack_fold_vpdpwssds_mask(<16 x i32>* %a0, <16 x i32> %a1, <1
 }
 
 define <16 x i32> @stack_fold_vpdpwssds_mask_commuted(<16 x i32>* %a0, <16 x i32> %a1, <16 x i32> %a2, i16 %mask) {
-  ;CHECK-LABEL: stack_fold_vpdpwssds_mask_commuted:
-  ;CHECK:       vpdpwssds {{-?[0-9]*}}(%rsp), {{%zmm[0-9][0-9]*}}, {{%zmm[0-9][0-9]*}} {{{%k[0-7]}}} {{.*#+}} 64-byte Folded Reload
+; CHECK-LABEL: stack_fold_vpdpwssds_mask_commuted:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vmovups %zmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
+; CHECK-NEXT:    #APP
+; CHECK-NEXT:    nop
+; CHECK-NEXT:    #NO_APP
+; CHECK-NEXT:    vmovaps (%rdi), %zmm2
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vpdpwssds {{[-0-9]+}}(%r{{[sb]}}p), %zmm0, %zmm2 {%k1} # 64-byte Folded Reload
+; CHECK-NEXT:    vmovaps %zmm2, %zmm0
+; CHECK-NEXT:    retq
   %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
   %2 = load <16 x i32>, <16 x i32>* %a0
   %3 = call <16 x i32> @llvm.x86.avx512.vpdpwssds.512(<16 x i32> %2, <16 x i32> %a2, <16 x i32> %a1)
@@ -102,8 +177,15 @@ define <16 x i32> @stack_fold_vpdpwssds_mask_commuted(<16 x i32>* %a0, <16 x i32
 }
 
 define <16 x i32> @stack_fold_vpdpwssds_maskz(<16 x i32> %a0, <16 x i32> %a1, <16 x i32> %a2, i16* %mask) {
-  ;CHECK-LABEL: stack_fold_vpdpwssds_maskz:
-  ;CHECK:       vpdpwssds {{-?[0-9]*}}(%rsp), {{%zmm[0-9][0-9]*}}, {{%zmm[0-9][0-9]*}} {{{%k[0-7]}}} {z} {{.*#+}} 64-byte Folded Reload
+; CHECK-LABEL: stack_fold_vpdpwssds_maskz:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vmovups %zmm2, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
+; CHECK-NEXT:    #APP
+; CHECK-NEXT:    nop
+; CHECK-NEXT:    #NO_APP
+; CHECK-NEXT:    kmovw (%rdi), %k1
+; CHECK-NEXT:    vpdpwssds {{[-0-9]+}}(%r{{[sb]}}p), %zmm1, %zmm0 {%k1} {z} # 64-byte Folded Reload
+; CHECK-NEXT:    retq
   %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
   %2 = call <16 x i32> @llvm.x86.avx512.vpdpwssds.512(<16 x i32> %a0, <16 x i32> %a1, <16 x i32> %a2)
   %3 = load i16, i16* %mask
@@ -113,8 +195,15 @@ define <16 x i32> @stack_fold_vpdpwssds_maskz(<16 x i32> %a0, <16 x i32> %a1, <1
 }
 
 define <16 x i32> @stack_fold_vpdpwssds_maskz_commuted(<16 x i32> %a0, <16 x i32> %a1, <16 x i32> %a2, i16* %mask) {
-  ;CHECK-LABEL: stack_fold_vpdpwssds_maskz_commuted:
-  ;CHECK:       vpdpwssds {{-?[0-9]*}}(%rsp), {{%zmm[0-9][0-9]*}}, {{%zmm[0-9][0-9]*}} {{{%k[0-7]}}} {z} {{.*#+}} 64-byte Folded Reload
+; CHECK-LABEL: stack_fold_vpdpwssds_maskz_commuted:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vmovups %zmm2, {{[-0-9]+}}(%r{{[sb]}}p) # 64-byte Spill
+; CHECK-NEXT:    #APP
+; CHECK-NEXT:    nop
+; CHECK-NEXT:    #NO_APP
+; CHECK-NEXT:    kmovw (%rdi), %k1
+; CHECK-NEXT:    vpdpwssds {{[-0-9]+}}(%r{{[sb]}}p), %zmm1, %zmm0 {%k1} {z} # 64-byte Folded Reload
+; CHECK-NEXT:    retq
   %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
   %2 = call <16 x i32> @llvm.x86.avx512.vpdpwssds.512(<16 x i32> %a0, <16 x i32> %a2, <16 x i32> %a1)
   %3 = load i16, i16* %mask