; AVX512VL-NEXT: shrq $48, %rdi
; AVX512VL-NEXT: # kill: %EDI<def> %EDI<kill> %RDI<kill>
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl %bx, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl %bp, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movl %r14d, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa {{[0-9]+}}(%rsp), %xmm1 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 {{[0-9]+}}(%rsp), %xmm1 # 16-byte Reload
; AVX512VL-NEXT: vinsertps {{.*#+}} xmm0 = xmm1[0],xmm0[0],xmm1[2,3]
; AVX512VL-NEXT: vinsertps $32, {{[0-9]+}}(%rsp), %xmm0, %xmm0 # 16-byte Folded Reload
; AVX512VL-NEXT: # xmm0 = xmm0[0,1],mem[0],xmm0[3]
; AVX512VL-NEXT: shrq $48, %rdi
; AVX512VL-NEXT: # kill: %EDI<def> %EDI<kill> %RDI<kill>
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl %bx, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl %bp, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movl %r14d, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa {{[0-9]+}}(%rsp), %xmm1 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 {{[0-9]+}}(%rsp), %xmm1 # 16-byte Reload
; AVX512VL-NEXT: vinsertps {{.*#+}} xmm0 = xmm1[0],xmm0[0],xmm1[2,3]
; AVX512VL-NEXT: vinsertps $32, {{[0-9]+}}(%rsp), %xmm0, %xmm0 # 16-byte Folded Reload
; AVX512VL-NEXT: # xmm0 = xmm0[0,1],mem[0],xmm0[3]
; AVX512VL-NEXT: shrq $48, %rdi
; AVX512VL-NEXT: # kill: %EDI<def> %EDI<kill> %RDI<kill>
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl %bx, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl %r14w, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movl %ebp, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movl %r13d, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl %r12w, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl %r15w, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movl {{[0-9]+}}(%rsp), %edi # 4-byte Reload
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa {{[0-9]+}}(%rsp), %xmm1 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 {{[0-9]+}}(%rsp), %xmm1 # 16-byte Reload
; AVX512VL-NEXT: vinsertps {{.*#+}} xmm0 = xmm1[0],xmm0[0],xmm1[2,3]
; AVX512VL-NEXT: vinsertps $32, {{[0-9]+}}(%rsp), %xmm0, %xmm0 # 16-byte Folded Reload
; AVX512VL-NEXT: # xmm0 = xmm0[0,1],mem[0],xmm0[3]
; AVX512VL-NEXT: vinsertps $48, {{[0-9]+}}(%rsp), %xmm0, %xmm0 # 16-byte Folded Reload
; AVX512VL-NEXT: # xmm0 = xmm0[0,1,2],mem[0]
-; AVX512VL-NEXT: vmovdqa {{[0-9]+}}(%rsp), %xmm1 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 {{[0-9]+}}(%rsp), %xmm1 # 16-byte Reload
; AVX512VL-NEXT: vinsertps $16, {{[0-9]+}}(%rsp), %xmm1, %xmm1 # 16-byte Folded Reload
; AVX512VL-NEXT: # xmm1 = xmm1[0],mem[0],xmm1[2,3]
; AVX512VL-NEXT: vinsertps $32, {{[0-9]+}}(%rsp), %xmm1, %xmm1 # 16-byte Folded Reload
; AVX512VL-NEXT: shrq $48, %rdi
; AVX512VL-NEXT: # kill: %EDI<def> %EDI<kill> %RDI<kill>
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl %r12w, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl %bp, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movl %r14d, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movl %ebx, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl %r13w, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl %r15w, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movl {{[0-9]+}}(%rsp), %edi # 4-byte Reload
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movq {{[0-9]+}}(%rsp), %rdi # 8-byte Reload
; AVX512VL-NEXT: # kill: %EDI<def> %EDI<kill> %RDI<kill>
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movq {{[0-9]+}}(%rsp), %rax # 8-byte Reload
; AVX512VL-NEXT: movzwl %ax, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movq {{[0-9]+}}(%rsp), %rax # 8-byte Reload
; AVX512VL-NEXT: movzwl %ax, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movl {{[0-9]+}}(%rsp), %edi # 4-byte Reload
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movq {{[0-9]+}}(%rsp), %rdi # 8-byte Reload
; AVX512VL-NEXT: # kill: %EDI<def> %EDI<kill> %RDI<kill>
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movq {{[0-9]+}}(%rsp), %rax # 8-byte Reload
; AVX512VL-NEXT: movzwl %ax, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movq {{[0-9]+}}(%rsp), %rax # 8-byte Reload
; AVX512VL-NEXT: movzwl %ax, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movl {{[0-9]+}}(%rsp), %edi # 4-byte Reload
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa {{[0-9]+}}(%rsp), %xmm1 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 {{[0-9]+}}(%rsp), %xmm1 # 16-byte Reload
; AVX512VL-NEXT: vinsertps {{.*#+}} xmm0 = xmm1[0],xmm0[0],xmm1[2,3]
; AVX512VL-NEXT: vinsertps $32, {{[0-9]+}}(%rsp), %xmm0, %xmm0 # 16-byte Folded Reload
; AVX512VL-NEXT: # xmm0 = xmm0[0,1],mem[0],xmm0[3]
; AVX512VL-NEXT: vinsertps $48, {{[0-9]+}}(%rsp), %xmm0, %xmm0 # 16-byte Folded Reload
; AVX512VL-NEXT: # xmm0 = xmm0[0,1,2],mem[0]
-; AVX512VL-NEXT: vmovdqa {{[0-9]+}}(%rsp), %xmm1 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 {{[0-9]+}}(%rsp), %xmm1 # 16-byte Reload
; AVX512VL-NEXT: vinsertps $16, {{[0-9]+}}(%rsp), %xmm1, %xmm1 # 16-byte Folded Reload
; AVX512VL-NEXT: # xmm1 = xmm1[0],mem[0],xmm1[2,3]
; AVX512VL-NEXT: vinsertps $32, {{[0-9]+}}(%rsp), %xmm1, %xmm1 # 16-byte Folded Reload
; AVX512VL-NEXT: vinsertps $48, {{[0-9]+}}(%rsp), %xmm1, %xmm1 # 16-byte Folded Reload
; AVX512VL-NEXT: # xmm1 = xmm1[0,1,2],mem[0]
; AVX512VL-NEXT: vinsertf32x4 $1, %xmm0, %ymm1, %ymm0
-; AVX512VL-NEXT: vmovdqa {{[0-9]+}}(%rsp), %xmm1 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 {{[0-9]+}}(%rsp), %xmm1 # 16-byte Reload
; AVX512VL-NEXT: vinsertps $16, {{[0-9]+}}(%rsp), %xmm1, %xmm1 # 16-byte Folded Reload
; AVX512VL-NEXT: # xmm1 = xmm1[0],mem[0],xmm1[2,3]
; AVX512VL-NEXT: vinsertps $32, {{[0-9]+}}(%rsp), %xmm1, %xmm1 # 16-byte Folded Reload
; AVX512VL-NEXT: # xmm1 = xmm1[0,1],mem[0],xmm1[3]
; AVX512VL-NEXT: vinsertps $48, {{[0-9]+}}(%rsp), %xmm1, %xmm1 # 16-byte Folded Reload
; AVX512VL-NEXT: # xmm1 = xmm1[0,1,2],mem[0]
-; AVX512VL-NEXT: vmovdqa {{[0-9]+}}(%rsp), %xmm2 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 {{[0-9]+}}(%rsp), %xmm2 # 16-byte Reload
; AVX512VL-NEXT: vinsertps $16, {{[0-9]+}}(%rsp), %xmm2, %xmm2 # 16-byte Folded Reload
; AVX512VL-NEXT: # xmm2 = xmm2[0],mem[0],xmm2[2,3]
; AVX512VL-NEXT: vinsertps $32, {{[0-9]+}}(%rsp), %xmm2, %xmm2 # 16-byte Folded Reload
; AVX512VL-NEXT: movq %rdi, %rbx
; AVX512VL-NEXT: movzwl 6(%rbx), %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl 4(%rbx), %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl (%rbx), %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, (%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, (%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl 2(%rbx), %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa (%rsp), %xmm1 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 (%rsp), %xmm1 # 16-byte Reload
; AVX512VL-NEXT: vinsertps {{.*#+}} xmm0 = xmm1[0],xmm0[0],xmm1[2,3]
; AVX512VL-NEXT: vinsertps $32, {{[0-9]+}}(%rsp), %xmm0, %xmm0 # 16-byte Folded Reload
; AVX512VL-NEXT: # xmm0 = xmm0[0,1],mem[0],xmm0[3]
; AVX512VL-NEXT: shrq $48, %rdi
; AVX512VL-NEXT: # kill: %EDI<def> %EDI<kill> %RDI<kill>
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl %bx, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl %bp, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movl %r14d, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa {{[0-9]+}}(%rsp), %xmm1 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 {{[0-9]+}}(%rsp), %xmm1 # 16-byte Reload
; AVX512VL-NEXT: vinsertps {{.*#+}} xmm0 = xmm1[0],xmm0[0],xmm1[2,3]
; AVX512VL-NEXT: vinsertps $32, {{[0-9]+}}(%rsp), %xmm0, %xmm0 # 16-byte Folded Reload
; AVX512VL-NEXT: # xmm0 = xmm0[0,1],mem[0],xmm0[3]
; AVX512VL-NEXT: movq %rdi, %rbx
; AVX512VL-NEXT: movzwl 6(%rbx), %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl 4(%rbx), %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl (%rbx), %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl 2(%rbx), %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl 14(%rbx), %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl 12(%rbx), %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl 8(%rbx), %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, (%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, (%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl 10(%rbx), %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa (%rsp), %xmm1 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 (%rsp), %xmm1 # 16-byte Reload
; AVX512VL-NEXT: vinsertps {{.*#+}} xmm0 = xmm1[0],xmm0[0],xmm1[2,3]
; AVX512VL-NEXT: vinsertps $32, {{[0-9]+}}(%rsp), %xmm0, %xmm0 # 16-byte Folded Reload
; AVX512VL-NEXT: # xmm0 = xmm0[0,1],mem[0],xmm0[3]
; AVX512VL-NEXT: vinsertps $48, {{[0-9]+}}(%rsp), %xmm0, %xmm0 # 16-byte Folded Reload
; AVX512VL-NEXT: # xmm0 = xmm0[0,1,2],mem[0]
-; AVX512VL-NEXT: vmovdqa {{[0-9]+}}(%rsp), %xmm1 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 {{[0-9]+}}(%rsp), %xmm1 # 16-byte Reload
; AVX512VL-NEXT: vinsertps $16, {{[0-9]+}}(%rsp), %xmm1, %xmm1 # 16-byte Folded Reload
; AVX512VL-NEXT: # xmm1 = xmm1[0],mem[0],xmm1[2,3]
; AVX512VL-NEXT: vinsertps $32, {{[0-9]+}}(%rsp), %xmm1, %xmm1 # 16-byte Folded Reload
; AVX512VL-NEXT: movq %rdi, %rbx
; AVX512VL-NEXT: movzwl 6(%rbx), %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl 4(%rbx), %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl (%rbx), %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl 2(%rbx), %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl 14(%rbx), %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl 12(%rbx), %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl 8(%rbx), %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl 10(%rbx), %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl 22(%rbx), %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl 20(%rbx), %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl 16(%rbx), %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl 18(%rbx), %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl 30(%rbx), %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl 28(%rbx), %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl 24(%rbx), %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, (%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, (%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl 26(%rbx), %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa (%rsp), %xmm1 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 (%rsp), %xmm1 # 16-byte Reload
; AVX512VL-NEXT: vinsertps {{.*#+}} xmm0 = xmm1[0],xmm0[0],xmm1[2,3]
; AVX512VL-NEXT: vinsertps $32, {{[0-9]+}}(%rsp), %xmm0, %xmm0 # 16-byte Folded Reload
; AVX512VL-NEXT: # xmm0 = xmm0[0,1],mem[0],xmm0[3]
; AVX512VL-NEXT: vinsertps $48, {{[0-9]+}}(%rsp), %xmm0, %xmm0 # 16-byte Folded Reload
; AVX512VL-NEXT: # xmm0 = xmm0[0,1,2],mem[0]
-; AVX512VL-NEXT: vmovdqa {{[0-9]+}}(%rsp), %xmm1 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 {{[0-9]+}}(%rsp), %xmm1 # 16-byte Reload
; AVX512VL-NEXT: vinsertps $16, {{[0-9]+}}(%rsp), %xmm1, %xmm1 # 16-byte Folded Reload
; AVX512VL-NEXT: # xmm1 = xmm1[0],mem[0],xmm1[2,3]
; AVX512VL-NEXT: vinsertps $32, {{[0-9]+}}(%rsp), %xmm1, %xmm1 # 16-byte Folded Reload
; AVX512VL-NEXT: vinsertps $48, {{[0-9]+}}(%rsp), %xmm1, %xmm1 # 16-byte Folded Reload
; AVX512VL-NEXT: # xmm1 = xmm1[0,1,2],mem[0]
; AVX512VL-NEXT: vinsertf32x4 $1, %xmm0, %ymm1, %ymm0
-; AVX512VL-NEXT: vmovdqa {{[0-9]+}}(%rsp), %xmm1 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 {{[0-9]+}}(%rsp), %xmm1 # 16-byte Reload
; AVX512VL-NEXT: vinsertps $16, {{[0-9]+}}(%rsp), %xmm1, %xmm1 # 16-byte Folded Reload
; AVX512VL-NEXT: # xmm1 = xmm1[0],mem[0],xmm1[2,3]
; AVX512VL-NEXT: vinsertps $32, {{[0-9]+}}(%rsp), %xmm1, %xmm1 # 16-byte Folded Reload
; AVX512VL-NEXT: # xmm1 = xmm1[0,1],mem[0],xmm1[3]
; AVX512VL-NEXT: vinsertps $48, {{[0-9]+}}(%rsp), %xmm1, %xmm1 # 16-byte Folded Reload
; AVX512VL-NEXT: # xmm1 = xmm1[0,1,2],mem[0]
-; AVX512VL-NEXT: vmovdqa {{[0-9]+}}(%rsp), %xmm2 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 {{[0-9]+}}(%rsp), %xmm2 # 16-byte Reload
; AVX512VL-NEXT: vinsertps $16, {{[0-9]+}}(%rsp), %xmm2, %xmm2 # 16-byte Folded Reload
; AVX512VL-NEXT: # xmm2 = xmm2[0],mem[0],xmm2[2,3]
; AVX512VL-NEXT: vinsertps $32, {{[0-9]+}}(%rsp), %xmm2, %xmm2 # 16-byte Folded Reload
; AVX512VL-NEXT: shrl $16, %ebx
; AVX512VL-NEXT: movzwl %ax, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movl %ebx, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
; AVX512VL-NEXT: vcvtss2sd %xmm0, %xmm0, %xmm0
; AVX512VL-NEXT: shrl $16, %ebp
; AVX512VL-NEXT: movzwl %ax, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movl %ebp, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl %bx, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movl %r14d, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
; AVX512VL-NEXT: vcvtss2sd %xmm0, %xmm0, %xmm0
; AVX512VL-NEXT: shrl $16, %ebx
; AVX512VL-NEXT: movzwl %ax, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movl %ebx, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
; AVX512VL-NEXT: vcvtss2sd %xmm0, %xmm0, %xmm0
; AVX512VL-NEXT: shrl $16, %ebp
; AVX512VL-NEXT: movzwl %ax, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movl %ebp, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl %bx, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movl %r14d, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
; AVX512VL-NEXT: vcvtss2sd %xmm0, %xmm0, %xmm0
; AVX512VL-NEXT: shrl $16, %ebp
; AVX512VL-NEXT: movzwl %ax, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movl %ebp, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl %r14w, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movl %ebx, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl %r13w, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movl %r15d, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl %r12w, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movq {{[0-9]+}}(%rsp), %rdi # 8-byte Reload
; AVX512VL-NEXT: # kill: %EDI<def> %EDI<kill> %RDI<kill>
; AVX512VL-NEXT: callq __gnu_h2f_ieee
; AVX512VL-NEXT: shrl $16, %ebp
; AVX512VL-NEXT: movzwl %ax, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movl %ebp, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movzwl %bx, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
-; AVX512VL-NEXT: vmovdqa %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, {{[0-9]+}}(%rsp) # 16-byte Spill
; AVX512VL-NEXT: movl %r14d, %edi
; AVX512VL-NEXT: callq __gnu_h2f_ieee
; AVX512VL-NEXT: vcvtss2sd %xmm0, %xmm0, %xmm0
; AVX512VL-NEXT: pushq %r14
; AVX512VL-NEXT: pushq %rbx
; AVX512VL-NEXT: subq $24, %rsp
-; AVX512VL-NEXT: vmovdqa %xmm0, (%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, (%rsp) # 16-byte Spill
; AVX512VL-NEXT: vmovshdup {{.*#+}} xmm0 = xmm0[1,1,3,3]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movw %ax, %bx
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movzwl %ax, %r14d
; AVX512VL-NEXT: orl %ebx, %r14d
-; AVX512VL-NEXT: vmovdqa (%rsp), %xmm0 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 (%rsp), %xmm0 # 16-byte Reload
; AVX512VL-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[3,1,2,3]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movw %ax, %bx
; AVX512VL-NEXT: shll $16, %ebx
-; AVX512VL-NEXT: vmovdqa (%rsp), %xmm0 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 (%rsp), %xmm0 # 16-byte Reload
; AVX512VL-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movzwl %ax, %eax
; AVX512VL-NEXT: pushq %r14
; AVX512VL-NEXT: pushq %rbx
; AVX512VL-NEXT: subq $24, %rsp
-; AVX512VL-NEXT: vmovdqa %xmm0, (%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, (%rsp) # 16-byte Spill
; AVX512VL-NEXT: vmovshdup {{.*#+}} xmm0 = xmm0[1,1,3,3]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movw %ax, %bx
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movzwl %ax, %r14d
; AVX512VL-NEXT: orl %ebx, %r14d
-; AVX512VL-NEXT: vmovdqa (%rsp), %xmm0 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 (%rsp), %xmm0 # 16-byte Reload
; AVX512VL-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[3,1,2,3]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movw %ax, %bx
; AVX512VL-NEXT: shll $16, %ebx
-; AVX512VL-NEXT: vmovdqa (%rsp), %xmm0 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 (%rsp), %xmm0 # 16-byte Reload
; AVX512VL-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movzwl %ax, %eax
; AVX512VL-NEXT: pushq %r14
; AVX512VL-NEXT: pushq %rbx
; AVX512VL-NEXT: subq $24, %rsp
-; AVX512VL-NEXT: vmovdqa %xmm0, (%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, (%rsp) # 16-byte Spill
; AVX512VL-NEXT: vmovshdup {{.*#+}} xmm0 = xmm0[1,1,3,3]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movw %ax, %bx
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movzwl %ax, %r14d
; AVX512VL-NEXT: orl %ebx, %r14d
-; AVX512VL-NEXT: vmovdqa (%rsp), %xmm0 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 (%rsp), %xmm0 # 16-byte Reload
; AVX512VL-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[3,1,2,3]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movw %ax, %bx
; AVX512VL-NEXT: shll $16, %ebx
-; AVX512VL-NEXT: vmovdqa (%rsp), %xmm0 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 (%rsp), %xmm0 # 16-byte Reload
; AVX512VL-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movzwl %ax, %eax
; AVX512VL-NEXT: pushq %r14
; AVX512VL-NEXT: pushq %rbx
; AVX512VL-NEXT: subq $32, %rsp
-; AVX512VL-NEXT: vmovdqu %ymm0, (%rsp) # 32-byte Spill
+; AVX512VL-NEXT: vmovdqu64 %ymm0, (%rsp) # 32-byte Spill
; AVX512VL-NEXT: vmovshdup {{.*#+}} xmm0 = xmm0[1,1,3,3]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movw %ax, %bx
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movzwl %ax, %r15d
; AVX512VL-NEXT: orl %ebx, %r15d
-; AVX512VL-NEXT: vmovdqu (%rsp), %ymm0 # 32-byte Reload
+; AVX512VL-NEXT: vmovdqu64 (%rsp), %ymm0 # 32-byte Reload
; AVX512VL-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[3,1,2,3]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movw %ax, %bx
; AVX512VL-NEXT: shll $16, %ebx
-; AVX512VL-NEXT: vmovdqu (%rsp), %ymm0 # 32-byte Reload
+; AVX512VL-NEXT: vmovdqu64 (%rsp), %ymm0 # 32-byte Reload
; AVX512VL-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movzwl %ax, %r14d
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movzwl %ax, %r15d
; AVX512VL-NEXT: orl %ebx, %r15d
-; AVX512VL-NEXT: vmovdqa (%rsp), %xmm0 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 (%rsp), %xmm0 # 16-byte Reload
; AVX512VL-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[3,1,2,3]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movw %ax, %bx
; AVX512VL-NEXT: shll $16, %ebx
-; AVX512VL-NEXT: vmovdqa (%rsp), %xmm0 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 (%rsp), %xmm0 # 16-byte Reload
; AVX512VL-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movzwl %ax, %eax
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: # kill: %AX<def> %AX<kill> %EAX<def>
; AVX512VL-NEXT: movl %eax, {{[0-9]+}}(%rsp) # 4-byte Spill
-; AVX512VL-NEXT: vmovdqa (%rsp), %xmm0 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 (%rsp), %xmm0 # 16-byte Reload
; AVX512VL-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: # kill: %AX<def> %AX<kill> %EAX<def>
; AVX512VL-NEXT: movl %eax, {{[0-9]+}}(%rsp) # 4-byte Spill
-; AVX512VL-NEXT: vmovdqa (%rsp), %xmm0 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 (%rsp), %xmm0 # 16-byte Reload
; AVX512VL-NEXT: vmovshdup {{.*#+}} xmm0 = xmm0[1,1,3,3]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: # kill: %AX<def> %AX<kill> %EAX<def>
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: # kill: %AX<def> %AX<kill> %EAX<def>
; AVX512VL-NEXT: movl %eax, {{[0-9]+}}(%rsp) # 4-byte Spill
-; AVX512VL-NEXT: vmovdqa (%rsp), %xmm0 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 (%rsp), %xmm0 # 16-byte Reload
; AVX512VL-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movw %ax, %r13w
-; AVX512VL-NEXT: vmovdqa (%rsp), %xmm0 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 (%rsp), %xmm0 # 16-byte Reload
; AVX512VL-NEXT: vmovshdup {{.*#+}} xmm0 = xmm0[1,1,3,3]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movw %ax, %bx
; AVX512VL-NEXT: vmovaps (%rsp), %xmm0 # 16-byte Reload
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movw %ax, %bp
-; AVX512VL-NEXT: vmovdqu {{[0-9]+}}(%rsp), %ymm0 # 32-byte Reload
+; AVX512VL-NEXT: vmovdqu64 {{[0-9]+}}(%rsp), %ymm0 # 32-byte Reload
; AVX512VL-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[3,1,2,3]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movw %ax, %r14w
-; AVX512VL-NEXT: vmovdqu {{[0-9]+}}(%rsp), %ymm0 # 32-byte Reload
+; AVX512VL-NEXT: vmovdqu64 {{[0-9]+}}(%rsp), %ymm0 # 32-byte Reload
; AVX512VL-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movw %ax, %r15w
; AVX512VL-NEXT: # kill: %XMM0<def> %XMM0<kill> %YMM0<kill>
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movw %ax, %r12w
-; AVX512VL-NEXT: vmovdqu {{[0-9]+}}(%rsp), %ymm0 # 32-byte Reload
+; AVX512VL-NEXT: vmovdqu64 {{[0-9]+}}(%rsp), %ymm0 # 32-byte Reload
; AVX512VL-NEXT: vmovshdup {{.*#+}} xmm0 = xmm0[1,1,3,3]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: # kill: %AX<def> %AX<kill> %EAX<def>
; AVX512VL-NEXT: pushq %rbx
; AVX512VL-NEXT: subq $24, %rsp
; AVX512VL-NEXT: movq %rdi, %rbx
-; AVX512VL-NEXT: vmovdqa %xmm0, (%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, (%rsp) # 16-byte Spill
; AVX512VL-NEXT: vmovshdup {{.*#+}} xmm0 = xmm0[1,1,3,3]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movl %eax, %r14d
-; AVX512VL-NEXT: vmovdqa (%rsp), %xmm0 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 (%rsp), %xmm0 # 16-byte Reload
; AVX512VL-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movl %eax, %r15d
-; AVX512VL-NEXT: vmovdqa (%rsp), %xmm0 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 (%rsp), %xmm0 # 16-byte Reload
; AVX512VL-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[3,1,2,3]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movl %eax, %ebp
; AVX512VL-NEXT: pushq %rbx
; AVX512VL-NEXT: subq $16, %rsp
; AVX512VL-NEXT: movq %rdi, %r14
-; AVX512VL-NEXT: vmovdqa %xmm0, (%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, (%rsp) # 16-byte Spill
; AVX512VL-NEXT: vmovshdup {{.*#+}} xmm0 = xmm0[1,1,3,3]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movw %ax, %bp
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movzwl %ax, %ebx
; AVX512VL-NEXT: orl %ebp, %ebx
-; AVX512VL-NEXT: vmovdqa (%rsp), %xmm0 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 (%rsp), %xmm0 # 16-byte Reload
; AVX512VL-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[3,1,2,3]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movw %ax, %bp
; AVX512VL-NEXT: shll $16, %ebp
-; AVX512VL-NEXT: vmovdqa (%rsp), %xmm0 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 (%rsp), %xmm0 # 16-byte Reload
; AVX512VL-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movzwl %ax, %eax
; AVX512VL-NEXT: pushq %rbx
; AVX512VL-NEXT: subq $16, %rsp
; AVX512VL-NEXT: movq %rdi, %r14
-; AVX512VL-NEXT: vmovdqa %xmm0, (%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, (%rsp) # 16-byte Spill
; AVX512VL-NEXT: vmovshdup {{.*#+}} xmm0 = xmm0[1,1,3,3]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movw %ax, %bp
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movzwl %ax, %ebx
; AVX512VL-NEXT: orl %ebp, %ebx
-; AVX512VL-NEXT: vmovdqa (%rsp), %xmm0 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 (%rsp), %xmm0 # 16-byte Reload
; AVX512VL-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[3,1,2,3]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movw %ax, %bp
; AVX512VL-NEXT: shll $16, %ebp
-; AVX512VL-NEXT: vmovdqa (%rsp), %xmm0 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 (%rsp), %xmm0 # 16-byte Reload
; AVX512VL-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movzwl %ax, %eax
; AVX512VL-NEXT: pushq %rbx
; AVX512VL-NEXT: subq $72, %rsp
; AVX512VL-NEXT: movq %rdi, %rbx
-; AVX512VL-NEXT: vmovdqu %ymm0, {{[0-9]+}}(%rsp) # 32-byte Spill
+; AVX512VL-NEXT: vmovdqu64 %ymm0, {{[0-9]+}}(%rsp) # 32-byte Spill
; AVX512VL-NEXT: vmovshdup {{.*#+}} xmm0 = xmm0[1,1,3,3]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movw %ax, {{[0-9]+}}(%rsp) # 2-byte Spill
-; AVX512VL-NEXT: vmovdqu {{[0-9]+}}(%rsp), %ymm0 # 32-byte Reload
+; AVX512VL-NEXT: vmovdqu64 {{[0-9]+}}(%rsp), %ymm0 # 32-byte Reload
; AVX512VL-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movw %ax, {{[0-9]+}}(%rsp) # 2-byte Spill
-; AVX512VL-NEXT: vmovdqu {{[0-9]+}}(%rsp), %ymm0 # 32-byte Reload
+; AVX512VL-NEXT: vmovdqu64 {{[0-9]+}}(%rsp), %ymm0 # 32-byte Reload
; AVX512VL-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[3,1,2,3]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movl %eax, %r12d
; AVX512VL-NEXT: vmovshdup {{.*#+}} xmm0 = xmm0[1,1,3,3]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movl %eax, %r13d
-; AVX512VL-NEXT: vmovdqa {{[0-9]+}}(%rsp), %xmm0 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 {{[0-9]+}}(%rsp), %xmm0 # 16-byte Reload
; AVX512VL-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movl %eax, %ebp
-; AVX512VL-NEXT: vmovdqa {{[0-9]+}}(%rsp), %xmm0 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 {{[0-9]+}}(%rsp), %xmm0 # 16-byte Reload
; AVX512VL-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[3,1,2,3]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movl %eax, %r14d
; AVX512VL-NEXT: vmovshdup {{.*#+}} xmm0 = xmm0[1,1,3,3]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movw %ax, {{[0-9]+}}(%rsp) # 2-byte Spill
-; AVX512VL-NEXT: vmovdqa {{[0-9]+}}(%rsp), %xmm0 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 {{[0-9]+}}(%rsp), %xmm0 # 16-byte Reload
; AVX512VL-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movw %ax, {{[0-9]+}}(%rsp) # 2-byte Spill
-; AVX512VL-NEXT: vmovdqa {{[0-9]+}}(%rsp), %xmm0 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 {{[0-9]+}}(%rsp), %xmm0 # 16-byte Reload
; AVX512VL-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[3,1,2,3]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movw %ax, {{[0-9]+}}(%rsp) # 2-byte Spill
; AVX512VL-NEXT: vmovshdup {{.*#+}} xmm0 = xmm0[1,1,3,3]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movw %ax, {{[0-9]+}}(%rsp) # 2-byte Spill
-; AVX512VL-NEXT: vmovdqu {{[0-9]+}}(%rsp), %ymm0 # 32-byte Reload
+; AVX512VL-NEXT: vmovdqu64 {{[0-9]+}}(%rsp), %ymm0 # 32-byte Reload
; AVX512VL-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movw %ax, {{[0-9]+}}(%rsp) # 2-byte Spill
-; AVX512VL-NEXT: vmovdqu {{[0-9]+}}(%rsp), %ymm0 # 32-byte Reload
+; AVX512VL-NEXT: vmovdqu64 {{[0-9]+}}(%rsp), %ymm0 # 32-byte Reload
; AVX512VL-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[3,1,2,3]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movw %ax, {{[0-9]+}}(%rsp) # 2-byte Spill
; AVX512VL-NEXT: vmovshdup {{.*#+}} xmm0 = xmm0[1,1,3,3]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movw %ax, {{[0-9]+}}(%rsp) # 2-byte Spill
-; AVX512VL-NEXT: vmovdqa {{[0-9]+}}(%rsp), %xmm0 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 {{[0-9]+}}(%rsp), %xmm0 # 16-byte Reload
; AVX512VL-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movl %eax, %r14d
-; AVX512VL-NEXT: vmovdqa {{[0-9]+}}(%rsp), %xmm0 # 16-byte Reload
+; AVX512VL-NEXT: vmovdqa64 {{[0-9]+}}(%rsp), %xmm0 # 16-byte Reload
; AVX512VL-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[3,1,2,3]
; AVX512VL-NEXT: callq __gnu_f2h_ieee
; AVX512VL-NEXT: movl %eax, %r15d
; AVX512VL: # BB#0:
; AVX512VL-NEXT: pushq %rbx
; AVX512VL-NEXT: subq $16, %rsp
-; AVX512VL-NEXT: vmovdqa %xmm0, (%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, (%rsp) # 16-byte Spill
; AVX512VL-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512VL-NEXT: callq __truncdfhf2
; AVX512VL-NEXT: movw %ax, %bx
; AVX512VL-NEXT: pushq %r14
; AVX512VL-NEXT: pushq %rbx
; AVX512VL-NEXT: subq $40, %rsp
-; AVX512VL-NEXT: vmovdqu %ymm0, (%rsp) # 32-byte Spill
+; AVX512VL-NEXT: vmovdqu64 %ymm0, (%rsp) # 32-byte Spill
; AVX512VL-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512VL-NEXT: callq __truncdfhf2
; AVX512VL-NEXT: movw %ax, %bx
; AVX512VL-NEXT: pushq %r14
; AVX512VL-NEXT: pushq %rbx
; AVX512VL-NEXT: subq $40, %rsp
-; AVX512VL-NEXT: vmovdqu %ymm0, (%rsp) # 32-byte Spill
+; AVX512VL-NEXT: vmovdqu64 %ymm0, (%rsp) # 32-byte Spill
; AVX512VL-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512VL-NEXT: callq __truncdfhf2
; AVX512VL-NEXT: movw %ax, %bx
; AVX512VL-NEXT: pushq %r14
; AVX512VL-NEXT: pushq %rbx
; AVX512VL-NEXT: subq $40, %rsp
-; AVX512VL-NEXT: vmovdqu %ymm0, (%rsp) # 32-byte Spill
+; AVX512VL-NEXT: vmovdqu64 %ymm0, (%rsp) # 32-byte Spill
; AVX512VL-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512VL-NEXT: callq __truncdfhf2
; AVX512VL-NEXT: movw %ax, %bx
; AVX512VL-NEXT: pushq %rbx
; AVX512VL-NEXT: subq $24, %rsp
; AVX512VL-NEXT: movq %rdi, %rbx
-; AVX512VL-NEXT: vmovdqa %xmm0, (%rsp) # 16-byte Spill
+; AVX512VL-NEXT: vmovdqa64 %xmm0, (%rsp) # 16-byte Spill
; AVX512VL-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512VL-NEXT: callq __truncdfhf2
; AVX512VL-NEXT: movl %eax, %ebp
; AVX512VL-NEXT: pushq %rbx
; AVX512VL-NEXT: subq $88, %rsp
; AVX512VL-NEXT: movq %rdi, %rbx
-; AVX512VL-NEXT: vmovdqu %ymm0, {{[0-9]+}}(%rsp) # 32-byte Spill
+; AVX512VL-NEXT: vmovdqu64 %ymm0, {{[0-9]+}}(%rsp) # 32-byte Spill
; AVX512VL-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512VL-NEXT: callq __truncdfhf2
; AVX512VL-NEXT: movl %eax, %r14d
; AVX512VL-NEXT: pushq %rbx
; AVX512VL-NEXT: subq $32, %rsp
; AVX512VL-NEXT: movq %rdi, %r14
-; AVX512VL-NEXT: vmovdqu %ymm0, (%rsp) # 32-byte Spill
+; AVX512VL-NEXT: vmovdqu64 %ymm0, (%rsp) # 32-byte Spill
; AVX512VL-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512VL-NEXT: callq __truncdfhf2
; AVX512VL-NEXT: movw %ax, %bp
; AVX512VL-NEXT: pushq %rbx
; AVX512VL-NEXT: subq $32, %rsp
; AVX512VL-NEXT: movq %rdi, %r14
-; AVX512VL-NEXT: vmovdqu %ymm0, (%rsp) # 32-byte Spill
+; AVX512VL-NEXT: vmovdqu64 %ymm0, (%rsp) # 32-byte Spill
; AVX512VL-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX512VL-NEXT: callq __truncdfhf2
; AVX512VL-NEXT: movw %ax, %bp