[X86] Add tests for FREM

[n-omer]

Summary

Introduces basic test coverage for frem on x86.

Split off from https://reviews.llvm.org/D125988

[RKSimon]

LGTM with one minor - nice to have some actual test coverage!

llvm/test/CodeGen/X86/frem-libcall.ll

4

Add a FIXME as this is currently broken

[n-omer]

Address review comments.

llvm/test/CodeGen/X86/frem-libcall.ll

+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+; RUN: llc -mtriple=x86_64-linux-gnu < %s  | FileCheck %s
+
+; Ensure vectorized FREMs are not widened/unrolled such that they get lowered
+; into libcalls on undef elements.
+
+define float @frem(<2 x float> %a0, <2 x float> %a1, <2 x float> %a2, <2 x float> *%p3) nounwind {
+; CHECK-LABEL: frem:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    pushq %rbx
+; CHECK-NEXT:    subq $80, %rsp
+; CHECK-NEXT:    movq %rdi, %rbx
+; CHECK-NEXT:    movaps %xmm2, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    shufps {{.*#+}} xmm0 = xmm0[3,3,3,3]
+; CHECK-NEXT:    shufps {{.*#+}} xmm1 = xmm1[3,3,3,3]
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    movaps %xmm0, (%rsp) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movhlps {{.*#+}} xmm0 = xmm0[1,1]
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    movhlps {{.*#+}} xmm1 = xmm1[1,1]
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    unpcklps (%rsp), %xmm0 # 16-byte Folded Reload
+; CHECK-NEXT:    # xmm0 = xmm0[0],mem[0],xmm0[1],mem[1]
+; CHECK-NEXT:    movaps %xmm0, (%rsp) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    movaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    shufps {{.*#+}} xmm0 = xmm0[1,1,1,1]
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    shufps {{.*#+}} xmm1 = xmm1[1,1,1,1]
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    unpcklps {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1]
+; CHECK-NEXT:    unpcklpd (%rsp), %xmm1 # 16-byte Folded Reload
+; CHECK-NEXT:    # xmm1 = xmm1[0],mem[0]
+; CHECK-NEXT:    divps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Folded Reload
+; CHECK-NEXT:    movaps %xmm1, %xmm0
+; CHECK-NEXT:    shufps {{.*#+}} xmm0 = xmm0[1,1],xmm1[1,1]
+; CHECK-NEXT:    addss %xmm1, %xmm0
+; CHECK-NEXT:    movlps %xmm1, (%rbx)
+; CHECK-NEXT:    addq $80, %rsp
+; CHECK-NEXT:    popq %rbx
+; CHECK-NEXT:    retq
+  %frem = frem <2 x float> %a0, %a1
+  %fdiv = fdiv <2 x float> %frem, %a2
+  %ex0 = extractelement <2 x float> %fdiv, i32 0
+  %ex1 = extractelement <2 x float> %fdiv, i32 1
+  %res = fadd float %ex0, %ex1
+  store <2 x float> %fdiv, <2 x float> *%p3
+  ret float %res
+}
+

llvm/test/CodeGen/X86/frem.ll

+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+; RUN: llc -mtriple=x86_64-linux-gnu < %s  | FileCheck %s
+
+; Basic test coverage for FREM
+
+define void @frem_f16(half %a0, half %a1, half *%p3) nounwind {
+; CHECK-LABEL: frem_f16:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    pushq %rbp
+; CHECK-NEXT:    pushq %rbx
+; CHECK-NEXT:    pushq %rax
+; CHECK-NEXT:    movq %rdx, %rbx
+; CHECK-NEXT:    movl %edi, %ebp
+; CHECK-NEXT:    movzwl %si, %edi
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl %bp, %edi
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, (%rbx)
+; CHECK-NEXT:    addq $8, %rsp
+; CHECK-NEXT:    popq %rbx
+; CHECK-NEXT:    popq %rbp
+; CHECK-NEXT:    retq
+  %frem = frem half %a0, %a1
+  store half %frem, half *%p3
+  ret void
+}
+
+define void @frem_f32(float %a0, float %a1, float *%p3) nounwind {
+; CHECK-LABEL: frem_f32:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    pushq %rbx
+; CHECK-NEXT:    movq %rdi, %rbx
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    movss %xmm0, (%rbx)
+; CHECK-NEXT:    popq %rbx
+; CHECK-NEXT:    retq
+  %frem = frem float %a0, %a1
+  store float %frem, float *%p3
+  ret void
+}
+
+define void @frem_f64(double %a0, double %a1, double *%p3) nounwind {
+; CHECK-LABEL: frem_f64:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    pushq %rbx
+; CHECK-NEXT:    movq %rdi, %rbx
+; CHECK-NEXT:    callq fmod@PLT
+; CHECK-NEXT:    movsd %xmm0, (%rbx)
+; CHECK-NEXT:    popq %rbx
+; CHECK-NEXT:    retq
+  %frem = frem double %a0, %a1
+  store double %frem, double *%p3
+  ret void
+}
+
+define void @frem_f80(x86_fp80 %a0, x86_fp80 %a1, x86_fp80 *%p3) nounwind {
+; CHECK-LABEL: frem_f80:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    pushq %rbx
+; CHECK-NEXT:    subq $32, %rsp
+; CHECK-NEXT:    movq %rdi, %rbx
+; CHECK-NEXT:    fldt {{[0-9]+}}(%rsp)
+; CHECK-NEXT:    fldt {{[0-9]+}}(%rsp)
+; CHECK-NEXT:    fstpt {{[0-9]+}}(%rsp)
+; CHECK-NEXT:    fstpt (%rsp)
+; CHECK-NEXT:    callq fmodl@PLT
+; CHECK-NEXT:    fstpt (%rbx)
+; CHECK-NEXT:    addq $32, %rsp
+; CHECK-NEXT:    popq %rbx
+; CHECK-NEXT:    retq
+  %frem = frem x86_fp80 %a0, %a1
+  store x86_fp80 %frem, x86_fp80 *%p3
+  ret void
+}
+
+define void @frem_f128(fp128 %a0, fp128 %a1, fp128 *%p3) nounwind {
+; CHECK-LABEL: frem_f128:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    pushq %rbx
+; CHECK-NEXT:    movq %rdi, %rbx
+; CHECK-NEXT:    callq fmodl@PLT
+; CHECK-NEXT:    movaps %xmm0, (%rbx)
+; CHECK-NEXT:    popq %rbx
+; CHECK-NEXT:    retq
+  %frem = frem fp128 %a0, %a1
+  store fp128 %frem, fp128 *%p3
+  ret void
+}
+
+define void @frem_v16f32(<16 x float> %a0, <16 x float> %a1, <16 x float> *%p3) nounwind {
+; CHECK-LABEL: frem_v16f32:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    pushq %rbx
+; CHECK-NEXT:    subq $160, %rsp
+; CHECK-NEXT:    movq %rdi, %rbx
+; CHECK-NEXT:    movaps %xmm7, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps %xmm6, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps %xmm5, (%rsp) # 16-byte Spill
+; CHECK-NEXT:    movaps %xmm4, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps %xmm3, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps %xmm2, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    shufps {{.*#+}} xmm0 = xmm0[3,3,3,3]
+; CHECK-NEXT:    movaps %xmm4, %xmm1
+; CHECK-NEXT:    shufps {{.*#+}} xmm1 = xmm1[3,3],xmm4[3,3]
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    movaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movhlps {{.*#+}} xmm0 = xmm0[1,1]
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    movhlps {{.*#+}} xmm1 = xmm1[1,1]
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    unpcklps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
+; CHECK-NEXT:    # xmm0 = xmm0[0],mem[0],xmm0[1],mem[1]
+; CHECK-NEXT:    movaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    movaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    shufps {{.*#+}} xmm0 = xmm0[1,1,1,1]
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    shufps {{.*#+}} xmm1 = xmm1[1,1,1,1]
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    unpcklps {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1]
+; CHECK-NEXT:    unpcklpd {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Folded Reload
+; CHECK-NEXT:    # xmm1 = xmm1[0],mem[0]
+; CHECK-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    shufps {{.*#+}} xmm0 = xmm0[3,3,3,3]
+; CHECK-NEXT:    movaps (%rsp), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    shufps {{.*#+}} xmm1 = xmm1[3,3,3,3]
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    movaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movhlps {{.*#+}} xmm0 = xmm0[1,1]
+; CHECK-NEXT:    movaps (%rsp), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    movhlps {{.*#+}} xmm1 = xmm1[1,1]
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    unpcklps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
+; CHECK-NEXT:    # xmm0 = xmm0[0],mem[0],xmm0[1],mem[1]
+; CHECK-NEXT:    movaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movaps (%rsp), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    movaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    shufps {{.*#+}} xmm0 = xmm0[1,1,1,1]
+; CHECK-NEXT:    movaps (%rsp), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    shufps {{.*#+}} xmm1 = xmm1[1,1,1,1]
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    unpcklps {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1]
+; CHECK-NEXT:    unpcklpd {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Folded Reload
+; CHECK-NEXT:    # xmm1 = xmm1[0],mem[0]
+; CHECK-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    shufps {{.*#+}} xmm0 = xmm0[3,3,3,3]
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    shufps {{.*#+}} xmm1 = xmm1[3,3,3,3]
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    movaps %xmm0, (%rsp) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movhlps {{.*#+}} xmm0 = xmm0[1,1]
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    movhlps {{.*#+}} xmm1 = xmm1[1,1]
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    unpcklps (%rsp), %xmm0 # 16-byte Folded Reload
+; CHECK-NEXT:    # xmm0 = xmm0[0],mem[0],xmm0[1],mem[1]
+; CHECK-NEXT:    movaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    movaps %xmm0, (%rsp) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    shufps {{.*#+}} xmm0 = xmm0[1,1,1,1]
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    shufps {{.*#+}} xmm1 = xmm1[1,1,1,1]
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    movaps (%rsp), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    unpcklps {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1]
+; CHECK-NEXT:    unpcklpd {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Folded Reload
+; CHECK-NEXT:    # xmm1 = xmm1[0],mem[0]
+; CHECK-NEXT:    movaps %xmm1, (%rsp) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    shufps {{.*#+}} xmm0 = xmm0[3,3,3,3]
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    shufps {{.*#+}} xmm1 = xmm1[3,3,3,3]
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    movaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movhlps {{.*#+}} xmm0 = xmm0[1,1]
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    movhlps {{.*#+}} xmm1 = xmm1[1,1]
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    unpcklps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
+; CHECK-NEXT:    # xmm0 = xmm0[0],mem[0],xmm0[1],mem[1]
+; CHECK-NEXT:    movaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    movaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    shufps {{.*#+}} xmm0 = xmm0[1,1,1,1]
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    shufps {{.*#+}} xmm1 = xmm1[1,1,1,1]
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    unpcklps {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1]
+; CHECK-NEXT:    unpcklpd {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Folded Reload
+; CHECK-NEXT:    # xmm1 = xmm1[0],mem[0]
+; CHECK-NEXT:    movaps %xmm1, 48(%rbx)
+; CHECK-NEXT:    movaps (%rsp), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movaps %xmm0, 32(%rbx)
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movaps %xmm0, 16(%rbx)
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movaps %xmm0, (%rbx)
+; CHECK-NEXT:    addq $160, %rsp
+; CHECK-NEXT:    popq %rbx
+; CHECK-NEXT:    retq
+  %frem = frem <16 x float> %a0, %a1
+  store <16 x float> %frem, <16 x float> *%p3
+  ret void
+}
+
+define void @frem_v8f32(<8 x float> %a0, <8 x float> %a1, <8 x float> *%p3) nounwind {
+; CHECK-LABEL: frem_v8f32:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    pushq %rbx
+; CHECK-NEXT:    subq $96, %rsp
+; CHECK-NEXT:    movq %rdi, %rbx
+; CHECK-NEXT:    movaps %xmm3, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps %xmm2, (%rsp) # 16-byte Spill
+; CHECK-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    shufps {{.*#+}} xmm0 = xmm0[3,3,3,3]
+; CHECK-NEXT:    movaps %xmm2, %xmm1
+; CHECK-NEXT:    shufps {{.*#+}} xmm1 = xmm1[3,3],xmm2[3,3]
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    movaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movhlps {{.*#+}} xmm0 = xmm0[1,1]
+; CHECK-NEXT:    movaps (%rsp), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    movhlps {{.*#+}} xmm1 = xmm1[1,1]
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    unpcklps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
+; CHECK-NEXT:    # xmm0 = xmm0[0],mem[0],xmm0[1],mem[1]
+; CHECK-NEXT:    movaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movaps (%rsp), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    movaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    shufps {{.*#+}} xmm0 = xmm0[1,1,1,1]
+; CHECK-NEXT:    movaps (%rsp), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    shufps {{.*#+}} xmm1 = xmm1[1,1,1,1]
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    unpcklps {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1]
+; CHECK-NEXT:    unpcklpd {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Folded Reload
+; CHECK-NEXT:    # xmm1 = xmm1[0],mem[0]
+; CHECK-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    shufps {{.*#+}} xmm0 = xmm0[3,3,3,3]
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    shufps {{.*#+}} xmm1 = xmm1[3,3,3,3]
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    movaps %xmm0, (%rsp) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movhlps {{.*#+}} xmm0 = xmm0[1,1]
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    movhlps {{.*#+}} xmm1 = xmm1[1,1]
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    unpcklps (%rsp), %xmm0 # 16-byte Folded Reload
+; CHECK-NEXT:    # xmm0 = xmm0[0],mem[0],xmm0[1],mem[1]
+; CHECK-NEXT:    movaps %xmm0, (%rsp) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    movaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    shufps {{.*#+}} xmm0 = xmm0[1,1,1,1]
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    shufps {{.*#+}} xmm1 = xmm1[1,1,1,1]
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    unpcklps {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1]
+; CHECK-NEXT:    unpcklpd (%rsp), %xmm1 # 16-byte Folded Reload
+; CHECK-NEXT:    # xmm1 = xmm1[0],mem[0]
+; CHECK-NEXT:    movaps %xmm1, 16(%rbx)
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movaps %xmm0, (%rbx)
+; CHECK-NEXT:    addq $96, %rsp
+; CHECK-NEXT:    popq %rbx
+; CHECK-NEXT:    retq
+  %frem = frem <8 x float> %a0, %a1
+  store <8 x float> %frem, <8 x float> *%p3
+  ret void
+}
+
+define void @frem_v4f32(<4 x float> %a0, <4 x float> %a1, <4 x float> *%p3) nounwind {
+; CHECK-LABEL: frem_v4f32:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    pushq %rbx
+; CHECK-NEXT:    subq $64, %rsp
+; CHECK-NEXT:    movq %rdi, %rbx
+; CHECK-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    shufps {{.*#+}} xmm0 = xmm0[3,3,3,3]
+; CHECK-NEXT:    shufps {{.*#+}} xmm1 = xmm1[3,3,3,3]
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    movaps %xmm0, (%rsp) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movhlps {{.*#+}} xmm0 = xmm0[1,1]
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    movhlps {{.*#+}} xmm1 = xmm1[1,1]
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    unpcklps (%rsp), %xmm0 # 16-byte Folded Reload
+; CHECK-NEXT:    # xmm0 = xmm0[0],mem[0],xmm0[1],mem[1]
+; CHECK-NEXT:    movaps %xmm0, (%rsp) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    movaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    shufps {{.*#+}} xmm0 = xmm0[1,1,1,1]
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    shufps {{.*#+}} xmm1 = xmm1[1,1,1,1]
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    unpcklps {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1]
+; CHECK-NEXT:    unpcklpd (%rsp), %xmm1 # 16-byte Folded Reload
+; CHECK-NEXT:    # xmm1 = xmm1[0],mem[0]
+; CHECK-NEXT:    movaps %xmm1, (%rbx)
+; CHECK-NEXT:    addq $64, %rsp
+; CHECK-NEXT:    popq %rbx
+; CHECK-NEXT:    retq
+  %frem = frem <4 x float> %a0, %a1
+  store <4 x float> %frem, <4 x float> *%p3
+  ret void
+}
+
+define void @frem_v8f64(<8 x double> %a0, <8 x double> %a1, <8 x double> *%p3) nounwind {
+; CHECK-LABEL: frem_v8f64:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    pushq %rbx
+; CHECK-NEXT:    subq $144, %rsp
+; CHECK-NEXT:    movq %rdi, %rbx
+; CHECK-NEXT:    movaps %xmm7, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps %xmm6, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps %xmm5, (%rsp) # 16-byte Spill
+; CHECK-NEXT:    movaps %xmm4, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps %xmm3, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps %xmm2, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps %xmm4, %xmm1
+; CHECK-NEXT:    callq fmod@PLT
+; CHECK-NEXT:    movaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movhlps {{.*#+}} xmm0 = xmm0[1,1]
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    movhlps {{.*#+}} xmm1 = xmm1[1,1]
+; CHECK-NEXT:    callq fmod@PLT
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    movlhps {{.*#+}} xmm1 = xmm1[0],xmm0[0]
+; CHECK-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movaps (%rsp), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    callq fmod@PLT
+; CHECK-NEXT:    movaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movhlps {{.*#+}} xmm0 = xmm0[1,1]
+; CHECK-NEXT:    movaps (%rsp), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    movhlps {{.*#+}} xmm1 = xmm1[1,1]
+; CHECK-NEXT:    callq fmod@PLT
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    movlhps {{.*#+}} xmm1 = xmm1[0],xmm0[0]
+; CHECK-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    callq fmod@PLT
+; CHECK-NEXT:    movaps %xmm0, (%rsp) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movhlps {{.*#+}} xmm0 = xmm0[1,1]
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    movhlps {{.*#+}} xmm1 = xmm1[1,1]
+; CHECK-NEXT:    callq fmod@PLT
+; CHECK-NEXT:    movaps (%rsp), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    movlhps {{.*#+}} xmm1 = xmm1[0],xmm0[0]
+; CHECK-NEXT:    movaps %xmm1, (%rsp) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    callq fmod@PLT
+; CHECK-NEXT:    movaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movhlps {{.*#+}} xmm0 = xmm0[1,1]
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    movhlps {{.*#+}} xmm1 = xmm1[1,1]
+; CHECK-NEXT:    callq fmod@PLT
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    movlhps {{.*#+}} xmm1 = xmm1[0],xmm0[0]
+; CHECK-NEXT:    movaps %xmm1, 48(%rbx)
+; CHECK-NEXT:    movaps (%rsp), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movaps %xmm0, 32(%rbx)
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movaps %xmm0, 16(%rbx)
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movaps %xmm0, (%rbx)
+; CHECK-NEXT:    addq $144, %rsp
+; CHECK-NEXT:    popq %rbx
+; CHECK-NEXT:    retq
+  %frem = frem <8 x double> %a0, %a1
+  store <8 x double> %frem, <8 x double> *%p3
+  ret void
+}
+
+define void @frem_v4f64(<4 x double> %a0, <4 x double> %a1, <4 x double> *%p3) nounwind {
+; CHECK-LABEL: frem_v4f64:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    pushq %rbx
+; CHECK-NEXT:    subq $80, %rsp
+; CHECK-NEXT:    movq %rdi, %rbx
+; CHECK-NEXT:    movaps %xmm3, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps %xmm2, (%rsp) # 16-byte Spill
+; CHECK-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps %xmm2, %xmm1
+; CHECK-NEXT:    callq fmod@PLT
+; CHECK-NEXT:    movaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movhlps {{.*#+}} xmm0 = xmm0[1,1]
+; CHECK-NEXT:    movaps (%rsp), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    movhlps {{.*#+}} xmm1 = xmm1[1,1]
+; CHECK-NEXT:    callq fmod@PLT
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    movlhps {{.*#+}} xmm1 = xmm1[0],xmm0[0]
+; CHECK-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    callq fmod@PLT
+; CHECK-NEXT:    movaps %xmm0, (%rsp) # 16-byte Spill
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movhlps {{.*#+}} xmm0 = xmm0[1,1]
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    movhlps {{.*#+}} xmm1 = xmm1[1,1]
+; CHECK-NEXT:    callq fmod@PLT
+; CHECK-NEXT:    movaps (%rsp), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    movlhps {{.*#+}} xmm1 = xmm1[0],xmm0[0]
+; CHECK-NEXT:    movaps %xmm1, 16(%rbx)
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movaps %xmm0, (%rbx)
+; CHECK-NEXT:    addq $80, %rsp
+; CHECK-NEXT:    popq %rbx
+; CHECK-NEXT:    retq
+  %frem = frem <4 x double> %a0, %a1
+  store <4 x double> %frem, <4 x double> *%p3
+  ret void
+}
+
+define void @frem_v2f64(<2 x double> %a0, <2 x double> %a1, <2 x double> *%p3) nounwind {
+; CHECK-LABEL: frem_v2f64:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    pushq %rbx
+; CHECK-NEXT:    subq $48, %rsp
+; CHECK-NEXT:    movq %rdi, %rbx
+; CHECK-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps %xmm0, (%rsp) # 16-byte Spill
+; CHECK-NEXT:    callq fmod@PLT
+; CHECK-NEXT:    movaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
+; CHECK-NEXT:    movaps (%rsp), %xmm0 # 16-byte Reload
+; CHECK-NEXT:    movhlps {{.*#+}} xmm0 = xmm0[1,1]
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    movhlps {{.*#+}} xmm1 = xmm1[1,1]
+; CHECK-NEXT:    callq fmod@PLT
+; CHECK-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 16-byte Reload
+; CHECK-NEXT:    movlhps {{.*#+}} xmm1 = xmm1[0],xmm0[0]
+; CHECK-NEXT:    movaps %xmm1, (%rbx)
+; CHECK-NEXT:    addq $48, %rsp
+; CHECK-NEXT:    popq %rbx
+; CHECK-NEXT:    retq
+  %frem = frem <2 x double> %a0, %a1
+  store <2 x double> %frem, <2 x double> *%p3
+  ret void
+}
+
+define void @frem_v32f16(<32 x half> %a0, <32 x half> %a1, <32 x half> *%p3) nounwind {
+; CHECK-LABEL: frem_v32f16:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    pushq %rbp
+; CHECK-NEXT:    pushq %r15
+; CHECK-NEXT:    pushq %r14
+; CHECK-NEXT:    pushq %r13
+; CHECK-NEXT:    pushq %r12
+; CHECK-NEXT:    pushq %rbx
+; CHECK-NEXT:    subq $248, %rsp
+; CHECK-NEXT:    movl %r9d, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl %r8d, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl %ecx, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl %edx, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl %esi, %r14d
+; CHECK-NEXT:    movl %edi, %ebx
+; CHECK-NEXT:    movq {{[0-9]+}}(%rsp), %r15
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %ebp
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %r13d
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %r12d
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %edi
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl %bx, %edi
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl %r12d, %edi
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl %r14w, %edi
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl %r13d, %edi
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 2-byte Folded Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl %ebp, %edi
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 2-byte Folded Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 2-byte Folded Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 2-byte Folded Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movl %eax, %ebp
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movl %eax, %r14d
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movl %eax, %r12d
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movl %eax, %r13d
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movl %eax, %ebx
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, 62(%r15)
+; CHECK-NEXT:    movw %bx, 60(%r15)
+; CHECK-NEXT:    movw %r13w, 58(%r15)
+; CHECK-NEXT:    movw %r12w, 56(%r15)
+; CHECK-NEXT:    movw %r14w, 54(%r15)
+; CHECK-NEXT:    movw %bp, 52(%r15)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 50(%r15)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 48(%r15)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 46(%r15)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 44(%r15)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 42(%r15)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 40(%r15)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 38(%r15)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 36(%r15)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 34(%r15)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 32(%r15)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 30(%r15)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 28(%r15)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 26(%r15)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 24(%r15)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 22(%r15)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 20(%r15)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 18(%r15)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 16(%r15)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 14(%r15)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 12(%r15)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 10(%r15)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 8(%r15)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 6(%r15)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 4(%r15)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 2(%r15)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, (%r15)
+; CHECK-NEXT:    addq $248, %rsp
+; CHECK-NEXT:    popq %rbx
+; CHECK-NEXT:    popq %r12
+; CHECK-NEXT:    popq %r13
+; CHECK-NEXT:    popq %r14
+; CHECK-NEXT:    popq %r15
+; CHECK-NEXT:    popq %rbp
+; CHECK-NEXT:    retq
+  %frem = frem <32 x half> %a0, %a1
+  store <32 x half> %frem, <32 x half> *%p3
+  ret void
+}
+
+define void @frem_v16f16(<16 x half> %a0, <16 x half> %a1, <16 x half> *%p3) nounwind {
+; CHECK-LABEL: frem_v16f16:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    pushq %rbp
+; CHECK-NEXT:    pushq %r15
+; CHECK-NEXT:    pushq %r14
+; CHECK-NEXT:    pushq %r13
+; CHECK-NEXT:    pushq %r12
+; CHECK-NEXT:    pushq %rbx
+; CHECK-NEXT:    subq $120, %rsp
+; CHECK-NEXT:    movl %r9d, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl %r8d, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl %ecx, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl %edx, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl %esi, %r15d
+; CHECK-NEXT:    movl %edi, %r14d
+; CHECK-NEXT:    movq {{[0-9]+}}(%rsp), %rbp
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %r13d
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %r12d
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %ebx
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %edi
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl %r14w, %edi
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl %ebx, %edi
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl %r15w, %edi
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl %r12d, %edi
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 2-byte Folded Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl %r13d, %edi
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 2-byte Folded Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 2-byte Folded Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 2-byte Folded Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movl %eax, %r13d
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movl %eax, %r12d
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movl %eax, %ebx
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movl %eax, %r14d
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movl %eax, %r15d
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, 30(%rbp)
+; CHECK-NEXT:    movw %r15w, 28(%rbp)
+; CHECK-NEXT:    movw %r14w, 26(%rbp)
+; CHECK-NEXT:    movw %bx, 24(%rbp)
+; CHECK-NEXT:    movw %r12w, 22(%rbp)
+; CHECK-NEXT:    movw %r13w, 20(%rbp)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 18(%rbp)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 16(%rbp)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 14(%rbp)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 12(%rbp)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 10(%rbp)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 8(%rbp)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 6(%rbp)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 4(%rbp)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, 2(%rbp)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, (%rbp)
+; CHECK-NEXT:    addq $120, %rsp
+; CHECK-NEXT:    popq %rbx
+; CHECK-NEXT:    popq %r12
+; CHECK-NEXT:    popq %r13
+; CHECK-NEXT:    popq %r14
+; CHECK-NEXT:    popq %r15
+; CHECK-NEXT:    popq %rbp
+; CHECK-NEXT:    retq
+  %frem = frem <16 x half> %a0, %a1
+  store <16 x half> %frem, <16 x half> *%p3
+  ret void
+}
+
+define void @frem_v8f16(<8 x half> %a0, <8 x half> %a1, <8 x half> *%p3) nounwind {
+; CHECK-LABEL: frem_v8f16:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    pushq %rbp
+; CHECK-NEXT:    pushq %r15
+; CHECK-NEXT:    pushq %r14
+; CHECK-NEXT:    pushq %r13
+; CHECK-NEXT:    pushq %r12
+; CHECK-NEXT:    pushq %rbx
+; CHECK-NEXT:    subq $56, %rsp
+; CHECK-NEXT:    movl %r9d, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl %r8d, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl %ecx, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl %edx, %ebx
+; CHECK-NEXT:    movl %esi, %r13d
+; CHECK-NEXT:    movl %edi, %r15d
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %eax
+; CHECK-NEXT:    movl %eax, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %r14d
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %ebp
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %r12d
+; CHECK-NEXT:    movzwl {{[0-9]+}}(%rsp), %edi
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl %r15w, %edi
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movw %ax, {{[-0-9]+}}(%r{{[sb]}}p) # 2-byte Spill
+; CHECK-NEXT:    movl %r12d, %edi
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl %r13w, %edi
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movl %eax, %r12d
+; CHECK-NEXT:    movl %ebp, %edi
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl %bx, %edi
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movl %eax, %r13d
+; CHECK-NEXT:    movl %r14d, %edi
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 2-byte Folded Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movl %eax, %r14d
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 2-byte Folded Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movl %eax, %r15d
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 2-byte Folded Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movl %eax, %ebp
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movl %eax, %ebx
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %edi # 4-byte Reload
+; CHECK-NEXT:    callq __gnu_h2f_ieee@PLT
+; CHECK-NEXT:    movss {{[-0-9]+}}(%r{{[sb]}}p), %xmm1 # 4-byte Reload
+; CHECK-NEXT:    # xmm1 = mem[0],zero,zero,zero
+; CHECK-NEXT:    callq fmodf@PLT
+; CHECK-NEXT:    callq __gnu_f2h_ieee@PLT
+; CHECK-NEXT:    movq {{[0-9]+}}(%rsp), %rcx
+; CHECK-NEXT:    movw %ax, 14(%rcx)
+; CHECK-NEXT:    movw %bx, 12(%rcx)
+; CHECK-NEXT:    movw %bp, 10(%rcx)
+; CHECK-NEXT:    movw %r15w, 8(%rcx)
+; CHECK-NEXT:    movw %r14w, 6(%rcx)
+; CHECK-NEXT:    movw %r13w, 4(%rcx)
+; CHECK-NEXT:    movw %r12w, 2(%rcx)
+; CHECK-NEXT:    movzwl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 2-byte Folded Reload
+; CHECK-NEXT:    movw %ax, (%rcx)
+; CHECK-NEXT:    addq $56, %rsp
+; CHECK-NEXT:    popq %rbx
+; CHECK-NEXT:    popq %r12
+; CHECK-NEXT:    popq %r13
+; CHECK-NEXT:    popq %r14
+; CHECK-NEXT:    popq %r15
+; CHECK-NEXT:    popq %rbp
+; CHECK-NEXT:    retq
+  %frem = frem <8 x half> %a0, %a1
+  store <8 x half> %frem, <8 x half> *%p3
+  ret void
+}
+
+define void @frem_v4f80(<4 x x86_fp80> %a0, <4 x x86_fp80> %a1, <4 x x86_fp80> *%p3) nounwind {
+; CHECK-LABEL: frem_v4f80:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    pushq %rbx
+; CHECK-NEXT:    subq $128, %rsp
+; CHECK-NEXT:    movq %rdi, %rbx
+; CHECK-NEXT:    fldt {{[0-9]+}}(%rsp)
+; CHECK-NEXT:    fstpt {{[-0-9]+}}(%r{{[sb]}}p) # 10-byte Folded Spill
+; CHECK-NEXT:    fldt {{[0-9]+}}(%rsp)
+; CHECK-NEXT:    fstpt {{[-0-9]+}}(%r{{[sb]}}p) # 10-byte Folded Spill
+; CHECK-NEXT:    fldt {{[0-9]+}}(%rsp)
+; CHECK-NEXT:    fstpt {{[-0-9]+}}(%r{{[sb]}}p) # 10-byte Folded Spill
+; CHECK-NEXT:    fldt {{[0-9]+}}(%rsp)
+; CHECK-NEXT:    fstpt {{[-0-9]+}}(%r{{[sb]}}p) # 10-byte Folded Spill
+; CHECK-NEXT:    fldt {{[0-9]+}}(%rsp)
+; CHECK-NEXT:    fstpt {{[-0-9]+}}(%r{{[sb]}}p) # 10-byte Folded Spill
+; CHECK-NEXT:    fldt {{[0-9]+}}(%rsp)
+; CHECK-NEXT:    fstpt {{[-0-9]+}}(%r{{[sb]}}p) # 10-byte Folded Spill
+; CHECK-NEXT:    fldt {{[0-9]+}}(%rsp)
+; CHECK-NEXT:    fldt {{[0-9]+}}(%rsp)
+; CHECK-NEXT:    fstpt {{[0-9]+}}(%rsp)
+; CHECK-NEXT:    fstpt (%rsp)
+; CHECK-NEXT:    callq fmodl@PLT
+; CHECK-NEXT:    fstpt {{[-0-9]+}}(%r{{[sb]}}p) # 10-byte Folded Spill
+; CHECK-NEXT:    fldt {{[-0-9]+}}(%r{{[sb]}}p) # 10-byte Folded Reload
+; CHECK-NEXT:    fstpt {{[0-9]+}}(%rsp)
+; CHECK-NEXT:    fldt {{[-0-9]+}}(%r{{[sb]}}p) # 10-byte Folded Reload
+; CHECK-NEXT:    fstpt (%rsp)
+; CHECK-NEXT:    callq fmodl@PLT
+; CHECK-NEXT:    fstpt {{[-0-9]+}}(%r{{[sb]}}p) # 10-byte Folded Spill
+; CHECK-NEXT:    fldt {{[-0-9]+}}(%r{{[sb]}}p) # 10-byte Folded Reload
+; CHECK-NEXT:    fstpt {{[0-9]+}}(%rsp)
+; CHECK-NEXT:    fldt {{[-0-9]+}}(%r{{[sb]}}p) # 10-byte Folded Reload
+; CHECK-NEXT:    fstpt (%rsp)
+; CHECK-NEXT:    callq fmodl@PLT
+; CHECK-NEXT:    fstpt {{[-0-9]+}}(%r{{[sb]}}p) # 10-byte Folded Spill
+; CHECK-NEXT:    fldt {{[-0-9]+}}(%r{{[sb]}}p) # 10-byte Folded Reload
+; CHECK-NEXT:    fstpt {{[0-9]+}}(%rsp)
+; CHECK-NEXT:    fldt {{[-0-9]+}}(%r{{[sb]}}p) # 10-byte Folded Reload
+; CHECK-NEXT:    fstpt (%rsp)
+; CHECK-NEXT:    callq fmodl@PLT
+; CHECK-NEXT:    fstpt 30(%rbx)
+; CHECK-NEXT:    fldt {{[-0-9]+}}(%r{{[sb]}}p) # 10-byte Folded Reload
+; CHECK-NEXT:    fstpt 20(%rbx)
+; CHECK-NEXT:    fldt {{[-0-9]+}}(%r{{[sb]}}p) # 10-byte Folded Reload
+; CHECK-NEXT:    fstpt 10(%rbx)
+; CHECK-NEXT:    fldt {{[-0-9]+}}(%r{{[sb]}}p) # 10-byte Folded Reload
+; CHECK-NEXT:    fstpt (%rbx)
+; CHECK-NEXT:    addq $128, %rsp
+; CHECK-NEXT:    popq %rbx
+; CHECK-NEXT:    retq
+  %frem = frem <4 x x86_fp80> %a0, %a1
+  store <4 x x86_fp80> %frem, <4 x x86_fp80> *%p3
+  ret void
+}