Skip to content

[LegalizeTypes][X86] Improve ExpandIntRes_FP_TO_SINT/ExpandIntRes_FP_… #3215

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Aug 31, 2021
Merged

[LegalizeTypes][X86] Improve ExpandIntRes_FP_TO_SINT/ExpandIntRes_FP_… #3215

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 6 additions & 0 deletions llvm/lib/CodeGen/SelectionDAG/LegalizeIntegerTypes.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3087,6 +3087,9 @@ void DAGTypeLegalizer::ExpandIntRes_FP_TO_SINT(SDNode *N, SDValue &Lo,
EVT NFPVT = TLI.getTypeToTransformTo(*DAG.getContext(), Op.getValueType());
Op = GetSoftPromotedHalf(Op);
Op = DAG.getNode(ISD::FP16_TO_FP, dl, NFPVT, Op);
Op = DAG.getNode(ISD::FP_TO_SINT, dl, VT, Op);
SplitInteger(Op, Lo, Hi);
return;
}

RTLIB::Libcall LC = RTLIB::getFPTOSINT(Op.getValueType(), VT);
Expand Down Expand Up @@ -3116,6 +3119,9 @@ void DAGTypeLegalizer::ExpandIntRes_FP_TO_UINT(SDNode *N, SDValue &Lo,
EVT NFPVT = TLI.getTypeToTransformTo(*DAG.getContext(), Op.getValueType());
Op = GetSoftPromotedHalf(Op);
Op = DAG.getNode(ISD::FP16_TO_FP, dl, NFPVT, Op);
Op = DAG.getNode(ISD::FP_TO_UINT, dl, VT, Op);
SplitInteger(Op, Lo, Hi);
return;
}

RTLIB::Libcall LC = RTLIB::getFPTOUINT(Op.getValueType(), VT);
Expand Down
46 changes: 38 additions & 8 deletions llvm/test/CodeGen/X86/half.ll
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -210,14 +210,23 @@ define i64 @test_fptosi_i64(half* %p) #0 {
;
; CHECK-I686-LABEL: test_fptosi_i64:
; CHECK-I686: # %bb.0:
; CHECK-I686-NEXT: subl $12, %esp
; CHECK-I686-NEXT: subl $28, %esp
; CHECK-I686-NEXT: movl {{[0-9]+}}(%esp), %eax
; CHECK-I686-NEXT: movzwl (%eax), %eax
; CHECK-I686-NEXT: movl %eax, (%esp)
; CHECK-I686-NEXT: calll __gnu_h2f_ieee
; CHECK-I686-NEXT: fstps (%esp)
; CHECK-I686-NEXT: calll __fixsfdi
; CHECK-I686-NEXT: addl $12, %esp
; CHECK-I686-NEXT: fstps {{[0-9]+}}(%esp)
; CHECK-I686-NEXT: flds {{[0-9]+}}(%esp)
; CHECK-I686-NEXT: fnstcw {{[0-9]+}}(%esp)
; CHECK-I686-NEXT: movzwl {{[0-9]+}}(%esp), %eax
; CHECK-I686-NEXT: orl $3072, %eax # imm = 0xC00
; CHECK-I686-NEXT: movw %ax, {{[0-9]+}}(%esp)
; CHECK-I686-NEXT: fldcw {{[0-9]+}}(%esp)
; CHECK-I686-NEXT: fistpll {{[0-9]+}}(%esp)
; CHECK-I686-NEXT: fldcw {{[0-9]+}}(%esp)
; CHECK-I686-NEXT: movl {{[0-9]+}}(%esp), %eax
; CHECK-I686-NEXT: movl {{[0-9]+}}(%esp), %edx
; CHECK-I686-NEXT: addl $28, %esp
; CHECK-I686-NEXT: retl
%a = load half, half* %p, align 2
%r = fptosi half %a to i64
Expand Down Expand Up @@ -295,14 +304,35 @@ define i64 @test_fptoui_i64(half* %p) #0 {
;
; CHECK-I686-LABEL: test_fptoui_i64:
; CHECK-I686: # %bb.0:
; CHECK-I686-NEXT: subl $12, %esp
; CHECK-I686-NEXT: subl $28, %esp
; CHECK-I686-NEXT: movl {{[0-9]+}}(%esp), %eax
; CHECK-I686-NEXT: movzwl (%eax), %eax
; CHECK-I686-NEXT: movl %eax, (%esp)
; CHECK-I686-NEXT: calll __gnu_h2f_ieee
; CHECK-I686-NEXT: fstps (%esp)
; CHECK-I686-NEXT: calll __fixunssfdi
; CHECK-I686-NEXT: addl $12, %esp
; CHECK-I686-NEXT: fstps {{[0-9]+}}(%esp)
; CHECK-I686-NEXT: movss {{.*#+}} xmm0 = mem[0],zero,zero,zero
; CHECK-I686-NEXT: movss {{.*#+}} xmm1 = mem[0],zero,zero,zero
; CHECK-I686-NEXT: ucomiss %xmm1, %xmm0
; CHECK-I686-NEXT: jae .LBB9_2
; CHECK-I686-NEXT: # %bb.1:
; CHECK-I686-NEXT: xorps %xmm1, %xmm1
; CHECK-I686-NEXT: .LBB9_2:
; CHECK-I686-NEXT: subss %xmm1, %xmm0
; CHECK-I686-NEXT: movss %xmm0, {{[0-9]+}}(%esp)
; CHECK-I686-NEXT: setae %al
; CHECK-I686-NEXT: flds {{[0-9]+}}(%esp)
; CHECK-I686-NEXT: fnstcw {{[0-9]+}}(%esp)
; CHECK-I686-NEXT: movzwl {{[0-9]+}}(%esp), %ecx
; CHECK-I686-NEXT: orl $3072, %ecx # imm = 0xC00
; CHECK-I686-NEXT: movw %cx, {{[0-9]+}}(%esp)
; CHECK-I686-NEXT: fldcw {{[0-9]+}}(%esp)
; CHECK-I686-NEXT: fistpll {{[0-9]+}}(%esp)
; CHECK-I686-NEXT: fldcw {{[0-9]+}}(%esp)
; CHECK-I686-NEXT: movzbl %al, %edx
; CHECK-I686-NEXT: shll $31, %edx
; CHECK-I686-NEXT: xorl {{[0-9]+}}(%esp), %edx
; CHECK-I686-NEXT: movl {{[0-9]+}}(%esp), %eax
; CHECK-I686-NEXT: addl $28, %esp
; CHECK-I686-NEXT: retl
%a = load half, half* %p, align 2
%r = fptoui half %a to i64
Expand Down