[llvm] r357549 - InstSimplify: Fold round intrinsics from sitofp/uitofp
Matt Arsenault via llvm-commits
llvm-commits at lists.llvm.org
Tue Apr 2 17:25:06 PDT 2019
Author: arsenm
Date: Tue Apr 2 17:25:06 2019
New Revision: 357549
URL: http://llvm.org/viewvc/llvm-project?rev=357549&view=rev
Log:
InstSimplify: Fold round intrinsics from sitofp/uitofp
https://godbolt.org/z/gEMRZb
Modified:
llvm/trunk/lib/Analysis/InstructionSimplify.cpp
llvm/trunk/test/Transforms/InstSimplify/round-intrinsics.ll
Modified: llvm/trunk/lib/Analysis/InstructionSimplify.cpp
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Analysis/InstructionSimplify.cpp?rev=357549&r1=357548&r2=357549&view=diff
==============================================================================
--- llvm/trunk/lib/Analysis/InstructionSimplify.cpp (original)
+++ llvm/trunk/lib/Analysis/InstructionSimplify.cpp Tue Apr 2 17:25:06 2019
@@ -4701,6 +4701,22 @@ static Value *simplifyUnaryIntrinsic(Fun
match(Op0, m_Intrinsic<Intrinsic::pow>(m_SpecificFP(10.0),
m_Value(X)))) return X;
break;
+ case Intrinsic::floor:
+ case Intrinsic::trunc:
+ case Intrinsic::ceil:
+ case Intrinsic::round:
+ case Intrinsic::nearbyint:
+ case Intrinsic::rint: {
+ // floor (sitofp x) -> sitofp x
+ // floor (uitofp x) -> uitofp x
+ //
+ // Converting from int always results in a finite integral number or
+ // infinity. For either of those inputs, these rounding functions always
+ // return the same value, so the rounding can be eliminated.
+ if (match(Op0, m_SIToFP(m_Value())) || match(Op0, m_UIToFP(m_Value())))
+ return Op0;
+ break;
+ }
default:
break;
}
Modified: llvm/trunk/test/Transforms/InstSimplify/round-intrinsics.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/Transforms/InstSimplify/round-intrinsics.ll?rev=357549&r1=357548&r2=357549&view=diff
==============================================================================
--- llvm/trunk/test/Transforms/InstSimplify/round-intrinsics.ll (original)
+++ llvm/trunk/test/Transforms/InstSimplify/round-intrinsics.ll Tue Apr 2 17:25:06 2019
@@ -4,8 +4,7 @@
define float @sitofp_floor(i32 %arg) {
; CHECK-LABEL: @sitofp_floor(
; CHECK-NEXT: [[CVT:%.*]] = sitofp i32 [[ARG:%.*]] to float
-; CHECK-NEXT: [[ROUND:%.*]] = call float @llvm.floor.f32(float [[CVT]])
-; CHECK-NEXT: ret float [[ROUND]]
+; CHECK-NEXT: ret float [[CVT]]
;
%cvt = sitofp i32 %arg to float
%round = call float @llvm.floor.f32(float %cvt)
@@ -15,8 +14,7 @@ define float @sitofp_floor(i32 %arg) {
define float @uitofp_floor(i32 %arg) {
; CHECK-LABEL: @uitofp_floor(
; CHECK-NEXT: [[CVT:%.*]] = uitofp i32 [[ARG:%.*]] to float
-; CHECK-NEXT: [[ROUND:%.*]] = call float @llvm.floor.f32(float [[CVT]])
-; CHECK-NEXT: ret float [[ROUND]]
+; CHECK-NEXT: ret float [[CVT]]
;
%cvt = uitofp i32 %arg to float
%round = call float @llvm.floor.f32(float %cvt)
@@ -26,8 +24,7 @@ define float @uitofp_floor(i32 %arg) {
define float @sitofp_trunc(i32 %arg) {
; CHECK-LABEL: @sitofp_trunc(
; CHECK-NEXT: [[CVT:%.*]] = sitofp i32 [[ARG:%.*]] to float
-; CHECK-NEXT: [[ROUND:%.*]] = call float @llvm.trunc.f32(float [[CVT]])
-; CHECK-NEXT: ret float [[ROUND]]
+; CHECK-NEXT: ret float [[CVT]]
;
%cvt = sitofp i32 %arg to float
%round = call float @llvm.trunc.f32(float %cvt)
@@ -37,8 +34,7 @@ define float @sitofp_trunc(i32 %arg) {
define float @uitofp_trunc(i32 %arg) {
; CHECK-LABEL: @uitofp_trunc(
; CHECK-NEXT: [[CVT:%.*]] = uitofp i32 [[ARG:%.*]] to float
-; CHECK-NEXT: [[ROUND:%.*]] = call float @llvm.trunc.f32(float [[CVT]])
-; CHECK-NEXT: ret float [[ROUND]]
+; CHECK-NEXT: ret float [[CVT]]
;
%cvt = uitofp i32 %arg to float
%round = call float @llvm.trunc.f32(float %cvt)
@@ -48,8 +44,7 @@ define float @uitofp_trunc(i32 %arg) {
define float @sitofp_ceil(i32 %arg) {
; CHECK-LABEL: @sitofp_ceil(
; CHECK-NEXT: [[CVT:%.*]] = sitofp i32 [[ARG:%.*]] to float
-; CHECK-NEXT: [[ROUND:%.*]] = call float @llvm.ceil.f32(float [[CVT]])
-; CHECK-NEXT: ret float [[ROUND]]
+; CHECK-NEXT: ret float [[CVT]]
;
%cvt = sitofp i32 %arg to float
%round = call float @llvm.ceil.f32(float %cvt)
@@ -59,8 +54,7 @@ define float @sitofp_ceil(i32 %arg) {
define float @uitofp_ceil(i32 %arg) {
; CHECK-LABEL: @uitofp_ceil(
; CHECK-NEXT: [[CVT:%.*]] = uitofp i32 [[ARG:%.*]] to float
-; CHECK-NEXT: [[ROUND:%.*]] = call float @llvm.ceil.f32(float [[CVT]])
-; CHECK-NEXT: ret float [[ROUND]]
+; CHECK-NEXT: ret float [[CVT]]
;
%cvt = uitofp i32 %arg to float
%round = call float @llvm.ceil.f32(float %cvt)
@@ -70,8 +64,7 @@ define float @uitofp_ceil(i32 %arg) {
define float @sitofp_round(i32 %arg) {
; CHECK-LABEL: @sitofp_round(
; CHECK-NEXT: [[CVT:%.*]] = sitofp i32 [[ARG:%.*]] to float
-; CHECK-NEXT: [[ROUND:%.*]] = call float @llvm.round.f32(float [[CVT]])
-; CHECK-NEXT: ret float [[ROUND]]
+; CHECK-NEXT: ret float [[CVT]]
;
%cvt = sitofp i32 %arg to float
%round = call float @llvm.round.f32(float %cvt)
@@ -81,8 +74,7 @@ define float @sitofp_round(i32 %arg) {
define float @uitofp_round(i32 %arg) {
; CHECK-LABEL: @uitofp_round(
; CHECK-NEXT: [[CVT:%.*]] = uitofp i32 [[ARG:%.*]] to float
-; CHECK-NEXT: [[ROUND:%.*]] = call float @llvm.round.f32(float [[CVT]])
-; CHECK-NEXT: ret float [[ROUND]]
+; CHECK-NEXT: ret float [[CVT]]
;
%cvt = uitofp i32 %arg to float
%round = call float @llvm.round.f32(float %cvt)
@@ -92,8 +84,7 @@ define float @uitofp_round(i32 %arg) {
define float @sitofp_nearbyint(i32 %arg) {
; CHECK-LABEL: @sitofp_nearbyint(
; CHECK-NEXT: [[CVT:%.*]] = sitofp i32 [[ARG:%.*]] to float
-; CHECK-NEXT: [[NEARBYINT:%.*]] = call float @llvm.nearbyint.f32(float [[CVT]])
-; CHECK-NEXT: ret float [[NEARBYINT]]
+; CHECK-NEXT: ret float [[CVT]]
;
%cvt = sitofp i32 %arg to float
%nearbyint = call float @llvm.nearbyint.f32(float %cvt)
@@ -103,8 +94,7 @@ define float @sitofp_nearbyint(i32 %arg)
define float @uitofp_nearbyint(i32 %arg) {
; CHECK-LABEL: @uitofp_nearbyint(
; CHECK-NEXT: [[CVT:%.*]] = uitofp i32 [[ARG:%.*]] to float
-; CHECK-NEXT: [[NEARBYINT:%.*]] = call float @llvm.nearbyint.f32(float [[CVT]])
-; CHECK-NEXT: ret float [[NEARBYINT]]
+; CHECK-NEXT: ret float [[CVT]]
;
%cvt = uitofp i32 %arg to float
%nearbyint = call float @llvm.nearbyint.f32(float %cvt)
@@ -114,8 +104,7 @@ define float @uitofp_nearbyint(i32 %arg)
define float @sitofp_rint(i32 %arg) {
; CHECK-LABEL: @sitofp_rint(
; CHECK-NEXT: [[CVT:%.*]] = sitofp i32 [[ARG:%.*]] to float
-; CHECK-NEXT: [[RINT:%.*]] = call float @llvm.rint.f32(float [[CVT]])
-; CHECK-NEXT: ret float [[RINT]]
+; CHECK-NEXT: ret float [[CVT]]
;
%cvt = sitofp i32 %arg to float
%rint = call float @llvm.rint.f32(float %cvt)
@@ -125,8 +114,7 @@ define float @sitofp_rint(i32 %arg) {
define float @uitofp_rint(i32 %arg) {
; CHECK-LABEL: @uitofp_rint(
; CHECK-NEXT: [[CVT:%.*]] = uitofp i32 [[ARG:%.*]] to float
-; CHECK-NEXT: [[RINT:%.*]] = call float @llvm.rint.f32(float [[CVT]])
-; CHECK-NEXT: ret float [[RINT]]
+; CHECK-NEXT: ret float [[CVT]]
;
%cvt = uitofp i32 %arg to float
%rint = call float @llvm.rint.f32(float %cvt)
More information about the llvm-commits
mailing list