RPCS3/llvm-mirror
1
0
Fork 0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2025-01-31 12:41:49 +01:00
Code Issues Projects Releases Wiki Activity

[InstSimplify] loosen FMF for sqrt(X) * sqrt(X) --> X

Browse Source 
As shown in the code comment, we don't need all of 'fast', 
but we do need reassoc + nsz + nnan.

Differential Revision: https://reviews.llvm.org/D43765

llvm-svn: 327796
This commit is contained in:
Sanjay Patel 2018-03-18 14:12:25 +00:00
parent 83f069b61f
commit 83d39ccfbf
2 changed files with 43 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
lib/Analysis/InstructionSimplify.cpp
View File

@ -4249,10 +4249,13 @@ static Value *SimplifyFMulInst(Value *Op0, Value *Op1, FastMathFlags FMF,
if (FMF.noNaNs() && FMF.noSignedZeros() && match(Op1, m_AnyZeroFP()))
return ConstantFP::getNullValue(Op0->getType());
// sqrt(X) * sqrt(X) --> X
// sqrt(X) * sqrt(X) --> X, if we can:
// 1. Remove the intermediate rounding (reassociate).
// 2. Ignore non-zero negative numbers because sqrt would produce NAN.
// 3. Ignore -0.0 because sqrt(-0.0) == -0.0, but -0.0 * -0.0 == 0.0.
Value *X;
if (FMF.isFast() && Op0 == Op1 &&
match(Op0, m_Intrinsic<Intrinsic::sqrt>(m_Value(X))))
if (Op0 == Op1 && match(Op0, m_Intrinsic<Intrinsic::sqrt>(m_Value(X))) &&
FMF.allowReassoc() && FMF.noNaNs() && FMF.noSignedZeros())
return X;
return nullptr;

39
test/Transforms/InstSimplify/fast-math.ll
View File

@ -323,7 +323,7 @@ define float @fdiv_neg_swapped2(float %f) {
}
; PR21126: http://llvm.org/bugs/show_bug.cgi?id=21126
; With unsafe/fast math, sqrt(X) * sqrt(X) is just X.
; With loose math, sqrt(X) * sqrt(X) is just X.
declare double @llvm.sqrt.f64(double)
@ -332,7 +332,42 @@ define double @sqrt_squared(double %f) {
; CHECK-NEXT: ret double [[F:%.*]]
;
%sqrt = call double @llvm.sqrt.f64(double %f)
%mul = fmul fast double %sqrt, %sqrt
%mul = fmul reassoc nnan nsz double %sqrt, %sqrt
ret double %mul
}
; Negative tests for the above transform: we need all 3 of those flags.
define double @sqrt_squared_not_fast_enough1(double %f) {
; CHECK-LABEL: @sqrt_squared_not_fast_enough1(
; CHECK-NEXT: [[SQRT:%.*]] = call double @llvm.sqrt.f64(double [[F:%.*]])
; CHECK-NEXT: [[MUL:%.*]] = fmul nnan nsz double [[SQRT]], [[SQRT]]
; CHECK-NEXT: ret double [[MUL]]
;
%sqrt = call double @llvm.sqrt.f64(double %f)
%mul = fmul nnan nsz double %sqrt, %sqrt
ret double %mul
}
define double @sqrt_squared_not_fast_enough2(double %f) {
; CHECK-LABEL: @sqrt_squared_not_fast_enough2(
; CHECK-NEXT: [[SQRT:%.*]] = call double @llvm.sqrt.f64(double [[F:%.*]])
; CHECK-NEXT: [[MUL:%.*]] = fmul reassoc nnan double [[SQRT]], [[SQRT]]
; CHECK-NEXT: ret double [[MUL]]
;
%sqrt = call double @llvm.sqrt.f64(double %f)
%mul = fmul reassoc nnan double %sqrt, %sqrt
ret double %mul
}
define double @sqrt_squared_not_fast_enough3(double %f) {
; CHECK-LABEL: @sqrt_squared_not_fast_enough3(
; CHECK-NEXT: [[SQRT:%.*]] = call double @llvm.sqrt.f64(double [[F:%.*]])
; CHECK-NEXT: [[MUL:%.*]] = fmul reassoc nsz double [[SQRT]], [[SQRT]]
; CHECK-NEXT: ret double [[MUL]]
;
%sqrt = call double @llvm.sqrt.f64(double %f)
%mul = fmul reassoc nsz double %sqrt, %sqrt
ret double %mul
}