[llvm] 9eb399b - [InstCombine] Support zext nneg in `foldLogicCastConstant` (#82355)
via llvm-commits
llvm-commits at lists.llvm.org
Tue Feb 20 07:09:04 PST 2024
Author: Yingwei Zheng
Date: 2024-02-20T23:09:00+08:00
New Revision: 9eb399b8548b835547947884b75c5fa79f977394
URL: https://github.com/llvm/llvm-project/commit/9eb399b8548b835547947884b75c5fa79f977394
DIFF: https://github.com/llvm/llvm-project/commit/9eb399b8548b835547947884b75c5fa79f977394.diff
LOG: [InstCombine] Support zext nneg in `foldLogicCastConstant` (#82355)
This patch extends [D36234](https://reviews.llvm.org/D36234) to handle
`zext nneg` instructions.
I found this while adding support for cast instructions in
`getFreelyInvertedImpl`.
Added:
Modified:
llvm/lib/Transforms/InstCombine/InstCombineAndOrXor.cpp
llvm/test/Transforms/InstCombine/cast.ll
Removed:
################################################################################
diff --git a/llvm/lib/Transforms/InstCombine/InstCombineAndOrXor.cpp b/llvm/lib/Transforms/InstCombine/InstCombineAndOrXor.cpp
index ba4649e50402af..bf8794bba8859f 100644
--- a/llvm/lib/Transforms/InstCombine/InstCombineAndOrXor.cpp
+++ b/llvm/lib/Transforms/InstCombine/InstCombineAndOrXor.cpp
@@ -1685,7 +1685,7 @@ static Instruction *foldLogicCastConstant(BinaryOperator &Logic, CastInst *Cast,
}
}
- if (match(Cast, m_OneUse(m_SExt(m_Value(X))))) {
+ if (match(Cast, m_OneUse(m_SExtLike(m_Value(X))))) {
if (Constant *TruncC = IC.getLosslessSignedTrunc(C, SrcTy)) {
// LogicOpc (sext X), C --> sext (LogicOpc X, C)
Value *NewOp = IC.Builder.CreateBinOp(LogicOpc, X, TruncC);
diff --git a/llvm/test/Transforms/InstCombine/cast.ll b/llvm/test/Transforms/InstCombine/cast.ll
index 1cda0e503ee939..85433a99f2caec 100644
--- a/llvm/test/Transforms/InstCombine/cast.ll
+++ b/llvm/test/Transforms/InstCombine/cast.ll
@@ -748,8 +748,8 @@ define i32 @test54(i64 %A) {
define i64 @test55(i32 %A) {
; ALL-LABEL: @test55(
; ALL-NEXT: [[TMP1:%.*]] = and i32 [[A:%.*]], 7224
-; ALL-NEXT: [[C:%.*]] = zext nneg i32 [[TMP1]] to i64
-; ALL-NEXT: [[D:%.*]] = or disjoint i64 [[C]], -32574
+; ALL-NEXT: [[TMP2:%.*]] = or disjoint i32 [[TMP1]], -32574
+; ALL-NEXT: [[D:%.*]] = sext i32 [[TMP2]] to i64
; ALL-NEXT: ret i64 [[D]]
;
%B = trunc i32 %A to i16
More information about the llvm-commits
mailing list