[SCEV] Disallow simplifying phi(undef, X) to X #115109
Merged
+106
−37
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
@llvm/pr-subscribers-llvm-analysis @llvm/pr-subscribers-llvm-transforms Author: Yingwei Zheng (dtcxzyw) ChangesSee the following case:
This patch disallows simplifying Full diff: https://github.com/llvm/llvm-project/pull/115109.diff 7 Files Affected:
diff --git a/llvm/lib/Analysis/ScalarEvolution.cpp b/llvm/lib/Analysis/ScalarEvolution.cpp
index 58e23e9556f144..b10811133770e1 100644
--- a/llvm/lib/Analysis/ScalarEvolution.cpp
+++ b/llvm/lib/Analysis/ScalarEvolution.cpp
@@ -6023,7 +6023,11 @@ const SCEV *ScalarEvolution::createNodeForPHI(PHINode *PN) {
if (const SCEV *S = createAddRecFromPHI(PN))
return S;
- if (Value *V = simplifyInstruction(PN, {getDataLayout(), &TLI, &DT, &AC}))
+ // We do not allow simplifying phi (undef, X) to X here, to avoid reusing the
+ // phi node for X.
+ if (Value *V = simplifyInstruction(
+ PN, {getDataLayout(), &TLI, &DT, &AC, /*CtxI=*/nullptr,
+ /*UseInstrInfo=*/true, /*CanUseUndef=*/false}))
return getSCEV(V);
if (const SCEV *S = createNodeFromSelectLikePHI(PN))
diff --git a/llvm/test/Transforms/IndVarSimplify/invalidate-modified-lcssa-phi.ll b/llvm/test/Transforms/IndVarSimplify/invalidate-modified-lcssa-phi.ll
index 856fc376204995..0538c1c64de34d 100644
--- a/llvm/test/Transforms/IndVarSimplify/invalidate-modified-lcssa-phi.ll
+++ b/llvm/test/Transforms/IndVarSimplify/invalidate-modified-lcssa-phi.ll
@@ -48,13 +48,14 @@ define i8 @test_pr52023(i1 %c.1, i1 %c.2) {
; CHECK-NEXT: br label [[LOOP_1:%.*]]
; CHECK: loop.1:
; CHECK-NEXT: [[INC79:%.*]] = phi i8 [ [[TMP0:%.*]], [[LOOP_1_LATCH:%.*]] ], [ 0, [[ENTRY:%.*]] ]
-; CHECK-NEXT: [[TMP0]] = add i8 [[INC79]], 1
+; CHECK-NEXT: [[TMP1:%.*]] = add i8 [[INC79]], 1
; CHECK-NEXT: br label [[LOOP_2:%.*]]
; CHECK: loop.2:
; CHECK-NEXT: br i1 [[C_1:%.*]], label [[LOOP_2_LATCH:%.*]], label [[LOOP_1_LATCH]]
; CHECK: loop.2.latch:
; CHECK-NEXT: br label [[LOOP_1_LATCH]]
; CHECK: loop.1.latch:
+; CHECK-NEXT: [[TMP0]] = phi i8 [ [[TMP1]], [[LOOP_2_LATCH]] ], [ undef, [[LOOP_2]] ]
; CHECK-NEXT: br i1 [[C_2:%.*]], label [[EXIT:%.*]], label [[LOOP_1]]
; CHECK: exit:
; CHECK-NEXT: [[INC_LCSSA_LCSSA:%.*]] = phi i8 [ [[TMP0]], [[LOOP_1_LATCH]] ]
diff --git a/llvm/test/Transforms/IndVarSimplify/no-iv-rewrite.ll b/llvm/test/Transforms/IndVarSimplify/no-iv-rewrite.ll
index 579b8536cedf03..5339468fe8036b 100644
--- a/llvm/test/Transforms/IndVarSimplify/no-iv-rewrite.ll
+++ b/llvm/test/Transforms/IndVarSimplify/no-iv-rewrite.ll
@@ -260,11 +260,13 @@ define void @identityphi(i32 %limit) nounwind {
; CHECK-NEXT: entry:
; CHECK-NEXT: br label [[LOOP:%.*]]
; CHECK: loop:
-; CHECK-NEXT: br i1 undef, label [[IF_THEN:%.*]], label [[CONTROL:%.*]]
+; CHECK-NEXT: [[IV:%.*]] = phi i32 [ 0, [[ENTRY:%.*]] ], [ [[IV_NEXT:%.*]], [[CONTROL:%.*]] ]
+; CHECK-NEXT: br i1 undef, label [[IF_THEN:%.*]], label [[CONTROL]]
; CHECK: if.then:
; CHECK-NEXT: br label [[CONTROL]]
; CHECK: control:
-; CHECK-NEXT: [[CMP:%.*]] = icmp slt i32 0, [[LIMIT:%.*]]
+; CHECK-NEXT: [[IV_NEXT]] = phi i32 [ [[IV]], [[LOOP]] ], [ undef, [[IF_THEN]] ]
+; CHECK-NEXT: [[CMP:%.*]] = icmp slt i32 [[IV_NEXT]], [[LIMIT:%.*]]
; CHECK-NEXT: br i1 [[CMP]], label [[LOOP]], label [[EXIT:%.*]]
; CHECK: exit:
; CHECK-NEXT: ret void
diff --git a/llvm/test/Transforms/LoopStrengthReduce/X86/2012-01-13-phielim.ll b/llvm/test/Transforms/LoopStrengthReduce/X86/2012-01-13-phielim.ll
index 492746615d8461..38f18f68e2db72 100644
--- a/llvm/test/Transforms/LoopStrengthReduce/X86/2012-01-13-phielim.ll
+++ b/llvm/test/Transforms/LoopStrengthReduce/X86/2012-01-13-phielim.ll
@@ -156,29 +156,29 @@ define fastcc void @test3(ptr nocapture %u) nounwind uwtable ssp {
; CHECK: for.inc8.us.i:
; CHECK-NEXT: br i1 true, label [[MESHBB1_LOOPEXIT:%.*]], label [[MESHBB:%.*]]
; CHECK: for.body3.us.i:
-; CHECK-NEXT: [[INDVARS_IV_I_SV_PHI:%.*]] = phi i64 [ [[INDVARS_IV_NEXT_I:%.*]], [[MESHBB]] ], [ 0, [[FOR_BODY3_LR_PH_US_I:%.*]] ]
+; CHECK-NEXT: [[TMP:%.*]] = phi i32 [ [[LSR_IV_NEXT:%.*]], [[MESHBB]] ], [ [[TMP3:%.*]], [[FOR_BODY3_LR_PH_US_I:%.*]] ]
+; CHECK-NEXT: [[SCEVGEP:%.*]] = phi ptr [ [[SCEVGEP1:%.*]], [[MESHBB]] ], [ [[U:%.*]], [[FOR_BODY3_LR_PH_US_I]] ]
; CHECK-NEXT: [[OPQ_SA_CALC12:%.*]] = sub i32 undef, 227
-; CHECK-NEXT: [[TMP0:%.*]] = add i64 [[LSR_IV:%.*]], [[INDVARS_IV_I_SV_PHI]]
-; CHECK-NEXT: [[TMP:%.*]] = trunc i64 [[TMP0]] to i32
; CHECK-NEXT: [[MUL_I_US_I:%.*]] = mul nsw i32 0, [[TMP]]
-; CHECK-NEXT: [[TMP1:%.*]] = shl nuw nsw i64 [[INDVARS_IV_I_SV_PHI]], 3
-; CHECK-NEXT: [[SCEVGEP:%.*]] = getelementptr i8, ptr [[U:%.*]], i64 [[TMP1]]
; CHECK-NEXT: [[TMP2:%.*]] = load double, ptr [[SCEVGEP]], align 8
; CHECK-NEXT: br i1 undef, label [[FOR_INC8_US_I:%.*]], label [[MESHBB]]
; CHECK: for.body3.lr.ph.us.i.loopexit:
-; CHECK-NEXT: [[LSR_IV_NEXT:%.*]] = add i64 [[LSR_IV]], 1
; CHECK-NEXT: br label [[FOR_BODY3_LR_PH_US_I]]
; CHECK: for.body3.lr.ph.us.i:
-; CHECK-NEXT: [[LSR_IV]] = phi i64 [ [[LSR_IV_NEXT]], [[FOR_BODY3_LR_PH_US_I_LOOPEXIT:%.*]] ], [ undef, [[MESHBB1]] ]
+; CHECK-NEXT: [[LSR_IV:%.*]] = phi i64 [ undef, [[MESHBB1]] ], [ [[INDVARS_IV8_I_SV_PHI24:%.*]], [[FOR_BODY3_LR_PH_US_I_LOOPEXIT:%.*]] ]
; CHECK-NEXT: [[ARRAYIDX_US_I:%.*]] = getelementptr inbounds double, ptr undef, i64 [[LSR_IV]]
+; CHECK-NEXT: [[TMP1:%.*]] = add i64 [[LSR_IV]], 1
+; CHECK-NEXT: [[TMP3]] = trunc i64 [[LSR_IV]] to i32
; CHECK-NEXT: br label [[FOR_BODY3_US_I:%.*]]
; CHECK: for.inc8.us.i2:
; CHECK-NEXT: unreachable
; CHECK: eval_At_times_u.exit:
; CHECK-NEXT: ret void
; CHECK: meshBB:
+; CHECK-NEXT: [[INDVARS_IV8_I_SV_PHI24]] = phi i64 [ undef, [[FOR_BODY3_US_I]] ], [ [[TMP1]], [[FOR_INC8_US_I]] ]
; CHECK-NEXT: [[MESHSTACKVARIABLE_PHI:%.*]] = phi i32 [ [[OPQ_SA_CALC12]], [[FOR_BODY3_US_I]] ], [ undef, [[FOR_INC8_US_I]] ]
-; CHECK-NEXT: [[INDVARS_IV_NEXT_I]] = add i64 [[INDVARS_IV_I_SV_PHI]], 1
+; CHECK-NEXT: [[SCEVGEP1]] = getelementptr i8, ptr [[SCEVGEP]], i64 8
+; CHECK-NEXT: [[LSR_IV_NEXT]] = add i32 [[TMP]], 1
; CHECK-NEXT: br i1 true, label [[FOR_BODY3_LR_PH_US_I_LOOPEXIT]], label [[FOR_BODY3_US_I]]
; CHECK: meshBB1.loopexit:
; CHECK-NEXT: br label [[MESHBB1]]
diff --git a/llvm/test/Transforms/LoopUnroll/pr114879.ll b/llvm/test/Transforms/LoopUnroll/pr114879.ll
new file mode 100644
index 00000000000000..4cd91a0f6c09f7
--- /dev/null
+++ b/llvm/test/Transforms/LoopUnroll/pr114879.ll
@@ -0,0 +1,62 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 5
+; RUN: opt -S -passes=loop-unroll-full -unroll-full-max-count=1 %s | FileCheck %s
+
+@GlobIntONE = global i32 0, align 4
+
+; Make sure we don't reuse the phi (undef, X) for X.
+
+define ptr @test() {
+; CHECK-LABEL: define ptr @test() {
+; CHECK-NEXT: [[ENTRY:.*:]]
+; CHECK-NEXT: br label %[[FOR_BODY_PEEL_BEGIN:.*]]
+; CHECK: [[FOR_BODY_PEEL_BEGIN]]:
+; CHECK-NEXT: br label %[[FOR_BODY_PEEL:.*]]
+; CHECK: [[FOR_BODY_PEEL]]:
+; CHECK-NEXT: br i1 true, label %[[CLEANUP_PEEL:.*]], label %[[CLEANUP_LOOPEXIT_PEEL:.*]]
+; CHECK: [[CLEANUP_LOOPEXIT_PEEL]]:
+; CHECK-NEXT: br label %[[CLEANUP_PEEL]]
+; CHECK: [[CLEANUP_PEEL]]:
+; CHECK-NEXT: [[RETVAL_2_PEEL:%.*]] = phi ptr [ undef, %[[FOR_BODY_PEEL]] ], [ @GlobIntONE, %[[CLEANUP_LOOPEXIT_PEEL]] ]
+; CHECK-NEXT: br i1 true, label %[[FOR_BODY_PEEL_NEXT:.*]], label %[[CLEANUP2:.*]]
+; CHECK: [[FOR_BODY_PEEL_NEXT]]:
+; CHECK-NEXT: br label %[[FOR_BODY_PEEL_NEXT1:.*]]
+; CHECK: [[FOR_BODY_PEEL_NEXT1]]:
+; CHECK-NEXT: br label %[[ENTRY_PEEL_NEWPH:.*]]
+; CHECK: [[ENTRY_PEEL_NEWPH]]:
+; CHECK-NEXT: br label %[[FOR_BODY:.*]]
+; CHECK: [[FOR_BODY]]:
+; CHECK-NEXT: br i1 false, label %[[CLEANUP:.*]], label %[[CLEANUP_LOOPEXIT:.*]]
+; CHECK: [[CLEANUP_LOOPEXIT]]:
+; CHECK-NEXT: br label %[[CLEANUP]]
+; CHECK: [[CLEANUP]]:
+; CHECK-NEXT: br i1 false, label %[[FOR_BODY]], label %[[CLEANUP2_LOOPEXIT:.*]], !llvm.loop [[LOOP0:![0-9]+]]
+; CHECK: [[CLEANUP2_LOOPEXIT]]:
+; CHECK-NEXT: [[RETVAL_2_LCSSA_PH:%.*]] = phi ptr [ @GlobIntONE, %[[CLEANUP]] ]
+; CHECK-NEXT: br label %[[CLEANUP2]]
+; CHECK: [[CLEANUP2]]:
+; CHECK-NEXT: [[RETVAL_2_LCSSA:%.*]] = phi ptr [ [[RETVAL_2_PEEL]], %[[CLEANUP_PEEL]] ], [ [[RETVAL_2_LCSSA_PH]], %[[CLEANUP2_LOOPEXIT]] ]
+; CHECK-NEXT: ret ptr [[RETVAL_2_LCSSA]]
+;
+entry:
+ br label %for.body
+
+for.body:
+ %retval.0 = phi ptr [ undef, %entry ], [ %retval.2, %cleanup ]
+ %cmp1.not = phi i1 [ true, %entry ], [ false, %cleanup ]
+ br i1 %cmp1.not, label %cleanup, label %cleanup.loopexit
+
+cleanup.loopexit:
+ br label %cleanup
+
+cleanup:
+ %retval.2 = phi ptr [ %retval.0, %for.body ], [ @GlobIntONE, %cleanup.loopexit ]
+ br i1 %cmp1.not, label %for.body, label %cleanup2
+
+cleanup2:
+ %retval.2.lcssa = phi ptr [ %retval.2, %cleanup ]
+ ret ptr %retval.2.lcssa
+}
+;.
+; CHECK: [[LOOP0]] = distinct !{[[LOOP0]], [[META1:![0-9]+]]}
+; CHECK: [[META1]] = !{!"llvm.loop.peeled.count", i32 1}
+;.
diff --git a/llvm/test/Transforms/LoopVectorize/uniform-blend.ll b/llvm/test/Transforms/LoopVectorize/uniform-blend.ll
index c9fc8beb006d9b..05919cdcf93c8c 100644
--- a/llvm/test/Transforms/LoopVectorize/uniform-blend.ll
+++ b/llvm/test/Transforms/LoopVectorize/uniform-blend.ll
@@ -9,16 +9,27 @@ define void @blend_uniform_iv_trunc(i1 %c) {
; CHECK-NEXT: [[ENTRY:.*]]:
; CHECK-NEXT: br i1 false, label %[[SCALAR_PH:.*]], label %[[VECTOR_PH:.*]]
; CHECK: [[VECTOR_PH]]:
+; CHECK-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <4 x i1> poison, i1 [[C]], i64 0
+; CHECK-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <4 x i1> [[BROADCAST_SPLATINSERT]], <4 x i1> poison, <4 x i32> zeroinitializer
; CHECK-NEXT: br label %[[VECTOR_BODY:.*]]
; CHECK: [[VECTOR_BODY]]:
; CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ]
-; CHECK-NEXT: [[TMP0:%.*]] = trunc i64 [[INDEX]] to i16
-; CHECK-NEXT: [[TMP1:%.*]] = add i16 [[TMP0]], 0
-; CHECK-NEXT: [[PREDPHI:%.*]] = select i1 [[C]], i16 [[TMP1]], i16 undef
+; CHECK-NEXT: [[VEC_IND:%.*]] = phi <4 x i16> [ <i16 0, i16 1, i16 2, i16 3>, %[[VECTOR_PH]] ], [ [[VEC_IND_NEXT:%.*]], %[[VECTOR_BODY]] ]
+; CHECK-NEXT: [[PREDPHI1:%.*]] = select <4 x i1> [[BROADCAST_SPLAT]], <4 x i16> [[VEC_IND]], <4 x i16> undef
+; CHECK-NEXT: [[PREDPHI:%.*]] = extractelement <4 x i16> [[PREDPHI1]], i32 0
; CHECK-NEXT: [[TMP2:%.*]] = getelementptr inbounds [32 x i16], ptr @dst, i16 0, i16 [[PREDPHI]]
-; CHECK-NEXT: [[TMP3:%.*]] = getelementptr inbounds i16, ptr [[TMP2]], i32 0
-; CHECK-NEXT: store <4 x i16> zeroinitializer, ptr [[TMP3]], align 2
+; CHECK-NEXT: [[TMP8:%.*]] = extractelement <4 x i16> [[PREDPHI1]], i32 1
+; CHECK-NEXT: [[TMP3:%.*]] = getelementptr inbounds [32 x i16], ptr @dst, i16 0, i16 [[TMP8]]
+; CHECK-NEXT: [[TMP9:%.*]] = extractelement <4 x i16> [[PREDPHI1]], i32 2
+; CHECK-NEXT: [[TMP5:%.*]] = getelementptr inbounds [32 x i16], ptr @dst, i16 0, i16 [[TMP9]]
+; CHECK-NEXT: [[TMP6:%.*]] = extractelement <4 x i16> [[PREDPHI1]], i32 3
+; CHECK-NEXT: [[TMP7:%.*]] = getelementptr inbounds [32 x i16], ptr @dst, i16 0, i16 [[TMP6]]
+; CHECK-NEXT: store i16 0, ptr [[TMP2]], align 2
+; CHECK-NEXT: store i16 0, ptr [[TMP3]], align 2
+; CHECK-NEXT: store i16 0, ptr [[TMP5]], align 2
+; CHECK-NEXT: store i16 0, ptr [[TMP7]], align 2
; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 4
+; CHECK-NEXT: [[VEC_IND_NEXT]] = add <4 x i16> [[VEC_IND]], <i16 4, i16 4, i16 4, i16 4>
; CHECK-NEXT: [[TMP4:%.*]] = icmp eq i64 [[INDEX_NEXT]], 32
; CHECK-NEXT: br i1 [[TMP4]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]]
; CHECK: [[MIDDLE_BLOCK]]:
@@ -71,15 +82,27 @@ define void @blend_uniform_iv(i1 %c) {
; CHECK-NEXT: [[ENTRY:.*]]:
; CHECK-NEXT: br i1 false, label %[[SCALAR_PH:.*]], label %[[VECTOR_PH:.*]]
; CHECK: [[VECTOR_PH]]:
+; CHECK-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <4 x i1> poison, i1 [[C]], i64 0
+; CHECK-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <4 x i1> [[BROADCAST_SPLATINSERT]], <4 x i1> poison, <4 x i32> zeroinitializer
; CHECK-NEXT: br label %[[VECTOR_BODY:.*]]
; CHECK: [[VECTOR_BODY]]:
; CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ]
-; CHECK-NEXT: [[TMP0:%.*]] = add i64 [[INDEX]], 0
-; CHECK-NEXT: [[PREDPHI:%.*]] = select i1 [[C]], i64 [[TMP0]], i64 undef
+; CHECK-NEXT: [[VEC_IND:%.*]] = phi <4 x i64> [ <i64 0, i64 1, i64 2, i64 3>, %[[VECTOR_PH]] ], [ [[VEC_IND_NEXT:%.*]], %[[VECTOR_BODY]] ]
+; CHECK-NEXT: [[PREDPHI1:%.*]] = select <4 x i1> [[BROADCAST_SPLAT]], <4 x i64> [[VEC_IND]], <4 x i64> undef
+; CHECK-NEXT: [[PREDPHI:%.*]] = extractelement <4 x i64> [[PREDPHI1]], i32 0
; CHECK-NEXT: [[TMP1:%.*]] = getelementptr inbounds [32 x i16], ptr @dst, i16 0, i64 [[PREDPHI]]
-; CHECK-NEXT: [[TMP2:%.*]] = getelementptr inbounds i16, ptr [[TMP1]], i32 0
-; CHECK-NEXT: store <4 x i16> zeroinitializer, ptr [[TMP2]], align 2
+; CHECK-NEXT: [[TMP2:%.*]] = extractelement <4 x i64> [[PREDPHI1]], i32 1
+; CHECK-NEXT: [[TMP8:%.*]] = getelementptr inbounds [32 x i16], ptr @dst, i16 0, i64 [[TMP2]]
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <4 x i64> [[PREDPHI1]], i32 2
+; CHECK-NEXT: [[TMP5:%.*]] = getelementptr inbounds [32 x i16], ptr @dst, i16 0, i64 [[TMP4]]
+; CHECK-NEXT: [[TMP6:%.*]] = extractelement <4 x i64> [[PREDPHI1]], i32 3
+; CHECK-NEXT: [[TMP7:%.*]] = getelementptr inbounds [32 x i16], ptr @dst, i16 0, i64 [[TMP6]]
+; CHECK-NEXT: store i16 0, ptr [[TMP1]], align 2
+; CHECK-NEXT: store i16 0, ptr [[TMP8]], align 2
+; CHECK-NEXT: store i16 0, ptr [[TMP5]], align 2
+; CHECK-NEXT: store i16 0, ptr [[TMP7]], align 2
; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 4
+; CHECK-NEXT: [[VEC_IND_NEXT]] = add <4 x i64> [[VEC_IND]], <i64 4, i64 4, i64 4, i64 4>
; CHECK-NEXT: [[TMP3:%.*]] = icmp eq i64 [[INDEX_NEXT]], 32
; CHECK-NEXT: br i1 [[TMP3]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]]
; CHECK: [[MIDDLE_BLOCK]]:
diff --git a/llvm/unittests/Analysis/ScalarEvolutionTest.cpp b/llvm/unittests/Analysis/ScalarEvolutionTest.cpp
index 37c61e4e4fa714..3450302f36f617 100644
--- a/llvm/unittests/Analysis/ScalarEvolutionTest.cpp
+++ b/llvm/unittests/Analysis/ScalarEvolutionTest.cpp
@@ -139,7 +139,7 @@ TEST_F(ScalarEvolutionsTest, SimplifiedPHI) {
auto *Ty = Type::getInt32Ty(Context);
auto *PN = PHINode::Create(Ty, 2, "", LoopBB->begin());
PN->addIncoming(Constant::getNullValue(Ty), EntryBB);
- PN->addIncoming(UndefValue::get(Ty), LoopBB);
+ PN->addIncoming(PoisonValue::get(Ty), LoopBB);
ScalarEvolution SE = buildSE(*F);
const SCEV *S1 = SE.getSCEV(PN);
const SCEV *S2 = SE.getSCEV(PN);
|
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
See the following case:
simplifyInstruction(%retval.2.peel)returns@GlobIntONE. Thus,ScalarEvolution::createNodeForPHIreturns SCEV expr@GlobIntONEfor%retval.2.peel.SimplifyIndvar::replaceIVUserWithLoopInvarianttries to replace the use of%retval.2.peelin%retval.2.lcssa.phwith@GlobIntONE.simplifyLoopAfterUnroll -> simplifyLoopIVs -> SCEVExpander::expandreuses%retval.2.peel = phi ptr [ undef, %for.body.peel ], [ @GlobIntONE, %cleanup.loopexit.peel ]to generate code for@GlobIntONE. It is incorrect.This patch disallows simplifying
phi(undef, X)toXby settingCanUseUndefto false.Closes #114879.