Loading CombinedParsers assert on 1.8 with "Illegal inttoptr" #45702
Assignees
Labels
compiler:llvm
For issues that relate to LLVM
regression
Regression in behavior compared to a previous version
upstream
The issue is with an upstream dependency, e.g. LLVM
Comments
KristofferC
added
the
regression
vtjnash
added
upstream
|
Looks like a magicptr is at fault here. Very simple to fix, assuming this won't cause regressions for anyone (which should be very unlikely since it is a very rare case, and early-cse, gvn, or newgvn will usually handle this better anyways): diff --git a/llvm/lib/Transforms/Utils/SimplifyCFG.cpp b/llvm/lib/Transforms/Utils/SimplifyCFG.cpp
index 399fc3baf7aa..53fa6f70a0a0 100644
--- a/llvm/lib/Transforms/Utils/SimplifyCFG.cpp
+++ b/llvm/lib/Transforms/Utils/SimplifyCFG.cpp
@@ -478,7 +478,7 @@ static bool dominatesMergePoint(Value *V, BasicBlock *BB,
static ConstantInt *GetConstantInt(Value *V, const DataLayout &DL) {
// Normal constant int.
ConstantInt *CI = dyn_cast<ConstantInt>(V);
- if (CI || !isa<Constant>(V) || !V->getType()->isPointerTy())
+ if (CI || !isa<Constant>(V) || !V->getType()->isPointerTy() || DL.isNonIntegralPointerType(V->getType()))
return CI;
// This is some kind of pointer constant. Turn it into a pointer-sized
diff --git a/llvm/test/Transforms/SimplifyCFG/unprofit2.ll b/llvm/test/Transforms/SimplifyCFG/unprofit2.ll
new file mode 100644
index 000000000000..00c295ddc5c6
--- /dev/null
+++ b/llvm/test/Transforms/SimplifyCFG/unprofit2.ll
@@ -0,0 +1,29 @@
+; RUN: opt -simplifycfg -verify -S < %s | FileCheck %s
+; RUN: opt -passes=simplifycfg,verify -S < %s | FileCheck %s
+
+target datalayout = "ni:1"
+
+define void @test_01(i64 addrspace(1)* align 8 %ptr) local_unnamed_addr #0 {
+; CHECK-LABEL: @test_01(
+; CHECK-NOT: ptrtoint
+; CHECK-NEXT: icmp eq i64 addrspace(1)* %ptr, null
+; CHECK-NOT: ptrtoint
+ %cond1 = icmp eq i64 addrspace(1)* %ptr, null
+ %cond2 = icmp eq i64 addrspace(1)* %ptr, null
+ br i1 %cond1, label %true1, label %false1
+
+true1:
+ br i1 %cond2, label %true2, label %false2
+
+false1:
+ store i64 1, i64 addrspace(1)* %ptr, align 8
+ br label %true1
+
+true2:
+ store i64 2, i64 addrspace(1)* %ptr, align 8
+ ret void
+
+false2:
+ store i64 3, i64 addrspace(1)* %ptr, align 8
+ ret void
+} |
KristofferC
pushed a commit
to KristofferC/llvm-project
that referenced
this issue
Jun 30, 2022
|
@vtjnash can you post this patch to upstream? |
|
Upstream: https://reviews.llvm.org/D128670 |
vchuravy
pushed a commit
to JuliaLang/llvm-project
that referenced
this issue
Jun 30, 2022
…ters SimplifyCFG expects to be able to cast both sides to an int, if either side can be case to an int, but this is not desirable or legal, in general, per D104547. Spotted in JuliaLang/julia#45702 Differential Revision: https://reviews.llvm.org/D128670 (cherry picked from commit b904e7e94c5cc9d86762b32211cd0e35670534cb)
vchuravy
pushed a commit
to JuliaLang/llvm-project
that referenced
this issue
Jun 30, 2022
…ters SimplifyCFG expects to be able to cast both sides to an int, if either side can be case to an int, but this is not desirable or legal, in general, per D104547. Spotted in JuliaLang/julia#45702 Differential Revision: https://reviews.llvm.org/D128670 (cherry picked from commit b904e7e94c5cc9d86762b32211cd0e35670534cb)
|
Picked up the fix and I am releasing a new binary for LLVM 13 JuliaPackaging/Yggdrasil#5092 |
vchuravy
pushed a commit
to llvm/llvm-project
that referenced
this issue
Aug 15, 2022
…ters SimplifyCFG expects to be able to cast both sides to an int, if either side can be case to an int, but this is not desirable or legal, in general, per D104547. Spotted in JuliaLang/julia#45702 Reviewed By: nikic Differential Revision: https://reviews.llvm.org/D128670
vchuravy
pushed a commit
to JuliaLang/llvm-project
that referenced
this issue
Aug 23, 2022
…ters SimplifyCFG expects to be able to cast both sides to an int, if either side can be case to an int, but this is not desirable or legal, in general, per D104547. Spotted in JuliaLang/julia#45702 Differential Revision: https://reviews.llvm.org/D128670 (cherry picked from commit b904e7e94c5cc9d86762b32211cd0e35670534cb)
thilinarmtb
pushed a commit
to nomp-org/llvm-project
that referenced
this issue
Aug 24, 2022
…ters SimplifyCFG expects to be able to cast both sides to an int, if either side can be case to an int, but this is not desirable or legal, in general, per D104547. Spotted in JuliaLang/julia#45702 Reviewed By: nikic Differential Revision: https://reviews.llvm.org/D128670
pchintalapudi
pushed a commit
to JuliaLang/llvm-project
that referenced
this issue
Oct 20, 2022
…ters SimplifyCFG expects to be able to cast both sides to an int, if either side can be case to an int, but this is not desirable or legal, in general, per D104547. Spotted in JuliaLang/julia#45702 Differential Revision: https://reviews.llvm.org/D128670 (cherry picked from commit b904e7e94c5cc9d86762b32211cd0e35670534cb)
vchuravy
pushed a commit
to JuliaLang/llvm-project
that referenced
this issue
Feb 14, 2023
…ters SimplifyCFG expects to be able to cast both sides to an int, if either side can be case to an int, but this is not desirable or legal, in general, per D104547. Spotted in JuliaLang/julia#45702 Differential Revision: https://reviews.llvm.org/D128670 (cherry picked from commit b904e7e94c5cc9d86762b32211cd0e35670534cb)
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
(Running with assertions on)
https://s3.amazonaws.com/julialang-reports/nanosoldier/pkgeval/by_hash/8b2e406_vs_742b9ab/CombinedParsers.primary.log
The text was updated successfully, but these errors were encountered: