[llvm] r359879 - [CodeGenPrepare] limit overflow intrinsic matching to a single basic block
Evgenii Stepanov via llvm-commits
llvm-commits at lists.llvm.org
Fri May 3 10:13:48 PDT 2019
Hi,
this crash might be caused by this commit:
http://lab.llvm.org:8011/builders/sanitizer-x86_64-linux-bootstrap-ubsan/builds/12660/steps/build%20clang%2Fubsan/logs/stdio
0. Program arguments:
/b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm_build0/bin/clang-9
-cc1 -triple x86_64-unknown-linux-gnu -emit-obj -disable-free
-main-file-name CommandLine.cpp -mrelocation-model pic -pic-level 2
-mthread-model posix -mdisable-fp-elim -fmath-errno -masm-verbose
-mconstructor-aliases -munwind-tables -fuse-init-array -target-cpu
x86-64 -dwarf-column-info -debug-info-kind=line-tables-only
-dwarf-version=4 -debugger-tuning=gdb -momit-leaf-frame-pointer
-ffunction-sections -fdata-sections -coverage-notes-file
/b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm_build_ubsan/lib/Support/CMakeFiles/LLVMSupport.dir/CommandLine.cpp.gcno
-resource-dir /b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm_build0/lib/clang/9.0.0
-dependency-file
lib/Support/CMakeFiles/LLVMSupport.dir/CommandLine.cpp.o.d
-sys-header-deps -MT
lib/Support/CMakeFiles/LLVMSupport.dir/CommandLine.cpp.o -D
GTEST_HAS_RTTI=0 -D _DEBUG -D _GNU_SOURCE -D __STDC_CONSTANT_MACROS -D
__STDC_FORMAT_MACROS -D __STDC_LIMIT_MACROS -I lib/Support -I
/b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm/lib/Support -I
/usr/include/libxml2 -I include -I
/b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm/include -U NDEBUG
-internal-isystem
/usr/lib/gcc/x86_64-linux-gnu/6.3.0/../../../../include/c++/6.3.0
-internal-isystem
/usr/lib/gcc/x86_64-linux-gnu/6.3.0/../../../../include/x86_64-linux-gnu/c++/6.3.0
-internal-isystem
/usr/lib/gcc/x86_64-linux-gnu/6.3.0/../../../../include/x86_64-linux-gnu/c++/6.3.0
-internal-isystem
/usr/lib/gcc/x86_64-linux-gnu/6.3.0/../../../../include/c++/6.3.0/backward
-internal-isystem /usr/local/include -internal-isystem
/b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm_build0/lib/clang/9.0.0/include
-internal-externc-isystem /usr/include/x86_64-linux-gnu
-internal-externc-isystem /include -internal-externc-isystem
/usr/include -O3 -Werror=date-time -Werror=unguarded-availability-new
-Wall -Wextra -Wno-unused-parameter -Wwrite-strings -Wcast-qual
-Wmissing-field-initializers -Wno-long-long -Wimplicit-fallthrough
-Wcovered-switch-default -Wno-class-memaccess -Wno-noexcept-type
-Wnon-virtual-dtor -Wdelete-non-virtual-dtor -Wstring-conversion
-pedantic -w -std=c++11 -fdeprecated-macro -fdebug-compilation-dir
/b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm_build_ubsan
-ferror-limit 19 -fmessage-length 0 -fvisibility-inlines-hidden
-fsanitize=alignment,array-bounds,bool,builtin,enum,float-cast-overflow,float-divide-by-zero,integer-divide-by-zero,nonnull-attribute,null,object-size,pointer-overflow,return,returns-nonnull-attribute,shift-base,shift-exponent,signed-integer-overflow,unreachable,vla-bound
-fsanitize-blacklist=/b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm/utils/sanitizers/ubsan_blacklist.txt
-fdepfile-entry=/b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm/utils/sanitizers/ubsan_blacklist.txt
-fno-rtti -fobjc-runtime=gcc -fdiagnostics-show-option
-fcolor-diagnostics -vectorize-loops -vectorize-slp -o
lib/Support/CMakeFiles/LLVMSupport.dir/CommandLine.cpp.o -x c++
/b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm/lib/Support/CommandLine.cpp
-faddrsig
1. <eof> parser at end of file
2. Code generation
3. Running pass 'Function Pass Manager' on module
'/b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm/lib/Support/CommandLine.cpp'.
4. Running pass 'CodeGen Prepare' on function
'@_ZN12_GLOBAL__N_117CommandLineParser23ParseCommandLineOptionsEiPKPKcN4llvm9StringRefEPNS5_11raw_ostreamE'
#0 0x000055d2f954521a llvm::sys::PrintStackTrace(llvm::raw_ostream&)
(/b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm_build0/bin/clang-9+0x298521a)
#1 0x000055d2f9543175 llvm::sys::RunSignalHandlers()
(/b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm_build0/bin/clang-9+0x2983175)
#2 0x000055d2f954328c SignalHandler(int)
(/b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm_build0/bin/clang-9+0x298328c)
#3 0x00007f0bd09f40e0 __restore_rt
(/lib/x86_64-linux-gnu/libpthread.so.0+0x110e0)
#4 0x000055d2f8fb84f0
llvm::ReplaceableMetadataImpl::getIfExists(llvm::Metadata&)
(/b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm_build0/bin/clang-9+0x23f84f0)
#5 0x000055d2f8fb8569 llvm::MetadataTracking::untrack(void*,
llvm::Metadata&)
(/b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm_build0/bin/clang-9+0x23f8569)
#6 0x000055d2f8f742b4 llvm::Instruction::~Instruction()
(/b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm_build0/bin/clang-9+0x23b42b4)
#7 0x000055d2f8fef140 llvm::Value::deleteValue()
(/b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm_build0/bin/clang-9+0x242f140)
#8 0x000055d2f8b53944 (anonymous
namespace)::CodeGenPrepare::runOnFunction(llvm::Function&)
(.part.1278) (/b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm_build0/bin/clang-9+0x1f93944)
#9 0x000055d2f8fa8710
llvm::FPPassManager::runOnFunction(llvm::Function&)
(/b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm_build0/bin/clang-9+0x23e8710)
#10 0x000055d2f8fa892c llvm::FPPassManager::runOnModule(llvm::Module&)
(/b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm_build0/bin/clang-9+0x23e892c)
#11 0x000055d2f8fa9454
llvm::legacy::PassManagerImpl::run(llvm::Module&)
(/b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm_build0/bin/clang-9+0x23e9454)
#12 0x000055d2f973ed37 (anonymous
namespace)::EmitAssemblyHelper::EmitAssembly(clang::BackendAction,
std::unique_ptr<llvm::raw_pwrite_stream,
std::default_delete<llvm::raw_pwrite_stream> >)
(/b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm_build0/bin/clang-9+0x2b7ed37)
#13 0x000055d2f974063c
clang::EmitBackendOutput(clang::DiagnosticsEngine&,
clang::HeaderSearchOptions const&, clang::CodeGenOptions const&,
clang::TargetOptions const&, clang::LangOptions const&,
llvm::DataLayout const&, llvm::Module*, clang::BackendAction,
std::unique_ptr<llvm::raw_pwrite_stream,
std::default_delete<llvm::raw_pwrite_stream> >)
(/b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm_build0/bin/clang-9+0x2b8063c)
#14 0x000055d2fa11f628
clang::BackendConsumer::HandleTranslationUnit(clang::ASTContext&)
(/b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm_build0/bin/clang-9+0x355f628)
#15 0x000055d2fa982319 clang::ParseAST(clang::Sema&, bool, bool)
(/b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm_build0/bin/clang-9+0x3dc2319)
#16 0x000055d2fa11e8b9 clang::CodeGenAction::ExecuteAction()
(/b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm_build0/bin/clang-9+0x355e8b9)
#17 0x000055d2f9bcfbbe clang::FrontendAction::Execute()
(/b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm_build0/bin/clang-9+0x300fbbe)
#18 0x000055d2f9b97b96
clang::CompilerInstance::ExecuteAction(clang::FrontendAction&)
(/b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm_build0/bin/clang-9+0x2fd7b96)
#19 0x000055d2f9c73271
clang::ExecuteCompilerInvocation(clang::CompilerInstance*)
(/b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm_build0/bin/clang-9+0x30b3271)
#20 0x000055d2f7a5db77 cc1_main(llvm::ArrayRef<char const*>, char
const*, void*) (/b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm_build0/bin/clang-9+0xe9db77)
#21 0x000055d2f79c812d main
(/b/sanitizer-x86_64-linux-bootstrap-ubsan/build/llvm_build0/bin/clang-9+0xe0812d)
On Fri, May 3, 2019 at 6:06 AM Sanjay Patel via llvm-commits
<llvm-commits at lists.llvm.org> wrote:
>
> Author: spatel
> Date: Fri May 3 06:09:18 2019
> New Revision: 359879
>
> URL: http://llvm.org/viewvc/llvm-project?rev=359879&view=rev
> Log:
> [CodeGenPrepare] limit overflow intrinsic matching to a single basic block
>
> Using/updating a dominator tree to match math overflow patterns may be very
> expensive in compile-time (because of the way CGP uses a DT), so just handle
> the single-block case.
>
> Also, we were restarting the iterator loops when doing the overflow intrinsic
> transforms by marking the dominator tree for update. That was done to prevent
> iterating over a removed instruction. But we can postpone the deletion using
> the existing "RemovedInsts" structure, and that means we don't need to update
> the DT.
>
> See post-commit thread for rL354298 for more details:
> http://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-20190422/646276.html
>
> Differential Revision: https://reviews.llvm.org/D61075
>
> Modified:
> llvm/trunk/lib/CodeGen/CodeGenPrepare.cpp
> llvm/trunk/test/CodeGen/X86/cgp-usubo.ll
> llvm/trunk/test/Transforms/CodeGenPrepare/X86/optimizeSelect-DT.ll
> llvm/trunk/test/Transforms/CodeGenPrepare/X86/overflow-intrinsics.ll
>
> Modified: llvm/trunk/lib/CodeGen/CodeGenPrepare.cpp
> URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/CodeGen/CodeGenPrepare.cpp?rev=359879&r1=359878&r2=359879&view=diff
> ==============================================================================
> --- llvm/trunk/lib/CodeGen/CodeGenPrepare.cpp (original)
> +++ llvm/trunk/lib/CodeGen/CodeGenPrepare.cpp Fri May 3 06:09:18 2019
> @@ -388,9 +388,9 @@ class TypePromotionTransaction;
> bool tryToSinkFreeOperands(Instruction *I);
> bool replaceMathCmpWithIntrinsic(BinaryOperator *BO, CmpInst *Cmp,
> Intrinsic::ID IID);
> - bool optimizeCmp(CmpInst *Cmp, bool &ModifiedDT);
> - bool combineToUSubWithOverflow(CmpInst *Cmp, bool &ModifiedDT);
> - bool combineToUAddWithOverflow(CmpInst *Cmp, bool &ModifiedDT);
> + bool optimizeCmp(CmpInst *Cmp);
> + bool combineToUSubWithOverflow(CmpInst *Cmp);
> + bool combineToUAddWithOverflow(CmpInst *Cmp);
> };
>
> } // end anonymous namespace
> @@ -484,9 +484,13 @@ bool CodeGenPrepare::runOnFunction(Funct
> if (!LargeOffsetGEPMap.empty())
> MadeChange |= splitLargeGEPOffsets();
>
> - // Really free removed instructions during promotion.
> - for (Instruction *I : RemovedInsts)
> + // Really free instructions removed during promotion or kept around to
> + // improve efficiency (see comments in overflow intrinsic transforms).
> + for (Instruction *I : RemovedInsts) {
> + if (I->getParent())
> + I->removeFromParent();
> I->deleteValue();
> + }
>
> EverMadeChange |= MadeChange;
> SeenChainsForSExt.clear();
> @@ -1177,6 +1181,20 @@ static bool OptimizeNoopCopyExpression(C
> bool CodeGenPrepare::replaceMathCmpWithIntrinsic(BinaryOperator *BO,
> CmpInst *Cmp,
> Intrinsic::ID IID) {
> + if (BO->getParent() != Cmp->getParent()) {
> + // We used to use a dominator tree here to allow multi-block optimization.
> + // But that was problematic because:
> + // 1. It could cause a perf regression by hoisting the math op into the
> + // critical path.
> + // 2. It could cause a perf regression by creating a value that was live
> + // across multiple blocks and increasing register pressure.
> + // 3. Use of a dominator tree could cause large compile-time regression.
> + // This is because we recompute the DT on every change in the main CGP
> + // run-loop. The recomputing is probably unnecessary in many cases, so if
> + // that was fixed, using a DT here would be ok.
> + return false;
> + }
> +
> // We allow matching the canonical IR (add X, C) back to (usubo X, -C).
> Value *Arg0 = BO->getOperand(0);
> Value *Arg1 = BO->getOperand(1);
> @@ -1186,45 +1204,28 @@ bool CodeGenPrepare::replaceMathCmpWithI
> Arg1 = ConstantExpr::getNeg(cast<Constant>(Arg1));
> }
>
> - Instruction *InsertPt;
> - if (BO->hasOneUse() && BO->user_back() == Cmp) {
> - // If the math is only used by the compare, insert at the compare to keep
> - // the condition in the same block as its users. (CGP aggressively sinks
> - // compares to help out SDAG.)
> - InsertPt = Cmp;
> - } else {
> - // The math and compare may be independent instructions. Check dominance to
> - // determine the insertion point for the intrinsic.
> - bool MathDominates = getDT(*BO->getFunction()).dominates(BO, Cmp);
> - if (!MathDominates && !getDT(*BO->getFunction()).dominates(Cmp, BO))
> - return false;
> -
> - BasicBlock *MathBB = BO->getParent(), *CmpBB = Cmp->getParent();
> - if (MathBB != CmpBB) {
> - // Avoid hoisting an extra op into a dominating block and creating a
> - // potentially longer critical path.
> - if (!MathDominates)
> - return false;
> - // Check that the insertion doesn't create a value that is live across
> - // more than two blocks, so to minimise the increase in register pressure.
> - BasicBlock *Dominator = MathDominates ? MathBB : CmpBB;
> - BasicBlock *Dominated = MathDominates ? CmpBB : MathBB;
> - auto Successors = successors(Dominator);
> - if (llvm::find(Successors, Dominated) == Successors.end())
> - return false;
> + // Insert at the first instruction of the pair.
> + Instruction *InsertPt = nullptr;
> + for (Instruction &Iter : *Cmp->getParent()) {
> + if (&Iter == BO || &Iter == Cmp) {
> + InsertPt = &Iter;
> + break;
> }
> -
> - InsertPt = MathDominates ? cast<Instruction>(BO) : cast<Instruction>(Cmp);
> }
> + assert(InsertPt != nullptr && "Parent block did not contain cmp or binop");
>
> IRBuilder<> Builder(InsertPt);
> Value *MathOV = Builder.CreateBinaryIntrinsic(IID, Arg0, Arg1);
> Value *Math = Builder.CreateExtractValue(MathOV, 0, "math");
> Value *OV = Builder.CreateExtractValue(MathOV, 1, "ov");
> +
> + // Delay the actual removal/deletion of the binop and compare for efficiency.
> + // If we delete those now, we would invalidate the instruction iterator and
> + // trigger dominator tree updates.
> BO->replaceAllUsesWith(Math);
> Cmp->replaceAllUsesWith(OV);
> - BO->eraseFromParent();
> - Cmp->eraseFromParent();
> + RemovedInsts.insert(BO);
> + RemovedInsts.insert(Cmp);
> return true;
> }
>
> @@ -1260,8 +1261,7 @@ static bool matchUAddWithOverflowConstan
>
> /// Try to combine the compare into a call to the llvm.uadd.with.overflow
> /// intrinsic. Return true if any changes were made.
> -bool CodeGenPrepare::combineToUAddWithOverflow(CmpInst *Cmp,
> - bool &ModifiedDT) {
> +bool CodeGenPrepare::combineToUAddWithOverflow(CmpInst *Cmp) {
> Value *A, *B;
> BinaryOperator *Add;
> if (!match(Cmp, m_UAddWithOverflow(m_Value(A), m_Value(B), m_BinOp(Add))))
> @@ -1281,13 +1281,10 @@ bool CodeGenPrepare::combineToUAddWithOv
> if (!replaceMathCmpWithIntrinsic(Add, Cmp, Intrinsic::uadd_with_overflow))
> return false;
>
> - // Reset callers - do not crash by iterating over a dead instruction.
> - ModifiedDT = true;
> return true;
> }
>
> -bool CodeGenPrepare::combineToUSubWithOverflow(CmpInst *Cmp,
> - bool &ModifiedDT) {
> +bool CodeGenPrepare::combineToUSubWithOverflow(CmpInst *Cmp) {
> // We are not expecting non-canonical/degenerate code. Just bail out.
> Value *A = Cmp->getOperand(0), *B = Cmp->getOperand(1);
> if (isa<Constant>(A) && isa<Constant>(B))
> @@ -1342,8 +1339,6 @@ bool CodeGenPrepare::combineToUSubWithOv
> if (!replaceMathCmpWithIntrinsic(Sub, Cmp, Intrinsic::usub_with_overflow))
> return false;
>
> - // Reset callers - do not crash by iterating over a dead instruction.
> - ModifiedDT = true;
> return true;
> }
>
> @@ -1413,14 +1408,14 @@ static bool sinkCmpExpression(CmpInst *C
> return MadeChange;
> }
>
> -bool CodeGenPrepare::optimizeCmp(CmpInst *Cmp, bool &ModifiedDT) {
> +bool CodeGenPrepare::optimizeCmp(CmpInst *Cmp) {
> if (sinkCmpExpression(Cmp, *TLI))
> return true;
>
> - if (combineToUAddWithOverflow(Cmp, ModifiedDT))
> + if (combineToUAddWithOverflow(Cmp))
> return true;
>
> - if (combineToUSubWithOverflow(Cmp, ModifiedDT))
> + if (combineToUSubWithOverflow(Cmp))
> return true;
>
> return false;
> @@ -6945,7 +6940,7 @@ bool CodeGenPrepare::optimizeInst(Instru
> }
>
> if (auto *Cmp = dyn_cast<CmpInst>(I))
> - if (TLI && optimizeCmp(Cmp, ModifiedDT))
> + if (TLI && optimizeCmp(Cmp))
> return true;
>
> if (LoadInst *LI = dyn_cast<LoadInst>(I)) {
>
> Modified: llvm/trunk/test/CodeGen/X86/cgp-usubo.ll
> URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/X86/cgp-usubo.ll?rev=359879&r1=359878&r2=359879&view=diff
> ==============================================================================
> --- llvm/trunk/test/CodeGen/X86/cgp-usubo.ll (original)
> +++ llvm/trunk/test/CodeGen/X86/cgp-usubo.ll Fri May 3 06:09:18 2019
> @@ -121,7 +121,7 @@ define i1 @usubo_ne_constant0_op1_i32(i3
> ret i1 %ov
> }
>
> -; Verify insertion point for multi-BB.
> +; This used to verify insertion point for multi-BB, but now we just bail out.
>
> declare void @call(i1)
>
> @@ -131,14 +131,17 @@ define i1 @usubo_ult_sub_dominates_i64(i
> ; CHECK-NEXT: testb $1, %cl
> ; CHECK-NEXT: je .LBB8_2
> ; CHECK-NEXT: # %bb.1: # %t
> -; CHECK-NEXT: subq %rsi, %rdi
> -; CHECK-NEXT: setb %al
> -; CHECK-NEXT: movq %rdi, (%rdx)
> +; CHECK-NEXT: movq %rdi, %rax
> +; CHECK-NEXT: subq %rsi, %rax
> +; CHECK-NEXT: movq %rax, (%rdx)
> ; CHECK-NEXT: testb $1, %cl
> -; CHECK-NEXT: jne .LBB8_3
> +; CHECK-NEXT: je .LBB8_2
> +; CHECK-NEXT: # %bb.3: # %end
> +; CHECK-NEXT: cmpq %rsi, %rdi
> +; CHECK-NEXT: setb %al
> +; CHECK-NEXT: retq
> ; CHECK-NEXT: .LBB8_2: # %f
> ; CHECK-NEXT: movl %ecx, %eax
> -; CHECK-NEXT: .LBB8_3: # %end
> ; CHECK-NEXT: retq
> entry:
> br i1 %cond, label %t, label %f
>
> Modified: llvm/trunk/test/Transforms/CodeGenPrepare/X86/optimizeSelect-DT.ll
> URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/Transforms/CodeGenPrepare/X86/optimizeSelect-DT.ll?rev=359879&r1=359878&r2=359879&view=diff
> ==============================================================================
> --- llvm/trunk/test/Transforms/CodeGenPrepare/X86/optimizeSelect-DT.ll (original)
> +++ llvm/trunk/test/Transforms/CodeGenPrepare/X86/optimizeSelect-DT.ll Fri May 3 06:09:18 2019
> @@ -14,11 +14,10 @@ define i1 @PR41004(i32 %x, i32 %y, i32 %
> ; CHECK-NEXT: br label [[SELECT_END]]
> ; CHECK: select.end:
> ; CHECK-NEXT: [[MUL:%.*]] = phi i32 [ [[REM]], [[SELECT_TRUE_SINK]] ], [ 0, [[ENTRY:%.*]] ]
> -; CHECK-NEXT: [[TMP0:%.*]] = call { i32, i1 } @llvm.usub.with.overflow.i32(i32 [[T1:%.*]], i32 1)
> -; CHECK-NEXT: [[MATH:%.*]] = extractvalue { i32, i1 } [[TMP0]], 0
> -; CHECK-NEXT: [[OV:%.*]] = extractvalue { i32, i1 } [[TMP0]], 1
> -; CHECK-NEXT: [[ADD:%.*]] = add i32 [[MATH]], [[MUL]]
> -; CHECK-NEXT: ret i1 [[OV]]
> +; CHECK-NEXT: [[NEG:%.*]] = add i32 [[T1:%.*]], -1
> +; CHECK-NEXT: [[ADD:%.*]] = add i32 [[NEG]], [[MUL]]
> +; CHECK-NEXT: [[TOBOOL:%.*]] = icmp eq i32 [[T1]], 0
> +; CHECK-NEXT: ret i1 [[TOBOOL]]
> ;
> entry:
> %rem = srem i32 %x, 2
>
> Modified: llvm/trunk/test/Transforms/CodeGenPrepare/X86/overflow-intrinsics.ll
> URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/Transforms/CodeGenPrepare/X86/overflow-intrinsics.ll?rev=359879&r1=359878&r2=359879&view=diff
> ==============================================================================
> --- llvm/trunk/test/Transforms/CodeGenPrepare/X86/overflow-intrinsics.ll (original)
> +++ llvm/trunk/test/Transforms/CodeGenPrepare/X86/overflow-intrinsics.ll Fri May 3 06:09:18 2019
> @@ -47,15 +47,16 @@ define i64 @uaddo3(i64 %a, i64 %b) nounw
> ret i64 %Q
> }
>
> +; TODO? CGP sinks the compare before we have a chance to form the overflow intrinsic.
> +
> define i64 @uaddo4(i64 %a, i64 %b, i1 %c) nounwind ssp {
> ; CHECK-LABEL: @uaddo4(
> ; CHECK-NEXT: entry:
> +; CHECK-NEXT: [[ADD:%.*]] = add i64 [[B:%.*]], [[A:%.*]]
> ; CHECK-NEXT: br i1 [[C:%.*]], label [[NEXT:%.*]], label [[EXIT:%.*]]
> ; CHECK: next:
> -; CHECK-NEXT: [[TMP0:%.*]] = call { i64, i1 } @llvm.uadd.with.overflow.i64(i64 [[B:%.*]], i64 [[A:%.*]])
> -; CHECK-NEXT: [[MATH:%.*]] = extractvalue { i64, i1 } [[TMP0]], 0
> -; CHECK-NEXT: [[OV:%.*]] = extractvalue { i64, i1 } [[TMP0]], 1
> -; CHECK-NEXT: [[Q:%.*]] = select i1 [[OV]], i64 [[B]], i64 42
> +; CHECK-NEXT: [[TMP0:%.*]] = icmp ugt i64 [[B]], [[ADD]]
> +; CHECK-NEXT: [[Q:%.*]] = select i1 [[TMP0]], i64 [[B]], i64 42
> ; CHECK-NEXT: ret i64 [[Q]]
> ; CHECK: exit:
> ; CHECK-NEXT: ret i64 0
> @@ -362,7 +363,7 @@ define i1 @usubo_ne_constant0_op1_i32(i3
> ret i1 %ov
> }
>
> -; Verify insertion point for multi-BB.
> +; This used to verify insertion point for multi-BB, but now we just bail out.
>
> declare void @call(i1)
>
> @@ -371,15 +372,14 @@ define i1 @usubo_ult_sub_dominates_i64(i
> ; CHECK-NEXT: entry:
> ; CHECK-NEXT: br i1 [[COND:%.*]], label [[T:%.*]], label [[F:%.*]]
> ; CHECK: t:
> -; CHECK-NEXT: [[TMP0:%.*]] = call { i64, i1 } @llvm.usub.with.overflow.i64(i64 [[X:%.*]], i64 [[Y:%.*]])
> -; CHECK-NEXT: [[MATH:%.*]] = extractvalue { i64, i1 } [[TMP0]], 0
> -; CHECK-NEXT: [[OV1:%.*]] = extractvalue { i64, i1 } [[TMP0]], 1
> -; CHECK-NEXT: store i64 [[MATH]], i64* [[P:%.*]]
> +; CHECK-NEXT: [[S:%.*]] = sub i64 [[X:%.*]], [[Y:%.*]]
> +; CHECK-NEXT: store i64 [[S]], i64* [[P:%.*]]
> ; CHECK-NEXT: br i1 [[COND]], label [[END:%.*]], label [[F]]
> ; CHECK: f:
> ; CHECK-NEXT: ret i1 [[COND]]
> ; CHECK: end:
> -; CHECK-NEXT: ret i1 [[OV1]]
> +; CHECK-NEXT: [[OV:%.*]] = icmp ult i64 [[X]], [[Y]]
> +; CHECK-NEXT: ret i1 [[OV]]
> ;
> entry:
> br i1 %cond, label %t, label %f
>
>
> _______________________________________________
> llvm-commits mailing list
> llvm-commits at lists.llvm.org
> https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-commits
More information about the llvm-commits
mailing list