diff --git a/llvm/lib/Transforms/Coroutines/CoroSplit.cpp b/llvm/lib/Transforms/Coroutines/CoroSplit.cpp --- a/llvm/lib/Transforms/Coroutines/CoroSplit.cpp +++ b/llvm/lib/Transforms/Coroutines/CoroSplit.cpp @@ -975,6 +975,37 @@ return false; } +// Check whether CI obeys the rules of musttail attribute. +static bool shouldBeMustTail(const CallInst &CI, const Function &F) { + if (CI.isInlineAsm()) + return false; + + // Match prototypes and calling conventions of resume function. + FunctionType *CalleeTy = CI.getFunctionType(); + if (!CalleeTy->getReturnType()->isVoidTy() || (CalleeTy->getNumParams() != 1)) + return false; + + Type *CalleeParmTy = CalleeTy->getParamType(0); + if (!CalleeParmTy->isPointerTy() || + (CalleeParmTy->getPointerAddressSpace() != 0)) + return false; + + if (CI.getCallingConv() != F.getCallingConv()) + return false; + + // CI should not has any ABI-impacting function attributes. + static const Attribute::AttrKind ABIAttrs[] = { + Attribute::StructRet, Attribute::ByVal, Attribute::InAlloca, + Attribute::InReg, Attribute::Returned, Attribute::SwiftSelf, + Attribute::SwiftError, Attribute::Alignment}; + AttributeList Attrs = CI.getAttributes(); + for (auto AK : ABIAttrs) + if (Attrs.hasParamAttribute(0, AK)) + return false; + + return true; +} + // Add musttail to any resume instructions that is immediately followed by a // suspend (i.e. ret). We do this even in -O0 to support guaranteed tail call // for symmetrical coroutine control transfer (C++ Coroutines TS extension). @@ -987,11 +1018,8 @@ SmallVector Resumes; for (auto &I : instructions(F)) if (auto *Call = dyn_cast(&I)) - if (auto *CalledValue = Call->getCalledValue()) - // CoroEarly pass replaced coro resumes with indirect calls to an - // address return by CoroSubFnInst intrinsic. See if it is one of those. - if (isa(CalledValue->stripPointerCasts())) - Resumes.push_back(Call); + if (shouldBeMustTail(*Call, F)) + Resumes.push_back(Call); // Set musttail on those that are followed by a ret instruction. for (CallInst *Call : Resumes) diff --git a/llvm/test/Transforms/Coroutines/coro-split-musttail2.ll b/llvm/test/Transforms/Coroutines/coro-split-musttail2.ll new file mode 100644 --- /dev/null +++ b/llvm/test/Transforms/Coroutines/coro-split-musttail2.ll @@ -0,0 +1,68 @@ +; Tests that coro-split will convert coro.resume followed by a suspend to a +; musttail call. +; RUN: opt < %s -coro-split -S | FileCheck %s +; RUN: opt < %s -passes=coro-split -S | FileCheck %s + +define void @fakeresume1(i8*) { +entry: + ret void; +} + +define void @fakeresume2(i64*) { +entry: + ret void; +} + +define void @g() #0 { +entry: + %id = call token @llvm.coro.id(i32 0, i8* null, i8* null, i8* null) + %alloc = call i8* @malloc(i64 16) #3 + %vFrame = call noalias nonnull i8* @llvm.coro.begin(token %id, i8* %alloc) + + %save = call token @llvm.coro.save(i8* null) + call fastcc void @fakeresume1(i8* null) + + %suspend = call i8 @llvm.coro.suspend(token %save, i1 false) + switch i8 %suspend, label %exit [ + i8 0, label %await.ready + i8 1, label %exit + ] +await.ready: + %save2 = call token @llvm.coro.save(i8* null) + call fastcc void @fakeresume2(i64* null) + + %suspend2 = call i8 @llvm.coro.suspend(token %save2, i1 false) + switch i8 %suspend2, label %exit [ + i8 0, label %exit + i8 1, label %exit + ] +exit: + call i1 @llvm.coro.end(i8* null, i1 false) + ret void +} + +; Verify that in the initial function resume is not marked with musttail. +; CHECK-LABEL: @g( +; CHECK-NOT: musttail call fastcc void @fakeresume1(i8* null) + +; Verify that in the resume part resume call is marked with musttail. +; CHECK-LABEL: @g.resume( +; CHECK: musttail call fastcc void @fakeresume2(i64* null) +; CHECK-NEXT: ret void + +declare token @llvm.coro.id(i32, i8* readnone, i8* nocapture readonly, i8*) #1 +declare i1 @llvm.coro.alloc(token) #2 +declare i64 @llvm.coro.size.i64() #3 +declare i8* @llvm.coro.begin(token, i8* writeonly) #2 +declare token @llvm.coro.save(i8*) #2 +declare i8* @llvm.coro.frame() #3 +declare i8 @llvm.coro.suspend(token, i1) #2 +declare i8* @llvm.coro.free(token, i8* nocapture readonly) #1 +declare i1 @llvm.coro.end(i8*, i1) #2 +declare i8* @llvm.coro.subfn.addr(i8* nocapture readonly, i8) #1 +declare i8* @malloc(i64) + +attributes #0 = { "coroutine.presplit"="1" } +attributes #1 = { argmemonly nounwind readonly } +attributes #2 = { nounwind } +attributes #3 = { nounwind readnone }