-
Notifications
You must be signed in to change notification settings - Fork 12.5k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Fix Objective-C++ Sret of non-trivial data types on Windows ARM64 #88671
Conversation
@llvm/pr-subscribers-clang-codegen @llvm/pr-subscribers-clang Author: Hugo Melder (hmelder) ChangesLinked to gnustep/libobjc2#289. More information can be found in issue: #88273. My solution involves creating a new message-send function for this calling convention when targeting MSVC. Additional information is available in the libobjc2 pull request. I am unsure whether we should check for a runtime version where objc_msgSend_stret2_np is guaranteed to be present or leave it as is, considering it remains a critical bug. What are your thoughts about this @davidchisnall? Full diff: https://github.com/llvm/llvm-project/pull/88671.diff 4 Files Affected:
diff --git a/clang/lib/CodeGen/CGCall.cpp b/clang/lib/CodeGen/CGCall.cpp
index 7a0bc6fa77b889..5ea26b0f594815 100644
--- a/clang/lib/CodeGen/CGCall.cpp
+++ b/clang/lib/CodeGen/CGCall.cpp
@@ -1585,6 +1585,11 @@ bool CodeGenModule::ReturnTypeUsesSRet(const CGFunctionInfo &FI) {
return RI.isIndirect() || (RI.isInAlloca() && RI.getInAllocaSRet());
}
+bool CodeGenModule::ReturnTypeHasInReg(const CGFunctionInfo &FI) {
+ const auto &RI = FI.getReturnInfo();
+ return RI.getInReg();
+}
+
bool CodeGenModule::ReturnSlotInterferesWithArgs(const CGFunctionInfo &FI) {
return ReturnTypeUsesSRet(FI) &&
getTargetCodeGenInfo().doesReturnSlotInterfereWithArgs();
diff --git a/clang/lib/CodeGen/CGObjCGNU.cpp b/clang/lib/CodeGen/CGObjCGNU.cpp
index 4e7f777ba1d916..5faecc6af33556 100644
--- a/clang/lib/CodeGen/CGObjCGNU.cpp
+++ b/clang/lib/CodeGen/CGObjCGNU.cpp
@@ -2911,12 +2911,25 @@ CGObjCGNU::GenerateMessageSend(CodeGenFunction &CGF,
"objc_msgSend_fpret")
.getCallee();
} else if (CGM.ReturnTypeUsesSRet(MSI.CallInfo)) {
+ StringRef name = "objc_msgSend_stret";
+
+ // The address of the memory block is be passed in x8 for POD type,
+ // or in x0 for non-POD type (marked as inreg).
+ bool shouldCheckForInReg =
+ CGM.getContext()
+ .getTargetInfo()
+ .getTriple()
+ .isWindowsMSVCEnvironment() &&
+ CGM.getContext().getTargetInfo().getTriple().isAArch64();
+ if (shouldCheckForInReg && CGM.ReturnTypeHasInReg(MSI.CallInfo)) {
+ name = "objc_msgSend_stret2_np";
+ }
+
// The actual types here don't matter - we're going to bitcast the
// function anyway
- imp =
- CGM.CreateRuntimeFunction(llvm::FunctionType::get(IdTy, IdTy, true),
- "objc_msgSend_stret")
- .getCallee();
+ imp = CGM.CreateRuntimeFunction(
+ llvm::FunctionType::get(IdTy, IdTy, true), name)
+ .getCallee();
} else {
imp = CGM.CreateRuntimeFunction(
llvm::FunctionType::get(IdTy, IdTy, true), "objc_msgSend")
diff --git a/clang/lib/CodeGen/CodeGenModule.h b/clang/lib/CodeGen/CodeGenModule.h
index 1cc447765e2c97..be43a18fc60856 100644
--- a/clang/lib/CodeGen/CodeGenModule.h
+++ b/clang/lib/CodeGen/CodeGenModule.h
@@ -1241,6 +1241,9 @@ class CodeGenModule : public CodeGenTypeCache {
/// Return true iff the given type uses 'sret' when used as a return type.
bool ReturnTypeUsesSRet(const CGFunctionInfo &FI);
+ /// Return true iff the given type has `inreg` set.
+ bool ReturnTypeHasInReg(const CGFunctionInfo &FI);
+
/// Return true iff the given type uses an argument slot when 'sret' is used
/// as a return type.
bool ReturnSlotInterferesWithArgs(const CGFunctionInfo &FI);
diff --git a/clang/test/CodeGenObjCXX/msabi-stret-arm64.mm b/clang/test/CodeGenObjCXX/msabi-stret-arm64.mm
new file mode 100644
index 00000000000000..0a17b5862b338e
--- /dev/null
+++ b/clang/test/CodeGenObjCXX/msabi-stret-arm64.mm
@@ -0,0 +1,77 @@
+// RUN: %clang_cc1 -triple aarch64-pc-windows-msvc -fobjc-runtime=gnustep-2.2 -fobjc-dispatch-method=non-legacy -emit-llvm -o - %s | FileCheck %s
+
+// Pass and return for type size <= 8 bytes.
+struct S1 {
+ int a[2];
+};
+
+// Pass and return hfa <= 8 bytes
+struct F1 {
+ float a[2];
+};
+
+// Pass and return for type size > 16 bytes.
+struct S2 {
+ int a[5];
+};
+
+// Pass and return aggregate (of size < 16 bytes) with non-trivial destructor.
+// Sret and inreg: Returned in x0
+struct S3 {
+ int a[3];
+ ~S3();
+};
+S3::~S3() {
+}
+
+
+@interface MsgTest { id isa; } @end
+@implementation MsgTest
+- (S1) smallS1 {
+ S1 x;
+ x.a[0] = 0;
+ x.a[1] = 1;
+ return x;
+
+}
+- (F1) smallF1 {
+ F1 x;
+ x.a[0] = 0.2f;
+ x.a[1] = 0.5f;
+ return x;
+}
+- (S2) stretS2 {
+ S2 x;
+ for (int i = 0; i < 5; i++) {
+ x.a[i] = i;
+ }
+ return x;
+}
+- (S3) stretInRegS3 {
+ S3 x;
+ for (int i = 0; i < 3; i++) {
+ x.a[i] = i;
+ }
+ return x;
+}
++ (S3) msgTestStretInRegS3 {
+ S3 x;
+ for (int i = 0; i < 3; i++) {
+ x.a[i] = i;
+ }
+ return x;
+}
+@end
+
+void test0(MsgTest *t) {
+ // CHECK: call {{.*}} @objc_msgSend
+ S1 ret = [t smallS1];
+ // CHECK: call {{.*}} @objc_msgSend
+ F1 ret2 = [t smallF1];
+ // CHECK: call {{.*}} @objc_msgSend_stret
+ S2 ret3 = [t stretS2];
+ // CHECK: call {{.*}} @objc_msgSend_stret2_np
+ S3 ret4 = [t stretInRegS3];
+ // CHECK: call {{.*}} @objc_msgSend_stret2_np
+ S3 ret5 = [MsgTest msgTestStretInRegS3];
+}
|
Seems like the formatting error in buildkite is unrelated to this PR. |
✅ With the latest revision this PR passed the C/C++ code formatter. |
@davidchisnall thank you for reviewing the PR. Do you have merge permissions? |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@davidchisnall Is this patch waiting for something? (I got a request to check on the progress of this patch.)
de9e216
to
b4d0255
Compare
…vm#88671) Linked to gnustep/libobjc2#289. More information can be found in issue: llvm#88273. My solution involves creating a new message-send function for this calling convention when targeting MSVC. Additional information is available in the libobjc2 pull request. I am unsure whether we should check for a runtime version where objc_msgSend_stret2_np is guaranteed to be present or leave it as is, considering it remains a critical bug. What are your thoughts about this @davidchisnall? (cherry picked from commit 3dcd2cc)
…vm#88671) Linked to gnustep/libobjc2#289. More information can be found in issue: llvm#88273. My solution involves creating a new message-send function for this calling convention when targeting MSVC. Additional information is available in the libobjc2 pull request. I am unsure whether we should check for a runtime version where objc_msgSend_stret2_np is guaranteed to be present or leave it as is, considering it remains a critical bug. What are your thoughts about this @davidchisnall? (cherry picked from commit 3dcd2cc)
Linked to gnustep/libobjc2#289.
More information can be found in issue: #88273.
My solution involves creating a new message-send function for this calling convention when targeting MSVC. Additional information is available in the libobjc2 pull request.
I am unsure whether we should check for a runtime version where objc_msgSend_stret2_np is guaranteed to be present or leave it as is, considering it remains a critical bug. What are your thoughts about this @davidchisnall?