Adding a Test to check if Swift refcounting is really working. #35368

nmahendru · 2024-10-17T17:49:01Z

39d68c46d92e77d3a7a19c86d3fd8bc15c666c4f

Adding a Test to check if Swift refcounting is really working.
https://bugs.webkit.org/show_bug.cgi?id=281608
rdar://problem/138056283

Reviewed by NOBODY (OOPS!).

Check that reference count is getting incremented in swift.

* Source/WebGPU/WebGPU.xcodeproj/project.pbxproj:
* Source/WebGPU/WebGPU/TestPlumbing.swift: Added.
(getRefCount(_:)):
* Source/WebGPU/WebGPU/WebGPU.h:
(retainCpp):
(releaseCpp):
* Source/WebGPU/WebGPU/WebGPUExt.h:
* Source/WebGPU/WebGPU/WebGPUSwift.mm: Added.
(WebGPUTest::CppRefCounted::getSwiftRefCount):
* Tools/TestWebKitAPI/Configurations/TestWebKitAPIBase.xcconfig:
* Tools/TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
* Tools/TestWebKitAPI/Tests/WebKitCocoa/WebGPUSwift.mm: Added.
(TEST(WebGPUSwiftTest, CppReferenceCounting)):

1b70bf0

Misc	iOS, visionOS, tvOS & watchOS	macOS	Linux	Windows
❌ 🧪 style	❌ 🛠 ios	❌ 🛠 mac	✅ 🛠 wpe	✅ 🛠 win
✅ 🧪 bindings	❌ 🛠 ios-sim	❌ 🛠 mac-AS-debug	✅ 🧪 wpe-wk2	❌ 🧪 win-tests
✅ 🧪 webkitperl	❌ 🧪 ios-wk2	❌ 🧪 api-mac	✅ 🧪 api-wpe
✅ 🧪 webkitpy	❌ 🧪 ios-wk2-wpt	❌ 🧪 mac-wk1	✅ 🛠 wpe-cairo
✅ 🛠 🧪 jsc	❌ 🧪 api-ios	❌ 🧪 mac-wk2	✅ 🛠 gtk
✅ 🛠 🧪 jsc-arm64	❌ 🛠 vision	❌ 🧪 mac-AS-debug-wk2	⏳ 🧪 gtk-wk2
	❌ 🛠 vision-sim	❌ 🧪 mac-wk2-stress	✅ 🧪 api-gtk
	⏳ 🧪 vision-wk2	❌ 🧪 mac-intel-wk2	✅ 🛠 jsc-armv7
	❌ 🛠 tv		✅ 🧪 jsc-armv7-tests
	❌ 🛠 tv-sim
	❌ 🛠 watch
	❌ 🛠 watch-sim

webkit-early-warning-system · 2024-10-17T17:49:09Z

EWS run on previous version of this PR (hash d0eac26)

Misc	iOS, visionOS, tvOS & watchOS	macOS	Linux	Windows
✅ 🧪 style	❌ 🛠 ios	❌ 🛠 mac	~~🛠 wpe~~	✅ 🛠 win
✅ 🧪 bindings	❌ 🛠 ios-sim	❌ 🛠 mac-AS-debug	~~🧪 wpe-wk2~~	✅ 🧪 win-tests
✅ 🧪 webkitperl	❌ 🧪 ios-wk2	❌ 🧪 api-mac	~~🧪 api-wpe~~
	❌ 🧪 ios-wk2-wpt	❌ 🧪 mac-wk1	~~🛠 wpe-cairo~~
	❌ 🧪 api-ios	❌ 🧪 mac-wk2	✅ 🛠 gtk
	❌ 🛠 vision	❌ 🧪 mac-AS-debug-wk2	⏳ 🧪 gtk-wk2
	❌ 🛠 vision-sim	❌ 🧪 mac-wk2-stress	⏳ 🧪 api-gtk
	⏳ 🧪 vision-wk2	❌ 🧪 mac-intel-wk2
	❌ 🛠 tv
	❌ 🛠 tv-sim
	~~🛠 watch~~
	❌ 🛠 watch-sim

webkit-early-warning-system · 2024-10-17T18:52:08Z

EWS run on previous version of this PR (hash e0d3936)

Misc	iOS, visionOS, tvOS & watchOS	macOS	Linux	Windows
✅ 🧪 style	~~🛠 ios~~	❌ 🛠 mac	~~🛠 wpe~~	~~🛠 win~~
✅ 🧪 bindings	~~🛠 ios-sim~~	❌ 🛠 mac-AS-debug	~~🧪 wpe-wk2~~	~~🧪 win-tests~~
✅ 🧪 webkitperl	~~🧪 ios-wk2~~	❌ 🧪 api-mac	~~🧪 api-wpe~~
	~~🧪 ios-wk2-wpt~~	❌ 🧪 mac-wk1	~~🛠 wpe-cairo~~
	~~🧪 api-ios~~	❌ 🧪 mac-wk2	~~🛠 gtk~~
	~~🛠 vision~~	❌ 🧪 mac-AS-debug-wk2	~~🧪 gtk-wk2~~
	~~🛠 vision-sim~~	❌ 🧪 mac-wk2-stress	~~🧪 api-gtk~~
	⏳ 🧪 vision-wk2	❌ 🧪 mac-intel-wk2
	~~🛠 tv~~
	~~🛠 tv-sim~~
	~~🛠 watch~~
	~~🛠 watch-sim~~

webkit-early-warning-system · 2024-10-17T19:00:49Z

EWS run on previous version of this PR (hash 39d68c4)

Misc	iOS, visionOS, tvOS & watchOS	macOS	Linux	Windows
✅ 🧪 style	✅ 🛠 ios	✅ 🛠 mac	~~🛠 wpe~~	✅ 🛠 win
✅ 🧪 bindings	✅ 🛠 ios-sim	✅ 🛠 mac-AS-debug	~~🧪 wpe-wk2~~	✅ 🧪 win-tests
✅ 🧪 webkitperl	✅ 🧪 ios-wk2	✅ 🧪 api-mac	~~🧪 api-wpe~~
	✅ 🧪 ios-wk2-wpt	✅ 🧪 mac-wk1	✅ 🛠 wpe-cairo
	~~🧪 api-ios~~	✅ 🧪 mac-wk2	✅ 🛠 gtk
	✅ 🛠 vision	✅ 🧪 mac-AS-debug-wk2	⏳ 🧪 gtk-wk2
	✅ 🛠 vision-sim	✅ 🧪 mac-wk2-stress	⏳ 🧪 api-gtk
	✅ 🧪 vision-wk2	✅ 🧪 mac-intel-wk2
	✅ 🛠 tv
	✅ 🛠 tv-sim
	✅ 🛠 watch
	✅ 🛠 watch-sim

rr-codes · 2024-10-17T21:06:19Z

Source/WebGPU/WebGPU/TestPlumbing.swift

+    }
+
+    public static func getRefCountOptimizedOut(obj: WebGPUTest.CppRefCounted) -> UInt32 {
+        return obj.refCountProxy()


the return here can be omitted with recent Swift versions

rr-codes · 2024-10-17T21:07:07Z

Source/WebGPU/WebGPU/TestPlumbing.swift

@@ -0,0 +1,36 @@
+/*
+ * Copyright (C) 2021 Apple Inc. All rights reserved.


the year should be updated

emw-apple · 2024-10-17T21:13:53Z

Source/WebGPU/WebGPU/TestPlumbing.swift

+ */
+
+public class TestPlumbing {
+    public static func getRefCount(obj: WebGPUTest.CppRefCounted) -> UInt32 {


I've seen this in some of your other PRs, but never asked: Is it intentional that you're using the fully qualified <module>.<type> type names in function declarations? Normally I only see fully-qualified types used to disambiguate.

I have never really found a way around it which is not:
typelias CppRefCounted = WebGPUTest.CppRefCounted.
I'll do that for now and ask around as to how C++ namespaces(WebGPUTest) are actually getting imported.

If this is how the interop layer imports them, that's fine with me. I think the typealias would be unnecessary. It's just a little weird since typically in Swift, once you import the module (import WebGPUTest), all these types would be able to be named without being fully-qualified.

rr-codes · 2024-10-17T21:16:21Z

Tools/TestWebKitAPI/Tests/WebKitCocoa/WebGPUSwift.mm

+#if ENABLE(WEBGPU_SWIFT)
+TEST(WebGPUSwiftTest, CppReferenceCounting)
+{
+    auto obj = WebGPUTest::CppRefCounted::create();


Suggested change

auto obj = WebGPUTest::CppRefCounted::create();

Ref obj = WebGPUTest::CppRefCounted::create();

(per smart ptr guidelines)

emw-apple · 2024-10-17T21:19:49Z

Tools/TestWebKitAPI/Configurations/TestWebKitAPIBase.xcconfig


 FRAMEWORK_SEARCH_PATHS = $(FRAMEWORK_SEARCH_PATHS_$(WK_COCOA_TOUCH));
 FRAMEWORK_SEARCH_PATHS_ = $(inherited) $(SYSTEM_LIBRARY_DIR)/PrivateFrameworks $(SYSTEM_LIBRARY_DIR)/Frameworks/WebKit.framework/Versions/A/Frameworks;
 FRAMEWORK_SEARCH_PATHS_cocoatouch = $(inherited);

 PROJECT_HEADER_SEARCH_PATHS = $(SRCROOT)/../../Source/WebKit/Platform $(SRCROOT)/../../Source/WebKit/Platform/IPC $(SRCROOT)/../../Source/WebKit/Platform/IPC/darwin $(SRCROOT)/../../Source/WebKit/Platform/IPC/cocoa $(SRCROOT)/../../Source/WebKit/Shared $(SRCROOT)/../../Source/WebKit/Shared/Cocoa $(SRCROOT)/../../Source/WebKit/Shared/cf $(SRCROOT)/../../Source/WebKit/Shared/API/C $(BUILT_PRODUCTS_DIR)/DerivedSources/WebKit $(inherited);
-
+HEADER_SEARCH_PATHS = $(inherited) $(OBJROOT)/WebGPU.build/WebGPU.build/DerivedSources $(OBJROOT)/WebGPU.build/$(CONFIGURATION)$(EFFECTIVE_PLATFORM_NAME)/WebGPU.build/DerivedSource


Could you remind me if we have precedent for reaching into other projects' intermediate build dir like this? It's pretty gross, and I worry that you are making assumptions about build directory structure that won't always be true (e.g. for a person who uses the Xcode-default derived data dir).

There is a precedent in PAL<=>WebCore. and yes, It is gross.
Although, I think for this PR this is not even needed anymore. Sorry for that. It's just a remnant of how I was thinking of doing this earlier which I should have removed.

webkit-early-warning-system · 2024-10-17T22:42:43Z

EWS run on previous version of this PR (hash d4b414f)

Misc	iOS, visionOS, tvOS & watchOS	macOS	Linux	Windows
✅ 🧪 style	✅ 🛠 ios	✅ 🛠 mac	✅ 🛠 wpe	✅ 🛠 win
✅ 🧪 bindings	✅ 🛠 ios-sim	✅ 🛠 mac-AS-debug	✅ 🧪 wpe-wk2	✅ 🧪 win-tests
✅ 🧪 webkitperl	✅ 🧪 ios-wk2	✅ 🧪 api-mac	✅ 🧪 api-wpe
	✅ 🧪 ios-wk2-wpt	✅ 🧪 mac-wk1	✅ 🛠 wpe-cairo
	✅ 🧪 api-ios	✅ 🧪 mac-wk2	✅ 🛠 gtk
	✅ 🛠 vision	❌ 🧪 mac-AS-debug-wk2	✅ 🧪 gtk-wk2
	✅ 🛠 vision-sim	✅ 🧪 mac-wk2-stress	✅ 🧪 api-gtk
	✅ 🧪 vision-wk2	✅ 🧪 mac-intel-wk2
	✅ 🛠 tv
	✅ 🛠 tv-sim
	✅ 🛠 watch
	✅ 🛠 watch-sim

geoffreygaren · 2024-10-18T16:32:28Z

Source/WebGPU/WebGPU/WebGPUSwift.mm

+unsigned WebGPUTest::CppRefCounted::getSwiftRefCount()
+{
+    return WebGPU::TestPlumbing::getRefCount(this);
+}
+
+unsigned WebGPUTest::CppRefCounted::getSwiftRefCountOptimizedOut()
+{
+    return WebGPU::TestPlumbing::getRefCountOptimizedOut(this);
+}


Is this set of wrappers required? I don't think it is. Empty abstraction makes code hard to read and edit.

I want to expose the swift functions out of WebGPU.framework as a C++ API so that I can call it in TestWebKitAPI. Do you know of another way to do it ? If I directly want to call TestPlumbing then I will have to enable Interop in the test binary also.. which I really did not want to do.

Why do we want to avoid enabling interop in our unit tests? Don't we want to unit test interop?

I think we want to unit test interop. And that's why this PR.
I just did not want to turn on interop in TestWebKitAPI yet unless we really have a pressing need for it. Usually, enabling interop involves yet another series of #ifndef __swift__ which just adds debt to cleanup later.
If you strongly feel that we should just enable interop in TestWebKitAPI now, then I can work on that.

geoffreygaren · 2024-10-18T16:33:40Z

Source/WebGPU/WebGPU/WebGPU.h

+
+inline void retainCpp(WebGPUTest::CppRefCounted* obj)
+{
+    WTF::retainRefCounted(obj);


You should rename WTF::retainRefCounted to ref(). Having two names for one thing is confusing. Also, you can remove the RELEASE_ASSERT, static_assert, and static_cast. Just have it call object->ref().

There are three of these:

RefCounted

ThreadSafeRefCounted

ThreadSafeRefCountedAndCanMakeThreadSafeWeakPtr

All three have Global wrappers(which Swift mandates). So I cannot rename all of them as ref ?

Also, curious if you are asking me to remove the asserts for perfomance ? I have just added them to make sure I am calling the right ref

I'm saying something way more basic: If obj is a refcounted WebKit object, it has a member function named 'ref'.

I want to remove the cast because it's not type-safe, and not needed. I want to remove the asserts because they're just there to guard the cast.

geoffreygaren · 2024-10-18T16:35:31Z

Source/WebGPU/WebGPU/WebGPU.h

+
+}
+
+inline void retainCpp(WebGPUTest::CppRefCounted* obj)


I would call this function 'ref', since it's a free function whose job is to call the 'ref' member function.

But also: If you make a free function in WTF, template<typename T> void ref(T* object) { object->ref(); } you may not need a separate function here.

I tried that when I was adding it to the RefCounted classes but templates don't work with swift.

Oh, right! So we do need separate function overloads per class, but we can still make them free functions named ref: void ref(WebGPUTest::CppRefCounted* object) { object->ref(); }

geoffreygaren · 2024-10-18T16:37:03Z

Source/WebGPU/WebGPU/WebGPU.h

+
+inline void releaseCpp(WebGPUTest::CppRefCounted* obj)
+{
+    WTF::releaseRefCounted(obj);


geoffreygaren · 2024-10-18T16:38:05Z

Source/WebGPU/WebGPU/WebGPU.h

+    unsigned refCountProxy() { return refCount(); }
+    WGPU_EXPORT unsigned getSwiftRefCount();
+    WGPU_EXPORT unsigned getSwiftRefCountOptimizedOut();


Why do we need any of this abstraction? Can't the caller just call refCount()?

to expose the API in Swift in C++ and use it in TestWebKitAPI without enabling interop there.

geoffreygaren · 2024-10-18T16:38:17Z

Source/WebGPU/WebGPU/WebGPU.h

+#include <wtf/RetainReleaseSwift.h>
+#if ENABLE(WEBGPU_SWIFT)
+namespace WebGPUTest {
+class CppRefCounted: public RefCounted<CppRefCounted> {


Can this class move to TestWebKitAPI?

Only if I enable interop there which I would rather not do and avoid the complexity.

If we want to write unit tests for Swift, we're going to need to make Swift unit testing possible.

Would it help to make Swift unit testing a separate build target? TestWebKitAPI supports that, and already has TestIPC, TestWGSL, TestWTF, etc.

geoffreygaren · 2024-10-18T16:40:49Z

Source/WebGPU/WebGPU/WebGPU.h

+    // Swift is not importing const methods and RefCount::refCount() is const
+    unsigned refCountProxy() { return refCount(); }


You can make an #if __swift__ non-const overload for refCount() in the RefCounted class, with a FIXME and a Radar for not importing const methods.

geoffreygaren · 2024-10-18T16:43:15Z

Source/WebGPU/WebGPU/TestPlumbing.swift

+public class TestPlumbing {
+    public static func getRefCount(obj: CppRefCounted) -> UInt32 {
+        // Just an API call to ensure swift does not optimize this out.
+        let refCount = obj.refCountProxy()
+        let _ = "The ref count is \(refCount)"
+        return refCount
+    }
+
+    public static func getRefCountOptimizedOut(obj: CppRefCounted) -> UInt32 {
+        obj.refCountProxy()
+    }


This seems kinda brittle. I don't think we need it?

The goal is to determine that the object is allocated with the right refcount, right? If so, we want to export a Swift function that will allocate and return the object, and then we can test the refcount of the result directly in our caller TestWebKitAPI code.

hmm.. okie.. I can remove that. I just wanted to highlight the optimization via a test but it's not important.

webkit-early-warning-system · 2024-10-19T01:00:23Z

EWS run on previous version of this PR (hash fff4bc9)

Misc	iOS, visionOS, tvOS & watchOS	macOS	Linux	Windows
❌ 🧪 style	~~🛠 ios~~	❌ 🛠 mac	~~🛠 wpe~~	~~🛠 win~~
✅ 🧪 bindings	❌ 🛠 ios-sim	~~🛠 mac-AS-debug~~	~~🧪 wpe-wk2~~	~~🧪 win-tests~~
✅ 🧪 webkitperl	❌ 🧪 ios-wk2	❌ 🧪 api-mac	~~🧪 api-wpe~~
✅ 🧪 webkitpy	❌ 🧪 ios-wk2-wpt	❌ 🧪 mac-wk1	~~🛠 wpe-cairo~~
~~🛠 🧪 jsc~~	❌ 🧪 api-ios	❌ 🧪 mac-wk2	~~🛠 gtk~~
✅ 🛠 🧪 jsc-arm64	~~🛠 vision~~	~~🧪 mac-AS-debug-wk2~~	~~🧪 gtk-wk2~~
	~~🛠 vision-sim~~	❌ 🧪 mac-wk2-stress	~~🧪 api-gtk~~
	⏳ 🧪 vision-wk2	❌ 🧪 mac-intel-wk2	✅ 🛠 jsc-armv7
	~~🛠 tv~~		✅ 🧪 jsc-armv7-tests
	~~🛠 tv-sim~~
	~~🛠 watch~~
	~~🛠 watch-sim~~

webkit-early-warning-system · 2024-10-19T01:40:48Z

EWS run on previous version of this PR (hash ef20a4e)

Misc	iOS, visionOS, tvOS & watchOS	macOS	Linux	Windows
❌ 🧪 style	❌ 🛠 ios	❌ 🛠 mac	✅ 🛠 wpe	✅ 🛠 win
✅ 🧪 bindings	❌ 🛠 ios-sim	❌ 🛠 mac-AS-debug	✅ 🧪 wpe-wk2	✅ 🧪 win-tests
✅ 🧪 webkitperl	❌ 🧪 ios-wk2	❌ 🧪 api-mac	✅ 🧪 api-wpe
✅ 🧪 webkitpy	❌ 🧪 ios-wk2-wpt	❌ 🧪 mac-wk1	✅ 🛠 wpe-cairo
✅ 🛠 🧪 jsc	❌ 🧪 api-ios	❌ 🧪 mac-wk2	✅ 🛠 gtk
✅ 🛠 🧪 jsc-arm64	❌ 🛠 vision	❌ 🧪 mac-AS-debug-wk2	✅ 🧪 gtk-wk2
	❌ 🛠 vision-sim	❌ 🧪 mac-wk2-stress	✅ 🧪 api-gtk
	⏳ 🧪 vision-wk2	❌ 🧪 mac-intel-wk2	✅ 🛠 jsc-armv7
	❌ 🛠 tv	🛠 mac-safer-cpp	✅ 🧪 jsc-armv7-tests
	❌ 🛠 tv-sim
	❌ 🛠 watch
	❌ 🛠 watch-sim

webkit-early-warning-system · 2024-11-04T19:00:58Z

EWS run on previous version of this PR (hash 3fbe91f)

Misc	iOS, visionOS, tvOS & watchOS	macOS	Linux	Windows
❌ 🧪 style	❌ 🛠 ios	❌ 🛠 mac	~~🛠 wpe~~	~~🛠 win~~
✅ 🧪 bindings	~~🛠 ios-sim~~	❌ 🛠 mac-AS-debug	~~🧪 wpe-wk2~~	~~🧪 win-tests~~
✅ 🧪 webkitperl	~~🧪 ios-wk2~~	❌ 🧪 api-mac	~~🧪 api-wpe~~
✅ 🧪 webkitpy	~~🧪 ios-wk2-wpt~~	❌ 🧪 mac-wk1	~~🛠 wpe-cairo~~
✅ 🛠 🧪 jsc	~~🧪 api-ios~~	❌ 🧪 mac-wk2	✅ 🛠 gtk
~~🛠 🧪 jsc-arm64~~	~~🛠 vision~~	❌ 🧪 mac-AS-debug-wk2	~~🧪 gtk-wk2~~
	~~🛠 vision-sim~~	❌ 🧪 mac-wk2-stress	~~🧪 api-gtk~~
	⏳ 🧪 vision-wk2	❌ 🧪 mac-intel-wk2	✅ 🛠 jsc-armv7
	~~🛠 tv~~		⏳ 🧪 jsc-armv7-tests
	~~🛠 tv-sim~~
	~~🛠 watch~~
	~~🛠 watch-sim~~

https://bugs.webkit.org/show_bug.cgi?id=281608 rdar://problem/138056283 Reviewed by NOBODY (OOPS!). Check that reference count is getting incremented in swift. * Source/WTF/wtf/RefCounted.h: (WTF::RefCountedBase::ref): (WTF::RefCountedBase::ref const): Deleted. * Source/WebGPU/WebGPU/Buffer.mm: (WebGPU::Buffer::copy): * Source/WebGPU/WebGPU/WebGPUExt.h: * Tools/Scripts/run-api-tests: (parse_args): * Tools/Scripts/webkitpy/port/darwin.py: (DarwinPort): * Tools/TestWebKitAPI/Configurations/TestWebGPU.xcconfig: Added. * Tools/TestWebKitAPI/Configurations/TestWebKitAPIBase.xcconfig: * Tools/TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: * Tools/TestWebKitAPI/Tests/WebGPU/RefCountTests.mm: Added. (TEST(TestWebGPU, ReferenceCounting)): * Tools/TestWebKitAPI/Tests/WebGPU/TestWebGPU.h: Added. (refCpp): (derefCpp): * Tools/TestWebKitAPI/Tests/WebGPU/TestWebGPU.mm: Added. (Cpp::CppRefCounted::getSwiftRefCount): * Tools/TestWebKitAPI/Tests/WebGPU/TestWebGPU.modulemap: Added. * Tools/TestWebKitAPI/Tests/WebGPU/Tests.swift: Added. (getRefCount(_:)):

webkit-early-warning-system · 2024-11-04T19:19:27Z

EWS run on current version of this PR (hash 1b70bf0)

Misc	iOS, visionOS, tvOS & watchOS	macOS	Linux	Windows
❌ 🧪 style	❌ 🛠 ios	❌ 🛠 mac	✅ 🛠 wpe	✅ 🛠 win
✅ 🧪 bindings	❌ 🛠 ios-sim	❌ 🛠 mac-AS-debug	✅ 🧪 wpe-wk2	❌ 🧪 win-tests
✅ 🧪 webkitperl	❌ 🧪 ios-wk2	❌ 🧪 api-mac	✅ 🧪 api-wpe
✅ 🧪 webkitpy	❌ 🧪 ios-wk2-wpt	❌ 🧪 mac-wk1	✅ 🛠 wpe-cairo
✅ 🛠 🧪 jsc	❌ 🧪 api-ios	❌ 🧪 mac-wk2	✅ 🛠 gtk
✅ 🛠 🧪 jsc-arm64	❌ 🛠 vision	❌ 🧪 mac-AS-debug-wk2	⏳ 🧪 gtk-wk2
	❌ 🛠 vision-sim	❌ 🧪 mac-wk2-stress	✅ 🧪 api-gtk
	⏳ 🧪 vision-wk2	❌ 🧪 mac-intel-wk2	✅ 🛠 jsc-armv7
	❌ 🛠 tv		✅ 🧪 jsc-armv7-tests
	❌ 🛠 tv-sim
	❌ 🛠 watch
	❌ 🛠 watch-sim

nmahendru self-assigned this Oct 17, 2024

nmahendru added the WebGPU For bugs in WebGPU label Oct 17, 2024

webkit-ews-buildbot added the merging-blocked Applied to prevent a change from being merged label Oct 17, 2024

nmahendru removed the merging-blocked Applied to prevent a change from being merged label Oct 17, 2024

webkit-ews-buildbot added the merging-blocked Applied to prevent a change from being merged label Oct 17, 2024

nmahendru force-pushed the eng/WebGPU-Swift-Reference-Counting-Tests branch from d0eac26 to e0d3936 Compare October 17, 2024 18:52

nmahendru removed the merging-blocked Applied to prevent a change from being merged label Oct 17, 2024

nmahendru force-pushed the eng/WebGPU-Swift-Reference-Counting-Tests branch from e0d3936 to 39d68c4 Compare October 17, 2024 19:00

nmahendru marked this pull request as ready for review October 17, 2024 21:01

nmahendru requested review from tadeuzagallo, mwyrzykowski and emw-apple as code owners October 17, 2024 21:01

rr-codes reviewed Oct 17, 2024

View reviewed changes

emw-apple reviewed Oct 17, 2024

View reviewed changes

rr-codes reviewed Oct 17, 2024

View reviewed changes

emw-apple reviewed Oct 17, 2024

View reviewed changes

nmahendru force-pushed the eng/WebGPU-Swift-Reference-Counting-Tests branch from 39d68c4 to d4b414f Compare October 17, 2024 22:42

webkit-ews-buildbot added the merging-blocked Applied to prevent a change from being merged label Oct 18, 2024

nmahendru requested review from rr-codes and emw-apple October 18, 2024 03:15

geoffreygaren requested changes Oct 18, 2024

View reviewed changes

nmahendru force-pushed the eng/WebGPU-Swift-Reference-Counting-Tests branch from d4b414f to fff4bc9 Compare October 19, 2024 01:00

nmahendru requested a review from JonWBedard as a code owner October 19, 2024 01:00

nmahendru force-pushed the eng/WebGPU-Swift-Reference-Counting-Tests branch from fff4bc9 to ef20a4e Compare October 19, 2024 01:40

nmahendru force-pushed the eng/WebGPU-Swift-Reference-Counting-Tests branch from ef20a4e to 3fbe91f Compare November 4, 2024 19:00

nmahendru force-pushed the eng/WebGPU-Swift-Reference-Counting-Tests branch from 3fbe91f to 1b70bf0 Compare November 4, 2024 19:19

mwyrzykowski marked this pull request as draft June 30, 2025 22:20

		@@ -0,0 +1,36 @@
		/*
		* Copyright (C) 2021 Apple Inc. All rights reserved.

	auto obj = WebGPUTest::CppRefCounted::create();
	Ref obj = WebGPUTest::CppRefCounted::create();

		// Swift is not importing const methods and RefCount::refCount() is const
		unsigned refCountProxy() { return refCount(); }

Adding a Test to check if Swift refcounting is really working. #35368

Are you sure you want to change the base?

Adding a Test to check if Swift refcounting is really working. #35368

Uh oh!

Conversation

nmahendru commented Oct 17, 2024 • edited by webkit-early-warning-system Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

39d68c46d92e77d3a7a19c86d3fd8bc15c666c4f

Uh oh!

webkit-early-warning-system commented Oct 17, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

webkit-early-warning-system commented Oct 17, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

webkit-early-warning-system commented Oct 17, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

webkit-early-warning-system commented Oct 17, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

webkit-early-warning-system commented Oct 19, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

webkit-early-warning-system commented Oct 19, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

nmahendru commented Oct 17, 2024 •

edited by webkit-early-warning-system

Loading

webkit-early-warning-system commented Oct 17, 2024 •

edited

Loading

webkit-early-warning-system commented Oct 17, 2024 •

edited

Loading

webkit-early-warning-system commented Oct 17, 2024 •

edited

Loading

webkit-early-warning-system commented Oct 17, 2024 •

edited

Loading

webkit-early-warning-system commented Oct 19, 2024 •

edited

Loading

webkit-early-warning-system commented Oct 19, 2024 •

edited

Loading

webkit-early-warning-system commented Nov 4, 2024 •

edited

Loading

webkit-early-warning-system commented Nov 4, 2024 •

edited

Loading