[Webkit-unassigned] [Bug 234008] ANGLE Metal framebuffer object clear performance regression

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Wed Dec 15 12:32:49 PST 2021


https://bugs.webkit.org/show_bug.cgi?id=234008

Kimmo Kinnunen <kkinnunen at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Blocks|                            |231180

--- Comment #2 from Kimmo Kinnunen <kkinnunen at apple.com> ---
I can reproduce slight regressions in perf of iPhone 11, iPad Pro 2015, nothing as drastic as reported. iPhone 11 fps == ~30, iPad Pro 2015 fps == ~10

All samples under 
Sample Count, Samples %, Normalized CPU %, Symbol
209, 4.4%, 3.2%, WebCore::GraphicsContextGLANGLE::clear(unsigned int) (in WebCore)
209, 4.4%, 3.2%,     gl::Clear(unsigned int) (in libANGLE-shared.dylib)
173, 3.6%, 2.6%,         gl::Context::clear(unsigned int) (in libANGLE-shared.dylib)
169, 3.5%, 2.6%,             gl::Context::syncState(angle::IterableBitSet<63ul> const&, angle::BitSetT<12ul, unsigned int, unsigned long> const&, gl::Command) (in libANGLE-shared.dylib)
166, 3.5%, 2.5%,                 rx::ContextMtl::syncState(gl::Context const*, angle::IterableBitSet<63ul> const&, angle::IterableBitSet<63ul> const&) (in libANGLE-shared.dylib)
165, 3.5%, 2.5%,                     rx::ContextMtl::onDrawFrameBufferChangedState(gl::Context const*, rx::FramebufferMtl*, bool) (in libANGLE-shared.dylib)
165, 3.5%, 2.5%,                         rx::ContextMtl::endEncoding(bool) (in libANGLE-shared.dylib)
165, 3.5%, 2.5%,                             rx::ContextMtl::endRenderEncoding(rx::mtl::RenderCommandEncoder*) (in libANGLE-shared.dylib)
163, 3.4%, 2.5%,                                 rx::mtl::RenderCommandEncoder::endEncodingImpl(bool) (in libANGLE-shared.dylib)


Referenced Bugs:

https://bugs.webkit.org/show_bug.cgi?id=231180
[Bug 231180] ANGLE Metal backend regressions metabug
-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.webkit.org/pipermail/webkit-unassigned/attachments/20211215/d6b1c33b/attachment-0001.htm>


More information about the webkit-unassigned mailing list