src


Log

Author Commit Date CI Message
Charlie Lao 8b6b1113 2024-08-30T16:06:17 Vulkan: Fallback to host unCached if no cached memory available Android studio may not have any host cached memory. This CL will make it use unCached memory in this case. User of this memory will still treat it like cached and try to flush/invalidate cache, but these flush/invalidate call will early out, so they should still work. Performance wont be good, but that is the best we can do to make it work instead of crash. Bug: b/363347052 Change-Id: I68ac917585f6f5672a75085c397858bdc28687e9 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5828132 Reviewed-by: Lars Harrison <slumpwuffle@google.com> Auto-Submit: Charlie Lao <cclao@google.com> Reviewed-by: Yuxin Hu <yuxinhu@google.com> Commit-Queue: Yuxin Hu <yuxinhu@google.com>
Cody Northrop 21b286a1 2024-08-28T15:06:10 Trace replay: Call glFinish before screenshots On some platforms we're getting partial or no results in the screenshot, even though it ends up displaying correctly on screen. Adding a glFinish before glReadPixels clears up the issue. Bug: b/42263221 Change-Id: Id5d2384c01fece1f7cd4d79b7cc599d082fb2864 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5824180 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Roman Lavrov <romanl@google.com> Auto-Submit: Cody Northrop <cnorthrop@google.com>
Charlie Lao 20e0b83a 2024-08-30T11:05:09 Vulkan: Expand preferSubmitAtFBOBoundary from ARM to all tiler GPUs On S24, enable preferSubmitAtFBOBoundary flag boost black_clover_m frame time significantly. The app uses occlusion query which was blocking the CPU to encode next frame until one of the previous frame's render pass is completed. Enable preferSubmitAtFBOBoundary makes the render pass being submitted early which reduces the query wait time, which in turn boost GPU utilization and bumps GPU clock. This reduces frame time of black_clover_m from 23.8ms to 6.4ms, even much better than native's 16ms. It could impact power usage as well, a little bit harder to tell without data that it will be better or worse. For now, I am enabling this for all tiler based GPUs and we can re-evaluate if any problems comes up. This would potentially also benefit other desktop GPU as well, but I am not touching these until we have data showing the benefit before enabling it. Bug: b/362555815 Change-Id: Ie7a5231350389c49091bb2771c542b4439c9c227 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5829781 Reviewed-by: Yuxin Hu <yuxinhu@google.com> Commit-Queue: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Cody Northrop 56ea2edc 2024-08-30T11:01:48 Traces: Update regex for power rails Bug: b/354006406 Change-Id: If014c49ccdbb6e0a57c0dc88833511049c3578ea Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5828131 Reviewed-by: Roman Lavrov <romanl@google.com> Commit-Queue: Cody Northrop <cnorthrop@google.com>
Charlie Lao e4be31ca 2024-08-29T14:30:11 Test: Add VertexAttribPointer test with different offset The first test draw with VertexAttribPointer with smaller offset and then followed by draw with VertexAttribPointer with larger offset. The second test draw with larger offset and then followed by draw with smaller offset. They will exercise logic in backend that may reuse the conversion buffers (when forced format fallback is used). Bug: b/357622380 Change-Id: If325a0547e5f80305125f448e9f987896f3e35bd Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5827285 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Yuxin Hu <yuxinhu@google.com> Commit-Queue: Charlie Lao <cclao@google.com>
Shahbaz Youssefi 4167a9a0 2024-08-29T23:38:05 Don't lock the context mutex if no flush in eglClientWaitSync Bug: b/362604439 Change-Id: Ic514bcb3824514b5fd82ebb14ab97286aeb6557c Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5828262 Reviewed-by: Igor Nazarov <i.nazarov@samsung.com> Reviewed-by: Charlie Lao <cclao@google.com> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Charlie Lao 4e3b4e0c 2024-08-21T16:56:47 Vulkan: Clear conversion buffer's dirtyRange if bufferData called If BufferData is called, the previous dirtyRange on the conversion buffer is no longer valid (the original data in BufferVk is no longer valid). This CL clears the dirty range in conversion buffers so that they do not gets unnecessarily converted. Bug: b/357622380 Change-Id: Ic97c867a8d2141659a0ffd269049f9d4168a421f Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5804531 Reviewed-by: Yuxin Hu <yuxinhu@google.com> Commit-Queue: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Shahbaz Youssefi 27a0a5ed 2024-08-29T23:23:19 Vulkan: Only access ContextVk in EGLSync if asked to flush Bug: b/362604439 Change-Id: Ifec413628de88e9048e8134c2c5e9497970f358c Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5828261 Reviewed-by: Igor Nazarov <i.nazarov@samsung.com> Reviewed-by: Charlie Lao <cclao@google.com> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Charlie Lao e4aab5cc 2024-08-13T14:54:19 Vulkan: Reuse the same conversion buffer by adjusting offset In some usage case where app calls bufferSubData and then calls glVertexAttribPointer to source vertex from different offset, we could still just reuse the existing VertexConversionBuffer instead of create a new one. The benefit with this is that all previous converted data are still valid, thus reduce the overall conversion, and reduce the number of conversion buffers and saves memory. Bug: b/357622380 Change-Id: Ifcf626427e2ed500d6ab617541b78a60839d9acd Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5785670 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Charlie Lao <cclao@google.com> Reviewed-by: Yuxin Hu <yuxinhu@google.com>
Yuly Novikov 4dfba10a 2024-08-30T15:05:44 Skip passmark_complex on Linux Intel Bug: b/362801312 Change-Id: I743261c7f5086d5aaf98d28c1b32fa494c0fc2e1 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5822715 Commit-Queue: Cody Northrop <cnorthrop@google.com> Reviewed-by: Cody Northrop <cnorthrop@google.com> Auto-Submit: Yuly Novikov <ynovikov@chromium.org> Commit-Queue: Yuly Novikov <ynovikov@chromium.org>
Shahbaz Youssefi 8d6f6c23 2024-08-29T14:56:31 Vulkan: Unsuppress fixed VVL errors Bug: angleproject:361600662 Change-Id: Ic4585f4b38261f69700a9703bdb10b60e18f638a Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5826762 Reviewed-by: Yuly Novikov <ynovikov@chromium.org> Commit-Queue: Yuly Novikov <ynovikov@chromium.org> Auto-Submit: Shahbaz Youssefi <syoussefi@chromium.org>
Alexey Knyazev d886a0fb 2024-08-29T00:00:00 Metal: Remove unused platform macros ANGLE_APPLE_AVAILABLE_XCI and ANGLE_APPLE_AVAILABLE_XI are no longer used. Bug: angleproject:360147119 Change-Id: I872fbe994a352673a6e9ad6468f0141c3ed98694 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5823423 Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com>
Kimmo Kinnunen 8c211b10 2024-06-20T13:51:20 Metal: Clear depth pbuffers to 1 for robust init Robust resource initialization would init depth pbuffers as 0.0 instead of 1.0. This would differ from the FBO initialization, where the uninitialized depth buffers would be initialized to 1.0. Bug: angleproject:348199042 Change-Id: I0f251e85d6e08ce992fe323b257a0906d237f5b1 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5642763 Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Kimmo Kinnunen <kkinnunen@apple.com> Auto-Submit: Kimmo Kinnunen <kkinnunen@apple.com> Reviewed-by: Quyen Le <lehoangquyen@chromium.org>
Alexey Knyazev e2afdd3c 2024-08-29T00:00:00 Metal: Fix depth/stencil FBO attachments validation Simplified and fixed depth/stencil attachments validation to match Metal behavior; added a new error message. Updated allowSeparateDepthStencilBuffers feature condition to enable it on Apple silicon Macs. Fixed: angleproject:362901599 Change-Id: Ib60020f157976240ebfd652254d63380341cddbd Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5823534 Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Kimmo Kinnunen <kkinnunen@apple.com> Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com>
Shahbaz Youssefi b43f8889 2024-08-29T12:34:46 Vulkan: Fix varyings reserved for xfb At some point in the past, the ANGLEXfbPosition varying was generated unconditionally when the xfb extension is available. The logic that reserved one varying for this was not updated and fails to account for it in recent drivers. Bug: angleproject:362545033 Change-Id: I65fb58cd912b96b2774a7dfbffa4aa656f694e81 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5826761 Auto-Submit: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Charlie Lao <cclao@google.com> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Mark Lobodzinski 28d82804 2024-08-29T10:05:09 Tests: Add Passmark benchmark Complex trace Add trace of the 'Complex' benchmark test of the Passmark perofrmance test suite. Test: angle_trace_tests --gtest_filter=*passmark_complex* Bug: b/362801312 Change-Id: Ic85874140fa22f97c13b1be043120e8ccc428c96 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5826921 Reviewed-by: Cody Northrop <cnorthrop@google.com> Commit-Queue: Cody Northrop <cnorthrop@google.com>
Alexey Knyazev e850656a 2024-08-29T00:00:00 Add EXT_separate_depth_stencil stubs Bug: angleproject:362780331 Change-Id: I378b5c9ff06a2c5e56d68ef93ed42e0351b5044e Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5823195 Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Mark Lobodzinski c3309dbd 2024-08-23T15:47:27 Tests: Add Passmark benchmark Simple trace Add trace of the 'Simple' benchmark case of the Passmark performance test suite. Test: angle_trace_tests --gtest_filter=*passmark_simple* Bug: b/362784518 Change-Id: I387efb5cc0d7b1ee8c731f2fdedc56806701169c Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5824181 Commit-Queue: Cody Northrop <cnorthrop@google.com> Reviewed-by: Cody Northrop <cnorthrop@google.com>
Gowtham Tammana 79729f7c 2024-08-27T17:13:57 Reapply "Vulkan: Expose float control properties as angle features" This reverts commit 5a8eab96c6b7b99b8568feb902b8b85e2df7e19a. This relands the commit 027cbe1a071f41a09d596c4baa243a0abc3d1a86 with the below changes - add back needed suppressions that were removed erroneously ``` // https://anglebug.com/361600662 "VUID-RuntimeSpirv-OpEntryPoint-08743", "VUID-RuntimeSpirv-OpEntryPoint-07754", "VUID-RuntimeSpirv-maintenance4-06817", ``` Bug: angleproject:360031000 Change-Id: I74ed80b459466bdd8ae638ff875677b089aaa360 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5818733 Reviewed-by: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Brian Sheedy c094d4bc 2024-08-26T18:00:50 Enable tests on linux-exp-nvidia Enables tests on the new linux-exp-nvidia builder. Bug: chromium:357573052, chromium:362545033, chromium:362486580 Bug: chromium:362728695 Change-Id: Ib2f1c2edb139be1b3b2225dd9a9ec88df2bf1a84 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5814931 Reviewed-by: Yuly Novikov <ynovikov@chromium.org> Commit-Queue: Brian Sheedy <bsheedy@chromium.org>
Gowtham Tammana 0929a8d7 2024-08-13T10:43:06 CL/Vulkan: Add utility function for clspv compile options Add a utility function for querying the options based on the vulkan renderer for the clspv compiler. Bug: angleproject:361717757 Change-Id: I9722b8a80bcad3f2b799d87b8fbe29c84a686a6e Signed-off-by: Gowtham Tammana <g.tammana@samsung.com> Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5810696 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Charlie Lao <cclao@google.com>
Geoff Lang 42ea8555 2024-08-28T11:37:39 GL: Sync default VAO attributes for external contexts. When ANGLE touches the default VAO, we need to save and restore all of its state for external contexts. This mirrors what Chrome does in ScopedAppGLStateRestoreImpl. This allows us to use the default VAO and client arrays for external contexts, improving performance by about 10X on client array heavy benchmarks. Bug: chromium:360151048, chromium:361699768, angleproject:355034868 Change-Id: If38eb1090b5b5fdfca183a7ec519d5a8c3df0b4c Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5823832 Commit-Queue: Geoff Lang <geofflang@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Geoff Lang 8522e835 2024-08-26T20:46:45 WebGPU: Avoid poping error scopes too many times The ErrorScope RAII wrapper would pop the webgpu error scope multiple times if the explicit PopScope funciton was called before the destructor. Bug: angleproject:362191737 Change-Id: I8caa7ab39addf03bc09a7cce517cde800c91653c Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5815568 Commit-Queue: Geoff Lang <geofflang@chromium.org> Reviewed-by: Matthew Denton <mpdenton@chromium.org>
Igor Nazarov 24a3d30d 2024-08-28T14:33:49 Possibly fix FixedQueue.ConcurrentPushPopWithResize flakiness Queue may be empty when `enqueueThreadFinished` become true. Bug: b/302739073 Change-Id: Idb636e3f87c1217520a9e68a69e749f5bcac4d0f Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5823039 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Igor Nazarov <i.nazarov@samsung.com> Reviewed-by: Charlie Lao <cclao@google.com>
Alexey Knyazev cd4b8adb 2024-08-20T00:00:00 Remove obsolete macOS availability macros MTLDevice.registryID * macOS 10.13 (min deploy 10.15) kCGLRPRegistryIDLow, kCGLRPRegistryIDHigh * macOS 10.13 (min deploy 10.15) Bug: angleproject:360147119 Change-Id: Ife691a844b73abf0fb877578ad2c29e77cc70397 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5814951 Reviewed-by: Kimmo Kinnunen <kkinnunen@apple.com> Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com> Reviewed-by: Kenneth Russell <kbr@chromium.org>
Alexey Knyazev 3f5ede8f 2024-08-20T00:00:00 Metal: Remove obsolete availability macros from context MTLSharedEvent * macOS 10.14 (min deploy 10.15) * iOS/tvOS 12.0 (min deploy 15.0) MTLDevice.hasUnifiedMemory * macOS 10.15 (min deploy 10.15) * iOS/tvOS 13.0 (min deploy 15.0) Bug: angleproject:360147119 Change-Id: I8c7e09b1fd9d16cf2bc9dcdca50a7fe577ad25a2 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5816833 Reviewed-by: Kimmo Kinnunen <kkinnunen@apple.com> Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com>
Alexey Knyazev e39d08b9 2024-08-20T00:00:00 Metal: Remove obsolete code from caps and extensions init MSL 2.1 * macOS 10.14 (min deploy 10.15) * iOS/tvOS 12.0 (min deploy 15.0) MSL 2.2 * macOS 10.15 (min deploy 10.15) * iOS/tvOS 13.0 (min deploy 15.0) Volume BC textures * macOS 10.15 (min deploy 10.15) Depth clipping mode * macOS 10.11 (min deploy 10.15) * iOS/tvOS 11.0 (min deploy 15.0) Dual-source blending * macOS 10.12 (min deploy 10.15) * iOS/tvOS 11.0 (min deploy 15.0) Bug: angleproject:360147119 Change-Id: I00d33d2b6b5c80a305bfbe9316c0ea670b9a4b04 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5814950 Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Kimmo Kinnunen <kkinnunen@apple.com> Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com>
Mark Lobodzinski fc4f0ac0 2024-08-21T12:54:13 formatutils: Allow 16-bit formats for renderbuffer with ext in GLES1.0 These format/use combinations should be available before ES 2.0 if the GL_OES_framebuffer_object extension is present: GL_RGBA4 GL_RGB565 GL_RGB5_A1 An E2E test is included. Test: angle_end2end_tests --gtest_filter=*FramebufrerObjectTest.RenderBufferObject* Bug: b/361370571 Change-Id: I0bc4cf43733a312c2847ebd63075338b365571d7 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5805221 Reviewed-by: Cody Northrop <cnorthrop@google.com> Commit-Queue: Mark Łobodziński <mark@lunarg.com> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Kimmo Kinnunen b512e42d 2024-08-20T15:12:19 Reland: Metal: Fix availability for kIOMainPortDefault Remove ad hoc deprecation disables for kIOMasterPortDefault. macCatalyst kIOMainPortDefault available 15.0, less than minimum 15.6. Remove the ifdefs. Reland: Guard macOS specific unittests with compile guards. The patch changes the tested function to be macOS only. Bug: angleproject:360147118 Change-Id: Icbcc811a066995e742825a364fc5f0b5a27a59bc Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5816832 Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Kimmo Kinnunen <kkinnunen@apple.com> Commit-Queue: Geoff Lang <geofflang@chromium.org> Auto-Submit: Kimmo Kinnunen <kkinnunen@apple.com>
Alexey Knyazev b52f281a 2024-08-20T00:00:00 Metal: Remove unused and obsolete availability macros * Both usages of ANGLE_APPLE_AVAILABLE_XC are obsolete. * ANGLE_APPLE_AVAILABLE_CI and ANGLE_APPLE_AVAILABLE_I are unused. * iOS and tvOS SDK version ranges are supported without redefinitions. * Mac Catalyst is defined in the supported SDK versions. Bug: angleproject:360147119 Change-Id: I8e7730c31f73c03d491373a0a5d316432513d8a2 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5817148 Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com> Reviewed-by: Kimmo Kinnunen <kkinnunen@apple.com> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Alexey Knyazev 2609e02c 2024-08-20T00:00:00 Metal: Remove unused branches from StartFrameCapture The used APIs are supported on all deployment targets. Bug: angleproject:360147119 Change-Id: Ieaa8662e97b71e97bdab3cd0f5c5ef6db3d1b305 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5817150 Reviewed-by: Kimmo Kinnunen <kkinnunen@apple.com> Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Alexey Knyazev 8cfb9ea2 2024-08-20T00:00:00 Metal: Align OS macro usage Consistenty used TARGET_OS_* macro instead of ANGLE_PLATFORM_* in the Metal backend code for better searchability. Bug: angleproject:360147119 Change-Id: I754ddc158daa6b1894b0b4fb0489d59f14605b66 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5816828 Reviewed-by: Kimmo Kinnunen <kkinnunen@apple.com> Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Alexey Knyazev be4532be 2024-08-20T00:00:00 Metal: Update availability macros in tests IsMetalTextureSwizzleAvailable is always false on Simulator. IsMetalCompressedTexture3DAvailable is always true on the supported platforms. Bug: angleproject:360147119 Change-Id: I297f7649071143b0cf3c20d5adc8fc3deca88910 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5816830 Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com> Reviewed-by: Kimmo Kinnunen <kkinnunen@apple.com> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Yuly Novikov 5a8eab96 2024-08-27T10:07:26 Revert "Vulkan: Expose float control properties as angle features" This reverts commit 027cbe1a071f41a09d596c4baa243a0abc3d1a86. Reason for revert: removes needed suppressions, breaking bots: https://ci.chromium.org/ui/p/angle/builders/ci/win-exp-test/569/overview https://ci.chromium.org/ui/p/angle/builders/ci/linux-exp-test/810/overview https://ci.chromium.org/ui/p/angle/builders/ci/android-arm64-exp-s22-test/1060/overview Original change's description: > Vulkan: Expose float control properties as angle features > > The different fp configuration that a vulkan backend supports are > exposed through `VkPhysicalDeviceFloatControlsProperties`. Expose these > as angle features that api's implementations can query. > > The autogen files are changes as reported by running `python3 > scripts/run_code_generation.py` script. > > Bug: angleproject:360031000 > Change-Id: I038cbf2c9e2b10916697994782ef73b35a1caf12 > Signed-off-by: Gowtham Tammana <g.tammana@samsung.com> > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5798053 > Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> > Reviewed-by: Charlie Lao <cclao@google.com> Bug: angleproject:360031000 Change-Id: I88c0a083725320b5e1a8302a50f67d41cbf5c089 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5816829 Auto-Submit: Yuly Novikov <ynovikov@chromium.org> Commit-Queue: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com> Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
angle-autoroll 2177be94 2024-08-26T17:32:22 Manual roll VK-GL-CTS from 21ff94426906 to 824d14748364 (13 revisions) Manual roll requested by syoussefi@google.com * KHR-GLES3: * Removed outdated test names to resolve the unused expectation warning. * (After test name changes, they no longer exist in the mustpass test list.) * KHR-GLES31: * Unsuppressed a test that no longer fails. https://chromium.googlesource.com/external/github.com/KhronosGroup/VK-GL-CTS.git/+log/21ff94426906..824d14748364 2024-08-23 ziga@lunarg.com Fix build on windows 2024-08-23 MengYang.Liu@amd.com Add test cases for maxPushConstantsSize queried from device 2024-08-23 mate@rastergrid.com Fix waiver mechanism for Vulkan SC testing 2024-08-23 gsweet@broadcom.com Check fillModeNonSolid before using non-solid fill modes 2024-08-23 gleese@broadcom.com Add extra fast-math decoration in OpPhi tests 2024-08-23 ziga@lunarg.com Test pipeline using compatible render pass without resolve attachment 2024-08-23 ziga@lunarg.com Fix crash in local read max input attachments test 2024-08-23 gleese@broadcom.com Revert "Test mismatched dimensionality" 2024-08-23 ewerness@nvidia.com Default --deqp-terminate-on-device-lost to enable 2024-08-23 kamil.goras@mobica.com Split KHR-GL46.shading_language_420pack* 2024-08-23 kamil.goras@mobica.com Split KHR-GL46.packed_pixels.* 2024-08-23 syoussefi@google.com Fix missing GL barriers in `LoadStoreMachine` tests 2024-08-23 lorenzo@khronosgroup.org Merge vk-gl-cts/vulkan-cts-1.3.9 into vk-gl-cts/main If this roll has caused a breakage, revert this CL and stop the roller using the controls here: https://autoroll.skia.org/r/vk-gl-cts-angle-autoroll Please CC abdolrashidi@google.com,angle-team@google.com,syoussefi@google.com on the revert to ensure that a human is aware of the problem. To file a bug in ANGLE: https://bugs.chromium.org/p/angleproject/issues/entry To report a problem with the AutoRoller itself, please file a bug: https://issues.skia.org/issues/new?component=1389291&template=1850622 Documentation for the AutoRoller is here: https://skia.googlesource.com/buildbot/+doc/main/autoroll/README.md Bug: b/350994515 Bug: angleproject:40096882 Tbr: abdolrashidi@google.com,syoussefi@google.com Change-Id: I9db95cedef583120733c34c8630864fbad3c36f6 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5812210 Reviewed-by: Yuxin Hu <yuxinhu@google.com> Commit-Queue: Amirali Abdolrashidi <abdolrashidi@google.com>
Gowtham Tammana 027cbe1a 2024-08-13T17:27:27 Vulkan: Expose float control properties as angle features The different fp configuration that a vulkan backend supports are exposed through `VkPhysicalDeviceFloatControlsProperties`. Expose these as angle features that api's implementations can query. The autogen files are changes as reported by running `python3 scripts/run_code_generation.py` script. Bug: angleproject:360031000 Change-Id: I038cbf2c9e2b10916697994782ef73b35a1caf12 Signed-off-by: Gowtham Tammana <g.tammana@samsung.com> Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5798053 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Charlie Lao <cclao@google.com>
Austin Annestrand a2602533 2024-08-16T15:31:43 CL/VK: Update zero-enqueue validation checks According to the spec: CL_INVALID_GLOBAL_WORK_SIZE if global_work_size is NULL or if any of the values specified in global_work_size[0], …​ global_work_size[work_dim - 1] are 0. Returning this error code under these circumstances is deprecated by version 2.1. Since ANGLE supports OpenCL 3.0, we need to remove our "gws[dim] != 0" assert since this is no longer an error. Also adding missing check for LWS[dim] == 0 case since validationCL does not check for this prior to performing a modulo operation. Bug: angleproject:361750102 Tests-Fixed: OCLCTS.test_api.zero_sized_enqueue Change-Id: I02f302fd988bc37e3b52aeea30586843a47cadfd Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5807300 Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Austin Annestrand <a.annestrand@samsung.com>
Austin Annestrand c34e8c75 2024-08-16T14:06:10 CL: Platform check for ValidateCreateContextFromType Bug: angleproject:42266872 Change-Id: Iaf19a2bb6ca1e1a1ce7b02852a3e4fd5290bd557 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5807299 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Austin Annestrand <a.annestrand@samsung.com> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Alexey Knyazev 6ef39db0 2024-08-20T00:00:00 Metal: Remove obsolete compile guards for texture helpers MTLDevice.supports32BitFloatFiltering - macOS 11.0 (min compile 12.0, min deploy 10.15) - iOS 14.0 (min compile 17.0, min deploy 15.0) - tvOS 16.0 (min compile 17.0, min deploy 17.0) MTLDevice.supportsBCTextureCompression - macOS 11.0 (min compile 12.0, min deploy 10.15) - iOS 16.4 (min compile 17.0, min deploy 15.0) - tvOS 16.4 (min compile 17.0, min deploy 17.0) MTLPixelFormatPVRTC* enums - macOS 11.0 (min compile 12.0) - iOS 8.0 (min compile 17.0) - tvOS 9.0 (min compile 17.0) Bug: angleproject:360147119 Change-Id: Iafb010efc233f9c677a2c6f65cb4efdf0bb9cf9c Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5805290 Reviewed-by: Kimmo Kinnunen <kkinnunen@apple.com> Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Alexey Knyazev 939fafc8 2024-08-20T00:00:00 Metal: Remove MSL version compile guards MTLLanguageVersion1_0 - Deprecated MTLLanguageVersion2_3 - macOS 11.0 (min compile 12.0, min deploy 10.15) - iOS/tvOS 14.0 (min compile 17.0, min deploy 15.0) MTLLanguageVersion2_4 - macOS 12.0 (min compile 12.0, min deploy 10.15) - iOS/tvOS 15.0 (min compile 17.0, min deploy 15.0) Bug: angleproject:360147119 Change-Id: I4a0da9d92ef86e899766bc860f3bc6e7e614f4a9 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5806332 Reviewed-by: Kimmo Kinnunen <kkinnunen@apple.com> Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com>
Geoff Lang e10e31ed 2024-08-26T12:56:42 WebGPU: Add error handling macros for debugging. Add a ANGLE_WGPU_SCOPED_DEBUG_TRY to create an error scope, execute some commands and pop the scope in builds with assertions enabled. A follow-up with full error handling at frame boundaries is still needed. Bug: angleproject:362191737 Change-Id: I4af9e1587f6d70bf14a3f6435b487e26424624ef Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5813006 Reviewed-by: Matthew Denton <mpdenton@chromium.org> Commit-Queue: Geoff Lang <geofflang@chromium.org>
Yuly Novikov 95c30276 2024-08-26T18:14:30 Don't use VK_GOOGLE_surfaceless_query on Samsung S24 Xclipse Apply S22 workaround on S24 as well, tentative fix for bad traces. Bug: b/362253922 Change-Id: I5a04741d7465139f86de5baadbf774eed172b4e5 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5812323 Auto-Submit: Yuly Novikov <ynovikov@chromium.org> Reviewed-by: Cody Northrop <cnorthrop@google.com> Commit-Queue: Yuly Novikov <ynovikov@chromium.org> Commit-Queue: Cody Northrop <cnorthrop@google.com>
Alexey Knyazev 966d0a1a 2024-08-20T00:00:00 Metal: Cleanup memoryless storage mode support detection Bug: angleproject:360147119 Change-Id: Ic47f1d84e748f65194aa819a39418db66b1948c3 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5802956 Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com> Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Kimmo Kinnunen <kkinnunen@apple.com>
Yuly Novikov 8281af52 2024-08-15T16:26:02 Skip timing out end2end tests on Samsung S24 Xclipse. Add GALAXYS24 expectation tag and skip: FramebufferTest_ES31.ChangeFBOSizeAndAttachmentsCount FramebufferTest_ES31.ChangeFBOSizeWithNoAttachments FramebufferTest_ES31.CreateNoAttachmentFBOWithDifferentSize FramebufferTest_ES31.RenderingLimitToDefaultFBOSizeWithNoAttachments Bug: angleproject:359904354 Change-Id: If88fa3876b7fcf534d99fee75b5aa4fa6da140c6 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5808948 Reviewed-by: Solti Ho <solti@google.com> Commit-Queue: Yuly Novikov <ynovikov@chromium.org> Reviewed-by: Yuxin Hu <yuxinhu@google.com>
Alexey Knyazev 003737c8 2024-08-20T00:00:00 Metal: Adjust texture storage mode compile guards Used managed mode only for macOS and Mac Catalyst compile targets. Fixed compilation for visionOS. Bug: angleproject:360147119 Change-Id: I88fae28ff6d1ecb5ba303d32c4a3874bf3f8aa87 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5802959 Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Kimmo Kinnunen <kkinnunen@apple.com> Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com>
Alexey Knyazev a36c459e 2024-08-20T00:00:00 Metal: Remove redundant symbol redeclarations MTLBlitOptionRowLinearPVRTC - macOS 11.0 (min compile 12.0) - iOS/tvOS 9.0 (min compile 17.0) MTLBarrierScope - macOS 10.14 (min compile 12.0) - iOS/tvOS 12.0 (min compile 17.0) MTLBarrierScopeRenderTargets - Moved compile guards to call sites MTLRenderStages, MTLRenderStageVertex, MTLRenderStageFragment - macOS 10.13 (min compile 12.0) - iOS/tvOS 10.0 (min compile 17.0) Bug: angleproject:360147119 Change-Id: I7dac46accde63be486a7de918284bbd6e8367a4e Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5806896 Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Kimmo Kinnunen <kkinnunen@apple.com> Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com>
Alexey Knyazev 1f1b7426 2024-08-20T00:00:00 Metal: Remove compile guards for preserveInvariance MTLCompileOptions.preserveInvariance - macOS 11.0 (min compile 12.0, min deploy 10.15) - iOS/tvOS 14.0 (min compile 17.0, min deploy 15.0) Bug: angleproject:360147119 Change-Id: I54bd75f3136e907ebb738b855f12b80f3c161b41 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5806898 Reviewed-by: Kimmo Kinnunen <kkinnunen@apple.com> Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com>
Charlie Lao acf63b9e 2024-08-22T09:52:41 Vulkan: call traceGpuEvent after queueSerial has been allocated traceGpuEvent() will write to mOutsideRenderPassCommands. This means if we destroy context immediately it will flush OutsideRenderPassCommands. This will trigger assertion that the queueSerialIndex is invalid. This CL moves queueSerialIndex allocation before traceGpuEvent. Bug: b/361570359 Change-Id: I70909ebd23c43c05cb793319b255f81e65a17a9d Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5806203 Commit-Queue: Charlie Lao <cclao@google.com> Reviewed-by: Greg Schlomoff <gregschlom@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Gowtham Tammana 3f1c47bf 2024-08-13T15:06:40 Vulkan: Add VK_KHR_variable_pointers ext if available Add VK_KHR_variable_pointers exetensions to enabled extension if it is supported. This is needed when SPIR-V 1.4 is supported. Bug: angleproject:360031000 Change-Id: Ia93604ae308c9f304acb3bc39457811401a0ca7b Signed-off-by: Gowtham Tammana <g.tammana@samsung.com> Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5789897 Reviewed-by: Charlie Lao <cclao@google.com> Commit-Queue: Austin Annestrand <a.annestrand@samsung.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Yuly Novikov 2e35835b 2024-08-23T11:28:54 Suppress new VUIDs VUID-RuntimeSpirv-OpEntryPoint-08743 VUID-RuntimeSpirv-OpEntryPoint-07754 VUID-RuntimeSpirv-maintenance4-06817 Started reporting after VVL roll https://chromium-review.googlesource.com/c/angle/angle/+/5807076 Bug: angleproject:361600662 Change-Id: Ibdf30bd2506366460009f18a4e1f63559d23eb97 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5803266 Reviewed-by: Cody Northrop <cnorthrop@google.com> Commit-Queue: Cody Northrop <cnorthrop@google.com> Auto-Submit: Yuly Novikov <ynovikov@chromium.org>
Charlie Lao 38d5b4d6 2024-08-21T15:49:33 Vulkan: Clean up the CovertVertexBuffer* API Some minor clean up of the API so that the arguments we passed into these two functions CovertVertexBufferGPU and CovertVertexBufferCPU are symmetrical. This CL also factor out the common logic in these two functions into a helper function CalculateOffsetAndVertexCountForConversion(). Bug: b/357622380 Change-Id: I5339850c99d733dfea5039ba74734290f8326bd3 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5804529 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Charlie Lao <cclao@google.com> Reviewed-by: Yuxin Hu <yuxinhu@google.com>
Charlie Lao 8fc6d8a8 2024-08-13T09:49:46 Vulkan: Limit vertexBufferConversion to the data within dirtyRange Right now when we do vertex buffer conversion, we always convert from the binding.offset all the way to the end of the buffer, because we do not know who much vertex will be drawn. According to the comment, if we choose to only convert the number of vertices this draw call will be using, then we may end up issuing a lot of conversions in the case that there are many draw calls with small vertex numbers. This CL intends to keep this behavior (i.e, still do the conversion of the data from binding offset to end of the buffer), but apply the optimization that pays attention of mDirtyRange and only convert the portion of data that has actually been modified. Note that the mDirtyRange is the union of all data ranges that has been modified, so may include the unmodified data between two modified ranges. Bug: b/357622380 Change-Id: I91f3dbf4cacb17d3fa4d8d599519865be765557f Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5785860 Commit-Queue: Charlie Lao <cclao@google.com> Reviewed-by: Yuxin Hu <yuxinhu@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Charlie Lao 6f97a8b3 2024-08-21T13:25:28 Vulkan: Add two tests for vertex conversion Two glBufferSubData calls and then use the buffer (get down to the vertex conversion code path, at least with forced fallback config). The test will exercise the logic regarding the conversion with multiple dirty ranges. Two tests are added in this CL, one with two overlapping range and another with non-overlapping range. Bug: b/357622380 Change-Id: Id8fa23a6e1511e2f03c002782ab99f167416ca02 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5805244 Reviewed-by: Yuxin Hu <yuxinhu@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Charlie Lao <cclao@google.com>
Charlie Lao 790e0162 2024-08-09T17:11:38 Vulkan: Add dirty range to VertexConversionBuffer class Previously, ConversionBuffer only has a boolean indicates it is dirty or not. This CL adds mDirtyRange to it to indicate which range of data has been modified. The existing dirty boolean has been changed to mEntireBufferDirty so that all the current code will still work. Right now mEntireBufferDirty is always set when we mark it dirty, which means entire buffer gets converted. mDirtyRange has not been used to reduce the data to be converted. Right now the range is always being merged to the existing range and not actually being used in this CL. It will be used in the next CL. Bug: b/357622380 Change-Id: Ibfa702b29011f4e26c511d5db85c07cbf2a4aefb Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5778347 Reviewed-by: Yuxin Hu <yuxinhu@google.com> Commit-Queue: Charlie Lao <cclao@google.com>
Alexey Knyazev 33c2ff07 2024-08-20T00:00:00 Metal: Remove obsolete watchOS compile guards MTLSamplerAddressModeMirrorClampToEdge: - macOS 10.11 (min compile 12.0, min deploy 10.15) - iOS 14.0 (min compile 17.0, min deploy 15.0) - tvOS 16.0 (min compile 17.0, min deploy 17.0) Bug: angleproject:360147119 Change-Id: I21457e7f0694d09857153199037929f9739eded0 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5803986 Reviewed-by: Kimmo Kinnunen <kkinnunen@apple.com> Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com>
Matthew Denton d263945a 2024-08-20T16:32:36 Translator: make replaceChildNode() a little more type-safe If we know the type the child node should be, instead of static_cast<>ing to the expected type, use getAs{ExpectedType}() and crash if nullptr. This would have saved me when the ReplaceVariables() AST transformaton replaced a TIntermSymbol in a TIntermGlobalQualifierDeclaration with a TIntermTyped that was not a TIntermSymbol. But TIntermGlobalQualifierDeclaration can only have a TIntermSymbol as a child, so this caused confusing crashes. Other node types have similar expectations of the types of their children. This doesn't enforce any type safety on TIntermAggregate children as those are a little more flexible and complicated. This CL also doesn't enforce non-nullptr children when the parent can't handle nullptr children. Bug: angleproject:42267100 Change-Id: Ic7fb41039705aa4f6daec052940d9bfc4974d955 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5800264 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Matthew Denton <mpdenton@chromium.org>
angle-autoroll 41309a2f 2024-08-20T08:01:27 Manual Roll VK-GL-CTS from 50e83129b558 to f105fd7ea388 (16 revs) Added deqp.gni changes for: gl3cPrimitiveRestart.hpp/cpp es3cDriverErrorTests.hpp/cpp https://chromium.googlesource.com/external/github.com/KhronosGroup/VK-GL-CTS.git/+log/50e83129b558..f105fd7ea388 2024-08-16 lorenzo@khronosgroup.org Merge vk-gl-cts/vulkan-cts-1.3.9 into vk-gl-cts/main 2024-08-16 ziga@lunarg.com Fix race condition in EGL multithreading tests 2024-08-16 younggwan.kim@arm.com Add coverage for depth bound test without depth attachment 2024-08-16 tomek.bury@broadcom.com Fix dEQP-EGL.functional.fence_sync.valid.egl_fence_persistent_buffer 2024-08-16 okuznetsov@google.com New tests group for known driver error tests 2024-08-16 marcin.hajder@mobica.com Port KC-CTS tests to VK-GL-CTS (primitive_restart_mode), PART 10 2024-08-16 piotr.byszewski@mobica.com Test DRLR remapping while using shader object 2024-08-16 ziga@lunarg.com Test drawing with stencil test and without vkCmdSetFrontFace 2024-08-16 rgarcia@igalia.com Reuse scratch buffer creating acceleration structures back-to-back 2024-08-16 rgarcia@igalia.com Test preemption with VK_KHR_global_priority 2024-08-16 rgarcia@igalia.com Removed ESO compute tests using Amber 2024-08-16 lorenzo@khronosgroup.org Merge vk-gl-cts/opengl-es-cts-3.2.11 into vk-gl-cts/main 2024-08-14 syoussefi@google.com Remove DE_NULL's definition 2024-08-05 lorenzo@khronosgroup.org Merge vk-gl-cts/opengl-cts-4.6.5 into vk-gl-cts/main 2024-08-05 ziga@lunarg.com Fix broken build on windows 2024-08-04 michal.jakubek@mobica.com VK_DEFINE_PLATFORM_TYPE macro doesn't accept nullptr If this roll has caused a breakage, revert this CL and stop the roller using the controls here: https://autoroll.skia.org/r/vk-gl-cts-angle-autoroll Please CC angle-team@google.com,cnorthrop@google.com on the revert to ensure that a human is aware of the problem. To file a bug in ANGLE: https://bugs.chromium.org/p/angleproject/issues/entry To report a problem with the AutoRoller itself, please file a bug: https://issues.skia.org/issues/new?component=1389291&template=1850622 Documentation for the AutoRoller is here: https://skia.googlesource.com/buildbot/+doc/main/autoroll/README.md Bug: None Tbr: cnorthrop@google.com Change-Id: I71223610821b102501a78eea2abecb95d10f4910 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5799167 Reviewed-by: Yuly Novikov <ynovikov@chromium.org> Commit-Queue: Cody Northrop <cnorthrop@google.com>
Alexey Knyazev 1dabe52c 2024-08-20T00:00:00 Metal: Remove texture swizzle compile guards Texture swizzle support: - macOS 10.15 (min compile 12.0, min deploy 10.15) - iOS/tvOS 13.0 (min compile 17.0, min deploy 15.0) Bug: angleproject:360147119 Change-Id: If7e59cc094e5bcdb145a54541662d0fe0a989e8d Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5803984 Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Alexey Knyazev f551a112 2024-08-20T00:00:00 Metal: Remove obsolete compile guards from the command buffer memoryBarrierWithScope and memoryBarrierWithResources: - macOS 10.14 (min compile 12.0, min deploy 10.15) - iOS/tvOS 12.0 (min compile 17.0, min deploy 15.0) dispatchThreads: - macOS 10.13 (min compile 12.0, min deploy 10.15) - iOS 11.0 (min compile 17.0, min deploy 15.0) - tvOS 14.5 (min compile 17.0, min deploy 15.0) Bug: angleproject:360147119 Change-Id: I05e4e5dac81b4e07c47e0f3a1b7d165ff9406cc3 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5804021 Reviewed-by: Kimmo Kinnunen <kkinnunen@apple.com> Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com>
Yuly Novikov 36f7e03a 2024-08-21T13:49:10 Revert "Metal: Fix availability for kIOMainPortDefault" This reverts commit a712626e940cbd3ff6f279a6e838362392a3fd1d. Reason for revert: breaks iOS https://chromium-review.googlesource.com/c/chromium/src/+/5802272 https://ci.chromium.org/ui/p/chromium/builders/try/ios-simulator/2048194/overview Original change's description: > Metal: Fix availability for kIOMainPortDefault > > Remove ad hoc deprecation disables for > kIOMasterPortDefault. > macCatalyst kIOMainPortDefault available 15.0, > less than minimum 15.6. Remove the ifdefs. > > Bug: angleproject:360147118 > Change-Id: I2ccc08814e267984ff53275cc2886e2bfa8d4c2a > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5797149 > Commit-Queue: Kimmo Kinnunen <kkinnunen@apple.com> > Reviewed-by: Geoff Lang <geofflang@chromium.org> > Auto-Submit: Kimmo Kinnunen <kkinnunen@apple.com> > Reviewed-by: Alexey Knyazev <lexa.knyazev@gmail.com> Bug: angleproject:360147118 Change-Id: Iab3e8d1dece4866732ecbebab4c83f274b99e034 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5803985 Commit-Queue: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com> Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com> Auto-Submit: Yuly Novikov <ynovikov@chromium.org>
Le Hoang Quyen f102d75a 2024-08-14T19:52:30 Metal: translate IOSurface pbuffer's GL_RGB to RGBX/BGRX format. Currently when IOSurface is bound to a pbuffer with GL_RGB internal format, we asign angle format=B8G8R8A8_UNORM/R8G8B8A8_UNORM to it. And then disable alpha write to the respective MTLTexture. However, when an implicit MSAA texture is created for the above pbuffer, it will be assigned B8G8R8A8_UNORM/R8G8B8A8_UNORM angle format but has no knowledge about alpha write disable. Consequently, when a render pass uses this implicit MSAA texture, its alpha channel won't be cleared properly and would contain garbage values. In order to avoid the above bug, we should use B8G8R8X8_UNORM/R8G8B8X8_UNORM angle format for the IOSurface pbuffer. This will fix the render corruptions found in chromium:358957665. Bug: angleproject:42261786 Bug: chromium:358957665 Change-Id: I8aa1e24ace2de80dd08e21711445515ba4b339ff Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5788801 Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Quyen Le <lehoangquyen@chromium.org>
Kimmo Kinnunen 5cf94ffa 2024-08-20T10:41:40 Metal: Use pipeline rasterSampleCount Use MTLRenderPipelineDescriptor rasterSampleCount instead of sampleCount. This avoids the use of the deprecation macro. Availability: iOS 8.0 (min: 15.0), mac Catalyst 13.1 (min: 16.5), macOS 10.11 (min: 10.15), tvOS 11.0 (min: 15.0), visionOS 1.0+ Bug: angleproject:360147118 Change-Id: I7009a7d7b39ecb763e23f4b97441fb45a64a891d Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5797134 Auto-Submit: Kimmo Kinnunen <kkinnunen@apple.com> Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Alexey Knyazev <lexa.knyazev@gmail.com> Commit-Queue: Kimmo Kinnunen <kkinnunen@apple.com>
Kimmo Kinnunen 25237423 2024-08-20T15:22:32 Metal: Remove ifdefs for primitive topology class MTLPrimitiveTopologyClass availability iOS 12.0, minimum runtime 15.0. Remove the ifdefs. Hardcoding ANGLE_IOS_DEPLOY_TARGET is dangerous, likely it has never been correct. Remove. Should use __IPHONE_OS_VERSION_MIN_REQUIRED. Bug: angleproject:360147118 Change-Id: I43d7a364fa0237af1ee1f546015ae884fc373b15 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5797141 Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Alexey Knyazev <lexa.knyazev@gmail.com> Commit-Queue: Kimmo Kinnunen <kkinnunen@apple.com> Auto-Submit: Kimmo Kinnunen <kkinnunen@apple.com>
Kimmo Kinnunen a712626e 2024-08-20T15:12:19 Metal: Fix availability for kIOMainPortDefault Remove ad hoc deprecation disables for kIOMasterPortDefault. macCatalyst kIOMainPortDefault available 15.0, less than minimum 15.6. Remove the ifdefs. Bug: angleproject:360147118 Change-Id: I2ccc08814e267984ff53275cc2886e2bfa8d4c2a Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5797149 Commit-Queue: Kimmo Kinnunen <kkinnunen@apple.com> Reviewed-by: Geoff Lang <geofflang@chromium.org> Auto-Submit: Kimmo Kinnunen <kkinnunen@apple.com> Reviewed-by: Alexey Knyazev <lexa.knyazev@gmail.com>
Charlie Lao 86508e20 2024-08-16T14:56:37 Vulkan: Make VertexConversionBuffer a class And wrap the cache key (i.e, formatID/stride/offset) into a CacheKey struct so that we can easily add more data members. This CL also changes ConversionBuffer from struct to class to have better encapsulation. No functional changes is expected here. Bug: b/357622380 Change-Id: Ieecf5c922b95a940137c8e54657ef3f458c55fc9 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5793921 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Charlie Lao <cclao@google.com> Reviewed-by: Yuxin Hu <yuxinhu@google.com>
Charlie Lao a3f9f6c3 2024-08-16T12:12:00 Vulkan: Move VertexConversionBuffer out of BufferVk class This is another preparation CL to make code review easier. In the future CL, VertexConversionBuffer will be used in other files. BufferVk::VertexConversionBuffer is a bit too verbose. Make VertexConversionBuffer a standalone class makes code a bit easier to read. We could potentially merge ConversionBuffer and VertexConversionBuffer into single class, but I am leaving it for future exercise. Bug: b/357622380 Change-Id: I2255a1c9b01f5d6c3846e8017cc667e6a9d66ffc Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5787504 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Yuxin Hu <yuxinhu@google.com> Commit-Queue: Charlie Lao <cclao@google.com>
Charlie Lao 7080b766 2024-08-16T11:31:39 Vulkan: Move LineLoopHelper from vk to rx namespace There is no line loop support in vulkan. LineLoopHelper is a utility function for backend, not a helper function for vulkan object. So it is better fit in rx namespace instead of vk namespace. This also helps my next CL where I am going to change initBufferForVertexConversion to take a ConversionBuffer instead of BufferHelper. LineLoopHelper uses initBufferForVertexConversion, which means I have to change LineLoopHelper to uses ConversionBuffer. This causes header inclusion problem that now vk namespace object end up have to include rx namespace header. This CL fixes this inclusion problem by moving it to the proper namespace. Bug: b/357622380 Change-Id: I6d6cf1aa926f726bb1b1ab1017bcab092eaf5d37 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5787502 Reviewed-by: Yuxin Hu <yuxinhu@google.com> Commit-Queue: Charlie Lao <cclao@google.com>
Kimmo Kinnunen e0f790c0 2024-08-20T12:08:40 Metal: Use present APIs for Metal availability Use non-deprecated APIs for checking Metal backend availability. macCatalyst and macOS needs MacFamily2 Availability: - macOS 10.15, minimum runtime 10.15 - macCatalyst 13.1, minimum runtime 16.5 iOS, minimum runtime 15.0, can always use the backend. Bug: angleproject:360147118 Change-Id: Id0b70c81bc8b54a22b94ba514aab148b202d58ef Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5797053 Commit-Queue: Geoff Lang <geofflang@chromium.org> Reviewed-by: Alexey Knyazev <lexa.knyazev@gmail.com> Auto-Submit: Kimmo Kinnunen <kkinnunen@apple.com> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Kimmo Kinnunen eb6c4bfe 2024-08-20T09:58:11 Metal: Fix availabilities of MTLGPUFamily use Fixes compile in case of targeting only the newer SDKS. [MTLDevice supportsFamily:] is supported on all minimum runtimes: - macOS 10.15 (minimum runtime 10.15) - macCatalyst 13.1 (minimum runtime 16.5) - iOS/tvOS 13.0 (minimum runtime 15.0) Remove the ifdefs and fallbacks. MTLGPUFamilyMacCatalyst1 is deprecated on macCatalyst 16.0. Replace with MTLGPUFamilyMac2. MTLGPUFamilyMac1 is deprecated on macOS 13.0. Replace with MTLGPUFamilyMac2. MTLGPUFamilyApple7 is present on iOS/tvOS 14.0, minimum runtime 15.0. Remove ifdefs. MTLGPUFamilyApple6,7 are present on macOS 11.0 SDK, minimum SDK 12.0. Remove ifdefs. Bug: angleproject:360147118 Change-Id: I337d1ad44940a2f39a57ba8a28de0e1d7387658c Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5797050 Auto-Submit: Kimmo Kinnunen <kkinnunen@apple.com> Reviewed-by: Alexey Knyazev <lexa.knyazev@gmail.com> Commit-Queue: Geoff Lang <geofflang@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Kimmo Kinnunen 2758d4dc 2024-08-20T16:00:07 Metal: Remove shared event related ifdefs Shared events are supported by iOS 12, macOS 10.14, included in the minimum versions iOS 15, macOS 10.15. Bug: angleproject:360147118 Change-Id: I5b62dade045f9d50864ef1906cb22a18fd09e4da Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5797150 Auto-Submit: Kimmo Kinnunen <kkinnunen@apple.com> Reviewed-by: Alexey Knyazev <lexa.knyazev@gmail.com> Commit-Queue: Geoff Lang <geofflang@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Geoff Lang 2f2cdbb8 2024-08-19T11:05:22 D3D11: Fix uniform setting for mat3 followed by float A single float following a float3x3 will pack tightly in the last row of the matrix. The uniform setting code for matricies would write a full row stride's worth of data for the last matrix row, overwriting the float packed at the end. Update ExpandMatrix to only write colsSrc worth of values in the last row. Bug: angleproject:42266878, chromium:345525082 Change-Id: Ieb616972387501cd8c30f02b267f86a7f81cf01a Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5796631 Reviewed-by: Yuxin Hu <yuxinhu@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Yuly Novikov <ynovikov@chromium.org> Commit-Queue: Geoff Lang <geofflang@chromium.org>
Le Hoang Quyen 1a191b1d 2024-08-16T17:31:38 Metal: add an option to emulate DontCare loadOp with randomized Clear. This is useful for testing undefined initial values caused by DontCare MTLLoadAction on non-tiled GPUs. Bug: angleproject:42261786 Change-Id: Id5c4a9ee8510a8a9143bdd67e276bfcf2bd7eaa5 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5791386 Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Quyen Le <lehoangquyen@chromium.org>
Alexey Knyazev 57a3e9a7 2024-08-15T00:00:00 Prevent compilation with unsupported Apple SDK versions Bug: angleproject:360147119 Change-Id: Ie9924cb49ffa729d454a2869a2d19d957500bd10 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5793513 Reviewed-by: Kimmo Kinnunen <kkinnunen@apple.com> Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Kimmo Kinnunen 62cc3790 2024-08-19T11:05:52 Metal: Remove UseResource old SDK fallbacks Fixes deprecation compile errors on newer SDKs. Remove the version guards, the useResource API is available on: macOS 10.15, minimum supported version 10.15 iOS 13, minimum supported version 15 macCatalyst 13.1, minimum supported version 16.6 (for macOS 13.5) Bug: angleproject:360147118 Change-Id: I6d2dd7cb510d6134b75d4479ae7159b8fbcfc96b Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5796002 Commit-Queue: Geoff Lang <geofflang@chromium.org> Auto-Submit: Kimmo Kinnunen <kkinnunen@apple.com> Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Alexey Knyazev <lexa.knyazev@gmail.com>
Gowtham Tammana 579a5855 2024-08-13T11:39:54 Vulkan: Add query for 64bit fp support feature check Add a feature check for shaderFloat64 support. If present expose the corresponding feature check. The following files are updated by running the autogen script - `run_auto_code_generation.py` - include/platform/autogen/FeaturesVk_autogen.h - util/autogen/angle_features_autogen.cpp - util/autogen/angle_features_autogen.h Bug: angleproject:360031000 Change-Id: If2a11190729550ad5b84714e0a3aa3b81956c66a signed-off-by: gowtham tammana <g.tammana@samsung.com> Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5789896 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Gowtham Tammana e90cd0c4 2024-08-12T17:07:47 Vulkan: Add check for int8 extension support Vulkan backend currently has a check for float16 part in VK_KHR_shader_float16_int8 extension. Add the int8 support as well. The changes in below files are generated through `scripts/run_code_generation.py` script - include/platform/autogen/FeaturesVk_autogen.h - util/autogen/angle_features_autogen.cpp - util/autogen/angle_features_autogen.h Bug: angleproject:360031000 Change-Id: I8811779299e3d24b4536d1871e1eadcbb87432c2 Signed-off-by: Gowtham Tammana <g.tammana@samsung.com> Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5792021 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Gowtham Tammana 319907d2 2024-04-05T16:05:33 Vulkan: Propagate the support of shaderInt16/64 feature Progate the support of shaderInt16/64 features in to enabled feature list. Bug: angleproject:360031000 Change-Id: I38915a764fb2f1dd31348c01bc2fcfa4913ebdec Signed-off-by: Gowtham Tammana <g.tammana@samsung.com> Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5792019 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Kimmo Kinnunen 78dd9a8e 2024-08-16T15:44:49 Metal: Support compiling without OpenGL on macOS Avoid OpenGL framework use unless CGL is enabled. It was used in active GPU resolution for older macOSes. Bug: angleproject:360159407 Change-Id: I6d820ea37290d5e75151db0b42dc266910bcba6a Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5793512 Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Kenneth Russell <kbr@chromium.org>
Roman Lavrov 2f28264e 2024-08-15T18:05:43 Trace perf: add origin/main revision to summary if != HEAD Bug: None Change-Id: Iacacd98993d7dc5fbea8489b46236e939f465689 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5792199 Reviewed-by: Cody Northrop <cnorthrop@google.com>
Kimmo Kinnunen 26c842fc 2024-08-15T14:51:22 Metal: use injectAsmStatementIntoLoopBodies on iOS The setting should be enabled for iOS. Bug: chromium:1513738 Change-Id: I5efd4126711eff6df8f5d84689f7f3310d7e8487 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5788428 Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Kimmo Kinnunen <kkinnunen@apple.com>
Austin Annestrand 83ca89e5 2024-08-08T19:53:32 CL: Update logic for default platform on context creation In ANGLE, when multiple backends are enabled (e.g. passthrough, Vulkan, etc.), each backend is its own platform. Existing validation code was failing in this case for clCreateContext in the event of user not manually specifying CL_CONTEXT_PLATFORM. Spec says implementation then decides which platform to be associated with that context. ANGLE has a Platform::GetDefault() routine that was used and it simply grabs the 1st platform from our platform list in ANGLE (passthrough mode platform object in this case). We erroneously fail the validation in the case where we pass Vulkan device in our device_list since default-platform != Vulkan-platform. Updating the logic to use device's platform (from list) as default case. For clCreateContextFromType, we can continue to use Platform::GetDefault() as this routine generates the appropriate device list for the user to query/get via clGetContextInfo. Bug: angleproject:42266872 Change-Id: I285459688309d1cc76ade5a8564164a5538e1eda Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5789652 Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Austin Annestrand <a.annestrand@samsung.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Yuly Novikov 31ad61c4 2024-08-15T16:31:23 Skip end2end WebGPU tests flaky crashing on Linux Intel Seen on UHD630 Ubuntu 22.04 Mesa 23.2.1 experiment. Bug: angleproject:360083331 Change-Id: I63f1eda1dc3ef7f64e3b5e37b678a31ac9a190af Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5786999 Commit-Queue: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com> Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com> Auto-Submit: Yuly Novikov <ynovikov@chromium.org> Commit-Queue: Yuly Novikov <ynovikov@chromium.org>
Austin Annestrand 21905014 2024-08-08T16:55:52 CL: Fix entrypoint validation for ICDs For some OpenCL ICD implementations, the clGetPlatformIDs API is left as nullptr in the ICD dispatch table. This is because the OpenCL ICD Loader instead uses the clIcdGetPlatformIDsKHR call instead. Updating the passthrough code to only require checking the clGetPlatformIDs API if the platform is not an ICD. Bug: angleproject:42266872 Change-Id: Ic723adbf3ad0f6c79389ef629dd24ae8a82ca9ba Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5789651 Commit-Queue: Austin Annestrand <a.annestrand@samsung.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Mark Lobodzinski 77616aa3 2024-08-14T13:22:32 Tests: Add Gwent Witcher Card Game trace Test: angle_trace_tests --gtest_filter="*gwent_witcher_card_game*" Bug: b/359910003 Change-Id: I6b93d36ac58df7639f07b7f21107f4d4522f6cee Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5789276 Commit-Queue: Cody Northrop <cnorthrop@google.com> Reviewed-by: Cody Northrop <cnorthrop@google.com>
angle-autoroll e6b46eae 2024-08-15T10:01:05 Roll vulkan-deps from f084a55f96da to 91058f2297c9 (4 revisions) Suppress new VUIDs: VUID-vkCmdDraw-Input-07939 VUID-vkCmdDrawIndexed-Input-07939 VUID-vkCmdDrawIndexedIndirect-Input-07939 VUID-vkCmdDrawIndirect-Input-07939 https://chromium.googlesource.com/vulkan-deps.git/+log/f084a55f96da..91058f2297c9 Also rolling transitive DEPS: https://chromium.googlesource.com/external/github.com/KhronosGroup/glslang/+log/3a08cd8de1f88d114c18670366f29226db453a9e..0dc6711e5a178e4d5643437af688c6b48f829f5c https://chromium.googlesource.com/external/github.com/KhronosGroup/SPIRV-Headers/+log/f013f08e4455bcc1f0eed8e3dd5e2009682656d9..1b75a4ae0b4289014b4c369301dc925c366f78a6 https://chromium.googlesource.com/external/github.com/KhronosGroup/Vulkan-ValidationLayers/+log/8d902c7f67edd9e54b0fa5f7d2c34f9abfeacba8..4eb3585f32ec194efe2d9cd88a2b2943ff41f922 If this roll has caused a breakage, revert this CL and stop the roller using the controls here: https://autoroll.skia.org/r/vulkan-deps-angle-autoroll Please CC angle-team@google.com,ynovikov@google.com on the revert to ensure that a human is aware of the problem. To file a bug in ANGLE: https://bugs.chromium.org/p/angleproject/issues/entry To report a problem with the AutoRoller itself, please file a bug: https://issues.skia.org/issues/new?component=1389291&template=1850622 Documentation for the AutoRoller is here: https://skia.googlesource.com/buildbot/+doc/main/autoroll/README.md Bug: chromium:359904720 Change-Id: Idf82bf9a210b953785fc2387e5c64ef157ec36e6 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5791052 Reviewed-by: Cody Northrop <cnorthrop@google.com> Commit-Queue: Yuly Novikov <ynovikov@chromium.org>
Gowtham Tammana 2003e062 2024-08-09T17:52:38 Vulkan: Set queue family index for oneoff command buffer Use the queue family index that the Renderer has chosen for creation of oneoff command buffers. Bug: angleproject:359623670 Change-Id: I0a36edf79a751133d477d433c9cfd8dbddaad221 Signed-off-by: Gowtham Tammana <g.tammana@samsung.com> Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5785864 Reviewed-by: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Austin Annestrand 2b723c13 2024-02-02T14:12:06 CL/VK: Add initial enqueueCopyBuffer Generalizing the copy buffer with appropriate barrier inserts and commandbuffer command. Bug: angleproject:42267074 Change-Id: I02fd36a221ea70f007aee78ceeef84c24fa85682 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5410935 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Gowtham Tammana ef6a3132 2024-06-25T17:30:03 CL: Return error code for not supported APIs The spec requires error codes to be reported for not supported API calls. Fixing to return relevant error codes. Bug: angleproject:359296611 Change-Id: Ia64bf2673446269c6557308c169f019c7dd67cc8 Signed-off-by: Gowtham Tammana <g.tammana@samsung.com> Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5782397 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Geoff Lang da572160 2024-07-23T16:36:10 Reland: GL: Forward client-side arrays to the driver when possible The OpenGL driver can handle client-side arrays when the context is OpenGL ES or a desktop GL compatibility profile. When in these situations, use the driver default VAO for all frontend context VAOs and forward client-side data directly to the driver. Fix synchronizing the default VAO state for external contexts. There is no valid VertexArrayStateGL for external VAOs so make sure it's nulled and the VAO dirty bits are set so the correct VAO state is reapplied. Disable syncing to the default VAO for external contexts. The only VAO that they can share with ANGLE's internal state is the default VAO so avoid having to save and restore its state. Bug: angleproject:355034686 Change-Id: I015bbbc854938fe4bc1e92d0ca8fe04628d0db16 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5743284 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Geoff Lang <geofflang@chromium.org>
Geoff Lang 6fddb18d 2024-07-26T12:44:23 GL: Disable heavily emulated extensions for non-WebGL. The multidraw and base vertex base instance extensions are heavily emulated in ANGLE's GL backend, manipulating vertex arrays multiple times per draw. This is often a performance improvement for deep pipelines such as WebGL but slower for shallow stacks like Chrome's rasterization. Mark these extensions as "emulated" and disable them in the frontend for non-WebGL contexts. Only EXT_base_instance can have true native support. Add a frontend feature to re-enable these extensions and use it in the existing end2end tests. Bug: angleproject:355645824 Change-Id: I897559fb270e2aa0f476d7784d25d97574e28f43 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5743837 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Geoff Lang <geofflang@chromium.org>
Shahbaz Youssefi 087afc91 2024-08-10T20:40:53 Vulkan: Prefer fallback to renderable texture than CPU copy When a texture copy is done, the transfer path is preferred. If transfer is not possible, ANGLE falls back to a draw-based copy. However, if the format is not renderable, ANGLE falls back to CPU copy. Instead, this change makes ANGLE try to fall back to a renderable format before giving up on the draw-based copy. Bug: b/351872449 Change-Id: I3d09dd88cc0b7feb5e92e58c4f19380973eb5e94 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5775592 Reviewed-by: Amirali Abdolrashidi <abdolrashidi@google.com> Reviewed-by: Charlie Lao <cclao@google.com> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Austin Annestrand 5967afe7 2024-04-02T17:45:48 CL: Update Platform ctor to handle rx errors Bug: angleproject:359294584 Change-Id: I7aad107e57ea6f29b8fc69205ba72c27de6a9428 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5421576 Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Austin Annestrand <a.annestrand@samsung.com>
Geoff Lang 072dc5d1 2024-08-13T13:53:52 Metal: Disable multisampled render to texture. Corruption has been seen on M1 Macs for WebGL content which uses a multisampled backbuffer. Bug: chromium:358957665, angleproject:42261786 Change-Id: Ia17581b15cd032e91e2642d80abace8d57ceb23b Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5783540 Commit-Queue: Geoff Lang <geofflang@chromium.org> Reviewed-by: Quyen Le <lehoangquyen@chromium.org>
Geoff Lang b36d94dc 2024-08-02T16:23:03 WebGPU: Implement viewport and scissor Tested using samples. dEQP tests covering viewport and scissor require more shader translation to be completed. Bug: angleproject:0 Change-Id: Ia76f2edeea5a07af1881704a21fafb64ab7df43a Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5756127 Reviewed-by: Liza Burakova <liza@chromium.org> Commit-Queue: Geoff Lang <geofflang@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Yuly Novikov c1767fd9 2024-08-13T12:37:18 Suppress deqp_khr_gles31 failure on Samsung S22 KHR-GLES31.core.shader_image_load_store.basic-allTargets-loadStoreCS regressed in http://ci.chromium.org/ui/p/angle/builders/ci/android-arm64-exp-s22-test/912 http://chromium-review.googlesource.com/c/angle/angle/+/5719258 Bug: angleproject:350994515 Change-Id: I342043208cd85acf2877399f9a528c3dd428631b Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5785959 Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com> Commit-Queue: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com> Commit-Queue: Yuly Novikov <ynovikov@chromium.org> Auto-Submit: Yuly Novikov <ynovikov@chromium.org>
Yuly Novikov 1735415a 2024-08-13T10:14:57 Remove unused deqp_khr_gles3 expectations KHR-GL33.cull_distance.functional was split into individual cases in https://chromium-review.googlesource.com/c/angle/angle/+/5710888 Also move SWIFTSHADER expectations added in that CL to be after VULKAN ones, to avoid conflicts due to VULKAN expectations applying to SWIFTSHADER as well. Bug: b/353358652 Change-Id: Idebfba9ef90b7e882d7072538722a591c9dc007b Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5785955 Auto-Submit: Yuly Novikov <ynovikov@chromium.org> Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com> Commit-Queue: Yuly Novikov <ynovikov@chromium.org> Commit-Queue: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Austin Annestrand 5b4517ec 2024-02-02T14:04:29 CL/VK: Add helper for inserting barriers Adding a helper function to insert execution and memory barriers when resource access commands are recorded into the command buffer in CLCommandQueueVk. An example of a copy buffer command with src and dst ``` set access.readBuffer <- src set access.writeBuffer <- dst insert barriers - onResourceAccess(access) add copy command - commandBuffer->copyBuffer() ``` Bug: angleproject:42267073 Change-Id: I61d8064fd015a4bf3dab018cbb83d147915ed627 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5409249 Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Austin Annestrand <a.annestrand@samsung.com>
Gowtham Tammana f8fc8ac3 2024-08-05T11:50:11 Vulkan: Remove dependency on ContextVk for CommandBufferHelper Following on the changes in [1], this makes the `CommandBufferHelperCommon` and `OutsideRenderPassCommandBufferHelper` interfaces independent of `ContextVk` state. Any dependency is made explicit. In addition, interfaces that are not specific to GLES context are also updated. [1]: Commit (bcf814fda5 Vulkan: Constrain the dependency on ContextVk in BufferHelper) Bug: angleproject:8544 Change-Id: I7d90ad915e8c14187ab5584453b9e8802bd91e2b Signed-off-by: Gowtham Tammana <g.tammana@samsung.com> Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5319147 Reviewed-by: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Yuly Novikov 3d6cf23c 2024-08-12T16:26:03 Skip end2end tests failing on iOS simulator Bug: angleproject:359136169 Change-Id: Ic1100d74f308c6d8a3d2f450e8b5f46e8539f2a5 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5782861 Auto-Submit: Yuly Novikov <ynovikov@chromium.org> Commit-Queue: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com> Commit-Queue: Yuly Novikov <ynovikov@chromium.org> Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Le Hoang Quyen ee25dddb 2024-08-09T01:03:26 Metal: Properly clear render pass' attachment with an emulated alpha. When a texture is created with emulated format such as RGB, its alpha channel is initialized to 1 and the write mask is configured such that the alpha value won't be overwritten or set to other values. However, if we use this texture as an attachment in a render pass, and DontCare loadAction is used, the alpha value would be initialized with garbage values. No draw calls within the render pass would change these garbage values because the configured write mask already prevents that. Thus the garbage alpha values would be stored in the memory after the render pass ends. This bug also affects MSRTT's unresolve step, because it uses a draw call to blit the resolve texture to the MSAA texture. However, the alpha write is disabled thus the MSAA texture will remain having garbage alpha values. The fix for this bug is that: if we detect that the texture has emulated format and its loadAction is DontCare, we change it to Clear action to clear the alpha channel to 1. Bug: angleproject:42261786 Change-Id: I994849bd4b3c3ab51698833d1cc520376ae6de44 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5773347 Reviewed-by: Kenneth Russell <kbr@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org> Auto-Submit: Quyen Le <lehoangquyen@chromium.org> Commit-Queue: Quyen Le <lehoangquyen@chromium.org>