kc3-lang/angle/src/libANGLE/renderer

src/libANGLE/renderer

Log

Author Commit Date CI Message
Amirali Abdolrashidi 23ec0620 2023-03-23T09:51:53 Vulkan: Add more data to blob cache header This is to potentially increase blob cache robustness against collisions and partially modified data. This CL adds 7 bytes per chunk (8 bytes in total). During decompression, the data will be checked in order to determine if the cache data can be used. * Added chunk index to the header. (1 byte) * Added CRC16 of the compressed data to the header. (2 bytes) * Added the original (uncompressed) data size to the header. (4 bytes) * Moved kBlobHeaderSize to the anonymous namespace. * Added the following GN flag to enable CRC calculation for the cache. * angle_enable_crc_for_pipeline_cache (only available in debug mode) * Corresponds to kEnableCRCForPipelineCache. * When it is false, the CRC is set to 0. Bug: b/246683126 Change-Id: I9aaf9cda52e0af07a2e1c2d0c39aca407e515701 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4378717 Commit-Queue: Amirali Abdolrashidi <abdolrashidi@google.com> Reviewed-by: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Kimmo Kinnunen 79a2f7e4 2023-04-06T10:42:19 EAGL: Remove the impl preprocessor guards Adhere to the project formatting: - Remove #if preprocessor guards from the EAGL implementation. - Order the includes as in style guide - Use include / import consistently - Move GLES_SILENCE_DEPRECIATION into .gn, similar to CGL - Remove or add platform.h include based on whether the file itself actually the platform.h defined macros Simplifies the nested preprocessor macros. Makes it simpler to further fix the preprocessor use. WebKit should instead have a non-ANGLE .mm file #importing the implementations with the proper include guards. Bug: angleproject:8121 Change-Id: I775bedd048fe3382f873abffb693f130f6104c5c Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4404158 Auto-Submit: Kimmo Kinnunen <kkinnunen@apple.com> Commit-Queue: Kenneth Russell <kbr@chromium.org> Reviewed-by: Dan Glastonbury <djg@apple.com> Reviewed-by: Kenneth Russell <kbr@chromium.org>
Geoff Lang 36296928 2023-03-31T16:44:35 Mark RGBX and BGRX formats as having 8 unused bits. This makes sure that pixelBytes ends up being 4 and fixes potential buffer size validation. Fix EGL configs using pixelBytes to compute EGL_BUFFER_SIZE which is not supposed to include unused bits. This is covered by dEQP-EGL.functional.query_config.constraints.color_buffer_size Bug: chromium:1404790 Change-Id: I47a6e825f3b47501fd0984c21554c394fef32eb6 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4390022 Reviewed-by: Cody Northrop <cnorthrop@google.com> Reviewed-by: Yuxin Hu <yuxinhu@google.com> Commit-Queue: Geoff Lang <geofflang@chromium.org>
Geoff Lang 1e9182a5 2023-04-05T12:54:30 Disable staged uploads to packed texture formats on AMD. The same formats are having upload failures on AMD drivers. Bug: chromium:1430857 Change-Id: I8364d5646ff6811db09710ac27d2ac0f6dfdc116 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4401891 Reviewed-by: Austin Eng <enga@chromium.org> Commit-Queue: Geoff Lang <geofflang@chromium.org>
Alexey Knyazev f33c16e3 2023-03-27T00:00:00 Add SampleVariablesTest * Increased required OpenGL context version to 4.2 to disable the extension on older platforms that do not pass the tests. Bug: angleproject:8097 Change-Id: I9158d895f0b0ec80db2516cc126e89ff5da753e1 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4401931 Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com> Reviewed-by: Kenneth Russell <kbr@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Igor Nazarov 9a2ef389 2023-04-04T14:57:09 Vulkan: Minor protected submitCommands() optimization. Bug: angleproject:3965 Change-Id: I52a7597f1dd3bdaf4a221c6b6338caa1577c08f9 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4392876 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Charlie Lao <cclao@google.com> Commit-Queue: Igor Nazarov <i.nazarov@samsung.com>
Shahbaz Youssefi 8b79410b 2023-04-02T22:25:12 Vulkan: Treat readonly SSBOs as readonly! Instead of assuming SSBOs are always written to, this change adds plumbing for the backend to know when an SSBO is declared readonly and marks the buffer readonly accordingly. With this change, BufferVk can optimize uploads and copies to and from the buffer with the knowledge that it can be safely mapped on the CPU for read while it's being used by the GPU. Bug: b/276002151 Change-Id: I75342148c07949a83436054a738395bbd88caec5 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4392720 Reviewed-by: Charlie Lao <cclao@google.com> Reviewed-by: Yuxin Hu <yuxinhu@google.com> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Shahbaz Youssefi 25e60197 2023-03-31T14:17:26 Vulkan: Unify buffer alloc strategy for uploads and GPU copies With this change, glCopyBufferSubData uses the same buffer allocation strategy as glBufferSubData. Only exception is with buffer self-copies which never allocate a new buffer for simplicity. Additionally, this change allows glCopyBufferSubData to be done on the CPU if possible, i.e. if the source buffer is not being written to by the GPU and whenever the equivalent glBufferSubData would have used a CPU upload. Bug: b/276002151 Change-Id: Ice8df5891c5516b148245d5d6fa9b19b787df4ce Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4390023 Reviewed-by: Charlie Lao <cclao@google.com> Reviewed-by: Yuxin Hu <yuxinhu@google.com> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Charlie Lao 41f0a321 2023-04-03T21:58:43 Revert "Vulkan: Reactivate already started render pass when possible" This reverts commit ad9537af7f2bb5e22bc73f4e833fd3789adaa217. Reason for revert: Suspected cause for flakiness. anglebug.com/8118 Original change's description: > Vulkan: Reactivate already started render pass when possible > > In some usage case (such as lineage_mobile), we are seeing in the middle > of render pass, app switch to another fbo just to issue a glClear() > call, which the clear call itself gets deferred. Application then switch > back to the original frame buffer. Before this CL, the render pass gets > recreated due to frame buffer binding change, even though the clear gets > deferred and new render pass and the previous render pass are > essentially the same. This CL detects this situation and reactivate the > current render pass instead of creating a new one. With this CL, > lineage_m app trace reduces frame time from 3.86ms to 3.7ms, and only > one render pass is used instead of two. > > This CL also allows the render pass started by BlitFramebuffer reused by > subsequent draw calls. Asphalt_9 is hitting this use pattern and this CL > reduces frame time by 0.1245 ms (from 5.6203 ms to 5.4958 on pixel 7 > pro) > > Bug: b/273808966 > Change-Id: I48c2671cbef3ff9d6cf59caae88c37c77828ee07 > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4348713 > Reviewed-by: Amirali Abdolrashidi <abdolrashidi@google.com> > Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> > Commit-Queue: Charlie Lao <cclao@google.com> Bug: b/273808966 Change-Id: I81cc2dcacb52466808b2ccf5819feda466c39fc5 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4396502 Reviewed-by: Charlie Lao <cclao@google.com> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com> Auto-Submit: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Shahbaz Youssefi 3aea3cfd 2023-04-03T16:38:29 Vulkan: Workaround depth bias constant factor on ANV Bug: b/249380591 Change-Id: Iaeda7faf5eb40e0e2086674d3e63bf5bc9911ab4 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4392893 Reviewed-by: Amirali Abdolrashidi <abdolrashidi@google.com> Reviewed-by: Yiwei Zhang <zzyiwei@chromium.org> Auto-Submit: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Charlie Lao 96cda1ac 2023-03-31T15:08:11 Vulkan: Switch to modified framebuffer should trigger submit The feature flag preferSubmitAtFBOBoundary intended to trigger submitCommands call when FBO is switched. Right now there is a bug that when FBO is switched, submitCommands did not get called even if feature is enabled. The reason is that when framebuffer changed, we first get FramebufferVk::syncState call, and if the FBO that we switched to is drity, we will end up calling ContextVk::flushCommandsAndEndRenderPass to immediate end the render pass. The problem with that is that later on when we get to ContextVk::syncState and saw DIRTY_BIT_DRAW_FRAMEBUFFER_BINDING dirty bit and try to issue a submit, we notice render pass already ended, so mHasDeferredFlush never gets set, which means we never call submitCommands. All apps render to system framebuffer in the last render pass, and system framebuffer is always dirty due to swap chain image rotation, we hit this almost with every app. This CL avoid flushCommandsAndEndRenderPass() call from FramebufferVk::syncState. We now relies on ContextVk::onFramebufferChanged (which calls onRenderPassFinished() which will set DIRTY_RENDER_PASS bit to trigger deferred endRenderPass) to end the current render pass. This CL also add a regression test to ensure the submit indeed occur. Bug: b/275624771 Change-Id: I92b95a7a6c435f242d6684cb7852172cf41896c3 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4390642 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Charlie Lao <cclao@google.com> Reviewed-by: Roman Lavrov <romanl@google.com>
Shahbaz Youssefi c5e9de23 2023-04-03T15:14:03 Revert "Vulkan: Use midRenderPass clear if RP has started but inactive" This reverts commit 98151770adfd990c533991da27615b4879494307. Reason for revert: Suspected cause for flakiness. anglebug.com/8118 Original change's description: > Vulkan: Use midRenderPass clear if RP has started but inactive > > This CL extends prior CL's optimization so that if clear is issued right > after blitFramebuffer call (this could make sense if blit and clear are > on different buffer), we can keep the started render pass and do the > midRenderPass clear instead of endRenderPass and start another > renderPass. > > Bug: b/273808966 > Change-Id: Ia2504e8e260867a6f797d42cd4c8a72f187280ef > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4374145 > Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Bug: b/273808966 Change-Id: I7a11635a6eceafb6f4fb3a0d95f6627ee98321c0 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4393497 Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Geoff Lang f1f80e1e 2023-03-22T12:54:08 Metal: Always prefer staged GPU uploads for textures. Uploading Metal texture data via MTLTexture::replaceRegion is a source of lots of CPU hangs and jank. There may be better heuristics to determine if we should do a CPU vs GPU upload but for now preferring GPU uploads has better overall performance. This also improves the MotionMark images test: - 40 -> 100 when using Metal. - 217 -> 235 when using Metal + https://chromium-review.googlesource.com/c/chromium/src/+/4091749 Bug: angleproject:8024, angleproject:8092, angleproject:8109 Change-Id: I36b5f585884391b4cc416365ae65f8542745beee Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4264963 Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Geoff Lang <geofflang@chromium.org> Reviewed-by: Jonah Ryan-Davis <jonahr@google.com>
Alexey Knyazev b24b5568 2023-03-27T00:00:00 Vulkan: Skip sample coverage for single sample rendering A new test sets sample coverage to zero and checks that it is applied only to multisampled rendering. Bug: angleproject:8102 Change-Id: I1a5649869e9b7ecf0543108fb99095bfaf6fd858 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4379839 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com>
Maksim Sisov 61cd24d4 2023-03-31T14:54:17 MTLPixelFormatRGB9E5Float is not renderable on simulator. See https://developer.apple.com/documentation/metal/ developing_metal_apps_that_run_in_simulator?language=objc Bug: angleproject:7994 Change-Id: I07e50c19aaf6d033c7c25d0dbd4b21a8c255c1d9 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4386403 Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Kenneth Russell <kbr@chromium.org>
Kimmo Kinnunen f176027f 2023-03-31T13:37:03 CGL: implement EGL_ANGLE_wait_until_work_scheduled It is as important to wait until scheduled in CGL, as it is in Metal. CGL waits until scheduled when glFlush() is called. Fixed: angleproject:8112 Change-Id: Id4a9e87804c6df1828b35cfd30c8427314820e52 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4386400 Commit-Queue: Kenneth Russell <kbr@chromium.org> Auto-Submit: Kimmo Kinnunen <kkinnunen@apple.com> Reviewed-by: Kenneth Russell <kbr@chromium.org>
Shahbaz Youssefi 552d4271 2023-03-31T10:59:30 Vulkan: Refactor buffer init logic Bug: b/276002151 Change-Id: I28d3fa34ab11340cc8b38743e87664a514870068 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4388547 Reviewed-by: Charlie Lao <cclao@google.com> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Yuxin Hu <yuxinhu@google.com>
Colin Blundell 6458cd95 2023-03-30T14:27:07 [GL backend] Remove over-zealous ASSERT in StateManagerGL.cpp StateManagerGL::bindVertexArray asserts that the passed-in |vaoState| pointer is non-null; however, we actually pass in null directly in [1]. This assert goes off when running a WebView test with ANGLE. [1] https://source.chromium.org/chromium/chromium/src/+/main:third_party/angle/src/libANGLE/renderer/gl/StateManagerGL.cpp;l=3385;drc=48340c1e35efad5fb0253025dcc36b3a9573e258;bpv=1;bpt=1 Bug: chromium:1168587 Change-Id: I14d5f9f317f3cbe2a828ab7637edc5f1911bf81c Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4381861 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Dean Jackson 5fcd027e 2023-03-29T08:54:34 Fix the watchOS build for the mirror clamp to edge extension The mirror clamp to edge extension relies on MTLSamplerAddressModeMirrorClampToEdge, which is not available on watchOS. Disable the extension there for the moment. Bug: angleproject:8103 Change-Id: I52b9d5217b39080cde6663e0d0fa839e53dacecf Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4379037 Reviewed-by: Kenneth Russell <kbr@chromium.org>
Charlie Lao 20f3df07 2023-03-30T11:36:29 Vulkan: ImageHelper::initExternal missing some variable init When ImageHelper::releaseImage is called and then re-initialized (due to mipmap change or usage change etc), right now we have a bug that mCurrentShaderReadStageMask and mLastNonShaderReadOnlyLayout are still left as is. And they do not get reset when image is initialized. This means if the image was reinitialized again, it may end up think the new image was previously used by fragment shader and insert an unnecessary barrier when new VkImage is used. This was discovered when I do the experimental work try to ghost VkImage for glTexSubImage call and realized that new VkImage was having a strong fragment->vertex dependency which it should not (since this is a new VkImage). This CL initialize mCurrentShaderReadStageMask and mLastNonShaderReadOnlyLayout and mCurrentQueueFamilyIndex from initExternal call since it is a new vkImage. Bug: b/273808966 Change-Id: I5bbeae5f1012f10c24620cfae8fa20365a7b5ab7 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4386136 Reviewed-by: Yuxin Hu <yuxinhu@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Charlie Lao 98151770 2023-03-27T09:49:33 Vulkan: Use midRenderPass clear if RP has started but inactive This CL extends prior CL's optimization so that if clear is issued right after blitFramebuffer call (this could make sense if blit and clear are on different buffer), we can keep the started render pass and do the midRenderPass clear instead of endRenderPass and start another renderPass. Bug: b/273808966 Change-Id: Ia2504e8e260867a6f797d42cd4c8a72f187280ef Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4374145 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Charlie Lao ad9537af 2023-03-16T16:03:29 Vulkan: Reactivate already started render pass when possible In some usage case (such as lineage_mobile), we are seeing in the middle of render pass, app switch to another fbo just to issue a glClear() call, which the clear call itself gets deferred. Application then switch back to the original frame buffer. Before this CL, the render pass gets recreated due to frame buffer binding change, even though the clear gets deferred and new render pass and the previous render pass are essentially the same. This CL detects this situation and reactivate the current render pass instead of creating a new one. With this CL, lineage_m app trace reduces frame time from 3.86ms to 3.7ms, and only one render pass is used instead of two. This CL also allows the render pass started by BlitFramebuffer reused by subsequent draw calls. Asphalt_9 is hitting this use pattern and this CL reduces frame time by 0.1245 ms (from 5.6203 ms to 5.4958 on pixel 7 pro) Bug: b/273808966 Change-Id: I48c2671cbef3ff9d6cf59caae88c37c77828ee07 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4348713 Reviewed-by: Amirali Abdolrashidi <abdolrashidi@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Charlie Lao <cclao@google.com>
Geoff Lang 3c8aee63 2023-03-27T13:24:47 D3D11: Fix streamed instanced attribute size calculation. D3D11 passes around a 0 instance count to signify "not an instanced draw call". This had an incorrect interaction with ComputeVertexBindingElementCount which checks for an instance count of 0 and skips instancing calculations. Force instance count to be at least 1 when computing buffer size needed. Bug: chromium:1425606 Change-Id: Ia10fa1c11abebce28c57ce4527e2b748e51516ce Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4375138 Reviewed-by: Jonah Ryan-Davis <jonahr@google.com>
Shahbaz Youssefi 48b431ca 2023-03-21T16:39:33 Vulkan: Fix stage component reservations w.r.t xfb Input components limit don't need to reserve any components. This change also fixes a bug where the tessellation evaluation output components didn't reserve a slot for gl_Position. Additionally, the input component count of stages are clamped by the output component count of the previous stages. Bug: b/274289482 Change-Id: Ib5b4243d7a4cb713ccea505a5f0064bf3377caa8 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4356115 Auto-Submit: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Yuxin Hu <yuxinhu@google.com> Reviewed-by: Yiwei Zhang <zzyiwei@chromium.org>
Yuxin Hu 29abd5a2 2023-03-06T11:35:45 Explicitly Add Aliased Memory Decoration in SpirV In GLSL, aliased memory qualifier is implicitly available, unless we explicitly specify the memory is restrict: https://www.khronos.org/opengl/wiki/Type_Qualifier_(GLSL)#Memory_qualifiers. However, in SpirV, aliased memory qualifier has to be explicitly specified: https://registry.khronos.org/SPIR-V/specs/unified1/SPIRV.html#AliasingSection This change adds the aliased memory decorations to SSBO if the GLSL does not specify the restrict memory qualifier. This is a temporary workaround to fix the deqp test failures on some android devices. Eventually we would like to waive this test for ANGLE due to alised memory in Vulkan does not suppose to affect SSBOs that have different set/binding, please refer to this Khronos ticket for more details: https://gitlab.khronos.org/Tracker/vk-gl-cts/-/issues/4321 Bug: b/266235549 Change-Id: Ic7afc417a5d421664b60e0413a011314787e14e7 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4312130 Reviewed-by: Yiwei Zhang <zzyiwei@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Shahbaz Youssefi 97897d92 2023-03-27T16:02:57 Vulkan: Work around driver bug with dynamic primitive restart This CL forces the state to be static on buggy drivers. Bug: b/275210062 Change-Id: Ia3391ecb19c3c9d19c05a83e11da8c718513a4e2 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4374104 Reviewed-by: Yuxin Hu <yuxinhu@google.com> Reviewed-by: Yiwei Zhang <zzyiwei@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@google.com>
Alexey Knyazev 6d12a280 2023-03-24T00:00:00 D3D11: Support NV_shader_noperspective_interpolation Fixed: angleproject:8096 Change-Id: Ic7c27c735661b2ef8095c81201b2ce0a82ad3ae7 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4368472 Reviewed-by: Geoff Lang <geofflang@chromium.org>
Geoff Lang da7dd31f 2023-03-27T11:29:10 GL: Disable extended dirty bit states for internal blits. New GL states have been added to the GL backend and not updated in ScopedGLState which is used to reset internal state for blits. This caused test failures when tests were run in specific orders. Ex: ClipDistanceTest.ThreeClipDistancesRedeclared/ES2_OpenGL:CopyTextureVariationsTest.CopySubTexture/ES2_OpenGL__AToBGRAFlipYUnmultiplyAlpha Bug: angleproject:5160 Bug: chromium:1410191 Change-Id: I805e39bc8c4c7d7d843b43fdc7813dd668919e8a Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4374101 Reviewed-by: Peng Huang <penghuang@chromium.org>
Geoff Lang c6ec59dc 2023-03-27T11:15:48 Explicitly pass the extended dirty bits to syncState. Add a the extended dirty bits and bit mask to syncState instead of calling gl::State::getAndResetExtendedDirtyBits when encountering DIRTY_BIT_EXTENDED. It disallowed us from masking the extended dirty bits and feels like an anti-pattern to modify the extended dirty bits in gl::State from the backend. This is a refactor only. Bug: chromium:1410191 Change-Id: I66fdec3eb57e3426cf0fda9ccb759700eafdda14 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4374100 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Yuxin Hu <yuxinhu@google.com>
Alexey Knyazev fc7cb00e 2023-03-24T00:00:00 Metal: Support NV_shader_noperspective_interpolation Fixed: angleproject:8095 Change-Id: Ieded1f704f82963984f1aee5072f152de0de374c Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4368473 Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Kramer Ge fbb16d64 2023-03-24T02:23:19 Mali: Disable disjointTimerQueries on ChromeOS with MaliT8xxOrOlder ChromeOS boards using MT8183 are encountering GL_OUT_OF_MEMORY errors when using disjointTimerQueries, which leads to meets visual effects freezing after a couple of seconds. Extend the workaround treatment to ChromeOS. Bug: chromium:1356053,b/269068358 Change-Id: I212f2b4a2fed1ddd8e8344bde4f4ca9069ea3393 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4369684 Commit-Queue: Kramer Ge <fangzhoug@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org> Auto-Submit: Kramer Ge <fangzhoug@chromium.org>
Colin Blundell abbbe970 2023-03-24T08:10:40 [GL backend] Get texture bindings size from frontend caps EGLContext currently sets the texture bindings size to be GL_MAX_COMBINED_TEXTURE_IMAGE_UNITS. StateManagerGL later copies these bindings into its own textures array [1]. However, the latter texture array is clamped to a size of max 96 [2], and thus this copy can overflow (see crash in crbug.com/1418842). This CL changes the texture bindings size to be set from the frontend caps rather than the native caps. The former have the same clamping as for StateManagerGL's textures. [1] https://source.chromium.org/chromium/chromium/src/+/main:third_party/angle/src/libANGLE/renderer/gl/StateManagerGL.cpp;drc=c85c710dec58dfdcf2f35af647a0a5155639edd5;l=3274 [2] https://source.chromium.org/chromium/chromium/src/+/main:third_party/angle/src/libANGLE/angletypes.h;drc=4a77b0f5237515960673b9bd1bc5a229f0eeaa95;l=952 [3] https://source.chromium.org/chromium/chromium/src/+/main:third_party/angle/src/libANGLE/Context.cpp;l=4050-4059;drc=b4bc946c63b2b95e1f05dec4e84adcadd10499c6 Bug: chromium:1418842 Change-Id: I66c10b640f70d9dc6c7d41925f12ffb7a3ba78e0 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4369699 Commit-Queue: Geoff Lang <geofflang@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Alexey Knyazev afc1e5dc 2023-03-27T00:00:00 GL: Support OES_sample_variables Bug: angleproject:8097 Change-Id: I983d6e83fe9ca942c91cbdf7274b46306e38f11f Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4368476 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com>
Hailin Zhang 1301402c 2023-03-23T21:56:45 Vulkan: improve etc to bc grident texture precison. ETC has gradient modes. Two endpoints in BC are RGB565; the color depth is lower than ETC. Two endpoints in RGB888 space may be quite different, but after conversion to RGB565 they may be projected to the same point. This conversion loses quite a lot of details. This change tries to change one endpoint to an adjacent one (not optimal) in RGB565, so that all the colors are interpolated from these two endpoints. With this change, the arknight image in b/259327729 improved PSNR from 39 to 49. This change includes some of work by @gregschlom 1) Use look up table instead of bit operation. 2) Use more vector operations. This may not improve the performance but makes the code short and neat. bug: b/259327729 Change-Id: Ic0720aff17af5d466e5598dc38ba953dc084a3cc Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4368021 Reviewed-by: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Hailin Zhang <hailinzhang@google.com>
Igor Nazarov 435e557c 2023-03-22T21:35:09 Vulkan: Do not enforce custom ANGLEs secondary command buffers. This change become possible after the commit: Vulkan: Fix Secondary Command Buffers with asyncCommandQueue. After the above change, `attach/detachAllocatorImpl<>()` will be skipped when using Vulkan secondary command buffers. Other places that use `mCommandAllocator` already no-op regardless of the allocator type. Therefore enforcing: ANGLE_USE_CUSTOM_VULKAN_OUTSIDE_RENDER_PASS_CMD_BUFFERS ANGLE_USE_CUSTOM_VULKAN_RENDER_PASS_CMD_BUFFERS when enabling `angle_enable_vulkan_shared_ring_buffer_cmd_alloc` is no longer required. This will remove possible confusion when configuring ANGLE build and allow mixing Vulkan secondary command buffers and ANGLEs custom secondary buffers that use `shared_ring_buffer_cmd_alloc`. Bug: angleproject:6401 Bug: b/256666069 Change-Id: I52a64065e6ce9e4aabbf76031c3f5443e0e26a3a Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4360071 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Igor Nazarov <i.nazarov@samsung.com> Reviewed-by: Charlie Lao <cclao@google.com>
Igor Nazarov aa8f2e4b 2023-03-22T19:53:04 Vulkan: Minor SecondaryCommandPool::destroy() optimization. Follow up for the commit: 9b6368ccfceddf1784ef41e95e57696453cca71d Vulkan: Fix freeing Secondary Command Buffers from wrong thread. Bug: angleproject:6100 Change-Id: I8ef34898941c0c5684d3ea4378332cdf6b87c0c8 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4359703 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Igor Nazarov <i.nazarov@samsung.com> Reviewed-by: Charlie Lao <cclao@google.com>
Igor Nazarov 85735eb4 2023-03-17T19:56:23 Vulkan: Enable async features for Secondary Command Buffers. Since, all known problems with Secondary Command Buffers and async features are fixed, it should be safe to allow enabling these features for testing. Locally, this CL passes custom build dEQP tests (EGL/GLES2/3/3.1) with asyncCommandQueue and Vulkan Secondary Command buffers enabled. Tests were executed on S906B that uses ANGLE as the system GLES driver. Bug: angleproject:6811 Bug: angleproject:6100 Change-Id: Id3ef6df56f0a3bd1986f2e6047f01e6eac165fc6 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4350270 Reviewed-by: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Igor Nazarov <i.nazarov@samsung.com>
Igor Nazarov e27e7c6a 2023-03-17T19:43:54 Vulkan: Retire Command Buffers before destroying the Pools. "VulkanSecondaryCommandBuffer"s may be still in "CommandQueue" when "ContextVk" destroys the Command Pools (asyncCommandBufferReset = true). Fixed by calling "retireFinishedCommandsLocked()" when appropriate. Normally, it is only required to retire Vulkan Secondary Buffers. However, some drivers may have bug and also require to reset Primary Buffers, that have Secondaries recorded, before destroying the Secondaries Pools. Bug: angleproject:6811 Bug: angleproject:6100 Change-Id: I891547c95cfbdfab44398980f939596af56ab57b Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4350269 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Igor Nazarov <i.nazarov@samsung.com> Reviewed-by: Charlie Lao <cclao@google.com>
Igor Nazarov 9524c639 2023-01-17T18:47:47 Vulkan: Fix Secondary Command Buffers with asyncCommandQueue. This change fixes following errors (asyncCommandQueue = true): 1. "vkEndCommandBuffer()" called from "asyncCommandQueue" thread. Call stack: vkEndCommandBuffer() VulkanSecondaryCommandBuffer::end() OutsideRenderPassCommandBufferHelper::flushToPrimary() CommandQueue::flushOutsideRPCommands() CommandProcessor::processTask() Fixed by calling "vkEndCommandBuffer()" from the Context thread in the new "OutsideRenderPassCommandBufferHelper::detachCommandPool()" method. 2. "vkAllocateCommandBuffers()/vkBeginCommandBuffer()" called from "asyncCommandQueue" thread. Call stack: vkAllocateCommandBuffers()/vkBeginCommandBuffer() VulkanSecondaryCommandBuffer::initialize() <*>CommandBufferHelper::initializeCommandBuffer() <*>CommandBufferHelper::reset() <*>CommandBufferHelper::flushToPrimary() CommandQueue::flush<*>Commands() CommandProcessor::processTask() Fixed by calling "vkAllocateCommandBuffers()/vkBeginCommandBuffer()" from the Context thread in the new "<*>CommandBufferHelper::attachCommandPool()" method. 3. "SecondaryCommandPool::collect()" called from "asyncCommandQueue" thread without synchronization. Call stack: SecondaryCommandPool::collect() rx::vk::RecycleCommandBufferHelper() CommandBufferRecycler<>::recycleCommandBufferHelper) RendererVk::recycle<*>CommandBufferHelper() CommandProcessor::processTask() No need for this call, because "SecondaryCommandPool" is already detached. Notes: This CL not only fixes errors, but also optimizes CommandBufferHelper recycling. Before, there was no recycling plus unnecessary "VulkanSecondaryCommandBuffer" allocation/freeing. Further optimization may add multiple "VkCommandPool"s to the "SecondaryCommandPool" to allow resetting buffers in the async thread. Bug: angleproject:6811 Bug: angleproject:6100 Change-Id: I7004c27a112e916c5c973b43b137193017d6aa3d Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4342189 Reviewed-by: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Igor Nazarov <i.nazarov@samsung.com>
Amirali Abdolrashidi f2c5ce4e 2023-03-16T17:44:56 Re-enable mutable texture upload for one context It was observed that in some apps, mutable textures are uploaded in a context, but that context's outside RP command buffer is not flushed at all, resulting in invalid (noise) or incorrect textures. * Added isEligibleForMutableTextureFlush() to be used to determine whether onMutableTextureUpload() should be called. * Restricted the use of mutable texture upload to single-context share groups. * When the number of contexts becomes greater than one, the original context's outside render pass command buffer is flushed. * Added related tests. * Added a test to make sure that textures can be uploaded in one thread, and used for draw calls in another, with the help of a sync object. * Added a test to make sure that textures can be uploaded in the main thread, and used in a new thread. * Added a test in which a new context is created after mutable mipmap textures are defined. Bug: b/264143971 Change-Id: I66c0d8b04d39bb7244e5752aac0e46a0192f012e Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4349156 Reviewed-by: Charlie Lao <cclao@google.com> Commit-Queue: Amirali Abdolrashidi <abdolrashidi@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Alexey Knyazev 997c4c7b 2023-03-21T00:00:00 D3D11: Implement EXT_conservative_depth Drive-by: * Hid the extension from ES 2.0 client contexts. Bug: angleproject:8046 Change-Id: I8bca4161dde4bda7ee75541b9164777884900d89 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4366784 Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com>
Alexey Knyazev 648f3657 2023-03-21T00:00:00 Metal: Implement EXT_conservative_depth Bug: angleproject:8046 Change-Id: I3b9f497da19bf005c7948e552f2e1c7415d4ac6d Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4363020 Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com> Reviewed-by: Kyle Piddington <kpiddington@apple.com> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Igor Nazarov 0e502f8c 2023-03-20T20:34:28 Vulkan: Update SharedCommandBlockPool::valid() method. Current logic returns `valid()` result whether `mAllocator` is attached or not. However this is not correct, because it is allowed to use some method even after allocator is detached. Since this class does not require initialization - it is always valid. Bug: angleproject:6401 Bug: b/256666069 Change-Id: I57f899ae7718762b7e2b68e9c3d1e04dccb2c351 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4355391 Reviewed-by: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Igor Nazarov <i.nazarov@samsung.com>
Kramer Ge 888ca8d9 2023-03-21T18:56:56 Bind FBO before timer query on Mali GL driver. glBeginQuery(GL_TIME_ELAPSED/GL_TIMESTAMP) on Mali implementation assumes a complete FrameBuffer. Without it glGetQueryObject will return a meaningless value, causing some applications to misbehave. This workaround caches and binds a default FBO in this case. Bug: chromium:1356053,b/269068358 Change-Id: I756ded948c2c5aada744f9dd428ad77c37a009c6 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4359032 Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Kramer Ge <fangzhoug@chromium.org>
Alexey Knyazev ad52f12e 2023-03-21T00:00:00 Implement EXT_conservative_depth Added translator frontend and GLSL backend support for gl_FragDepth redeclaration and layout qualifiers. Added mappings to DepthGreater, DepthLess, and DepthUnchanged SPIR-V execution modes. Bug: angleproject:8046 Change-Id: I23f19ff54380741107970a44055ea269eef179f6 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4355028 Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Greg Schlomoff a621ea88 2023-03-17T10:21:56 Adding a trace point for texture metrics. Bug: b/236121838 Change-Id: I28e7977f7b5dc6e24cb5a2be10689c223851ba0b Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4348737 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Doug Horn <doughorn@google.com> Reviewed-by: Charlie Lao <cclao@google.com>
Alexey Knyazev ba845fcf 2023-03-20T00:00:00 Metal: Implement EXT_clip_control * Skipped vertex shader depth correction when clip depth mode is set to GL_ZERO_TO_ONE. * Adjusted front face mode and pre-fragment vertical flip driver uniform based on the clip origin state. * Added more clip control end2end tests. * D3D11: Fixed scissored draws with upper-left clip origin. Fixed: angleproject:8065 Change-Id: Ife82fa2ad39fad463c23ec10b49bff409272ebf2 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4356661 Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com> Reviewed-by: Kyle Piddington <kpiddington@apple.com>
Alexey Knyazev e809e7bd 2023-03-13T00:00:00 Reland "Implement EXT_depth_clamp" This is a reland of commit f8c1418319ac2aef4b3101e322005b1d0f73120f Host GPU bugs are observable in iOS Simulator Original change's description: > Implement EXT_depth_clamp > > * Added depthClamp to the RasterizerState > * Added DepthWriteTest end2end tests covering > both clipped and clamped depth writes > > Capture > * Updated serialized rasterizer state > * Updated CaptureMidExecutionSetup > > OpenGL > * Requires GL 3.2 or ARB_depth_clamp > on desktop contexts > * Maps to EXT_depth_clamp on ES > > D3D11 > * Maps to the opposite of > D3D11_RASTERIZER_DESC.DepthClipEnable > * The new tests uncover several edge cases where > a workaround is needed to implement unextended > OpenGL semantics on top of D3D > > Metal > * Maps to the setDepthClipMode command > > Bug: angleproject:8047 > Bug: angleproject:8077 > Change-Id: I1b3448e5b84443e4be18af9bc22d2f8495ac8267 > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4347753 > Reviewed-by: Geoff Lang <geofflang@chromium.org> > Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com> Bug: angleproject:8047 Bug: angleproject:8077 Change-Id: I8c5f8304276c97c51b2c3382cd2764592ee0c3fe Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4349938 Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com>
Yuxin Hu 4a77b0f5 2023-03-18T00:16:24 Revert "Implement EXT_depth_clamp" This reverts commit f8c1418319ac2aef4b3101e322005b1d0f73120f. Reason for revert: This change breaks angle_end2end_tests on Metal backend: https://ci.chromium.org/ui/p/chromium/builders/ci/ios-angle-intel/26035/overview Original change's description: > Implement EXT_depth_clamp > > * Added depthClamp to the RasterizerState > * Added DepthWriteTest end2end tests covering > both clipped and clamped depth writes > > Capture > * Updated serialized rasterizer state > * Updated CaptureMidExecutionSetup > > OpenGL > * Requires GL 3.2 or ARB_depth_clamp > on desktop contexts > * Maps to EXT_depth_clamp on ES > > D3D11 > * Maps to the opposite of > D3D11_RASTERIZER_DESC.DepthClipEnable > * The new tests uncover several edge cases where > a workaround is needed to implement unextended > OpenGL semantics on top of D3D > > Metal > * Maps to the setDepthClipMode command > > Bug: angleproject:8047 > Bug: angleproject:8077 > Change-Id: I1b3448e5b84443e4be18af9bc22d2f8495ac8267 > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4347753 > Reviewed-by: Geoff Lang <geofflang@chromium.org> > Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com> Bug: angleproject:8047 Bug: angleproject:8077 Change-Id: I829add68c006c72b7b4acf03aee3efa8a9a16fac No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4350876 Reviewed-by: Alexis Hétu <sugoi@google.com> Reviewed-by: Yuxin Hu <yuxinhu@google.com> Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com> Commit-Queue: Yuxin Hu <yuxinhu@google.com>
Igor Nazarov 7d1a401b 2023-01-17T18:45:55 Vulkan: Fix freeing not completed Secondary Command Buffers. Problem: - Protected Context flushes its commands to the Protected Primary Command Buffer; - Unprotected Context flushes its commands to the Unprotected Primary Command Buffer; - Context with different "egl::ContextPriority" may flush commands into different Primary Command Buffers. - Secondary Command Buffers from all Contexts end-up in the single "CommandBufferRecycler::mSecondaryCommandBuffersToReset" list; - One of the Contexts submits its Primary Command Buffer, and attaches current "mSecondaryCommandBuffersToReset" list to the "CommandBatch"; - Secondary Command Buffers of other Contexts may be collected and later freed by "SecondaryCommandPool" without submitting/completion corresponding Primary Command Buffers. Fix: - Moving "mSecondaryCommandBuffersToReset" to the new "SecondaryCommandBufferCollector" class. - Separate "SecondaryCommandBufferCollector" instance is stored in the "CommandQueue" for each current Primary Command Buffer. Additionally fixes "asyncCommandQueue" related problem: "releaseCommandBuffersToReset()" may get outdated results if flush is not yet executed in the "asyncCommandQueue" thread. Bug: angleproject:6100 Change-Id: I7df161ac1f999fb34d4eccaebb603c58ecb1ac11 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4334579 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Charlie Lao <cclao@google.com> Commit-Queue: Igor Nazarov <i.nazarov@samsung.com>
Igor Nazarov 9b6368cc 2023-03-14T14:48:30 Vulkan: Fix freeing Secondary Command Buffers from wrong thread. Problem: - Secondary Command Buffers are freed in the "CommandQueue" class. - This may happen from any Context thread that calls "checkCompletedCommands()" or "finish<*>()" methods. - As the result, one Command Buffer may be freed from one thread, while other Command Buffer from the same "VkCommandPool" is allocated/reset/recorded in the other thread. Vulkan spec demands external "VkCommandPool" synchronization for any modifications (begin/end/reset/free/cmd) on its "VkCommandBuffer"s. Fix: - Added new "rx::vk::SecondaryCommandPool" class that replaces the "rx::vk::CommandPool" wrapper. - This class has "collect()" method for storing "VkCommandBuffer"s. Collected buffers are freed from the correct thread on the next "allocate()" call. This CL only fixes the problem, keeping Secondary Command Buffer memory management as is (allocate/free single buffer without reuse). In the future CLs this behavior may be changed (reuse buffers, reset/free entire pools). Bug: angleproject:6100 Change-Id: If938416c4df4fe55f0cfb418b6759721ac53098b Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4334577 Reviewed-by: Charlie Lao <cclao@google.com> Commit-Queue: Igor Nazarov <i.nazarov@samsung.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Igor Nazarov 29f80eda 2023-03-15T14:59:26 Vulkan: Hot fix crash when using Invalid VkSemaphore Object. Problem started after the commit: b194c21ad30145ca15153a93425a37a8211df373 Vulkan: Enforce ContextPriority in ShareGroup and with EGLImage Happens only with enabled "asyncCommandQueue" feature: - "RendererVk::submitPriorityDependency()" creates "Vk::Semaphore" object on the stack. - Then submits this Semaphore with "queueSubmitWaitSemaphore()" method. - "vk::Semaphore" passed to the "CommandProcessor"s "enqueueSubmitOneOffCommands()" method where stored in the "CommandProcessorTask". - Stack "vk::Semaphore" object garbage collected by the "RendererVk" and then destroyed. - "CommandProcessorTask" now points to the invalid memory. This CL changes usage of "vk::Semaphore" and "vk::Fence" pointers in the "CommandProcessor.h" classes to Vulkan Handles. This will make API consistent, fix this problem, and avoid similar problems in the future. Issue discovered while testing ANGLE as a system driver on S906B with "asyncCommandQueue" enabled. Several system processes crashed with VVL: VUID-VkSubmitInfo-pWaitSemaphores-parameter(ERROR / SPEC):msgNum: -1328048864 - Validation Error: [ VUID-VkSubmitInfo-pWaitSemaphores-parameter ] Object 0: handle = 0xb4000078c1aab430, type = VK_OBJECT_TYPE_INSTANCE; | MessageID = 0xb0d79520 | Invalid VkSemaphore Object 0xb400007831740b98. The Vulkan spec states: If waitSemaphoreCount is not 0, pWaitSemaphores must be a valid pointer to an array of waitSemaphoreCount valid VkSemaphore handles (https://www.khronos.org/registry/vulkan/specs/1.3-extensions/html/vkspec.html#VUID-VkSubmitInfo-pWaitSemaphores-parameter) Bug: angleproject:8039 Change-Id: I8818288b9783b5c5a7970bf82ec721452ae57471 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4339758 Reviewed-by: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Igor Nazarov <i.nazarov@samsung.com>
Igor Nazarov 12b3d52d 2023-03-10T17:02:55 Prevent bugs in "FastVector" class. FastVector does not call destructors when resizing down and constructors when resizing up. This may cause reuse of previous values after resizing up and prevent releasing resource in destructors. Above problems only relevant for non trivially constructible/destructible types. For performance reasons (chromium:1417087) this CL disables using trivially destructible types at all and adds special resizing methods for non trivially constructible types. Almost all uses of FastVector was already using trivially destructible types. Except "angle::FixedVector" used in "rx::vk::SubpassVector<>". However, "angle::FixedVector" is unnecessarily calls "clear()" in the destructor. This CL removed "clear()" and made the destructor trivial. All non trivial constructor cases are limited to "angle::spirv::BoxedUint32<>" used in "angle::spirv::FastVectorHelper<>". No problems found because of the possibility to reuse previous values. All "resize(count)" methods replaced with special versions. Therefore, this CL does not fix actual bugs, but rather prevents incorrect use in the future. Bug: angleproject:8021 Bug: chromium:1417087 Change-Id: Id65d75575f2f582450b1cc45dc6b1f2bf3bc5289 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4328286 Reviewed-by: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Igor Nazarov <i.nazarov@samsung.com>
Alexey Knyazev f8c14183 2023-03-13T00:00:00 Implement EXT_depth_clamp * Added depthClamp to the RasterizerState * Added DepthWriteTest end2end tests covering both clipped and clamped depth writes Capture * Updated serialized rasterizer state * Updated CaptureMidExecutionSetup OpenGL * Requires GL 3.2 or ARB_depth_clamp on desktop contexts * Maps to EXT_depth_clamp on ES D3D11 * Maps to the opposite of D3D11_RASTERIZER_DESC.DepthClipEnable * The new tests uncover several edge cases where a workaround is needed to implement unextended OpenGL semantics on top of D3D Metal * Maps to the setDepthClipMode command Bug: angleproject:8047 Bug: angleproject:8077 Change-Id: I1b3448e5b84443e4be18af9bc22d2f8495ac8267 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4347753 Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com>
Yiwei Zhang 12aefbc0 2023-03-16T04:43:38 Vulkan: add MESA Virtio-GPU Venus driver feature conditions Bug: b/267576238 Change-Id: I5f0c479b23cd3465ca7560966bb4d98edf4c40df Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4342819 Auto-Submit: Yiwei Zhang <zzyiwei@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Charlie Lao f4e71351 2023-03-14T14:55:04 Vulkan: Switch acquireAndUpdate to use Buddy pool Based on survey of all app traces we have, it is common that we end up with BufferVk::acquireAndUpdate even though the buffer was created with STATIC usage. This is mostly due to glBufferSubData call on the STATIC usage buffers and on ARM we most likely end up with acquireAndUpdate. Similarly, we also getting into ghostMappedBuffer and mapRangeImpl with STATIC usage buffers, even though with less app traces. Since the usage pattern usually repeats, using generic allocation algorithm has performance penalty. This CL moves these usage to buddy algorithm to ensure alloc/free are fast. This CL and previous CL crrev.com/c/4327290, reduces efootball_pes-2021 frame time from 4.2 ms to 2.87 ms, achieves parity with native GLES on pixel 7 pro. Bug: b/271915956 Change-Id: I56e0195181c77a3130513c74ec8a5075b2b29ea4 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4321870 Commit-Queue: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Amirali Abdolrashidi <abdolrashidi@google.com>
Charlie Lao 4c157b4b 2023-03-14T18:45:38 Vulkan: Switch staging buffer to Buddy algorithm Staging buffer should be considered as dynamic and uses buddy allocator. Bug: b/271915956 Change-Id: I7cbe3765fdae120582034b24376560043e007e67 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4327290 Commit-Queue: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Amirali Abdolrashidi <abdolrashidi@google.com>
Austin Eng a8401f03 2023-03-15T09:29:14 Mark the context as needing flush in TextureGL set*Image operations Bug: chromium:1181068, chromium:1418291 Change-Id: I70b158dcec54edf20b44fcc6169d1fc1d86a0de7 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4341190 Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Austin Eng <enga@chromium.org>
Charlie Lao 4982b903 2023-03-14T19:56:51 Revert "Vulkan: Remove inUseAndRespecifiedWithoutData from BufferVk" This reverts commit 755bfe471d23bc2aac5e78493537801dc5f90792. Reason for revert: Causing flaky on pixel 6 angleproject:8082 Original change's description: > Vulkan: Remove inUseAndRespecifiedWithoutData from BufferVk > > BufferVk::setDataWithMemoryType() has one optimization that it tries to > detect glBufferData(target, size, nullptr, usage) and if existing > storage is busy, it immediately reallocate storage. With the > optimization in previous CL (crrev.com/c/4317488), the storage reuse > logic should detect if we can reuse the storage or not. If the size > matches the existing storage's size, then there is no reason we can not > reuse existing storage. Later on when glBufferSubData or > glMapBufferRange is called, there are optimization in those calls that > will detect if we should reallocate storage or not as the further > optimization. This CL removes this check and replies on the other > optimization to handle the storage reallocate (shadowing) if necessary. > This simplifies code and also potentially avoids storage reallocation in > certain usage cases. > > This CL also fixes a test bug in > BufferDataTestES3.BufferDataWithNullFollowedByMap that was calling > glMapBufferRange with MAP_UNSYNCHRONIZED_BIT but incorrectly expecting > GL to do synchronization. > > Bug: b/271915956 > Change-Id: I7901687b3e3e262e77699f14eb8602d8a57eda3e > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4322048 > Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> > Commit-Queue: Charlie Lao <cclao@google.com> > Reviewed-by: Amirali Abdolrashidi <abdolrashidi@google.com> Bug: b/271915956 Change-Id: Ie5716b609ab96b96afbe5927f20dfcf2bf5d4db6 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4338263 Auto-Submit: Charlie Lao <cclao@google.com> Commit-Queue: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com> Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Geoff Lang a8720455 2023-03-10T13:48:03 D3D11: Add logic to disassociate EGL image storages. The TextureStorage classes for External and EGLImages were missing the logic to disassociate from images. This lead to the images continuing to hold references to deleted storages. Bug: chromium:1415330 Change-Id: I8303f6751d87a9b0a52993c7d4e9509b086b93f3 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4328347 Reviewed-by: Peng Huang <penghuang@chromium.org> Commit-Queue: Geoff Lang <geofflang@chromium.org>
Austin Eng 9c167fd2 2023-03-14T10:37:39 Mark the context as needing flush in TextureGL copy/blit operations Bug: chromium:1181068, chromium:1418291 Change-Id: Ibce7d27ba33e7ed3a05ea0226ef60aa31682286a Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4337957 Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Geoff Lang <geofflang@chromium.org>
Chris Dalton 513ca723 2023-03-13T14:30:20 Delete GL_LOAD_OP_DISABLE_ANGLE from PLS This load op makes the PLS spec and WebGL implementations unnecessarily complex. Bug: chromium:1421437 Change-Id: Iab02a8b02083899c6cc345ecb25b88c5871611c9 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4335148 Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Chris Dalton <chris@rive.app> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Alexey Knyazev b26b01a2 2023-03-09T00:00:00 Implement EXT_render_snorm Bug: angleproject:8048 Change-Id: Id01beaea9565f8ab374c732fef70ec0ac0d8743e Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4334303 Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com>
Alexey Knyazev bf5e9dbc 2023-03-09T00:00:00 GL: Reset clip origin before scissored clears Clip origin must not affect scissor box but some drivers flip it for clear operations. Added capture/replay support for ClipControl. Bug: angleproject:8066 Change-Id: I9292cb4945b49c56c80da4c5813e89df3453b6b6 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4328267 Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com>
Charlie Lao 755bfe47 2023-03-08T14:31:33 Vulkan: Remove inUseAndRespecifiedWithoutData from BufferVk BufferVk::setDataWithMemoryType() has one optimization that it tries to detect glBufferData(target, size, nullptr, usage) and if existing storage is busy, it immediately reallocate storage. With the optimization in previous CL (crrev.com/c/4317488), the storage reuse logic should detect if we can reuse the storage or not. If the size matches the existing storage's size, then there is no reason we can not reuse existing storage. Later on when glBufferSubData or glMapBufferRange is called, there are optimization in those calls that will detect if we should reallocate storage or not as the further optimization. This CL removes this check and replies on the other optimization to handle the storage reallocate (shadowing) if necessary. This simplifies code and also potentially avoids storage reallocation in certain usage cases. This CL also fixes a test bug in BufferDataTestES3.BufferDataWithNullFollowedByMap that was calling glMapBufferRange with MAP_UNSYNCHRONIZED_BIT but incorrectly expecting GL to do synchronization. Bug: b/271915956 Change-Id: I7901687b3e3e262e77699f14eb8602d8a57eda3e Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4322048 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Charlie Lao <cclao@google.com> Reviewed-by: Amirali Abdolrashidi <abdolrashidi@google.com>
Charlie Lao d6a25bfa 2023-03-07T15:06:10 Vulkan: Optimize glBufferData call to improve storage reuse If app calls glBufferData with certain size, then calls it again with size 0, and then call it again with same old size again, we should try to reuse the existing storage. When size is zero, with the existing logic, we never free the storage. When glBufferData is called third time with the same size as the first glBufferData call, we expect to reuse the existing storage. But because of the storage reuse logic is comparing buffer's new size to the old size (which is 0), we missed the opportunity to reuse the existing storage. This CL update the reuse logic so that it checks the new size against storage's size (instead of OpenGLES buffer's size) and if we will end up with same sized allocation and same pool and memory type, then we reuse instead of reallocate. This reduces efootball_pes_2021 frame time from 4.670 ms to 4.277 ms on pixel 7 pro. Bug: b/271915956 Change-Id: I6f91e3e85b104eca215b28e7d0bea413ecc4401c Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4317488 Reviewed-by: Amirali Abdolrashidi <abdolrashidi@google.com> Commit-Queue: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Charlie Lao 67ad3ddc 2023-03-06T16:44:36 Vulkan: Relax size limit for dynamicBuffer to pick buddy algorithm If glBufferData's usage is one of the dynamic usage, app may keep calling glBufferData frequently, which means get into suballocation code frequently. There are two suballocation algorithms today: buddy algorithm (faster) and generic (slower). Right now the decision of which algorithm (i.e, which pool) to use is purely based on size or memory type. This CL also utilize usage information so that dynamic usage will pick buddy algorithm with bigger size threshold. mSmallBufferPool is removed and replaced with the BufferPoolPointerArray that gets picked based on allocation algorithm. This CL reduces average frame time of efootball_pes_2021 from 7.518 ms to 4.670 ms on pixel 7 Pro. Bug: b/271915956 Change-Id: I1c2f270ac49f56e6f405501d20691cfbab49e7eb Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4313685 Reviewed-by: Amirali Abdolrashidi <abdolrashidi@google.com> Commit-Queue: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Igor Nazarov 4edccb15 2023-01-17T16:17:35 Vulkan: Fixed Context Priority mixing problems. Problem details: - Each "egl::ContextPriority" may have separate "VkQueue". - "CommandQueue" only has two "PrimaryCommandBuffer"s for normal and protected content. - Commands from multiple "ContextVk" may be written to a single "PrimaryCommandBuffer". - That "PrimaryCommandBuffer" may be randomly submitted to different "VkQueue"s. - As the result - Commands from a single "ContextVk" may be submitted to multiple "VkQueue"s. Fix details: - Created separate "PrimaryCommandBuffer" (lazily allocated) for each "egl::ContextPriority". - Commands with different priorities can't be mixed in a single "PrimaryCommandBuffer". - Therefore - Commands from a single "ContextVk" will be submitted to a single "VkQueue". - No difference for applications that use single "egl::ContextPriority" for all Contexts. Notes: Another problem when resource is used in multiple Contexts with different "VkQueue" (Priority). One solution is to use Semaphores. Another is to enfore same Priority for all Contexts in a Share Group and Default Priority when using EGLImage. This solution was submitted in the previous CL: Vulkan: Enforce ContextPriority in ShareGroup and with EGLImage Below test fails on G996B without this CL. Bug: angleproject:8039 Test: angle_end2end_tests --gtest_filter=MultithreadingTestES3.ContextPriorityMixing* Change-Id: Iaa57826ca55956944f922813fcfac42f1a764dbb Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4194183 Reviewed-by: Charlie Lao <cclao@google.com> Commit-Queue: Igor Nazarov <i.nazarov@samsung.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Igor Nazarov cd901cdd 2023-03-03T17:04:27 Vulkan: Add and use rx::vk::ReleasableResource class. This is a follow up for CL: Vulkan: Enforce ContextPriority in ShareGroup and with EGLImage Bug: angleproject:8039 Change-Id: I9e654557d4a1ce9aee4b04f1211eeb6ae3f0e482 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4306721 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Charlie Lao <cclao@google.com> Commit-Queue: Igor Nazarov <i.nazarov@samsung.com>
Igor Nazarov 233c128b 2023-01-17T19:21:58 Vulkan: Fix UBs when deleted attachment is used in a RenderPass. Problem: - "RenderbufferVk"/"TextureVk" with "mOwnsImage == false" used as RenderPass attachment. - "RenderbufferVk"/"TextureVk" deleted. - Owning resource is destroyed ("EGLImage" and all siblings / "EGLSurface"). - Crash (UB) may happen when ending RenderPass, flushing or executing commands. Fix adds tracking of "vk::ImageSourceID" value in "vk::RenderPassAttachment" - IDs of objects, that originally provide "vk::ImageHelper" images. This is necessary, because when using EGLImage, there may be multiple "TextureVk" objects with same "vk::ImageHelper", and need to call "finalizeImageLayout()" for the correct attachment. Bug: angleproject:8032 Test: angle_end2end_tests --gtest_filter=ImageTest*DeletedWhileInUse* Test: angle_end2end_tests --gtest_filter=PbufferTest.UseAsFramebufferColorThenDestroy* Change-Id: I50fdd9d6b6a9677adad2262373303b46de1dee4c Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4296014 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Charlie Lao <cclao@google.com> Commit-Queue: Igor Nazarov <i.nazarov@samsung.com>
Igor Nazarov eb1cb31d 2023-02-21T14:09:49 Vulkan: Remove code left after introduce of "vk::SharedGarbage" This logic calls "flushImpl()" each time some "ImageVk" is orphaned, regardless if it used in the RenderPass or not. Such undesired flushes negatively affect CPU and GPU performance. This flush was added in the very old commit: e755a5374f7eb24da579fdc9862b01e3c3c04721 Vulkan: Add a new garbage type gated by fences. Flush was necessary to grab a proper Fence. However, after commit: f10bf6bf55a78669bff7bb5cdd3ae0954a87661e Vulkan: Implement multi-threaded GL. Fence was replaced by "vk::SharedGarbage" and "vk::SharedResourceUse". But "flushImpl()" was not removed along with misleading comment, that it is necessary "to make sure the fence has been submitted". This CL removes this leftover code. Any regressions should be fixed in a better way. Bug: angleproject:2464 Change-Id: I640bb2b9519c15a47adf30e0de845a3125ceab42 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4272834 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Igor Nazarov <i.nazarov@samsung.com> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Igor Nazarov a1bf828d 2023-02-13T21:30:38 Vulkan: Rename "RendererVk::waitFor*ToBeSubmitted()" methods. Following "RendererVk" methods are a little bit confusing: - submitCommands() - hasResourceUseSubmitted() - hasQueueSerialSubmitted() - waitForResourceUseToBeSubmitted() - waitForQueueSerialToBeSubmitted() Because after "RendererVk::submitCommands(..., submitQueueSerial)" call "hasQueueSerialSubmitted(submitQueueSerial)" will always return "true". And it is not clear why need to call "waitForResourceUseToBeSubmitted()" method, if it already "Submitted". It is even more: it is technically illegal to call "waitFor*ToBeSubmitted()" if "has*Submitted" is "false". This refactoring suggests adding "ToDevice" to the methods names: - waitForResourceUseToBeSubmittedToDevice() - waitForQueueSerialToBeSubmittedToDevice() So that: - "Submitted" - will mean to the RendererVk (and maybe Device) - "SubmittedToDevice" - definitely submitted to the Device. Bug: b/267348918 Change-Id: I12323be3ddc0cbcff4667e52a37089b187b63fe8 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4245423 Reviewed-by: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Igor Nazarov <i.nazarov@samsung.com>
Igor Nazarov b194c21a 2023-02-24T15:41:00 Vulkan: Enforce ContextPriority in ShareGroup and with EGLImage This CL enforces single Context Priority for all Contexts in a Share Group. This is necessary until Vulkan Semaphores will be used to automatically synchronize Resource access between Contexts. Contexts Priority updated when new Contexts is added to the Share Group. New Priority will be the highest among all ever existed Contexts (except if Priority is locked). When Contexts Priority changes, all flushed commands are submitted to the old VkQueue and semaphore is inserted into the new VkQueue. Currently opened RenderPasses and commands will not be flushed. When EGLImage is used in a Context, all Contexts in that Share Group locked (forever) to the Default Priority (Medium). This is done to simplify the implementation and because of the current limitations (lack of mutex protection across Context Share Groups). Notes: - the EGL_CONTEXT_PRIORITY_LEVEL_IMG will report initial priority. - below tests fail on G996B without this CL. Bug: angleproject:8039 Test: angle_end2end_tests --gtest_filter=MultithreadingTestES3.RenderThenSampleDifferentContextPriority* Test: angle_end2end_tests --gtest_filter=MultithreadingTestES3.RenderThenSampleInNewContextWithDifferentPriority* Test: angle_end2end_tests --gtest_filter=MultithreadingTestES3.RenderThenSampleDifferentContextPriorityUsingEGLImage* Change-Id: Ia6a2f0084d39168a58fd7ec33edc90ece9cead05 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4289750 Reviewed-by: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Igor Nazarov <i.nazarov@samsung.com>
Roman Lavrov 63a60225 2023-03-09T12:36:57 Ignore VUID-VkGraphicsPipelineCreateInfo-None-06573. Occurring in certain trace tests after vulkan-deps update. Bug: angleproject:8076 Change-Id: I267ff7ff5a41cdcfec44b11893877120d02a89ce Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4324721 Commit-Queue: Roman Lavrov <romanl@google.com> Auto-Submit: Roman Lavrov <romanl@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Roman Lavrov 7ee61366 2023-03-09T16:18:56 Revert "Vulkan: SurfaceVk should only wait for GPU work that uses it" This reverts commit 5b63e1dc1d5e96018d0ad30582265a612d309f3b. Reason for revert: angle_deqp_internal_main_presubmit device lost in android rolls https://anglebug.com/8073 Original change's description: > Vulkan: SurfaceVk should only wait for GPU work that uses it > > Right now when we destroy swapchain, we call mRenderer->finish() to > finish everything, even though the work is unrelated to this surface. > This CL changes it to only wait for ResourceUse of all images in the > swapChain. > > Bug: b/267806287 > Change-Id: I33d136ad50961fbf5fbb200ff0f89f1dbf23585d > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4220723 > Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> > Commit-Queue: Charlie Lao <cclao@google.com> Bug: b/267806287 Bug: angleproject:8073 Change-Id: Id0bfe7be670b7d44ed91e61f2c98a189bd1f214c Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4324718 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Roman Lavrov <romanl@google.com>
Dan Glastonbury aa5b97de 2023-03-06T17:59:43 ANGLE_metal_shared_event_sync: Control signaling external events It was assumed that the external MTLSharedEvent passed to eglCreateSync should be signaled by the GL. This change adds EGL_SYNC_METAL_SHARED_EVENT_SIGNALED_ANGLE, which when passed as the value for EGL_SYNC_CONDITION during eglCreateSync, changes the behavior to not insert a fence command into the command stream. Test: angle_end2end_tests --gtest_filter=EGLSyncTestMetalSharedEvent.AngleMetalSharedEventSync_WaitSync_ExternallySignaled Bug: angleproject:8064 Change-Id: Ia1b8615b976f293d411b7d2be506b0ac87d64dee Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4307152 Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Kenneth Russell <kbr@chromium.org> Reviewed-by: Sunny Sachanandani <sunnyps@chromium.org>
Igor Nazarov 244e1931 2023-01-17T19:22:10 Vulkan: Fix use of pending Outside RenderPass CommandBuffer. Regression from this commit: 730c127102b540ce2c4ec086b037c8b732706e26 "Vulkan: Submit queue more often for texture data" Bug: angleproject:6354 Test: angle_end2end_tests --gtest_filter=*SubmittingOutsideCommandBufferAssertIsOpen* Change-Id: I5f72f499cd7153c94c8e5f8a3415df2182726c8e Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4296802 Commit-Queue: Igor Nazarov <i.nazarov@samsung.com> Reviewed-by: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Amirali Abdolrashidi 156efe9e 2023-02-23T14:07:23 Fix the alloc error with enabled backtrace feature When angle_enable_unwind_backtrace_support is enabled, an assertion occurs on memory allocation with the following message: "constructed value does not match the lookup key" The assertion comes from PolicyTraits::apply() in the hash map. * In MemoryTracking.h, the type of mMemoryAllocationRecord is now std::unordered_map instead of angle::HashMap. Bug: b/262029018 Change-Id: I11a6d8e99a129759c046e37cef9e74f7db193066 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4289947 Reviewed-by: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Amirali Abdolrashidi <abdolrashidi@google.com>
Amirali Abdolrashidi b0d99f72 2023-02-22T13:48:16 Move the memory tracking classes to new files * Moved the classes, functions, and constants related to memory tracking to MemoryTracking.h and MemoryTracking.cpp. Main classes include the following: * MemoryAllocationTracker * MemoryReport * MemoryAllocationType * MemoryAllocationInfo * MemoryLogSeverity * New static function added in RendererVk to get the Vulkan object type name (GetVulkanObjectTypeName()). Bug: b/262029018 Change-Id: I619001e3c24114c4fe7bf024498338bce146fced Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4284639 Reviewed-by: Charlie Lao <cclao@google.com> Commit-Queue: Amirali Abdolrashidi <abdolrashidi@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Saifuddin Hitawala 877cd04c 2023-03-02T10:24:14 Reland "Add vulkan format image fallback for R16G16B16" Suppressed misbehaving trace test in https://crrev.com/c/4317088 This is a reland of commit 71f6d54c8d3662693283281651e57c994a10cf08 Original change's description: > Add vulkan format image fallback for R16G16B16 > > This change adds a vulkan format image mapping fallback from R16G16B16 > to R16G16B16A16, R32G32B32_FLOAT and R32G32B32A32_FLOAT for both > UNORM and SNORM variants. > > This is done because in Chrome we want to use R16/RG16 formats which > are exposed to Skia over the EXT_texture_norm16. Currently, > EXT_texture_norm16 requires RGB16_EXT which if not present is not > supported even if R16_EXT and RG16_EXT are supported. This fallback > helps us support R16/RG16 as well over RGBA16. > > It also updates validationES checks for GL_RGBA signed and unsigned > normalized checking if type is GL_SHORT or GL_UNSIGNED_SHORT. > > It adds a method LoadToFloat that allows a type (GLushort or GLshort) > to be loaded into float format types. This is then used as part of > fallbacks for load_functions_data.json. > > Bug: None > Change-Id: I5c6879cd2ed5dd6e3440877f4891f269d96d88a1 > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4294694 > Reviewed-by: Vasiliy Telezhnikov <vasilyt@chromium.org> > Reviewed-by: Geoff Lang <geofflang@chromium.org> > Commit-Queue: Geoff Lang <geofflang@chromium.org> Bug: None Change-Id: I101dd9e8c62e6794692c9f89c4944297e195f710 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4317089 Reviewed-by: Cody Northrop <cnorthrop@google.com> Auto-Submit: Roman Lavrov <romanl@google.com> Commit-Queue: Cody Northrop <cnorthrop@google.com>
Alexey Knyazev 1174582a 2023-03-06T00:00:00 GL: Implement EXT_clip_control The extension is trivially exposed if the current context supports it. * Added packed clip control enums * Removed unused state query code * Aligned symbol names with the specs Bug: angleproject:8066 Change-Id: I9d106f39800658ecc75f4525ee93cb534dc49f9e Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4306770 Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Igor Nazarov ad7949c6 2023-01-17T19:24:59 Vulkan: Remove "rx::vk::ImageHelper" move constructor. Move constructor was only used in the "rx::SwapchainImage" structure. This CL replaces "image" member with "std::unique_ptr<vk::ImageHelper>". This will remove source of bugs. Bug: angleproject:8052 Change-Id: Ic16f674095233baaa56fbe8a8fb7ef3e323a7331 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4294905 Reviewed-by: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Igor Nazarov <i.nazarov@samsung.com>
Igor Nazarov 267c556e 2023-03-06T18:03:36 Vulkan: Move PersistentCommandPool out of the CommandsState. Refactoring before the actual fix. Bug: angleproject:8039 Change-Id: Ib240cb9a2e9b92eab120c946b15ac285b939c7db Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4307875 Commit-Queue: Igor Nazarov <i.nazarov@samsung.com> Reviewed-by: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Alexey Knyazev 8abf7153 2023-03-01T00:00:00 GL: Complete EXT_blend_func_extended * Increased test threshold to 2 to avoid failures on some GPUs * Skipped desktop name bindings when running on OpenGL ES * Adjusted extension exposure conditions to avoid failures on older platforms * Adjusted webgl_FragData array size Bug: angleproject:1085 Bug: angleproject:2833 Change-Id: Ic72ba42b024de276d3586446e03013e6063d15ba Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4307122 Reviewed-by: Kenneth Russell <kbr@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com>
Igor Nazarov ac8513fa 2023-03-03T16:10:32 Vulkan: Add and use ScopedQueueSerialIndex helper class. This is a follow up for CL: eb0475c05425d94a4516b708c1c9075e3bc423a7 Vulkan: Cleanup RendererVk::allocateQueueSerialIndex method. Made private "allocateQueueSerialIndex(SerialIndex *)" method and renamed to "allocateQueueSerialIndexImpl" to limit incorrect API usage. Bug: b/267806287 Change-Id: I4d1a9875d454c29f930a5e844161d2a10dc2675b Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4306720 Commit-Queue: Igor Nazarov <i.nazarov@samsung.com> Reviewed-by: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Alexey Knyazev 4256c022 2023-03-01T00:00:00 Metal: Implement ANGLE_stencil_texturing Used texture views to sample stencil values from combined depth/stencil textures. Texture2DDepthStencilTestES3.TexSampleModes* tests were split into swizzled and non-swizzled variants to suppress only swizzled cases on some platforms. Added a new avoidStencilTextureSwizzle workaround to skip creating swizzled texture views of stencil-only textures on platforms that fail to sample from them. Fixed: angleproject:8051 Change-Id: I0b1148f8d30fc6459239efcdaeee6c0364633cc8 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4304058 Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com> Reviewed-by: Kenneth Russell <kbr@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Roman Lavrov b3178411 2023-03-07T15:08:46 Revert "Add vulkan format image fallback for R16G16B16" This reverts commit 71f6d54c8d3662693283281651e57c994a10cf08. Reason for revert: crashing win-trace https://ci.chromium.org/ui/p/angle/builders/ci/win-trace/3527/overview Also reproducible on Linux: % src/tests/capture_replay_tests.py --gtest_filter="Texture2DNorm16TestES3.TextureNorm16R16 Original change's description: > Add vulkan format image fallback for R16G16B16 > > This change adds a vulkan format image mapping fallback from R16G16B16 > to R16G16B16A16, R32G32B32_FLOAT and R32G32B32A32_FLOAT for both > UNORM and SNORM variants. > > This is done because in Chrome we want to use R16/RG16 formats which > are exposed to Skia over the EXT_texture_norm16. Currently, > EXT_texture_norm16 requires RGB16_EXT which if not present is not > supported even if R16_EXT and RG16_EXT are supported. This fallback > helps us support R16/RG16 as well over RGBA16. > > It also updates validationES checks for GL_RGBA signed and unsigned > normalized checking if type is GL_SHORT or GL_UNSIGNED_SHORT. > > It adds a method LoadToFloat that allows a type (GLushort or GLshort) > to be loaded into float format types. This is then used as part of > fallbacks for load_functions_data.json. > > Bug: None > Change-Id: I5c6879cd2ed5dd6e3440877f4891f269d96d88a1 > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4294694 > Reviewed-by: Vasiliy Telezhnikov <vasilyt@chromium.org> > Reviewed-by: Geoff Lang <geofflang@chromium.org> > Commit-Queue: Geoff Lang <geofflang@chromium.org> Bug: None Change-Id: I8373aee3a1d2c2279a3882ff7203d88483e29f4e No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4316422 Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com> Commit-Queue: Roman Lavrov <romanl@google.com>
Shahbaz Youssefi cee1237c 2023-03-06T15:39:03 Vulkan: Fix reserved UBOs for default uniforms Only 3 were reserved, which is not correct with tessellation support included. The limit is now adjusted based on whether geometry and/or tessellation is supported. Bug: angleproject:6858 Change-Id: I7530a60b4b6bc9d4f4561303615c52c63bab1045 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4312559 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Amirali Abdolrashidi <abdolrashidi@google.com> Reviewed-by: Yuxin Hu <yuxinhu@google.com>
Shahbaz Youssefi 868b63ab 2023-03-06T15:21:25 Vulkan: Remove reserved UBO for driver uniforms from limits ANGLE no longer uses UBO for driver uniforms, but uses push constants instead. Bug: angleproject:6858 Change-Id: I3e01c75a19bb333428f2dd1cd732022f1e740c4c Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4306885 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Yuxin Hu <yuxinhu@google.com> Reviewed-by: Amirali Abdolrashidi <abdolrashidi@google.com>
Charlie Lao 991fca06 2023-03-06T10:31:48 Vulkan: Minor clean up in CommandProcessor.cpp Minor clean up per the feedback from already merged CLs. Bug: b/255411748 Change-Id: I3d0148700cf79fa597e260e7192d419198f9749f Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4311756 Commit-Queue: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Roman Lavrov 5331491b 2023-03-06T17:26:30 SYNC-HAZARD-READ-AFTER-WRITE: VkNonDispatchableHandle on x86 Also link to the new bug. Bug: angleproject:8054 Change-Id: I2421d625f64a2530768ea35497a45091a0317025 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4312523 Auto-Submit: Roman Lavrov <romanl@google.com> Reviewed-by: Amirali Abdolrashidi <abdolrashidi@google.com> Commit-Queue: Amirali Abdolrashidi <abdolrashidi@google.com> Commit-Queue: Roman Lavrov <romanl@google.com>
Charlie Lao 5b63e1dc 2023-02-03T15:14:00 Vulkan: SurfaceVk should only wait for GPU work that uses it Right now when we destroy swapchain, we call mRenderer->finish() to finish everything, even though the work is unrelated to this surface. This CL changes it to only wait for ResourceUse of all images in the swapChain. Bug: b/267806287 Change-Id: I33d136ad50961fbf5fbb200ff0f89f1dbf23585d Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4220723 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Charlie Lao <cclao@google.com>
Roman Lavrov 569a881f 2023-03-06T16:43:18 Make SYNC-HAZARD-READ-AFTER-WRITE case less specific. SYNC_VERTEX_SHADER_SHADER_STORAGE_READ added in https://crrev.com/c/4306772 Different prior_usage in TransformFeedbackTest Bug: angleproject:8054 Change-Id: Ia79fc0edd52dff49255906158a5621d4860d7df5 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4313319 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Roman Lavrov <romanl@google.com>
Roman Lavrov 8ba78abd 2023-02-28T11:16:08 Reland "MSRTSS uses AppendToPNextChain due to non-NULL pNext." Also fix msrtss->pNext being set to a pointer to a pointer (void* yay!) This is a reland of commit 33df630f9c8944061902b2e38fe65b280f731802 Original change's description: > MSRTSS uses AppendToPNextChain due to non-NULL pNext. > > As discussed in: > https://chromium-review.googlesource.com/c/angle/angle/+/4116675/comment/3097cb31_16922d39/ > > Not currently causing issues as it requires very recent drivers but I > saw the `ASSERT(ptr->pNext == nullptr);` in AddToPNextChain fail > somewhere. > > Bug: angleproject:7899 > Change-Id: Id46162a5aacd3d8599382ce1dfca25aca5e730e1 > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4296801 > Auto-Submit: Roman Lavrov <romanl@google.com> > Commit-Queue: Yuxin Hu <yuxinhu@google.com> > Reviewed-by: Yuxin Hu <yuxinhu@google.com> Bug: angleproject:7899 Change-Id: I597b6633123f161e05e1a5a28b2e8c6c61835e29 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4306827 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Igor Nazarov dbece66f 2023-01-17T19:25:24 Vulkan: Fix move constructor/assignment of Resource classes. Bug in "DynamicallyGrowingPool<Pool>::PoolResource" causes real problems with queries. "QueryPool" may be reused without waiting for the previous use. Bugs "QueryHelper" may affect "mInFlightGpuEventQueries" (not used in "QueryVk"). Updated "FramebufferHelper" move assignment so it uses "Resource" assignment instead of protected member access. Bug: angleproject:8053 Change-Id: I441b62102fcf232456027fb42eefa97ed8958676 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4300050 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Charlie Lao <cclao@google.com> Commit-Queue: Igor Nazarov <i.nazarov@samsung.com>
Charlie Lao 7eb6869a 2022-08-30T16:28:08 Vulkan: Change ResourceAccess::Write to ResourceAccess::ReadWrite AS a preparation for the next CL which will optimize for WriteOnly access, this CL changes Write to ReadWrite and adds WriteOnly access (but not used yet). Mechanical changes only and no function difference is expected. Bug: b/243711628 Change-Id: I509d6045ae87635e24076b646af42f35d88d52cf Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3866672 Commit-Queue: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Yuxin Hu <yuxinhu@google.com>
Igor Nazarov 4f87f4e9 2023-03-03T13:19:09 Vulkan: Add useResetCommandBufferBitForSecondaryPools feature. Currently ANGLE does not use "vkResetCommandBuffer()" on Vulkan Secondary Command Buffers. Instead it uses "vkFreeCommandBuffers()" and "vkAllocateCommandBuffers". According to spec, "VK_COMMAND_POOL_CREATE_RESET_COMMAND_BUFFER_BIT" is required only for command buffer reset operations. However, some ARM drivers may not free memory in "vkFreeCommandBuffers()" without this flag. Bug: angleproject:8059 Change-Id: Ibfe45bca345dc48484b625c450369d30805cec77 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4306722 Reviewed-by: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Igor Nazarov <i.nazarov@samsung.com>
Igor Nazarov 0eea2893 2023-03-03T18:55:21 Vulkan: Use *_POOL_CREATE_TRANSIENT_BIT in OneOffCommandPool Bug: angleproject:8061 Change-Id: Ib1f851e83e681ee369cac8da0ff52cd3951c5749 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4306724 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Igor Nazarov <i.nazarov@samsung.com> Reviewed-by: Charlie Lao <cclao@google.com>
Roman Lavrov 390fa116 2023-03-06T10:51:24 Suppress another SYNC-HAZARD-READ-AFTER-WRITE case. Starts at: https://github.com/KhronosGroup/Vulkan-ValidationLayers/commit/dd401219bdb982046bb235160c29643c597d6f16 --gtest_filter='MemoryBarrierBufferTest.TransformFeedbackBitWriteThenCapture/ES3_1_Vulkan_SwiftShader*' Bug: angleproject:8054 Change-Id: Ie57665a757c6664a29d351d43fbad9e4743b57f9 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4306772 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Saifuddin Hitawala 71f6d54c 2023-03-02T10:24:14 Add vulkan format image fallback for R16G16B16 This change adds a vulkan format image mapping fallback from R16G16B16 to R16G16B16A16, R32G32B32_FLOAT and R32G32B32A32_FLOAT for both UNORM and SNORM variants. This is done because in Chrome we want to use R16/RG16 formats which are exposed to Skia over the EXT_texture_norm16. Currently, EXT_texture_norm16 requires RGB16_EXT which if not present is not supported even if R16_EXT and RG16_EXT are supported. This fallback helps us support R16/RG16 as well over RGBA16. It also updates validationES checks for GL_RGBA signed and unsigned normalized checking if type is GL_SHORT or GL_UNSIGNED_SHORT. It adds a method LoadToFloat that allows a type (GLushort or GLshort) to be loaded into float format types. This is then used as part of fallbacks for load_functions_data.json. Bug: None Change-Id: I5c6879cd2ed5dd6e3440877f4891f269d96d88a1 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4294694 Reviewed-by: Vasiliy Telezhnikov <vasilyt@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Geoff Lang <geofflang@chromium.org>
kmx.io     mail     discord kmxgit v0.4.0