kc3-lang/angle/src/libANGLE

src/libANGLE

Log

Author Commit Date CI Message
Jamie Madill 9d668d6c 2022-01-07T18:18:23 Revert "Frontend: separate lock in swap prep" This reverts commit 40c5cb255c0a07bdab574aa076ee603e7d791ab3. Reason for revert: Regression in TSAN caused by this CL: https://ci.chromium.org/ui/p/angle/builders/ci/linux-tsan-test/352/overview Original change's description: > Frontend: separate lock in swap prep > > Swapchain-based backends like Vulkan might block a lot in > vkAcquireNextImageKHR, which is bad for overall fast progress if we also > hold the global EGL lock there. > > This CL starts to split the global EGL lock. We release the EGL lock > when performing vkAcquireNextImageKHR, and only maintain a lock for > surfaces. > > Bug: angleproject:6851 > Change-Id: I329d5c4c579718a4980c4261590f77099ce1400e > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3361249 > Reviewed-by: Charlie Lao <cclao@google.com> > Reviewed-by: Geoff Lang <geofflang@chromium.org> > Commit-Queue: Lingfeng Yang <lfy@google.com> Bug: angleproject:6851 Change-Id: Ie03b784021f7b8b5c1ef95a911ef7da4029abd46 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3373165 Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Tim Van Patten 462cdb08 2022-01-06T18:46:21 Capture/Replay: Reset FBOs, RBOs, VAOs Add resetting framebuffers, renderbuffers, and vertex arrays to MaybeResetResources(). This is required to prevent leaks in app launch traces, such as Angry Birds 2 (com.rovio.baba). Bug: angleproject:4599 Change-Id: I34b45ecb1a1b5144f1e23b6bc56315f7713fed94 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3371881 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Cody Northrop <cnorthrop@google.com> Commit-Queue: Tim Van Patten <timvp@google.com>
Lingfeng Yang 40c5cb25 2022-01-06T13:31:54 Frontend: separate lock in swap prep Swapchain-based backends like Vulkan might block a lot in vkAcquireNextImageKHR, which is bad for overall fast progress if we also hold the global EGL lock there. This CL starts to split the global EGL lock. We release the EGL lock when performing vkAcquireNextImageKHR, and only maintain a lock for surfaces. Bug: angleproject:6851 Change-Id: I329d5c4c579718a4980c4261590f77099ce1400e Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3361249 Reviewed-by: Charlie Lao <cclao@google.com> Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Lingfeng Yang <lfy@google.com>
Alexey Knyazev 1e2856fd 2022-01-07T13:04:39 Update freetype-py wheel Bug: chromium:1282168 Change-Id: I7daac60c540d6f67620fba2823ac0d4d36845a13 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3372783 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Charlie Lao caf6433a 2022-01-04T14:37:29 Vulkan: Rename releaseStagingBuffer to releaseStagedUpdate There is no more per ImageHelper dynamic buffer for staging, and this function is not just removing staging buffer, but removing all staged updates. Rename it to reflect what it actually does. Bug: b/208323792 Change-Id: I5a2182291848a9e137ed29b888f31215664b46fd Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3367078 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Charlie Lao <cclao@google.com>
Charlie Lao 93159c02 2021-12-30T16:01:52 Vulkan: Remove DynamicShadowBuffer implementation This is no longer used since mShadowBuffer was removed from BufferVk in https://chromium-review.googlesource.com/c/angle/angle/+/3313333 Bug: b/208323792 Change-Id: I0bb1d66c2f0f58bac74139a274c957681edf2ad4 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3362504 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Charlie Lao <cclao@google.com>
Charlie Lao 3263eb01 2021-12-28T10:10:50 Vulkan: Switch GPU translated vertex buffer to NonHostVisible When we translate using GPU, there is no need to use host visible memory. Use device local memory instead. Bug: b/205337962 Change-Id: Ic76dcb28bde2f079f6ac406d846518bf5f369d74 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3340553 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Charlie Lao <cclao@google.com>
Charlie Lao 8589c456 2021-12-22T17:13:33 Vulkan: Remove mStagingBuffer from ImageHelper and ContextVk This CL removes ImageHelper::mStagingBuffer and ContextVk::mStagingBuffer and uses per shared group staging buffer pool instead. Bug: b/208323792 Change-Id: I46b5fd542697eb860def9c6d592dd147ad4dc973 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3354090 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Charlie Lao <cclao@google.com>
Charlie Lao 6da1ac81 2021-12-22T10:13:43 Vulkan: Switch ReadPixels from DynamicBuffer to shared pool FramebufferVk::readPixelsImpl() and ImageHelper::copyImageDataToBuffer() use per FramebufferVk DynamicBuffer. This CL removes this and uses shared buffer pool to allocate a temporary staging buffer for readPixels as needed and frees it immediately afterwards. Bug: b/208323792 Change-Id: I65ddf9bf9f1f14578d9def63f5287cb1a4121dff Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3354038 Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Charlie Lao <cclao@google.com>
Charlie Lao ad27d5d6 2021-12-21T11:22:30 Reland "Vulkan: Consolidate all vertex conversion buffers to shared pool" This is a reland of cca412cd8b349b7281727c50f2a59d115fd90a05 Further inspection shows it was red-herring. The original CL does not have the un-intended diff that I saw in the commit email. This is try to reland the original CL without any modification. Original change's description: > Vulkan: Consolidate all vertex conversion buffers to shared pool > > There are various conversion buffers that holds converted vertex or > element or index data. They are DynamicBuffer for now. This CL switches > them to use the shared group buffer pool. With this change, all > allocation is represented by a BufferHelper object instead of an offset. > I am able to remove the offset arguments from a lot of APIs. > > Bug: b/208323792 > Change-Id: Ib611beb0c16cddbdd9ddf7b8961c439da9fa5180 > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3352489 > Reviewed-by: Tim Van Patten <timvp@google.com> > Reviewed-by: Jamie Madill <jmadill@chromium.org> > Commit-Queue: Charlie Lao <cclao@google.com> Bug: b/208323792 Change-Id: I90852ad38c2b9ac423800bb6854757bcc17cd166 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3370602 Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Charlie Lao <cclao@google.com>
Tim Van Patten a61a7f59 2021-12-28T18:16:40 Vulkan: Use correct aspectMask Update ImageHelper::stageSubresourceUpdateFromImage() to use the correct aspectMask flags based on the image's format, rather than assuming only color formats. Additionally, only call ensureMutable() for sRGB textures that indicate srgbDecode == GL_SKIP_DECODE_EXT. Bug: angleproject:6852 Test: DepthStencilFormatsTest.DepthTextureRender Change-Id: Iaf1f167612144f836c5574b0f561cc1b58c3cb6a Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3360090 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Cody Northrop <cnorthrop@google.com> Commit-Queue: Tim Van Patten <timvp@google.com>
Charlie Lao 4e85bdd9 2022-01-06T17:06:25 Revert "Vulkan: Consolidate all vertex conversion buffers to shared pool" This reverts commit cca412cd8b349b7281727c50f2a59d115fd90a05. Reason for revert: There is accidental code merge bug left in. Original change's description: > Vulkan: Consolidate all vertex conversion buffers to shared pool > > There are various conversion buffers that holds converted vertex or > element or index data. They are DynamicBuffer for now. This CL switches > them to use the shared group buffer pool. With this change, all > allocation is represented by a BufferHelper object instead of an offset. > I am able to remove the offset arguments from a lot of APIs. > > Bug: b/208323792 > Change-Id: Ib611beb0c16cddbdd9ddf7b8961c439da9fa5180 > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3352489 > Reviewed-by: Tim Van Patten <timvp@google.com> > Reviewed-by: Jamie Madill <jmadill@chromium.org> > Commit-Queue: Charlie Lao <cclao@google.com> Bug: b/208323792 Change-Id: I18bba207d1d8bb76dff32d9855a744dba93bc6d6 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3370601 Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com> Commit-Queue: Charlie Lao <cclao@google.com>
Charlie Lao cca412cd 2021-12-21T11:22:30 Vulkan: Consolidate all vertex conversion buffers to shared pool There are various conversion buffers that holds converted vertex or element or index data. They are DynamicBuffer for now. This CL switches them to use the shared group buffer pool. With this change, all allocation is represented by a BufferHelper object instead of an offset. I am able to remove the offset arguments from a lot of APIs. Bug: b/208323792 Change-Id: Ib611beb0c16cddbdd9ddf7b8961c439da9fa5180 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3352489 Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Charlie Lao <cclao@google.com>
Tim Van Patten fd014271 2021-12-16T20:44:42 Vulkan: Match chromaFilter to min/mag for YUV AHBs From the VVL: VUID-VkSamplerCreateInfo-minFilter VkCreateSampler: VK_FORMAT_FEATURE_SAMPLED_IMAGE_YCBCR_CONVERSION_SEPARATE_RECONSTRUCTION_FILTER_BIT specifies that the format can have different chroma, min, and mag filters. However, VK_FORMAT_FEATURE_SAMPLED_IMAGE_YCBCR_CONVERSION_SEPARATE_RECONSTRUCTION_FILTER_BIT is not supported for VkSamplerYcbcrConversionCreateInfo.format = VK_FORMAT_UNDEFINED so chromaFilter needs to be equal to minFilter/magFilter. We don't know what the min/mag filters are when the AHB is imported, so we need to update the YcbcrConversionDesc with the current min/mag filter value before creating the VkSamplerYcbcrConversion when the ImageView is created. Additionally, if the min/mag filters are updated later, TextureVk::syncState() needs to recreate the ImageViews to ensure the chromaFilter matches the new min/magFilters. Test: atest android.media.cts.DecodeAccuracyTest#testGLViewDecodeAccuracy[0] Test: ImageTestES3.SourceYUVAHBTargetExternalYUVSampleLinearFiltering Bug: b/210526871 Change-Id: I95dbd9738f6e3fd0870e484518eee105e995f93a Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3346394 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Tim Van Patten <timvp@google.com>
Kyle Piddington abf17f01 2021-12-10T17:10:53 Metal: Support depth filtering queries for D32, D24S8 Port changes from https://bugs.webkit.org/show_bug.cgi?id=232071 Bug: angleproject:6779 Change-Id: Id2d7a0c215e4db072b8830aedd90b878f3817967 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3331666 Reviewed-by: Gregg Tavares <gman@chromium.org> Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Kyle Piddington <kpiddington@apple.com>
Kyle Piddington f13f440e 2021-12-15T15:11:05 Reland: Metal: Fix Webkit warnings in ANGLE build Fix nullable warnings, unused functions Add in additional function specification to FormatStringIntoVector to correctly fix warning Bug: angleproject:6781 Change-Id: I26af2d698f14d353832802aa7b5ce34c5a1f4b95 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3366796 Reviewed-by: Kenneth Russell <kbr@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Kyle Piddington <kpiddington@apple.com>
Tim Van Patten 7126c3a3 2021-12-23T16:58:25 Capture/Replay: Limit caps for QC/SwS On Pixel 4 (Qualcomm), GL_MAX_ATOMIC_COUNTER_BUFFER_BINDINGS = 6, so limit mState.mCaps.maxAtomicCounterBufferBindings to 6 to make traces more portable. For SwiftShader, MAX_SHADER_STORAGE_BUFFER_BINDINGS = 12, so limit mState.mCaps.maxShaderStorageBufferBindings to 12 to make traces more portable. These limits affect collecting traces for Angry Birds 2 (com.rovio.baba). Bug: angleproject:6848 Test: TracePerfTest.Run/*_angry_birds_2_launch Change-Id: Ie8364c8b04d1b82655c4a95b3983f618f478e23c Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3355555 Reviewed-by: Cody Northrop <cnorthrop@google.com> Commit-Queue: Tim Van Patten <timvp@google.com>
Charlie Lao 0e49a3dd 2022-01-04T11:23:54 Vulkan: Add std::move support for BufferHelper There are needs to support std::move for BufferHelpers in other CLs (See crrev.com/c/3352489). Without this support, we can not store BufferHelper into std::vector. This CL adds move support for BufferHelper class. Bug: b/208323792 Change-Id: I93f79490715750abc1bcedd41b683ad0c2460ebb Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3366855 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Charlie Lao <cclao@google.com>
Charlie Lao 43b0e92b 2021-12-13T15:38:32 Vulkan: Consolidate mHostVisibleBufferPool and mMapInvalidate BufferVk::mHostVisibleBufferPool is allocated when BufferVk::mBuffer is not hostvisible and we need to map it. In that case mHostVisibleBufferPool is allocated and data copied from mBuffer to it and the pointer to mHostVisibleBufferPool is returned to user. BufferVk::mMapInvalidateRangeStagingBuffer is used when map is called on a small range. In this case we allocate memory for the small range of buffer and return that intead of waiting for entire buffer for GPU to finish. Also when BufferSubData is called, we also needs to allocate a staging buffer and issue a copyBuffer from staging buffer to main buffer. This CL consolidate all these three usage cases into one mStagingBuffer. It removes mHostVisibleBufferPool and mMapInvalidateRangeStagingBuffer from BufferVk class. This makes overall logic of managing data consistency much simpler as well since we only have two buffers: The main buffer storage mBuffer or mStagingBuffer. And mIsStagingBufferMapped tracks if mStagingBuffer is the one actually mapped to user or not so that at unmap time we know if we should flush the data to mBuffer or not. Bug: b/208323792 Change-Id: I4f0c79a2d86da1a43844ed2ba83ddeb7dd4a5c0b Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3335216 Reviewed-by: Lingfeng Yang <lfy@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Charlie Lao <cclao@google.com>
Jamie Madill cabc3b98 2021-12-30T14:33:31 Move atomic counter uniform range counter to executable. This makes the code more consistent with the other ranges. It will make way for future refactorings to moving more uniform code to the executable. Bug: angleproject:3570 Change-Id: Ie365c0e184039f813f19d3f38b401ad8a0b72120 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3359001 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Peng Huang e9e9a2c6 2022-01-04T20:14:48 Do not pass mInstance to vkGetInstanceProcAddr for vkEnumerateInstanceVersion Bug: angleproject:6869 Change-Id: Ifc93a067824c3467003a47bad240b3be10874ab0 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3366038 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Peng Huang <penghuang@chromium.org>
Maksim Sisov beda2459 2021-12-09T14:24:08 Add support for GL_MESA_framebuffer_flip_y 2/* This is a second CL that adds actual implementation for GL_MESA_framebuffer_flip_y extension. Also, some tests are added to verify the functionality. Please note that bots do not support this extension yet, and the tests were verified by running them locally. Bug: chromium:1231934 Change-Id: Iea483aa13a298df6b5cf0b7b5ffb795a4666e3bd Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3329603 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Jonah Ryan-Davis <jonahr@google.com> Commit-Queue: Maksim Sisov <msisov@igalia.com>
Lubosz Sarnecki c984dad5 2021-12-03T16:02:49 FrameCapture: Don't remove protection from shared pages. Don't remove protection from pages that are shared when a persistent coherent buffer storage is unmapped. The shared page can still be relevant when the the other buffer is kept and written to. Add a PageSharingBuffers test that produces this behaviour. Test: angle_end2end_tests --gtest_filter="BufferStorageTestES3.PageSharingBuffers/*" Bug: angleproject:5857 Change-Id: I6927f25229d2dfe9f68ba9a993e9d3e994bc7ce0 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3306623 Reviewed-by: Cody Northrop <cnorthrop@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Lubosz Sarnecki <lubosz.sarnecki@collabora.com>
Kyle Piddington cfc4db34 2021-11-04T16:54:15 Metal: Provoking vertex support for Metal DrawArrays Add in provoking vertex rewrite support for drawArrays command, enabling 6 dEQP tests. GLES3/functional_rasterization_flatshading_line_loop GLES3/functional_rasterization_flatshading_line_strip GLES3/functional_rasterization_flatshading_lines GLES3/functional_rasterization_flatshading_triangle_fan GLES3/functional_rasterization_flatshading_triangle_strip GLES3/functional_rasterization_flatshading_triangles This patch adds a new shader to the Provoking Vertex helper that generates draw commands for all simple data types. Fix bug in provoking vertex helper that caused reused index buffers to suffer from allocation issues. Also fix Provoking vertex for triangle fan generation Bug: angleproject:5325 Change-Id: I7a1211dfcd99329868269ea0666eef1915d487b5 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3261635 Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Kyle Piddington <kpiddington@apple.com>
Jamie Madill 0331738b 2021-12-29T14:47:32 Fix extension download script, and update data. This includes the update to the NV 1660s, and Intel driver update, and automatically gathered SwiftShader data. Bug: angleproject:6853 Change-Id: I481918105fd0e0224a2e2ddf5175b38e97918195 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3360784 Reviewed-by: Cody Northrop <cnorthrop@google.com> Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Shahbaz Youssefi a400e76d 2021-12-22T12:24:13 Vulkan: Don't expose MSRTT without VK_KHR_depth_stencil_resolve Previously, ANGLE exposed the EXT_multisampled_render_to_texture extension even if VK_KHR_depth_stencil_resolve was not available. This was due to scarcity of drivers on the bots with support for this extension and allowed partial testing. This is no longer true. Additionally, this cleans up the expectations because old and buggy drivers are now simultaneously filtered out. Bug: angleproject:6845 Change-Id: I8c285a5fa5d9beeb98b48d8b056cdc8779cce6fd Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3353895 Reviewed-by: Yuly Novikov <ynovikov@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Lingfeng Yang 6f1c39b5 2021-12-16T16:10:49 Vulkan: Restore color attachment mask-related blend states When we end up resetting blend state such as when changing to a depth-only FBO and back, we can end up adding attachment states. They need to be updated with the current GL blend state. Bug: b/210543392 Bug: angleproject:6828 Change-Id: I3419acb8081eb7d662576a75382bb8684cc0d29f Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3346320 Reviewed-by: Cody Northrop <cnorthrop@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Lingfeng Yang <lfy@google.com>
Gert Wollny 0a28c6c6 2021-12-15T15:49:33 Capture/Replay: Capture some MultiDrawIndirect parameters It seems some new tests were added that fail Capture/Replay. Bug: angleproject:6180 Change-Id: I26ccc1e9defc7b7d66c7e26538df286230b47def Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3340945 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Cody Northrop <cnorthrop@google.com> Commit-Queue: Gert Wollny <gert.wollny@collabora.com>
Gert Wollny 5f53268c 2021-11-04T17:56:48 Capture/Replay: Gather Program sources directly from call With CreateShaderProgramv the sources were collected from the attached shaders, wich gave complications, because normally the the shaders were detached right after program linking and a different handling was needed when capture was enabled, handling that let the replay tests fail. To overcome this gather the shader source directly from the call parameters instead of pulling it from the attached shaders. Bug: angleproject:6180 Bug: angleproject:6659 Change-Id: I248c1aa0f58b61519b938fce1a77fb25dc0a7cfc Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3262133 Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Cody Northrop <cnorthrop@google.com> Commit-Queue: Gert Wollny <gert.wollny@collabora.com>
Tim Van Patten 542f7d80 2021-12-22T19:36:27 Vulkan: Copy buffer data to correct cube map face TextureVk::copyBufferDataToImage() was failing to set region.imageSubresource.baseArrayLayer to the correct cube map face value, causing every glTexSubImage2D()-type call using GL_PIXEL_UNPACK_BUFFER to only copy to the GL_TEXTURE_CUBE_MAP_POSITIVE_X face. In the case of PUBG Mobile, this left all other faces with "random" data, resulting in corrupted reflections on things like water, the airplane, etc. Bug: b/194140619 Test: TextureCubeTestES3.CubeMapPixelUnpackBuffer Change-Id: I93a9219ce098985f9f4b906f8fe2e05850fb8d43 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3354091 Reviewed-by: Ian Elliott <ianelliott@google.com> Reviewed-by: Cody Northrop <cnorthrop@google.com> Commit-Queue: Tim Van Patten <timvp@google.com>
Cody Northrop 2ebd583e 2021-12-22T11:24:43 Android: Remove setenv from common path In Android production stress testing, the setenv call was causing a race condition. To fix, only use setenv in the paths that need it as a performance optimization. Specifically, FrameCapture makes use of it every frame. Bug: angleproject:6822 Change-Id: I452f1ef8607ca9baf15d69246932c47454a3ce46 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3353893 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Ian Elliott <ianelliott@google.com> Commit-Queue: Cody Northrop <cnorthrop@google.com>
Charlie Lao 2f3e6cc0 2021-12-13T17:40:18 Vulkan: Remove mShadowBuffer from BufferVk class. The shadow buffer was initially designed to avoid synchronization in glMapBuffer call while buffer itself is still busy. There are many optimization done inside BufferVk::mapImpl that try to avoid wait for GPU as much as we can by distinguish GPU write versus read, by detecting map call read/write intention by checking access bit, and finally by allocating a staging buffer to return a CPU friendly copy of data to caller. This shadow buffer implementation also have known bugs that are not keeping data in sync. With all these optimization added after initial mShadowBuffer implementation, I believe we do not have a good reason to still keep mShadowBuffer. And this has been disabled for months in main branch. This CL removes this code path completely which makes code a lot simpler. Bug: b/208323792 Change-Id: Ie5999e38b6120a371ec2e969f196e4754ebd0f8d Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3313333 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: mohan maiya <m.maiya@samsung.com> Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Charlie Lao <cclao@google.com>
Charlie Lao bc9d2d7d 2021-12-07T16:16:47 Vulkan: Consolidate BufferHelper's code paths into one There are cases that BufferHelper may not sub-allocate from the pool. For example, when buffer is created from external memory. Right now we have two different code paths in the BufferHelper object and checking if suballocation object is valid or not and pick different code path. This CL consolidates both code path into one by making non sub-allocated case also creates BufferSUbAllocation object and owns a BufferBlock object. Bug: b/205337962 Change-Id: Iac82bdb0b69d424e4147b52d458ced6274e106a1 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3322100 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Charlie Lao <cclao@google.com>
Shahbaz Youssefi 4114fe23 2021-11-20T22:05:56 Vulkan: Simplify transform feedback emulation logic In [1], the program is transformed to not output transform feedback logic when transform feedback is inactive. With that change, it's no longer necessary to dedicate a driver uniform to indicate if transform feedback is active. [1]: https://chromium-review.googlesource.com/c/angle/angle/+/3294661 Bug: chromium:1209285 Change-Id: Ica725d0da08e2676f442fe6307962d367b443535 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3294702 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Jamie Madill ab288fea 2021-07-09T13:02:08 Upgrade restricted traces. (Re-land) Includes removing capture to temporary ResourceMaps, and including all source files in the JSON. Bug: angleproject:5133 Change-Id: I0508dd46a9f613a96ca3635f82b6904c0f2523a7 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3345222 Reviewed-by: Cody Northrop <cnorthrop@google.com> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Jamie Madill c1f478fe 2021-12-20T12:37:33 Frame Capture: Allow for no FBO when running MEC. Multi-context captures use the surfaceless context extension. That was crashing MEC because we were assuming the context would have a Framebuffer. Bug: angleproject:5133 Change-Id: I6be12d332f4754f80c088d88973219327eca251a Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3352084 Reviewed-by: Cody Northrop <cnorthrop@google.com> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Jamie Madill eba2e620 2021-12-15T12:29:12 Capture/Replay: Fix uniform program indexing. We were using the "native" version of the program ID in the update call, which could overflow the bounds of our fixed-size program maps. Instead use the captured (and upper-bounded) version of the ID and only get the mapped ID when calling GetUniformLocation. Bug: angleproject:6431 Bug: angleproject:6816 Change-Id: I807bd07c9dcff6ba5a50991e28b0a58d16f7bc63 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3345220 Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Cody Northrop <cnorthrop@google.com> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Jamie Madill 1da7e3af 2021-12-21T15:55:38 Revert "Metal: Fix Webkit warnings in ANGLE build" This reverts commit 6706799186683dd3733c4610de09d84721aa08c8. Reason for revert: Includes pragma that disables a warning. Original change's description: > Metal: Fix Webkit warnings in ANGLE build > > Fix nullable warnings, unused functions > > Bug: angleproject:6781 > Change-Id: I063331e60d31a55b3cc9df0b41ace014d7d13659 > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3343174 > Reviewed-by: Kenneth Russell <kbr@chromium.org> > Commit-Queue: Kenneth Russell <kbr@chromium.org> Bug: angleproject:6781 Change-Id: Id5fce2afd0381030a974871b99e8adf075677bd1 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3352086 Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Charlie Lao eca427b2 2021-12-20T10:55:41 Vulkan: Add mutex to protect mVirtualBlock mVirtualBlock could be accessed from multiple thread when asyncCommandQueue is enabled, where the free call could come from the submission thread. This CL adds a mVirtualBlockMutex in BufferBlock and always take the lock when mVirtualBlock is been accessed. This CL also adds ConditionalMutex class for the general usage that a mutex is only used based on a boolean. Bug: angleproject:6840 Change-Id: Ib647b4ff12ebfc08f2a70192d39f16e15a1fc5f5 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3350798 Reviewed-by: Ian Elliott <ianelliott@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Charlie Lao <cclao@google.com>
Shahbaz Youssefi c1598640 2021-11-30T23:44:30 Vulkan: Improve mid-RP clear warning message One app does: - Draw - Draw - Clear depth, then disable it - Draw In this case, the message generated for mid-RP clear, suggesting the clear be done earlier is not meaningful as the previous draw calls did use the depth/stencil image. The message now includes an alternative suggestion to invalidate the depth/stencil image instead of clearing it. Note that the app may still legitimately do multiple passes in one render pass where depth/stencil is cleared in between, so the warning is not applicable in all cases. It's still useful to notice issues in more common scenarios. Bug: angleproject:2472 Change-Id: I3abbecf8c83b7b856c2430675e69b1471e91c0c4 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3308920 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Ian Elliott <ianelliott@google.com>
Shahbaz Youssefi cf8ac00a 2021-12-20T11:57:25 Vulkan: Fix MSRTT w.r.t stencil-only unresolve The stencil bit was being silently masked out in a bitset in FramebufferDesc that tracked whether the framebuffer requires any unresolve. If only stencil needs unresolving, this mask was zero, leading to an incorrect framebuffer getting pulled from the cache. A follow up change will add an ASSERT in BitSetT to catch such errors in the future. This issue was only reproducible on SwiftShader and AMD as the only implementers of VK_EXT_shader_stencil_export. Bug: angleproject:6840 Bug: angleproject:6324 Change-Id: I4f055982ebd75f621ec1e34b0d60eaa497c27b17 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3349979 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Kimmo Kinnunen aabfe874 2021-12-17T15:43:22 Metal: Use the context device to Texture::resize Makes the resized texture to be attributed to the owner. Bug: angleproject:6832 Change-Id: Ia1eae1bb6c9e83e23fae51121f88d9ce37dc948b Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3347642 Reviewed-by: John Cunningham <johncunningham@apple.com> Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Shahbaz Youssefi 34ea8111 2021-10-28T11:07:53 Reland "Vulkan: Don't expose MSRTT on SwiftShader" This is a reland of fb77afd3c76301e11f5c7397d9a63d283f86036c Original change's description: > Vulkan: Don't expose MSRTT on SwiftShader > > Emulating this extension on SwiftShader is neither useful nor efficient > (on the contrary, it hurts performance (which is true for all non-tiling > architectures)). This extension was exposed purely to increase testing > coverage, but is being disabled now as OOMs resulting from allocating > the implict multisampled image is confusing Chrome. Note that > SwiftShader exposes only a limited amount of memory. > > Bug: chromium:1263046 > Change-Id: Ia69b7fc63804169431885352a8f637a95ab08bd0 > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3251581 > Reviewed-by: Geoff Lang <geofflang@chromium.org> > Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Bug: chromium:1263046 Change-Id: I88415f3c1a7532eda5473da6acbd22acf14ae11c Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3260330 Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Kimmo Kinnunen 624e3c3d 2021-12-17T15:30:54 Fix dangerous use of AutoObjCPtr after ownership identity change Metal new* methods create objects with +1 retain count. AutoObjCPtr is intended to adopt this reference. Otherwise, like before this patch, the AutoObjCPtr holds object with +2 count. Before this patch, some but not all call sites donated the extra retain count to autoreleasepool "out of band", and as such the code did not leak that much. Bug: angleproject:6831 Change-Id: I72bcbc712f2cadbcbc6148c6aedfa7e151314518 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3347641 Reviewed-by: John Cunningham <johncunningham@apple.com> Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Alexey Knyazev 4a148e9d 2021-12-20T13:39:48 Revert "Vulkan: Suppress VUID-vkCmdDraw-blendEnable" This reverts commit 797ebdde9894b4411ed304a83fdaeb8d8a91fe56. Reason for revert: The VVL bug has been fixed. Original change's description: > Vulkan: Suppress VUID-vkCmdDraw-blendEnable > > A new VVL causes failures in BlendIntegerTest.MRT* > > Bug: angleproject:6355 > Change-Id: If62b1a7934574acaeb95601d7fb0a61dc5d9c99a > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3140159 > Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> > Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Bug: angleproject:6355 Change-Id: I769d34fada1c2db00513f9912ef178dcd1eb4ee9 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3347648 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Kimmo Kinnunen ede0b8b7 2021-12-17T14:35:47 Avoid ANGLE_MTL_AUTORELEASE inside ASSERT, it leaks on Release Avoid ANGLE_MTL_AUTORELEASE inside ASSERT, it leaks on Release. The create functions already return retained, instead of autoreleasepool retained, so owning the reference via adopting AutoObjCObj is as efficient as autoreleasepool on possible future ARC. Thus use AutoObjCObj, that is safer than autoreleasepool with C++ and works towards being able to perhaps remove use of autoreleasepool. Bug: angleproject:6830 Change-Id: I80f523d9a6846097a593fe94b5ea9715fa88ea15 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3347640 Reviewed-by: Kenneth Russell <kbr@chromium.org> Reviewed-by: John Cunningham <johncunningham@apple.com> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Kenneth Russell 001221f0 2021-12-16T21:41:29 Metal: upstream fix from WebKit for texture leak. Upstream fix by Kimmo Kinnunen for a leak of MTLTexture instances for render targets from: REGRESSION (r287068): [macOS arm64] TestWebKitAPI.GraphicsContextGLCocoaTest. UnrecycledDisplayBuffersNoLeaks is failing https://bugs.webkit.org/show_bug.cgi?id=234356 The leak would happen through a temporary render pass descriptor that would ref the render target texture. The descriptor was created without an autoreleasepool in call stack. Also fixes the error string leaks. Bug: angleproject:6827 Bug: angleproject:6829 Change-Id: I07ec961886a9fb8caad6df334293358a175eaf15 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3346682 Reviewed-by: Kyle Piddington <kpiddington@apple.com> Commit-Queue: Kenneth Russell <kbr@chromium.org>
Charlie Lao 6c894e82 2021-11-04T14:49:41 Vulkan: Replace BufferVk::getBufferAndOffset() with getBuffer() Now BufferHelper class already keeps offset information. There is no reason for BufferVk to have that information any more. Bug: b/205337962 Change-Id: I6e014fb480bfcd5018ef9231b0fb87a50021f179 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3266147 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Charlie Lao <cclao@google.com>
Lingfeng Yang a93a3930 2021-12-08T20:36:07 Vulkan: Track enabled in VertexArray Previously in sequences such as glEnableVertexAttribArray(0) glDrawElements() glDisableVertexAttribArray(0) glEnableVertexAttribArray(0) glDrawElements() the second glDrawElements() would always be called with dirty enable. On some backends like Vulkan this could result in unnecessary pipeline changes. Bug: angleproject:6796 Change-Id: Ib44877b6030e8d0acede7cf5c591936ba3844b69 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3325208 Reviewed-by: Charlie Lao <cclao@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Lingfeng Yang <lfy@google.com>
Charlie Lao 83a670ab 2021-10-29T09:12:26 Vulkan: Implement BufferPool using VMA's virtual allocator VMA's allocation calls used to be sub-allocating a pool of memory. What we really want is sub-allocate a VkBuffer object. VMA recently added support to expose the underlying range allocation algorithm via APIs, which user can use it to sub-allocate any object. This CL uses that new virtual allocation API to sub-allocate from a pool of VkBuffers. In this CL we only switched BufferVk::mBuffer to sub-allocate from the BufferPool object. Bug: b/205337962 Change-Id: Ia6ef00c22e58687e375b31bc12ac515fd89f3488 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3266146 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Charlie Lao <cclao@google.com>
Lingfeng Yang 83679716 2021-12-14T19:02:48 Vulkan: Dirty vertex buffers if default attrib values changed In the call_break_offline trace, attribute index 4 is disabled then enabled before a draw. This redundant operation is currently not blocked and makes it to the Vulkan backend, resulting in handleDirtyGraphicsVertexBuffers. However, at the same time, we change the default attrib value for index 3. Which makes handleDirtyGraphicsVertexBuffers not redundant. We'll need to take into account cases when other attribute indices are not dirty and yet the default attrib value changes. Bug: angleproject:6802 Change-Id: I4a3ec411aa42f73a209fd0f06976d7ef2d21e027 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3331412 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Charlie Lao <cclao@google.com> Commit-Queue: Lingfeng Yang <lfy@google.com>
Charlie Lao 543f5750 2021-12-15T16:58:46 Vulkan: Disable protected memory for ARM ARM driver reports protected memory feature support. But when we enable the feature, even if the feature is not been used (i.e., no memory or queue is created with protected bit), we still see quite large performance drop, mainly due to excessive load/store unit activity. This CL disables this feature on ARM platform until we can find a good solution for it. Bug: angleproject:6817 Change-Id: I57e8b3f568f3818080e56b5d6ae9d1623732d236 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3343671 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Charlie Lao <cclao@google.com>
Maksim Sisov c7446ac4 2021-12-16T16:00:08 Add support for GL_MESA_framebuffer_flip_y 1/* LaCrOS, which is same as ChromeOS, relies on this extension and uses that with WebGL content. However, when LaCrOS uses ANGLE, the extension is not exposed and the WebGL content is y-inverted as direct scanout is always used. There are two bugs - 1) ANGLE should expose the extension so that direct scanout is possible (this CL fixes the issue. 2) LaCrOS shouldn't rely on the extension if it's not available (a bug to be fixed in Chromium). This particular CL consists of autogen files with framebufferParameteriMESA && getFramebufferParameterivMESA in Context marked as UNIMPLEMENTED. And also unimplemented validation. Bug: chromium:1231934 Change-Id: Id617846b0412f579df5b4ace6cb33c1ec6d403df Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3343492 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Maksim Sisov <msisov@igalia.com>
Peng Huang 6e5fd816 2021-12-15T18:31:13 Support more ImageLayout in ConvertImageLayoutToGLImageLayout Bug: chromium:1264439 Change-Id: I35938a65981edf495bd9aafb8bfb17aa3ec9a66c Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3341640 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Peng Huang <penghuang@chromium.org>
Gregg Tavares b26bb13a 2021-12-15T15:46:00 Metal: Remove references to xfbActiveUnpaused uniform Metal chooses a shader with or without support for transformfeedback so there's no point in checking in the shader if it's enabled or not. Bug: angleproject:6823 Change-Id: Iacd26b1e55b311a7cc7987dcd8b2f8c0c8727d62 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3343179 Reviewed-by: Kenneth Russell <kbr@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Gregg Tavares <gman@chromium.org>
Alexey Knyazev c4d434b9 2021-12-15T22:57:13 Metal: Enable OES_fbo_render_mipmap Bug: angleproject:6813 Change-Id: I643c114d7b37ca0532d2dda1964090a15e470842 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3342568 Reviewed-by: Kyle Piddington <kpiddington@apple.com> Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Alexey Knyazev <lexa.knyazev@gmail.com>
Kyle Piddington 67067991 2021-12-15T15:11:05 Metal: Fix Webkit warnings in ANGLE build Fix nullable warnings, unused functions Bug: angleproject:6781 Change-Id: I063331e60d31a55b3cc9df0b41ace014d7d13659 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3343174 Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Kenneth Russell <kbr@chromium.org>
Gregg Tavares bda49325 2021-12-10T16:11:15 Add guards for 64 bit types on 32bit systems Bug: angleproject:6785 Change-Id: Iff043a22d398f5a7a1a46ae3cc4d2de5a6e3797e Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3331639 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Gregg Tavares <gman@chromium.org>
Amirali Abdolrashidi 89e11878 2021-12-06T09:42:46 Vulkan: Fix the UAF issue with BufferData * Fixed the use-after-free issue with stale buffer handles after calling BeginTransformFeedback. * Added an observer for TransformFeedbackVk to update the buffer handles when buffer's storage is changed and the buffer update type is StorageRedefined. * Added a function to TransformFeedbackVk::onDestroy() to release the counter buffers in order to avoid crash due to TransformFeedbackVk::end() not being called, e.g., as a result of no glEndTransformFeedback() calls. Bug: chromium:1274316 Change-Id: I8ed477f36e6ff89dd4764bb59af564c69efe33e2 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3321789 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Amirali Abdolrashidi <abdolrashidi@google.com>
Gregg Tavares ee1d3cf7 2021-12-10T12:29:16 Metal: Integrate Metal Binary Shader generation Bug: angleproject:6801 Change-Id: I2eb7072ed342f2b4cb5d9ee50ca9db75c5c7df69 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3331206 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Kyle Piddington <kpiddington@apple.com> Commit-Queue: Gregg Tavares <gman@chromium.org>
Kyle Piddington 3e7e2758 2021-12-14T15:48:42 Metal: Fix macOS ANGLE build on Catalina (10.15) SDK macOS Catalina does not have an enum for AppleGPUFamily6 In addition, platform includes for macOS 11.0 should use 110000, not 101600 (10.16) Bug: angleproject:6814 Change-Id: I9c6a0587aa22fb2e58d02fcb1779060de2632963 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3340104 Reviewed-by: Alexey Knyazev <lexa.knyazev@gmail.com> Commit-Queue: Kyle Piddington <kpiddington@apple.com> Reviewed-by: Kenneth Russell <kbr@chromium.org>
Amirali Abdolrashidi 05a44dd2 2021-11-29T12:16:58 Vulkan: Set content undefined on eglSwapBuffers * Invalidate the swap buffer content if the swap behavior is EGL_BUFFER_DESTROYED. * In that event, WindowSurfaceVk::getBufferAge() sets `age` to 0 and returns with no errors. * Unit tests in EGLBufferAgeTest.cpp updated to reflect this. * Minor cleanup in the test file. Bug: angleproject:6363 Change-Id: Ibab7ccb99c1be149588e6aad16bcc2d6dd989295 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3307776 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Amirali Abdolrashidi <abdolrashidi@google.com>
Shahbaz Youssefi 5275f487 2021-12-14T21:03:54 Vulkan: Disable async command queue with Vulkan SCBs There is currently a threading hazard with ContextVk::mCommandPool, where ContextVk records to command buffers allocated from the pool while the async command queue thread is using other command buffers allocated from the same pool. While experimenting with Vulkan SCBs, disable the async command queue feature when using Vulkan SCBs. It is unknown whether async command queue with Vulkan SCBs is even worthwhile; if it proves to be, we could make ContextVk maintain multiple pools and rotate between them. Bug: angleproject:6811 Change-Id: Iac2e9337b5b109b2c7cb359109afe6ee386e4c34 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3340018 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Mohan Maiya feb599ad 2021-12-09T20:17:14 Invalidate EGL handles during eglTerminate EGL 1.5 spec says - Termination marks all EGL-specific resources, such as contexts and surfaces, associated with the specified display for deletion. Handles to all such resources are invalid as soon as eglTerminate returns Move EGL object handles to another set, tracking invalid objects, during display terminate. Destroy these invalid objects during eglReleaseThread Bug: angleproject:6798 Test: EGLMultiContextTest.NegativeTestAfterEglTerminate* Test: EGLMultiContextTest.RepeatedEglInitAndTerminate* Test: Android CTS WrapperTest.testThreadCleanup Change-Id: Ie160212c98367493e645d9d1c8260e7a30649386 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3329273 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: mohan maiya <m.maiya@samsung.com>
Shahbaz Youssefi 90c3038e 2021-12-14T16:46:13 Vulkan: Fix desc set updates with Vulkan SCBs Since Vulkan SCBs are a command buffer in their own, we cannot aggregate desc set updates to be done once after all the binds have been recorded. This change makes sure that desc set updates are issued before bind, but this problem could potentially also be solved with VK_DESCRIPTOR_BINDING_UPDATE_AFTER_BIND_BIT, by actually issuing the writes at the end of the secondary CB, just before closing it. Bug: angleproject:6811 Change-Id: I18e150b73a5881258679ba53a2d84acd2fc44edd Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3338390 Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Yuxin Hu 800c6971 2021-12-08T01:08:17 Modified buffer fallback format to smaller storage format Bug: angleproject:6726 Change-Id: I475d99fc0143f265e66cc4f7aa13fa0de8745169 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3315916 Reviewed-by: Lingfeng Yang <lfy@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Yuxin Hu <yuxinhu@google.com>
Mohan Maiya e70a81b9 2021-12-14T13:41:28 Vulkan: Rectify typo introduced in 36fcf80b1f2a99 The method should have been updateYcbcrConversionDesc Bug: angleproject:6732 Test: Texture2DTestES3.*Yuv*Vulkan Change-Id: I67ab35ac07a6d751cbe3c3c29bdde21a5a7f5834 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3339605 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: mohan maiya <m.maiya@samsung.com>
Kyle Piddington 76d42082 2021-12-14T09:30:24 Metal: ETC and ASTC are missing on M1 Platform guards are incorrect in mtl_utils.mm. We need to account for macOS post 10.15, which now supports the apple GPU family query. Bug: angleproject:6809 Change-Id: Iae219a11ffe0ef0715681009aa24cbbc6fd2c61e Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3338624 Reviewed-by: Gregg Tavares <gman@chromium.org> Reviewed-by: Alexey Knyazev <lexa.knyazev@gmail.com> Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Kenneth Russell <kbr@chromium.org>
Mohan Maiya 36fcf80b 2021-12-11T20:55:53 Vulkan: Consolidate SamplerYcbcrConversionCache Now that SamplerDesc and ImageHelper have the fully qualified description of a VkSamplerYcbcrConversion, we can lazy allocate the conversion handle along with streamlining the interface to the SamplerYcbcrConversionCache class. Bug: angleproject:6732 Test: Texture2DTestES3.*Yuv*Vulkan Change-Id: I7deaa0220031bc4c500d88021459017bf937218c Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3333629 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: mohan maiya <m.maiya@samsung.com>
Hailin Zhang 7b13a9ac 2021-12-09T18:37:59 Vulkan: Fix dynamic partial update buffer data issue. add test case for dynamic update buffer data. Signed-off-by: Hailin Zhang<hailinzhang@google.com> Bug: b/207714894 Change-Id: I8c1e93d152847c3162c0e2dd49abe3d899c859a0 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3328869 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Charlie Lao <cclao@google.com> Commit-Queue: Hailin Zhang <hailinzhang@google.com>
Peng Huang 32f572b7 2021-12-03T17:58:51 Support creating EGLImage from VkImage Bug: chromium:1264439 Change-Id: I520182143e748f25b44d0725f3f171b7b33a85d8 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3311131 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Peng Huang <penghuang@chromium.org>
Jamie Madill 7fc6c07c 2021-09-29T14:02:29 Capture/Replay: Update process for trace upgrading. Includes changes to the retracing script. Also includes documentation on how the process works. Bug: angleproject:5133 Change-Id: I1acfe338f3fe0282a0461c314274c761ed04bd2f Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3193418 Reviewed-by: Cody Northrop <cnorthrop@google.com> Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Lubosz Sarnecki d4b8cd5e 2021-09-14T16:34:31 FrameCapture: Implement capture of coherent bufferstorages. Track buffers with the GL_MAP_COHERENT_BIT_EXT access flag in a CoherentBufferTracker on a per page basis using the PageFaultHandler and protection functions from system_utils. Check for dirty memory and capture it on draw calls. Bug: angleproject:5857 Change-Id: Ib098f96952db7583d3e6517570c179e7754f54b2 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3168629 Reviewed-by: Cody Northrop <cnorthrop@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Lubosz Sarnecki <lubosz.sarnecki@collabora.com>
Amirali Abdolrashidi 5b858acf 2021-12-09T14:22:14 Vulkan: Fix the offset issue in multiDrawIndirect * In MultiDrawElementsIndirect, it resets the index buffer offset (mCurrentIndexBufferOffset) to 0. * Added a unit test for MultiDrawElementsIndirect to make sure that the index buffer offset is reset properly. * Updated the test program in MultiDrawIndirectTest to include a multi-colored mode. * D3D11 and Metal tests regarding the following functions are disabled so more test coverage can be added: * glMultiDrawElementsIndirectEXT * glMultiDrawElementsInstancedBaseVertexBaseInstanceANGLE Bug: angleproject:6751 Change-Id: I81bb7b3aa9b6ad8a9283dd50728ec6250c3668ad Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3330737 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Amirali Abdolrashidi <abdolrashidi@google.com>
John Cunningham 42bd4fc2 2021-12-10T14:58:17 Metal: add ANGLE_metal_create_context_ownership_identity. This Metal-specific EGL extension allows a given context and the GPU resources it allocates to be associated with a particular task ID on the system, for system-level bookkeeping purposes. Bug: angleproject:6795 Change-Id: I19ee0993564169b01c4a450e63dcfacd339b98b5 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3335172 Reviewed-by: Jonah Ryan-Davis <jonahr@google.com> Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Kenneth Russell <kbr@chromium.org>
Jeff Vigil 1fc9b7c6 2021-12-08T14:10:36 EGL: Add queries for EGL_KHR_protected_content Add surface and context query for EGL_PROTECTED_CONTENT_EXT attribute. Bug: angleproject:6693 Change-Id: Iad3d8d7e0db89414c44a0101c26cda6e60308e71 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3324963 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Brandon Schade <b.schade@samsung.com>
Gregg Tavares afca80a8 2021-12-09T17:48:49 Metal:Add debug info for shaders This should fix the WEBGL_debug_shader_info CTS test Bug: angleproject:6788 Change-Id: I89e18f94ea3199149c252355372648b5e328e1ed Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3329204 Reviewed-by: Yuly Novikov <ynovikov@chromium.org> Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Gregg Tavares <gman@chromium.org>
Jamie Madill f60b152c 2021-12-06T10:19:53 Frame Capture: Merge small frames into single sources. By buffering up the replay functions before writing them out into sources, we can reduce the number of replay files by a large margin. Bug: angleproject:5133 Change-Id: I0a556fd1a3a5f2dcc8a26b761e885c9b7e464cf5 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3316200 Reviewed-by: Cody Northrop <cnorthrop@google.com> Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Shahbaz Youssefi b7a42c88 2021-12-13T15:53:07 Vulkan: Remove VK_EXT_swapchain_colorspace check We don't make use of this extension since https://chromium-review.googlesource.com/c/angle/angle/+/3208689. Bug: angleproject:2514 Change-Id: I8b8edf5006ac1c3eeeefcadbe77249a5a6a5dda7 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3335070 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Jamie Madill f40e83cb 2021-09-10T19:15:17 Reduced "rx::vk::kInFlightCommandsLimit" from "100" to "50". The "100" value is too high and causes "S21 Adreno" block in "vkQueueSubmit()". This is a temp solution. Patch authored by Igor Nazarov <i.nazarov@samsung.com> Bug: angleproject:6401 Change-Id: Ie7f9c8ef5311b17325640cae0f613ece768c0fa3 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3165666 Reviewed-by: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Lingfeng Yang ea03bf47 2021-12-03T08:48:50 Vulkan: fix crash when clearing stencil Previously, we cast the originally-32-bit mask to 8 bits, and then crashed ourselves in the assert. Also optimize away the stencil based on 8 bit explicitly in the frontend. Bug: chromium:1275858 Change-Id: I0b03a17e72f5a4540b5c4605353eb1fde898057d Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3315158 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Charlie Lao <cclao@google.com> Commit-Queue: Lingfeng Yang <lfy@google.com>
Jamie Madill e8b186c9 2021-12-09T09:45:20 Frame Capture: Don't unmap regenned buffers in reset. These buffers are deleted and recreated, so unmapping them gives a "resource not mapped" error during replay. Found when working on the "world_war_doh" trace. Bug: angleproject:5133 Change-Id: Id1e6318c44a2a09fa643ef0a72f1d03d8396e5fb Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3327862 Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Cody Northrop <cnorthrop@google.com> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Jamie Madill a24919e7 2021-12-09T10:04:25 Frame Capture: Clear array buffer binding in MEC. This should allow us to properly reset buffer state when capturing vertex attribute pointer state in the per-context (not shared) setup function. Also adds a missing replay state buffer binding update. Fixes a state serialization difference in "badland". Bug: angleproject:5133 Change-Id: I8509dbf7211726ff47f4e7e5ff86a0eb8726d75c Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3327861 Reviewed-by: Cody Northrop <cnorthrop@google.com> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Gregg Tavares a1b9eb68 2021-12-10T16:36:46 Replace IOSurfaceSurfaceEAGL.mm table with less strict padding Bug: angleproject:6783 Change-Id: Id67f581af0742b47fc5ea38de616391e57b763de Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3331641 Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Gregg Tavares <gman@chromium.org>
Kyle Piddington 194687e7 2021-12-06T13:40:24 Metal: Fix M1 and iOS synchronization for readPixels ReadPixels does not implicitly flush context before reading on iOS and macOS hardware due to shared memory. look for any pending GPU Work, and flush the command buffer if needed. Bug: angleproject:6803 Change-Id: I5688b031ecffcd7427263041d4e33ae256f68af6 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3318592 Reviewed-by: Kenneth Russell <kbr@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Kyle Piddington <kpiddington@apple.com>
Gregg Tavares 2d6caae4 2021-12-10T17:25:58 Metal: Add maximum render target size validation Bug: angleproject:6790 Change-Id: I4a7e7285e040cc68e4fc16bb653f6d1512f541c9 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3331671 Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Gregg Tavares <gman@chromium.org>
Gregg Tavares cf9ab647 2021-12-10T15:42:36 Add early out validation in drawArraysCommon Bug: angleproject:6791 Change-Id: I60de6806dae9c917c6db24a18f586e3b0aa2154e Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3331414 Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Gregg Tavares <gman@chromium.org>
Gregg Tavares d97b1c4a 2021-12-10T16:27:08 Metal: Refactor restartRange cache Bug: angleproject:6784 Change-Id: I8a2e1df2ac621415d75708c0bba16f8ca17b0ce6 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3331675 Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Gregg Tavares <gman@chromium.org>
Gregg Tavares 1642b4a3 2021-12-10T16:01:10 Metal: Fix Instanced Draw feature support on Simulator Bug: angleproject:6786 Change-Id: I4c567bf2ec61df581ae0e8eeb1c36c7d37e15d04 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3331637 Reviewed-by: Kenneth Russell <kbr@chromium.org> Reviewed-by: Kyle Piddington <kpiddington@apple.com> Commit-Queue: Gregg Tavares <gman@chromium.org>
Gregg Tavares 2e669566 2021-12-09T18:21:02 Metal: Refactors from Webkit Mostly renaming, typos, cleanup bug: angleproject:6787 Change-Id: Ic71c296d7849244cae2ca29b264a10cb96bca612 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3329261 Reviewed-by: Jonah Ryan-Davis <jonahr@google.com> Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Gregg Tavares <gman@chromium.org>
Gregg Tavares 5068c040 2021-12-09T17:55:01 Metal: Merge changes to mtlPixelFormatTable Bug: angleproject:6789 Change-Id: I9893093fdff1138fc8ae35ac4363201fa1e0b825 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3329206 Reviewed-by: Jonah Ryan-Davis <jonahr@google.com> Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Gregg Tavares <gman@chromium.org>
Amirali Abdolrashidi 4d0de3d2 2021-12-03T17:09:06 Update the arg types for glMultiDraw*IndirectEXT() * Updated the auto-generated types for arguments "mode" and "type" for glMultiDraw*Indirect and glMultiDrawIndirectEXT functions to PrimitiveMode and DrawElementsType, similar to the other glMultiDraw* functions. Bug: angleproject:6770 Change-Id: If02a66b6fb2f81848ab6e43a6cd406d5956bf331 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3321792 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Amirali Abdolrashidi <abdolrashidi@google.com>
Peng Huang 933f066e 2021-12-10T15:04:53 Enable GL_ANGLE_texture_usage extension This extension doesn't change ANGLE behavior with Vulkan backend, since it is just a hint for ANGLE to optimize texture allocation. Bug: angleproject:6771,chromium:1264439 Change-Id: Ie2411894a02bbcb0a5304c635f1f5648520ad22a Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3329694 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Peng Huang <penghuang@chromium.org>
Peng Huang cc516f97 2021-12-10T02:48:54 Use STL container for vulkan memory allocator There is a crash related to VMA containers, so switch to STL containers to see if it can workaround the issue. Bug: chromium:1266412 Change-Id: I8416e4b4a27588ac2d710a4ef7600cd91b080535 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3327744 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Peng Huang <penghuang@chromium.org>
Hailin Zhang 3be551d7 2021-12-08T16:44:56 fix directUpdate buffer pointer issue. inside mapWithOffset already add the mBufferOffset Signed-off-by: Hailin Zhang<hailinzhang@google.com> Bug: b/207714894 Change-Id: Ia400bccbef1abc756cd8155e93a775338a30e8b9 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3325025 Reviewed-by: Lingfeng Yang <lfy@google.com> Reviewed-by: Charlie Lao <cclao@google.com> Commit-Queue: Hailin Zhang <hailinzhang@google.com>
Gregg Tavares 04110e4e 2021-12-01T14:06:03 Metal: Fix Intel backend fails with tall texture Copied test from the WebGL conformance suite. A 1x<MAX_TEXTURE_SIZE> texture fails on Intel in the Metal backend. This was fixed in WebKit so just applying that fix. Bug: angleproject:6754 Change-Id: Id1cafb3b560f14630684d55b0e61cea4c7f33a6d Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3311941 Reviewed-by: Kenneth Russell <kbr@chromium.org> Reviewed-by: Jonah Ryan-Davis <jonahr@google.com> Commit-Queue: Gregg Tavares <gman@chromium.org>
Charlie Lao 494a0ec2 2021-12-08T16:07:50 Vulkan: Append the actual buffer size when binding's size is 0 When app calls glBindBufferBase, the size of binding is set to 0. The spec says the actual size should be size of buffer when buffer is been referenced. But when we generate ShaderBuffersDescriptorDesc, we are directly using binding's size. But the actual descriptor set has the buffer's actual size. This causing the ShaderBuffersDescriptorDesc have a false cache hit (because size is always 0) when the actual descriptor set should be different. This CL checks bindging.getSize() and if it is 0, we use the actual buffer size. Bug: angleproject:6792 Change-Id: I3e044dc5ff113b4fae1b5c7d5e78645c6e98edf8 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3325024 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Charlie Lao <cclao@google.com>
Jamie Madill 9b23be91 2021-12-08T13:55:22 Frame Capture: Fix client buffers in MEC. In some cases we would neglect to reset the currently bound array buffer to "null" before capturing a client array buffer in MEC. Noticed in the "badland" trace context setup. Bug: angleproject:5133 Change-Id: I693edce763b449720a7f2123dcaf460854142e91 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3323683 Reviewed-by: Cody Northrop <cnorthrop@google.com> Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Tim Van Patten 16ce5226 2021-12-01T16:58:49 Vulkan: Re-enable RGB8 for pbuffers. pbuffers support was removed from RGB8 EGL configs to prevent mismatched surface formats for surfaces backed by external memory, since ANGLE overrides RGB8->RGBA8 for all surfaces. However, ANGLE no longer supports surfaces backed by external memory with the Vulkan backend, so we can re-enable RGB8 for pbuffers. Bug: angleproject:6651 Change-Id: I7d8b666ed5e3226f067479758f44d9041dcb67b0 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3312361 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Ian Elliott <ianelliott@google.com> Commit-Queue: Tim Van Patten <timvp@google.com>
Shahbaz Youssefi 55840e90 2021-12-03T15:24:00 Vulkan: Fix deferred flush vs UtilsVk Take the following scenario: 1. Draw 2. Flush (this is deferred) 3. Get image view (this is retain()ed) 4. Pass view to a draw-based UtilsVk function 5. Flush 6. Delete image view At step 4, UtilsVk may start a new render pass and use the image view from step 3. Since the flush at step 2 is deferred, it will be performed at this step, and so the serial of the image view is set to the previous submission. When step 4 uses this view, it doesn't retain it. Step 5 submits the new command buffer using this image view. At step 6, if the previous submission has finished, it will destroy the view immediately even though it's in use by the new submission. One solution could have been to make sure render pass closure originating from UtilsVk doesn't incur a flush. However, due to the current design where the render pass is immediately recorded in RendererVk's primary command buffer, it's possible that an unrelated context would perform the flush anyway. This change makes sure instead that the render pass is closed before any views are allocated/retained to be used by UtilsVk. Bug: chromium:1272266 Change-Id: I5bdefb34e03c368511c4c174cf7965fda158d2b8 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3315976 Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
kmx.io     mail     discord kmxgit v0.4.0