kc3-lang/angle/src

src

Log

Author Commit Date CI Message
Shahbaz Youssefi f99ccb08 2021-01-06T23:29:21 Vulkan: Add a perf warning when dropping superseded updates If an update to an image subresource is entirely superseded by a following update, the former update is dropped. This change adds a perf warning for this. Bug: angleproject:3461 Change-Id: Iaf4984ba52f3dfc86d5cc87f8be82123badebd04 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2613203 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Jonah Ryan-Davis a392a81d 2021-01-06T12:26:44 GLX: Avoid creating child window when X visual ID is specified. ANGLE's GLX backend creates a child ID because the visual of the X window must match that of the GLX window, and we can't always be certain. EGL_ANGLE_x11_visual allows applications to specify the visual ID of the window passed to ANGLE via EGL_X11_VISUAL_ID_ANGLE. When this is the case, we don't need to make a child window. Since Chrome always passes this information, this may help optimize ANGLE's GLX usage in Chrome, because we don't have to poll the parent window to manage the child window's properties. Bug: chromium:1132827 Change-Id: If8082d2d07469905afffab01dde2ec9fca8d4eb9 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2611556 Commit-Queue: Jonah Ryan-Davis <jonahr@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Tim Van Patten 2d1406a8 2020-12-29T18:11:33 Capture/Replay: Capture glBindBufferBase during MEC setup "World War Doh" binds uniform buffers via glBindBufferBase() calls before any frames have rendered and then uses those bound buffers for all subsequent frames. ANGLE's frame capture was failing to perform these calls during the mid-execution capture setup phase, leading to the replay generating VVL errors: VUID-vkCmdDispatch-None-02699 Descriptor in binding #0 index 0 is being used in draw but has never been updated via vkUpdateDescriptorSets() or a similar call. This CL adds those calls to the MEC setup phase, allowing the replay to execute without any errors. Bug: angleproject:5495 Test: angle_perftests --gtest_filter="*world_war_doh*" Change-Id: I751c9c6f60bf78c13428a9d1d6a06dbfe600c24b Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2606806 Reviewed-by: Cody Northrop <cnorthrop@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Tim Van Patten <timvp@google.com>
Shahbaz Youssefi a7158eb5 2021-01-06T21:33:24 Vulkan: Fix image leak in staged updates in ImageHelper When removing superseding updates, the superseded update was not released, causing a memory leak. This change also makes SubresourceUpdate non-copyable and correctly implements the move assignment operator such that swap between different update types are correct. As a result, the destructor can now ASSERT that the image is not leaked. Bug: chromium:1146516 Bug: chromium:1163354 Change-Id: I7531c91d8559c23b2e09159118fe645d12fc601f Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2613201 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Shahbaz Youssefi 157ddfdc 2021-01-06T23:20:36 Build compiler_tests only if GLSL or ESSL build is enabled Most of these tests use TranslatorGLSL or TranslatorESSL (often through ShaderCompileTreeTest). In specialized builds that disable GLSL and ESSL shader generation, disable these unit tests. Bug: chromium:1161513 Change-Id: Ib87e651706f141a41ffdaebfb0cbe5168582e341 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2613202 Reviewed-by: Yuly Novikov <ynovikov@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Peng Huang acdb63b8 2020-12-23T12:57:08 Remove the checking before accessing to function pointers Bug: angleproject:5490 Bug: chromium:1161513 Change-Id: I96c5c463a9fd5a337d09f6cefe7836ff52bea90f Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2601071 Reviewed-by: Jonah Ryan-Davis <jonahr@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Peng Huang <penghuang@chromium.org>
Shahbaz Youssefi e2a8a69a 2021-01-04T23:52:02 Vulkan: Move xfb emulation offset calculation to translator A new function is added by the translator to calculate the transform feedback offsets in the emulation path. This function makes the generated code for transform feedback smaller. Bug: angleproject:3606 Change-Id: I01460f907e20e2887cb720bddad96697fdcb0cf3 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2607492 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Shahbaz Youssefi 07d619cb 2021-01-06T11:36:40 Vulkan: Fix missing xfb buffer rebind on new command buffer When a command buffer is ended, bindings are lost. This adds a dirty bit to `mNewGraphicsCommandBufferDirtyBits` to make sure transform feedback buffers are bound again on the new command buffer. Bug: angleproject:5428 Change-Id: I7733c93b1eb5d33a77cbee231a83199be950e19f Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2611552 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Shahbaz Youssefi 5e280545 2021-01-06T14:13:29 Move RegenerateStructNames AST transformation to gl/ This AST transformation is used by the chromium validating command decoder on Linux/AMD/GL as well as MacOS. Bug: chromium:1161513 Change-Id: I67e40215ae5576819ba5a58987ceb1776c86c731 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2613195 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Yuly Novikov <ynovikov@chromium.org>
Shahbaz Youssefi ca94d1da 2021-01-04T21:51:33 Vulkan: Refactor generated code in xfb emulation path The index expression was regenerated for every component written to the output. This change refactors the generated code such that common values are calculated once. This is in preparation for moving some of this code to the translator and reducing the amount of code generated at link time. Bug: angleproject:3606 Change-Id: Ib180d245218f8f6f18c57c0ea092e8c45cbffd1a Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2607490 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com>
Yuly Novikov 46d3f02a 2021-01-05T17:56:24 Suppress dEQP.KHR_GLES31/core_vertex_attrib_binding_basicinputcase5 on Vulkan Pixel2 and SwiftShader. Ignore device and vendor in test config if we are running on SwiftShader. Bug: angleproject:5500 Change-Id: Ib07240dfe1c2ae1d6d081ffa389d041d97b22b36 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2611309 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Ian Elliott <ianelliott@google.com> Commit-Queue: Yuly Novikov <ynovikov@chromium.org>
James Darpinian ebf00703 2020-12-29T16:25:54 angle_end2end_tests passes on iOS! Miscellaneous test skips and fixes for iOS. Bug: angleproject:5417 Bug: angleproject:5491 Change-Id: Id0785e6243949fc756e4d7923dbbe77a411052f3 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2606656 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: James Darpinian <jdarpinian@chromium.org>
Shahbaz Youssefi 2762641a 2021-01-05T23:00:14 Use is_apple instead of is_mac and is_ios everywhere Bug: chromium:1161513 Change-Id: Id7d582f127f4b79ea6fc5b8ad39fff768c45f477 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2611312 Reviewed-by: James Darpinian <jdarpinian@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Tim Van Patten 56479b06 2020-12-21T13:03:50 Tests: Add FIFA Mobile trace Test: angle_perftests --gtest_filter=TracePerfTest.Run/*fifa_mobile* Bug: angleproject:4048 Change-Id: I5cc731ada3ddd9adcb247d7c6f8cf5d823bdc2a3 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2600098 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Cody Northrop <cnorthrop@google.com> Commit-Queue: Tim Van Patten <timvp@google.com>
Jonah Ryan-Davis 956a358d 2020-12-23T15:43:00 GL: Add function pointers for EXT/NV_framebuffer_blit Allow the GL backend to call BlitFramebuffer through these extensions. Bug: angleproject:5474 Change-Id: Ic19b0535056c3e2b098dedc60e2eb8da5f862433 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2602702 Commit-Queue: Jonah Ryan-Davis <jonahr@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill dc2cf6fc 2021-01-06T11:35:53 Roll third_party/vulkan-deps/ b08eace32..cbab63796 (17 commits; 1 trivial rolls) https://chromium.googlesource.com/vulkan-deps/+log/b08eace32e9c..cbab637962d0 $ git log b08eace32..cbab63796 --date=short --no-merges --format='%ad %ae %s' 2021-01-06 chromium-autoroll Roll SPIRV-Tools from d630e5f8c10c to 37c03859de9e (1 revision) 2021-01-06 chromium-autoroll Roll Vulkan-ValidationLayers from 2468aa9630c2 to 98382948dfea (1 revision) 2021-01-06 chromium-autoroll Roll SPIRV-Tools from 1bb80d2778a3 to d630e5f8c10c (1 revision) 2021-01-06 chromium-autoroll Roll glslang from 9325cc013e3d to f77085c307fc (1 revision) 2021-01-05 chromium-autoroll Roll Vulkan-ValidationLayers from 57255a663ed9 to 2468aa9630c2 (1 revision) 2021-01-05 chromium-autoroll Roll Vulkan-Loader from 1f0c7868a438 to 9ba835885ffe (1 revision) 2021-01-05 chromium-autoroll Roll Vulkan-ValidationLayers from 56a96658b750 to 57255a663ed9 (3 revisions) 2021-01-05 chromium-autoroll Roll Vulkan-Loader from 294480cb7f3e to 1f0c7868a438 (1 revision) 2021-01-05 chromium-autoroll Roll Vulkan-Tools from 7554321b88ae to 4cc0932495ec (2 revisions) 2021-01-05 chromium-autoroll Roll Vulkan-Loader from 66a23ac8c566 to 294480cb7f3e (1 revision) 2021-01-05 chromium-autoroll Roll Vulkan-Headers from 85470b32ad5d to 9efc4a631161 (1 revision) 2021-01-05 chromium-autoroll Roll Vulkan-ValidationLayers from aa076dae88e2 to 56a96658b750 (25 revisions) 2021-01-05 chromium-autoroll Roll glslang from f426f652299d to 9325cc013e3d (1 revision) 2021-01-05 chromium-autoroll Roll SPIRV-Cross from e50f7d1ce8e1 to 7b7a21c4058c (3 revisions) 2021-01-05 chromium-autoroll Roll SPIRV-Tools from 17ffa89097b2 to 1bb80d2778a3 (1 revision) 2020-12-21 jmadill Rolling 8 dependencies Created with: roll-dep third_party/vulkan-deps Bug: angleproject:5390 Change-Id: I6f4a499099211e8990ed038984f41bc5f881c71f Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2601340 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill d9669322 2020-12-30T14:33:02 Remove "init" from VaryingPacking. Instead we can pass the pack mode and size to the collectAndPack method. This cleans up the interface and also allows us to merge two separate code blocks in Program and ProgramPipeline. Bug: angleproject:5496 Change-Id: I390b5d2e8a3b033374ccc5a250597be1f03dec96 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2606531 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com>
Jamie Madill d33ffb22 2020-12-30T11:14:00 Organize VaryingPacking methods. This clarifies the difference between "collect" and "pack" methods. Previously the naming was overlapping between the two. Also makes the "packUserVaryings" method private. Refactoring change only. Will enable further improvements for capturing varyings with multiple shader stages. Bug: angleproject:5496 Change-Id: I854590ceab39f3a0e7a785516e0d1fd44e4ccc98 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2606529 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Jamie Madill 92e7bc89 2020-12-23T15:18:42 Remove ProgramLinkedResources from ProgramExecutable. Instead of storing the entire LinkedResources struct, we can keep it only for the duration of the linking calls. Refactoring change only. It sets the stage for more refactoring. This change also switches the link call to use LinkingState's ProgramLinkedResources directly to avoid the need to copy the varying packing or use a pointer. Bug: angleproject:4514 Bug: angleproject:5496 Change-Id: Iefea3c16a33213dc338cc54efaa7c3064ea6ae08 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2601403 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Tim Van Patten e09e947d 2020-12-29T18:21:36 Capture/Replay: Don't force built-in attribs to have a location CaptureMidExecutionSetup() hits an assert while capturing "World War Doh": ANGLE : FATAL: FrameCapture.cpp:2876 (CaptureMidExecutionSetup): ! Assert failed [[[...]]] attrib.location != -1 This is due to CaptureMidExecutionSetup() enforcing that all attributes, including built-ins, have a location. However, in the case of "World War Doh", the built-in 'gl_GlobalInvocationID' does not have a location. Moving the assert to after the gl::IsBuiltInName() check allows the capture to complete without hitting any asserts. Bug: angleproject:4048 Test: Capture "World War Doh" Change-Id: Ice96dec629b1477b11a710283c4ec4bf1c95c6e2 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2606807 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Cody Northrop <cnorthrop@google.com> Commit-Queue: Tim Van Patten <timvp@google.com>
Shahbaz Youssefi bf6cc88a 2021-01-05T22:17:41 Vulkan: Remove duplicate shaderType parameter to SPIR-V transformer The passed-in GlslangSpirvOptions already has the shaderType, so this extra parameter was not necessary. Bug: angleproject:4524 Change-Id: Ibbc7442b3173b26825f3cb9e2365789fade92cfc Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2611311 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Shahbaz Youssefi 382bf288 2020-12-24T23:56:39 Organize AST transforms per backend Most of the AST transforms are written as a workaround to an issue that affects a single backend. This change identifies such transforms and organizes them by backend. They are then only built if the respective backend is. Additionally, about half of the GL transforms are due to mac workarounds, including the large RewriteRowMajorMatrices transform. Mac-specific workarounds are additionally only built on said platform. This change reduces the ANGLE binary size: - 106KB in a Vulkan-only build on Linux - 27KB in a GL-only build on Android (60KB on Linux) Bug: chromium:1084580 Bug: chromium:1161513 Change-Id: I64b334332c0d4f848756c6538af0d8d96864c7e9 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2601346 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com>
Jamie Madill e91097bf 2020-12-29T14:05:56 Vulkan: Clean up "actual"/"intended" naming. Clarifies that the GL internal format is an "intended" format and the Vulkan formats are "actual" formats. This makes all the format fields use the same consistent naming pattern. Bug: angleproject:5438 Change-Id: I935a49895109e9e06eae5ef98d5614dfd1128ff8 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2605728 Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Shahbaz Youssefi 973bd4ba 2020-12-21T13:45:36 Complete I/O block GLSL tests The tests meant to perform a draw call and validate the correctness of the shaders. Bug: angleproject:3580 Change-Id: I3a008d204ae3d7dceb731ba7a815ab54f38c2920 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2599938 Reviewed-by: Jonah Ryan-Davis <jonahr@google.com> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Shahbaz Youssefi aff17499 2020-12-23T14:40:06 Vulkan: Directly capture non-gl_Postion builtins Building on support for transform feedback capture of I/O block members, this change optimizes capture of builtins other than gl_Position by directly decorating members of gl_PerVertex. Most importantly, this allows us to reserve only one varying for transform feedback, instead of as many builtins there could be (up to 4). Bug: angleproject:3606 Change-Id: Ie0957802c657ed6c5aac538d92b860582ed6da45 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2601072 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
James Darpinian b6ea6edc 2020-12-29T16:30:29 Remove tabs from volk.c/h WebKit's SVN continues to abhor tabs. Bug: angleproject:3439 Change-Id: Ib97fc97c0eb7401f73ea493a7b679f78afc1a8c4 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2606658 Reviewed-by: Kenneth Russell <kbr@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: James Darpinian <jdarpinian@chromium.org>
James Darpinian 6c608e94 2020-12-29T16:24:08 iOS sRGB mipmap generation is broken too Enable the encodeAndDecodeSRGBForGenerateMipmap workaround for iOS. Bug: angleproject:5417 Change-Id: I39d9e7c371f0ee5ec5119669f3b855251691f9e2 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2606657 Reviewed-by: Kenneth Russell <kbr@chromium.org> Reviewed-by: Jonah Ryan-Davis <jonahr@google.com> Commit-Queue: James Darpinian <jdarpinian@chromium.org>
Jonah Ryan-Davis fb35201a 2020-12-22T13:14:41 GL: Expose NV_framebuffer_blit for ES2 contexts when available ANGLE_framebuffer_blit is already exposed but has restrictions on scaling/flipping and mismatched color buffer formats. When NV_framebuffer_blit, EXT_framebuffer_blit, or an ES3 context is available we can expose a less restrictive BlitFramebuffer. Bug: angleproject:5474 Bug: chromium:1157057 Change-Id: I916ee39c8d6120216f91461080eb3ee9ca777e29 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2601165 Commit-Queue: Jonah Ryan-Davis <jonahr@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Qin Jiajia de62b251 2020-12-29T17:07:06 Make sure the queried resources are active In GLES 3.1 spec, sesion 7.3.1 program interfaces, it says that 'Resources referenced in shader code are considered active unless the compiler and linker can conclusively determine that they have no observable effect on the results produced by the executable code of the program.'. Variables might be considered inactive if they are used only in computations of temporary variables having no effect on any shader output. So we should use atomicCounterIncrement instead of atomicCounter in case the resources are marked as inactive on some platforms. Bug: angleproject:5494 Change-Id: Ic04bd4a2936e97776fc680b5ef8273f0944d7480 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2605917 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Jiajia Qin <jiajia.qin@intel.com>
Peng Huang 47ee6a7b 2020-12-23T13:31:23 Add build flags to disable desktop gl backend This change saves about 100KB on Android Bug: angleproject:5490 Change-Id: Ieca0e597423f74a917197f381560af5409a9e874 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2601341 Reviewed-by: Jonah Ryan-Davis <jonahr@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Peng Huang <penghuang@chromium.org>
Jamie Madill 093250e0 2020-12-23T12:07:46 Vulkan: Make shader variable info maps a class. Instead of using a map type directly we abstract the info maps into an encapsulated class. We can enforce a specific API set instead of using the same API as the map class. This also cleans up a few of the APIs related to these maps. This change will allow future changes to the way the variables are stored in the class without drastically changing the interface. Bug: angleproject:3572 Bug: angleproject:4524 Change-Id: Ic1a63e1776c39f49b895a1274bae8282d7a6b9b5 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2600080 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com>
Tom Anderson c2847bbc 2020-12-23T15:59:59 [GLX] Don't request exposure events for child windows After [1], Chrome and ANGLE use separate X11 connections, so in order for Chrome to receive exposure events, it must select for them its own connection, which is done in [2]. The events selected by ANGLE's connection are never dispatched, so they queue up indefinitely and leak until the connection is closed (when happens when Chrome closes). This change removes the request to receive exposure events. [1] https://chromium-review.googlesource.com/c/chromium/src/+/2492177 [2] https://chromium-review.googlesource.com/c/chromium/src/+/2602441 Change-Id: Ie47c3700855aeeecc8efa1b379df60c344b1b70a Bug: chromium:1158170 R=jonahr Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2602705 Reviewed-by: Jonah Ryan-Davis <jonahr@google.com> Commit-Queue: Jonah Ryan-Davis <jonahr@google.com>
Tim Van Patten 220642a9 2020-12-16T21:10:37 Allow Drawing with Immutable Persistent Mapped Buffers From the EXT_buffer_storage overview: The GL_EXT_texture_storage extension added immutable storage for texture objects (and was subsequently incorporated into OpenGL ES 3.0). This extension further applies the concept of immutable storage to buffer objects. [T]his extension introduces the concept of persistent client mappings of buffer objects, which allow clients to retain pointers to a buffer's data store returned as the result of a mapping, and to issue drawing commands while those mappings are in place. The initial implementation of EXT_buffer_storage didn't enable this portion of the extension, so ANGLE is generating errors while attempting to draw with an immutable buffer mapped with the GL_MAP_PERSISTENT_BIT flag. This CL enables that functionality, since apps (e.g., FIFA Soccer) rely on it. Bug: angleproject:5473 Change-Id: Icf1c0597156044a342aac5e4d2abbc29b34f46b2 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2596957 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Cody Northrop <cnorthrop@google.com> Commit-Queue: Tim Van Patten <timvp@google.com>
Shahbaz Youssefi c6e63da8 2020-12-24T21:32:42 Suppress IOBlocksSeparate xfb test on AMD/Linux/GL Bug: angleproject:5487 Bug: angleproject:5493 Change-Id: I6e15c35eacafef0ac4b7c7789619470501e31402 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2601345 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Yuly Novikov <ynovikov@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Yuly Novikov 9f0748ea 2020-12-24T16:05:24 Skip TransformFeedbackTestES31.IOBlocks* on Linux AMD Vulkan Bug: angleproject:5493 Change-Id: I5e0b0ec478807ebfe7a0ec114adaba13ea103ea0 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2601344 Reviewed-by: Yuly Novikov <ynovikov@chromium.org> Commit-Queue: Yuly Novikov <ynovikov@chromium.org>
James Darpinian cde0e3ab 2020-12-22T16:38:47 Don't use optional BGRA_EXT by default It's not currently exposed on iOS. Bug: angleproject:5417 Change-Id: I2564d3363f8f8bddad4da7d8a9d7dac9bea15897 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2600454 Reviewed-by: Kenneth Russell <kbr@chromium.org> Reviewed-by: Jonah Ryan-Davis <jonahr@google.com> Commit-Queue: James Darpinian <jdarpinian@chromium.org>
James Darpinian 939195a0 2020-12-23T15:12:30 Skip tests using unimplemented features on iOS Bug: angleproject:5485 Change-Id: I3248126060b9957596f65221c60f42f213e9a9dc Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2601377 Reviewed-by: Kenneth Russell <kbr@chromium.org> Reviewed-by: Jonah Ryan-Davis <jonahr@google.com> Commit-Queue: James Darpinian <jdarpinian@chromium.org>
James Darpinian 7061af26 2020-12-22T10:43:45 Fix BlitFramebuffer issues with large rects on iOS Enable adjustSrcDstRegionBlitFramebuffer workaround on iOS and fix a conflict when it is enabled in combination with clipSrcRegionBlitFramebuffer. Bug: angleproject:5417 Change-Id: I1de0f9f392643fbcf6900bad86d83084cb6d4009 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2601106 Reviewed-by: Kenneth Russell <kbr@chromium.org> Reviewed-by: Jonah Ryan-Davis <jonahr@google.com> Commit-Queue: James Darpinian <jdarpinian@chromium.org>
Jonah Ryan-Davis 88fec1e8 2020-12-23T11:53:55 Code generation for NV_framebuffer_blit Generate the entry points for NV_framebuffer_blit. Bug: angleproject:5474 Bug: chromium:1157057 Change-Id: Iadffa24ce7368d8ab6c4bf9d5b6c016276e762f5 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2602242 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Jonah Ryan-Davis <jonahr@google.com>
James Darpinian 0dbe7cdd 2020-12-08T12:24:35 iOS testing support angle_white_box_tests build and runs and passes on the iOS simulator with this change. angle_end2end_tests builds and runs but crashes. Bug: angleproject:4256 Bug: angleproject:5417 Change-Id: I8817e46415c4598cbfae49804727a2e9b21baff1 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2600361 Commit-Queue: James Darpinian <jdarpinian@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Kenneth Russell <kbr@chromium.org>
Shahbaz Youssefi eeb14308 2020-12-23T01:45:35 Vulkan: Support xfb capture of I/O block fields In the emulation path, it's ensured that the generated code references the I/O block field correctly (using the instance name if provided, and without it otherwise). In the extension path, the info map is augmented with an array of xfb decorations for its fields. Then when `OpDecorate %IOBlockId Block` is encountered, the transform feedback decorations on the fields are inserted: OpMemberDecorate %IOBlockId MemberN XfbBuffer buffer OpMemberDecorate %IOBlockId MemberN XfbStride stride OpMemberDecorate %IOBlockId MemberN Offset offset Future work includes removing the duplicate varying added for gl_PointSize and use this mechanism to decorate gl_PerVertex directly. Bug: angleproject:3606 Change-Id: I6fed0b1ee7245fe695337043b40b281fb01a1fb0 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2599953 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Shahbaz Youssefi 8065aa82 2020-12-22T23:38:42 Front-end support for xfb capture of I/O block members Validation and generation of transform feedback varyings that specify an I/O block member are implemented in this change. The GL backend is able to pass the added tests. Bug: angleproject:3606 Change-Id: I66d02bed8ca9161555d0d1e7a32ae9ef4d9e813f Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2599952 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com>
Jamie Madill b91afcfd 2020-12-11T10:58:54 Bump implementation max textures/images limit to 96. We need the higher limit for tessellation shaders and ES 3.2. Unfortunately this means we are no longer using a packed 64-bit bitset for the active textures and images. We fallback to using std::bitset. Also keeps the 64 texture limit if tessellation is not available. The higher limits was causing timeouts in WebGL tests. Bug: angleproject:3572 Change-Id: I1953955600b56d7c66178bd610de53453151dc8f Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2586996 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com>
James Darpinian c2a620b0 2020-12-22T11:44:32 Fix out of bounds indices in transform feedback test Type confusion on the index buffer. It should be unsigned shorts. Fixes a crash on iOS. Bug: angleproject:4992 Bug: angleproject:5417 Change-Id: I18179a89dd81fff2582636496ea9684e432f4400 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2601162 Reviewed-by: Kenneth Russell <kbr@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
angle-autoroll 5c0ba543 2020-12-22T09:32:42 Roll VK-GL-CTS from 54509765b18d to b29bf0434c16 (15 revisions) Manually update deqp.gni to include new es3cNumberParsingTests files. https://chromium.googlesource.com/external/github.com/KhronosGroup/VK-GL-CTS.git/+log/54509765b18d..b29bf0434c16 2020-12-19 alexander.galazin@arm.com Merge vk-gl-cts/vulkan-cts-1.2.5 into vk-gl-cts/master 2020-12-18 alexander.galazin@arm.com Raise min CMake version to 3.10.2 2020-12-18 jeremyg@lunarg.com Fix Renderpass storeOp synchronization hazards 2020-12-18 jeremyg@lunarg.com Fix back-to-back transfer command synchronization hazards 2020-12-18 gleese@broadcom.com Simplify code for subgroup builtin mask tests 2020-12-18 jeremyg@lunarg.com Fix end of Renderpass synchronization hazards 2020-12-18 jeremyg@lunarg.com Fix additional RenderPass loadOp/stencilLoadOp sync hazards. 2020-12-18 jeremyg@lunarg.com Fix RenderPass stencilLoadOp synchronization hazards. 2020-12-18 jeremyg@lunarg.com Fix RenderPass loadOp synchronization hazards. 2020-12-17 mikko.tiusanen@siru.fi Tests for GLES3 number parsing 2020-12-17 mikko.tiusanen@siru.fi Tests for multiple clears within render pass 2020-12-17 rgarcia@igalia.com Enable nullDescriptor for AS ray query tests 2020-12-15 piotr.byszewski@mobica.com Test dynamic indexing of AS 2020-12-15 rgarcia@igalia.com Test dynamic indexing of ray queries 2020-12-13 piotr.byszewski@mobica.com Test procedural geometry with complex BB sets If this roll has caused a breakage, revert this CL and stop the roller using the controls here: https://autoroll.skia.org/r/vk-gl-cts-angle-autoroll Please CC ynovikov@google.com,angle-bots+autoroll-info@google.com on the revert to ensure that a human is aware of the problem. To report a problem with the AutoRoller itself, please file a bug: https://bugs.chromium.org/p/skia/issues/entry?template=Autoroller+Bug Documentation for the AutoRoller is here: https://skia.googlesource.com/buildbot/+doc/master/autoroll/README.md Bug: None Tbr: ynovikov@google.com,angle-bots+autoroll-info@google.com Change-Id: I59e4aabcfbf496d02e18ba7b7aa06391ba90f434 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2600639 Reviewed-by: Yuly Novikov <ynovikov@chromium.org> Reviewed-by: angle-autoroll <angle-autoroll@skia-public.iam.gserviceaccount.com> Commit-Queue: Yuly Novikov <ynovikov@chromium.org>
Shahbaz Youssefi 74788951 2020-12-22T13:11:17 Vulkan: Fix assertion in xfb of indirect calls An assertion was fired which was only valid when emulating transform feedback. The relevant block is conditioned to the respective feature. Bug: angleproject:3571 Change-Id: I287ef6d94c920ccda742e4032bfc389409a38e1f Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2599951 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Shahbaz Youssefi 9d45f527 2020-12-21T14:23:11 Vulkan: GS tests that demonstrate gl_Position bugs Bug: angleproject:5478 Bug: angleproject:5479 Change-Id: If7ae7b6e0960c66dde697726381bc68c6809b3c6 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2599940 Reviewed-by: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Jamie Madill 00643e7d 2020-07-22T20:58:49 Use Android API to get storage path. This pipes through to a system call via JNI. This will allow Chromium changes to land that were prevented because ANGLE hard-coded some parts of this path. This in turn will allow us to more easily override these paths for changes needed for Android R support. Bug: chromium:1094062 Change-Id: I20d75b8ee40d418ba5c057f618640ef896248299 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2315483 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Jonah Ryan-Davis <jonahr@google.com>
Shahbaz Youssefi 85707f7f 2020-12-21T14:52:38 Vulkan: Pass in SPIR-V transform options in a struct Clean up change in preparation for changes that add more options. Bug: angleproject:5478 Change-Id: Id35825b337dba153a5c28dfcc311b344ce257f78 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2599941 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com>
Charlie Lao 0c5a55a5 2020-12-17T14:52:59 Vulkan: MapBufferRange should avoid wait if INVALIDATE_BUFFER is set If glMapBufferRange is called with GL_MAP_INVALIDATE_BUFFER_BIT bit set, caller indicates that it don't care about the previous content. If the buffer is busy, instead of wait for GPU to finish and then map the buffer, we should just allocate a new memory and return it. brawl_stars is hitting this case. With this CL, the frame time is cutting to half on the pixel device. Bug: b/175905404 Change-Id: If1220f07ebf53dd28fe6a4732eaba84e2e57598e Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2597784 Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Charlie Lao <cclao@google.com>
Shahbaz Youssefi 4e2b6d6b 2020-12-21T10:48:12 Fix PackedVarying's move constructor The move constructor was calling the move assignment operator, which used std::swap on the members. However, `this` was not constructed, so std::swap was operating on uninitialized values. Caught by UBSAN. Bug: chromium:1159444 Change-Id: I65f2f0affd1461f880c59780f31a368779411f24 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2599936 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Ian Elliott <ianelliott@google.com> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Le Hoang Quyen ed9213bc 2020-12-18T12:12:10 Metal: explicitly convert XFB vertex shader to void function Previously, spirv-cross would automatically do this (converting to void function) if shader writes to XFB buffers, however this behavior is changed in latest spirv-cross version. So the rasterization disable must be explicitly specified via spirv-cross’s options now. Bug: chromium:1159974 Change-Id: I9f34e1cd9d2f1386e8a8d030a800fbd7341eb68f Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2596914 Commit-Queue: Le Hoang Quyen <le.hoang.q@gmail.com> Reviewed-by: Jamie Madill <jmadill@chromium.org>
jchen10 0322fc1a 2020-12-16T11:02:13 Disable syncFramebufferBindingsOnTexImage on Intel/Win/GL The driver fix is available already. Bug: angleproject:2906 Change-Id: Id607fa5217360afd99e5de0d2e5f35f79615d160 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2594323 Commit-Queue: Jie A Chen <jie.a.chen@intel.com> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Kenneth Russell 011e0ad1 2020-12-16T17:34:50 Promote RGBA4444 and RGB565 to RGB[A]8 on macOS with AMD GPUs. There are bugs in Apple's OpenGL driver when uploading to textures of these formats via pixel buffer objects. Because automatic graphics switching might activate the AMD GPU at any time, use this workaround on dual-GPU MacBook Pros even if the Intel GPU is active. Fixes the following WebGL 2.0 conformance tests: deqp/functional/gles3/texturespecification/ teximage2d_pbo_2d_00.html teximage2d_pbo_2d_01.html texsubimage2d_pbo_2d_00.html texsubimage2d_pbo_2d_01.html texsubimage2d_pbo_cube_01.html Bug: angleproject:5469 Change-Id: I7d65dc5f664159b4b59d46400a8d9c492f67040b Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2595245 Commit-Queue: Kenneth Russell <kbr@chromium.org> Reviewed-by: Jonah Ryan-Davis <jonahr@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 653f6196 2020-12-14T12:55:55 Import Vulkan DEPS from the new vulkan-deps repo. This will allow the entire set of Vulkan DEPS to be rolled as one. Bug: angleproject:5390 Change-Id: I10ca4f966f986baf3aab7250935802a3ed1065f1 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2587671 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Yuly Novikov <ynovikov@chromium.org>
jchen10 c0bbe6cf 2020-12-16T11:56:52 Unsuppress TextureCubeTest.CubeMapFBO on Intel/Win/Vulkan This disables forceCpuPathForCubeMapCopy as well Bug: angleproject:2822 Bug: angleproject:3055 Change-Id: Iecb37d0c2d109e319c67310dd64ad8d4bbc14fd8 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2594707 Commit-Queue: Jie A Chen <jie.a.chen@intel.com> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Tim Van Patten de703db5 2020-12-16T20:39:59 Fix ValidateBufferStorageEXT() ValidateBufferStorageEXT() has a copy/paste error where the flag GL_MAP_PERSISTENT_BIT_EXT is enabled twice while GL_MAP_COHERENT_BIT_EXT is not enabled at all during parameter checking, causing ANGLE to reject valid calls to glBufferStorageEXT(). Bug: angleproject:5473 Test: BufferStorageTestES3.BufferStorageFlagsPersistentCoherentWrite Change-Id: I05596f54d48118f609ef1c88ca222d9fcdb2dd3a Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2596956 Commit-Queue: Tim Van Patten <timvp@google.com> Reviewed-by: Charlie Lao <cclao@google.com> Reviewed-by: Cody Northrop <cnorthrop@google.com>
Jonah Ryan-Davis 937bc82b 2020-12-17T17:57:50 Revert "GLX: Expose EGL_ANGLE_window_fixed_size" This reverts commit fa9b803e078fb833143df7a2b3f88c8694fa8f2f. Reason for revert: https://bugs.chromium.org/p/chromium/issues/detail?id=1157748 Original change's description: > GLX: Expose EGL_ANGLE_window_fixed_size > > Right now the GLX backend checks for resize on every SwapBuffers > call. If EGL_ANGLE_window_fixed_size is used, it will only resize > when signaled by Chrome. > > Bug: chromium:1132827 > Change-Id: Ia4ddbbbf6dcf13f46ad564208eccc517de8be33c > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2572886 > Reviewed-by: Jamie Madill <jmadill@chromium.org> > Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> > Commit-Queue: Jonah Ryan-Davis <jonahr@google.com> TBR=syoussefi@chromium.org,jonahr@google.com,jmadill@chromium.org # Not skipping CQ checks because original CL landed > 1 day ago. Bug: chromium:1132827 Change-Id: I1110c5dd5d806b3c6c59308d4865f0d3aa57fc3f Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2595197 Reviewed-by: Jonah Ryan-Davis <jonahr@google.com> Commit-Queue: Jonah Ryan-Davis <jonahr@google.com>
Tim Van Patten 909ea88b 2020-11-20T13:07:53 Reland "Vulkan: Ignore glFlush to reduce vkQueueSubmits in Asphalt 9" This is a reland of 5cf7472dd161bbda329dfc5e4e65bb6ce0c06fbd The ShareGroupVk::mResourceUseLists was not being cleared each call to RendererVk::submitFrame(), so it was growing indefinitely. Each vk::ResourceUseList within it was cleared, so it was holding an essentially "infinite" list of empty lists, but that caused the loop in RendererVk::submitFrame() to take more and more time until the tests timed out. The fix is to do 'resourceUseLists.clear()' once the loop to release all resources has completed, like releaseResourceUsesAndUpdateSerials() does for each individual list. Additionally, ASSERTs are added to guarantee that the lists are empty when the ContextVk and ShareGroupVk are destroyed. Original change's description: > Vulkan: Ignore glFlush to reduce vkQueueSubmits in Asphalt 9 > > Multithreaded apps can use the following pattern: > > glDrawElements() > glFenceSync() > glFlush() > glWaitSync() > > This currently results in a vkQueueSubmit for every glFlush() to ensure > that the work has landed in the command queue in the correct order. > However, ANGLE can instead avoid the vkQueueSubmit during the glFlush() > in this situation by instead flushing the ContextVk's commands and > ending the render pass to ensure the commands are submitted in the > correct order to the renderer. This improves performance for Asphalt 9 > by reducing frame times from 150-200msec to 35-55msec. > > Specifically, ANGLE will call flushCommandsAndEndRenderPass() when > there is a sync object pending a flush or if the ContextVk is currently > shared. > > Additionally, on all devices except Qualcomm, ANGLE can ignore all other > glFlush() calls entirely and return immediately. For Qualcomm devices, > ANGLE is still required to perform a full flush (resulting in a > vkQueueSubmit), since ignoring the glFlush() reduces the Manhattan 3.0 > offscreen score by ~3%. > > Bug: angleproject:5306 > Bug: angleproject:5425 > Change-Id: I9d747caf5bf306166be0fec630a78caf41208c27 > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2552718 > Commit-Queue: Tim Van Patten <timvp@google.com> > Reviewed-by: Charlie Lao <cclao@google.com> > Reviewed-by: Jamie Madill <jmadill@chromium.org> Bug: angleproject:5306 Bug: angleproject:5425 Bug: angleproject:5470 Change-Id: I14ee424d032f22e5285d67accbec078ad1955dd0 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2595811 Commit-Queue: Tim Van Patten <timvp@google.com> Reviewed-by: Charlie Lao <cclao@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Mohan Maiya 33fb6e38 2020-12-16T13:01:52 Vulkan: Unskip tests with passthrough GLSL function The 2 end2end tests added by a62ee4d1 no longer fail with latest Pixel ICDs. Enable those tests as regression checks. Tests: angle_end2end_tests --gtest_filter=*SamplerPassthrough*Vulkan* Bug: angleproject:5457 Change-Id: I53fb06fffa3fc76ef6f9a9e811ef5b833577ed67 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2595389 Reviewed-by: Cody Northrop <cnorthrop@google.com> Commit-Queue: Mohan Maiya <m.maiya@samsung.com>
jchen10 9bb06f37 2020-12-16T10:24:29 Unsuppress GLSLTest.PointCoordConsistency on Intel/Win/GL The test can pass on 26.20.100.8141 now. Bug: angleproject:2805 Change-Id: I8049c43ed7eca6c2001a8d8745aa6f9c2a2192a5 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2594349 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Corentin Wallez a19bd601 2020-12-16T13:04:38 Revert "Vulkan: Ignore glFlush to reduce vkQueueSubmits in Asphalt 9" This reverts commit 5cf7472dd161bbda329dfc5e4e65bb6ce0c06fbd. Reason for revert: causes timeouts, see anglebug.com/5470 Original change's description: > Vulkan: Ignore glFlush to reduce vkQueueSubmits in Asphalt 9 > > Multithreaded apps can use the following pattern: > > glDrawElements() > glFenceSync() > glFlush() > glWaitSync() > > This currently results in a vkQueueSubmit for every glFlush() to ensure > that the work has landed in the command queue in the correct order. > However, ANGLE can instead avoid the vkQueueSubmit during the glFlush() > in this situation by instead flushing the ContextVk's commands and > ending the render pass to ensure the commands are submitted in the > correct order to the renderer. This improves performance for Asphalt 9 > by reducing frame times from 150-200msec to 35-55msec. > > Specifically, ANGLE will call flushCommandsAndEndRenderPass() when > there is a sync object pending a flush or if the ContextVk is currently > shared. > > Additionally, on all devices except Qualcomm, ANGLE can ignore all other > glFlush() calls entirely and return immediately. For Qualcomm devices, > ANGLE is still required to perform a full flush (resulting in a > vkQueueSubmit), since ignoring the glFlush() reduces the Manhattan 3.0 > offscreen score by ~3%. > > Bug: angleproject:5306 > Bug: angleproject:5425 > Change-Id: I9d747caf5bf306166be0fec630a78caf41208c27 > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2552718 > Commit-Queue: Tim Van Patten <timvp@google.com> > Reviewed-by: Charlie Lao <cclao@google.com> > Reviewed-by: Jamie Madill <jmadill@chromium.org> TBR=timvp@google.com,jmadill@chromium.org,cclao@google.com Change-Id: I9886bf901a835d408b6a4b8be7ea408fa2121be0 No-Presubmit: true No-Tree-Checks: true No-Try: true Bug: angleproject:5306 Bug: angleproject:5425 Bug: angleproject:5470 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2595032 Reviewed-by: Corentin Wallez <cwallez@chromium.org> Commit-Queue: Corentin Wallez <cwallez@chromium.org>
Mohan Maiya 98b56e60 2020-12-12T16:28:21 Vulkan: Accumulate internal cache stats in renderer The CacheStats of all internal caches are accumulated by the renderer. In order to see the hit ratios of all caches, the following GN args must be enabled: is_debug = true angle_enable_perf_counter_output = true Bug: angleproject:5447 Test: Manual verification with angle_end2end_tests Change-Id: Iaca3249192e9e4e130d8291b7759c459d79b06ee Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2588430 Commit-Queue: Mohan Maiya <m.maiya@samsung.com> Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Tim Van Patten 5cf7472d 2020-11-20T13:07:53 Vulkan: Ignore glFlush to reduce vkQueueSubmits in Asphalt 9 Multithreaded apps can use the following pattern: glDrawElements() glFenceSync() glFlush() glWaitSync() This currently results in a vkQueueSubmit for every glFlush() to ensure that the work has landed in the command queue in the correct order. However, ANGLE can instead avoid the vkQueueSubmit during the glFlush() in this situation by instead flushing the ContextVk's commands and ending the render pass to ensure the commands are submitted in the correct order to the renderer. This improves performance for Asphalt 9 by reducing frame times from 150-200msec to 35-55msec. Specifically, ANGLE will call flushCommandsAndEndRenderPass() when there is a sync object pending a flush or if the ContextVk is currently shared. Additionally, on all devices except Qualcomm, ANGLE can ignore all other glFlush() calls entirely and return immediately. For Qualcomm devices, ANGLE is still required to perform a full flush (resulting in a vkQueueSubmit), since ignoring the glFlush() reduces the Manhattan 3.0 offscreen score by ~3%. Bug: angleproject:5306 Bug: angleproject:5425 Change-Id: I9d747caf5bf306166be0fec630a78caf41208c27 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2552718 Commit-Queue: Tim Van Patten <timvp@google.com> Reviewed-by: Charlie Lao <cclao@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Paul Thomson 4abf3788 2020-12-15T13:44:53 Tests: update WhatsApp trace The WhatsApp trace was previously captured with the GL_NV_shader_noperspective_interpolation extension enabled. This re-capture was made with the extension disabled so the trace can replayed on devices that don't support the extension. Test: angle_perftests --gtest_filter="*whatsapp*" Bug: b/174256233 Change-Id: I9dbf580ed5849b919a72944879be35193632e34a Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2593191 Commit-Queue: Cody Northrop <cnorthrop@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Cody Northrop <cnorthrop@google.com>
Charlie Lao 64508f44 2020-12-12T18:00:10 Vulkan: Reduce driver uniform data to minimum if specConst is used If specialization constant is used, driver uniform data structure should be reduced to minimum to increase cache locality. Bug: b/175479076 Change-Id: I1fc50666542c6763c60bfe011cde5bc77ccc08e1 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2588549 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Charlie Lao <cclao@google.com>
Mohan Maiya 8f9e17d5 2020-12-14T14:11:11 Add EXT_clip_cull_distance extension autogen code Addition of shader autogen code for EXT_clip_cull_distance extension. Bug: angleproject:5458 Change-Id: I299528957c0e747fd84987c420588c314f54aae0 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2590989 Commit-Queue: Mohan Maiya <m.maiya@samsung.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Mohan Maiya 46eaba7f 2020-12-12T10:31:26 Vulkan: Add support for internal cache hit and miss counts Add a CacheStats class that provides cache hit and miss bookkeeping. All internal caches make use of this class to keep track of its stats. This provides a means to profile cache hit ratios a.k.a Vulkan object reuse for any application. Bug: angleproject:5447 Test: Manual verification with angle_end2end_tests Change-Id: I44eeb0c2b9b291ec1cdd156fb2be4a5fe80d2848 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2580111 Commit-Queue: Mohan Maiya <m.maiya@samsung.com> Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Mohan Maiya f32fbb51 2020-12-14T14:42:58 Add EXT_shader_framebuffer_fetch_non_coherent entry points Addition of the entry points for EXT_shader_framebuffer_fetch_non_coherent extension. Bug: angleproject:5454 Change-Id: I2b8ba5141eed61a0c3ba58aeb2eec12a2071e7f1 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2590991 Commit-Queue: Mohan Maiya <m.maiya@samsung.com> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Mohan Maiya 9229b805 2020-12-14T14:10:28 Vulkan: Add the tests for EXT_clip_cull_distance Before supporting EXT_clip_cull_distance extension, some tests are added to validate the implementation of EXT_clip_cull_distance extension. These tests are implemented based on the tests for APPLE_clip_distance Bug: angleproject:5458 Tests: angle_end2end_tests --gtest_filter=Clip*DistanceTest* angle_unittests --gtest_filter=*Clip*Distance* Change-Id: I018c72ae8f0aff616c9d2600e63246e9609cf3de Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2585986 Commit-Queue: Mohan Maiya <m.maiya@samsung.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Tim Van Patten e8c5525c 2020-12-09T09:41:21 Vulkan: flushCommandsAndEndRenderPass during glFenceSync() The description for glFenceSync states: glFenceSync — create a new sync object and insert it into the GL command stream ANGLE needs to break the render pass and flush any prior commands to ensure that the newly created Fence object lands with the correct ordering. This will be done by calling ContextVk::flushCommandsAndEndRenderPass() before creating the Vulkan Event object. Bug: angleproject:5306 Change-Id: I3a460bb559acf8eb4b3041065ec743938d44d823 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2582199 Reviewed-by: Charlie Lao <cclao@google.com> Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Tim Van Patten <timvp@google.com>
Paul Thomson c47b951e 2020-12-15T13:17:28 Capture: Disable GL_NV_shader_noperspective_interpolation Skia uses the extension GL_NV_shader_noperspective_interpolation, which is implemented by ANGLE. This leads to traces with shaders that use the extension. Such traces do not replay on some native GLES drivers, as the extension is not available. The fix is to disable the GL_NV_shader_noperspective_interpolation extension when capturing, so that apps do not use this extension. Bug: b/174256233 Change-Id: I5265cab2a56be488669ed31aa2ea8ec7b7fad8fd Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2593190 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Paul Thomson <paulthomson@google.com>
Jamie Madill 4798c8cf 2020-12-14T16:35:27 Introduce GN variables for Vulkan repos. This will allow them to be seamlessly overrideen in other repos. Bug: angleproject:5390 Change-Id: I973cae48b5683e39ea3b07898b95956511260319 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2591107 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Le Hoang Quyen cb8903b1 2020-12-08T01:08:00 Metal: Ignore OS's internal shader cache when testing. Internal shader cache caused timeout in some dEQP tests. Work-around: ignore the cache by hooking fopen function and return null when the cache related files are accessed. Bug: angleproject:5354 Change-Id: I12ca228540925e67454bf24ce1ba83d703882c87 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2580918 Commit-Queue: Le Hoang Quyen <le.hoang.q@gmail.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Kenneth Russell <kbr@chromium.org>
Corentin Wallez fd7733e9 2020-12-14T19:04:28 GeometryTest: Expand suppressions to AMD Intel OpenGL TBR=syoussefi@chromium.org Bug: angleproject:5463 Change-Id: I3bdd17b5f9fd413e8fcbe734c6cb6f2110364e30 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2590185 Commit-Queue: Corentin Wallez <cwallez@chromium.org> Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Mohan Maiya a62ee4d1 2020-12-11T14:28:42 Vulkan: Add test for Pixel bug with passthrough GLSL function Passing in a sampler2D's values using texture2D through a function generates unexpected results on a Pixel device with the Vulkan backend. 1. SamplerPassthroughFailedLink - program fails to link. 2. SamplerPassthroughIncorrectColor - output color is incorrect. These tests have no issues with SwiftShader ICD. Tests: angle_end2end_tests --gtest_filter=*SamplerPassthrough*Vulkan* Bug: angleproject:5457 Change-Id: I9c43062e98c7b3637bb8a2b00a957141da821f0a Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2586059 Commit-Queue: Cody Northrop <cnorthrop@google.com> Reviewed-by: Cody Northrop <cnorthrop@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Xiaoxuan Liu d0f99d54 2020-12-03T11:06:56 Vulkan: Fix build issue in vulkan display/headless backend. In Linux + Vulkan display/headless backend, angle tests such as angle_unittests/angle_perftests/angle_white_box_perf_tests could not be compiled successfully, fix build issues on vulkan display/headless backend in angle tests. 1. Add EGL_NO_X11 flag into vulkan display/headless backend. 2. Disable angle_white_box_perf_tests on vulkan display/headless backend. Bug: angleproject:5260 Change-Id: I579aee8b0bfaa4c0d7fdf9a6df91c1b78ec4373e Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2585256 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Shahbaz Youssefi 282fb409 2020-12-11T16:09:10 Fix link validation of I/O block members Location and struct name matching for fields was missing as they only apply to I/O blocks and not varyings of struct type. Bug: angleproject:3580 Change-Id: I69083f39088458da72828b418be3068187a30fcc Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2587456 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Mohan Maiya da8c2261 2020-12-12T16:39:55 Vulkan: Enable FramebufferVk cache on Apple This effectively reverts ff60aba. The crash no longer occurs on Apple. Bug: angleproject:4442 Change-Id: I4aa745c80a482eb99311f3810e34124afe950cfe Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2588429 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Mohan Maiya <m.maiya@samsung.com>
Shahbaz Youssefi 56330564 2020-12-10T00:46:04 Vulkan: Support layered framebuffers This feature is introduced by geometry shaders, where all the layers of a texture can be attached to a framebuffer. The geometry shader would use gl_Layer to decide which layer the primitive should be rendered to. Bug: angleproject:3571 Change-Id: Ib2ae8e227b226295f9e2f62f6b230839070bc95c Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2582711 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com>
Shahbaz Youssefi 8326b26a 2020-12-11T13:59:40 Fix link validation with ambiguous instanceless interface blocks The following interface blocks should fail link: VS: layout(binding=0) buffer BufferBlockNameA { mediump float variable; }; FS: layout(binding=1) buffer BufferBlockNameB { mediump float variable; }; Because `variable` is ambiguous. Bug: angleproject:3580 Change-Id: I29576a6f152780819af0e9fb63249dbee7d9f2fc Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2587450 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Shahbaz Youssefi d9318acc 2020-12-11T10:40:21 Roll VK-GL-CTS from 41331850eb21 to 54509765b18d (18 revisions) https://chromium.googlesource.com/external/github.com/KhronosGroup/VK-GL-CTS.git/+log/41331850eb21..54509765b18d 2020-12-10 boris.zanin@mobica.com Basic tests for VK_EXT_conservative_rasterization 2020-12-10 ari.suonpaa@siru.fi Add test for image copy special case 2020-12-10 mikko.tiusanen@siru.fi Add tests for GLES3 vector construction from other types 2020-12-10 alexander.galazin@arm.com Merge vk-gl-cts/opengl-es-cts-3.2.6 into vk-gl-cts/master 2020-12-09 alexander.galazin@arm.com Merge vk-gl-cts/vulkan-cts-1.2.5 into vk-gl-cts/master 2020-12-09 ari.suonpaa@siru.fi Add drawing tests for Android hardware buffer based color buffer 2020-12-09 rgarcia@igalia.com Test data spilling and unspilling around RT shader calls 2020-12-09 dominik.witczak@amd.com Add new tests that verify IgnoreIntersection/TerminateRay corner cases. 2020-12-09 rgarcia@igalia.com Test maxPipelineRayRecursionDepth can be zero 2020-12-09 gleese@broadcom.com Enable YCbCr filtering tests for non-separable chroma 2020-12-07 slawomir.cygan@intel.com Fix clang 10.0 build of ray tracing control flow tests 2020-12-07 slawomir.cygan@intel.com Remove merge conflict markers from mandatory_features.txt 2020-12-07 jan.mroczkowski@mobica.com Vulkan Portability support 2020-12-07 piotr.byszewski@mobica.com Disable conflicting fragment shading rate features 2020-12-07 alexander.galazin@arm.com Update Vulkan headers 2020-12-07 tzlatinski@nvidia.com Select SPIR-V for the robustness' compute shaders 2020-12-07 rgarcia@igalia.com Fix buffer offsets in push descriptor tests 2020-12-07 rgarcia@igalia.com Destroy deferred ops before destroying devices Bug: angleproject:5461 Change-Id: I86f70ea9b77c114ed9495215d83a6fee61adc76a Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2586995 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Shahbaz Youssefi ec1ff8f5 2020-12-09T14:25:56 Add a scissored MSRTT test Covers a bug Skia's implementation of unresolve uncovered. ANGLE's implementation is slightly different (in that the unresolve pass is full screen), so it's not affected by this bug. Bug: angleproject:4836 Change-Id: Ibc8010b0b5ef6912a45fd5f2612fb8b619e4407e Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2582707 Reviewed-by: Jonah Ryan-Davis <jonahr@google.com> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Charlie Lao c75473c2 2020-11-23T14:26:32 Vulkan: Generalize FlipRotationSpecConst to SpecializationConstant Specialization constant are used not just for flip/rotation. It also used for other things. This CL merges all specialization constant usage (lineRasterEmulation, flip, rotation, halfRenderArea) into one class and rename FlipRotationSpecConst to SpecConst. Bug: b/173800146 Change-Id: I8dc3354b6caedbb183cec29855fc1c301ec8872a Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2555812 Commit-Queue: Charlie Lao <cclao@google.com> Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Courtney Goeltzenleuchter dcc0bebf 2020-12-10T13:43:03 Vulkan: clean up comments, rename per request Bug: b/170312581 Change-Id: Iefecb5776cac07477266e6a2e77af7d3c9f38ad6 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2585745 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Courtney Goeltzenleuchter <courtneygo@google.com>
Courtney Goeltzenleuchter 6d94471b 2020-11-19T16:53:09 Vulkan: Eliminate fence wait from SyncVk Waiting on a fence is problematic if using threaded worker to submit work since code needs to know if the work has been submitted or not before doing the wait. Eliminate the wait by using a serial if available or if no serial check the file descriptor directly. If no serial it's because we are waiting on an imported file descriptor. Could call fence.wait in that case but can also wait on the file descriptor. Test: angle_end2end_tests --gtest_filter=EGLSyncTest.*/ES2_Vulkan Bug: b/170312581 Change-Id: I392a5e73ac8f851d0d5bc53f06063568c0c90d2c Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2579042 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Courtney Goeltzenleuchter <courtneygo@google.com>
Jamie Madill f6df8692 2020-12-09T12:46:35 Vulkan: Support XFB in non-Vertex stages. This updates the code in several places to support Geometry and Tessellation Evaluation shaders to output transform feedback. Does not turn on any new tests but enables support for XFB when we turn on GS/TS. Bug: angleproject:3571 Bug: angleproject:3572 Change-Id: I6dcb768f2df4eeee81a4a500e999fcf16716f58f Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2581941 Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Jamie Madill 4d9157a5 2020-12-09T16:41:31 Increase internal max uniform blocks limit to 60. This will make space for the tessellation / ES 3.2 limit of 60 blocks. Bug: angleproject:3572 Change-Id: Ie6cc6b24bdce2f5f67f62d11cdeefb10f3bd01d0 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2582708 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Xiaoxuan Liu 108b759e 2020-12-04T15:15:49 EGL: Update EGL headers/xml The new EGL headers introduced 'EGL_NO_X11' which we could use for ANGLE vulkan display/headless backend. Changes in CL: 1. Updated include/EGL/egl*.h and scripts/egl.xml based on latest EGL repo: https://github.com/KhronosGroup/EGL-Registry Note: local modifications to the file were preserved in eglext.h, search keyword 'eglext_angle.h' for detail 2. run scripts to update entry_points/loader scripts/generate_entry_points.py scripts/generate_loader.py scripts/run_code_generation.py 3. Update ANGLE code on API 'eglSwapBuffersWithDamage' 4. Format with 'git cl format' Bug: angleproject:5260 Change-Id: I70ed0dccecf0426929ef8b4775605554d66c5724 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2576314 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Charlie Lao 067a12f0 2020-11-19T16:47:25 Reland "Vulkan: Use specialization constant for halfRenderAreaWidth" This is a reland of ff38106d62fca2e36ec2db1cd580b546fd53ae43 Original change's description: > Vulkan: Use specialization constant for halfRenderAreaWidth > > halfRenderArea is used by fragment shader to adjust gl_FragCoord for > rotation and yflip compensation. This CL bakes halfRenderAreaWidth into > shader via specialization constant, thus allow compiler to consolidate > the calculation into one MAD instruction. > > Bug: b/173800146 > Change-Id: Id66301278e3389e2582369b695825e632bccecee > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2551541 > Commit-Queue: Charlie Lao <cclao@google.com> > Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> > Reviewed-by: Tim Van Patten <timvp@google.com> > Reviewed-by: Jamie Madill <jmadill@chromium.org> Bug: b/173800146 Change-Id: Ifc36086f4c5d6a44fb5456aa6e31bc8a783bba71 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2579648 Commit-Queue: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Charlie Lao 1788355b 2020-12-08T16:12:34 Vulkan: Add dirty bits instead of direct access FBO state The CL crrev/c/2551541 causes intermittent Bot failure. I believe the problem is that from ContextVk::onMakeCurrent call stack, the object state may not in sync or even not completely specified. So instead of accessing fbo state object here to update mGraphicsPipelineDesc, we should insert dirty bit and let state validation code to handle it. Bug: b/175157604 Change-Id: I5f6b9fd901cc4187068f6161bd12836c8bbf8e87 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2580914 Commit-Queue: Ian Elliott <ianelliott@google.com> Commit-Queue: Charlie Lao <cclao@google.com> Reviewed-by: Ian Elliott <ianelliott@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Shahbaz Youssefi f9f569e2 2020-12-09T11:37:10 Suppress test requiring VK_EXT_transform_feedback on Nvidia/win7 Bug: angleproject:5450 Bug: angleproject:5435 Change-Id: I1925a4f926d10c81e5c69cfc62fcb26da2cec81d Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2581940 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Yuly Novikov <ynovikov@chromium.org>
Shahbaz Youssefi 8789457d 2020-12-09T00:21:08 Fix varying packing of I/O block fields The sorting algorithm didn't keep the fields in order, which made them receive arbitrary locations. The SPIR-V transformer assigns the location on the whole block instead of individual members, which was incorrect in this situation. The SPIR-V transformer could have been modified to decorate each field of the I/O block with a location. This change instead sorts the fields in such a way that I/O block fields are allocated contiguously, which allows the SPIR-V transformer to function unchanged. Bug: angleproject:3580 Change-Id: I27df9e8122dd4207835bad448ffb8015692a1635 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2581076 Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Xinghua Cao 378653f8 2020-11-20T12:24:41 D3D: throw a perf warning for uniform block We had translated an uniform block only containing a large array member into StructuredBuffer instead of cbuffer on D3D backend for slow fxc compile performance issue with dynamic uniform indexing. This patch throw a warning if a uniform block containing a large array member fails to hit the optimization. Bug: angleproject:3682 Change-Id: I33459b559923f16a8dfb70c6f46ec52f68d96e06 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2552365 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Jiajia Qin <jiajia.qin@intel.com>
Shahbaz Youssefi 8b851562 2020-12-08T16:51:51 Don't output matrix qualifiers in I/O blocks row_major and column_major only apply to uniform and buffer interface blocks per GLSL ES 3.2 spec. Bug: angleproject:3580 Change-Id: Iad0afeffc7ddb7bff2ebaec0ea0ec4eda26171a7 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2580191 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Shahbaz Youssefi 36f74334 2020-12-03T21:26:28 Vulkan: Fix query pause on framebuffer binding change When a render pass is closed, render pass queries are paused. The code that pauses queries however is conditioned to the render pass being open. In ContextVk::syncState, when processing gl::State::DIRTY_BIT_DRAW_FRAMEBUFFER_BINDING, `onRenderPassFinished()` is called. Later on, when the render pass is actually finished, the queries are not paused. This change moves the logic to pause render pass queries to onRenderPassFinished(). Bug: angleproject:5427 Change-Id: I3a87db2e4543ff698803ac5e154a370e85ac7985 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2573581 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Charlie Lao <cclao@google.com>
Jiawei Shao 54d95994 2020-12-09T09:36:00 Enable ComputeShaderTest/DrawDispatchDispatchDraw on Intel/Win/Vulkan This patch enables ComputeShaderTest/DrawDispatchDispatchDraw on Intel/Win/Vulkan after the upgrade of Chromium try bots. Bug: angleproject:3871 Change-Id: I3e8d36020cbb01304a7b024d5b74923640402b4e Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2580112 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Jiajia Qin <jiajia.qin@intel.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Shahbaz Youssefi f691b3b5 2020-12-02T13:11:54 Vulkan: Support PrimitivesGenerated query This query uses the Vulkan transform feedback extension. In GL, GL_TRANSFORM_FEEDBACK_PRIMITIVES_WRITTEN and GL_PRIMITIVES_GENERATED queries can be independently begun/ended. However, Vulkan requires that queries from pools of the same type can only be active one at a time. This forbids the two GL queries from being handled by two VK queries when they are simultaneously begun. This change makes these queries share their QueryHelper objects. The Vulkan transform feedback queries unconditionally retrieve both results anyway, so this is just a matter of making sure the two GL queries are merged as one when they are simultaneously used. The change fixes a number of issues as collateral: - TransformFeedbackPrimitivesWritten queries when !emulated were not released - Stashed queries were never released - If no render pass is open when a query ends, then getResult(no_wait) ended up waiting Bug: angleproject:5404 Change-Id: I8ce13ea76ffd31b3152ded7c713c6466d0315504 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2573580 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Charlie Lao <cclao@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Jonah Ryan-Davis fa9b803e 2020-12-03T17:53:12 GLX: Expose EGL_ANGLE_window_fixed_size Right now the GLX backend checks for resize on every SwapBuffers call. If EGL_ANGLE_window_fixed_size is used, it will only resize when signaled by Chrome. Bug: chromium:1132827 Change-Id: Ia4ddbbbf6dcf13f46ad564208eccc517de8be33c Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2572886 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Jonah Ryan-Davis <jonahr@google.com>
Courtney Goeltzenleuchter aafcb504 2020-12-08T09:32:10 Vulkan: Add ensureSubmission to queueSubmitOneOff Some callers of queueSubmitOneOff require that the command being queued to have been sent to the GPU. The new ensureSubmission parameter indicates that behavior and when running with threaded worker will wait for the worker queue to empty before returning. Bug: b/170312581 Change-Id: Ib620fb37f4b9b4431451ccbd10807c0dff1842af Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2579041 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Cody Northrop <cnorthrop@google.com> Commit-Queue: Courtney Goeltzenleuchter <courtneygo@google.com>
kmx.io     mail     discord kmxgit v0.4.0