kc3-lang/angle/src/libANGLE/renderer/d3d/d3d11/TextureStorage11.h

src/libANGLE/renderer/d3d/d3d11/TextureStorage11.h

Log

Author Commit Date CI Message
Jiajie Hu e7d27705 2020-04-30T05:50:21 Fix corruption when changing the base level of a framebuffer texture attachment In the D3D renderer, changing the base level may trigger re-allocation of the texture storage, for example if the new base level has a different aspect ratio. During the process, image contents in the texture storage should be backed up properly. The D3D11 backend does this if an image has been associated with the texture storage, but it may happen such an association has never been established, and corruption will be observed then. The proposed patch mitigates the problem by introducing a new method named findRenderTarget(), with which one can tell if a mip level has been used as the render target. This works based on the fact that render targets are cached in the texture storage object. Hence all mip levels of interest can be found, without relying on the association between images and texture storage. Bug: angleproject:2291 Change-Id: Ic73af7b603be25c65760928f276bec16df003baf Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2158830 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Mingyu Hu 3b43647f 2019-11-15T14:17:55 EXT_multisampled_render_to_texture clean up and prep for 2DMultisampleArray. 1) Added new Multisampled Image Index. For the upcoming multisampled array case, we cannot automatically create the same MS image index during resolve. So the correct MS image index is created and saved during the initial creation of the MS Texture. It seems to be more streamlined as well, only creating the MS image index once, and not using hardcoded values to get the subresourceIndex. 2) reordering resolveTextureHelper to remove duplicated code. Bug: angleproject:3107 Change-Id: Ifd91136a64efb5e7bd296e09cd876a6e6eb8e77e Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/1919555 Commit-Queue: Rafael Cintron <rafael.cintron@microsoft.com> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Mingyu Hu 5f4db6a7 2019-11-11T14:52:37 Fix performance issue with EXT_multisampled_render_to_texture Adding a dirty flag to the mMSTexInfo struct so that the multisampled texture does not get unnecessarily resolved on each call. Bug: chromium:1019199 Change-Id: Ia9b24cd0a7f2d82f83f20f7e60f9f2b23383ec6f Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/1910531 Reviewed-by: Rafael Cintron <rafael.cintron@microsoft.com> Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Mingyu Hu 9122bec2 2019-10-21T11:03:48 Fix EXT_multisample_render_to_texture perf regression EXT_multisampled_render_to_texture was causing perf degradation through multiple unnecessary creation/destruction of the shadow multisampled texture. This is solved by keeping around the multisampled texture, so both will be available. Bug: chromium:1015031 Change-Id: I44deee67c02c19bc8749d3ed0d45661f52902e72 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/1872618 Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Mingyu Hu 2d0e5b55 2019-08-27T13:49:07 GL_EXT_multisampled_render_to_texture extension. Part 2. For textures that use this extension, a multisampled texture is implicitly created for the texture. Upon write or read, the multisampled texture is either return to be drawn to or resolved and returned as a single sampled texture. This is the functionality change with end2end tests. Bug: angleproject:980428 Change-Id: I5776875a132fed7a3f4f00fb02f9e8e250684630 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/1773717 Commit-Queue: Rafael Cintron <rafael.cintron@microsoft.com> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Mingyu Hu 7e44ec26 2019-08-26T15:59:48 GL_EXT_multisampled_render_to_texture extension. Part 1. Adding new parameters for extension without adding any real code change. Since no new code paths were added, we expect all tests to pass as before. Bug: angleproject:980428 Change-Id: I551b46a66f422eabd357fd021e00cf266a991efb Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/1772377 Commit-Queue: Rafael Cintron <rafael.cintron@microsoft.com> Reviewed-by: Rafael Cintron <rafael.cintron@microsoft.com> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Stuart Morgan 9d737966 2019-08-14T12:25:12 Standardize copyright notices to project style For all "ANGLE Project" copyrights, standardize to the format specified by the style guide. Changes: - "Copyright (c)" and "Copyright(c)" changed to just "Copyright". - Removed the second half of date ranges ("Y1Y1-Y2Y2"->"Y1Y1"). - Fixed a small number of files that had no copyright date using the initial commit year from the version control history. - Fixed one instance of copyright being "The ANGLE Project" rather than "The ANGLE Project Authors" These changes are applied both to the copyright of source file, and where applicable to copyright statements that are generated by templates. BUG=angleproject:3811 Change-Id: I973dd65e4ef9deeba232d5be74c768256a0eb2e5 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/1754397 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Qin Jiajia e7da32f1 2018-12-27T15:31:47 Fix the failure in UpdateImageTextureInUse If we update the texture data between two dispatch calls, ReadPixels can't get right result after the second dispatch call. The failure reason is that after the first dispatch, ReadPixels will sync framebufer state which will update color render target. Finally, TextureD3D::ensureRenderTarget is reached. However, we are in compute pipeline. mTexStorage->isRenderTarget() will be false. That results the current texture will create a new render target storage. But the UAV is still bound with the previous texture storage. If there is no texture dirty bit between these two dispatch calls, applyTexturesForCompute won't be called. After the second dispatch, readPixels will read data from the new texture storage which is not updated. Bug: angleproject:3015 Change-Id: Ib2494ab8bf6e12faefc0a7370719d383526c36ba Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/1390710 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Jiajia Qin <jiajia.qin@intel.com>
Geoff Lang 48d040e8 2019-01-18T10:43:45 D3D11: Fix reads and writes to EGL image textures with mip offsets. RenderTarget11 knows how to render to a specific mip of a texture but if TextureStorage11 attempts to read or write to the backing resource directly it does not select the correct subresource index. This manifested in calls to glTexSubImage going to the wrong mip of the resource. Fix TextureStorage11::getSubresourceIndex to look up the correct subresource in TextureStorage11_EGLImage. Add a test to cover this case in ImageTest.MipLevels. BUG=angleproject:2668 Change-Id: Id9270ab3bf9f37c7c671639bb8b7ce9578daaed6 Reviewed-on: https://chromium-review.googlesource.com/c/1422058 Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Geoff Lang <geofflang@chromium.org>
Jamie Madill b980c563 2018-11-27T11:34:27 Reformat all cpp and h files. This applies git cl format --full to all ANGLE sources. Bug: angleproject:2986 Change-Id: Ib504e618c1589332a37e97696cdc3515d739308f Reviewed-on: https://chromium-review.googlesource.com/c/1351367 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Olli Etuaho efbabde2 2018-08-29T14:37:36 Support multisample array textures on D3D11 This implements complete support for multisample array textures on the D3D11 backend. There's a specialized TextureD3D class as well as a TextureStorage11 class for multisample array textures that have the bulk of the functionality. BUG=angleproject:2775 TEST=angle_end2end_tests, angle_deqp_gles31_tests Change-Id: I68116635c01cef0e48c089fd9da05a844ef2d802 Reviewed-on: https://chromium-review.googlesource.com/1216003 Reviewed-by: Corentin Wallez <cwallez@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>
Jamie Madill 8370e6bc 2018-09-11T17:56:23 Rename constant to kCubeFaceCount. This corrects the naming style for this constant variable. Also updates many usages in TextureD3D.cpp and TextureGL.cpp. Bug: angleproject:2470 Change-Id: I029dd98e925cabf23a4bb26467a62afe6c5bb6e0 Reviewed-on: https://chromium-review.googlesource.com/1220726 Reviewed-by: Yuly Novikov <ynovikov@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Olli Etuaho 480edb8c 2018-08-30T18:25:43 Clean up TextureStorage11_2DMultisample This de-duplicates functionality from TextureStorage11_EGLImage and TextureStorage11_2DMultisample. Neither of those supports changing the associated images or image access in shaders. Multisampled textures also cannot be compressed, so the size doesn't need to be adjusted according to block size. BUG=angleproject:2275 TEST=angle_end2end_tests Change-Id: I45bc573584065bf4f528f883435fadfbd7281077 Reviewed-on: https://chromium-review.googlesource.com/1216002 Reviewed-by: Corentin Wallez <cwallez@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>
Till Rathmann c1551dc2 2018-08-15T17:04:49 D3D11 fix: Mipmapping was used although disabled via Sampler When using a sampler the texture was created as if it has mipmaps, regardless what you specified in GL_TEXTURE_MIN_FILTER via glSamplerParameteri() -- mistakenly the default value GL_NEAREST_MIPMAP_LINEAR or the value set via glTexParameteri() was evaluated. If you didn't provide mipmaps and didn't let the driver generate them this led to not sampling your texture data when minification occurred. BUG=angleproject:2772 TEST=angle_end2end_tests.Texture2DTestES3.MinificationWithSamplerNoMipmapping Change-Id: I195147e208129c6fa8686bb0d9aea6931b2f20b1 Reviewed-on: https://chromium-review.googlesource.com/1175910 Commit-Queue: Geoff Lang <geofflang@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill ec1fe5b7 2018-08-10T10:05:52 D3D: Use angle::Result error pattern. This completes the refactor for the D3D9/D3D11 back-ends. Bug: angleproject:2752 Change-Id: Ie35a925f75c902d8f9fdab6bc33b3bb9c937f85b Reviewed-on: https://chromium-review.googlesource.com/1167209 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Jamie Madill 306b6c16 2018-07-27T08:12:49 D3D11: Use angle::Result error pattern. 1/3 This CL improves performance on the draw call microbenchmark by 10% when no-oping driver calls. Bug: angleproject:2738 Change-Id: I4f5c11db90d9056ce4557b2a4432bc55b42b5bba Reviewed-on: https://chromium-review.googlesource.com/1150093 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Yuly Novikov <ynovikov@chromium.org>
Jamie Madill b1565903 2018-07-27T08:12:48 D3D: Pass gl::Context to many more functions. This makes the Context more available for logging errors. Also includes more refactoring to VertexDataManager to ensure we can access the gl::Context. Bug: angleproject:2738 Change-Id: Iae3d22a1403078d236bfe63a3e2d203c13678dc4 Reviewed-on: https://chromium-review.googlesource.com/1151449 Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Yuly Novikov <ynovikov@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Jamie Madill 2160821e 2018-07-24T23:19:43 Add gl::TexLevelArray. This moves a D3D specific type into a shared location. Bug: chromium:867089 Change-Id: Ieac9f1aa4a9de11fbc86550eee30fb242a2cb98e Reviewed-on: https://chromium-review.googlesource.com/1149080 Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Jamie Madill daf2e913 2018-07-24T11:06:48 TextureStorage11: Pass more Context around. This will make it possible to record errors in the context immediately. Bug: angleproject:2378 Change-Id: Ib7ccdbcfa2bf51e3cc876496267c61cf5a26b7bd Reviewed-on: https://chromium-review.googlesource.com/1148464 Reviewed-by: Yuly Novikov <ynovikov@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Jamie Madill 7b4ef3da 2018-07-24T11:06:48 TextureStorage11: Don't use gl::ErrorOrResult. Bug: angleproject:2738 Change-Id: I062b2fe3ff2b53bb361539c408b61626ac6c2f25 Reviewed-on: https://chromium-review.googlesource.com/1148463 Reviewed-by: Yuly Novikov <ynovikov@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Jamie Madill cc129377 2018-04-12T09:13:18 ImageIndex: Consolidate layer/cube face. In terms of the Texture or Image resource, a cube face refers to a layer of a 2D texture. This layer has a special meaning for cube textures, but it is represented as a layer with a layer index. Cube array textures are no different, they just use a different indexing scheme for the array layers. This also cleans up the ImageIndex helper to have a class structure with private data, and cleans up a few cases to use generic Make functions and iterators where they were setting properties of the index directly. This will make it easier to have ImageIndexes address entire levels of a Cube map in the future, and makes the layer count logic in Vulkan cleaner. Bug: angleproject:2318 Change-Id: Iea9842e233f974a9896282ca224cb001f7882bd1 Reviewed-on: https://chromium-review.googlesource.com/987525 Reviewed-by: Luc Ferron <lucferron@chromium.org> Reviewed-by: Yuly Novikov <ynovikov@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Xinghua Cao 26143fdd 2017-11-01T18:19:05 ES31: Support bindImageTexture on Texture2D for compute shaders on D3D BUG=angleproject:1987 TEST=angle_end2end_tests Change-Id: I3b0afb441a41dbd7f204b1d1bba7884c8d203ce1 Reviewed-on: https://chromium-review.googlesource.com/749004 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill acf2f3ad 2017-11-21T19:22:44 Apply Chromium style fixes. This addresses several minor code quality issues that are validated in Chromium, but not yet applied to ANGLE: * constructors and destructors must be defined out-of-line * auto is not allowed for simple pointer types * use override everywhere instead of virtual * virtual functions must also be defined out-of-line Slightly reduces binary size for me (~2k on Win, 150k on Linux). Bug: angleproject:1569 Change-Id: I073ca3365188caf5f29fb28d9eb207903c1843e6 Reviewed-on: https://chromium-review.googlesource.com/779959 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Geoff Lang 92019431 2017-11-20T13:09:34 Make conversion from GL types to native bools consistant. Some places would compare with "== GL_TRUE" and others with "!= GL_FALSE". This behaviour is not in the OpenGL spec but "!= GL_FALSE" is the most standard and follows the same rules as C and C++. Remove un-necessary validation that params are either GL_TRUE or GL_FALSE. Update some internal storage from GLboolean to bool. BUG=angleproject:2258 Change-Id: I12adbe2d24318a206521ca6ad1099ee7e2bf677e Reviewed-on: https://chromium-review.googlesource.com/779799 Reviewed-by: Corentin Wallez <cwallez@chromium.org> Commit-Queue: Geoff Lang <geofflang@chromium.org>
Jamie Madill 05b35b21 2017-10-03T09:01:44 D3D11: Lazy robust resource init. This patch moves the robust resource init logic to the GL front-end. Instead of initializing texture resources immediately on creation in D3D11, it defers the clear until before a draw call in some cases, or skips the update if we can determine if a texture (or other resource) has been fully initialized. Currently lazy init is only implemented for Textures, Renderbuffers, and Surfaces. Various places where lazy resource init is triggered: * Framebuffer operations (Draw, Blit, CopyTexImage, Clear, ReadPixels) * Texture operations (SubImage, GenerateMipmap, CopyTexImage) Some efficiency gains remain to be implemented, such as when a SubImage call fills the entire object. Similarly for Blit, and a few other operations. In these cases we can skip lazy init as an optimization. Edge cases with EGLImage are mostly untested. BUG=angleproject:2107 Change-Id: I2bf3a69b1eae0d4feeb5b17daca23451f1037be8 Reviewed-on: https://chromium-review.googlesource.com/576058 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Jamie Madill ea84f6ff 2017-09-20T13:20:30 Pass Context to Framebuffer11 dirty callback. This allows us to call StateManager11::invalidateRenderTarget from the Framebuffer11::signal function, which will then trigger state refresh on the next draw call. This requires passing Context through a few more Renderbuffer methods, and reorganizing the RenderTarget signalling so that it doesn't signal dirty in the destructor. Instead they are signaled as they are destroyed in the containing classes. BUG=angleproject:2151 Change-Id: I4cf575e4a01b48275ff78d75bc55b2d1fced591d Reviewed-on: https://chromium-review.googlesource.com/673139 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Frank Henigman <fjhenigman@chromium.org>
Jamie Madill 8ec383e6 2017-09-20T13:20:29 Use array template aliases in TextureStorage11. This cleans up some of the iteration logic. BUG=angleproject:2151 Change-Id: I8d80a8d732ee808babbb4859290b648b4fa67b4d Reviewed-on: https://chromium-review.googlesource.com/673138 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Jamie Madill 136a2742 2017-09-20T12:29:44 TextureStorage11: Use unique_ptr for RenderTargets. This ensures they are auto-freed when the Storage is destroyed. Also remove the 'delete this' design pattern. BUG=angleproject:2151 Change-Id: I784b94c7125a1bbc15f5d6ae90e55317e199faa6 Reviewed-on: https://chromium-review.googlesource.com/673137 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Frank Henigman <fjhenigman@chromium.org>
Martin Radev 18b75bad 2017-08-15T15:50:40 D3D11: Handle Clear* commands for layered framebuffers According to the ANGLE_multiview spec, Clear* commands only affect the range of attached layers to the multi-view layered framebuffer. The patch extends ImageIndex so that the range of attached layers is tracked and a render target view can be created with that range of texture array slices attached. The special case of scissored clears for depth and stencil attachments is handled by instancing the same number of quads as there are views and selecting the layer within a geometry shader. BUG=angleproject:2062 TEST=angle_end2end_tests Change-Id: Ibea248b980513f83d918652030a72c62c7ecd88b Reviewed-on: https://chromium-review.googlesource.com/632256 Commit-Queue: Martin Radev <mradev@nvidia.com> Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org>
JiangYizhou 34bc315d 2017-03-29T14:56:01 ES31: Implement multisampled Textures for D3D part. Implement TexStorage2DMultisample api for d3d part. BUG=angleproject:1590 TEST=angle_end2end_tests --gtest_filter=TextureMultisampleTest* TEST=angle_deqp_gles31_tests --deqp-case=dEQP-GLES31.functional.texture.multisample.negative.fbo_* Change-Id: Icbfba45b9c2965af02b54dd4060b7b49970cb74b Reviewed-on: https://chromium-review.googlesource.com/457161 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 4928b7ca 2017-06-20T12:57:39 Proliferate gl::Context everywhere. This gives the D3D back-end access to the GL state almost anywhere. This uses the onDestroy hook for Textures to push errors up from destructors, although they still don't quite make it to the Context. There are places, such as in EGL object (Context/Surface) destruction, where we end up calling through to GL implementation internals without having access to a gl::Context. We handle this via a proxy Context to a Display, basically a null context, that has access to impl-side state like the Renderer pointer if necessary. It does not have access to the normal GL state. Also Pass gl::Context to RefCountObject::release(). Since we're using destroy() methods now, we should not ever call the destructor directly. BUG=angleproject:1156 Change-Id: Ie4c32ad6bf6caaff0289901f30b5c6bafa2ce259 Reviewed-on: https://chromium-review.googlesource.com/529707 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Jamie Madill fe54834f 2017-06-19T11:13:24 Proliferate gl::Context. This enables a few small things: it will enable making the platform a property of the Display rather than a global. The same goes for the global logging annotator. Also it ensures all back-end implementations have access to the GL / EGL state when available. Also introduces a smart pointer helper class to angleutils for objects that prefer to be destroyed with a context (gl::Context/egl::Display) parameter. We were using std::unique_ptr in a few places that would not work well with these objects. BUG=angleproject:1156 Change-Id: I59e288a3d6f766ff8a0f4b48ff3a1fbf7489daba Reviewed-on: https://chromium-review.googlesource.com/529706 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Jamie Madill 0a82f2fa 2017-05-30T15:42:00 D3D11: Don't copy SharedSRVs around. This was causing a very large performance regression in Texture re-binding. Instead pass by const &, which should be fine, since values in std::map (and unordered_map) are not reallocated when the map changes. Also make the SharedResource type non-copyable, and add an explicit makeCopy method for when we need to clone the shared pointer. BUG=angleproject:2034 BUG=chromium:727318 Change-Id: I39508a6ca4b41e4da31fe68899caa4464138cada Reviewed-on: https://chromium-review.googlesource.com/517670 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Jamie Madill 2c479d6e 2017-05-24T14:37:22 D3D11: Use TextureHelper11 everywhere. This consolidates all texture allocation into ResourceManager11. It removes a lot of error checking and resource management code. In a few places we're storing some redundant information, like in TextureStorage11, we might store the Format pointer in two places, or the Texture's size. BUG=angleproject:2034 Change-Id: I9369e76925a67632c444c662e5667c5ed7875547 Reviewed-on: https://chromium-review.googlesource.com/503252 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Jamie Madill e0e009fe 2017-05-19T14:13:03 D3D11: Consolidate SRV allocation. This change also moves all SRV storage to use the custom smart pointer classes in ResourceManager11. Every single SRV is now allocated in ResourceManager11. For TextureStorage11 and RenderTarget11, we use a shared pointer type since the SRVs use shared ownership by both classes. BUG=angleproject:2034 Change-Id: I90666fbbbc23ff10870b7401f3da09ef79258136 Reviewed-on: https://chromium-review.googlesource.com/503250 Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Corentin Wallez <cwallez@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Jamie Madill 403c1682 2017-05-18T11:11:46 D3D11: Add resource manager classes. The resource manager tracks the allocations and deallocations of all D3D11 resources that have device memory: Buffers and Textures (which are either 2D or 3D). It also tracks the number of active Views (DSV, RTV, and SRV, potentially UAV with ES 3.1). A new smart pointer type will wrap the resource deallocation so that the object notifies the manager when ANGLE is done with the resouce. This allows us to track precisely how much GPU memory we think we're using at any point, and will help prevent resource leaks for these object types. It also makes initialization and releasing much more trivial. The base class for a resource uses a template template parameter so that we can use a unique or shared pointer depending on if the object in question needs unique or shared ownership. For some resources (in our case, SRVs are shared between the TextureStorage11 and RenderTarget11 classes, and Textures are shared in many places) we need to have the ability to have shared ownership. Unique ownership is a little bit more efficient so supporting both can be helpful. In this patch RenderTargetView allocation is moved to use the unique smart pointer. BUG=angleproject:2034 Change-Id: Idb1245c24cd66733b8b5ca524c727350b2d2c745 Reviewed-on: https://chromium-review.googlesource.com/503248 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Jiajia Qin f6b986c8 2017-04-18T14:52:31 Refactor TextureStorage11 The CL includes below changes: 1. Change bool isAssociatedImageValid to void verifyAssociatedImageValid since we always required that the validation check should never be false. So ASSERT() is enough. Same to Image11::isAssociatedStorageValid->Image11::verifyAssociatedStorageValid. 2. Remove the unnecessary if checking after ASSERT 3. Use override instead of virtual if the function is virtual and is overriding a virtual function from the base class. BUG=angleproject:2006 Change-Id: I036666ae1ed4bfcaa8cef9e0e9626d375cd81a27 Reviewed-on: https://chromium-review.googlesource.com/480015 Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Corentin Wallez <cwallez@chromium.org> Commit-Queue: Corentin Wallez <cwallez@chromium.org>
Jamie Madill 408293f8 2016-12-20T10:11:45 D3D11: Fix stale SRVs with tiny depth textures. When the tiny depth texture would get re-created on change, it was not freeing any cached SRVs which were pointing to the old tex. Fix this by erasing the cache entry if the texure is recreated. Also include a test which hooks into the workarounds to force testing on every platform. Also narrow the workaround to only apply to depth/stencil textures which use these mips in GL, rather than D3D, where we always create a full mip chain. BUG=angleproject:1664 Change-Id: If0ac396b8847e1bf9b50165e5332da573e9bb3e4 Reviewed-on: https://chromium-review.googlesource.com/421567 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Jamie Madill 9e3d7aa0 2016-09-02T15:19:43 D3D11: Work around small mipped stencil textures. AMD has a bug with 2x2 and 1x1 mips of depth/stencil textures. Formats D24S8 and D32FS8 both seem to render black for the very small mips. We can work around this by selectively caching a copy of the texture data that only has depth information, since stencil isn't filterable in OpenGL ES. BUG=angleproject:1493 BUG=chromium:638323 Change-Id: Iebef42c4680018d9854dbe789d677823167213dd Reviewed-on: https://chromium-review.googlesource.com/380037 Reviewed-by: Corentin Wallez <cwallez@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Jamie Madill dea2777c 2016-09-02T15:09:09 D3D11: Rename TextureStorage swizzle invalidation. Since we'll also be storing a cache of depth textures as well as swizzle info, generalize the invalidation methods. BUG=angleproject:1493 BUG=chromium:638323 Change-Id: I8fef049b01c81483fcf28196866a22d5cfed32a1 Reviewed-on: https://chromium-review.googlesource.com/380036 Reviewed-by: Corentin Wallez <cwallez@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Jamie Madill 8cf70d55 2016-08-09T11:10:35 D3D11: Rename ANGLEFormatSet to Format. BUG=angleproject:1455 Change-Id: I896b3ed2d5e4ff3ad72de9a3a4b554841129e4e9 Reviewed-on: https://chromium-review.googlesource.com/367093 Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Jamie Madill c98cd326 2016-08-09T11:10:33 D3D11: Merge d3d11::TextureFormat and ANGLEFormatSet. BUG=angleproject:1455 Change-Id: I7fc2640ec9d73528feb526b4d8b89cabccda4d17 Reviewed-on: https://chromium-review.googlesource.com/367092 Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Jamie Madill 17b10a9a 2016-08-03T14:10:06 D3D11: Store more format info by-reference. This CL stores the d3d11::TextureFormat by reference in the texture storage. Adding the internalFormat to the TextureFormat allows us to store a single ref instead of three per TextureStorage11. Also store the format sets in a d3d11::TextureFormat by-ref instead of by-pointer, making the code a bit cleaner. BUG=angleproject:1455 Change-Id: I3c0e966d948c694435577d7d45dc0cd156480cdb Reviewed-on: https://chromium-review.googlesource.com/365412 Reviewed-by: Olli Etuaho <oetuaho@nvidia.com> Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Jamie Madill 3416ff3e 2016-08-04T10:13:21 D3D: Check for error before createTextureStorageEGLImage. Previously we were just dumping any unexpected errors here. Fix this by passing in the RT explicitly in the constructor, although we still store the EGLImage pointer in case the RT changes. BUG=angleproject:1455 Change-Id: Ie4d4a07c3a7b3a5ccc0a43c4e80bfc6ee43a676b Reviewed-on: https://chromium-review.googlesource.com/365411 Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Jamie Madill 11b30612 2016-08-02T12:31:58 D3D11: Store ANGLEFormatSets instead of ANGLEFormats. This saves us looking up the FormatSet repeatedly. BUG=angleproject:1455 Change-Id: I77890c1eb427e7d087ceaf194a5001b8b03585d4 Reviewed-on: https://chromium-review.googlesource.com/359084 Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
John Bauman 9585a784 2016-07-08T15:12:34 Clear SRV cache on keyed mutex texture unbind from TextureStorage11_External This fixes the same issue as https://chromium-review.googlesource.com/328400, except with NV12 external storage textures. The SRV cache thinks the texture is still bound, but releasing the keyed mutex has unbound it. BUG=625900,626524 Change-Id: I991cb3eeaaea0a1c4570b88de2ba873ae868ec2a Reviewed-on: https://chromium-review.googlesource.com/359430 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: John Bauman <jbauman@chromium.org>
Olli Etuaho e8528d89 2016-05-16T17:50:52 Fix determining texture mipmap completeness The code didn't previously take base level properly into account when determining how many levels to check when checking for texture completeness. The code is refactored so that the "q" value in spec, that is the maximum mipmap level for determining completeness, can be queried from TextureState. This value is used now for checking completeness. This requires moving ImageDescs under TextureState. Functions that operate on the ImageDesc array are also moved from Texture to TextureState. TextureState members are also renamed to start with the "m" prefix and made private. Also handle levels outside the base/max level range consistently in eglCreateImageKHR validation. We interpret the spec so that if the level used for the EGL image is not a part of the texture levels that affect texture completeness, an error is generated. BUG=angleproject:596 TEST=angle_end2end_tests Change-Id: I038ef24aa83e0a6905ca3c0bbada5989eecb00d9 Reviewed-on: https://chromium-review.googlesource.com/344880 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>
Olli Etuaho 77ae8d57 2016-05-06T14:19:01 Share TextureState structure with TextureImpl This refactoring patch removes the need to pass texture state to methods of TextureImpl in some cases. It also adds target value to TextureState, and moves TextureState definition to Texture.h. The effective base level can now also be queried from TextureState, which reduces the need to pass it around. Two different code paths that dealt with the TextureStorage11 SRV cache are combined into one. Besides refactoring, this patch fixes applying mTopLevel twice when determining the amount of mip levels TextureStorage11. BUG=angleproject:596 TEST=angle_end2end_tests, angle_unittests Change-Id: I1add3d9ad847bec56774e394125156cf9cb0fc2a Reviewed-on: https://chromium-review.googlesource.com/342940 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>
Ian Ewell bda75597 2016-04-18T17:25:54 Finish NV12 support via streams. The main functionality for NV12 texture support through EGL streams has been added. Updates to the compiler, texture code, and stream code were added to support binding to external D3D11 NV12 textures. An end2end test was also added to test sampling of YUV textures and converting to RGB. There is also a new script to convert BMP files to an NV12 texture ready to load into D3D11 for testing purposes. BUG=angleproject:1332 Change-Id: I39b6ec393ea338e2c843fb911acc1b36cd1158a0 Reviewed-on: https://chromium-review.googlesource.com/339454 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Ian Ewell <ewell@google.com> Reviewed-on: https://chromium-review.googlesource.com/341254 Reviewed-by: Ian Ewell <ewell@google.com>
Corentin Wallez 9670b03e 2016-04-29T09:47:47 Revert "Finish NV12 support via streams." Broke Windows Clang compilation, see https://build.chromium.org/p/chromium.gpu.fyi/builders/GPU%20Win%20Clang%20Builder%20%28dbg%29/builds/3583/steps/compile/logs/stdio and search for TextureStorage11.h This reverts commit 9b8b359fa3615be7c7492239a48f61103b2e4fcc. Change-Id: I6e54305eba02b40927a35577594df39e951adb32 Reviewed-on: https://chromium-review.googlesource.com/341430 Reviewed-by: Corentin Wallez <cwallez@chromium.org> Commit-Queue: Corentin Wallez <cwallez@chromium.org>
Ian Ewell 9b8b359f 2016-04-18T17:25:54 Finish NV12 support via streams. The main functionality for NV12 texture support through EGL streams has been added. Updates to the compiler, texture code, and stream code were added to support binding to external D3D11 NV12 textures. An end2end test was also added to test sampling of YUV textures and converting to RGB. There is also a new script to convert BMP files to an NV12 texture ready to load into D3D11 for testing purposes. BUG=angleproject:1332 Change-Id: I098940e6f25e113dcc4fc8d22ffed4b5a16fd860 Reviewed-on: https://chromium-review.googlesource.com/339454 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Ian Ewell <ewell@google.com>
Olli Etuaho 1a4523f3 2016-03-18T15:33:55 Avoid copying of texture format info structures Use const pointers to the statically allocated structures instead of copying them in TextureStorage11. This avoids the cost of copying and saves a little bit of memory. BUG=angleproject:1244 TEST=angle_end2end_tests Change-Id: Ib59fddd68ba9bc53e491d55683416c0661f26e0e Reviewed-on: https://chromium-review.googlesource.com/333930 Reviewed-by: Corentin Wallez <cwallez@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>
Olli Etuaho 7b591905 2016-02-26T14:37:57 D3D11: Use blit SRV format for blits blitSRVFormat stores the format that is used with ANGLE's internal blit shaders. By default, it is the same as the normal SRV format. For integer textures with a red channel, the RTV format is used instead. This makes it possible to change the storage format and the SRV format for the integer textures without affecting the blit format. The blitSRVFormat is used when doing blits in Blit11::copyTexture(). An exception is made for depth/stencil renderbuffer blit - in these cases it is okay to assume that the regular SRV format works for blitting. In the future the regular SRV format for integer textures will be changed to be different from their blit SRV format. TEST=angle_end2end_tests dEQP-GLES3.functional.fbo.blit.* (no regression) dEQP-GLES3.functional.texture.swizzle.* (no regression) BUG=angleproject:1244 Change-Id: Ie0e790e58ec054b64ef5983a09dbfc7754f269ca Reviewed-on: https://chromium-review.googlesource.com/327104 Reviewed-by: Corentin Wallez <cwallez@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>
Olli Etuaho 114d129d 2016-02-26T12:33:00 D3D11: Get color read function from ANGLE format This is done to make it possible to change some of the DXGI formats of integer textures without affecting their color read function associations. The packPixels function gets the ANGLE format from the texture helper, which now requires the ANGLE format to be passed in at the time of creation. BUG=angleproject:1244 TEST=angle_end2end_tests, dEQP-GLES3.functional.fbo.* (no regressions), dEQP-GLES3.functional.pbo.* (all pass) Change-Id: I368337cfe5f8c86ff3292009ccf29e9d01409a07 Reviewed-on: https://chromium-review.googlesource.com/329213 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>
Olli Etuaho bc49758c 2016-02-23T14:43:19 Choose D3D11 swizzle formats in gen_texture_format_table.py Making swizzle format data come from the ANGLE format table will make it easier to make changes that affect both swizzle formats and regular texture formats. Swizzle format is now specified manually for some tricky formats, but for most it can be determined automatically from the ANGLE texture format info. The ANGLE texture format info in texture_format_table.json is changed to facilitate this. The componentType field now captures only whether the data is normalized, int or float and its signedness, but not the width of the data type. Bit widths of the individual channels are recorded in a separate "bits" object for each ANGLE format entry. Also, a new 16-bit RGBA UNORM ANGLE format is added to support swizzling 16-bit normalized depth formats. This change is mostly just refactoring, but it fixes swizzling for formats which have less bits for alpha than other channels: - RGB10_A2 - RGB10_A2UI - RGB5_A1 BUG=angleproject:1322 BUG=angleproject:1244 TEST=angle_end2end_tests, dEQP-GLES3.functional.texture.swizzle.* (all pass) Change-Id: I87d8a9cc0b6569191f50c41754d77b20ca6afef9 Reviewed-on: https://chromium-review.googlesource.com/329074 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>
Olli Etuaho f434906c 2016-02-22T14:53:26 Group D3D11 DXGI format info under a struct This patch refactors how DXGI format info is stored. The goal is to make it easier to make changes that affect both swizzle formats and regular texture formats, and make it easier to pass the format sets around. BUG=angleproject:1244 TEST=angle_end2end_tests Change-Id: I1cc220bccbbdde9200a41829fdc37c8ec123c6a1 Reviewed-on: https://chromium-review.googlesource.com/329072 Reviewed-by: Corentin Wallez <cwallez@chromium.org> Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>
John Bauman 24a504ec 2016-02-18T13:53:00 Clear SRV cache on keyed mutex texture unbind from TextureStorage11 If the calling code releases the keyed mutex after the unbind then D3D will unbind the SRV or RTV from the pipeline without notifying the cache. This allows the SRV to be destroyed, which means that a subsequently-created SRV could have the same address. This would cause the new SRV to not be correctly bound to the pipeline if its slot is the same as the unbound SRV. BUG=588000 Change-Id: I0ad1ba3706a6bcc1fee50745e6148c9b23e578f6 Reviewed-on: https://chromium-review.googlesource.com/328400 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: John Bauman <jbauman@chromium.org>
Geoff Lang 69cce580 2015-09-17T13:20:36 Split the SamplerState struct into SamplerState and TextureState. SamplerState is now only the members that are overridden by a sampler object, this makes it easy to update those specific members. Opted for getters and setters for each member in Texture and Sampler because it will be required to enable dirty bits for these states. Added maxAnisotropy to the SamplerState instead of texture state. The sampler objects extension mentions it should be there. BUG=angleproject:1162 Change-Id: I5aa6d702bd5915ee9df1976afef3c8c1f69d27c8 Reviewed-on: https://chromium-review.googlesource.com/300490 Tryjob-Request: Geoff Lang <geofflang@chromium.org> Reviewed-by: Corentin Wallez <cwallez@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org> Tested-by: Geoff Lang <geofflang@chromium.org>
Geoff Lang 362c0a79 2015-08-12T14:44:38 Implement EGL image classes for the D3D renderers. BUG=angleproject:970 Change-Id: I24d393fcd75dd4ee510785e8475b093868701c77 Reviewed-on: https://chromium-review.googlesource.com/295152 Reviewed-by: Jamie Madill <jmadill@chromium.org> Tryjob-Request: Geoff Lang <geofflang@chromium.org> Tested-by: Geoff Lang <geofflang@chromium.org>
Austin Kinross ba8a0bf8 2015-05-13T09:48:59 Add plumbing for D3D11 device caps BUG=angleproject:1002 Change-Id: Id24783c75377ea92a73a43f2605693e07f63dc15 Reviewed-on: https://chromium-review.googlesource.com/270545 Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Austin Kinross <aukinros@microsoft.com>
Gregoire Payen de La Garanderie 752ce192 2015-04-14T11:11:12 D3D11: Use DX generateMips to generate mipmaps whenever possible. BUG=angleproject:974 Change-Id: I95937fe7a0833de77c52f838ebb3ecba55dfbf8a Reviewed-on: https://chromium-review.googlesource.com/265640 Tested-by: Gregoire Payen de La Garanderie <Gregory.Payen@imgtec.com> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill f4bf3811 2015-04-01T16:15:32 Use GetAs/GetImplAs whenever possible. This patch cleans up the rest of our custom casting helper functions. Change-Id: I41975c736765fca855c4498acca31116df3e8317 Reviewed-on: https://chromium-review.googlesource.com/263477 Tested-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Brandon Jones <bajones@chromium.org>
Jamie Madill f0d10f89 2015-03-31T12:56:52 Replace non-copyable macro with a helper class. This class provides a simpler scheme for blocking default copy and assignment operators. It also reduces the amount of code needed since it's inherited to child classes. This also fixes the conflict between our macro and the same-named macro in Chromium code. BUG=angleproject:956 Change-Id: If0dc72aa3f63fbc7b8fa34907418821c64c39e2f Reviewed-on: https://chromium-review.googlesource.com/263257 Tested-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Brandon Jones <bajones@chromium.org>
Austin Kinross 62815bf4 2015-01-15T16:32:36 Improve D3D11 FL9_3 zero-LOD workaround (e.g. TextureCubes) D3D11 Feature Level 9_3 can't disable mipmaps on a mipmapped texture, and sample from level zero of it. A previous commit added a workaround for this in ANGLE to Texture2Ds. This commit fixes some minor issues in that commit, and extends the workaround to apply to TextureCubes too. Change-Id: Ic97321af6f8bbf7ad5d96e58655c342db3978a6a Reviewed-on: https://chromium-review.googlesource.com/241944 Tested-by: Austin Kinross <aukinros@microsoft.com> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Geoff Lang c2e75afa 2015-01-05T14:26:24 Rename RenderTarget to RenderTargetD3D and move it into the d3d folder. BUG=angle:681 Change-Id: I1946e01ce09d99405c318723c254fe300cc5ac53 Reviewed-on: https://chromium-review.googlesource.com/238471 Reviewed-by: Geoff Lang <geofflang@chromium.org> Tested-by: Geoff Lang <geofflang@chromium.org>
Geoff Lang b4dedf3e 2015-01-05T14:08:53 Merge the Image class into ImageD3D. BUG=angle:681 Change-Id: I0c0d41fb3ff9592b08ede58a8c2ec2bd2d94a0b2 Reviewed-on: https://chromium-review.googlesource.com/238470 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Brandon Jones <bajones@chromium.org> Tested-by: Geoff Lang <geofflang@chromium.org>
Austin Kinross 215b37a6 2014-12-22T12:56:07 Work around zero-LOD mipmap issue on D3D11 Feature Level 9_3 In OpenGL ES, it is possible to sample from level 0 of a mipmapped texture by setting GL_TEXTURE_MIN_FILTER to GL_NEAREST. This is possible in D3D9 and D3D11 Feature Level 10_0+ via various methods. It's not possible in D3D11 Feature Level 9_3, though. This change works around this restriction by creating two copies of each texture on 9_3. The textures are identical, except one has mipmaps and one doesn't. The D3D11 renderer figures out which texture to use at the right time, and keeps the textures in sync with each other as necessary. Note: each texture is only created when it's needed. It's possible that only one (or even neither) D3D texture will be created for a given GL texture. Change-Id: I3c17137b4f63b9158b3abf067ad8e5d5c49d2191 Reviewed-on: https://chromium-review.googlesource.com/234522 Tested-by: Austin Kinross <aukinros@microsoft.com> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Austin Kinross 049743a9 2014-12-23T13:05:11 Restrict depth buffer formats on D3D11 Feature Level 9_3 *_TYPELESS formats weren't supported in D3D10Level9 until Windows 8. Some Win8 D3D9 drivers don't support them either. To workaround this, we avoid _TYPELESS formats on D3D11 FL9_3. BUG=angle:856 BUG=435726 Change-Id: I280dc7f87e3a2c737c14284ebb744188e7f10616 Reviewed-on: https://chromium-review.googlesource.com/237292 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org> Tested-by: Austin Kinross <aukinros@microsoft.com> Tested-by: Jamie Madill <jmadill@chromium.org>
Geoff Lang 0a73dd85 2014-11-19T16:18:08 Fix include guards. BUG=angle:733 Change-Id: I08b2c11c4831f1161c178c1842b10e807185aced Reviewed-on: https://chromium-review.googlesource.com/230831 Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Geoff Lang <geofflang@chromium.org>
Geoff Lang 2b5420c0 2014-11-19T14:20:15 Merge libGLESv2 and libEGL classes into libANGLE. BUG=angle:733 Change-Id: Ic491c971411fe82c56cd97c5c8325ac14ec218df Reviewed-on: https://chromium-review.googlesource.com/230830 Reviewed-by: Geoff Lang <geofflang@chromium.org> Tested-by: Geoff Lang <geofflang@chromium.org>
kmx.io     mail     discord kmxgit v0.4.0