kc3-lang/angle/src/libANGLE/renderer/d3d/ProgramD3D.cpp

src/libANGLE/renderer/d3d/ProgramD3D.cpp

Log

Author Commit Date CI Message
Jamie Madill 53ea9cc6 2016-05-17T10:12:52 Replace rx::Renderer with rx::ContextImpl. Previously Context had no Impl class, but had a special relationship with the instanced Renderer class. Having a ContextImpl backing every Context will allow new designs to enable things like multithreading (where each ContextImpl stores a Context-specific device) or non- virtual Contexts on Android or other platforms where it is more efficient. A large refactoring patch that touches every back-end. BUG=angleproject:1363 Change-Id: Icb73a7d37447f08a664eeb499a310ba05d71a57e Reviewed-on: https://chromium-review.googlesource.com/342052 Reviewed-by: Corentin Wallez <cwallez@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Jamie Madill 48ef11b2 2016-04-27T15:21:52 Rename gl::Framebuffer::Data to gl::FramebufferState. Moving this out of the Framebuffer class allows us to forward- declare it. BUG=angleproject:1363 Change-Id: I91971c37a92151df508cdf7f0eb8c3e93506d112 Reviewed-on: https://chromium-review.googlesource.com/340741 Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Jamie Madill 9082b982 2016-04-27T15:21:51 Rename gl::Data to gl::ContextState. Part of the new world order of renaming the Obj::Data classes to ObjState. BUG=angleproject:1363 Change-Id: I15cf002b8b093d687f540b9e86f045874af24a7e Reviewed-on: https://chromium-review.googlesource.com/340740 Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Olli Etuaho 9696316d 2016-03-21T11:54:33 Support ESSL structs containing samplers on D3D Since HLSL can't natively handle samplers in structs, samplers need to be extracted out of structs into separate variables in the translated shader code. In HLSL 4.1, samplers that were in structs go into the normal sampler arrays and are identified by index constants. In other HLSL versions, samplers that were in structs are translated as uniform variables. These transformations are done inside the HLSL output classes, not as tree transformations. This helps to keep the uniform API provided by the shader translator intact. Wherever a struct containing samplers is passed into a user-defined function, the translated HLSL code passes the separate sampler variables alongside a struct where the samplers have been removed. The D3D backend in libANGLE queries the uniform registers of any samplers that were in uniform structs, and adds them to the register maps, so that correct sampler state gets assigned to them. The extracted sampler variables are prefixed with "angle_" instead of the usual "_" to prevent any name conflicts between them and regular variables. BUG=angleproject:504 TEST=angle_end2end_tests, dEQP-GLES*.functional.shaders.struct.uniform.* (all pass), dEQP-GLES*.functional.uniform_api.* (most now pass) Change-Id: Ib79cba2fa0ff8257a973d70dfd917a64f0ca1efb Reviewed-on: https://chromium-review.googlesource.com/333743 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Corentin Wallez <cwallez@chromium.org> Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>
Jamie Madill 8047c0d2 2016-03-07T13:02:12 D3D11: Clean up InputLayoutCache. This change does a couple things. First, it uses the 'active attribs' mask in the gl::Program to sort the translated attributes, instead of checking the translated attribute themselves. This means we don't have to consult the 'active' field of the translated attributes, which in turns means we don't have to update the active field in the attributes, which breaks the dependency of the attributes on the gl::Program. Second, use a dynamically sized array for storing the cached vertex attributes in the InputLayoutCache. This is nice because it means we don't have to store the size of the array separately. Also some other refactoring cleanups. Refactoring change only. BUG=angleproject:1327 Change-Id: Iab22de92840b30674b92eca72e450673ed9f6d6d Reviewed-on: https://chromium-review.googlesource.com/330172 Reviewed-by: Zhenyao Mo <zmo@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Jamie Madill bdec2f4e 2016-03-02T16:35:32 D3D11: Fix overlapping vertex shader signatures. For the case of drawing with un-normalized integer vertex attributes, we need to do some dynamic conversion in the VS. However, each attribute can either be signed or unsigned, and our shader signature code would treat both as a match, giving rise to a warning in the D3D11 Debug runtime. It's unclear if this would give incorrect results, but it certainly should not produce a warning. BUG=angleproject:1329 Change-Id: I302d11b44e8a0ef981e89c181aefac5451a899b7 Reviewed-on: https://chromium-review.googlesource.com/329998 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Corentin Wallez <cwallez@chromium.org> Reviewed-by: Zhenyao Mo <zmo@chromium.org>
Olli Etuaho 618bebcd 2016-01-15T16:40:00 Pass texture base level to shaders on D3D11 The base level is passed to shaders in an array included in the driver uniform block. This is done on feature levels above 9_3, which treat samplers as indices to sampler arrays in shaders. A separate uniform block couldn't be used for the sampler metadata, since that would bring the number of available uniform blocks down to below minimum level defined by GLES 3.0. BUG=angleproject:596 TEST=angle_end2end_tests Change-Id: Ie040521402f9996d51a978aeeba9222e9dd761ce Reviewed-on: https://chromium-review.googlesource.com/326290 Reviewed-by: Corentin Wallez <cwallez@chromium.org> Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>
Jamie Madill 39046169 2016-02-08T15:05:17 CollectVariables: Don't include block name in field name. The spec mandates that the instance name of a block determines how the active uniform name for this field is reported. However, our handling of this was a bit bugged. We would include the proper prefix on the compiler-side, but this mangled the hashing, and was also not strictly needed. We now also expose the instance name, so we can determine the proper prefix for variable linking on the GL-side of things. This also is consistent with how we handle other spec issues, where the GL-side handles the GL-API specific functionality. This also allows us to fix name hashing of instanced uniform blocks, which was previously broken because we would hash the full name of the active uniform, instead of just the field. BUG=angleproject:1306 Change-Id: I06ace6dbc3f75fdd8129677360dcc142aa89136e Reviewed-on: https://chromium-review.googlesource.com/326681 Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Austin Kinross 2a63b3f8 2016-02-08T12:29:08 Re-land "Implement EGL_experimental_present_path_angle" - Re-land with clang fix. This allows ANGLE to render directly onto a D3D swapchain in the correct orientation when using the D3D11 renderer. The trick is to add an extra uniform to each shader which takes either the value +1.0 or -1.0. When rendering to a texture, ANGLE sets this value to -1.0. When rendering to the default framebuffer, ANGLE sets this value to +1.0. ANGLE multiplies vertex positions by this value in the VS to invert rendering when appropriate. It also corrects other state (e.g. viewport/scissor rect) and shader built-in values (e.g. gl_FragCoord). This saves a substantial amount of GPU time and lowers power consumption. For example, the old method (where ANGLE renders all content onto an offscreen texture, and then copies/inverts this onto the swapchain at eglSwapBuffers() time) uses about 20% of the GPU each frame on a Lumia 630. Verification: + dEQP GL ES2 tests pass when "present path fast" is enabled + all ANGLE_end2end_tests pass when "present path fast" is enabled BUG=angleproject:1219 Change-Id: I56b339897828753a616d7bae837a2f354dba9c63 Reviewed-on: https://chromium-review.googlesource.com/326730 Tryjob-Request: Austin Kinross <aukinros@microsoft.com> Reviewed-by: Corentin Wallez <cwallez@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Jamie Madill d1c46228 2016-02-08T14:51:18 Revert "Implement EGL_experimental_present_path_angle" Compile failure on Clang/Win: The reason for reverting is: FAILED: ninja -t msvc -e environment.x86 -- "..\..\third_party/llvm-build/Release+Asserts/bin/clang-cl" -m32 /nologo /showIncludes /FC @obj\third_party\angle\src\tests\egl_tests\angle_end2end_tests.EGLPresentPathD3D11Test.obj.rsp /c ..\..\third_party\angle\src\tests\egl_tests\EGLPresentPathD3D11Test.cpp /Foobj\third_party\angle\src\tests\egl_tests\angle_end2end_tests.EGLPresentPathD3D11Test.obj /Fdobj\gpu\angle_end2end_tests.cc.pdb In file included from ..\..\third_party\angle\src\tests\egl_tests\EGLPresentPathD3D11Test.cpp:7: In file included from ..\..\third_party\angle\src\tests\test_utils/ANGLETest.h:13: ..\..\testing\gtest\include\gtest/gtest.h(1392,16) : error: comparison of integers of different signs: 'const int' and 'const unsigned int' [-Werror,-Wsign-compare] if (expected == actual) { ~~~~~~~~ ^ ~~~~~~ ..\..\testing\gtest\include\gtest/gtest.h(1422,12) : note: in instantiation of function template specialization 'testing::internal::CmpHelperEQ<int, unsigned int>' requested here return CmpHelperEQ(expected_expression, actual_expression, expected, ^ ..\..\third_party\angle\src\tests\egl_tests\EGLPresentPathD3D11Test.cpp(281,9) : note: in instantiation of function template specialization 'testing::internal::EqHelper<false>::Compare<int, unsigned int>' requested here ASSERT_EQ(mWindowWidth * 4, mappedSubresource.RowPitch); ^ ..\..\testing\gtest\include\gtest/gtest.h(1960,32) : note: expanded from macro 'ASSERT_EQ' # define ASSERT_EQ(val1, val2) GTEST_ASSERT_EQ(val1, val2) ^ ..\..\testing\gtest\include\gtest/gtest.h(1943,67) : note: expanded from macro 'GTEST_ASSERT_EQ' EqHelper<GTEST_IS_NULL_LITERAL_(expected)>::Compare, \ ^ BUG=angleproject:1219 This reverts commit 6b3c1db5170450bbc4946d8f18ba0d8619da43a0. Change-Id: Ia67ab82dd13295dc03235d57fa417c73f20a49e6 Reviewed-on: https://chromium-review.googlesource.com/326680 Reviewed-by: Jamie Madill <jmadill@chromium.org>
Austin Kinross 6b3c1db5 2015-12-18T14:01:46 Implement EGL_experimental_present_path_angle This allows ANGLE to render directly onto a D3D swapchain in the correct orientation when using the D3D11 renderer. The trick is to add an extra uniform to each shader which takes either the value +1.0 or -1.0. When rendering to a texture, ANGLE sets this value to -1.0. When rendering to the default framebuffer, ANGLE sets this value to +1.0. ANGLE multiplies vertex positions by this value in the VS to invert rendering when appropriate. It also corrects other state (e.g. viewport/scissor rect) and shader built-in values (e.g. gl_FragCoord). This saves a substantial amount of GPU time and lowers power consumption. For example, the old method (where ANGLE renders all content onto an offscreen texture, and then copies/inverts this onto the swapchain at eglSwapBuffers() time) uses about 20% of the GPU each frame on a Lumia 630. Verification: + dEQP GL ES2 tests pass when "present path fast" is enabled + all ANGLE_end2end_tests pass when "present path fast" is enabled BUG=angleproject:1219 Change-Id: Ib6eeea46bafa6ebce4adada0ae9db3a433b8fc4c Reviewed-on: https://chromium-review.googlesource.com/321360 Reviewed-by: Geoff Lang <geofflang@chromium.org> Tryjob-Request: Austin Kinross <aukinros@microsoft.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Geoff Lang <geofflang@chromium.org>
Geoff Lang 00140f4a 2016-02-03T18:47:33 Revert "Pass texture base level to shaders on D3D11" This reverts commit 3026829e155bd89b5ca5b7b7c5267699b9192557. Change-Id: I0b4c3f5b1453b993b149423bb1ce407f4918cf54 Reviewed-on: https://chromium-review.googlesource.com/325435 Reviewed-by: Geoff Lang <geofflang@chromium.org>
Olli Etuaho 3026829e 2016-01-15T16:40:00 Pass texture base level to shaders on D3D11 The base level is passed to shaders in a uniform block created specifically for passing sampler metadata. This is done on feature levels above 9_3, which treat samplers as indices to sampler arrays in shaders. BUG=angleproject:596 TEST=angle_end2end_tests Change-Id: I846f2fc195ab1fd884052824ffd3c1d65083c0fb Reviewed-on: https://chromium-review.googlesource.com/322122 Reviewed-by: Zhenyao Mo <zmo@chromium.org> Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>
Jamie Madill 60ec6ea7 2016-01-22T15:27:19 Implement dirty bits for Framebuffer. The dirty bits set the stage for performance improvements in D3D, but don't actually reduce any of the redundant work just yet. BUG=angleproject:1260 Change-Id: Ib84e6a9b7aa40c37c41790f492361b22faaf4742 Reviewed-on: https://chromium-review.googlesource.com/318730 Reviewed-by: Geoff Lang <geofflang@chromium.org> Tryjob-Request: Jamie Madill <jmadill@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Austin Kinross 88829e84 2016-01-12T13:04:41 Don't compile D3D geometry shader when using instanced point emulation If ANGLE is using instanced point sprites (e.g. on Feature Level 9_3) then it still tries to compile the point sprite geometry shader. It shouldn't do this. Change-Id: I6e7753dd337c385893fc4bea6c8438aeaf5e8427 Reviewed-on: https://chromium-review.googlesource.com/321397 Tryjob-Request: Austin Kinross <aukinros@microsoft.com> Tested-by: Austin Kinross <aukinros@microsoft.com> Reviewed-by: Corentin Wallez <cwallez@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Olli Etuaho e5df3066 2015-11-18T13:45:19 D3D: Run applyTextures before applyShaders applyShaders will update uniform data by uploading the constant buffer. In the future uniform data will contain the base level of textures so that shaders can use it to adjust results of texture functions. The most natural place to set the base level in uniform data is applyTextures, so it needs to be run before applyShaders. updateSamplerMapping needs to be called before applyTextures, but genericDraw* functions already do that. Remove redundant updateSamplerMapping call from ProgramD3D::applyUniforms. BUG=angleproject:596 BUG=angleproject:1092 BUG=angleproject:1261 TEST=angle_end2end_tests Change-Id: I1828f7bdc75999872a26941a02cb4b6c6a66094d Reviewed-on: https://chromium-review.googlesource.com/322121 Reviewed-by: Jamie Madill <jmadill@chromium.org> Tryjob-Request: Olli Etuaho <oetuaho@nvidia.com> Tested-by: Olli Etuaho <oetuaho@nvidia.com> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Geoff Lang c5629759 2015-12-07T16:29:04 Implement full program binary support for ES3. Refactor validation to be used in both the OES and ES3 entry points. BUG=angleproject:600 BUG=angleproject:1101 Change-Id: I2008c4ea04ce07910f03ae0b997f8a77b66203d8 Reviewed-on: https://chromium-review.googlesource.com/316620 Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Geoff Lang <geofflang@chromium.org>
Jacek Caban 2302e69a 2015-12-01T12:44:43 Fixed compilation with mingw. Change-Id: I8102ce4dd934ac247a29e31140d3a1eca691802f Reviewed-on: https://chromium-review.googlesource.com/315060 Reviewed-by: Geoff Lang <geofflang@chromium.org> Tested-by: Geoff Lang <geofflang@chromium.org>
Jamie Madill 6328667d 2015-11-24T13:00:08 Add gl_FragDepth support for D3D11/ES3. An incorrect check was blocking existing support we added for the extension. BUG=angleproject:1226 Change-Id: I335053b1bebb37fd9296b36702031bd34c1bb59e Reviewed-on: https://chromium-review.googlesource.com/313992 Tryjob-Request: Jamie Madill <jmadill@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Corentin Wallez <cwallez@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 847638a6 2015-11-20T13:01:41 Re-land "D3D: Generate more shader debug info by default." Re-land with fix for shader size query. The debug info can be very useful to WebGL developers. Instead of hiding the feature behind an unsupported and broken flag, always enable it so that WebGL devs can access the useful translated info. Change the debug info policy to build the full info (including generated assembly) in Debug, and all info excepth the assembly in Release. BUG=angleproject:1179 Change-Id: I812b2c9ba98cb8c9d5ec95721441c70e8990b626 Reviewed-on: https://chromium-review.googlesource.com/313600 Tryjob-Request: Jamie Madill <jmadill@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 57d116a4 2015-11-20T17:59:18 Revert "D3D: Generate more shader debug info by default." Has a bug where we would return the wrong size for the query. BUG=angleproject:1179 This reverts commit 6de4d49477986339a92b542a18df22c49d10cf14. Change-Id: Ib58aaf0b58d8f20765f7267b29acd8c62d287525 Reviewed-on: https://chromium-review.googlesource.com/313552 Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 6de4d494 2015-11-19T14:08:34 D3D: Generate more shader debug info by default. The debug info can be very useful to WebGL developers. Instead of hiding the feature behind an unsupported and broken flag, always enable it so that WebGL devs can access the useful translated info. Change the debug info policy to build the full info (including generated assembly) in Debug, and all info excepth the assembly in Release. BUG=angleproject:1179 Change-Id: I25656b76b8437cb2edc9d88cd33f2c4d19b6a3f0 Reviewed-on: https://chromium-review.googlesource.com/313160 Tryjob-Request: Jamie Madill <jmadill@chromium.org> Reviewed-by: Kenneth Russell <kbr@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Jamie Madill 55c25d0c 2015-11-18T13:08:08 D3D11: Fix varying packing with structs. Previously we would try to pass an entire struct through HLSL's shader interface. Instead, split this off as if each field was its own variable, which seems to be spec compliant (see ESSL 3.10). In the future we may want to fix register packing to use specific components of float4/int4/uint4 HLSL registers. This could also fix the remaining bugs in the SM3 packing. TEST=dEQP-GLES3.functional.shaders.linkage.varying.* BUG=angleproject:910 BUG=angleproject:1202 Change-Id: I1fd8b4505abc39bd2385ed5c088c316d55d0bc2c Reviewed-on: https://chromium-review.googlesource.com/311242 Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Corentin Wallez <cwallez@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 9fc3682c 2015-11-18T13:08:07 D3D: Rework varying packing code. In D3D we pack varyings by making a register map, and using the recommended GLSL ES algorithm to reserve register space. We use this map to assign row and column slots to each varying and then produce a semantic index value. The existing scheme had a number of bugs, and was failing several angle_end2end_tests. The new design cleans up the code somewhat and uses a different counting scheme for the semantic indexes: just sort the varyings in packing order and use a simple incrementing semantic index per varying. In SM4+, the HLSL compiler sorts and packs the varyings correctly itself, and in SM3, handle the cases we don't support by returning an error instead of a D3D compiler link error. Also refactor how we store varying information for TF Feedback/ StreamOut. Only store the necessary D3D information, instead of extra information like the name and type. This fixes several tests in GLSLTest/*. This also will allow us to fix interpolation qualifier packing and the structure packing in HLSL, which seems to work differently than the rest of the varying types. BUG=angleproject:1202 TEST=bots,dEQP-GLES3.functional.transform_feedback.* Change-Id: Ie5bfbb4f71d8bf97f39115fc46d2e61b131df639 Reviewed-on: https://chromium-review.googlesource.com/311241 Reviewed-by: Geoff Lang <geofflang@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill e39a3f0a 2015-11-17T20:42:15 ProgramD3D: Move some common code into a Metadata class. This metadata class captures some of our commonly referenced but also complex flats into a shared place. We can then re-use them in the semantic code, the DynamicHLSL linking code, and the program code. Refactoring patch only. BUG=angleproject:1202 Change-Id: I8b6088cfa5488c5173a6f06c15abab5a4ead4cb8 Reviewed-on: https://chromium-review.googlesource.com/311700 Tryjob-Request: Jamie Madill <jmadill@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Jamie Madill 65345da4 2015-11-13T11:25:23 Re-land "D3D: Move some varying packing code into a new file." Re-land with fix for compile errors. Refactoring patch which moves code only. Work and full description will follow in a subsequent CL. BUG=angleproject:1202 Change-Id: Ib7c7dbd32491cdd9512dbecdc0e1e90303d8f28d Reviewed-on: https://chromium-review.googlesource.com/312481 Tryjob-Request: Jamie Madill <jmadill@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Jamie Madill bbdeeb18 2015-11-12T15:42:16 Revert "D3D: Move some varying packing code into a new file." Compile error: c:\b\build\slave\gpu_win_builder\build\src\third_party\angle\src\libangle\renderer\d3d\dynamichlsl.cpp(508) : error C3861: 'getSemanticInfo': identifier not found c:\b\build\slave\gpu_win_builder\build\src\third_party\angle\src\libangle\renderer\d3d\dynamichlsl.cpp(642) : error C3861: 'getSemanticInfo': identifier not found BUG=angleproject:1202 This reverts commit d992cde27ea6ffd71205399c3ddb88c7085c99b4. Change-Id: I11f34d2ac4cc047c397e45348cb9502676380be0 Reviewed-on: https://chromium-review.googlesource.com/312480 Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill d992cde2 2015-11-10T14:42:08 D3D: Move some varying packing code into a new file. Refactoring patch which moves code only. Work and full description will follow in a subsequent CL. BUG=angleproject:1202 Change-Id: I60346c516209a0517fc2bac38e0d04ff56bcbc3c Reviewed-on: https://chromium-review.googlesource.com/311260 Reviewed-by: Corentin Wallez <cwallez@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 28afae5d 2015-11-09T15:07:57 Rename LinkedVarying to D3DVarying. Also move this type to D3D-only world. It was only used in the D3D renderer and has specific stuff like register/semantic indexes. Refactoring patch to clean up further work with Varyings. BUG=angleproject:1202 Change-Id: I4b1d6899e9eef356efc7d11e9cd6cf88b234aa76 Reviewed-on: https://chromium-review.googlesource.com/311240 Reviewed-by: Geoff Lang <geofflang@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 3e14e2b1 2015-10-29T14:38:53 D3D11: Fix basic provoking vertex flat shading cases. The enables geometry shaders that correct for the flat shading on provoking vertexes. It does not fix it for triangle strips, or in conjunction with primitive restart (which is not yet implemented in D3D11). Also ensure we do not regress with flat shading enabled and transform feedback. In cases where we use flat shading, do a double draw call, first with an untransformed vertex stream, and no geometry shader, then with the geometry shader enabled. This also fixes the dEQP fragment output tests with ints. BUG=angleproject:754 Change-Id: Ib37e8ec32d19db17ea5d4dc88158cdae0c956bfc Reviewed-on: https://chromium-review.googlesource.com/309155 Tryjob-Request: Jamie Madill <jmadill@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Corentin Wallez <cwallez@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 4e31ad55 2015-10-29T10:32:57 D3D11: Add dynamic geometry shaders. The geometry shader we want will depend on our current draw mode, and if we're using flat shading in the shader. Without flat shading, we'll still be using them only for point sprites, but for other primitive types with flat shading enabled, we'll be using them to correct the provoking vertex order with D3D11. Note: no new features in this CL, those are turned on in follow-ups. BUG=angleproject:754 Change-Id: Iabf13ffd582f5a7200ee0df5aa9c3671aa7b6ed4 Reviewed-on: https://chromium-review.googlesource.com/309154 Reviewed-by: Geoff Lang <geofflang@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 76f8fa66 2015-10-29T10:32:56 Refactor Geometry Shader support to enable pass-through. This should be a refactoring change only. The new code is exercised in follow-up CLs. BUG=angleproject:754 Change-Id: I99285e1e7772cae467013102f25c911ebc9f54a9 Reviewed-on: https://chromium-review.googlesource.com/309153 Reviewed-by: Geoff Lang <geofflang@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 334d615f 2015-10-22T14:00:28 Fix up the style in DynamicHLSL a bit. Use std::stringstream when possible, and use Context's caps instead of the Renderer's. BUG=angleproject:754 Change-Id: I2dc773709bbd612ab7ea372a358337c0a81869a3 Reviewed-on: https://chromium-review.googlesource.com/307872 Reviewed-by: Corentin Wallez <cwallez@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 4a3c2341 2015-10-08T12:58:45 Program: Clean up UBO info gathering. The data size & block member info was getting messy, so clean up how we query this from the Impl layer. Also remove the register information from gl::UniformBlock, moving it into the D3D-only world. BUG=angleproject:1172 Change-Id: I40af658ebbd6b7c1a4251906a387ebcbb621cf77 Reviewed-on: https://chromium-review.googlesource.com/304150 Reviewed-by: Corentin Wallez <cwallez@chromium.org> Tryjob-Request: Jamie Madill <jmadill@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Geoff Lang 2b280922 2015-10-01T17:05:15 Remove the unused ProgramImpl::getShaderVersion method. BUG=angleproject:882 BUG=angleproject:1123 Change-Id: Iace1d6c2c594f0b00c18d595f14cef9bce6f3595 Reviewed-on: https://chromium-review.googlesource.com/303826 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Corentin Wallez <cwallez@chromium.org> Tested-by: Geoff Lang <geofflang@chromium.org>
Geoff Lang 5d124a69 2015-09-15T13:03:27 Support UBOs in RendererGL and ProgramGL. Reland: fix compile warnings on Windows. BUG=angleproject:882 BUG=angleproject:883 Change-Id: If4844d50d43ff9b765f64606e8f08d65a1afd9a4 Reviewed-on: https://chromium-review.googlesource.com/303431 Tryjob-Request: Geoff Lang <geofflang@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org> Tested-by: Geoff Lang <geofflang@chromium.org>
Geoff Lang cf2dbd91 2015-10-01T15:48:22 Revert "Support UBOs in RendererGL and ProgramGL." Compile error. This reverts commit 97d82b724b71be464f2aff377f056cade44fca94. Change-Id: I71ae10cb01b8ded0297eaf4742fb1098ce02a9bb Reviewed-on: https://chromium-review.googlesource.com/303303 Reviewed-by: Geoff Lang <geofflang@chromium.org> Tested-by: Geoff Lang <geofflang@chromium.org>
Geoff Lang 97d82b72 2015-09-15T13:03:27 Support UBOs in RendererGL and ProgramGL. BUG=angleproject:882 BUG=angleproject:883 Change-Id: I36f8ef42d87e289658a6ba4899380bc72b9bcebf Reviewed-on: https://chromium-review.googlesource.com/299871 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Corentin Wallez <cwallez@chromium.org> Tested-by: Geoff Lang <geofflang@chromium.org>
Jamie Madill 91445bce 2015-09-23T16:47:53 Make a shader Shader::Data state structure. This design follows the similar designs for Program, Framebuffer, etc. Because of the current design, share a mutable pointer with the Impl so the patch becomes a bit smaller and easier to review. In a follow- up patch we can move the shared code into the GL layer. BUG=angleproject:1159 Change-Id: Ib243e74779f23be51cdca80f1b5c6e5f3e36059d Reviewed-on: https://chromium-review.googlesource.com/299876 Reviewed-by: Geoff Lang <geofflang@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 3d3d2f20 2015-09-23T16:47:51 Re-re-land "Move sampler validation to the GL layer." This previously was D3D-only, but is required for every draw call. This completes the work of removing the D3D-specific Impl methods from ProgramImpl. Also add several regression tests to cover texture and sampler validation. Re-land with a fix for duplicate sampler active uniforms. Re-re-land with a fix for a test comparison warning on Linux. BUG=angleproject:1123 Change-Id: Iaf7b33861c07b9ceed4bd53ac2f010d35f05df45 Reviewed-on: https://chromium-review.googlesource.com/301712 Tryjob-Request: Jamie Madill <jmadill@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 6d246510 2015-09-23T18:04:10 Revert "Re-land "Move sampler validation to the GL layer."" Build failure on Linux: In file included from ../../third_party/angle/src/tests/gl_tests/UniformTest.cpp:7: In file included from ../../third_party/angle/src/tests/test_utils/ANGLETest.h:10: ../../testing/gtest/include/gtest/gtest.h:1392:16: error: comparison of integers of different signs: 'const int' and 'const unsigned int' [-Werror,-Wsign-compare] if (expected == actual) { ~~~~~~~~ ^ ~~~~~~ ../../testing/gtest/include/gtest/gtest.h:1422:12: note: in instantiation of function template specialization 'testing::internal::CmpHelperEQ<int, unsigned int>' requested here return CmpHelperEQ(expected_expression, actual_expression, expected, ^ ../../third_party/angle/src/tests/gl_tests/UniformTest.cpp:487:5: note: in instantiation of function template specialization 'testing::internal::EqHelper<false>::Compare<int, unsigned int>' requested here EXPECT_EQ(GL_SAMPLER_2D, type); ^ BUG=angleproject:1123 This reverts commit 6cbf4385280c4c1dd97f0882ecb18dbc4c341fd4. Change-Id: I95279b37d253e3ea78faa53f3773f2dc3d17df95 Reviewed-on: https://chromium-review.googlesource.com/302030 Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 6cbf4385 2015-09-22T19:12:11 Re-land "Move sampler validation to the GL layer." This previously was D3D-only, but is required for every draw call. This completes the work of removing the D3D-specific Impl methods from ProgramImpl. Also add several regression tests to cover texture and sampler validation. Re-land with a fix for duplicate sampler active uniforms. BUG=angleproject:1123 Change-Id: Iefef06e7901873c98bf2ba7864efd16a4c6435d3 Reviewed-on: https://chromium-review.googlesource.com/301581 Tryjob-Request: Jamie Madill <jmadill@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Corentin Wallez <cwallez@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Jamie Madill 3aa23a7c 2015-09-22T19:03:33 Revert "Move sampler validation to the GL layer." Probably causing failures in the dEQP-GLES2 GL back-end tests: dEQP-GLES2.functional.uniform_api.info_query.basic.sampler2D_both BUG=angleproject:1123 This reverts commit 6fa156b66da2090c98f1a044a948694c7ac83324. Change-Id: I3cc993c34aae4a108037ce81709b71d16ece814e Reviewed-on: https://chromium-review.googlesource.com/301580 Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 6fa156b6 2015-09-22T10:17:01 Move sampler validation to the GL layer. This previously was D3D-only, but is required for every draw call. This completes the work of removing the D3D-specific Impl methods from ProgramImpl. Also add several regression tests to cover texture and sampler validation. BUG=angleproject:1123 Change-Id: I167cbc8f47b4a397d2ae6cc507bdac35168bcacd Reviewed-on: https://chromium-review.googlesource.com/299400 Reviewed-by: Geoff Lang <geofflang@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Dian Xiang a492883b 2015-09-15T10:11:17 Optimizing updatecachedInputLayout to not update inactive vertex attribs BUG=angleproject:1158 part a We only need to iterate and update the active attributes in a program. This can further be optimized by updating the cache more intelligently, possibly by marking attributes as "need update" when necessary Change-Id: I41e9c033116582c82a56b8d20ea245f06e8fa3d1 Reviewed-on: https://chromium-review.googlesource.com/299758 Reviewed-by: Jamie Madill <jmadill@chromium.org> Tryjob-Request: Jamie Madill <jmadill@chromium.org> Reviewed-by: Zhenyao Mo <zmo@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Geoff Lang 98c56da4 2015-09-15T15:45:34 Fix warning about out-of-order field initialization. Change-Id: I38a3f93ea07c9bbd779d18bf7d3e2a5a6ee4113a Reviewed-on: https://chromium-review.googlesource.com/299772 Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Geoff Lang <geofflang@chromium.org> Tested-by: Nico Weber <thakis@chromium.org>
Jamie Madill 62d31cb6 2015-09-11T13:25:51 Re^6-land "Move Uniform and UBO info to the gl::Program layer." This data was previously stored entirely in the Impl level. Move as much as possible to the GL level, using a read-only view in the Impl level. Some information in D3D-specific, and should be stored separately in the Impl. This patch has a lot of refactoring that splits the D3D and GL info, and moves as much validation as possible to the GL layer, where it is shared between the back-ends. Re-land with fix for dEQP unused uniforms. The fix involves storing a local copy of all uniform data in the GL layer. This will also let us validate sampler indexes during draw calls at the GL layer. Re-re-land with a fix for multiply defined symbols on Clang. Re-re-re-land with a fix for boolean uniforms and Uniform{1234}f. Re^4-land with a fix for boolean uniform arrays and UBO uniforms. Re^5-land with a fix for a test warning on Linux. Re^6-land with a fix for transposed matrix uniform arrays. BUG=angleproject:1123 TEST=end2end_tests, bots, dEQP GLES3.ubo and GLES2.uniform_api Change-Id: Ie6fcde1c16eb05d67191b629338b88302a2563f5 Reviewed-on: https://chromium-review.googlesource.com/298971 Tryjob-Request: Jamie Madill <jmadill@chromium.org> Reviewed-by: Corentin Wallez <cwallez@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill fb53603c 2015-09-11T13:19:49 Revert "Re^5-land "Move Uniform and UBO info to the gl::Program layer."" Failing dEQP-GLES3.functional.uniform_api.random.22 and 23: There's a bug with arrays of tranpsosed matrix uniforms. BUG=angleproject:1123 This reverts commit 78d356929bd0441d81e2dd8a63130cd6788e2fde. Change-Id: If39b5908af39671dfe98965e6a1ba77fd18ea8fc Reviewed-on: https://chromium-review.googlesource.com/299320 Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 78d35692 2015-09-11T08:32:51 Re^5-land "Move Uniform and UBO info to the gl::Program layer." This data was previously stored entirely in the Impl level. Move as much as possible to the GL level, using a read-only view in the Impl level. Some information in D3D-specific, and should be stored separately in the Impl. This patch has a lot of refactoring that splits the D3D and GL info, and moves as much validation as possible to the GL layer, where it is shared between the back-ends. Re-land with fix for dEQP unused uniforms. The fix involves storing a local copy of all uniform data in the GL layer. This will also let us validate sampler indexes during draw calls at the GL layer. Re-re-land with a fix for multiply defined symbols on Clang. Re-re-re-land with a fix for boolean uniforms and Uniform{1234}f. Re^4-land with a fix for boolean uniform arrays and UBO uniforms. Re^5-land with a fix for a test warning on Linux. BUG=angleproject:1123 TEST=end2end_tests, bots, dEQP GLES3.ubo and GLES2.uniform_api Change-Id: I03afcd446d9e74573c4d4c712ed7407b91d8001c Reviewed-on: https://chromium-review.googlesource.com/299180 Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill a2eb02c9 2015-09-11T12:31:41 Revert "Re^4-land "Move Uniform and UBO info to the gl::Program layer."" Test has a warning on Linux: In file included from ../../third_party/angle/src/tests/test_utils/ANGLETest.h:13: ../../testing/gtest/include/gtest/gtest.h:1392:16: error: comparison of integers of different signs: 'const int' and 'const unsigned int' [-Werror,-Wsign-compare] if (expected == actual) { ~~~~~~~~ ^ ~~~~~~ ../../testing/gtest/include/gtest/gtest.h:1422:12: note: in instantiation of function template specialization 'testing::internal::CmpHelperEQ<int, unsigned int>' requested here return CmpHelperEQ(expected_expression, actual_expression, expected, ^ ../../third_party/angle/src/tests/gl_tests/UniformBufferTest.cpp:357:5: note: in instantiation of function template specialization 'testing::internal::EqHelper<false>::Compare<int, unsigned int>' requested here EXPECT_EQ(GL_FLOAT, type); ^ BUG=angleproject:1123 This reverts commit 2ae1ee6161c24aab065190b1528f6195e80e34d0. Change-Id: Ic3dabea54068377d1f2c5f722ba52966cf1a0491 Reviewed-on: https://chromium-review.googlesource.com/299170 Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 2ae1ee61 2015-09-10T10:04:24 Re^4-land "Move Uniform and UBO info to the gl::Program layer." This data was previously stored entirely in the Impl level. Move as much as possible to the GL level, using a read-only view in the Impl level. Some information in D3D-specific, and should be stored separately in the Impl. This patch has a lot of refactoring that splits the D3D and GL info, and moves as much validation as possible to the GL layer, where it is shared between the back-ends. Re-land with fix for dEQP unused uniforms. The fix involves storing a local copy of all uniform data in the GL layer. This will also let us validate sampler indexes during draw calls at the GL layer. Re-re-land with a fix for multiply defined symbols on Clang. Re-re-re-land with a fix for boolean uniforms and Uniform{1234}f. Re^4-land with a fix for boolean uniform arrays and UBO uniforms. BUG=angleproject:1123 TEST=end2end_tests, bots, dEQP GLES3.ubo and GLES2.uniform_api Change-Id: I4c9f5ed31b81380507bef7981f97086d642801ae Reviewed-on: https://chromium-review.googlesource.com/298451 Tryjob-Request: Jamie Madill <jmadill@chromium.org> Reviewed-by: Corentin Wallez <cwallez@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 61b8dd97 2015-09-09T19:04:04 Revert "Re-re-re-land "Move Uniform and UBO info to the gl::Program layer."" Seems to be failing the dEQP-GLES2.functional.uniform_api tests relating to boolean arrays. BUG=angleproject:1123 This reverts commit 892a6a4b17b35e89898be8b4605c1ee595d3ae13. Change-Id: Ifc4ce3ca806ef88a02ac6f693334ac37ce098a88 Reviewed-on: https://chromium-review.googlesource.com/298520 Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 892a6a4b 2015-09-09T10:17:15 Re-re-re-land "Move Uniform and UBO info to the gl::Program layer." This data was previously stored entirely in the Impl level. Move as much as possible to the GL level, using a read-only view in the Impl level. Some information in D3D-specific, and should be stored separately in the Impl. This patch has a lot of refactoring that splits the D3D and GL info, and moves as much validation as possible to the GL layer, where it is shared between the back-ends. Re-land with fix for dEQP unused uniforms. The fix involves storing a local copy of all uniform data in the GL layer. This will also let us validate sampler indexes during draw calls at the GL layer. Re-re-land with a fix for multiply defined symbols on Clang. Re-re-re-land with a fix for boolean uniforms and Uniform{1234}f. BUG=angleproject:1123 TEST=end2end_tests,Canary WebGL (D3D11/GL), dEQP-GLES2+3.functional.uniform_api Change-Id: Ia40820a5ce2f34ec2d27648b1dc940a8955e9999 Reviewed-on: https://chromium-review.googlesource.com/298440 Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Corentin Wallez <cwallez@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill b654f659 2015-09-08T20:40:00 Revert "Re-re-land "Move Uniform and UBO info to the gl::Program layer."" Failes the gles2_conform_test: GLES2ConformTest.GL2Tests_glGetUniform_input_run Possibly also WebGL failures, will investigate. BUG=angleproject:1123 This reverts commit 10750cb936288d8dd09d49fadd592904c06c56f9. Change-Id: I1ae59325e1831589019bc5a27ffc2091d3994a65 Reviewed-on: https://chromium-review.googlesource.com/298200 Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 10750cb9 2015-09-04T14:23:52 Re-re-land "Move Uniform and UBO info to the gl::Program layer." This data was previously stored entirely in the Impl level. Move as much as possible to the GL level, using a read-only view in the Impl level. Some information in D3D-specific, and should be stored separately in the Impl. This patch has a lot of refactoring that splits the D3D and GL info, and moves as much validation as possible to the GL layer, where it is shared between the back-ends. Re-land with fix for dEQP unused uniforms. The fix involves storing a local copy of all uniform data in the GL layer. This will also let us validate sampler indexes during draw calls at the GL layer. Re-re-land with a fix for multiply defined symbols on Clang. BUG=angleproject:1123 TEST=end2end_tests,Canary WebGL (D3D11/GL), dEQP-GLES2+3.functional.uniform_api Change-Id: I7b55e4964fbe64e0106c37928c894146937a2476 Reviewed-on: https://chromium-review.googlesource.com/298110 Reviewed-by: Corentin Wallez <cwallez@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 2119dd05 2015-09-08T19:30:37 Revert "Re-land "Move Uniform and UBO info to the gl::Program layer."" Problems with multiply defined symbols in the Linux builder. BUG=angleproject:1123 This reverts commit ff1abe63c3d8af28e9755c8d0c95b2aeca27a0d9. Change-Id: Ifcb2ad885d0841c128ffd69afb6594e37eba17fc Reviewed-on: https://chromium-review.googlesource.com/298100 Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill ff1abe63 2015-09-04T14:23:52 Re-land "Move Uniform and UBO info to the gl::Program layer." This data was previously stored entirely in the Impl level. Move as much as possible to the GL level, using a read-only view in the Impl level. Some information in D3D-specific, and should be stored separately in the Impl. This patch has a lot of refactoring that splits the D3D and GL info, and moves as much validation as possible to the GL layer, where it is shared between the back-ends. Re-land with fix for dEQP unused uniforms. The fix involves storing a local copy of all uniform data in the GL layer. This will also let us validate sampler indexes during draw calls at the GL layer. BUG=angleproject:1123 TEST=end2end_tests,Canary WebGL (D3D11/GL), dEQP-GLES2+3.functional.uniform_api Change-Id: I9074c057c2c67d4d3221ec11c5556a26e07c7ceb Reviewed-on: https://chromium-review.googlesource.com/297070 Tryjob-Request: Jamie Madill <jmadill@chromium.org> Reviewed-by: Corentin Wallez <cwallez@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 6ba5d4dd 2015-09-04T14:23:50 Fix GetUniform rounding in D3D. Also add a test. The test will help cover a bug that I found when doing the refactor of gl::Program. BUG=angleproject:1123 Change-Id: Id353ffaffef83ad1538580ad01668c915bf9853f Reviewed-on: https://chromium-review.googlesource.com/297031 Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Corentin Wallez <cwallez@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Cooper Partin 69f9b2c8 2015-08-20T13:25:41 Fixed GL_MAX_VERTEX_UNIFORM_VECTORS and GL_MAX_FRAGMENT_UNIFORM_VECTORS to report accurate maximum value. Change-Id: Ic581e1ef0c6be2bf1ff0092b373b3c480b6fbd8a Reviewed-on: https://chromium-review.googlesource.com/294755 Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Cooper Partin <coopp@microsoft.com> Tryjob-Request: Jamie Madill <jmadill@chromium.org>
Jamie Madill 48705cad 2015-09-02T20:40:51 Revert "Move Uniform and UBO info to the gl::Program layer." Seems to be failing dEQP-GLES2.functional.uniform_api.value.unused_uniforms.* on Linux (possibly Win as well) BUG=angleproject:1123 This reverts commit 54f882c9167b1aff5b3700187a2048bd886e8b17. Change-Id: I7dbbf40aae8dd9ebd35895df0dd338a3b6b9cc96 Reviewed-on: https://chromium-review.googlesource.com/297051 Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 54f882c9 2015-09-02T14:33:48 Move Uniform and UBO info to the gl::Program layer. This data was previously stored entirely in the Impl level. Move as much as possible to the GL level, using a read-only view in the Impl level. Some information in D3D-specific, and should be stored separately in the Impl. This patch has a lot of refactoring that splits the D3D and GL info, and moves as much validation as possible to the GL layer, where it is shared between the back-ends. BUG=angleproject:1123 TEST=end2end_tests,Canary WebGL (D3D11/GL), dEQP-GLES2+3.functional.uniform_api Change-Id: I8b7e1380a61590a46ccee646a47ffe9bda287f6b Reviewed-on: https://chromium-review.googlesource.com/295124 Reviewed-by: Geoff Lang <geofflang@chromium.org> Tryjob-Request: Jamie Madill <jmadill@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill f5f4ad24 2015-09-02T18:32:38 Revert "Revert "Clean up interface to ProgramImpl::link."" Should be fixed now with the fix the prior CL. BUG=angleproject:1123 This reverts commit 63069dfc20bc34fe6962dee694953893ee6fc663. Change-Id: Ie1826673820e3a24f9bb3e4ff5c723ac47570041 Reviewed-on: https://chromium-review.googlesource.com/297050 Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill ca03b35c 2015-09-02T12:38:13 Re-land "Compute packed varyings in ProgramD3D only." Instead of storing varying information in the shader, use a temporary set when linking a D3D program. This also means we won't have to modify information in the Shader object when linking a D3D program. This completes the refactoring for PackedVaryings. Re-land with fix for missing init of PackedVarying::vertexOnly. BUG=angleproject:1123 Change-Id: If110809c3817d88b0370ac575d739d7385b067d9 Reviewed-on: https://chromium-review.googlesource.com/296731 Reviewed-by: Corentin Wallez <cwallez@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill c437046f 2015-09-01T17:27:40 Revert "Compute packed varyings in ProgramD3D only." Likely the cause of Windows bot failures. BUG=angleproject:1123 This reverts commit 532061bbfb160586a06f0e47b6bbb5350e735203. Change-Id: Ia4f0161b97bfbf2adb4cafaa5fa5484f04fad245 Reviewed-on: https://chromium-review.googlesource.com/296710 Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 63069dfc 2015-09-01T17:26:41 Revert "Clean up interface to ProgramImpl::link." Probably causing bot failures in a prior dependent patch. BUG=angleproject:1123 This reverts commit b83592750d00ca96413575016b0e27bc9d861f58. Change-Id: I7642c4a66b190dda49cbb16611abc893f3eb6422 Reviewed-on: https://chromium-review.googlesource.com/296700 Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill b8359275 2015-08-26T15:57:14 Clean up interface to ProgramImpl::link. We no longer need to pass mutable shaders with the changed to packed varyings. BUG=angleproject:1123 Change-Id: I0a4bbf5d770ff48e10e3de2146840d8987b1e826 Reviewed-on: https://chromium-review.googlesource.com/295194 Reviewed-by: Corentin Wallez <cwallez@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 532061bb 2015-08-31T15:16:45 Compute packed varyings in ProgramD3D only. Instead of storing varying information in the shader, use a temporary set when linking a D3D program. This also means we won't have to modify information in the Shader object when linking a D3D program. This completes the refactoring for PackedVaryings. BUG=angleproject:1123 Change-Id: I241610e87f7d14f3e18b0d8bd84f1a3509c05dfd Reviewed-on: https://chromium-review.googlesource.com/295193 Reviewed-by: Geoff Lang <geofflang@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Jamie Madill 4cff2477 2015-08-21T16:53:18 Make PackedVarying a D3D-only type. The register assignment stuff only applies to the D3D back-end. Cleans up the GL back-ends use of PackedVarying, and will lead to future cleanups relating to packing varyings. BUG=angleproject:1123 Change-Id: Iaaa5fc03577e5b61ea6ae76ee1e15ad608037f34 Reviewed-on: https://chromium-review.googlesource.com/295190 Reviewed-by: Corentin Wallez <cwallez@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 80a6fc03 2015-08-21T16:53:16 Make output variables part of Program's shared data. Also initialize this structure within Program instead of DynamicHLSL. This should have benefits for other back-ends. Also these variables weren't being serialized and de-serialized with the program binary, which could mess up WebGL apps that use MRT. BUG=angleproject:1123 Change-Id: Ic0dd4840f26441a1bee8527dfa178b24daf82f8a Reviewed-on: https://chromium-review.googlesource.com/294571 Reviewed-by: Geoff Lang <geofflang@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 63805b47 2015-08-25T13:17:39 Make the Program's semantic index D3D-only. This concept isn't strictly necessary for GL-side validation. Instead we can use a bitset to track active attribs, and determine is a particular location is active. BUG=angleproject:1123 Change-Id: If7a920a3071672116bafffb3368671f721723b65 Reviewed-on: https://chromium-review.googlesource.com/294570 Reviewed-by: Geoff Lang <geofflang@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Geoff Lang 1528e56b 2015-08-24T15:10:58 Bind native GL attributes to match the locations in gl::Program. Simplifies a lot of logic when we don't have to maintain mappings between the driver and gl-layer locations. BUG=angleproject:882 BUG=angleproject:1123 Change-Id: Ia94257a322f768fdfa3167000a46a0715820ef4d Reviewed-on: https://chromium-review.googlesource.com/295231 Reviewed-by: Corentin Wallez <cwallez@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Geoff Lang <geofflang@chromium.org>
Jamie Madill c349ec03 2015-08-21T16:53:12 Re-land "Move shader attributes into Program shared data." Re-land with a fix for a warning in 32-bit Windows. Making the Program own the attribs, and the Impl only see a read-only copy cleans up the Impl object. It also allows us to more cleanly isolate certain coded into D3D. BUG=angleproject:1123 Change-Id: I73bda4b6d9e675d87d087d44757c598437d607f2 Reviewed-on: https://chromium-review.googlesource.com/295191 Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 4e10722d 2015-08-24T14:12:17 Revert "Move shader attributes into Program shared data." Once again a signed/unsigned mismatch warning in 32-bit. src\libangle\renderer\gl\programgl.cpp(190) : warning C4018: '<' : signed/unsigned mismatch BUG=angleproject:1123 This reverts commit 2d7731838722a53102e5086dba445e37f6e98d7e. Change-Id: Icd26906ead1eaa06b4bd3ff7fc2b10bef4f46022 Reviewed-on: https://chromium-review.googlesource.com/295241 Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 2d773183 2015-08-18T10:27:28 Move shader attributes into Program shared data. Making the Program own the attribs, and the Impl only see a read-only copy cleans up the Impl object. It also allows us to more cleanly isolate certain coded into D3D. BUG=angleproject:1123 Change-Id: I469051eb066fc56e55282affa2d5398b394ab8d2 Reviewed-on: https://chromium-review.googlesource.com/293826 Tested-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Jamie Madill 36cfd6a3 2015-08-18T10:46:20 Clean up Program::validate. This method should really be an Impl method, since it corresponds to a specific OpenGL call. Making it an Impl method also allows us to remove applyUniforms, which doesn't correspond to a GL call. BUG=angleproject:1123 Change-Id: I2abee3cfaa7393cad44566782d51ed701e84846b Reviewed-on: https://chromium-review.googlesource.com/293825 Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Corentin Wallez <cwallez@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill d1fe1646 2015-08-21T16:26:04 Move UBO bindings into Program shared state. Sharing the bindings with the Impl (read-only) allows us to clean out the applyUniformBuffers method from the Impl class, and make it D3D- only. BUG=angleproject:1123 Change-Id: Icb51a90cb227bbbdd83319ea308b68aa1c4c325e Reviewed-on: https://chromium-review.googlesource.com/293824 Reviewed-by: Corentin Wallez <cwallez@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 31c8c56a 2015-08-19T14:08:03 Make ProgramImpl::compileProgramExecutables D3D-only. Finally we can isolate this method into the D3D-only code, and remove the interface from ProgramImpl. BUG=angleproject:1123 Change-Id: I253b15d44d4cf8bbb5dbeaa78bbcc431a5267bf7 Reviewed-on: https://chromium-review.googlesource.com/293765 Reviewed-by: Corentin Wallez <cwallez@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill ccdf74b8 2015-08-18T10:46:12 Move TF Feedback linked varyings into D3D-only. The LinkedVarying structure is a D3D-specific type, and the GL back-end doesn't need the extra info. Isolate this into the D3D back-end so we can clean up the Impl inteface. BUG=angleproject:1123 Change-Id: I76d77ac505876d865e3e02f47acbfd6665a9507e Reviewed-on: https://chromium-review.googlesource.com/293764 Tested-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Jamie Madill e473deee 2015-08-18T14:49:01 Clean up uniform block Program Impl APIs. The D3D-specific code we can split into D3D-only methods, and the GL-level shared code we can place in Program instead of making Impl methods. This cleans up the Impl inteface significantly. BUG=angleproject:1123 Change-Id: Ibcb7d07733eb939adf5bb4f5395a661875a60238 Reviewed-on: https://chromium-review.googlesource.com/293763 Reviewed-by: Corentin Wallez <cwallez@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill ea918db9 2015-08-18T14:48:59 Clean up Program::linkUniforms. This method previously relied on output from the D3D back-end from ProgramImpl::link. Split the D3D-specific code into its own method, and do the linking validation before we call link. BUG=angleproject:1123 Change-Id: I582671e21af605af9f4451b02fc96ad29c923048 Reviewed-on: https://chromium-review.googlesource.com/293762 Reviewed-by: Corentin Wallez <cwallez@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill ada9ecc3 2015-08-17T12:53:37 Make TF Feedback buffer mode a GL-level variable. Don't query this as an Impl method, since it exists on the GL level. Also some related refactorings and cleanups. BUG=angleproject:1123 Change-Id: I3610bc0db2bcaa96408506e06a65a2f4dab93150 Reviewed-on: https://chromium-review.googlesource.com/293761 Reviewed-by: Corentin Wallez <cwallez@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 5c6b7bfe 2015-08-17T12:53:35 Add a Program::Data shared state structure. Similar to the Framebuffer and other classes, this gives the Impl class a read-only view of the object's state. BUG=angleproject:1123 Change-Id: I580eaebe2de236adf8131d6e3f54633cecce8c25 Reviewed-on: https://chromium-review.googlesource.com/293760 Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Corentin Wallez <cwallez@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Cooper Partin 4d61f7ed 2015-08-12T10:56:50 Reland Fixed compiler warning C4267 'conversion from 'size_t' to 'type', possible loss of data' Additional warnings found with more testing and added C4267 warning disable only for angle_libpng BUG=angleproject:1120 Change-Id: Ic403dcff5a8018056fa51a8c408e64207f3362eb Reviewed-on: https://chromium-review.googlesource.com/293028 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org> Tested-by: Geoff Lang <geofflang@chromium.org>
Jamie Madill b195643c 2015-08-12T17:35:20 Revert "Fixed compiler warning C4267 'conversion from 'size_t' to 'type', possible loss of data'" Seems to have quite a few warnings in 64-bit on my machine. BUG=angleproject:1120 This reverts commit c5cf9bc47d0ee028adbbf9e9f94ca567eec601dc. Change-Id: I86768b900aeba52e7a2242d9ae8949f93f1a5ba9 Reviewed-on: https://chromium-review.googlesource.com/293280 Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Cooper Partin c5cf9bc4 2015-08-06T10:46:48 Fixed compiler warning C4267 'conversion from 'size_t' to 'type', possible loss of data' BUG=angleproject:1120 Change-Id: I01ef10bea7f487c2b394d030c76628f38d2ea645 Reviewed-on: https://chromium-review.googlesource.com/292780 Tested-by: Cooper Partin <coopp@microsoft.com> Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill bd136f9a 2015-08-10T14:51:37 Reland of "D3D11: Fix draw perf regression with input layouts." f8dd7b10 fixed unnecessary shader compiles, but introduced a draw call perf regression. Fix the regression robustly by storing a minimal set of input layout elements/vertex attribute info, but also handle the case where the vector sizes mismatch between a cached and new vertex layout. Reland with fix for comparison warning. BUG=510151 Change-Id: I578ddbb5b5bb12e7c1a901f23c7b5fcbd64b5d23 Reviewed-on: https://chromium-review.googlesource.com/292460 Tested-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Jamie Madill 00c782e8 2015-08-10T17:54:30 Revert "D3D11: Fix draw perf regression with input layouts." size_t comparison warning in Win32. src\libangle\renderer\d3d\programd3d.cpp(2084) : warning C4018: '<' : signed/unsigned mismatch BUG=510151 This reverts commit bf3f780c0a281cfe6192aa5635b02a01314f7fd3. Change-Id: I5565b697ca29270375049f3e855437fa19157de9 Reviewed-on: https://chromium-review.googlesource.com/292331 Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill bf3f780c 2015-08-10T10:47:57 D3D11: Fix draw perf regression with input layouts. f8dd7b10 fixed unnecessary shader compiles, but introduced a draw call perf regression. Fix the regression robustly by storing a minimal set of input layout elements/vertex attribute info, but also handle the case where the vector sizes mismatch between a cached and new vertex layout. BUG=510151 Change-Id: If5e9ea4a4c9edaa7a78c8daa44bbe7d9927dc4e4 Reviewed-on: https://chromium-review.googlesource.com/292390 Reviewed-by: Geoff Lang <geofflang@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill f8dd7b10 2015-08-05T13:50:08 Refactor input layout & vertex signatures. Always size input layouts to gl::MAX_VERTEX_ATTRIBS, and use '1' bits to signal where we have a GPU conversion. The simplification allows us to more cleanly match the vertex executable signatures and makes our default VertexExecutable hit much more often. BUG=510151 TEST=angle_end2end_tests,Canary WebGL, manual testing with Chromium Change-Id: I5009323c4e7e208e7a2595be46658c344517a4ff Reviewed-on: https://chromium-review.googlesource.com/290740 Reviewed-by: Geoff Lang <geofflang@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Geoff Lang 6941a556 2015-07-27T11:06:45 Rename Workarounds to WorkaroundsD3D. It is not referenced outside of the d3d classes. BUG=angleproject:884 Change-Id: Ibeec034b1b78b569bea2f1c77d6a686b58b8e010 Reviewed-on: https://chromium-review.googlesource.com/288582 Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Geoff Lang <geofflang@chromium.org>
Jamie Madill 969194d4 2015-07-20T14:36:56 ES3-D3D11: Fix UBO vertex caching. There was a typo in the vertex shader cache which could trigger an overflow bug. TEST=dEQP-GLES3.functional.ubo.random.scalar_types.* BUG=angleproject:1077 Change-Id: I5652cf9675155b627f84531e09c01b42e29278fc Reviewed-on: https://chromium-review.googlesource.com/286775 Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Corentin Wallez <cwallez@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Geoff Lang 95137842 2015-06-02T15:38:43 Use a map to store uniform locations instead of a vector. Some intel drivers use seemingly arbitrary integers as uniform locations. This causes the location vector to be resized to arbitrarily large sizes. BUG=angleproject:882 Change-Id: I8db89178907dd6206eb8e646a0b426aa9bac0832 Reviewed-on: https://chromium-review.googlesource.com/274816 Reviewed-by: Kenneth Russell <kbr@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Geoff Lang <geofflang@chromium.org>
Austin Kinross 02df796f 2015-07-01T10:03:42 Centralize renderer limitations for non-conformant renderers Some renderer configurations (e.g. D3D11 Feature Level 9_3) have some limitations and aren't quite conformant. This change generates errors when applications hit these limitations, and informs developers that they must work around them. BUG=angleproject:1055 Change-Id: I6a4a9e5cc71288ca366a54c769ca0eb82e79a7f7 Reviewed-on: https://chromium-review.googlesource.com/282814 Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill d3dfda2b 2015-07-06T08:28:49 Refactor how we store vertex formats. Instead of storing a vertex format as a struct with the full info, instead use an enum, and look up the info when we need it. This saves a lot of constructor initialization time, operator comparison time, and storage. It also will allow us to look up D3D format info more quickly. BUG=angleproject:959 Change-Id: I202fd1ea96981073bc1b5b232b1ec3efa91485cb Reviewed-on: https://chromium-review.googlesource.com/277289 Reviewed-by: Geoff Lang <geofflang@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 03260faa 2015-06-22T13:57:22 D3D11: Optimize checks for active UBOs. We can reduce the amount of iterations we use by using std::vectors for storing UBO indexes, and resizing them to the max active index. This allows us to skip inactive ranges. Reduces draw call overhead. BUG=angleproject:959 Change-Id: I2577e5dbdce0e5ab102ca916eb7f1f35cf03f9ad Reviewed-on: https://chromium-review.googlesource.com/277287 Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Kenneth Russell <kbr@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Daniel Cheng f33ab835 2015-06-30T19:08:16 Fix some clang warnings with -Wmissing-braces in ANGLE. Clang warns if there are missing braces around a subobject initializer. The most common idiom that triggers this is: STRUCT s = {0}; which can be more simply written as: STRUCT s = {}; BUG=505297 Change-Id: Ib4513ee195d43e7c4a4a2f46c328866f0023a2a5 Reviewed-on: https://chromium-review.googlesource.com/282921 Reviewed-by: Kenneth Russell <kbr@chromium.org> Tested-by: Kenneth Russell <kbr@chromium.org>
Austin Kinross 137b1517 2015-06-17T16:14:53 Improve D3D GetProgramBinary chipset validation - Check chipset identifying info before trying to compile shaders - Check device feature level when loading a binary - Use chipset VendorID/DeviceID etc instead of LUID so that program binaries remain valid across system reboots Change-Id: I88ba4543bb990956d1d8fb324abf9784d72950cd Reviewed-on: https://chromium-review.googlesource.com/280428 Reviewed-by: Jamie Madill <jmadill@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 476682e6 2015-06-30T10:04:29 Use std::vector for TranslatedAttribs. This allows us to cache a std::vector between calls, and avoids us calling allocation/constructors for locals, which saves us some time. It also allows us to use the vector's size to limit the range of attribs we look at. BUG=angleproject:959 Change-Id: I799ed6c92fa8fca96e92e235b125a11d2d551aab Reviewed-on: https://chromium-review.googlesource.com/277286 Reviewed-by: Geoff Lang <geofflang@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill f9327d33 2015-06-22T13:57:16 D3D11: Optimize ProgramD3D::sortAttributesByLayout. We can use pointer math here, instead of copying values. BUG=angleproject:959 Change-Id: I3b87956224d0846c9011f5d8edb811bc5e4f2b85 Reviewed-on: https://chromium-review.googlesource.com/277119 Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Kenneth Russell <kbr@chromium.org> Tested-by: Jamie Madill <jmadill@chromium.org>
Geoff Lang 0ca53786 2015-05-07T13:49:39 Fix issues with ProgramGL. * Don't re-create the native GL program every link, some program state should persist between re-linking such as bound attribute locations. * Forward glBindAttribLocation calls to the ProgramImpl, fixes some chromium rendering issues because chromium always binds attribute locations, sometimes with gaps. * Query the real attrib location before inserting it into the list of attribs. It was unsafe to rely on the attrib having the same location as its index into the active attributes. BUG=angleproject:882 Change-Id: If14b4c4c2f5ebcdaa4f7c5a890b9519d6d4e6e43 Reviewed-on: https://chromium-review.googlesource.com/269991 Reviewed-by: Geoff Lang <geofflang@chromium.org> Tested-by: Geoff Lang <geofflang@chromium.org>
kmx.io     mail     discord kmxgit v0.4.0