summaryrefslogtreecommitdiff
path: root/tests/texturing
AgeCommit message (Collapse)AuthorFilesLines
2014-07-14tex-miplevel-selection: test all variants of textureProjGradOffsetMarek Olšák1-4/+16
2014-07-14tex-miplevel-selection: test all variants of textureProjGradMarek Olšák1-3/+14
2014-07-14tex-miplevel-selection: test all variants of textureGradOffsetMarek Olšák1-18/+36
2014-07-14tex-miplevel-selection: test all variants of textureGradMarek Olšák1-7/+56
2014-07-14tex-miplevel-selection: test all variants of textureProjLodOffsetMarek Olšák1-15/+42
+ Use level 0 texture coordinates to really verify the Lod fetch works. It was using scaled coordinates, so textureLodOffset could have been replaced by textureOffset and the test would pass. Now the test will fail if drivers attempt to fake it that way.
2014-07-14tex-miplevel-selection: test all variants of textureProjLodMarek Olšák1-3/+14
2014-07-14tex-miplevel-selection: test all variants of textureLodOffsetMarek Olšák1-5/+17
2014-07-14tex-miplevel-selection: test all variants of textureProjOffset with biasMarek Olšák1-5/+22
+ a comment about offsets
2014-07-14tex-miplevel-selection: simplify shader code generationMarek Olšák1-234/+85
2014-07-14tex-miplevel-selection: test all variants of textureProjOffsetMarek Olšák1-3/+36
2014-07-14tex-miplevel-selection: test all variants of textureProj with biasMarek Olšák1-2/+35
2014-07-14tex-miplevel-selection: test all variants of textureProjMarek Olšák1-11/+90
2014-07-14tex-miplevel-selection: test all variants of textureOffset with biasMarek Olšák1-9/+56
+ some bug fixes
2014-07-14tex-miplevel-selection: test all variants of textureOffsetMarek Olšák1-86/+212
2014-07-14tex-miplevel-selection: test all variants of texture() with biasMarek Olšák1-46/+157
2014-07-14tex-level-selection: deduplicate code, improve outputMarek Olšák1-58/+115
2014-07-14tex-miplevel-selection: test all variants of textureLodMarek Olšák1-93/+457
2014-06-09depth-tex-modes-rg: fix initial window sizeBrian Paul1-2/+6
Use the larger of the default window size and the size required by the test parameters. Reviewed-by: José Fonseca <jfonseca@vmware.com>
2014-06-09tex-skipped-unit: use default window sizeBrian Paul1-3/+0
Reviewed-by: José Fonseca <jfonseca@vmware.com>
2014-05-01teximage-errors: Make changes to support core profileAnuj Phogat1-1/+1
Signed-off-by: Anuj Phogat <anuj.phogat@gmail.com>
2014-05-01teximage-errors: Test the combinations of depth and depth-stencil formatsAnuj Phogat1-1/+75
This patch is to verify a bug fix in mesa commit 079bff5. This commit allowed GL_DEPTH_COMPONENT and GL_DEPTH_STENCIL combinations in glTexImage{123}D functions. V2: Add extension check for GL_EXT_texture_array Signed-off-by: Anuj Phogat <anuj.phogat@gmail.com> Reviewed-by: Brian Paul <brianp@vmware.com>
2014-05-01teximage-errors: Run all the tests before returning resultAnuj Phogat1-6/+4
Signed-off-by: Anuj Phogat <anuj.phogat@gmail.com> Reviewed-by: Brian Paul <brianp@vmware.com>
2014-04-30tex3d-npot: assorted clean-upsBrian Paul1-27/+24
As with the previous patch for the tex3d test. Reviewed-by: José Fonseca <jfonseca@vmware.com>
2014-04-30tex3d: assorted clean-upsBrian Paul1-36/+33
Don't call piglit_report_result() - return result from piglit_display(). Use default window size and fix "row wrapping" to avoid drawing off the edge of the window. Use bool instead of int. Use piglit_get_gl_enum_name(). Use GLubyte instead of unsigned char. Set texture env mode to replace. Reviewed-by: José Fonseca <jfonseca@vmware.com>
2014-04-30util: Don't redefine math.h functions on MSVC 2013.José Fonseca1-7/+0
MSVC 2013 provides all math.h functions that were missing in previous versions. Reviewed-by: Brian Paul <brianp@vmware.com>
2014-04-29texsubimage: use default window sizeBrian Paul1-2/+0
Reviewed-by: José Fonseca <jfonseca@vmware.com>
2014-04-29texture-al: use default window sizeBrian Paul1-2/+0
Reviewed-by: José Fonseca <jfonseca@vmware.com>
2014-04-29fragment-and-vertex-texturing: use default window sizeBrian Paul1-2/+0
Reviewed-by: José Fonseca <jfonseca@vmware.com>
2014-04-29texture-rg: use default window sizeBrian Paul1-2/+0
Reviewed-by: José Fonseca <jfonseca@vmware.com>
2014-04-29texredefine: use default window sizeBrian Paul1-2/+0
Reviewed-by: José Fonseca <jfonseca@vmware.com>
2014-04-29tex3d-maxsize: use default window sizeBrian Paul1-2/+0
Reviewed-by: José Fonseca <jfonseca@vmware.com>
2014-04-29tex-srgb: use default window sizeBrian Paul1-2/+0
Reviewed-by: José Fonseca <jfonseca@vmware.com>
2014-04-29proxy-texture: use default window sizeBrian Paul1-2/+0
Reviewed-by: José Fonseca <jfonseca@vmware.com>
2014-04-29incomplete-texture: use default window sizeBrian Paul1-2/+0
Reviewed-by: José Fonseca <jfonseca@vmware.com>
2014-04-29textureGather: Avoid variable length arrays.José Fonseca1-2/+5
Not supported by MSVC. Reviewed-by: Brian Paul <brianp@vmware.com>
2014-04-26remove old tex-swizzle testBrian Paul2-853/+0
It didn't test the GL_ONE and GL_ZERO terms and was needlessly complicated. Replaced by ext_texture_swizzle-swizzle test. Reviewed-by: Eric Anholt <eric@anholt.net>
2014-04-17textureGather: make sure to test out min/max offsetsIlia Mirkin1-42/+62
This increases the texture size and uses min/max allowable offsets instead of just -8/7. (Unfortunately this also slows the test down, as it will create a 64x64 texture for a -32/31 min/max.) Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu> Reviewed-by: Chris Forbes <chrisf@ijw.co.nz>
2014-04-16depth-tex-compare: Generally increase sanityChris Forbes1-32/+32
- Take 4 samples rather than 3 horizontally within each test. - Make EQUAL / NOTEQUAL cases more like the others, so we can see visually if they are wrong, and to also remove the tendency for spurious failures. Signed-off-by: Chris Forbes <chrisf@ijw.co.nz> Reviewed-by: Brian Paul <brianp@vmware.com>
2014-04-16depth-tex-compare: Get rid of handrolled enum stringsChris Forbes1-11/+2
Signed-off-by: Chris Forbes <chrisf@ijw.co.nz> Reviewed-by: Brian Paul <brianp@vmware.com>
2014-04-01getteximage-targets: use MAX2() macro instead of hand-rolled max()Brian Paul1-9/+3
Fixes MSVC build. Bugzilla: http://bugs.freedesktop.org/show_bug.cgi?id=76849
2014-03-26getteximage-targets: Add read texture image into PBO case for each targetJon Ashburn1-6/+40
Reviewed-by: Chia-I Wu <olv@lunarg.com>
2014-03-26getteximage-targets: Refactor and cleanup with generalized texture targetsJon Ashburn1-96/+132
Reviewed-by: Chia-I Wu <olv@lunarg.com>
2014-03-19texwrap: add ARB_depth_texture dependencyIlia Mirkin1-1/+2
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu> Reviewed-by: Marek Olšák <marek.olsak@amd.com>
2014-03-19copyteximage: add ARB_depth_texture dependenciesIlia Mirkin1-1/+6
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu> Reviewed-by: Marek Olšák <marek.olsak@amd.com>
2014-03-04max-texture-size: fix size reported in error messageBrian Paul1-1/+1
If texInitData() failed, it was with maxSide/2, not maxSide. Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
2014-03-04max-texture-size: test glTexSubImage2D() for GL_TEXTURE_RECTANGLE tooBrian Paul1-0/+3
All the other texture targets test glTexSubImage(), so do it for this target too. Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
2014-02-27s3tc-errors: Fix GCC compiler warnings.Vinson Lee1-3/+3
The params argument in glGetTexLevelParameteriv is of type GLint*. This patch fixes these GCC warnings introduced with commit 4f9d98dbaea90c11a930fd774382ff364892ce3f. s3tc-errors.c:156:6: warning: passing argument 4 of ‘piglit_dispatch_glGetTexLevelParameteriv’ from incompatible pointer type [enabled by default] &is_compressed); ^ s3tc-errors.c:156:6: note: expected ‘GLint *’ but argument is of type ‘_Bool *’ s3tc-errors.c:158:6: warning: pointer targets in passing argument 4 of ‘piglit_dispatch_glGetTexLevelParameteriv’ differ in signedness [-Wpointer-sign] &format); ^ s3tc-errors.c:158:6: note: expected ‘GLint *’ but argument is of type ‘GLenum *’ s3tc-errors.c:161:6: warning: pointer targets in passing argument 4 of ‘piglit_dispatch_glGetTexLevelParameteriv’ differ in signedness [-Wpointer-sign] &compressed_size); ^ s3tc-errors.c:161:6: note: expected ‘GLint *’ but argument is of type ‘GLuint *’ Signed-off-by: Vinson Lee <vlee@freedesktop.org> Reviewed-by: Brian Paul <brianp@vmware.com> Reviewed-by: Daniel Kurtz <djkurtz@chromium.org>
2014-02-17texturing/s3tc-errors: double check GL_TEXTURE parameters and allocDaniel Kurtz1-4/+42
If the GL does not actually support one of the compression formats in s3tc_formats, it may silently choose a different actual internalformat. In such a case, the resulting compressed image size may be larger, leading to memory corruption when the bigger-than-expected image is read back with glGetCompressedTexImage() into a buffer that was allocated to the expected, smaller, size. Instead, first confirm the texture has really been compressed, that the format is as expected, and that the size matches our pre-computed expected size. If any of these fail, set pass to false and spit an error, but keep going, using the actual format and size returned by the GL. Ian Romanick pointed out that "if we're going to all this effort, we should probably go all the way. A follow-on patch should check that glGetCompressedTexImage doesn't over-run the buffer." v2: * Fixed parameter check block format per Ian Romanick review * Updated commit message title per Brian Paul review * Function call continuation lines align with open '(' (like in rest of file). Signed-off-by: Daniel Kurtz <djkurtz@chromium.org> Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
2014-02-11textureGather: add textureGatherOffsets to execution test (v2.1)Dave Airlie1-19/+54
Test for checking textureGatherOffsets from ARB_gpu_shader5 works, this is tested on the nvidia binary driver and passes. also add to all.py. v2.1 fix VS bits + whitespace Reviewed-by: Chris Forbes <chrisf@ijw.co.nz> Signed-off-by: Dave Airlie <airlied@redhat.com>
2014-02-05max-texture-size: Cut SubImage data allocation by 1/4.Kenneth Graunke1-1/+1
I honestly have no idea why, but the original test calls TexSubImage with a width and height of *half* the texture size. Yet it allocates enough temporary pixel data for a full texture. This makes no sense, and clearly one or the other should change. Without understanding why the existing test does what it does, I chose to not change it and simply cut the memory allocation. Cutting memory usage also may reduce the likelihood of GL_OUT_OF_MEMORY errors. Signed-off-by: Kenneth Graunke <kenneth@whitecape.org> Reviewed-by: Matt Turner <mattst88@gmail.com>
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-02 19:02:28 +0000