path: root/src
AgeCommit message (Collapse)AuthorFilesLines
2010-03-06gallivm: Add a placeholder for TGSI_FILE_PREDICATE registers.José Fonseca1-0/+5
2010-03-06gallivm: Integrate udis86 on scons builds.José Fonseca1-0/+2
2010-03-06gallivm: Generate valid IR for LLVM 2.7.José Fonseca1-1/+14
The signatures for pack intrinsics were made more consistent in this version.
2010-03-06dri: drop MINIGLX_SOURCES (2)George Sapountzis12-7545/+0
2010-03-06dri: drop MINIGLX_SOURCESGeorge Sapountzis15-35/+0
2010-03-05mesa: bump version to 7.9Brian Paul1-3/+3
Now that the 7.8 branch has been created Mesa/master will be version 7.9
2010-03-05glx: Move initialization of ext_list_first_time to where it's storage is ↵Jon TURNEY4-9/+3
allocated Move the initialization of ext_list_first_time from all of the DRI loader's CreateScreen routines, to where the storage for the screen config is allocated. It needs to get set in the screen-config even if DRI is forced off using LIBGL_ALWAYS_INDIRECT, so that psc->direct_support is initialized correctly, otherwise __glXExtensionBitIsEnabled() always returns FALSE Specifically, this causes a problem with an X server which advertises GLX<=1.2, and the GLX_SGIX_fbconfig extension. glXGetFBConfigFromVisualSGIX() uses __glXExtensionBitIsEnabled() to check if the GLX_SGIX_fbconfig extension is available, but that function won't return correct information because that data has never been initialized, because ext_list_first_time was never set... Signed-off-by: Jon TURNEY <> Signed-off-by: Brian Paul <>
2010-03-05meta: Use the DrawBuffer's stencil sizeIan Romanick1-1/+1
Previously the code was erroneously using the stencil size of the context instead of the stencil size of the DrawBuffer. With FBOs these may be different. As a result, clearing the stencil buffer of an FBO bound to a context that doesn't have stencil would fail. Signed-off-by: Ian Romanick <>
2010-03-05swrast: Remove redundant test of the visual's accumulation buffer sizeIan Romanick1-5/+1
If the visual doesn't have an accumulation buffer, the renderbuffer passed into _swrast_clear_accum_buffer will be NULL anyway. There is no reason the check the visual. Moreover, the test erroneously checks the context's visual instead of the visual of the current DrawBuffer. With FBOs these may be different. Signed-off-by: Ian Romanick <>
2010-03-05swrast: Adjust colors based on ReadBuffer visual, not context visualIan Romanick1-9/+9
In the presence of FBOs, the visual of the context may not match the, possibly fake, visual of the current ReadBuffer. Note that the caller of adjust_colors correctly uses the visual of the ReadBuffer. Signed-off-by: Ian Romanick <>
2010-03-05swrast: Remove stray comment that mentions ctx->VisualIan Romanick1-1/+0
Signed-off-by: Ian Romanick <>
2010-03-05gallivm: checkpoint: code gen for mipmap selectionBrian Paul1-0/+65
2010-03-05llvmpipe: added code to set texture depth, max mipmap levels infoBrian Paul4-5/+23
2010-03-05gallivm: added methods for getting texture depth, num mipmap levelsBrian Paul1-0/+12
2010-03-05gallivm: checkpoint: texture LOD computation code genBrian Paul1-3/+126
2010-03-05gallivm: added field for sampler lodBrian Paul2-4/+6
2010-03-05i965: Only use the effective subset of the WM sampler key.Eric Anholt1-4/+9
The whole structure is 836 bytes, but if only the first one or two samplers are used (as is common), the part that matters is only 56 or 108 bytes. By using just that subset as the key (since the key size is also part of the key), we improve firefox-talos-gfx performance by 1%.
2010-03-05intel: Remove support for the DRI1 TFP extension.Eric Anholt7-144/+39
2010-03-05i965: Remove the paths for making a const surface object for a NULL BO.Eric Anholt2-20/+13
2010-03-05i965: Remove the unused SDC BO pointer from the sampler state key.Eric Anholt1-1/+0
Since it's a BO pointer, it already lives outside the key in brw->wm.sdc_bo[] which is used for sampler state lookup and setup.
2010-03-05i965: Rename SNB brw_*_key structs to gen6_*_key structs.Eric Anholt1-8/+8
2010-03-05i965: Pack gen6 blend_state_key structure.Eric Anholt1-2/+1
2010-03-05intel: Only do batchbuffer debug if --enable-debug is used.Eric Anholt1-0/+10
This saves 6.6KB on the 965 driver, and appears to speed firefox-talos-gfx up by 1-2%. Unlike many other asserts in the driver, when we make a mistake that would trigger one of these it generally shows up all the time for developers, so turning it off for release seems fine.
2010-03-05intel: Replace batch macro contents with function calls.Eric Anholt1-28/+39
This manages to cut down another 3800 bytes.
2010-03-05intel: Move the assertions about reloc delta from the macros to the function.Eric Anholt2-2/+4
Cuts another 1800 bytes from the driver.
2010-03-05intel: Check that the batch is mapped per BEGIN, not each OUT.Eric Anholt1-1/+1
Shaves 800 bytes off the driver.
2010-03-05i965: Try to assign SNB SF inputs to outputs.Eric Anholt1-2/+41
2010-03-05i965: Fix the SNB SF input size.Eric Anholt1-1/+1
2010-03-05Always return VOLATILE for ObjectPurgeable(VOLATILE)Chris Wilson1-4/+16
Fixes fdo bug 26128. The spec mandates that VOLATILE is returned from ObjectPurgeable(VOLATILE) irrespective of the actual status of the object upon completion of marking it purgeable. Conform to the spec, even though it seems wrong. Signed-off-by: Chris Wilson <>
2010-03-05llvmpipe: add pipe_thread_wait() callsBrian Paul1-0/+6
Wait for threads to exit before cleaning up per-thread data. Fixes hang on context destruction with glean makeCurrent test. See fd.o bug 26536.
2010-03-05gallium/docs: document additional wrap mode restrictionsBrian Paul1-0/+3
2010-03-05gallivm: minor arithmetic improvementsBrian Paul1-9/+12
2010-03-05gallivm: implement non-normalized texture wrap modesBrian Paul1-39/+68
Note that only the PIPE_TEX_WRAP_CLAMP,CLAMP_TO_EDGE,CLAMP_TO_BORDER modes work with non-normalized texcoords.
2010-03-05softpipe: implement PIPE_TEX_WRAP_CLAMP_TO_EDGE for unorm texture samplingBrian Paul1-6/+43
CLAMP_TO_BORDER and CLAMP_TO_EDGE were doing the same thing.
2010-03-05mesa: minor reformatting, new commentsBrian Paul1-2/+9
2010-03-05DRI2: expose OML_sync_control if possibleJesse Barnes1-0/+3
If the server supports the OML related protocol, enable support for the extension.
2010-03-05GLX: remove debug prints from glXWaitForMscOML dispatchJesse Barnes1-4/+0
Leftover from earlier commit.
2010-03-05nouveau: s/TEXTURE_USAGE_PRIMARY/TEXTURE_USAGE_SCANOUTChristoph Bumiller3-5/+5
2010-03-05gallium/docs: updated/improve sampler state documentationBrian Paul1-21/+77
2010-03-05mesa: whitespace fixes, 80-column wrapping, etc.Brian Paul1-84/+102
2010-03-05Merge branch 'object-purgeable'Chris Wilson23-4964/+5762
Acked-by: Brian Paul <>
2010-03-05APPLE_object_purgeable: intelChris Wilson2-1/+136
Implement support for purgeable objects by using the GEM madvise ioctl. Signed-off-by: Chris Wilson <>
2010-03-05APPLE_object_purgeable: coreChris Wilson8-0/+405
Signed-off-by: Chris Wilson <>
2010-03-05mesa/es: Update GLAPI and APISpec for GL_OES_EGL_image.Chia-I Wu4-0/+46
This generates the entrypoints and dispatches for GL_OES_EGL_image. There is no real support yet.
2010-03-05st/egl: Add support for EGL_KHR_image_base and EGL_KHR_image_pixmap.Chia-I Wu4-3/+202
These extensions are not quite useful until the client APIs gain support for the respective EGLImage extensions.
2010-03-05st/egl: Add get_param to native display interface.Chia-I Wu4-0/+72
get_param can be used to query the parameters of a native display. There is only NATIVE_PARAM_USE_NATIVE_BUFFER right now. It queries whether the window/pixmap surfaces use the native buffers instead of private buffers.
2010-03-05st/egl: Move modeset interface to native_modeset.h.Chia-I Wu2-55/+89
native.h is getting more and more complex. Move the independent modeset interface to native_modeset.h to simplify native.h a little.
2010-03-05st/egl: Add event support to the native display interface.Chia-I Wu8-14/+106
There is only invalid_surface event right now. When EGL receives the event, it sets the force_validate flag of the context binding to the surface. This helps skip an unnecessary check.
2010-03-04Revert "mesa: Remove pointless comparison of unsigned integer with a ↵Vinson Lee1-4/+9
negative constant." This reverts commit a05fdbcb719ac64e6be842372813f0f4ca2f4f93. Removing the comparison is wrong. The comparison with -1 should be changed to another value (probably PROGRAM_UNDEFINED) along with another change in the shader assembler. Conflicts: src/mesa/shader/prog_execute.c
2010-03-04intel: Remove non-kernel-exec-fencing support.Eric Anholt19-638/+94
Shaves 60k off the driver from removing the broken spans code. This means we now require 2.6.29, which seems fair given that it's a year old and we've removed support for non-KMS already in the last release of 2D.