summaryrefslogtreecommitdiff
path: root/src/gallium
AgeCommit message (Collapse)Author
2010-02-12wgl: Do not reach out and destroy contexts on cleanup.José Fonseca
Simply skip cleanup when contexts are still active. This addresses two issues: - in some situations the ICD DLL may be unloaded before the DLL that is using GL contexts is, so we may receive GL calls after stw_cleanup. - when aborting (exception, or control-c) the contexts may have been left in an inconsistent state and attempting to destroy can cause unpredictable results.
2010-02-12gdi: Never fail at DLL load time.José Fonseca
Windows doesn't really expect things to fail at this point -- it will try many times until it finally gives up, worse, something bad happens. The WGL state tracker will gracefully decline to do work even when it is loaded.
2010-02-12wgl: Fail gracefully whenever 3D could not be enabled for some reason.José Fonseca
2010-02-12svgadump: Several improvements to shader dumping.José Fonseca
2010-02-12nouveau: use AVAIL_RING()Ben Skeggs
2010-02-12nouveau: switch to nouveau_bo_new_tile()Ben Skeggs
2010-02-11svga: check min_index, max_index before assertionsBrian Paul
It's possible for min_index and max_index to be ~0 if the min/max values were not actually set or computed in the state tracker. Skip some assertions in that case. This only effects the debug build error checking. (cherry picked from commit 408f32dc165b36eb5ec0e2bbc7dcecd28f3a62d3)
2010-02-12gallium: make max_anisotropy a unsigned bitfield memberRoland Scheidegger
saves us a dword in sampler state, hw can't do non-integer aniso degree anyway. To allow aniso 1x (which seems of dubious value but some hardware (radeons) have such a mode, and even d3d allows specifiying it) redefine anisotropic filtering as disabled only if max_anistropy is 0.
2010-02-11nouveau: fix compile errors...Roland Scheidegger
2010-02-11nv50: adapt to stencil ref changesRoland Scheidegger
2010-02-11nv40: adapt to stencil ref changesRoland Scheidegger
2010-02-11nv30: adapt to stencil ref changesRoland Scheidegger
not sure if this has any hope of working
2010-02-11svga: implement PIPE_CAP_MAX_COMBINED_SAMPLERS queryBrian Paul
2010-02-11gallium: make u_blitter code save/restore stencil ref valuesRoland Scheidegger
2010-02-11r300g: adapt to stencil ref changesRoland Scheidegger
2010-02-11cell: adapt to stencil ref changesRoland Scheidegger
not betting this compiles, some of the code handling stencil test seems incomplete and/or never called?
2010-02-11i965g: adapt to stencil ref changesRoland Scheidegger
while there also potentially fix logic ops (were never enabled before)
2010-02-11i915g: adapt to stencil ref changesRoland Scheidegger
2010-02-11gallium: give pipe_stencil_ref its own cso_save/restore functionsRoland Scheidegger
seems cleaner, and other dynamic state like viewport is handled that way too
2010-02-11Revert "llvmpipe: Move tile coordinate assertion into a more useful place."José Fonseca
This reverts commit 8c0391726a89414356c705fcc2d5a38d12d0fe2b. Too pendantic.
2010-02-11scons: Fix drm/intel build.José Fonseca
2010-02-11llvmpipe: Handle TGSI_TOKEN_TYPE_PROPERTY.José Fonseca
Avoids assertion failures with certain shaders.
2010-02-11llvmpipe: Move tile coordinate assertion into a more useful place.José Fonseca
2010-02-11llvmpipe: Clamp both extremes of triangles to framebuffer size.José Fonseca
Fix segmentation fault when triangles crossed the axis.
2010-02-11gallivm: TGSI_OPCODE_CONT is not deprecated.José Fonseca
Note that with FIXME instead of an assertion failure. Addresses fdo 25956.
2010-02-11nouveau: remove u_simple_screen.h includesBen Skeggs
2010-02-11gallium: also save/restore stencil_ref in cso_save/restore_depth_stencil_alphaRoland Scheidegger
makes life of state trackers easier
2010-02-10i965g: ifdef out unused functions.Vinson Lee
2010-02-10r300g: Work around "defect" in r300compiler.Corbin Simpson
r300compiler doesn't handle half swizzles for vert shaders, which don't have them. So, for now, disable them.
2010-02-10os: Do not use Pthreads barrier functions on Mac OS X.Vinson Lee
Pthreads barrier functions are not available on some POSIX platforms.
2010-02-10gallium: cast to silence waringBrian Paul
2010-02-10gallium: use os_time.hBrian Paul
2010-02-10gallium: use os_time.h in pb_bufmgr_cache.cBrian Paul
Untested, but seems straightforward.
2010-02-10gallium: use os_time.h in u_timed_winsys.cBrian Paul
2010-02-10r300g: Fix typo.Corbin Simpson
Thanks to Uros Bizjak for catching it.
2010-02-10svga: adapt to stencil ref changesRoland Scheidegger
also add some disabled code to handle blendcolor?
2010-02-10llvmpipe: adapt to stencil ref changesRoland Scheidegger
since the driver doesn't actually handle stencil yet the code won't do much.
2010-02-10identity: adapt to stencil ref changesRoland Scheidegger
2010-02-10fo: adapt to stencil ref changesRoland Scheidegger
2010-02-10llvmpipe: Map rendertargets when bound as textures.José Fonseca
Fixes gearbox segfault (fdo 25960). Corruption remains.
2010-02-10dri: Allow selective generation of accum. buffer configsIan Romanick
Modify the interface to driCreateConfigs allowing drivers to not expose configs with an accumuation buffer. All of the drivers calling function have been updated to pass true for the accumulation selector. This maintains the current behavior. Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> Reviewed-by: Corbin Simpson <MostAwesomeDude@gmail.com>
2010-02-10trace: adapt to stencil ref changesRoland Scheidegger
2010-02-10softpipe: adapt to stencil ref changesRoland Scheidegger
2010-02-10gallium: adapt state trackers to stencil ref changeRoland Scheidegger
2010-02-10st/xorg: fix a silly offset bug in xvZack Rusin
we were multiplying the difference in the wrong direction.
2010-02-10llvmpipe: Fix null pointer dereference when no depth buffer is bound.José Fonseca
2010-02-10llvmpipe: Leave depth buffer in swizzled format.José Fonseca
This fixes several assertion failures due to only Z32 being supported.
2010-02-10llvmpipe: Remove unnecessary headers.Vinson Lee
2010-02-10Drop incorrectly revived files during the 'gallium-nopointsizeminmax' merge.Francisco Jerez
2010-02-09r300g: One less assert.Corbin Simpson
This entire codepath needs to be cleaned to be more pliant towards shader failure.