summaryrefslogtreecommitdiff
path: root/src
AgeCommit message (Collapse)AuthorFilesLines
2009-09-02st/mesa: tweak mesa/gallium RGB format selection/translationBrian Paul1-3/+4
Mesa's _mesa_texformat_bgr888 is a better match for PIPE_FORMAT_R8G8B8_UNORM.
2009-09-02mesa: additional texstore assertionsBrian Paul1-0/+13
When we call _mesa_make_temp_chan_image() we're typically passing dstFormat->BaseFormat as the textureBaseFormat. This parameter controls the components per pixels for the new/temp image. The subsequent loops that pack the temp pixels into the texture buffer expect a certain number of components.
2009-09-02gallium/util: add tile support for PIPE_FORMAT_R8G8B8_UNORMBrian Paul1-1/+53
2009-09-02mesa: the temp image in _mesa_texstore_argb8888() has to be GL_RGBABrian Paul1-1/+1
If we're storing an xrgb888 image the texFormat->BaseFormat will be GL_RGB but in the loop after _mesa_texstore_argb8888() we're expecting GL_RGBA. Just pass GL_RGBA to _mesa_make_temp_chan_image() all the time. Fixes bad texture seen in progs/tests/packedpixels.c for the GL_RGB case.
2009-09-02mesa: initial check-in of xRGB formatsKeith Whitwell6-22/+165
2009-09-02mesa: fix the non-GNU C bit-field caseBrian Paul1-2/+2
2009-09-02mesa: silence gcc bit-field warningGary Wong1-2/+2
2009-09-02mesa: replace 8 with NUM_UNITSBrian Paul1-1/+4
2009-09-02mesa: remove accidentally committed printfBrian Paul1-1/+0
2009-09-02mesa: added #ifdef __GNUC__ around GLubyte bitfield usageBrian Paul1-0/+5
It would be nice if there were a #pragma or something to disable the warnings: main/texenvprogram.c:87: warning: type of bit-field ‘Source’ is a GCC extension main/texenvprogram.c:88: warning: type of bit-field ‘Operand’ is a GCC extension but there doesn't appear to be a way to do that.
2009-09-02mesa: Compact state key for TexEnv program cacheChris Wilson1-3/+5
By rearranging the bitfields within the key we can reduce the size of the key from 644 to 196 bytes, reducing the cost of both the hashing and equality tests.
2009-09-02st/xorg: Remove code no longer support by xserver Git master.Michel Dänzer4-59/+4
Shouldn't be necessary with older versions either.
2009-09-02st/xorg: Add support for xserver Git master DRI2 interface.Michel Dänzer2-81/+151
Reorganize to handle both interfaces with common code.
2009-09-02st/dri: Slight consolidation of DRI2 buffer handling.Michel Dänzer2-30/+27
Always use / assume a combined depth/stencil buffer. Also update drawable formats from surfaces returned by the driver.
2009-09-02st/xorg: Add EXA UploadToScreen and DownloadFromScreen hooks.Michel Dänzer1-0/+64
2009-09-02i965: CS FENCE in URB_FENCE is 11-bits wideXiang, Haihao1-2/+2
2009-09-02i965: validate sf stateXiang, Haihao1-0/+1
2009-09-01mesa: Make MultiDrawElements submit multiple primitives at once.Eric Anholt10-33/+200
Previously, MultiDrawElements just called DrawElements a bunch of times. By sending several primitives down the pipeline at once, we avoid a bunch of validation. On my GL demo, this improves fps by 2.5% (+/- .41%) and reduces CPU usage by 70.5% (+/- 2.9%) (n=3). Reviewed by: Ian Romanick <ian.d.romanick@intel.com>
2009-09-02nv50: SWZ is the same as MOV from our perspectiveBen Skeggs1-0/+1
2009-09-01mesa: skip bitmap drawing code if width==0 or height==0Brian Paul1-19/+22
2009-09-01intel: use _mesa_expand_bitmap() to skip an intermediate bufferBrian Paul1-21/+6
2009-09-01st/mesa: use new _mesa_expand_bitmap() functionBrian Paul1-57/+4
2009-09-01mesa: new _mesa_expand_bitmap() functionBrian Paul2-0/+92
2009-09-01mesa: remove redundant assignmentsBrian Paul1-8/+1
2009-09-01mesa: more clean-upsBrian Paul1-24/+31
2009-09-01mesa: change conditional to match the previous oneBrian Paul1-1/+1
2009-09-01mesa: updated #includesBrian Paul1-2/+1
2009-09-01mesa: remove unused texenv_fragment_program::ctx fieldBrian Paul1-2/+0
2009-09-01mesa: remove unused ureg::abs fieldBrian Paul1-5/+2
2009-09-01mesa: remove unused ureg:negateabs fieldBrian Paul1-4/+1
2009-09-01mesa: more comments, clean-upsBrian Paul1-10/+10
2009-09-01mesa: simplify translate_tex_src_bit()Brian Paul1-20/+3
2009-09-01mesa: minor code clean-ups, commentsBrian Paul1-24/+34
2009-09-01mesa: replace 8 with MAX_TEXTURE_UNITSBrian Paul1-1/+1
2009-09-01dri: remove unused meta_clear_tris()Brian Paul2-266/+2
2009-09-01intel: use BUFFER_BITS_COLORBrian Paul1-1/+1
2009-09-01intel: fix incorrect parameter type for intel_bufferobj_map_range()Brian Paul1-1/+1
2009-09-01radeon: trim down #includesBrian Paul1-28/+0
2009-09-01radeon: use _mesa_meta_clear()Brian Paul1-2/+2
2009-09-01mesa: obey stencil write mask in _mesa_meta_draw_pixels()Brian Paul1-6/+8
2009-09-01intel: set Length/Offset fields in intel_bufferobj_map()Brian Paul1-0/+3
2009-09-01intel: use _mesa_meta_copy_pixels() when do_blit_copypixels() failsBrian Paul1-5/+1
Also, trim down #includes.
2009-09-01intel: trim down #includesBrian Paul1-8/+0
2009-09-01intel: use _mesa_meta_draw_pixels()Brian Paul1-147/+4
The textured quad path is slightly faster and will work with POT textures on i945.
2009-09-01intel: trim down #includesBrian Paul1-17/+0
2009-09-01intel: use _mesa_meta_clear(), it's a bit fasterBrian Paul1-1/+2
2009-09-02radeon: Fix OQ to set ful lstate as dirty too.Pauli Nieminen1-0/+1
2009-09-02radeon: Fix debug output to filter out less critical messages instead of ↵Pauli Nieminen1-1/+1
more critical.
2009-09-01st/xorg: getting indexes right tends to be helpfulZack Rusin1-3/+3
2009-09-01st/xorg: redo the vertex shaderZack Rusin1-14/+35