Age | Commit message (Collapse) | Author | |
---|---|---|---|
2007-05-11 | r300: Cleaned up the blend factor function. | Oliver McFadden | |
2007-05-11 | r300: Removed the r300BindProgram function; mesa default does better checking. | Oliver McFadden | |
2007-05-11 | r300: Removed deprecated/disabled VBO code from r300_mem.c | Oliver McFadden | |
Also removed a couple of unused fields from the r300_memory_manager structure. | |||
2007-05-11 | r300: Indented r300_mem.c; I forgot this because it used to be radeon_mm.c. | Oliver McFadden | |
2007-05-11 | r300: Use __FUNCTION__ not __func__. | Oliver McFadden | |
Just for consistency; most of the code already uses __FUNCTION__. | |||
2007-05-11 | r300: Assert if the primitive type is unknown; this can't really happen. | Oliver McFadden | |
2007-05-11 | r300: Moved some code around in r300_render.c and general clean up. | Oliver McFadden | |
2007-05-11 | user-declared uniform structs not supported yet (see bug 10908) | Brian | |
2007-05-11 | don't ignore return value of _slang_codegen_global_variable() | Brian | |
2007-05-11 | remove some unneeded code in init_machine() | Brian | |
2007-05-11 | r300: Removed some checking in r300NumVerts that is not needed. | Oliver McFadden | |
According to Aapo Tahkola the OpenGL specification defines the behaviour when there are not enough vertices for the primitive type, thus DRI drivers do not need to perform verification on the number of vertices per primitive. | |||
2007-05-11 | r300: Function naming correction in r300_texmem.c. | Oliver McFadden | |
2007-05-11 | r300: Cleaned up function names in r300_state.c. | Oliver McFadden | |
2007-05-11 | r300: Cleaned up function naming in r300_emit.c. | Oliver McFadden | |
2007-05-11 | r300: Renamed r300_maos.c to r300_emit.c; it contains mostly emit code now. | Oliver McFadden | |
2007-05-11 | Remove unused FB_* tokens, re-indent code. | Brian | |
2007-05-11 | When feeding back texcoords, don't divide by W. See bug 10913. | Brian | |
2007-05-11 | Fix some bugs related to loop counters and conditional branching. | Patrick Baggett | |
2007-05-10 | more indentation fixes, remove 'register' keywords | Brian | |
2007-05-10 | re-indent some code | Brian | |
2007-05-10 | document exp(), mod() fixes | Brian | |
2007-05-10 | regenerated | Brian | |
2007-05-10 | Implement exp() in terms of __asm float_power. Fix typo in mod(vec4) function. | Brian | |
exp() was using __asm float_exp (OPCODE_EXP) but that computes base two, not e. See bug 10907. | |||
2007-05-10 | Fix reversed enable logic in enable_texture | Ian Romanick | |
Fix bug inserted in commit c9e723e5013443df984cb3987ffa3a9ba3384b89. Discovered by Oliver McFadden (z3ro). | |||
2007-05-10 | r300: Added TODO comment regarding texture tiling; I'm not sure about this. | Oliver McFadden | |
2007-05-10 | r300: Corrected a compile error introduced by one of the previous commits. | Oliver McFadden | |
2007-05-10 | r300: Removed obsolete start_index16_packet/start_index32_packet. | Oliver McFadden | |
It's all in r300_render.c now. | |||
2007-05-10 | r300: Moved some more emit code into r300_render.c. | Oliver McFadden | |
2007-05-10 | r300: Moved some more of the emit code into r300_render.c. | Oliver McFadden | |
2007-05-10 | document tex sampler bug fix, code changes | Brian | |
2007-05-10 | Replace `pkg-config --cflags libdrm` with LIBDRM_CFLAGS, remove disabled ↵ | Brian | |
lines, remove obsolete comments. | |||
2007-05-10 | Convert "bit" parameters to GLbitfield. Fix cut-and-paste bug in _mesa_IsEnabled | Ian Romanick | |
These changes are based on patch review comments from Brian Paul, Alan Hourihane, and vehemens. | |||
2007-05-10 | Refactor the loop in unbind_texobj_from_texunits. | Ian Romanick | |
Common code was pulled out of the per-target if-statment and put at the end of the for-loop. The common code is guarded by a new variable, curr, that is set to point to the unit's current target in each if-statement. | |||
2007-05-10 | Refactor queries of GL_(SOURCE|OPERAND)[012]_(ALPHA|RGB). | Ian Romanick | |
Most switch-statements that have cases for these enums already use code like: const GLuint idx = pname - GL_SOURCE0_RGB; ... texUnit->Combine.SourceRGB[idx] ... This patch just brings the remaining bits up to speed. | |||
2007-05-10 | Refactor the way TestProxyTexImage is called in texture_error_check. | Ian Romanick | |
2007-05-10 | Refactor Enable / Disable and IsEnabled bits related to texture targets. | Ian Romanick | |
2007-05-09 | r300: Name the render functions in r300_render.c consistently. | Oliver McFadden | |
2007-05-09 | r300: Return -1 for error. | Oliver McFadden | |
2007-05-09 | r300: Probably a good idea to always check the vertices; it's a WARN_ONCE so the | Oliver McFadden | |
performance impact is extremely minimal. | |||
2007-05-09 | r300: Added some more Doxygen documentation and made some functions static. | Oliver McFadden | |
2007-05-09 | r300: Removed the "texmicrotile" variable; the tiling code is disabled via a | Oliver McFadden | |
compiler conditional anyway; probably broken? | |||
2007-05-09 | r300: Converted a few "if (0)" into "if (RADEON_DEBUG & DEBUG_TEXTURE)". | Oliver McFadden | |
2007-05-09 | Tweak the shell scripting for descending into and building subdirs. | Brian | |
In general, use this: @for dir in $(SUBDIRS) ; do \ if [ -d $$dir ] ; then \ (cd $$dir && $(MAKE)) || exit 1; \ fi \ done Basically, silently skip missing subdirs but generate an error and stop if there's a compilation or install problem. This was done inconsistantly before. In once case, a missing subdir was causing us to go into an infinte loop! | |||
2007-05-09 | r300: Cleaned up radeon_context.h slightly; mostly Indent. | Oliver McFadden | |
2007-05-09 | r300: radeon_span.h is now common code; radeon_span.c still needs some work. | Oliver McFadden | |
2007-05-09 | r300: Renamed the R300 memory manager from "radeon" to "r300"; it's R300 | Oliver McFadden | |
specific. | |||
2007-05-09 | r300: Indented radeon_span.h | Oliver McFadden | |
2007-05-09 | r300: Further reduced the radeon_span.c diff. | Oliver McFadden | |
2007-05-09 | r300: I have no idea why this was forced on, but it seems to work fine when | Oliver McFadden | |
restored to the previous behaviour. | |||
2007-05-09 | r300: Removed the dangerous R300_SPAN_DISABLE_LOCKING flag. | Oliver McFadden | |
Enabling R300_SPAN_DISABLE_LOCKING would probably cause more lockups; I think there are a couple of other little bugs in this code, too. It's best to remove it. |