aboutsummaryrefslogtreecommitdiffstats
path: root/src/mesa
Commit message (Collapse)AuthorAgeFilesLines
* r300: rewrite and hopefully simplify RS setupMaciej Cencora2009-03-063-213/+225
| | | | | | Testing and regression fixes by Markus Amsler Signed-off-by: Nicolai Haehnle <[email protected]>
* r300: add few macros for RS setupMaciej Cencora2009-03-061-0/+6
| | | | Signed-off-by: Nicolai Haehnle <[email protected]>
* r300: silence valgrindMaciej Cencora2009-03-061-1/+1
| | | | Signed-off-by: Nicolai Haehnle <[email protected]>
* r300: Print reg address when debugging is enabledMaciej Cencora2009-03-061-4/+14
| | | | Signed-off-by: Nicolai Haehnle <[email protected]>
* r300: don't crash on sw tcl hw if point size vertex attrib is sentMaciej Cencora2009-03-061-2/+2
|
* mesa: Reads must also be done with lock held.José Fonseca2009-03-061-3/+5
| | | | Otherwise two threads might think each made the refcount go zero.
* mesa: Fix typo.José Fonseca2009-03-061-1/+1
| | | | Windows threads block if one over-unlocks them.
* intel: Fix bpp setting of blits to 8bpp targets.Eric Anholt2009-03-051-0/+2
| | | | | This was causing hangs in cairogears, as we would blit to the 8bpp target (A8 texture) as 16bpp, and stomp over state objects.
* i965: fix 3DPRIMITIVE batch decode of the vertex count field.Eric Anholt2009-03-051-1/+1
|
* i965: Stop dumping programs after the first all-zeroes entry.Eric Anholt2009-03-051-0/+8
|
* intel: Add always_flush_batch driconf option for making small batchbuffers.Eric Anholt2009-03-056-1/+25
| | | | | This can improve debugging with INTEL_DEBUG=batch,sync by giving smaller batchbuffers.
* intel: Add always_flush_cache driconf option for debugging cache flush failure.Eric Anholt2009-03-057-2/+42
| | | | | I keep wanting to hack this knob in as a one-time thing, so it seemed useful to have all the time.
* i965: Add a note about why the _NEW_STENCIL is required in draw_buffers.Eric Anholt2009-03-051-0/+5
|
* intel: Remove a gratuitous MI_FLUSH after clearing with a blit.Eric Anholt2009-03-051-1/+0
| | | | | The 3D destination shares the same cache so we don't have any trouble with the later commands needing the writes flushed inside of the same batchbuffer.
* i965: Remove dead flushing code.Eric Anholt2009-03-054-23/+0
|
* i965: comments and formatting fixesBrian Paul2009-03-051-4/+14
|
* i965: fix emit_math1() function used for scalar instructionsBrian Paul2009-03-051-9/+32
| | | | | | | | | Instructions such as RCP, RSQ, LOG must smear the result of the function across the dest register's X, Y, Z and W channels (subject to write masking). Before this change, only the X component was getting written. Among other things, this fixes cube map texture sampling in GLSL shaders (since cube lookups involve normalizing the texcoord).
* mesa: added some assertionsBrian Paul2009-03-051-0/+2
|
* mesa: when printing src regs, use |reg| for absolute valueBrian Paul2009-03-051-7/+16
| | | | And check opcode number to avoid crashing on driver-private opcodes.
* i965: fix screen depth test in intel_validate_framebuffer)_Brian Paul2009-03-051-1/+2
| | | | front_region may be null.
* i965: init dest reg CondMask = COND_TR (the proper default)Brian Paul2009-03-051-2/+2
| | | | Plus fix up a debug printf.
* i965: add software fallback for conformant 3D textures and GL_CLAMPRobert Ellison2009-03-044-10/+40
| | | | | | | | | | | | | | | | | The i965 hardware cannot do GL_CLAMP behavior on textures; an earlier commit forced a software fallback if strict conformance was required (i.e. the INTEL_STRICT_CONFORMANCE environment variable was set) and 2D textures were used, but it was somewhat flawed - it could trigger the software fallback even if 2D textures weren't enabled, as long as one texture unit was enabled. This fixes that, and adds software fallback for GL_CLAMP behavior with 1D and 3D textures. It also adds support for a particular setting of the INTEL_STRICT_CONFORMANCE environment variable, which forces software fallbacks to be taken *all* the time. This is helpful with debugging. The value is: export INTEL_STRICT_CONFORMANCE=2
* mesa: call _mesa_get_cpu_string() to get CPU info for GL_RENDERER stringBrian Paul2009-03-041-66/+8
|
* mesa: call _mesa_get_cpu_features() during one-time-initBrian Paul2009-03-041-0/+3
|
* mesa: new cpuinfo.c file to init/query cpu info and extensionsBrian Paul2009-03-044-0/+158
|
* mesa: _mesa_get_x86_features() only needs to do its thing onceBrian Paul2009-03-041-0/+8
|
* mesa: Add x86/common_x86.c to libmesagallium.aJakob Bornecrantz2009-03-051-0/+1
| | | | Thanks to Brian for the refactor commit
* mesa: move sse.h and 3dnow.h #includesBrian Paul2009-03-043-44/+9
|
* mesa: remove unneeded x86.h headerBrian Paul2009-03-042-2/+0
|
* mesa: make _mesa_init_x86_transform_asm() staticBrian Paul2009-03-042-2/+2
|
* mesa: rename common_x86_macros.h -> x86_xform.hBrian Paul2009-03-045-6/+6
|
* mesa: rename x86.c -> x86_xform.cBrian Paul2009-03-043-2/+2
|
* mesa: refactor x86 codeBrian Paul2009-03-043-52/+77
| | | | | | Move _mesa_init_all_x86_transform_asm() into x86.c so that common_x86.c has no dependencies on the vertex transformation code. Plus some comments and clean-ups.
* mesa: Follow ARB_map_buffer_range more stricly.José Fonseca2009-03-043-12/+37
| | | | | Namelly, FlushMappedBufferRange takes a subrange relative to the original range.
* Merge commit 'origin/master' into gallium-map-rangeJosé Fonseca2009-03-042-1/+10
|\
| * vbo: second attempt - avoid getting buffer_ptr and buffer_map out of syncKeith Whitwell2009-03-042-1/+10
| |
| * mesa: Clear cache->trans after destroying the transfer.José Fonseca2009-03-041-0/+1
| |
| * Revert "vbo: avoid getting buffer_ptr and buffer_map out of sync"Keith Whitwell2009-03-042-8/+1
| | | | | | | | | | | | This fixed a minor bug but broke everything else. This reverts commit 579ef8ab1984d895867f547afa60b3bec4c4599a.
| * vbo: avoid getting buffer_ptr and buffer_map out of syncKeith Whitwell2009-03-042-1/+8
| |
* | mesa: Clear cache->trans after destroying the transfer.José Fonseca2009-03-041-0/+1
| |
* | Merge commit 'origin/master' into gallium-map-rangeJosé Fonseca2009-03-0423-86/+104
|\|
| * vbo: use FLUSH_UPDATE_CURRENT flag to indicate whether the vbo module is activeKeith Whitwell2009-03-041-6/+4
| | | | | | | | | | Add asserts for expected values on wakeup and flush. Remove cases where this flag is set or cleared except when waking up and flushing vbo module.
| * mesa: include mfeatures.hBrian Paul2009-03-041-0/+3
| | | | | | | | See bug 20319.
| * gallium: Remove some superfluous instances of #include "p_inlines.h".Michel Dänzer2009-03-046-6/+0
| |
| * gallium: Unify reference counting.Michel Dänzer2009-03-0415-62/+58
| | | | | | | | | | | | | | | | | | | | | | | | | | | | The core reference counting code is centralized in p_refcnt.h. This has some consequences related to struct pipe_buffer: * The screen member of struct pipe_buffer must be initialized, or pipe_buffer_reference() will crash trying to destroy a buffer with reference count 0. u_simple_screen takes care of this, but I may have missed some of the drivers not using it. * Except for rare exceptions deep in winsys code, buffers must always be allocated via pipe_buffer_create() or via screen->*buffer_create() rather than via winsys->*buffer_create().
| * mesa: include main/dd.h in t_vb_points.c and t_vp_build.cBrian Paul2009-03-032-0/+3
| | | | | | | | | | For some reason makedepend doesn't seem to find the dependency on this header in these two files. Directly including the header is a work-around.
| * gallium: remove old state tracker renderbuffer attachment assertionBrian Paul2009-03-031-2/+0
| | | | | | | | | | Not needed since renderbuffers are ref-counted. Fixes progs/demos/shadowtex.
| * xlib: code to force fixed function -> shader translation (for debug, disabled)Brian Paul2009-03-031-0/+8
| |
| * mesa: fix sw fallback state validation bugBrian Paul2009-03-031-10/+28
| | | | | | | | | | | | | | | | When a hw driver fell back to swrast, swrast wasn't always getting informed of program changes. When fixed function is translated into shaders, flags like _NEW_LIGHT, _NEW_TEXTURE, etc. should really signal _NEW_PROGRAM. In this case, swrast wasn't seeing _NEW_PROGRAM when new fragment shaders were generated.
* | mesa: Implement and use FlushMappedBufferRange.José Fonseca2009-03-042-2/+20
| |
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-05 00:53:12 +0000