summaryrefslogtreecommitdiffstats
path: root/src/mesa
Commit message (Collapse)AuthorAgeFilesLines
* More GLSL code:Michal Krol2006-04-1139-583/+882
| | | | | | | | | | | | - use macros to access and modify render inputs bit-field; - un-alias generic vertex attributes for ARB vertex calls; - use MAX_VERTEX_PROGRAM_ATTRIBS (NV code) or MAX_VERTEX_ATTRIBS (ARB code) in place of VERT_ATTRIB_MAX; - define VERT_ATTRIB_GENERIC0..15 for un-aliased vertex attributes for ARB_vertex_shader; - fix generic attribute index range check in arbprogparse.c; - interface GLSL varyings between vertex and fragment shader; - use 64-bit optimised bitset (bitset.h) for render inputs;
* get fog going(for real)Aapo Tahkola2006-04-111-1/+0
|
* rectangular fp16 fixes (Ewald Snel)Aapo Tahkola2006-04-111-0/+18
|
* Fog support (Ewald Snel)Aapo Tahkola2006-04-116-24/+177
|
* use _mesa_debug() instead of message() - fixes poor formatting of outputBrian Paul2006-04-101-55/+34
|
* Destroy the GL context after driDestroyTextureHeap, because the driver'sEric Anholt2006-04-093-12/+12
| | | | DestroyTexObj has a dependence on the glCtx and may segfault otherwise.
* fix some conflict happenings at context destroy by freeing buffers ↵Aapo Tahkola2006-04-082-8/+14
| | | | immediately. array locking can be done with vbos for vtxfmt_a so we dont really need to try to reuse them.
* regenerate SPARC assembly dispatch (bug 6484)Brian Paul2006-04-072-6/+17
|
* replace >&/dev/null with >/dev/null so it works with all shellsBrian Paul2006-04-071-1/+1
|
* bump version to 6.5.1Brian Paul2006-04-071-4/+4
|
* Fix some warnings on x86_64Alan Hourihane2006-04-078-10/+16
|
* make miniglx use libdrm properlyDave Airlie2006-04-071-3/+2
|
* Rename CLIP_ALL_BITS to CLIP_FRUSTUM_BITS to reflect the fact that the valueBrian Paul2006-04-067-9/+11
| | | | | only includes the 6 frustum bits, not the user-clip plane bit, nor the vertex cull bit.
* Fix line clipping bug 6512.Brian Paul2006-04-061-24/+28
| | | | This fix is to discard the line if both dot products are negative.
* Set DRIVER_DEFINES instead of DEFINES to fix duplicate flags problemBrian Paul2006-04-063-4/+4
|
* Replace DEFINES with DRIVER_DEFINES. If a driver needs to set extra flags,Brian Paul2006-04-061-4/+4
| | | | | do so with DRIVER_DEFINES. This eliminates the duplicated flags we were passing to each cc command.
* remove unused varBrian Paul2006-04-061-2/+0
|
* update a couple MESA_FORMAT_Zxxx occurancesBrian Paul2006-04-061-2/+2
|
* Replace MESA_FORMAT_DEPTH_COMPONENT_FLOAT32 with 32-bit integer format.Brian Paul2006-04-0612-129/+299
| | | | | | | This allows render to depth texture (we don't support floating pt. Z buffers). Rename MESA_FORMAT_DEPTH_COMPONENT16/32 as MESA_FORMAT_Z16/32. Software fallback for glCopyTexImage now uses integer temporary image instead of float, eliminates a lot of float/int conversions.
* do additional checking for setting up depth/stencil wrappersBrian Paul2006-04-061-3/+7
|
* Remove the code that checks for duplicated framebuffer attachments (whichBrian Paul2006-04-051-21/+1
| | | | | generated the error GL_FRAMEBUFFER_INCOMPLETE_DUPLICATE_ATTACHMENT_EXT). This was removed in version 117 of the GL_EXT_framebuffer_object spec.
* add tiling/fence support for the miniglx driverDave Airlie2006-04-052-20/+292
|
* don't waste the reserved memory, take an approximation of the poolDave Airlie2006-04-051-19/+89
| | | | allocator from the X.org driver - not as complicated
* just some comments/docsBrian Paul2006-04-051-3/+19
|
* Check/use texture border color when sampling depth textures. (bug 6498)Brian Paul2006-04-051-9/+16
| | | | Silence some warnings.
* Silence minor compiler warnings (-Wextra).Brian Paul2006-04-057-12/+22
|
* We always have a back buffer. Fixes visual problems.Alan Hourihane2006-04-041-1/+1
|
* use gart base from drm not try to work it out ourselves.. bump radeon clientDave Airlie2006-04-041-27/+22
| | | | to need 1.6 drm where this first appeared.. this fixes r300 on PCIE
* switch to card_type rather than IsPCI, doesn't change any functionality yetDave Airlie2006-04-045-24/+27
|
* Fix texcoord interpolation for fragment shaders.Michal Krol2006-04-041-9/+12
|
* More GLSL code:Michal Krol2006-04-048-59/+295
| | | | | | | - add support for varyings; GLSL fixes: - pow was wrongly computed in x86 back-end;
* Bit set.Michal Krol2006-04-041-0/+72
|
* Return GLSL version string (1.10).Michal Krol2006-04-041-3/+11
|
* Define MAX_VERTEX_ATTRIBS to 16.Michal Krol2006-04-041-2/+3
|
* Bug #5754: Allocate texture memory correctly. (Richard Drummond)Adam Jackson2006-04-031-3/+4
|
* Fix problem reported by Bill Newman in read_rgba_pixels for 16 and 32 bit ↵Karl Schultz2006-04-031-2/+2
| | | | (GDI driver). Bug would cause the wrong pixel to be read if more than one pixel requested and could cause memory access violations.
* add miniglx sources to i915Dave Airlie2006-04-031-1/+2
|
* if DDX is -1 don't do DDX checkDave Airlie2006-04-031-2/+3
|
* use WidthDave Airlie2006-04-031-0/+1
|
* add pitch support and use new WidthDave Airlie2006-04-031-2/+19
|
* Coverity #459: Silly off-by-one error.Adam Jackson2006-04-031-1/+1
|
* Coverity #476: Avoid walking off the end of ->vtx.attrsz, it's declared toAdam Jackson2006-04-021-1/+1
| | | | | | | | be _TNL_ATTRIB_MAX long so that's not a valid index (woo zero based indexing). This code still looks wrong though, the asymmetry between _tnl_copy_to_current and _tnl_copy_from_current has me worried.
* Coverity #567: Fix a memory leak on a failure path.Adam Jackson2006-04-021-0/+1
|
* Coverity #650: Fix a memory leak when near OOM.Adam Jackson2006-04-021-0/+1
|
* Initialize renderbuffer size.Daniel Borca2006-04-011-0/+3
|
* These messages really should be printed all the time, users want toDave Airlie2006-04-011-5/+5
| | | | know this not have to set debug for it
* Don't index box by the box's index within numClipRects; just dereference it.mesa_6_5Eric Anholt2006-04-011-4/+4
| | | | Fixes drawing with more than one cliprect. (Keith Packard)
* Call driUpdateFramebufferSize when we've found an updated DRI drawable stamp,Eric Anholt2006-04-011-0/+2
| | | | | like other drivers. Failure to do so resulted in incorrect buffer sizes for resized windows.
* DMesa now uses OSMesa as a back-end.Daniel Borca2006-03-3114-2396/+1673
|
* Enough testing alreadyAapo Tahkola2006-03-311-1/+0
|
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-16 23:27:40 +0000