Age | Commit message (Collapse) | Author | |
---|---|---|---|
2006-07-19 | For glGetTexImage(), move the _mesa_validate_pbo_access() error check into | Brian Paul | |
_mesa_GetTexImage() so it's not needed in the fallback or driver functions. | |||
2006-07-19 | Several loops over the map1/2 evaluator arrays were incorrect. | Brian Paul | |
There's only 16 evaluator maps. Define new _TNL_NUM_EVAL constant for those loops. Also, fix priority of GL_NV_vertex_program generic maps - they override the conventional maps. See bug 7564. | |||
2006-07-18 | indentation fix | Brian Paul | |
2006-07-18 | add missing 'return' after a few _mesa_error() calls in _mesa_GetTexImage() | Brian Paul | |
2006-07-18 | Compute ctx->Color._LogicOpEnabled in one place instead of four places! | Brian Paul | |
2006-07-15 | fix a memleak (that code block has been commented out for years for unknown ↵ | Tilman Sauerbeck | |
reasons) | |||
2006-07-15 | A small logic error. | Rune Petersen | |
We need to call do_swizzle() when the scaler _or_ vector unit have a swizzle other 000/ZERO/111/ONE. This ensures that the gloss on banners in ut2004 is not angel-dependent (since when was banners glossy?). | |||
2006-07-13 | define FLT_MAX_EXP to 128 if undefined | Brian Paul | |
2006-07-13 | Move initialization of server_support from __glXProcessServerString to | Ian Romanick | |
__glXCalcualteUsableExtensions. By having both callers of __glXProcessServerString initialize the bit-fields passed in, we guarantee that the correct amount of data is initialized: This fixes bugzilla #7353. | |||
2006-07-13 | LIB_DIR is now just 'lib' or 'lib64' | Brian Paul | |
Replaced $(LIB_DIR) with $(TOP)/$(LIB_DIR), use LIB_DIR in install targets. Patch by Hanno Böck. | |||
2006-07-12 | New config var DRI_DRIVER_SEARCH_DIR to specify where libGL should search | Brian Paul | |
(at runtime) for the DRI drivers. See bug 2372. | |||
2006-07-12 | fix another memleak | Tilman Sauerbeck | |
2006-07-12 | fix a memleak | Tilman Sauerbeck | |
2006-07-11 | free array objects properly (bug 7493) | Brian Paul | |
2006-07-11 | free GL2Objects unconditionally | Tilman Sauerbeck | |
2006-07-11 | free the ArrayObjects hash table | Tilman Sauerbeck | |
2006-07-11 | only access ss->BufferObjects if it's declared | Tilman Sauerbeck | |
2006-07-11 | _mesa_destroy_context() frees radeon->glCtx so we must not access it afterwards | Tilman Sauerbeck | |
2006-07-11 | Revert the SiS and Unichrome DRM version check changes. | Thomas Hellström | |
2006-07-10 | The SiS and Unichrome drivers assume that the handle to the | Thomas Hellström | |
AGP map is equivalent to the AGP physical address. Fix this. | |||
2006-07-09 | Fix #7195 | Aapo Tahkola | |
2006-07-06 | Don't include unused header files. | Kristian Høgsberg | |
2006-07-05 | R300_RR_ROUTE_1 still needs to be initialized. | Aapo Tahkola | |
2006-07-04 | Add R300_VAP_CNTL 0x2140 and cosmetic cleanup. | Jerome Glisse | |
2006-07-04 | enable hw antialiasing (Martijn van Oosterhout) | Brian Paul | |
2006-07-03 | additional clipping/viewport fixes (Martijn van Oosterhout) | Brian Paul | |
2006-07-03 | some minor simplifications (same as in r300 driver) since mesa's internally ↵ | Roland Scheidegger | |
used swizzle and writemask flags are identical to the bits used on the hardware, use a macro to error out if these assumptions are no longer true. | |||
2006-07-03 | check if radeon_drm.h from libdrm is new enough, otherwise print an error to ↵ | Roland Scheidegger | |
make it easier to figure out why it won't compile... | |||
2006-07-03 | Add copyright header and minor cleanup. | Jerome Glisse | |
2006-07-03 | disable a new mem leak check until we're sure it's valid | Brian Paul | |
2006-07-03 | fix clipping bug on context change (Martijn van Oosterhout) | Brian Paul | |
2006-07-01 | initialize offset to avoid bogus warnings from valgrind | Tilman Sauerbeck | |
2006-06-30 | Vertex/fragment program patches (Rune Petersen) | Brian Paul | |
2006-06-30 | clipping fixes (Martijn van Oosterhout) | Brian Paul | |
2006-06-30 | clean-ups | Brian Paul | |
2006-06-30 | use _mesa_HashDeleteAll() | Brian Paul | |
2006-06-30 | In free_shared_state(), use new _mesa_HashDeleteAll() function to delete | Brian Paul | |
textures, display lists, programs, VBOs, etc. Less code and more efficient. | |||
2006-06-30 | Added _mesa_delete_list() to delete a display list, but don't remove from | Brian Paul | |
hash table. Make _mesa_destroy_list() static. | |||
2006-06-30 | Added _mesa_HashDeleteAll() to delete all entries in a hash table with a | Brian Paul | |
callback function. Added _mesa_HashWalk() to walk over all entries in a hash table with callback. | |||
2006-06-29 | Fix typo. | Jerome Glisse | |
2006-06-29 | destroy left-over VBOs on shutdown | Tilman Sauerbeck | |
2006-06-29 | instead of messing with _TNL_LAST_GENERIC and _TNL_FIRST_GENERIC, use ↵ | Tilman Sauerbeck | |
_TNL_NUM_GENERIC. this fixes an off-by-one error (we didn't handle the final attribute) | |||
2006-06-29 | cleanup, avoid magic values | Tilman Sauerbeck | |
2006-06-29 | added _TNL_NUM_GENERIC | Tilman Sauerbeck | |
2006-06-29 | only assert front/back pitch if double-buffered (bug 7663) | Brian Paul | |
2006-06-28 | propagate errors from parse_generic_attrib_num in parse_attrib_binding (-> ↵ | Tilman Sauerbeck | |
don't segfault when you're trying to access invalid generic attributes | |||
2006-06-27 | Don't use nested struct definitions in tnl_clipspace_fastpath. | Brian Paul | |
Make attr_type into a top-level type: tnl_attr_type See bug 7340. | |||
2006-06-27 | Remove the extraneous GL_ prefix from the enums for GL_EXT_timer_query and | Ian Romanick | |
GL_EXT_framebuffer_blit. Enumerants in the XML schema get the GL_ prefix added to them automatically. This resulted in things like "GL_GL_TIME_ELAPSED_EXT" in enums.c. | |||
2006-06-27 | Fix stencil clears that have caused problems with redbook/stencil, ↵ | Aapo Tahkola | |
tests/stencilwrap and ut2k4 adrenaline pills. | |||
2006-06-27 | Fix generic vertex attribs (Rune Petersen) | Aapo Tahkola | |