Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | texenvprogram: Mark textures using ARB_shadow as ShadowSamplers | Nicolai Haehnle | 2008-06-14 | 1 | -2/+6 |
| | | | | | | | Since ARB_fragment_program and friends are defined to ignore the setting of the GL_TEXTURE_COMPARE_FUNC parameter, we have to explicitly enable the shadow comparison by marking the texture unit in ShadowSamplers when appropriate. | ||||
* | [texenvprogram] Fix refcounting-related memory leak. | Nicolai Haehnle | 2008-06-02 | 1 | -1/+2 |
| | | | | | | | | All newly created programs have RefCount == 1, but the fragment program cache added an additional reference in cache_item, with the result being that none of the programs were ever freed. Solve the problem by not creating the additional reference in cache_item. | ||||
* | fix refcounting bugs in tnl/tex program caches | Brian Paul | 2008-05-07 | 1 | -26/+33 |
| | |||||
* | replace // comment with /* */ (bug 14916) | Brian | 2008-03-09 | 1 | -1/+1 |
| | |||||
* | i965: use _Current pointer instead of Current pointer. | Xiang, Haihao | 2008-02-29 | 1 | -1/+2 |
| | | | | fix double free issue(bug#14710). It also corrects glsl/bitmap demo behavior. | ||||
* | fix bug with generated fragment programs if vertex shader is active | Roland Scheidegger | 2008-02-15 | 1 | -2/+2 |
| | | | | | | | When generating a fragment program from fixed function, checking texUnit->_ReallyEnabled is not sufficient, need texUnit->Enabled too since the former also represents texture enables from an active vertex shader. | ||||
* | Merge branch '965-glsl' | Zou Nan hai | 2007-10-26 | 1 | -1/+1 |
|\ | | | | | | | | | | | | | Conflicts: src/mesa/drivers/dri/i965/brw_sf.h src/mesa/drivers/dri/i965/intel_context.c | ||||
| * | Only vertex program fix, bypass tnl vertex program | Zou Nan hai | 2007-10-08 | 1 | -1/+1 |
| | | |||||
* | | properly init dst reg's CondMask/Swizzle fields | Brian | 2007-10-23 | 1 | -2/+2 |
| | | |||||
* | | bump up MAX_INSTRUCTIONS and add an assertion to catch emitting too many ↵ | Brian | 2007-10-23 | 1 | -4/+7 |
| | | | | | | | | instructions | ||||
* | | Brian's fix for bug9829 | Xiang, Haihao | 2007-08-17 | 1 | -1/+1 |
|/ | |||||
* | Be more consistant with paths in #includes. Eventually, eliminate a bunch ↵ | Brian | 2007-07-04 | 1 | -4/+4 |
| | | | | of -I flags. | ||||
* | fix GL_DOT3_RGBA texture combiner mode in generated fragment programs (bug ↵ | Roland Scheidegger | 2007-07-03 | 1 | -1/+1 |
| | | | | #11030) | ||||
* | merge of glsl-compiler-1 branch | Brian | 2007-03-26 | 1 | -40/+58 |
|\ | |||||
| * | Undo some changes to _mesa_UpdateTexEnvProgram(). Fixes broken i915 texturing. | Brian | 2007-02-24 | 1 | -3/+9 |
| | | |||||
| * | s/GLint/gl_state_index/, length is now 5 not 6 | Brian | 2007-02-23 | 1 | -9/+7 |
| | | |||||
| * | fix g++ warnings/errors | Brian | 2007-01-23 | 1 | -2/+3 |
| | | |||||
| * | Moved NumTexInstructions, NumTexIndirections, etc. into gl_program since | Brian | 2007-01-09 | 1 | -9/+9 |
| | | | | | | | | they can now apply to vertex programs. | ||||
| * | Lots of assorted changes for new GLSL compiler backend. | Brian | 2006-12-15 | 1 | -26/+40 |
| | | | | | | | | New datatypes, constants, variables. | ||||
* | | use _mesa_copy_instructions() | Brian | 2007-03-22 | 1 | -4/+3 |
| | | |||||
* | | fixes for C++ warnings/errors | Brian | 2007-01-23 | 1 | -2/+3 |
|/ | |||||
* | Add a size parameter to _mesa_add_unnamed_constant() and | Brian Paul | 2006-11-15 | 1 | -1/+1 |
| | | | | | _mesa_add_named_constant() to indicate vector size (1, 2, 3 or 4). Always 4 for now... | ||||
* | remove debug code | Brian Paul | 2006-10-30 | 1 | -7/+0 |
| | |||||
* | Rather than always allocating 100 instructions, put instructions into a | Brian Paul | 2006-10-29 | 1 | -18/+42 |
| | | | | | temporary buffer, then copy into a malloc'd array of the correct size. According to Glean texCombine, we never need more than 21 instructions. | ||||
* | Change _mesa_init_instruction() to initialize an array of instructions. | Brian Paul | 2006-10-29 | 1 | -1/+1 |
| | |||||
* | Fix some breakage after the PROGRAM_UNDEFINED change | Alan Hourihane | 2006-08-10 | 1 | -2/+2 |
| | |||||
* | Some structure renaming. Prefix vertex/fragment-related structs with | Brian Paul | 2006-07-20 | 1 | -8/+9 |
| | | | | "gl_" to match other structs. | ||||
* | remove redundant 'const' | Brian Paul | 2006-05-30 | 1 | -1/+1 |
| | |||||
* | Add const qualifiers in a number of places. | Brian Paul | 2006-05-23 | 1 | -11/+12 |
| | |||||
* | In cache_item(), there was sizeof(void) expression. | Brian Paul | 2006-05-23 | 1 | -1/+1 |
| | | | | Replace void *key parameter with const struct state_key * | ||||
* | memory usage fixes for glean/conform, use a better hash function | Keith Whitwell | 2006-05-22 | 1 | -29/+49 |
| | |||||
* | Fix typo | Keith Whitwell | 2006-05-22 | 1 | -1/+1 |
| | |||||
* | Better caching for texenv programs. | Keith Whitwell | 2006-05-22 | 1 | -15/+60 |
| | | | | Initialize some values correctly. | ||||
* | casts to fix some -pedantic warnings | Brian Paul | 2006-04-13 | 1 | -4/+4 |
| | |||||
* | fix #6072 (Keith Whitwell) | Aapo Tahkola | 2006-03-28 | 1 | -12/+17 |
| | |||||
* | fix using a tmp without asking for one | Jerome Glisse | 2006-01-24 | 1 | -0/+1 |
| | |||||
* | fix broken get_half & get_zero | Aapo Tahkola | 2006-01-24 | 1 | -4/+4 |
| | |||||
* | use GLbitfield instead of GLuint | Brian Paul | 2005-12-14 | 1 | -6/+3 |
| | |||||
* | silence warning, bug 5318 | Brian Paul | 2005-12-12 | 1 | -0/+2 |
| | |||||
* | C++ fixes, mostly casts (Stephane Conversy) | Brian Paul | 2005-12-06 | 1 | -1/+1 |
| | |||||
* | Make Saturate a 2-bit field again, renamed to SaturateMode with three | Brian Paul | 2005-11-20 | 1 | -1/+1 |
| | | | | possible values: SATURATE_OFF, SATURATE_ZERO_ONE and SATURATE_PLUS_MINUS_ONE. | ||||
* | additional clean-ups and improvements | Brian Paul | 2005-11-19 | 1 | -21/+38 |
| | |||||
* | Bunch of little fixes: | Brian Paul | 2005-11-19 | 1 | -35/+31 |
| | | | | | | | | Fix mem leaks in _mesa_TexEnvProgramCacheDestroy(). Check if ctx->Driver.BindProgram is non-null before calling. s/unsigned/GLuint/ Use MAX_INSTRUCTIONS instead of magic 100, check program length after it's made. Use _mesa_init_instruction() instead of _mesa_memset(). | ||||
* | Remove the _mesa_parse_arb_vertex/fragment_program() functions into | Brian Paul | 2005-11-19 | 1 | -1/+0 |
| | | | | | | | arbprogparse.c and remove arbvertparse.[ch] and arbfragparse.[ch]. Clean up quite a bit of the arb parsing code. Rewrite parser extension code to avoid a mess of string operations every time the parser was used. | ||||
* | Reduce stderr noise. | Aapo Tahkola | 2005-11-18 | 1 | -2/+2 |
| | |||||
* | Move stuff common to vertex/fragment_program into the base class, including: | Brian Paul | 2005-11-12 | 1 | -10/+9 |
| | | | | | | Instructions, InputsRead, OutputsWritten, and Parameters. Also, added debug functions: _mesa_print_instruction(), _mesa_print_program_parameters() and revamp _mesa_print_program(). | ||||
* | s/_mesa_debug_fp_inst/_mesa_print_program/ | Brian Paul | 2005-11-08 | 1 | -1/+1 |
| | |||||
* | Unify vertex/fragment program instuctions. | Brian Paul | 2005-11-05 | 1 | -30/+30 |
| | | | | Based on patch by Ian (#4967) but also unify instruction opcodes. | ||||
* | Rename FRAG_OUTPUT_* tokens to FRAG_RESULT_* to match vertex program convention | Brian Paul | 2005-11-02 | 1 | -3/+3 |
| | |||||
* | Several fp and vp tweaks: | Keith Whitwell | 2005-11-01 | 1 | -20/+29 |
| | | | | | | | | | | | | | | | | | | | | | | - Renumber VERT_RESULT_* values so that they match the numbers of the corresponding FRAG_ATTRIB_ values. - Add ctx->VertexProgram._Current and FragmentProgram._Current values which point to either the current client-supplied program if enabled, or otherwise to the current mesa-internal program. Thus this program is always the correct one for the current state, providing that the mesa flags to turn on automatic generation are enabled. - Add callbacks to ctx->Driver.BindProgram() in texenvprogram.c and t_vp_build.c so that the driver knows when the generated program changes. This is cleaner than trying to code all the possible _NEW_* flags into the driver, and more precise as well. - Add a UsesKill flag to identify fragment programs with that instruction, as these can require special treatment. - Move the FRAG_OUTPUT values to mtypes.h, near to similar defn's. |