summaryrefslogtreecommitdiffstats
path: root/src/mesa/swrast
Commit message (Collapse)AuthorAgeFilesLines
* init swrast->CurrentBufferBrian Paul2002-10-111-1/+6
|
* Another round of glRead/DrawBuffer() clean-ups and simplifications.Brian Paul2002-10-115-67/+40
| | | | | | | Replaced ctx->Color._DriverDrawBuffer with swrast->CurrentBuffer. Replaced ctx->Pixel._DriverReadBuffer with ctx->Pixel._ReadSrcMask. swrast->Driver.SetBuffer() takes FRONT/BACK_LEFT/RIGHT_BIT values now. Added tokens and code for GL_AUX buffers, for completeness.
* finally get rid of ctx->Texture._ReallyEnabled fieldBrian Paul2002-10-081-3/+3
|
* remove const storage class specifier for the decl of a var that isn'tKarl Schultz2002-10-041-2/+2
| | | | a const.
* Changed a number of context fields from GLchan to GLfloat (such as ClearColor).Brian Paul2002-10-044-54/+58
| | | | | | Also changed parameter types for some driver functions (like ctx->Driver.Clear- Color). Updated all the device drivers. Someday, we want to support 8, 16 and 32-bit channels dynamically at runtime.
* multiple GL_POINTS can now be rendered together into one fragment spanBrian Paul2002-10-044-125/+240
|
* finished up GL_EXT_stencil_two_sideBrian Paul2002-10-025-15/+19
|
* new texture compression infrastructureBrian Paul2002-09-271-98/+48
|
* merge gl_texture_image RowStride from DRI 4.0.4Brian Paul2002-09-232-4/+7
|
* GL_MESA_pack_invertBrian Paul2002-09-211-2/+9
|
* updates from 4.0.4 (MESA_ycbcr_texture, APPLE_client_storage, etc)Brian Paul2002-09-211-2/+3
|
* pull-in changes from DRI/Mesa-4.0.4Brian Paul2002-09-172-2/+25
|
* fixed _mesa_generate_mipmaps() target problemBrian Paul2002-09-161-8/+8
|
* GL_EXT_stencil_two_side extension, not 100% complete yet.Brian Paul2002-09-063-87/+95
|
* re-check-in - something didn't work on the previous check-inBrian Paul2002-08-071-164/+164
|
* struct sw_span is again allocated on the stack, but the arrays of spanBrian Paul2002-08-0723-908/+1007
| | | | | | | data are broken out into a new struct span_arrays which is allocated per-context (to avoid huge stack allocations - a problem on Windows). This lets us use span.redStep instead of span->redStep (for example) to hopefully get slightly better performance in the triangle functions.
* Overhaul of glRead/DrawBuffer() code. Now, swrast->Driver.SetBuffer()Brian Paul2002-07-0911-196/+244
| | | | | indicates the read AND draw color buffer for all software rasterization. Lots of related clean-ups. See RELNOTES-4.1 for details.
* blending fixes for CHAN_TYPE==GL_FLOAT (Gerk Huisma)Brian Paul2002-06-301-1/+38
|
* fix CHAN_TYPE==GL_FLOAT compile problemBrian Paul2002-06-261-3/+5
|
* Implemented GL_NV_texture_rectangle extension.Brian Paul2002-06-1511-51/+248
| | | | | | | | | Replace struct gl_texure_object's Dimension w/ Target field. Added _EnabledUnits to struct gl_texture_attrib - the _ReallyEnabled field is obsolete, but still present for now. This effectively removes the 8-texture units limit, 32 units now possible, but unlikely! New TEXTURE_1D/2D/3D/CUBE/RECT_BIT tokens for unit->_ReallyEnabled field. Updated device drivers to use ctx->Texture._EnabledUnits.
* Added ctx parameter to _mesa_debug()Brian Paul2002-06-156-40/+119
| | | | | | Added _mesa_printf() Updated SetDrawBuffer() function in all drivers (ala 4.0.3) Import 4.0.3/DRI changes.
* more removal of fprintf() callsBrian Paul2002-06-131-28/+30
|
* replaced experimental MESA_sprite_point with NV_point_spriteBrian Paul2002-05-272-69/+44
|
* Implemented GL_ARB_texture_env_crossbar.Brian Paul2002-05-026-171/+213
| | | | Simplification of some of the texture application code.
* fixed Width/Height cut&paste typoBrian Paul2002-04-202-6/+6
|
* Added min and max blend functions.Jose Fonseca2002-04-191-3/+15
| | | | Slight optimization of _mesa_mmx_blend_add.
* Allocate a sw_span struct in the swrast context instead of allocating itBrian Paul2002-04-1915-897/+906
| | | | | on the stack frame in the point/line/triangle functions. (Klaus Niederkrueger) This should solve the performance problem Karl found on Windows.
* MMX add blending function added.Jose Fonseca2002-04-191-2/+8
|
* MMX modulate belding function added.Jose Fonseca2002-04-191-2/+8
| | | | Factorization of more common MMX code.
* removed a stray debug printfBrian Paul2002-04-191-2/+1
|
* casts to fix GLint/GLuint mismatchesBrian Paul2002-04-199-63/+74
|
* restored writeAll = GL_FALSE test/assignmentBrian Paul2002-04-181-2/+6
|
* fixes for corrected GL_DOT3_RGB[A]_EXT token valuesBrian Paul2002-04-121-7/+32
|
* updated a few commentsBrian Paul2002-04-121-4/+5
|
* Klaus Niederkrueger's latest changes: use INIT_SPAN() to init span primitiveBrian Paul2002-04-1217-209/+240
| | | | type, width, interp mask and array mask.
* new MMX blend code (Jose Fonseca)Brian Paul2002-04-101-5/+3
|
* Texture combine alpha subtraction was wrong. Fixed 16-bit channel glitchesBrian Paul2002-04-041-8/+18
|
* use NULL lambda array for pixel textureBrian Paul2002-04-041-8/+9
|
* fixed 16 bit/channel problem in blend_modulate()Brian Paul2002-04-041-5/+14
|
* fixed 16/32-bit channel problems in CopyColor[Sub]Table()Brian Paul2002-04-041-5/+5
|
* blend_add() modified for GLfloat channelsBrian Paul2002-04-041-5/+15
|
* more blending updates from JoseBrian Paul2002-04-021-19/+9
|
* faster blending (Jose Fonseca)Brian Paul2002-04-021-1/+13
|
* reset span.end to zero after rendering the spanBrian Paul2002-03-281-1/+2
|
* fixed inequality in an assertionBrian Paul2002-03-271-2/+2
|
* removed a bad assertionBrian Paul2002-03-251-2/+1
|
* Test implementation of proposed GL_EXT_shadow_funcs extension. This justBrian Paul2002-03-231-55/+103
| | | | | generalizes the R/texture comparision operators to include all eight of the depth test comparisons.
* Use MESA_PBUFFER_ALLOC/FREE macros to allocate all framebuffer and textureBrian Paul2002-03-194-19/+20
| | | | | | | memory. These can be overridden by applications which need to manage this memory specially. Contributed by Gerk Huisma. Also, new code for 8-bit -> 16-bit/channel texture image storage which fills in the least-significant bits properly.
* fix Intel C compiler warnings (Gerk Huisma)Brian Paul2002-03-191-3/+3
|
* New mipmap lambda calculation. Previously, trilinear filtering couldBrian Paul2002-03-167-327/+265
| | | | | result in _very_ blurry textures. Still need to do some optimization of the new code in s_span.c