Age | Commit message (Collapse) | Author | |
---|---|---|---|
2009-04-17 | i965: updated CURBE allocation code | Brian Paul | |
Now that we have real constant buffers, the demands on the CURBE are lessened. When we use real VS/WM constant buffers we only use the CURBE for clip planes. | |||
2009-04-16 | Merge branch 'register-negate' | Brian Paul | |
2009-04-16 | intel: Add support for argb1555, argb4444 FBOs and fix rgb565 fbo readpixels. | Eric Anholt | |
Also enable them all regardless of screen bpp, as 32 bpp what I've been testing against, and haven't been able to detect any screen bpp-specific troubles with them. | |||
2009-04-16 | i965: disable using immediate values for MOV instructions | Brian Paul | |
For some reason, MOV instructions using immediate src values don't seem to work reliably on the GLSL path. Disable them for now (falling back to const buffer reads). This fixes a bunch of glean glsl1 failures. | |||
2009-04-16 | i965: minor debug output changes | Brian Paul | |
2009-04-16 | i965: const buffer debug code (disabled) | Brian Paul | |
2009-04-16 | i965: implement relative addressing for VS constant buffer reads | Brian Paul | |
A scatter-read should be possible, but we're just using two READs for the time being. | |||
2009-04-16 | i965: handle address reg in get_dst() | Brian Paul | |
2009-04-16 | i965: fix const buffer temp register clobbering | Brian Paul | |
Calls to release_tmps() were causing the temps holding constants to get recycled. | |||
2009-04-15 | i965: Clean up output of WM SS state dump, and add format output. | Eric Anholt | |
2009-04-14 | mesa: merge the prog_src_register::NegateBase and NegateAbs fields | Brian Paul | |
There's really no need for two negation fields. This came from the GL_NV_fragment_program extension. The new, unified Negate bitfield applies after the absolute value step. | |||
2009-04-14 | i965: fix VS constant buffer reads | Brian Paul | |
This mostly came down to finding the right MRF incantation in the brw_dp_READ_4_vs() function. Note: this feature is still disabled (but getting close to done). | |||
2009-04-14 | i965: checkpoint commit: VS constant buffers | Brian Paul | |
Hook up a constant buffer, binding table, etc for the VS unit. This will allow using large constant buffers with vertex shaders. The new code is disabled at this time (use_const_buffer=FALSE). | |||
2009-04-10 | i965: remove unused var | Brian Paul | |
2009-04-10 | i965: clean-up in prepare_wm_surfaces() | Brian Paul | |
2009-04-10 | i965: added null const_buffer pointer check in update_constant_buffer() | Brian Paul | |
2009-04-09 | i965: re-org of some of the new constant buffer code | Brian Paul | |
Plus, begin the new code for vertex shader const buffers. | |||
2009-04-09 | i965: new SURF_INDEX_ macros | Brian Paul | |
Used to map drawables, textures and constant buffers to surface binding table indexes. | |||
2009-04-09 | i965: free shader's constant buffer in brwDeleteProgram() | Brian Paul | |
Fixes mem leak observed with texcombine test. | |||
2009-04-08 | i965: set BRW_MASK_DISABLE flag in "send" instruction in brw_dp_READ_4() | Brian Paul | |
This fixes the random results that were seen when fetching a constant inside an IF/ELSE clause. Disabling the execution mask ensures that all the components of the register are written. | |||
2009-04-08 | i965: clean-ups, debug code in brw_wm_glsl.c | Brian Paul | |
2009-04-08 | i965: init current_const[i].index = -1 | Brian Paul | |
2009-04-08 | i965: move the fetch_constants() call before setting conditional mod state | Brian Paul | |
Before, the instruction's CondUpdate field was mistakenly effecting the constant-fetch operation. Fixes progs/glsl/bump.c demo. But there are some other issues related to condition flags and IF/ELSE that need investigation... | |||
2009-04-06 | i965: Use GTT maps when available to upload vertex arrays and system VBOs. | Eric Anholt | |
This speeds up OA on my GM45 by 21% (more than the original CPU cost of the upload path). We might still be able to squeeze a few more percent out by avoiding repeatedly mapping/unmapping buffers as we upload elements into them. | |||
2009-04-03 | i965: remove unused var | Brian Paul | |
2009-04-03 | i965: more const buffer debug code | Brian Paul | |
2009-04-03 | i965: added brw_same_reg() | Brian Paul | |
2009-04-03 | i965: s/GL_FALSE/BRW_COMPRESSION_NONE/ | Brian Paul | |
2009-04-03 | i965: another checkpoint commit of new constant buffer support | Brian Paul | |
Everything is in place now for using a true constant buffer for GLSL fragment shaders. Still some bugs to find though. | |||
2009-04-03 | i965: fix response length param in brw_dp_READ_4() | Brian Paul | |
We were accidentally clobbering the next register. | |||
2009-04-03 | i965: change args to get_src_reg() to prep for new constant buffer support | Brian Paul | |
2009-04-03 | i965: check-point commit of new constant buffer support | Brian Paul | |
Currently, shader constants are stored in the GRF (loaded from the CURBE prior to shader execution). This severly limits the number of constants and temps that we can support. This new code will support (practically) unlimited size constant buffers and free up registers in the GRF. We allocate a new buffer object for the constants and read them with "Read" messages/instructions. When only a small number of constants are used, we can still use the old method. The code works for fragment shaders only (and is actually disabled) for now. Need to do the same thing for vertex shaders and need to add the necessary code-gen to fetch the constants which are referenced by the shader instructions. | |||
2009-04-03 | i965: remove unused code for sampling a constant buffer | Brian Paul | |
2009-04-03 | i965: code to setup a constant buffer sampler | Brian Paul | |
This code won't actually be used and will be removed in a subsequent commit. Just committing for posterity. | |||
2009-04-03 | i965: do negation and Abs in get_src_reg_imm() | Brian Paul | |
Fixes regression seen with progs/glsl/bump.c | |||
2009-04-03 | i965: comments | Brian Paul | |
2009-04-03 | i965: code to debug/dump instruction immediates | Brian Paul | |
2009-04-03 | i965: minor code movement, new comment | Brian Paul | |
2009-04-03 | i965: added new brw_dp_READ_4() function | Brian Paul | |
Used to read float[4] vectors from the constant buffer/surface. | |||
2009-04-03 | i965: new and updated comments | Brian Paul | |
2009-04-03 | i965: comments for brw_SAMPLE() | Brian Paul | |
2009-04-03 | i965: comments for sampling code | Brian Paul | |
2009-04-03 | i965: add support for float literal instruction operands | Brian Paul | |
Call the get_src_reg_imm() function when it's permissible to generate a literal value src register. | |||
2009-04-03 | i965: remove 'nr' param from get_src/dst_reg() functions | Brian Paul | |
The value was always 1. | |||
2009-04-03 | i965: fix comment typo | Brian Paul | |
2009-04-03 | i965: comments, clean-up in prepare_wm_surfaces() | Brian Paul | |
2009-04-03 | i965: remove unneeded #includes | Brian Paul | |
2009-04-03 | i965: formatting clean-ups | Brian Paul | |
2009-04-03 | i965: comments, whitespace changes | Brian Paul | |
2009-04-03 | i965: rename scratch_buffer -> scratch_bo to be consistant with other buffers | Brian Paul | |