summaryrefslogtreecommitdiffstats
path: root/src
Commit message (Collapse)AuthorAgeFilesLines
* Code reorganization: move files into their places.José Fonseca2008-02-15342-0/+0
| | | | This is in a separate commit to ensure renames are properly preserved.
* vastly simplify indexingZack Rusin2008-02-154-130/+59
|
* redo indirectionZack Rusin2008-02-154-31/+130
| | | | make all load's respect indirection
* redo the way immediates are handledZack Rusin2008-02-155-57/+190
| | | | | implement madd start implementing arl
* rename 'argument' to 'load' because that's what it doesZack Rusin2008-02-153-14/+14
|
* implement swizzling on writesZack Rusin2008-02-153-54/+56
|
* a call was missed during the last renameZack Rusin2008-02-151-1/+1
|
* gallium: Fix memory leak.Michal Krol2008-02-151-0/+3
|
* gallium: Fix memory leak.Michal Krol2008-02-151-0/+1
|
* gallium: Use align_free() to free aligned memories.Michal Krol2008-02-151-2/+2
|
* gallium: Fix build on Windows.Michal Krol2008-02-153-3/+3
|
* gallium: Use MALLOC().Michal Krol2008-02-151-2/+2
|
* gallium: Fix MSVC compiler warnings.José Fonseca2008-02-151-0/+4
|
* gallium: remove some debug assertions in vertex format validationBrian2008-02-141-12/+11
| | | | | | | If a fragment shader references an input for which there's no vertex shader output (ex: texcoord3), use vertex output 0 by default. Basically, the attribute's value will be undefined. The shader writer should never rely on undefined fragment shader inputs anyway.
* gallium: rearrange vertex info/layout validationBrian2008-02-144-76/+134
| | | | | | | Delay validation until someone really needs the vertex layout (vbuf alloc vertex buffer or point/line/tri setup/rendering). This will allow the vertex size to change depending on whether we're drawing points, lines or triangles.
* gallium: include draw_context.h to silence warningBrian2008-02-141-0/+1
|
* gallium: minor function renamingBrian2008-02-141-2/+2
|
* gallium: new tgsi_transform_shader() functionBrian2008-02-142-0/+292
| | | | | Used to apply transformations to TGSI shaders, such as register search and replace.
* gallium: call draw_flush() in softpipe_flush()Brian2008-02-141-0/+2
| | | | Without this, we might not get any rendering at SwapBuffers time when using the vbuf path.
* gallium: fix some "instruction"/"declaration" mix-ups in tgsi_exec_prepare().Brian2008-02-141-3/+3
|
* gallium: replace "interpolate" terminology with "eval" to better reflect ↵Brian2008-02-141-11/+22
| | | | what's being done.
* gallium: changes to polygon mode weren't detected in draw_unfilled stage.Brian2008-02-141-0/+2
| | | | | Need to reset stage->tri = unfilled_first_try in unfilled_flush() so that the front/back state is picked up.
* gallium: rename draw_free_tmps->draw_free_temp_verts, ↵Brian2008-02-1410-20/+20
| | | | draw_alloc_tmps->draw_alloc_temp_verts
* galllium: comments, minor clean-upsBrian2008-02-143-4/+22
|
* More name typeo fixes.Ian Romanick2008-02-142-16/+16
|
* s/spu_/spe_/g Ugh.Ian Romanick2008-02-141-99/+99
|
* Cell: pass pointers to stored memory valuesIan Romanick2008-02-142-65/+98
| | | | | | | | | | Several routines use shuffle patterns that are stored in memory. For code gen, it is difficult to directly access the data segments. The routines have been modified to be passed a pointer to a global table of shuffle patterns. This *should* be the last change to this file before switching over to code gen.
* Cell: Fix unintended breakage from commit ↵Ian Romanick2008-02-141-1/+1
| | | | 1d62a057bcb3ee4ef6ebedd93f62ed2e0d8061ba
* Real-time assembler for Cell SPE.Ian Romanick2008-02-143-0/+662
|
* Fetch routines convert and transpose all 4 vertices at once.Ian Romanick2008-02-142-83/+71
|
* Remove open-coded fetch functionsIan Romanick2008-02-141-30/+4
|
* Initial pass at vertex cache, more vertex fetch changesIan Romanick2008-02-142-119/+368
| | | | This is just another step towards dynamic generate of vertex fetch code.
* Vectorize a couple fetch functionsIan Romanick2008-02-141-4/+42
| | | | | | These C-coded vectorized functions are expected to be short lived. They're basically a proof-of-concept for dynamically generated fetch routines.
* intel_winsys: Call st_notify_swapbuffers_complete() after buffer swap.Michel Dänzer2008-02-141-0/+1
|
* gallium: Cleanups related to clears.Keith Whitwell2008-02-141-32/+85
|
* make swizzling on incoming arguments workZack Rusin2008-02-133-23/+67
|
* get rid of the terrible auto-generated entry pointZack Rusin2008-02-134-1002/+3
| | | | it's not used anymore.
* implement immediates and make them workZack Rusin2008-02-132-1/+71
|
* try to load the consts correctlyZack Rusin2008-02-132-0/+22
|
* x86: remove debugKeith Whitwell2008-02-131-2/+0
|
* x86: emit absolute calls, as reallocating exec mem breaks relative onesKeith Whitwell2008-02-133-2/+19
|
* tgsi: fail gracefully when sse code can't emit shaderKeith Whitwell2008-02-131-74/+87
|
* x86: reallocate exec mem when we run outKeith Whitwell2008-02-132-35/+61
|
* tgsi: include more of the register info in debug dumpsKeith Whitwell2008-02-131-0/+10
|
* gallium: pipe->surface_copy can flip the contents vertically when necessary.Michel Dänzer2008-02-1315-11/+28
| | | | Fixes gears being upside down on the box in demos/gearbox.
* handle temporaries in llvm code generated pathsZack Rusin2008-02-136-25/+50
|
* implement addZack Rusin2008-02-131-4/+9
|
* implement mulZack Rusin2008-02-132-1/+21
|
* get mov workingZack Rusin2008-02-134-9/+12
|
* fix fetching input elements and generate a ret on endZack Rusin2008-02-133-3/+6
|