summaryrefslogtreecommitdiff
path: root/src/gallium/drivers/i965
AgeCommit message (Collapse)Author
2009-12-24i965g: strict aliasing changesKeith Whitwell
2009-12-24i965g: calculate depth min/maxKeith Whitwell
Previously hard-wired to 0..1
2009-12-24gallium: adapt drivers to pipe_constant_buffer removalRoland Scheidegger
2009-12-23i965g: Rename {pf->util_format}_is_depth_or_stencil()Tomas Carnecky
2009-12-23[i965] Rename {pf -> util}_format_is_depth_or_stencil()Tomas Carnecky
The function was moved and renamed in 0bed834b. Signed-off-by: Tomas Carnecky <tom@dbservice.com>
2009-12-22i965g: fix for edgeflag changes (untested)Roland Scheidegger
2009-12-22i965g: fix bo_referenceKeith Whitwell
2009-12-22i965g: update for u_format changesKeith Whitwell
2009-12-21i965g: fixes to build after merge of masterKeith Whitwell
2009-12-21i965g: remove half-finished change to tgsi_parseKeith Whitwell
2009-12-21i965g: hook vertex state emit up to PIPE_NEW_VERTEX_BUFFERKeith Whitwell
2009-12-21i965g: keep refcounts to bound vertex buffersKeith Whitwell
2009-12-21i965g: add DEBUG_WINSYS flagKeith Whitwell
2009-12-21i965g: remove duplicate set_viewport_state, fixes samples/depthKeith Whitwell
2009-12-10gallium: Silence compiler warnings.michal
2009-12-10i965: Fix after sampler view changes.michal
2009-12-01i965g: nasty hack for clearing y-tiled surfacesKeith Whitwell
2009-12-01i965g: add missing headerKeith Whitwell
2009-12-01i965g: don't reference unused vars in RSQKeith Whitwell
Probably overly pedantic
2009-11-30i965g: turn on texture tiling by defaultKeith Whitwell
2009-11-30i965g: pass backbuffer tiling information to driverKeith Whitwell
The gem winsys gets this information, needs to pass it on.
2009-11-30brw: add dumping to gem winsysKeith Whitwell
2009-11-30i965g: remove redundant nr_attrs memberKeith Whitwell
2009-11-29i965g: remove surface from views list before freeingDave Airlie
this fixes a crash with the xorg state tracker, however it then locks up the GPU once rendering is enabled but at least it doesn't crash. Signed-off-by: Dave Airlie <airlied@linux.ie>
2009-11-23i965g: gs nr_attrs also tracks nr fragment shader inputsKeith Whitwell
Or, equivalently, nr of outputs of active vertex shader varient.
2009-11-23i965g: correct test for unfilled modesKeith Whitwell
2009-11-23i965g: use correct key size for vs uploadKeith Whitwell
2009-11-23i965g: apply linear math to both linear and perspective attrsKeith Whitwell
2009-11-21i965g: get basic texturing working againKeith Whitwell
Revert to fixed-layout surface binding table -- it's probably the best way to do this. Pass sampler and texture numbers separately even though we're always keeping them the same at present.
2009-11-19i965g: fix typo converting wm src regsKeith Whitwell
2009-11-19i965g: fragment shader immediates workingKeith Whitwell
2009-11-19i965g: first pass at vs immediates in curbeKeith Whitwell
2009-11-19i965g: get linear vs perspective interpolation working againKeith Whitwell
2009-11-19i965g: include interpolation info in fs signatureKeith Whitwell
2009-11-19i965g: special case setup when fs has no inputsKeith Whitwell
2009-11-19i965g: add new state flag tracking fs signature changesKeith Whitwell
2009-11-19i965g: fix typo in previous commitKeith Whitwell
2009-11-19i965g: setup nr_attrs should track nr fragment shader inputsKeith Whitwell
This would be equivalent to taking the nr outputs of the vp varient
2009-11-19i965g: get fragment constants sort-of workingKeith Whitwell
2009-11-19i965g: make the load-before-use vs immediate path workKeith Whitwell
2009-11-19i965g: fix initialization of texture width/height/depth arraysKeith Whitwell
Will remove these arrays in another branch - they're completely redundant.
2009-11-18i965g: remove more references to nr_vp_outputs in wm compilationKeith Whitwell
We're really more concerned about how many inputs the fragment shader is expecting.
2009-11-17i965g: handle special vs outputs speciallyKeith Whitwell
Where vs output semantic tags indicate an output is signficant for fixed function processing (such as clipping, unfilled modes, etc), retain information about that output so that we can get to it easily later on. Fix up the unfilled processing, but hard-wire edgeflag to one for now. With this change, trivial/tri-unfilled works.
2009-11-17i965g rename offset_edge to offset_edgeflagKeith Whitwell
2009-11-14i965g: restore check on line smooth stateKeith Whitwell
2009-11-11i965g: consult fs inputs when laying out vs output regsKeith Whitwell
Vertex shader now emits just the FS inputs, in the positions and order expected by the fragment shader. This means potentially regenerating the vertex shader to match different fragment shader's input layouts.
2009-11-11i965g: skip over vertex position output when preallocating FS inputsKeith Whitwell
2009-11-11i965g: some more cases where file_max[] is used without obligatory +1Keith Whitwell
2009-11-11i965g: smaller upload buffer for index dataKeith Whitwell
2009-11-11i965g: fix off-by-one on curbe uploadKeith Whitwell