summaryrefslogtreecommitdiff
path: root/src/gallium/drivers
AgeCommit message (Collapse)Author
2011-02-18svga: Ensure LRP's restrictions are observed in all uses.José Fonseca
The dst reg must be a temporary, and not be the same as src0 or src2.
2011-02-18svga: Preserve src swizzles in submit_op2/3/4.José Fonseca
Several opcodes require scalar swizzle, and this requirement was being was not being observed when creating temporaries for other reasons.
2011-02-18r300g: remove tracking whether vertex buffers need to be validatedMarek Olšák
This was getting hard to maintain and didn't really bring any real benefits. Instead, validate buffers when the vertex array state is dirty.
2011-02-18svga: Disable surface cache for texturesJakob Bornecrantz
Signed-off-by: Jakob Bornecrantz <jakob@vmware.com>
2011-02-18svga: Describe svga_sampler_views for refcnt debuggingJakob Bornecrantz
Signed-off-by: Jakob Bornecrantz <jakob@vmware.com>
2011-02-18svga: Make sure that refcnt debugger gets the correct backtrace for createJakob Bornecrantz
Signed-off-by: Jakob Bornecrantz <jakob@vmware.com>
2011-02-18gallium/svga: Fix unnecessary swtnl fallbacksThomas Hellstrom
When we drop the in_swtnl_draw flag, we must force a rerun of update_need_swtnl to reset the need_swtnl flag to its correct value outside of a swtnl vbo draw. Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
2011-02-18r600g: reorganise rgtc pieces.Dave Airlie
when the cs checker fixes go upstream a lot of this can disappear into a drm version check. Signed-off-by: Dave Airlie <airlied@redhat.com>
2011-02-18r600g: Start a new TEX clause if the texture lookup address was fetched in ↵Fabian Bieler
the current clause Signed-off-by: Dave Airlie <airlied@redhat.com>
2011-02-18r600g: Add support to dump vertex- and texture-fetch clausesFabian Bieler
Signed-off-by: Dave Airlie <airlied@redhat.com>
2011-02-18r600g: add BC4/5 to RGTC conversionDave Airlie
this doesn't do anything much since the rest of mesa doesn't support RGTC yet.
2011-02-17svga: Don't use more than one constant per IFC instruction.José Fonseca
2011-02-17softpipe: rename env vars to be consistentBrian Paul
2011-02-17r600g: get s3tc working on cards with crappy 64/128 bit types.Dave Airlie
Some cards don't appear to work correctly with the UNORM type, so switch to the integer type, however since gallium has no integer types yet from what I can see we need to do a hack to workaround it for the blitter. Signed-off-by: Dave Airlie <airlied@redhat.com>
2011-02-17r600g: add missing type to color buffer swap.Dave Airlie
2011-02-16svga: flush when transitioning between HW and SW rendering pathsBrian Paul
To avoid mixing HW and SW rendering with the same vertex buffer.
2011-02-16svga: use TRUE/FALSE instead of 0/1Brian Paul
Some fields are booleans, others are bitmasks. Use TRUE/FALSE to clarify what's what.
2011-02-16svga: fix incorrect commentBrian Paul
2011-02-16svga: dimension the dirty[] array with SVGA_STATE_MAXBrian Paul
2011-02-16svga: disable a debug_printf() callBrian Paul
2011-02-16svga: Proper redefine_user_buffer implementation.José Fonseca
Unfortunately still not enough to make GoogleEarth happy.
2011-02-16r300g: fix a race between CS and SET_TILING ioctlsMarek Olšák
2011-02-16r300g: fix blitting NPOT compressed texturesMarek Olšák
2011-02-16svga: Don't fake DXT compression ability.José Fonseca
2011-02-16nvc0: fix blend factor mappingChristoph Bumiller
2011-02-16nvc0: fix emit_dfdx,dfdyChristoph Bumiller
2011-02-16nvc0: don't swap sources if either value is not in a GPRChristoph Bumiller
The memory / immediate source should already be in the only valid position.
2011-02-16nvc0: add missing break statements in constant_operandChristoph Bumiller
2011-02-16nvc0: fix clipping and use VIEWPORT instead of SCISSORChristoph Bumiller
2011-02-16nvc0: demagic the clear flags and fix region clearsChristoph Bumiller
The CLIP_RECTs always affect dedicated clears, and it's nicer than having to mark the viewport or scissor state dirty after it.
2011-02-16nvc0: front stencil mask and func mask methods are swappedChristoph Bumiller
2011-02-16nvc0: clone memory values with multiple refs before modifying themChristoph Bumiller
2011-02-16nvc0: disable early fragment tests if KIL is usedChristoph Bumiller
Early-Z pass raises the occlusion counter.
2011-02-16nvc0: force vertex data through FIFO if we need to convert itChristoph Bumiller
We may want to put the converted vertex buffer in persistent storage instead, but these are rare corner cases.
2011-02-16nvc0: use format from the template on surface creationChristoph Bumiller
Fixes piglit/fbo-srgb.
2011-02-16nvc0: update the set of formats supported by the 2D engineChristoph Bumiller
2011-02-16nvc0: fix user vertex buffer updatesChristoph Bumiller
2011-02-16r600g: fix typo in previous s3tc commitDave Airlie
pointed out by Marek on irc.
2011-02-16r300g: fix texture border color for float formatsMarek Olšák
2011-02-16r600g: add L8A8 SRGB formats.Dave Airlie
this fixes the piglit mipmap generation sRGB on my rv730. Signed-off-by: Dave Airlie <airlied@redhat.com>
2011-02-16r600g: add L4A4 support.Dave Airlie
this fixes piglit fbo-generatemipmap-formats on my rv730. Signed-off-by: Dave Airlie <airlied@redhat.com>
2011-02-16r600g: fix s3tc-texsubimageDave Airlie
we need to translate the destination box as well. fixes piglit's s3tc-texsubimage test. Signed-off-by: Dave Airlie <airlied@redhat.com>
2011-02-16r300g: disable linear filtering for float texturesMarek Olšák
2011-02-16r600g: add srgb compressed formats to the list.Dave Airlie
2011-02-15gallium: add usage parameter to pipe_buffer_createMarek Olšák
And set a reasonable buffer usage flag everywhere instead of just PIPE_USAGE_DEFAULT.
2011-02-15r600g: Fix RGB10_A2 format handlingFabian Bieler
2011-02-15r600g: fix regression in r6/7xx since mipmap reworkDave Airlie
I typod this when copy-pasting.
2011-02-15nvc0: Fix uninitialized variable warning.Vinson Lee
Fixes this GCC warning. nvc0_tgsi_to_nc.c: In function 'bld_tex': nvc0_tgsi_to_nc.c:1392: warning: 'dim' may be used uninitialized in this function
2011-02-15r300g: offload the CS ioctl to another threadMarek Olšák
This is a multi-threading optimization which hides the kernel overhead behind a thread. It improves performance in CPU-limited apps by 2-15%. Of course you must have at least 2 cores for it to make any difference. It can be disabled with: export RADEON_THREAD=0
2011-02-15r600g: add support for s3tc formats.Dave Airlie
On r600, s3tc formats require a 1D tiled texture format, so we have to do uploads using a blit, via the 64-bit and 128-bit formats Based on the r600c code we use a 64 and 128-bit type to do the blits. Still requires R600_ENABLE_S3TC until the kernel fixes are in, this has only been tested on evergreen where the kernel doesn't yet get in the way.