aboutsummaryrefslogtreecommitdiffstats
path: root/src/gallium/drivers
Commit message (Collapse)AuthorAgeFilesLines
* r600g: flush FMASK and CMASK at the end of CSMarek Olšák2013-01-061-1/+3
|
* r300g: implement MSAAMarek Olšák2013-01-0613-148/+456
| | | | | | | | | | | | | | | | | | | | | | This is not as optimized as r600g - the MSAA compression is missing, so r300g needs a lot of bandwidth (more than r600g to do the same thing). However, if the bandwidth is not an issue for you, you can enjoy this unoptimized MSAA support. The only other missing optimization for MSAA is the fast color clear. MSAA is enabled on r500 only, because that's the only GPU family I tested. That said, MSAA should work on r300 and r400 as well (but you must set RADEON_MSAA=1 to allow it, then turn MSAA on in your app or set GALLIUM_MSAA=n, n >= 2, n <= 6) I will enable the support by default on r300-r400 once someone (other than me) tests those chipsets with piglit. The supported modes are 2x, 4x, 6x. The supported MSAA formats are RGBA8, BGRA8, and RGBA16F (r500 only). Those 3 formats are used for all GL internal formats. Tested with piglit. (I have ported all MSAA tests to GL2.1)
* r300g: simplify DSA state, add ability to patch FG_ALPHA_FUNC while emittingMarek Olšák2013-01-063-79/+46
| | | | Preparation for MSAA and alpha-to-coverage.
* r300g/compiler: add shader emulation for the alpha_to_one stateMarek Olšák2013-01-065-0/+41
|
* radeon/llvm: Remove backend code from MesaTom Stellard2013-01-0499-19168/+0
| | | | | | | | | | | | This code now lives in an external tree. For the next Mesa release fetch the code from the master branch of this LLVM repo: http://cgit.freedesktop.org/~tstellar/llvm/ For all subsequent Mesa releases, fetch the code from the official LLVM project: www.llvm.org
* Support LLVM >= 3.2 on radeonsi and opencl.Johannes Obermayr2013-01-041-0/+7
| | | | | | Tom Stellard: - Backend now has same name for all LLVM versions - Add missing LLVM_VERSION_INT definition
* r300g: don't check for vertex and index buffer bind flagsMarek Olšák2013-01-042-4/+2
|
* r300g/swtcl: use memcpy to emit indicesMarek Olšák2013-01-041-10/+11
|
* r300g/swtcl: simplify vertex uploadingMarek Olšák2013-01-045-88/+42
| | | | | | | | - skip the vertex buffer reallocation in flush and just use the unsynchronized flag to get new memory. - remove the cruft needed to get around the issues with the vertex buffer reallocation in flush - use pb_buffer instead of pipe_resource
* r300g/swtcl: fix crash when setting vertex buffersMarek Olšák2013-01-041-0/+3
| | | | Broken by e73bf3b805de78299f1a652668ba4e6eab9bac94.
* r300g: don't set PIPE_BIND flags for internal texturesMarek Olšák2013-01-042-21/+1
|
* r600g: implement 3D transfersMarek Olšák2013-01-041-22/+32
| | | | That means we can map and read multiple slices with one transfer_map call.
* gallium/u_blitter: unify some parameters into a dstbox parameter in blit_genericMarek Olšák2013-01-042-9/+13
| | | | Reviewed-by: Brian Paul <[email protected]>
* gallium/u_blitter: remove useless parameter from blitter_default_dst_textureMarek Olšák2013-01-042-2/+2
| | | | Reviewed-by: Brian Paul <[email protected]>
* gallium/radeon: send the END_OF_FRAME flag to the DRMMarek Olšák2013-01-043-3/+7
|
* gallium: extend pipe_context::flush for it to accept an END_OF_FRAME flagMarek Olšák2013-01-0418-28/+42
| | | | | | | | | | | | | | | | | Usage with pipe_context: pipe->flush(pipe, NULL, PIPE_FLUSH_END_OF_FRAME); Usage with st_context_iface: st->flush(st, ST_FLUSH_END_OF_FRAME, NULL); The flag is only a hint for drivers. Radeon will use it for buffer eviction heuristics in the kernel (e.g. for queries like how many frames have passed since a buffer was used). The flag is currently only generated by st/dri on SwapBuffers. Reviewed-by: Brian Paul <[email protected]> Reviewed-by: Stéphane Marchesin <[email protected]>
* radeonsi: fix int->bool conversion in fence_signalledMarek Olšák2013-01-041-1/+1
|
* r600g: fix int->bool conversion in fence_signalledMarek Olšák2013-01-041-1/+1
| | | | NOTE: This is a candidate for the stable branches.
* r300g: Fix visibility CFLAGS in automakeAdam Jackson2013-01-021-0/+1
| | | | | | Note: this is a candidate for the 9.0 stable branch. Signed-off-by: Adam Jackson <[email protected]>
* galahad, noop: Fix visibility CFLAGS in automakeAdam Jackson2013-01-021-0/+2
| | | | | | Note: this is a candidate for the 9.0 stable branch. Signed-off-by: Adam Jackson <[email protected]>
* r600g: rename GPU_FLUSH -> INVAL_READ_CACHESMarek Olšák2012-12-225-11/+11
| | | | because that's what it does.
* r600g: remove redundant parameter alloc_bo from r600_texture_create_objectMarek Olšák2012-12-221-6/+5
| | | | alloc_bo == !buf
* nouveau: deal with tbo cap for now.Dave Airlie2012-12-223-0/+3
| | | | | | This fixes the printk running apps against master. Signed-off-by: Dave Airlie <[email protected]>
* r600g: always use a tiled resource as the destination of MSAA resolveMarek Olšák2012-12-213-8/+16
| | | | | | | | | i.e. we have to allocate a temporary tiled resource if dst isn't tiled. This fixes hardlocks on r6xx-r7xx, though using a linear resource is forbidden on later asics as well. NOTE: This is a candidate for the stable branches.
* r600g: remove a false commentMarek Olšák2012-12-211-1/+0
|
* r600g: don't suspend TIME_ELAPSED queries during flushingMarek Olšák2012-12-214-54/+8
| | | | | According to the GL spec, the result should be equivalent to comparing two timestamps.
* r600g: add cs tracing infrastructure for lockup pin pointingJerome Glisse2012-12-205-2/+106
| | | | | | | | It's a build time option you need to set R600_TRACE_CS to 1 and it will print to stderr all cs along as cs trace point value which gave last offset into a cs process by the GPU. Signed-off-by: Jerome Glisse <[email protected]>
* r600g: add htile support v16Jerome Glisse2012-12-2010-24/+221
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | htile is used for HiZ and HiS support and fast Z/S clears. This commit just adds the htile setup and Fast Z clear. We don't take full advantage of HiS with that patch. v2 really use fast clear, still random issue with some tiles need to try more flush combination, fix depth/stencil texture decompression v3 fix random issue on r6xx/r7xx v4 rebase on top of lastest mesa, disable CB export when clearing htile surface to avoid wasting bandwidth v5 resummarize htile surface when uploading z value. Fix z/stencil decompression, the custom blitter with custom dsa is no longer needed. v6 Reorganize render control/override update mecanism, fixing more issues in the process. v7 Add nop after depth surface base update to work around some htile flushing issue. For htile to 8x8 on r6xx/r7xx as other combination have issue. Do not enable hyperz when flushing/uncompressing depth buffer. v8 Fix htile surface, preload and prefetch setup. Only set preload and prefetch on htile surface clear like fglrx. Record depth clear value per level. Support several level for the htile surface. First depth clear can't be a fast clear. v9 Fix comments, properly account new register in emit function, disable fast zclear if clearing different layer of texture array to different value v10 Disable hyperz for texture array making test simpler. Force db_misc_state update when no depth buffer is bound. Remove unused variable, rename depth_clearstencil to depth_clear. Don't allocate htile surface for flushed depth. Something broken the cliprect change, this need to be investigated. v11 Rebase on top of newer mesa v12 Rebase on top of newer mesa v13 Rebase on top of newer mesa, htile surface need to be initialized to zero, somehow special casing first clear to not use fast clear and thus initialize the htile surface with proper value does not work in all case. v14 Use resource not texture for htile buffer make the htile buffer size computation easier and simpler. Disable preload on evergreen as its still troublesome in some case v15 Cleanup some comment and remove some left over v16 Define name for bit 20 of CP_COHER_CNTL Signed-off-by: Pierre-Eric Pelloux-Prayer <[email protected]> Signed-off-by: Alex Deucher <[email protected]> Signed-off-by: Jerome Glisse <[email protected]>
* r600g: rework flusing and synchronization pattern v7Jerome Glisse2012-12-208-162/+89
| | | | | | | | | | | | | | | This bring r600g allmost inline with closed source driver when it comes to flushing and synchronization pattern. v2-v4: history lost somewhere in outer space v5: Fix compute size of flushing, use define for flags, update worst case cs size requirement for flush, treat rs780 and newer as r7xx when it comes to streamout. v6: Fix num dw computation for framebuffer state, remove dead code, use define instead of hardcoded value. v7: Remove dead code Signed-off-by: Jerome Glisse <[email protected]>
* llvmpipe: Drop PIPE_QUERY_TIME_ELAPSED support.José Fonseca2012-12-203-13/+1
| | | | | | | | | | | | | | | | | It was slightly wrong: we were computing the longest duration of the query among all the rasterizer tasks. Regardless, for tile-based implementations such as llvmpipe, time differences will never be very useful, because rendering before/during/after the query is all interleaved. And this is expected, see ARB_timer_query spec, issue 10. In particular, piglit ext_timer_query-time-elapsed still fails, because it makes assumptions that don't hold true in in tiled architectures. Not sure how to fix that though. Reviewed-by: Dave Airlie <[email protected]> Reviewed-by: Marek Olšák <[email protected]> Reviewed-by: Brian Paul <[email protected]>
* gallium: s/PIPE_CAP_TIMER_QUERY/PIPE_CAP_QUERY_TIME_ELAPSED/José Fonseca2012-12-2011-11/+11
| | | | | | | | To better reflect what it is being advertised. Reviewed-by: Dave Airlie <[email protected]> Reviewed-by: Marek Olšák <[email protected]> Reviewed-by: Brian Paul <[email protected]>
* r600g: add assertions to prevent creation of invalid surfacesMarek Olšák2012-12-203-16/+18
|
* r600g: refactor and make streamout dumping more informativeMarek Olšák2012-12-201-12/+21
| | | | Reviewed-by: Dave Airlie <[email protected]>
* r600g: try to fix streamout for the cases where BURST_COUNT > 0Marek Olšák2012-12-201-1/+1
| | | | | | | | | | The burst was incorrectly used, because ELEM_SIZE was always 0. I don't know if the burst works, because I don't know of any test which uses it. NOTE: This is a candidate for the stable branches. Reviewed-by: Dave Airlie <[email protected]>
* r600g: lower stream outputs with dst_offset < start_componentMarek Olšák2012-12-201-7/+47
| | | | | | This fixes streamout breakage caused by the varying packing. Reviewed-by: Dave Airlie <[email protected]>
* r600g: use r600_get_temp to get temporaries for CLIPDIST shader outputsMarek Olšák2012-12-201-3/+11
| | | | | | I need this to be able to use r600_get_temp in the function later. Reviewed-by: Dave Airlie <[email protected]>
* softpipe: fix up FS variant unbinding / deletionBrian Paul2012-12-193-11/+9
| | | | | | | | | | | | The old call to tgsi_exec_machine_bind_shader() in softpipe_delete_fs_state() was never called since the shader's original tokens are never passed to the tgsi interpreter (only shader _variant_ tokens are). Now, unbind the variant's tokens from the tgsi interpreter when we free the variant. This doesn't fix any known bugs but it's the right thing to do. Note: This is a candidate for the stable branches.
* softpipe: fix unreliable FS variant binding bugBrian Paul2012-12-193-14/+10
| | | | | | | | | | | | | In exec_prepare() we were comparing pointers to see if the fragment shader variant had changed before calling tgsi_exec_machine_bind_shader(). This didn't work reliably when there was a lot of shader token malloc/ freeing going on because the memory might get reused. Instead, bind the shader variant during regular state validation. Fixes http://bugs.freedesktop.org/show_bug.cgi?id=40404 (fixes a couple of piglit's glsl-max-varyings test) Note: This is a candidate for the stable branches.
* Revert "r600g: work around ddx over alignment"Jerome Glisse2012-12-191-9/+3
| | | | | | This reverts commit d8287bac1fd4a77abc2db38de134f14176740d23. Cause more issue than it fix. Need to think of a proper solution.
* r600g: work around ddx over alignmentJerome Glisse2012-12-181-3/+9
| | | | | | | | | This force surface allocated from ddx to be consider as height aligned on 8 and fix 1D->2D tiling transition that result from this. Signed-off-by: Jerome Glisse <[email protected]> Reviewed-by: Marek Olšák <[email protected]>
* radeon/llvm: improve cube map handlingVadim Girlin2012-12-184-22/+84
| | | | | | | | Add support for TEX2, TXB2, TXL2, fix SHADOWCUBE Signed-off-by: Vadim Girlin <[email protected]> Reviewed-by: Michel Dänzer <[email protected]> Tested-by: Michel Dänzer <[email protected]>
* radeon/llvm: fix TXQ_LZ handling for cube mapsVadim Girlin2012-12-181-2/+4
| | | | Signed-off-by: Vadim Girlin <[email protected]>
* r600g: initialize inst_mod in r600_tex_from_byte_streamVadim Girlin2012-12-181-0/+2
| | | | Signed-off-by: Vadim Girlin <[email protected]>
* r600g: fixup offset types for printingDave Airlie2012-12-162-4/+4
| | | | | | This allows the debug code to at least show the sign properly. Signed-off-by: Dave Airlie <[email protected]>
* draw/llvmpipe: fix transform feedback position + enable other extensionsDave Airlie2012-12-141-1/+1
| | | | | | | | | | | | | | | | This builds on the previous draw/softpipe patch. So llvmpipe does streamout calls after clip/viewport stages, but we have the pre-clip position stored for later use, so when we are doing transform feedback, and its the position vertex grab the vertex from the stored pre clip position. The perfect fix is too probably add a codegen transform feedback stage in between shader and clip stages, but this is good enough for now. Reviewed-by: Roland Scheidegger <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* draw: add support for later transform feedback extensionsDave Airlie2012-12-141-1/+1
| | | | | | | | | | | | | | | | | This adds support to draw for the new features of transform feedback. a) fix count_from_stream_output, using max_index+1 for now but it looks like it should be valid as its derived from the vertex elements/vbo. b) fix striding and dst offsets in output buffers - was just wrong before. c) fix crash if tfb is suspended (so.num_targets == 0) This also enables the new features on softpipe. It should be possible to enable them on llvmpipe as well after this commit, but would need to schedule piglit runs. Signed-off-by: Dave Airlie <[email protected]>
* r600g/radeonsi: Silence warningsMaxence Le Dore2012-12-135-30/+49
| | | | Reviewed-by: Tom Stellard <[email protected]>
* r600g: use load_ar checks with llvm output.Vincent Lejeune2012-12-131-0/+6
| | | | Reviewed-by: Tom Stellard <[email protected]>
* softpipe: remove unused corner0 variableBrian Paul2012-12-121-1/+0
|
* llvmpipe: remove unneeded draw_flush() callBrian Paul2012-12-121-2/+0
| | | | | | | | | | This is redundant since we're calling draw_bind_fragment_shader() which already does a flush. v2: the redundant flush in llvmpipe_set_constant_buffer() has already been removed by commit 3427466e6dbbb8db7c1ecda6b3859ca1cc5827a3 Reviewed-by: José Fonseca <[email protected]>
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-11 16:13:55 +0000