Age | Commit message (Collapse) | Author | |
---|---|---|---|
2011-02-15 | gallium: add usage parameter to pipe_buffer_create | Marek Olšák | |
And set a reasonable buffer usage flag everywhere instead of just PIPE_USAGE_DEFAULT. | |||
2011-02-15 | r300g: offload the CS ioctl to another thread | Marek 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 | |||
2011-02-15 | r300g: implement fences using dummy relocations | Marek Olšák | |
So finally we have them. | |||
2011-02-15 | r300g: fix SIGFPE on debug builds | Marek Olšák | |
2011-02-15 | r300g: inline some of the pipe_buffer_map/unmap calls | Marek Olšák | |
2011-02-14 | r300g: do not track whether occlusion queries have been flushed | Marek Olšák | |
The winsys takes care of flushing automatically. | |||
2011-02-14 | gallium: notify drivers about possible changes in user buffer contents | Marek Olšák | |
Also implement the redefine_user_buffer hook in the drivers. | |||
2011-02-14 | r300g: Remove redundant initialization. | Vinson Lee | |
Remove redundant initialization from commit 3b01b52bd78e3d2fc857feacebd815a5fae00c94 noticed by tstellar. | |||
2011-02-14 | r300g: put indices in CS if there's just a few of them and are in user memory | Marek Olšák | |
2011-02-14 | r300g: cleanup the draw functions | Marek Olšák | |
2011-02-14 | r300g: do not create a user buffer struct for misaligned ushort indices fallback | Marek Olšák | |
2011-02-14 | r300g: fix fallback for misaligned ushort indices with num vertices >= 65535 | Marek Olšák | |
2011-02-14 | r300g: consolidate emission of common draw regs | Marek Olšák | |
2011-02-14 | r300g: Move declaration before code. | Vinson Lee | |
Fixes SCons build. | |||
2011-02-14 | r300g: emit 3D_LOAD_VBPNTR only when necessary | Marek Olšák | |
I thought I couldn't skip emitting this packet in some cases. Well it looks like I can. | |||
2011-02-14 | r300g: fix reference counting when translating indices | Marek Olšák | |
2011-02-14 | r300g: remove the relocation after AARESOLVE_PITCH | Marek Olšák | |
2011-02-13 | r300g: correctly determine if a texture is blittable in texture_get_transfer | Marek Olšák | |
2011-02-12 | r300g: when printing shader linker errors to stderr, report it's not a bug | Marek Olšák | |
2011-02-12 | r300g: add debug options nozmask and nohiz which disable some hyper-z features | Marek Olšák | |
2011-02-11 | r300g: prevent NULL pointer dereference in r300_buffer_create | Marek Olšák | |
Should fix: https://bugs.freedesktop.org/show_bug.cgi?id=33185 | |||
2011-02-11 | r300g: fix warning | Marek Olšák | |
2011-02-11 | r300g: import the last bits of libdrm and cleanup the whole thing | Marek Olšák | |
Based on Dave's branch. The majority of this commit is a cleanup, mainly renaming things. There wasn't much code to import, just ioctl calls. Also done: - implemented unsynchronized bo_map (important optimization!) - radeon_bo_is_referenced_by_cs is no longer a refcount hack - dropped the libdrm_radeon dependency I'm surprised that this has resulted in less code in the end. | |||
2011-02-10 | r300g: implement accelerated copy_region for compressed formats | Marek Olšák | |
2011-02-10 | r300g: add a way to change texture properties arbitrarily | Marek Olšák | |
So that we can implement resource_copy on arbitrary data. | |||
2011-02-10 | r300g: consolidate buffers and textures to r300_resource | Marek Olšák | |
Transfers and create/destroy are still handled separately. | |||
2011-02-10 | r300g: simplify WRITE_RELOC API and cleanup | Marek Olšák | |
2011-02-10 | r300g: use format from pipe_surface instead of pipe_resource | Marek Olšák | |
2011-02-10 | Revert "r300g: support sRGB colorbuffers" | Marek Olšák | |
This partially reverts commit 91eba2567eab9409d94efc3c1f07a4a3731d0047. Conflicts: src/gallium/drivers/r300/r300_blit.c | |||
2011-02-08 | r300g: use the same upload buffer for vertices and indices | Marek Olšák | |
2011-02-08 | u_vbuf_mgr: add a way to specify the BIND flag for the upload buffer | Marek Olšák | |
2011-02-07 | r300g: use the new vertex buffer manager | Marek Olšák | |
2011-02-06 | r300g: do not flush the uploader | Marek Olšák | |
We don't have to unmap and recreate the upload buffer when a flush occurs. This should also prevent buffer allocations from failing. | |||
2011-02-06 | r300g: RS400 doesn't have ZMASK | Marek Olšák | |
2011-02-03 | r300g: Make the buffer and texture vbtls static const. | Henri Verbeet | |
2011-01-30 | r300g: upload translated indices via the uploader | Marek Olšák | |
2011-01-30 | r300g: rework vertex format fallback | Marek Olšák | |
1) Only translate the [min_index, max_index] range. 2) Upload translated vertices via the uploader. 3) Rename valid_vertex_buffer[] to real_vertex_buffer[] | |||
2011-01-28 | r300g: fix resource_copy_region for DXT SRGB formats | Marek Olšák | |
2011-01-28 | r300g: handle PIPE_CAP_ARRAY_TEXTURES | Marek Olšák | |
2011-01-28 | r300g: 8x8-compressed zbuffer can only be point-sampled | Marek Olšák | |
2011-01-28 | r300g: fix and re-enable 8x8 zbuffer compression mode | Marek Olšák | |
Also cleanup the whole thing. | |||
2011-01-27 | r300g: print driver info if RADEON_DEBUG=info | Marek Olšák | |
2011-01-27 | r300g: add winsys flag CAN_AACOMPRESS | Marek Olšák | |
2011-01-27 | r300g: rename flag squaretiling -> drm_2_1_0 | Marek Olšák | |
2011-01-27 | r300g: fix some bugs with zbuffer compression (v4) | Marek Olšák | |
This drops the memblock manager for ZMASK. Instead, only one zbuffer can be compressed at a time. Note that this does not necessarily have to be slower. When there is a large number of zbuffers, compression might be used more often than it was before. It's also easier to debug. How it works: 1) 'clear' turns the compression on. 2) If some other zbuffer is set or the currently-bound zbuffer is used for texturing, the driver decompresses it and then turns the compression off. Notes: - The ZMASK clear has been refactored, so that only one packet3 is used to clear ZMASK. - The 8x8 compression mode is disabled. I couldn't make it work without issues. - Also removed driver-specific stuff from u_blitter. Driver status: - RV530 and R580 appear to just work (finally). - RV570 should work, but there may be an issue that we don't correctly calculate the number of dwords to clear, resulting in a partially uninitialized zbuffer. - RS690 misrenders as if no ZMASK clear happened. No idea what's going on. - RV350 may even hardlock. This issue was already present and this patch doesn't fix it. I think we are still missing some hardware info we need to make the zbuffer compression work fully. Note that there is also an issue with HiZ, resulting in a sort of blocky zigzagged corruption around some objects. | |||
2011-01-24 | r300g: remove unused function | Marek Olšák | |
2011-01-24 | r300g: remove any traces of depth_clamp | Marek Olšák | |
I couldn't make it work. GB_TILE_CONFIG.Z_EXTENDED, which enables per-pixel Z clamping, and VAP_CLIP_CNTL.CLIP_DISABLE, which disables clipping, do help, but they also add regressions like random graphics corruptions in some games. | |||
2011-01-24 | r300g: handle PIPE_CAP_INSTANCED_DRAWING in get_param | Marek Olšák | |
2011-01-24 | Revert "r300g/swtcl: re-enable LLVM" | Jakob Bornecrantz | |
This reverts commit 88550083b3857184445075e70fed8b2eed4952a1. | |||
2011-01-23 | r300g: Increase fragment shader limits for r400 cards | Tom Stellard | |
r400 fragment shaders now support up to 64 temporary registers, 512 ALU instructions, and 512 TEX instructions. |