summaryrefslogtreecommitdiffstats
path: root/src/gallium
Commit message (Collapse)AuthorAgeFilesLines
* nvfx: emit bo relocations only when neededLuca Barbieri2010-08-238-20/+65
| | | | Should improve performance, possibly significantly.
* nvfx: match Gallium's gl_PointCoord brokennessLuca Barbieri2010-08-231-2/+5
| | | | | | | Gallium always puts gl_PointCoord in GENERIC[0] if point_quad_rasterization is enabled. This is silly, but for now it makes mesa-demos/glsl/pointcoord work.
* nvfx: support clip planes sensibly and fix them on nv30Luca Barbieri2010-08-238-116/+166
| | | | | | | | | | Before, we were discarding the compiled vertex program on each vertex program change. Now we compile the program as if there were 6 clip planes and dynamically patch in an "end program" bit at the right place. Also, nv30 should now work.
* r300g: avoid stall in no-tcl drawing when mapping vboDave Airlie2010-08-233-15/+26
| | | | | | | | | the current code reuses the same vbo over and over, however after a flush we'd stall and wait for mapping on the vbo when we should just fire and forget. On a gears test this brings me from ~620 to ~750 on my rv530 in swtcl mode. Signed-off-by: Dave Airlie <[email protected]>
* util: implement depth blitting in u_blitMarek Olšák2010-08-221-17/+43
| | | | Signed-off-by: Brian Paul <[email protected]>
* nvfx: fix minor memory leakLuca Barbieri2010-08-231-1/+1
|
* nvfx: support both sprite coord originsLuca Barbieri2010-08-233-43/+108
| | | | | | | Now we lie less when claiming OpenGL 2 support. Also, first piglit result group is now all green, except for fdo25614-genmipmap, which seems mesa/st's fault.
* nvfx: use 64-bit bitmasks for tempsLuca Barbieri2010-08-231-8/+8
|
* r600g: fix DB decompressionJerome Glisse2010-08-226-75/+97
| | | | Signed-off-by: Jerome Glisse <[email protected]>
* nvfx: Include missing header in nvfx_vertprog.c.Vinson Lee2010-08-221-0/+2
| | | | | | | | | | | Include draw_context.h for draw_*_vertex_shader symbols. Fixes the following GCC warning. nvfx_vertprog.c: In function 'nvfx_vp_state_create': nvfx_vertprog.c:1276: warning: implicit declaration of function 'draw_create_vertex_shader' nvfx_vertprog.c:1276: warning: assignment makes pointer from integer without a cast nvfx_vertprog.c: In function 'nvfx_vp_state_delete': nvfx_vertprog.c:1298: warning: implicit declaration of function 'draw_delete_vertex_shader'
* translate_sse: add R32G32B32A32_FLOAT -> X8X8X8X8_UNORM for EMIT_4UBJakob Bornecrantz2010-08-221-0/+26
| | | | Changed by me to use movd instead of movss to avoid penalties.
* translate_sse: refactor constant managementLuca Barbieri2010-08-221-81/+76
|
* nvfx: refactor to support multiple fragment program versionsLuca Barbieri2010-08-225-76/+127
|
* nvfx: move stuff aroundLuca Barbieri2010-08-225-84/+95
|
* r600g: depth buffer likely needs decompression when used as textureJerome Glisse2010-08-228-34/+919
| | | | | | | | Before using depth buffer as texture, it needs to be decompressed (tile pattern of db are different from one used for colorbuffer like texture) Signed-off-by: Jerome Glisse <[email protected]>
* glx/xlib: configurable strict/non-strict buffer size invalidateKeith Whitwell2010-08-223-4/+42
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | Introduce a new configuration option XMESA_STRICT_INVALIDATE to switch between swapbuffers-based and glViewport-based buffer invalidation. Default strict invalidate to false, ie glViewport-based invalidation, aka ST_MANAGER_BROKEN_INVALIDATE. This means we will not call XGetGeometry after every swapbuffers, which allows swapbuffers to remain asynchronous. For apps running at 100fps with synchronous swapping, a 10% boost is typical. For gears, I see closer to 20% speedup. Note that the work of copying data on swapbuffers doesn't disappear - this change just allows the X server to execute the PutImage asynchronously without us effectively blocked until its completion. This applies even to llvmpipe's threaded rasterization as the swapbuffers operation was a large part of the serial component of an llvmpipe frame. The downside of this is correctness - applications which don't call glViewport on window resizes will get incorrect rendering, unless XMESA_STRICT_INVALIDATE is set. The ultimate solution would be to have per-frame but asynchronous invalidation. Xcb almost looks as if it could provide this, but the API doesn't seem to quite be there.
* llvmpipe: reduce size of fragment shader variant keyKeith Whitwell2010-08-222-6/+22
| | | | Don't spend as much time comparing them.
* llvmpipe: remove unused member from lp_fragment_shader_variant_keyKeith Whitwell2010-08-221-4/+0
|
* llvmpipe: don't clear unused binsKeith Whitwell2010-08-221-2/+5
| | | | | If bins outside the current scene bounds are being corrupted, we'll need to fix that separately. Currently seems ok though.
* draw: reduce the size of the llvm variant keyKeith Whitwell2010-08-224-38/+100
|
* glx/xlib: remove another XSyncKeith Whitwell2010-08-221-1/+0
| | | | | With this change, xmesa_get_window_size still does one round trip, but that's better than doing two.
* glx/xlib: no need to call XSync from XMesaFlushKeith Whitwell2010-08-221-1/+1
| | | | Try to eliminate some unnecessary X server round trips.
* nvfx: simplify and correct fragment program update logicLuca Barbieri2010-08-226-94/+149
| | | | | | | This version should hopefully be much clearer and thus less likely to be subtly broken. Also fixes point sprites on nv40 and possibly some other bugs too.
* nvfx: make stipple setting independent of enableLuca Barbieri2010-08-224-27/+3
|
* nvfx: fix vertex programsLuca Barbieri2010-08-221-0/+2
|
* nvfx: use relocations array for vp constantsLuca Barbieri2010-08-224-35/+34
|
* r600g: Don't blindly unmap NULL->size.Henri Verbeet2010-08-221-1/+3
| | | | | There may actually be something mapped in that range, especially for large buffers like e.g. the GL Drawable.
* svga: Do not shortcut NULL surface relocations with SVGA3D_INVALID_ID.José Fonseca2010-08-222-6/+9
| | | | | How to cope with NULL surface relocations should be entirely at winsys' discretion.
* libgl-xlib: Include missing header in xlib.c.Vinson Lee2010-08-221-0/+1
| | | | Include st_api.h for st_api_create_OpenGL symbol.
* nvfx: Silence unused variable warning.Vinson Lee2010-08-221-0/+1
| | | | The variable is used but only in the body of an assert.
* util: Use #ifdef instead of #if.Vinson Lee2010-08-211-1/+1
| | | | This is a typo fix of earlier commit 0f3b3751b8643352dcc242567b3696bd1505df1d.
* util: Define dump_cpu only for DEBUG builds.Vinson Lee2010-08-211-0/+2
| | | | | | | dump_cpu is used only when DEBUG is defined. Fixes the following GCC warning on builds without DEBUG defined. util/u_cpu_detect.c:76: warning: 'debug_get_option_dump_cpu' defined but not used
* translate_sse: Silence uninitialized variable warnings.Vinson Lee2010-08-211-0/+14
| | | | Initialize variables on error paths.
* nvfx: Silence uninitialized variable warnings.Vinson Lee2010-08-212-0/+6
| | | | Variables weren't initialized on the error paths.
* i965g: Silence printf format warnings on 64-bit builds.Vinson Lee2010-08-211-4/+4
|
* nvfx: Silence uninitialized variable warnings.Vinson Lee2010-08-211-4/+4
| | | | | | | | | Silence the following i686-apple-darwin10-gcc-4.2.1 warnings. nv04_2d.c: In function 'nv04_region_copy_cpu': nv04_2d.c:560: warning: 'dswy' may be used uninitialized in this function nv04_2d.c:559: warning: 'dswx' may be used uninitialized in this function nv04_2d.c:562: warning: 'sswy' may be used uninitialized in this function nv04_2d.c:561: warning: 'sswx' may be used uninitialized in this function
* nv50: Silence incompatible pointer type initialization warning.Vinson Lee2010-08-211-2/+3
| | | | | Silence the following GCC warning. warning: initialization from incompatible pointer type
* nv50: Disable unused code.Vinson Lee2010-08-211-0/+4
| | | | | Disable release_hw and emit_mov_from_pred functions as they are currently not being used.
* i965g: Fix printf format warning on 32-bit platforms.Vinson Lee2010-08-211-1/+1
| | | | | Fixes the following GCC warning on 32-bit platforms. warning: format '%li' expects type 'long int', but argument 4 has type 'int'
* r600g: partialy fix texturing from depth buffer + initial support for untilingJerome Glisse2010-08-217-66/+165
| | | | | | | | | | | | | Partialy fix texturing from depth buffer, depth buffer is tiled following different tile organisation that color buffer. This properly set the tile type & array mode field of texture sampler when sampling from db resource. Add initial support to untiling buffer when transfering them, it's kind of broken by corruption the vertex buffer of previous draw. Signed-off-by: Jerome Glisse <[email protected]>
* draw: Don't assert if indices point outside vertex buffer.José Fonseca2010-08-221-2/+7
| | | | | | | | | | | | This is valid input, and asserting here does causes the test suites that verify this to crash. Also, the assert was wrongly accepting the case max_index == vert_info->count which, IIUC, is the first vertex outside the buffer. Assuming the vert_info->count is precise (which often is not the case).
* util: Silence uninitialized variable warnings.Vinson Lee2010-08-211-0/+4
|
* nvfx: actually fix it properlyLuca Barbieri2010-08-211-1/+1
|
* nvfx: fix incorrect assertLuca Barbieri2010-08-211-6/+9
|
* util: Move loop variable declaration outside for loop.Vinson Lee2010-08-211-1/+2
| | | | Fixes build error with MSVC.
* nvfx: Fix SCons build.Vinson Lee2010-08-217-56/+105
| | | | | Move declarations before code. Fix void pointer arithmetic.
* nvfx: fix warningsLuca Barbieri2010-08-213-5/+2
|
* gallivm: Emit DIVPS instead of RCPPS.José Fonseca2010-08-211-12/+24
| | | | | | | See comments for detailed rationale. Thanks to Michal Krol and Zack Rusin for detecting and investigating this in detail.
* nvfx: enable translate_sseLuca Barbieri2010-08-211-1/+1
|
* auxiliary: Add missing files to SCons build.Vinson Lee2010-08-211-14/+16
| | | | | Add u_linear.c and u_linkages.c to SCons build. Reorder list of files to be more alphabetical.
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-21 03:17:26 +0000