4b56e86e67
llvmpipe: allow tri_3_16 at any 4-aligned location within a tile
...
Doesn't require 16-alignment, so catch more cases.
2010-09-12 15:03:49 +01:00
26b663c2aa
llvmpipe: refactor tri_3_16
...
Keep step array as a set of four m128i's and reuse throughout the
rasterization.
2010-09-12 15:03:49 +01:00
67b957781d
llvmpipe: pass linear masks to fragment shader
...
Fragment shader can extract the correct bits for each quad.
2010-09-12 15:03:49 +01:00
4b99b9f5ff
llvmpipe: fix warnings on both 32 and 64 bit builds
2010-09-12 15:01:41 +01:00
51b1d4f03c
llvmpipe: fix wierd performance regression in isosurf
...
I really don't understand the mechanism behind this, but it
seems like the way data blocks for a scene are malloced, and in
particular whether we treat them as stack or a queue, and whether
we retain the most recently allocated or least recently allocated
has a real affect (~5%) on isosurf framerates...
This is probably specific to my distro or even just my machine,
but none the less, it's nicer not to see the framerates go in the
wrong direction.
2010-09-12 14:58:43 +01:00
67763488b1
pb: Fix the build, and add notes.
2010-09-12 10:37:06 +01:00
853953dc3c
llvmpipe: Only generate the whole shader specialization for opaque shaders.
...
If not opaque, then the color buffer will have to be read any way,
therefore the specialization is pointless.
2010-09-12 10:15:48 +01:00
b5fcf0c8e0
pb: add void * for flush ctx to mapping functions
...
If the buffer we are attempting to map is referenced by the unsubmitted
command stream for this context, we need to flush the command stream,
however to do that we need to be able to access the context at the lowest
level map function, currently we set the buffer in the toplevel map, but this
racy between context. (we probably have a lot more issues than that.)
I'll look into a proper solution as suggested by jrfonseca when I get some time.
2010-09-12 13:32:43 +10:00
95555ed03e
nv30: fix breakage due to 10 texcoord support on nv40
2010-09-11 21:11:03 +02:00
c34225974b
Add missing files to the tarball file lists.
2010-09-12 02:31:33 +08:00
19b2cfd6f6
mesa: Fix depend.es[12] generation when LLVM is enabled.
...
"llvm-config --cflags" outputs -f options, which conflict makedepend.
Clean up compiler flags and append LLVM_CFLAGS to the new xxx_CFLAGS
instead of xxx_CPPFLAGS, where xxx may be MESA, ES1, or ES2.
2010-09-12 02:31:33 +08:00
33b1d14913
r600g: Undo bo placement change.
...
This reverts a part of e795ca8f31
2010-09-11 18:40:45 +02:00
3ad3cbfb12
llvmpipe: Silence some warnings.
2010-09-11 13:47:58 +01:00
552e2b5065
gallivm: nr_channels is only valid for formats with plain layout.
...
This is erroneously throwing non plain formats out of the faster
AoS sampling path.
Doing 8bit interpolation for single channels such as L8 should be no
worse than with floating point. But this may need more investigation.
2010-09-11 13:34:54 +01:00
74f6edaee4
gallivm: Use const keyword on swizzles.
2010-09-11 13:31:35 +01:00
697fdf3fff
gallivm: Allow to TGSI AoS translation to happen in BGRA ordering.
...
Or any ordering.
2010-09-11 13:31:17 +01:00
277d768d10
llvmpipe: Don't store display the alpha ref value in the key.
...
It's never used.
2010-09-11 13:30:14 +01:00
dcc5d7f672
gallivm: Add a new debug flag to warn about performance issues.
2010-09-11 13:29:29 +01:00
c0d41d0404
gallivm: Helper functions for pointer indirection.
2010-09-11 13:27:38 +01:00
225530f1f2
gallivm: Cleanup the TGSI <-> sampler interface.
2010-09-11 13:27:27 +01:00
775edbfaa1
gallivm: Add some utility functions to set/get array elements too.
2010-09-11 13:27:22 +01:00
93158622e2
gallivm: Basic AoS TGSI -> LLVM IR.
...
Essentially a variation of the SoA version.
2010-09-11 13:25:49 +01:00
58daea741f
gallivm: Move the texture modifiers to the header.
...
Useful to pass these around.
2010-09-11 13:25:46 +01:00
34ea50f672
gallivm: s/lp_build_broadcast_aos/lp_build_swizzle_scalar_aos/
...
More accurate description of this function purpose.
2010-09-11 13:01:31 +01:00
96c6e7e73e
Add a test for the KIL opcode
...
This is a simple test for the KIL opcode. It should render a 6 sided figure
with a colored interior.
2010-09-11 13:01:20 +01:00
afd328afa8
llvmpipe: restore larger command blocks
2010-09-11 10:04:53 +01:00
10d0a835bc
llvmpipe: move some debug to DEBUG_SCENE
2010-09-11 10:04:53 +01:00
bc759cd37e
llvmpipe: add DEBUG_MEM option
2010-09-11 10:04:53 +01:00
b7a54b3dc6
llvmpipe: allow bigger scenes
2010-09-11 10:04:52 +01:00
4749429a4a
r300/compiler: Reorganize presub_helper()
2010-09-10 18:18:10 -07:00
3fffcb317c
r300/compiler: Don't use presubtract in TEX instructions
2010-09-10 18:18:10 -07:00
8b67159e9a
r300/compiler: Print the presub subtract operation in the correct order
2010-09-10 18:18:10 -07:00
e3fc210a4f
r300/compiler: Fix dataflow bug in presub_helper()
2010-09-10 18:18:09 -07:00
d442c7f172
r300/compiler: Replace asserts with error messages
2010-09-10 18:18:09 -07:00
66b1e716fb
r300/compiler: Fix copy propigation for some presub instructions
2010-09-10 18:18:09 -07:00
564653b9f1
r300/compiler: Add peephole optimization for the 'sub' presubtract operation
2010-09-10 18:18:09 -07:00
a64b4a05af
r300/compiler: Add peephole optimization for the 'add' presubtract operation
2010-09-10 18:18:09 -07:00
598e220f95
r300/compiler: Clean up rc_pair_alloc_source()
2010-09-10 18:18:09 -07:00
63432ecfce
r300/compiler: Enable presubtract sources
...
The r300 compiler can now emit instructions that select from the presubtract
source. A peephole optimization has been added to convert instructions like:
ADD Temp[0].x, none.1, -Temp[1].x into the INV (1 - src0) presubtract
operation.
2010-09-10 18:18:08 -07:00
d8a3662008
mesa: Remove unused Emit flags from gl_shader_compiler_options
2010-09-10 13:10:26 -07:00
55aaee6020
intel: Remove noise opcode support from i915 and i965 drivers
...
With recent changes to the GLSL compiler, these opcode should never be
seen in these drivers.
2010-09-10 12:33:15 -07:00
58eaade662
r600c: add missing header
2010-09-10 14:14:12 -04:00
9cced19125
r600c: add OQ support for evergreen
2010-09-10 13:26:10 -04:00
7ffc5d7c2e
r600c: oq updates
2010-09-10 13:13:08 -04:00
4e53d6e6c9
r600c: add blit support for evergreen
...
driver was previously calling the r600 blit code
which won't work on evergreen.
2010-09-10 13:01:23 -04:00
886c6c09ea
r600c: emit start3d packet on evergreen
2010-09-10 13:01:23 -04:00
f35b728acc
r600c: fix some typos
2010-09-10 13:01:23 -04:00
a712db73ce
r600c: fix type in cb setup on evergreen
2010-09-10 13:01:23 -04:00
c7843280f0
r600c: add support for more rendering formats on evergreen
2010-09-10 13:01:22 -04:00
78109952d4
r600: set correct initial point_minmax values
2010-09-10 14:55:10 +03:00
Keith Whitwell
José Fonseca
Dave Airlie
Luca Barbieri
Chia-I Wu
Tilman Sauerbeck
Alex Corscadden
Tom Stellard
Ian Romanick
Alex Deucher
Andre Maasikas