1673bb38bb
tgsi: replace 0xf with TGSI_WRITEMASK_XYZW
2010-04-18 10:02:11 -06:00
10e4ec448e
softpipe: fix computation of fragment[FACE] attribute
...
In TGSI, front facing is +1 and back-facing is -1. We were computing
this attribute as +1 and 0 before. However, the value isn't actually
used anywhere because we machine->Face attribute overrides it in
tgsi_exec.c. That could be changed, removing some special-case code...
2010-04-18 10:02:11 -06:00
6756c07b66
softpipe: clean-up and comments
2010-04-18 10:02:11 -06:00
b806c3056f
softpipe: document MAX_QUADS
2010-04-18 10:02:11 -06:00
24ceef7a69
r300g: more solid is_buffer_referenced implementation
...
This fixes crash from 7a1b5c937f
2010-04-18 17:49:12 +02:00
cea7b9f937
u_surfaces: silence warning
2010-04-18 17:35:17 +02:00
69d8b6d5fb
Revert "r300g: save vertex buffers around u_blitter calls"
...
Not sure why it crashes.
2010-04-18 17:02:39 +02:00
5acee66023
nvfx: move declarations before code
2010-04-18 08:43:42 -06:00
27322eeb73
nvfx: fix void * arithmetic warning
2010-04-18 08:41:51 -06:00
b74f4bc6b8
nvfx: move declarations before code
2010-04-18 08:41:25 -06:00
d84b8cb1a8
llvmpipe: update comment
2010-04-18 08:38:19 -06:00
d7c336dc49
nvfx: move declarations before code to silence warnings
2010-04-18 08:38:03 -06:00
f9ea7a2ea9
auxiliary: add util_dirty_surfaces to track dirty render temporaries
2010-04-18 16:23:11 +02:00
ac4b8db62d
auxiliary: add util_surfaces to track a resource's pipe_surface structs
2010-04-18 16:23:11 +02:00
24af25df2b
u_inlines: add function to initialize pipe_surface
2010-04-18 16:23:11 +02:00
2ac088141a
nvfx: fix fragment program constant updates
2010-04-18 16:23:11 +02:00
7a1b5c937f
r300g: do not tell st that a buffer is referenced by hw
...
It saves a few libdrm calls and unnecessary flushes.
2010-04-18 16:09:21 +02:00
bc8fa67926
r300g: save vertex buffers around u_blitter calls
...
Not strictly needed by the Mesa state tracker, but needed in general.
Only compile tested.
2010-04-18 15:29:23 +02:00
64aa67a360
u_blitter: add support for saving vertex buffers
...
Currently r300g does not save vertex buffer on blitter calls.
It gets away with it because the current Mesa state tracker usually
resets vertex buffers on every draw calls.
However, this is wrong.
nvfx won't be lucky because it needs to use the blitter inside draw
calls.
2010-04-18 15:29:23 +02:00
5df6c43061
p_compiler: add likely/unlikely macros
...
Defined like the Linux kernel.
A explanatory comment is included.
2010-04-18 15:29:21 +02:00
851edfe3c9
r300g: fixup 0 stride buffer properly.
...
Just set the max index to 1, this lets doom3 run and seems correct,
though it would be better to just emit a constant like SVGA does.
Signed-off-by: Dave Airlie <airlied@redhat.com >
2010-04-18 18:51:10 +10:00
4b148bcf5e
llvmpipe: Emit only the vertex attributes necessary for the FS, and ensure the first one is always position.
...
With this we correctly handle vertex shaders whose output position is not
in index zero.
2010-04-18 10:36:28 +02:00
4272c01fbf
softpipe: Preserve/check the signal of the number of scanlines in an edge.
...
It can become negative in some weird triangles.
2010-04-18 10:36:28 +02:00
b609cfc7c9
draw: Cosmetic cleanups and comments.
2010-04-18 10:36:28 +02:00
d036e9312f
retrace: Fix vertex/index/constant buffer dumping.
2010-04-18 10:36:28 +02:00
53c3c914a7
trace: Handle transfers with zero strides correctly.
2010-04-18 10:36:28 +02:00
4b95c9549d
st/python: Set the bind flag for the default texture.
2010-04-18 10:36:28 +02:00
5c898af870
retrace: Remove deprecated code.
2010-04-18 10:36:28 +02:00
6aae062bdf
mesa: Restore comment too.
2010-04-18 10:36:28 +02:00
4e98116c23
glsl: Fix handling of OPCODE_PRINT for no registers case.
...
A register file value is unsigned so could never be -1. A
value of 0 also aliased to PROGRAM_TEMPORARY.
If an OPCODE_PRINT has no registers to print, set the register
file value to PROGRAM_UNDEFINED and check for that value when
executing this instruction.
2010-04-18 00:11:49 -07:00
4485dd8a3e
i965: Assert that an OPCODE_IF was seen before an OPCODE_ELSE.
2010-04-17 23:48:37 -07:00
010f81f32b
st/dri: Check for NULL before dereference and assignment.
2010-04-17 23:40:14 -07:00
63c6d6add1
nvfx: Initialize variables.
2010-04-17 23:25:47 -07:00
0268e8984c
r300g: Cast rbuf->user_buffer to 'uint8_t *' before arithmetic.
...
Fixes this SCons build error.
"pointer of type 'void *' uses in arithmetic"
2010-04-17 22:44:35 -07:00
fe52b585f2
nvfx: Remove unused variables.
2010-04-17 22:23:25 -07:00
891e9d6f3f
nv50: Initialize variables.
...
Silences GCC uninitialized variable warnings.
2010-04-17 21:21:31 -07:00
b054d7912d
nv50: Remove unused variable.
2010-04-17 20:39:03 -07:00
5db285b742
i915g: Remove broken power of two function
2010-04-18 02:50:35 +01:00
d14c4ca209
i915g: Add some helper functions
...
Mostly to make the lines not so long.
2010-04-18 02:50:35 +01:00
ed94e744d3
i915g: Fix cubemap layouts
2010-04-18 02:50:35 +01:00
762ed4db47
i915g: Clean and add some debug printing
...
Disabled by default
2010-04-18 02:50:35 +01:00
29ff6e87f5
gallium/tests: Make trivial apps depend on libs
2010-04-18 02:50:35 +01:00
0b4df63609
r300g: various other cleanups
...
It seems to me that the resource management code is the most blurry.
Also some of my notes about fastfill are wrong, removing them too.
2010-04-18 02:56:09 +02:00
d3e7dfc1fd
r300g: inline screen_buffer functions
2010-04-18 02:56:09 +02:00
9f2f6096a8
r300g: remove unused declarations
2010-04-18 02:56:09 +02:00
5b609badf7
r300g: add debugging options "notiling" and "noimmd" (for testing)
...
notiling = Disable texture tiling
noimmd = Disable immediate mode (this optimization was really worth it!)
2010-04-18 02:56:09 +02:00
4185d9770f
r300g: remove r300_transfer::ctx
2010-04-18 02:56:09 +02:00
bb4f5fff0c
r300g: remove use of c++ keywords
2010-04-18 02:56:09 +02:00
c383defdaa
r300g: always create microtiled zbuffer regardless of texture dimensions
2010-04-18 02:56:09 +02:00
65a3c316ba
i915g: Remove unnecessary headers.
2010-04-17 15:47:45 -07:00
Brian Paul