Brian Paul
4414a1a73c
llvmpipe: assorted clean-ups in texture code
2010-01-14 14:19:16 -07:00
Brian Paul
2ba1c8189a
llvmpipe: use one loop in lp_rast_clear_zstencil()
...
This is just a tiny bit faster.
2010-01-13 18:58:42 -07:00
Brian Paul
f94a99170e
llvmpipe: optimize lp_rast_clear_color() for non-gray colors
...
This makes a big difference in progs that clear to a non-gray color.
Some demos are 30-50% faster.
2010-01-13 18:54:50 -07:00
Brian Paul
f19f218e7a
llvmpipe: minor assorted clean-ups
2010-01-13 18:01:45 -07:00
Brian Paul
018b78ad64
llvmpipe: check render targets before other textures
2010-01-13 15:32:55 -07:00
Brian Paul
1287277446
llvmpipe: also check render target textures in lp_setup_is_texture_referenced()
2010-01-13 15:30:42 -07:00
Brian Paul
0b279c5382
llvmpipe: indentation fixes
2010-01-13 15:10:57 -07:00
Brian Paul
db58192cfb
llvmpipe: re-get scene pointer after flushing
2010-01-13 15:10:57 -07:00
Brian Paul
d59fe44896
llvmpipe: check for texture usage in all scenes
2010-01-13 15:10:57 -07:00
Brian Paul
a27b12171d
llvmpipe: add scene texture referencing code
2010-01-13 15:10:57 -07:00
Brian Paul
592e40aa7b
llvmpipe: added scene functions for texture reference counting
...
When a texture is used in the scene we add it to a list of texture
references. The lp_scene_is_textured_referenced() function tells
us if a texture is referenced by the scene.
2010-01-13 15:10:57 -07:00
Brian Paul
4769328fe1
llvmpipe: comments
2010-01-13 15:10:57 -07:00
Brian Paul
a36395d487
llvmpipe: comments
2010-01-13 15:10:57 -07:00
Brian Paul
9be1feacf2
llvmpipe: added llvmpipe_texture_const()
2010-01-13 15:10:56 -07:00
Brian Paul
ddb94661a4
llvmpipe: minor comment upgrades
2010-01-13 15:10:56 -07:00
José Fonseca
7df4c88088
llvmpipe: Opaque shader implies complete colormask too.
2010-01-13 22:07:24 +00:00
José Fonseca
a1acbff299
llvmpipe: Reset the bin when shading a whole tile with an opaque shader.
2010-01-13 21:51:47 +00:00
Keith Whitwell
4231006e29
llvmpipe: add bin debugger
...
Adjust definition of empty_bin according to what's actually in empty
bins. We often have a state packet before/after load commands.
Still need to do something about the fence packets.
2010-01-13 20:14:04 +00:00
Keith Whitwell
95ee14f147
llvmpipe: implement lp_rast_load_zstencil
...
Load zbuffer contents for binned scenes that don't start with a clear
and which have a bound zbuffer.
2010-01-13 16:52:44 +00:00
Brian Paul
0bb5c3060f
llvmpipe: added debug warning
2010-01-13 09:32:21 -07:00
Brian Paul
4439aab7b7
llvmpipe: comments and LLVMValueRef naming
2010-01-13 09:32:21 -07:00
Brian Paul
212f3a6cb3
llvmpipe: tweak subpixel_snap() arithmetic
...
This adjustment fixes some rasterization differences between llvmpipe
and softpipe (and other renderers).
2010-01-13 09:32:21 -07:00
Keith Whitwell
db83ad4b43
llvmpipe: improve empty-bin test further
...
Remove unused param, add comments. Thanks to Brian for review.
2010-01-13 16:29:39 +00:00
Keith Whitwell
f4b29e6ad3
llvmpipe: improve empty-bin test
...
We emit at most two clear packets (color and z respectively).
2010-01-13 15:49:24 +00:00
Keith Whitwell
da45f49cc6
llvmpipe: quick hack to short-circuit empty bins
2010-01-13 15:18:32 +00:00
Brian Paul
7e4c75c040
llvmpipe: fix indentation, comment typo
2010-01-12 17:12:49 -07:00
Brian Paul
ec9cfac768
llvmpipe: debug checks: make sure scene is empty at key points
2010-01-12 17:12:07 -07:00
Brian Paul
c560b97b17
llvmpipe: assert that we're putting data into a valid bin
2010-01-12 17:11:40 -07:00
Brian Paul
214ffad015
llvmpipe: clamp maxx,maxy to framebuffer size (in terms of tiles)
...
In some corner cases the right-most / bottom-most vertex can be
right on the edge of the framebuffer. Because the maxx, maxy vals
are computed with a series of float/int, pixel/tile transformations
we can end up with maxx >= scene->x_tiles or maxy >= scene->y_tiles.
This leads to putting data into bins that never get processed, or
reset. This becomes stale data that can lead to segfaults.
Clamping fixes this.
2010-01-12 17:08:07 -07:00
Brian Paul
de10168a46
llvmpipe: added lp_scene_is_empty()
2010-01-12 17:06:19 -07:00
Brian Paul
4061ca02dd
llvmpipe: silence unused var warnings
2010-01-12 13:01:32 -07:00
Brian Paul
a32e9b2a2d
llvmpipe: remove unused #define
2010-01-12 10:11:36 -07:00
Brian Paul
5cf4630969
llvmpipe: disable the all in/out test code for now
...
It's still faster not to try to special case the "all pixels are
known to be inside the triangle" case.
2010-01-11 15:30:56 -07:00
Brian Paul
9a10d14a44
llvmpipe: move, update comments
2010-01-11 15:30:17 -07:00
Brian Paul
3b5d849268
llvmpipe: refactor generate_fragment() code
...
This will make it easier to generate multiple versions of the fragment
code per variant.
2010-01-11 13:16:02 -07:00
Brian Paul
46b5bd6cad
llvmpipe: do the all-in test on the scalar c0 instead of vector c0
...
This still isn't faster, but committing it for posterity.
2010-01-11 12:59:39 -07:00
Keith Whitwell
86f450060d
llvmpipe: force constant interpolation of flatshade colors
...
Nice speedup for gears.
2010-01-11 12:12:59 +00:00
Keith Whitwell
ad74ea2869
st/mesa: early exit on error path
...
Can't rely on asserts having any effect on flowcontrol for release
builds.
2010-01-11 12:02:53 +00:00
Keith Whitwell
094525fb23
llvmpipe: remove opencoded constant
2010-01-11 11:06:15 +00:00
Keith Whitwell
16c1ad54bc
llvmpipe: remove scissor cliprect derived state
...
Was previously calculating the intersection of the scissor rectangle
and the framebuffer dimensions. Rendering is already restricted to
framebuffer dimensions by other means, so scissor testing (when
implemented) can just use the scissor state directly.
2010-01-11 11:06:15 +00:00
Keith Whitwell
c1a0441602
llvmpipe: initial mrt support
...
Non-mrt apps work, and the code looks correct, but not many mrt test apps
handy atm...
2010-01-10 17:22:09 +00:00
Brian Paul
f4321fbd96
llvmpipe: optimize case when all four pixels are inside the triangle
...
When the incoming c0,c1,c2 values are equal to INT_MIN it means that
all pixels are inside the triangle. Thus we can skip the detailed
pixel inside/outside triangle tests. Use the new lp_build_if()/endif()
functions to generate the branching code.
The code is disabled ATM however because it's actually a little slower
than the original code. A little more tuning may fix that though...
2010-01-08 14:49:34 -07:00
Brian Paul
5208af7853
llvmpipe: fix more if/else/endif design bugs
2010-01-08 12:47:30 -07:00
Brian Paul
855d7f51e4
llvmpipe: move some fields to the private lp_build_flow_if struct
2010-01-08 11:32:36 -07:00
Brian Paul
af31e65b55
llvmpipe: free the phi array
2010-01-08 11:20:38 -07:00
Brian Paul
70b8d59792
llvmpipe: checkpoint if/else/endif contructs work
...
The LLVM IR looks correct now. Basic blocks are where they're supposed
to be and the Phi functions have the right (var,block) information.
2010-01-08 11:06:16 -07:00
José Fonseca
080c40ab32
Merge remote branch 'origin/master' into lp-binning
...
Conflicts:
src/gallium/auxiliary/util/u_surface.c
src/gallium/drivers/llvmpipe/Makefile
src/gallium/drivers/llvmpipe/SConscript
src/gallium/drivers/llvmpipe/lp_bld_arit.c
src/gallium/drivers/llvmpipe/lp_bld_flow.c
src/gallium/drivers/llvmpipe/lp_bld_interp.c
src/gallium/drivers/llvmpipe/lp_clear.c
src/gallium/drivers/llvmpipe/lp_context.c
src/gallium/drivers/llvmpipe/lp_context.h
src/gallium/drivers/llvmpipe/lp_draw_arrays.c
src/gallium/drivers/llvmpipe/lp_jit.c
src/gallium/drivers/llvmpipe/lp_jit.h
src/gallium/drivers/llvmpipe/lp_prim_vbuf.c
src/gallium/drivers/llvmpipe/lp_setup.c
src/gallium/drivers/llvmpipe/lp_setup_point.c
src/gallium/drivers/llvmpipe/lp_state.h
src/gallium/drivers/llvmpipe/lp_state_blend.c
src/gallium/drivers/llvmpipe/lp_state_derived.c
src/gallium/drivers/llvmpipe/lp_state_fs.c
src/gallium/drivers/llvmpipe/lp_state_sampler.c
src/gallium/drivers/llvmpipe/lp_state_surface.c
src/gallium/drivers/llvmpipe/lp_tex_cache.c
src/gallium/drivers/llvmpipe/lp_tex_cache.h
src/gallium/drivers/llvmpipe/lp_tex_sample.h
src/gallium/drivers/llvmpipe/lp_tile_cache.c
2010-01-08 15:42:57 +00:00
José Fonseca
9cdf6f025b
scons: Set the default windows platform to be windows userspace.
...
I thought I had done this ages ago.
2010-01-08 14:31:25 +00:00
Dan Nicholson
a1de400e8d
mklib: Extract archives into temporary directories
...
When static libraries are created from other archives, objects are
extracted and then deleted when the static library is done. This can
race when there are multiple static libraries being created from the
same archives as with libmesa.a and libmesagallium.a.
Should fix this issue with parallel jobs:
make[5]: *** No rule to make target
> `../../../../../../src/mesa/libmesagallium.a', needed by
> `radeon_dri.so'. Stop
Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
Reported-and-Tested-by: Sedat Dilek <sedat.dilek@gmail.com>
2010-01-08 06:01:22 -08:00
Michal Krol
dddd5a36ac
tgsi: Cleanup dot-product opcodes in interpreter.
2010-01-08 12:03:55 +01:00