Move _glapi_* symbols from libGLESv1_CM.so and libGLESv2.so to
libglapi.so. This makes sure an app will get only one copy of glapi in
its address space.
Note that with this change, libGLES* and libglapi must be built from the
same source tree and distributed together. This requirement comes from
the fact that the dispatch offsets used by these libraries are
re-assigned whenever GLAPI XMLs are changed.
In bridge mode, mapi no longer implements glapi.h. It becomes a user of
glapi.h. Imagine an app that uses both libGL.so and libGLESv2.so.
There will be two copies of glapi in the app's memory. It is possible
that _glapi_get_dispatch does not return what _glapi_set_dispatch set,
if they access different copies of the global variables. The solution
to this situation to build either one of the libraries as a bridge to
the other. Or build both libraries as bridges to another shared
glapi library.
When MAPI_MODE_GLAPI is defined, u_current_table is renamed to
_glapi_Dispatch or _glapi_tls_Dispatch. The ASM dispatchers should not
use hardcoded name.
The new implementation is based on mapi. No new script is needed. As
noted in sources.mk, the way to use it is to compile MAPI_GLAPI_SOURCES
with MAPI_MODE_GLAPI defined.
Fix GLAPIPrinter, ES1APIPrinter, and ES2APIPrinter to output files that
are ready for compilation. Since gl_and_es_API.xml is based on
gl_API.xml, the hidden and handcode attributes of entries have to be
overridden for ES1APIPrinter and ES2APIPrinter.
gl_and_es_API.xml defines OpenGL ES 1.1 and 2.0 API as well as OpenGL
API. It consists of gl_API.xml and the newly added es_EXT.xml,
ARB_get_program_binary.xml, OES_single_precision.xml, and
OES_fixed_point.xml.
This fixes a bunch of unnecessary barriers due to the scheduler not
knowing what that arbitrary register description refers to when trying
to reason about its dependencies.
The result is rescheduling in the convolution kernel shader in
Lightsmark, which results in avoiding register spilling and increasing
the performance of the first scene from 6-7 fps midway through the
panning to 11fps. The register spilling was a regression from Mesa
7.9 to Mesa 7.10.
Improves performance of my GLSL demo by 5.1% (+/- 1.4%, n=7). It also
reschedules the giant multiply tree at the end of
glsl-fs-convolution-1 so that we end up not spilling registers,
producing the expected level of performance.
Can happen during swrast fallbacks if a buffer is somehow bound as
a render target and a texture.
Fixes gnome-shell on nv20, and gets it mostly working on nv10.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
sw clears were being used and not getting the correct offsets in the span
code.
also not emitting correct offsets for CB draws to texture levels.
(I've no idea why I'm playing with r100).
This is a candidate for 7.9 and 7.10
According to R700 ISA we have only two channels for cfile constants.
This patch makes piglit tests "glsl1-constant array with constant
indexing" happy on RV710.
Fixes the following Piglit tests:
glslparsertest/shaders/array2.frag
glslparsertest/shaders/dataType6.frag
NOTE: This is a candidate for the 7.9 and 7.10 branches.
This fixes a potential failure when a begin/end_query is the first
thing to happen after flushing the scene.
NOTE: This is a candidate for the 7.10 and 7.9 branches.
This revealed a bug in ra_get_spill_benefit where we only considered
the benefit of the first adjacency we were to remove, explaining some
of the ugly spilling I've seen in shaders. Because of the reduced
spilling, it reduces the runtime of glsl-fs-convolution-1 36.9% +/-
0.9% (n=5).
This was recommended in the original paper, but I figued "make it run"
before "make it fast". Now we make it fast. Reduces the runtime of
glsl-fs-convolution-1 by 12.7% +/- 0.6% (n=5).
Our use of the register allocator in i965 is somewhat unusual.
Whereas most architectures would have a smaller set of registers with
fewer register classes and reuse that across compilation, we have 1,
2, and 4-register classes (usually) and a variable number up to 128
registers per compile depending on how many setup parameters and push
constants are present. As a result, when compiling large numbers of
programs (as with glean texCombine going through ff_fragment_shader),
we spent much of our CPU time in computing the q[] array. By keeping
a separate list of what the conflicts are for a particular reg, we
reduce glean texCombine time 17.0% +/- 2.3% (n=5).
We don't expect this optimization to be useful for 915, which will
have a constant register set, but it would be useful if we were switch
to this register allocator for Mesa IR.
Chia-I Wu
Kenneth Graunke
twied
Henri Verbeet
Dave Airlie
Brian Paul
Eric Anholt
Christian König
Ben Skeggs
Chad Versace
José Fonseca
Andre Maasikas