Christian König [Fri, 24 Feb 2012 15:12:27 +0000 (16:12 +0100)]
vl: use SwapBuffers instead of CopyBuffers
This should speed things up a bit, but also shows
some bugs with the kernel implementation.
v2: require xcb-dri2 version 1.8
Signed-off-by: Christian König <deathsimple@vodafone.de>
Christian König [Fri, 24 Feb 2012 14:33:54 +0000 (15:33 +0100)]
st/xvmc: remove dst buffer texture workaround
Signed-off-by: Christian König <deathsimple@vodafone.de>
Christian König [Thu, 23 Feb 2012 10:13:57 +0000 (11:13 +0100)]
vl: rewrite vl DRI backend using XCB
v2: also set array_size of texture
Signed-off-by: Christian König <deathsimple@vodafone.de>
José Fonseca [Thu, 1 Mar 2012 06:37:18 +0000 (06:37 +0000)]
egl: Don't include strings.h on Windows.
strings.h is needed for the declaration of strcasecmp(), but only on
non-Windows platforms.
Yuanhan Liu [Mon, 19 Sep 2011 10:25:54 +0000 (18:25 +0800)]
mesa: let GL3 buf obj queries not depend on opengl major version
While the ARB_map_buffer_range extension spec says nothing about these
queries -- they were added in GL 3.0 --, it seems like this could be an
error in the extension spec. This is one of the extensions, like
ARB_framebuffer_object, that "back ports" OpenGL 3.0 functionality to
previous versions. These extensions are supposed to provide identical
functionality to OpenGL 3.0. The other cases of mismatches have been
determined to be bugs in the extension specs.
And tools like apitrace rely on such queries to function properly.
Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Signed-off-by: José Fonseca <jfonseca@vmware.com>
Acked-by: Brian Paul <brianp@vmware.com>
Acked-by: Ian Romanick <ian.d.romanick@intel.com>
José Fonseca [Sat, 18 Feb 2012 12:02:17 +0000 (12:02 +0000)]
gallivm: Update comments and prototype of vector-selects.
No runtime behavior change. As vector selects are still not very well
supported by LLVM.
Brian Paul [Thu, 1 Mar 2012 03:53:47 +0000 (20:53 -0700)]
docs: remove mention of components no longer in main Mesa tree
Brian Paul [Thu, 1 Mar 2012 03:53:47 +0000 (20:53 -0700)]
st/mesa: remove old assertion
Brian Paul [Thu, 1 Mar 2012 03:53:46 +0000 (20:53 -0700)]
r300g: init chan=0 to silence warning
Kenneth Graunke [Mon, 27 Feb 2012 19:43:10 +0000 (11:43 -0800)]
i965: Disable PrimitiveID upload.
We currently don't support gl_PrimitiveID, and I believe asking the
hardware to generate it results in vertex cache invalidations.
This could result in slowdowns for applications that use gl_InstanceID,
which would be counter-productive. Just turn it off for now.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
Kenneth Graunke [Thu, 9 Feb 2012 00:40:31 +0000 (16:40 -0800)]
i965: Enable the GL_ARB_draw_instanced extension.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
Kenneth Graunke [Thu, 9 Feb 2012 00:59:39 +0000 (16:59 -0800)]
i965: Fix swizzles for system values such as gl_InstanceID.
visit(ir_variable *) sets dst_reg::writemask to the appropriate channel
for system values. Unfortunately, visit(ir_dereference_variable *) then
calls swizzle_for_size, which for a float, sets the swizzle to .x.
This works for gl_VertexID, since we store it in the .x component (see
brw_draw_upload.c:732 - VID), but fails for gl_InstanceID (IID) since we
store it in the .y channel.
To fix this, avoid calling swizzle_for_size on ir_var_system_values.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
Ian Romanick [Wed, 29 Feb 2012 16:29:39 +0000 (08:29 -0800)]
glsl: Make gl_InstanceID available with GL_ARB_draw_instanced extension
Originally ARB_draw_instanced only specified that ARB decorated name.
Since no vendor actually implemented that behavior and some apps use
the undecorated name, the extension now specifies that both names are
available.
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Marek Olšák <maraeo@gmail.com>
Eric Anholt [Tue, 28 Feb 2012 21:33:53 +0000 (13:33 -0800)]
mesa: Add missing error check for first < 0 in glDrawArraysInstanced().
Fixes piglit GL_ARB_draw_instanced/negative-arrays-first-negative.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Eric Anholt [Tue, 28 Feb 2012 21:33:52 +0000 (13:33 -0800)]
mesa: Fix display lists for draw_elements_base_vertex with draw_instanced.
Fixes piglit GL_ARB_draw_elements_base_vertex/dlist-arb_draw_instanced
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Eric Anholt [Tue, 28 Feb 2012 21:33:51 +0000 (13:33 -0800)]
mesa: Fix display list handling for GL_ARB_draw_instanced.
When you called them in a display list compile before, you would just
end up calling through NULL.
Fixes piglit GL_ARB_draw_instanced/dlist.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Eric Anholt [Tue, 28 Feb 2012 21:33:50 +0000 (13:33 -0800)]
mesa: Fix typo in comment.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Kenneth Graunke [Wed, 29 Feb 2012 20:32:41 +0000 (12:32 -0800)]
i965: Fix Gen6+ dynamic state upper bound on older kernels.
Kernels prior to
271d81b84171d84723357ae6d172ec16b0d8139c (March 2011)
don't support relocations outside of the target buffer object. Rather
than guarding this with a I915_PARAM_HAS_RELAXED_DELTA check, just
smash the bound to 0xfffff001 like we do on Ironlake.
This effectively gives us no upper bound check, just like we did prior
to commit
271d81b84171d84723357ae6d172ec16b0d8139c.
Daniel Vetter would also like to mention that this relies on the guard
page at the end of the GTT.
NOTE: This is a candidate for release branches.
Fixes a regression since
271d81b84171d84723357ae6d172ec16b0d8139c.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=46766
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Eric Anholt [Mon, 6 Feb 2012 11:12:53 +0000 (12:12 +0100)]
egl/drivers: Convert to automake.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Eric Anholt [Sun, 5 Feb 2012 09:46:13 +0000 (10:46 +0100)]
egl/main: Convert to automake.
The drivers/ walk-through-subdirs makefile is converted as well so I
didn't need to keep EGL_DRIVERS_DIRS along with the per-driver
HAVE_EGL_DRIVER_WHATEVER.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Eric Anholt [Sun, 5 Feb 2012 06:54:38 +0000 (07:54 +0100)]
egl: Clean up some flow of EGL platform handling.
The default case code was set up in a separate way, while this makes
it more normal. I wanted to add code to the explicit x11 platform and
default x11 platform cases in the next commit.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Eric Anholt [Sun, 5 Feb 2012 06:06:56 +0000 (07:06 +0100)]
egl: Drop the non-builtin-drivers support.
The two drivers both did builtin mode, so drop the non-builtin build
system.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Eric Anholt [Sun, 5 Feb 2012 06:01:17 +0000 (07:01 +0100)]
egl: Drop _EGL_MAIN entrypoint obfuscation.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Eric Anholt [Sun, 5 Feb 2012 05:10:56 +0000 (06:10 +0100)]
glx: Convert to automake.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Mathias Fröhlich [Wed, 29 Feb 2012 17:19:35 +0000 (18:19 +0100)]
mesa: Push the shine table into the tnl module.
All users of the shine table outside of the tnl module
are gone. Move the implementation into the tnl module and
prefix the public functions with _tnl.
Reviewed-by: Alex Deucher <alexdeucher@gmail.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Signed-off-by: Mathias Froehlich <Mathias.Froehlich@web.de>
Mathias Fröhlich [Wed, 29 Feb 2012 17:19:34 +0000 (18:19 +0100)]
mesa: Do not invalidate shine tables in compute light positions.
Since the shine tables are now only used in the tnl lighting stage, where
they are validated through the tnl driver function NotifyMaterialChange
called in tnl/t_vb_light.c, we can not omit calling
_mesa_validate_all_lighting_tables (which only validates the shine tables)
in main/light.c.
Reviewed-by: Alex Deucher <alexdeucher@gmail.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Signed-off-by: Mathias Froehlich <Mathias.Froehlich@web.de>
Mathias Fröhlich [Wed, 29 Feb 2012 17:19:34 +0000 (18:19 +0100)]
mesa: Avoid revalidating shine tables in _tnl_RasterPos.
Now that _tnl_RasterPos no longer uses the shine tables, avoid
revalidating them.
Reviewed-by: Alex Deucher <alexdeucher@gmail.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Signed-off-by: Mathias Froehlich <Mathias.Froehlich@web.de>
Mathias Fröhlich [Wed, 29 Feb 2012 17:19:34 +0000 (18:19 +0100)]
mesa: Avoid using the shine table for _tnl_RasterPos
Use direct computation of pow for computing the shininess
in _tnl_RasterPos. Since the _tnl_RasterPos function is still
used by plenty drivers that do only need the shine table for
_tnl_RasterPos but do not make use of swtnl computations, this
enables pushing down the shine table computation and validation
into the tnl module, which will happen in a followup change.
Reviewed-by: Alex Deucher <alexdeucher@gmail.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Signed-off-by: Mathias Froehlich <Mathias.Froehlich@web.de>
Mathias Fröhlich [Wed, 29 Feb 2012 17:19:34 +0000 (18:19 +0100)]
mesa: Avoid explicit invalidation of shine tables.
Since the shine tables are implicitly invalidated by having
a different shininess value than the current one, we can
omit the explicit invalidation of the shine table.
Reviewed-by: Alex Deucher <alexdeucher@gmail.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Signed-off-by: Mathias Froehlich <Mathias.Froehlich@web.de>
Mathias Fröhlich [Wed, 29 Feb 2012 17:19:33 +0000 (18:19 +0100)]
mesa: Remove gl_light_attrib::_Flags.
This variable is only used locally in _mesa_update_lighting.
Reviewed-by: Alex Deucher <alexdeucher@gmail.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Signed-off-by: Mathias Froehlich <Mathias.Froehlich@web.de>
Mathias Fröhlich [Wed, 29 Feb 2012 17:19:33 +0000 (18:19 +0100)]
mesa: Remove _CosCutoffNeg from light state.
It is only used as a temporary variable during computation of
_CosCutoff. So, don't store it.
Reviewed-by: Alex Deucher <alexdeucher@gmail.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Signed-off-by: Mathias Froehlich <Mathias.Froehlich@web.de>
Brian Paul [Tue, 28 Feb 2012 14:48:34 +0000 (07:48 -0700)]
util: replace format equality test with compatibility test in blit code
This lets us use the resource_copy_region() path when blitting from
R8G8B8A8 to R8G8B8x8, for example.
v2: be smarter when src_format==dst_format
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: José Fonseca <jfonseca@vmware.com>
Brian Paul [Tue, 28 Feb 2012 14:45:17 +0000 (07:45 -0700)]
util: better comment for util_is_format_compatible()
Brian Paul [Tue, 28 Feb 2012 03:28:09 +0000 (20:28 -0700)]
mesa: use arrayObj var in more places in client_state()
Brian Paul [Tue, 28 Feb 2012 03:28:09 +0000 (20:28 -0700)]
mesa: asst. fixes for texture arrays and borders
For 1D array textures, there is no border on the height dimension.
For 2D array textures, there is no border on the depth dimension.
Brian Paul [Tue, 28 Feb 2012 03:28:09 +0000 (20:28 -0700)]
mesa: check for no state change in VertexAttribDivisor()
Reviewed-by: José Fonseca <jfonseca@vmware.com>
Brian Paul [Tue, 28 Feb 2012 03:28:09 +0000 (20:28 -0700)]
util: fix assertions in u_blitter.c code
Assertions of the form assert(a && b) should be written as separate assertions
so that you can actually tell which part is false when there's a failure.
Reviewed-by: Marek Olšák <maraeo@gmail.com>
Brian Paul [Tue, 28 Feb 2012 03:28:09 +0000 (20:28 -0700)]
mesa: reorder things in mtypes.h
Move structs, enums, etc so they're in more logical order. In particular,
the shader and transform feedback-related structs/enums were pretty
scattered around.
Brian Paul [Wed, 29 Feb 2012 14:52:18 +0000 (07:52 -0700)]
svga: fix max_lod clamping in update_tss_binding()
After biasing we need to clamp to be sure we don't exceed the number of
levels in the mipmap. This fixes an assertion at svga_sampler_view.c:70
v2: simplify the biasing, clamping code per Jose's suggestion.
Reviewed-by: José Fonseca <jfonseca@vmware.com>
Christian König [Tue, 21 Feb 2012 13:09:42 +0000 (14:09 +0100)]
st/vdpau: fix use of *.o in Makefile.vdpau
Signed-off-by: Christian König <deathsimple@vodafone.de>
Christian König [Tue, 21 Feb 2012 13:08:14 +0000 (14:08 +0100)]
st/xvmc: fix use of *.o in Makefile.xvmc
Signed-off-by: Christian König <deathsimple@vodafone.de>
Eric Anholt [Mon, 27 Feb 2012 17:52:05 +0000 (09:52 -0800)]
i965: Avoid blocking on the GPU for setting the HiZ op vertex data.
We need to allocate new space every time to avoid blocking on the last
HiZ op completing. There are two easy ways to do this:
brw_state_batch() and intel_upload_data(). brw_state_batch() is
simpler and avoids another buffer allocation.
Improves Unigine Tropics performance 0.376416% +/- 0.148722% (n=7).
Reviewed-by: Chad Versace <chad.versace@linux.intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Kenneth Graunke [Fri, 10 Feb 2012 04:33:44 +0000 (20:33 -0800)]
glcpp: Don't strlen() the output for every token being printed.
The ralloc string appending functions were originally intended for
simple, non-hot-path uses like printing to an info log.
Cuts Unigine Tropics load time by around 20% (6 seconds).
v2: Avoid strlen() on every newline, too.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> [v1]
Acked-by: José Fonseca <jfonseca@vmware.com> [v1]
Kenneth Graunke [Fri, 10 Feb 2012 04:03:36 +0000 (20:03 -0800)]
ralloc: Make rewrite_tail increase "start" by the new text's length.
Both callers of rewrite_tail immediately compute the new total string
length by adding the (known) length of the existing string plus the
length of the newly appended text. Unfortunately, callers generally
won't know the length of the new text, as it's printf-formatted.
Since ralloc already computes this length, it makes sense to add it in
and save the caller the effort. This simplifies both existing callers,
but more importantly, will allow for cheap-appending in the next commit.
v2: The link_uniforms code needs both the old and new length.
Apply the obvious fix (which sadly makes it less of a cleanup).
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> [v1]
Acked-by: José Fonseca <jfonseca@vmware.com> [v1]
Dave Airlie [Mon, 6 Feb 2012 16:32:59 +0000 (16:32 +0000)]
gallivm: add major integer opcodes to the tgsi action handler
This adds support for all the opcodes needed for native integer
support with GLSL 1.20 enabled, and some of the ones for GLSL1.30
support.
I've split them between non-cpu and cpu along the same lines
Tom's code did for the other ones I think, but I'm open to review
on which ones should go where.
With instance ids fixed I get no regressions on my box here
with LLVM 2.8, will test with later LLVMs as well.
Signed-off-by: Dave Airlie <airlied@redhat.com>
Dave Airlie [Tue, 28 Feb 2012 11:00:03 +0000 (11:00 +0000)]
gallivm: drop deprecated opcodes
These are integer opcodes not deprecated ones.
Signed-off-by: Dave Airlie <airlied@redhat.com>
Dave Airlie [Tue, 28 Feb 2012 11:23:04 +0000 (11:23 +0000)]
gallivm: only do rcp/mul for floating
rcp asserts on type.floating so don't go passing non-floating
things into it.
Signed-off-by: Dave Airlie <airlied@redhat.com>
José Fonseca [Mon, 27 Feb 2012 11:12:12 +0000 (11:12 +0000)]
svga: Advertise SVGA3D_DEVCAP_MAX_POINT_SIZE.
Backends usually advertise a SVGA3D_DEVCAP_MAX_POINT_SIZE between 63 and
256, so an hardcoded max point size of 80 is often incorrect.
This limitation does not apply for anti-aliased points (as they are done
via draw module) but we still advertise the same limit for both, because
all others pipe drivers do.
Reviewed-by: Brian Paul <brianp@vmware.com>
Neil Roberts [Sun, 26 Feb 2012 01:33:40 +0000 (01:33 +0000)]
mesa: Don't disable fast path for normalized types
Mesa has a fast path for the generic fallback when using glReadPixels
for RGBA data which uses memcpy. However it was really difficult to
hit this case because it would not be used if any transferOps are
enabled. Any type apart from floating point or non-normalized integer
types (so any of the common types) would force enabling clamping so
the fast path could not be used. This patch makes it ignore clamping
when determining whether to use the fast path if the data type of the
buffer is an unsigned normalized type because in that case clamping
will not have any effect anyway.
https://bugs.freedesktop.org/show_bug.cgi?id=46631
NOTE: This is a candidate for the 8.0 branch.
Signed-off-by: Brian Paul <brianp@vmware.com>
Brian Paul [Tue, 28 Feb 2012 14:54:43 +0000 (07:54 -0700)]
gallium: remove trailing comma to silence warning
Brian Paul [Tue, 28 Feb 2012 14:51:01 +0000 (07:51 -0700)]
mesa: minor comment, whitespace fixes in teximage.c
Dave Airlie [Tue, 28 Feb 2012 10:43:14 +0000 (10:43 +0000)]
gallivm: add frem support to the lp_build_mod helper.
for completeness.
Signed-off-by: Dave Airlie <airlied@redhat.com>
Dave Airlie [Fri, 17 Feb 2012 19:02:21 +0000 (19:02 +0000)]
gallivm: add bitarit xor and not ops.
Signed-off-by: Dave Airlie <airlied@redhat.com>
Dave Airlie [Thu, 9 Feb 2012 19:44:55 +0000 (19:44 +0000)]
st/mesa: don't unreference user attribs up front.
postpone unreferences until end of function, as the ones in use will
get naturally dereferenced.
Reviewed-by: Brian Paul <brianp@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Dave Airlie [Thu, 9 Feb 2012 19:44:56 +0000 (19:44 +0000)]
mesa/vbo: inline vbo_sizeof_ib_type.
Can't see any reason this wouldn't be better off as an inline.
Reviewed-by: Brian Paul <brianp@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Dave Airlie [Fri, 17 Feb 2012 19:02:23 +0000 (19:02 +0000)]
gallivm: add integer and unsigned mod arit functions. (v2)
use a single entry point, as per Jose's suggestion.
Signed-off-by: Dave Airlie <airlied@redhat.com>
Brian Paul [Tue, 28 Feb 2012 03:28:09 +0000 (20:28 -0700)]
tgsi: remove trailing comma to silence warning
Brian Paul [Tue, 28 Feb 2012 03:28:09 +0000 (20:28 -0700)]
xlib: silence unused var warning
José Fonseca [Mon, 27 Feb 2012 18:52:00 +0000 (18:52 +0000)]
svga: Remove unused SVGA_TEX_UNITS constant.
José Fonseca [Mon, 27 Feb 2012 11:21:32 +0000 (11:21 +0000)]
svga: Clamp advertised PIPE_SHADER_CAP_MAX_TEMPS to SVGA3D_TEMPREG_MAX.
Some backends may advertise more temps than SVGA3D_TEMPREG_MAX, but the
driver is hardwired to only support up to the value defined by
SVGA3D_TEMPREG_MAX, so clamp to it.
Reviewed-by: Brian Paul <brianp@vmware.com>
Marek Olšák [Sun, 26 Feb 2012 18:15:45 +0000 (19:15 +0100)]
r600g: use u_default_transfer_flush_region for all resource types
Marek Olšák [Sun, 26 Feb 2012 18:00:26 +0000 (19:00 +0100)]
r600g: use u_default_transfer_inline_write for all resource types
Marek Olšák [Sun, 26 Feb 2012 17:56:51 +0000 (18:56 +0100)]
gallium/util: add fast path for buffers in u_default_transfer_inline_write
v2: fix indentation, add assertions
Marek Olšák [Sun, 26 Feb 2012 17:48:42 +0000 (18:48 +0100)]
gallium/util: set correct usage flags in u_default_transfer_inline_write
The DISCARD flags should improve performance in drivers which handle them.
Marek Olšák [Sun, 26 Feb 2012 16:07:52 +0000 (17:07 +0100)]
r600g: fix streamout cache flush for r600
Figured out by trial and error.
Christian König [Mon, 27 Feb 2012 11:57:24 +0000 (12:57 +0100)]
vl: fix a douple free in xsp winsys backend
There are a couple of more bugs, but it is only
useful for debugging anyway.
Signed-off-by: Christian König <deathsimple@vodafone.de>
Christian König [Mon, 27 Feb 2012 11:53:49 +0000 (12:53 +0100)]
vl: adjust matrix and median filter to removal of PIPE_SHADER_CAP_OUTPUT_READ
Signed-off-by: Christian König <deathsimple@vodafone.de>
Marek Olšák [Sun, 26 Feb 2012 18:44:45 +0000 (19:44 +0100)]
r600g: cleanup r600_transfer_struct
Especially rename staging_texture to staging and change its type
to r600_resource. I will reuse it for buffers later.
Marek Olšák [Thu, 23 Feb 2012 23:38:33 +0000 (00:38 +0100)]
r600g: check for R600_STREAMOUT env var in winsys
Marek Olšák [Fri, 24 Feb 2012 16:17:57 +0000 (17:17 +0100)]
r600g: move initialization of use_surface flag into screen_create
Also change the type to bool and give it a less ambiguous name.
Marek Olšák [Fri, 24 Feb 2012 16:14:59 +0000 (17:14 +0100)]
r600g: properly check whether texture is busy in get_transfer
Marek Olšák [Fri, 24 Feb 2012 16:13:19 +0000 (17:13 +0100)]
r600g: rename r600_resource_texture::depth to bool is_depth
It's used as a boolean.
Marek Olšák [Thu, 23 Feb 2012 22:44:36 +0000 (23:44 +0100)]
gallium: remove PIPE_SHADER_CAP_OUTPUT_READ
r600g is the only driver which has made use of it. The reason the CAP was
added was to fix some piglit tests when the GLSL pass lower_output_reads
didn't exist.
However, not removing output reads breaks the fallback for glClampColorARB,
which assumes outputs are not readable. The fix would be non-trivial
and my personal preference is to remove the CAP, considering that reading
outputs is uncommon and that we can now use lower_output_reads to fix
the issue that the CAP was supposed to workaround in the first place.
Marek Olšák [Fri, 24 Feb 2012 18:29:48 +0000 (19:29 +0100)]
gallium/rtasm: properly detect SSE and SSE2
This should fix crashes on ancient processors.
Tom Stellard [Mon, 20 Feb 2012 00:38:58 +0000 (19:38 -0500)]
r300g: Use automake to generate Makefile v3
v2:
- s/$(top_builddir)/$(top_srcdir)/
- Always generate Makefile.in
v3:
- Fixes from Matt Turner
- Use Mesa CFLAGS
Tom Stellard [Fri, 24 Feb 2012 14:00:01 +0000 (09:00 -0500)]
r300g: Reorganize the compiler unit tests
Tom Stellard [Sun, 26 Feb 2012 20:12:36 +0000 (15:12 -0500)]
r300/compiler: Schedule KIL instructions before output writes
Tom Stellard [Tue, 14 Feb 2012 02:26:15 +0000 (21:26 -0500)]
r300/compiler: Use the smart scheduler for r300 cards
Tom Stellard [Tue, 14 Feb 2012 02:27:28 +0000 (21:27 -0500)]
r300/compiler: Fix bug when lowering KILP on r300 cards
KILP instruction inside IF blocks were being lowered to an unconditional
KIL. Since r300 doesn't support branching, when the IF's were lowered
to conditional moves, the KIL would always be executed. This is not a
problem with the mesa state tracker, because the GLSL compiler handles
lowering IF's, but this bug was appearing in the VDPAU state tracker,
which does not use the GLSL compiler.
Note: This is a candidate for the stable branches.
Christian König [Sat, 25 Feb 2012 11:14:58 +0000 (12:14 +0100)]
vl/compositor: fix a simple typo
Otherwise the dirty area tracking won't work correctly.
Signed-off-by: Christian König <deathsimple@vodafone.de>
Christian König [Wed, 15 Feb 2012 16:30:58 +0000 (17:30 +0100)]
st/xvmc: move xvmc state tracker out of xorg subdir
The xvmc state tracker is completely seperate and
doesn't shares code or anything else with the
xorg state tracker.
Signed-off-by: Christian König <deathsimple@vodafone.de>
Christian König [Wed, 15 Feb 2012 16:20:50 +0000 (17:20 +0100)]
vl: rework winsys interface
Throw out all the old and now unneeded stuff.
Signed-off-by: Christian König <deathsimple@vodafone.de>
Christian König [Wed, 15 Feb 2012 13:57:35 +0000 (14:57 +0100)]
vl: cleanup dri winsys abstraction
There was way to much dead code in it.
Signed-off-by: Christian König <deathsimple@vodafone.de>
Vinson Lee [Sun, 19 Feb 2012 07:43:03 +0000 (23:43 -0800)]
scons: Don't build the assembly sources on Mac OS X.
This patch allows the Mac OS X SCons build to complete. The assembly
sources contain psuedo-ops that not are supported on Mac OS X.
Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Reviewed-by: José Fonseca <jfonseca@vmware.com>
Stéphane Marchesin [Fri, 24 Feb 2012 23:17:27 +0000 (15:17 -0800)]
i915g: Fix fallout from
8e4540ec2a82e72be491bc8fe23c10551d29a96c
Fixes piglit regressions from that change.
Zack Rusin [Fri, 24 Feb 2012 19:26:41 +0000 (14:26 -0500)]
svga: Fix stencil op mapping
We were inverting the meaning of the stencil op flags: in svga/d3d
the normal incr/decr wraps and the SAT ops clamp.
This fixes piglit failures (at least stencil-twoside and stencil-wrap).
We should backport this everywhere we can.
Reviewed-by: Brian Paul <brianp@vmware.com>
Brian Paul [Fri, 24 Feb 2012 16:46:44 +0000 (17:46 +0100)]
svga: fix use of SVGA3D_x vs. PIPE_FORMAT_x in svga_is_format_supported()
Two of the switch cases used PIPE_FORMAT_ tokens instead of SVGA3D_ tokens.
As it happens, the token values are equal for these formats so there's no
net change.
Reviewed-by: José Fonseca <jfonseca@vmware.com>
Reviewed-by: Jakob Bornecrantz <jakob@vmware.com>
Brian Paul [Fri, 24 Feb 2012 16:27:35 +0000 (09:27 -0700)]
mesa/gdi: include swrast.h to fix compilation
Brian Paul [Tue, 21 Feb 2012 15:26:30 +0000 (08:26 -0700)]
st/mesa: initialize the MaxViewport, MaxRenderbufferSize constants
Use the max 2D/rect texture size as the limit. If that's not true
for some devices we'll need new PIPE_CAP_ queries.
Brian Paul [Mon, 20 Feb 2012 20:13:55 +0000 (13:13 -0700)]
mesa: move more swrast-related #defines out of core Mesa
Brian Paul [Mon, 20 Feb 2012 20:01:17 +0000 (13:01 -0700)]
mesa: remove STENCIL_BITS use
Brian Paul [Tue, 21 Feb 2012 15:01:05 +0000 (08:01 -0700)]
xlib: remove STENCIL_BITS
Brian Paul [Mon, 20 Feb 2012 20:00:17 +0000 (13:00 -0700)]
st/glx: remove STENCIL_BITS, DEFAULT_SOFTWARE_DEPTH_BITS
Remove some Mesa/swrast stuff.
Brian Paul [Mon, 20 Feb 2012 18:07:01 +0000 (11:07 -0700)]
mesa: remove last of MAX_WIDTH, MAX_HEIGHT
Define new MAX_VIEWPORT_WIDTH/HEIGHT and MAX_RENDERBUFFER_SIZE values
instead.
Brian Paul [Mon, 20 Feb 2012 18:07:01 +0000 (11:07 -0700)]
mesa: remove MAX_WIDTH from prog_execute.h
define a PROG_MAX_WIDTH var instead. It has to match MAX_WIDTH in
swrast. More elaborate refactoring could fix that (someday).
Brian Paul [Mon, 20 Feb 2012 18:07:01 +0000 (11:07 -0700)]
st/glx: remove MAX_WIDTH/HEIGHT usage
Brian Paul [Mon, 20 Feb 2012 18:07:01 +0000 (11:07 -0700)]
intel: remove MAX_WIDTH usage in intelInitContext()
Brian Paul [Mon, 20 Feb 2012 18:07:00 +0000 (11:07 -0700)]
swrast: check max renderbuffer size against SWRAST_MAX_WIDTH
Brian Paul [Mon, 20 Feb 2012 18:07:00 +0000 (11:07 -0700)]
osmesa: use SWRAST_MAX_WIDTH/HEIGHT
Brian Paul [Mon, 20 Feb 2012 18:07:00 +0000 (11:07 -0700)]
dri/swrast: use SWRAST_MAX_WIDTH/HEIGHT