Brian Paul [Fri, 12 Feb 2010 22:40:57 +0000 (15:40 -0700)]
st/mesa: improved draw_range_elements code
Don't use pipe->draw_range_elements() if min_index=max_index=~0 since
that doesn't provide any useful info.
Also, implement the loop around pipe->draw_range_elements() when
nr_prims > 1.
Brian Paul [Fri, 12 Feb 2010 21:46:51 +0000 (14:46 -0700)]
st/mesa: rename state -> tgsi, updated comments
Brian Paul [Fri, 12 Feb 2010 21:46:10 +0000 (14:46 -0700)]
st/mesa: remove unused var
Vinson Lee [Fri, 12 Feb 2010 22:36:11 +0000 (14:36 -0800)]
intel: Remove unused variable.
Brian Paul [Fri, 12 Feb 2010 22:24:54 +0000 (15:24 -0700)]
cell: fix-up C, CPP flags
Flags like -Wmissing-prototypes don't apply to g++.
Brian Paul [Fri, 12 Feb 2010 22:24:24 +0000 (15:24 -0700)]
cell: implement PIPE_CAP_MAX_COMBINED_SAMPLERS query
Brian Paul [Fri, 12 Feb 2010 19:35:18 +0000 (12:35 -0700)]
cell: remove -mcpu=cell from CFLAGS
This reverts part of commit
ebe2f7609533645e7e9c8af4a55bc0127b9bc515
and allows things to build here.
Kristian Høgsberg [Thu, 11 Feb 2010 22:19:34 +0000 (17:19 -0500)]
intel: Avoid dri_bo_flink() in intel_update_renderbuffers()
Now that we track the global name in struct intel_region, we can just
look it up there.
Kristian Høgsberg [Thu, 11 Feb 2010 22:18:01 +0000 (17:18 -0500)]
intel: Track named regions and make sure we only have one region per named bo
Kristian Høgsberg [Thu, 11 Feb 2010 22:10:19 +0000 (17:10 -0500)]
dri: Store the loader private passed in at dri context creation
We just threw it away before, but we haven't had a use for it yet.
Zack Rusin [Thu, 11 Feb 2010 03:15:24 +0000 (22:15 -0500)]
st/dri: don't expose visuals we'll have trouble supporting
without getBuffersWithFormat some visuals will just cause headaches (crashes),
so if we're running on an older system simply don't advertise them.
Zack Rusin [Thu, 11 Feb 2010 00:00:35 +0000 (19:00 -0500)]
st/dri, st/xorg: fix buffers that have attachements of different depth
we actually need to specify the formats for different attachements, otherwise
if the color buffer is 24bpp and the app asks for 16bpp depth buffer than
we end up fetching the depth from the drawable which is 24bpp and end up
creating the wrong depth buffer. use the new getBuffersWithFormat extension
to pass the depth correctly.
Roland Scheidegger [Fri, 12 Feb 2010 20:43:45 +0000 (21:43 +0100)]
svga: enable code to emit blend color
Roland Scheidegger [Fri, 12 Feb 2010 20:39:29 +0000 (21:39 +0100)]
Merge branch 'gallium-dynamicstencilref'
Vinson Lee [Fri, 12 Feb 2010 19:59:37 +0000 (11:59 -0800)]
progs/tests: Add checks for GL_NV_vertex_program.
Jerome Glisse [Fri, 12 Feb 2010 18:25:16 +0000 (19:25 +0100)]
Revert "r600c: substract vbo offset to vbo size"
This reverts commit
325614a882f5371da512536e07c443a172ffb87c.
Ok this was already fixed this commit actualy broke things
Patrice Mandin [Fri, 12 Feb 2010 18:22:41 +0000 (19:22 +0100)]
nouveau: nv30: it may have 8 texture samplers (not 4), but shader engine has 16 texture units
Signed-off-by: Patrice Mandin <patmandin@gmail.com>
Brian Paul [Fri, 12 Feb 2010 18:07:01 +0000 (11:07 -0700)]
cell: adjust structure padding
The size of some core Gallium state structures has changed so the
padding on some cell structs needed to be adjusted to keep them
a multiple of 16 bytes.
Brian Paul [Fri, 12 Feb 2010 18:06:19 +0000 (11:06 -0700)]
cell: remove unused var
Brian Paul [Fri, 12 Feb 2010 18:06:04 +0000 (11:06 -0700)]
cell: include cell_context.h
Brian Paul [Fri, 12 Feb 2010 16:47:34 +0000 (09:47 -0700)]
cell: re-fix undeclared var
Brian Paul [Fri, 12 Feb 2010 16:43:18 +0000 (09:43 -0700)]
cell: declare stencil_value variable
Jerome Glisse [Fri, 12 Feb 2010 16:06:19 +0000 (17:06 +0100)]
r600c: substract vbo offset to vbo size
To avoid kernel complaining and to set proper boundary on
vbo, substract the vbo offset to the vbo size.
Andre Maasikas [Fri, 12 Feb 2010 14:20:50 +0000 (16:20 +0200)]
r600: be more exact in vb size calculation
to make kernel cs checker happier, last attribs don't need full stride of
space. Calculate as count-1*stride + size of attrib
Andre Maasikas [Fri, 12 Feb 2010 14:16:56 +0000 (16:16 +0200)]
r600: support GL_SHORT attributes
normalized seems to work with this setting.
Michal Krol [Fri, 12 Feb 2010 13:45:51 +0000 (14:45 +0100)]
Revert "util: Fix descriptors for R32_FLOAT and R32G32_FLOAT formats."
This reverts commit
aa0b671422880b99dc178d43d1e4e1a3f766bf7f.
Michal Krol [Fri, 12 Feb 2010 12:32:35 +0000 (13:32 +0100)]
util: Fix descriptors for R32_FLOAT and R32G32_FLOAT formats.
Fill in ones for missing components.
Patrice Mandin [Fri, 12 Feb 2010 12:23:48 +0000 (13:23 +0100)]
nouveau: nv30: Define number of texture samplers
Signed-off-by: Patrice Mandin <patmandin@gmail.com>
José Fonseca [Fri, 12 Feb 2010 11:00:22 +0000 (11:00 +0000)]
wgl: Do not reach out and destroy contexts on cleanup.
Simply skip cleanup when contexts are still active.
This addresses two issues:
- in some situations the ICD DLL may be unloaded before the DLL that is
using GL contexts is, so we may receive GL calls after stw_cleanup.
- when aborting (exception, or control-c) the contexts may have been left
in an inconsistent state and attempting to destroy can cause
unpredictable results.
José Fonseca [Thu, 11 Feb 2010 16:34:43 +0000 (16:34 +0000)]
gdi: Never fail at DLL load time.
Windows doesn't really expect things to fail at this point -- it
will try many times until it finally gives up, worse, something bad
happens.
The WGL state tracker will gracefully decline to do work even when it
is loaded.
José Fonseca [Thu, 11 Feb 2010 16:32:23 +0000 (16:32 +0000)]
wgl: Fail gracefully whenever 3D could not be enabled for some reason.
José Fonseca [Thu, 11 Feb 2010 14:09:53 +0000 (14:09 +0000)]
svgadump: Several improvements to shader dumping.
Vinson Lee [Fri, 12 Feb 2010 08:41:01 +0000 (00:41 -0800)]
glu/sgi: Initialize member of class primStream.
Vinson Lee [Fri, 12 Feb 2010 08:36:00 +0000 (00:36 -0800)]
glu/sgi: Initialize members of class Arc.
Vinson Lee [Fri, 12 Feb 2010 08:28:44 +0000 (00:28 -0800)]
glu/sgi: Initialize member of class Bin.
Ben Skeggs [Fri, 12 Feb 2010 05:34:33 +0000 (15:34 +1000)]
nouveau: use AVAIL_RING()
Ben Skeggs [Fri, 12 Feb 2010 02:50:24 +0000 (12:50 +1000)]
nouveau: switch to nouveau_bo_new_tile()
Brian Paul [Thu, 11 Feb 2010 23:54:48 +0000 (16:54 -0700)]
svga: check min_index, max_index before assertions
It's possible for min_index and max_index to be ~0 if the min/max
values were not actually set or computed in the state tracker.
Skip some assertions in that case.
This only effects the debug build error checking.
(cherry picked from commit
408f32dc165b36eb5ec0e2bbc7dcecd28f3a62d3)
Karl Schultz [Thu, 11 Feb 2010 23:49:41 +0000 (16:49 -0700)]
main: Fix compiler warning. No need to convert/store depthScale as GLfloat just to convert it back to GLuint to call unpack_depth_span.
Also removes a difference between the 24/8 and 8/24 texstore routines.
Roland Scheidegger [Thu, 11 Feb 2010 23:43:38 +0000 (00:43 +0100)]
gallium: make max_anisotropy a unsigned bitfield member
saves us a dword in sampler state, hw can't do non-integer aniso degree anyway.
To allow aniso 1x (which seems of dubious value but some hardware (radeons)
have such a mode, and even d3d allows specifiying it) redefine anisotropic
filtering as disabled only if max_anistropy is 0.
Karl Schultz [Thu, 11 Feb 2010 22:52:22 +0000 (15:52 -0700)]
Turn off 64-bit portability problems detection.
Roland Scheidegger [Thu, 11 Feb 2010 22:11:03 +0000 (23:11 +0100)]
nouveau: fix compile errors...
Kristian Høgsberg [Thu, 11 Feb 2010 22:04:06 +0000 (17:04 -0500)]
glapi: Avoid #including gl.h
It's only used for a couple of integer types and and might conflict
with other client API header files.
Kristian Høgsberg [Thu, 11 Feb 2010 21:50:46 +0000 (16:50 -0500)]
i915: Remove always-true irq_active struct intel_screen field
Kristian Høgsberg [Thu, 11 Feb 2010 21:47:46 +0000 (16:47 -0500)]
i915: Remove a few DRI1 era struct intel_screen fields
Kristian Høgsberg [Thu, 11 Feb 2010 21:44:21 +0000 (16:44 -0500)]
i915: Drop intelScreenPrivate typedef and just call it struct intel_screen
Kristian Høgsberg [Thu, 11 Feb 2010 21:38:23 +0000 (16:38 -0500)]
i915: Remove unused intelRegion struct
Phew, sure is nice to only have one struct called intel region.
Roland Scheidegger [Thu, 11 Feb 2010 21:48:39 +0000 (22:48 +0100)]
nv50: adapt to stencil ref changes
Roland Scheidegger [Thu, 11 Feb 2010 21:48:31 +0000 (22:48 +0100)]
nv40: adapt to stencil ref changes
Kristian Høgsberg [Thu, 11 Feb 2010 21:34:11 +0000 (16:34 -0500)]
i915: Remove left-over drmUnmap()
Roland Scheidegger [Thu, 11 Feb 2010 19:55:33 +0000 (20:55 +0100)]
nv30: adapt to stencil ref changes
not sure if this has any hope of working
Brian Paul [Thu, 11 Feb 2010 18:34:28 +0000 (11:34 -0700)]
mesa: additional constant/limit assertions
Brian Paul [Thu, 11 Feb 2010 18:33:17 +0000 (11:33 -0700)]
svga: implement PIPE_CAP_MAX_COMBINED_SAMPLERS query
Roland Scheidegger [Thu, 11 Feb 2010 18:01:10 +0000 (19:01 +0100)]
gallium: make u_blitter code save/restore stencil ref values
Karl Schultz [Thu, 11 Feb 2010 17:54:18 +0000 (10:54 -0700)]
Add git tips
Roland Scheidegger [Thu, 11 Feb 2010 17:48:57 +0000 (18:48 +0100)]
r300g: adapt to stencil ref changes
Roland Scheidegger [Thu, 11 Feb 2010 17:18:11 +0000 (18:18 +0100)]
cell: adapt to stencil ref changes
not betting this compiles, some of the code handling stencil test seems
incomplete and/or never called?
Roland Scheidegger [Thu, 11 Feb 2010 17:16:52 +0000 (18:16 +0100)]
i965g: adapt to stencil ref changes
while there also potentially fix logic ops (were never enabled before)
Roland Scheidegger [Thu, 11 Feb 2010 17:16:09 +0000 (18:16 +0100)]
i915g: adapt to stencil ref changes
Roland Scheidegger [Thu, 11 Feb 2010 17:15:10 +0000 (18:15 +0100)]
gallium: give pipe_stencil_ref its own cso_save/restore functions
seems cleaner, and other dynamic state like viewport is handled that way too
Brian Paul [Thu, 11 Feb 2010 16:40:49 +0000 (09:40 -0700)]
st/mesa: fix gallium texture level selection for RTT
Fixes invalid texture level when rendering to a texture where
GL_BASE_LEVEL > 0.
This will be cherry-picked to the 7.7 branch after additional
testing.
Brian Paul [Thu, 11 Feb 2010 16:04:03 +0000 (09:04 -0700)]
mesa: remove darwin-x86ppc and dangling darwin refs in Makefile
Brian Paul [Thu, 11 Feb 2010 16:02:53 +0000 (09:02 -0700)]
docs: remove old CVS references
José Fonseca [Thu, 11 Feb 2010 12:14:01 +0000 (12:14 +0000)]
Revert "llvmpipe: Move tile coordinate assertion into a more useful place."
This reverts commit
8c0391726a89414356c705fcc2d5a38d12d0fe2b.
Too pendantic.
José Fonseca [Thu, 11 Feb 2010 11:51:32 +0000 (11:51 +0000)]
scons: Fix drm/intel build.
José Fonseca [Thu, 11 Feb 2010 11:08:30 +0000 (11:08 +0000)]
llvmpipe: Handle TGSI_TOKEN_TYPE_PROPERTY.
Avoids assertion failures with certain shaders.
José Fonseca [Thu, 11 Feb 2010 10:59:41 +0000 (10:59 +0000)]
llvmpipe: Move tile coordinate assertion into a more useful place.
José Fonseca [Thu, 11 Feb 2010 10:58:46 +0000 (10:58 +0000)]
llvmpipe: Clamp both extremes of triangles to framebuffer size.
Fix segmentation fault when triangles crossed the axis.
José Fonseca [Thu, 11 Feb 2010 10:08:10 +0000 (10:08 +0000)]
gallivm: TGSI_OPCODE_CONT is not deprecated.
Note that with FIXME instead of an assertion failure.
Addresses fdo 25956.
Ben Skeggs [Thu, 11 Feb 2010 05:21:31 +0000 (15:21 +1000)]
nouveau: remove u_simple_screen.h includes
Roland Scheidegger [Thu, 11 Feb 2010 03:06:01 +0000 (04:06 +0100)]
gallium: also save/restore stencil_ref in cso_save/restore_depth_stencil_alpha
makes life of state trackers easier
Vinson Lee [Thu, 11 Feb 2010 03:01:05 +0000 (19:01 -0800)]
i965g: ifdef out unused functions.
Corbin Simpson [Thu, 11 Feb 2010 02:38:53 +0000 (18:38 -0800)]
r300g: Work around "defect" in r300compiler.
r300compiler doesn't handle half swizzles for vert shaders, which don't
have them. So, for now, disable them.
Vinson Lee [Thu, 11 Feb 2010 02:37:46 +0000 (18:37 -0800)]
st/mesa: Silence uninitialized variable warning in st_cb_bitmap.c.
Vinson Lee [Thu, 11 Feb 2010 02:30:46 +0000 (18:30 -0800)]
swrast: Silence uninitialized variable warnings in ss_tritmp.h.
Vinson Lee [Thu, 11 Feb 2010 02:12:10 +0000 (18:12 -0800)]
os: Do not use Pthreads barrier functions on Mac OS X.
Pthreads barrier functions are not available on some POSIX platforms.
Vinson Lee [Thu, 11 Feb 2010 00:14:28 +0000 (16:14 -0800)]
ffb: Fix build.
Brian Paul [Wed, 10 Feb 2010 23:44:05 +0000 (16:44 -0700)]
gallium: cast to silence waring
Brian Paul [Wed, 10 Feb 2010 23:43:50 +0000 (16:43 -0700)]
gallium: use os_time.h
Brian Paul [Wed, 10 Feb 2010 23:41:41 +0000 (16:41 -0700)]
gallium: use os_time.h in pb_bufmgr_cache.c
Untested, but seems straightforward.
Brian Paul [Wed, 10 Feb 2010 23:37:18 +0000 (16:37 -0700)]
gallium: use os_time.h in u_timed_winsys.c
Brian Paul [Wed, 10 Feb 2010 23:34:24 +0000 (16:34 -0700)]
scons: add another prefix for mingw32 on Fedora
(cherry picked from commit
f1afb352daf0e74751c99254592eef863c64392a)
Ian Romanick [Wed, 10 Feb 2010 23:16:24 +0000 (15:16 -0800)]
dri: Fix one last copy-and-paste brain damage from a previous commit
José Fonseca [Tue, 9 Feb 2010 23:16:26 +0000 (23:16 +0000)]
scons: User friendly message for code generated files
Karl Schultz [Wed, 10 Feb 2010 22:22:07 +0000 (15:22 -0700)]
Merge branch 'master' of git+ssh://git.freedesktop.org/git/mesa/mesa
Karl Schultz [Wed, 10 Feb 2010 22:21:06 +0000 (15:21 -0700)]
Update Visual Studio Project files to compile the shader compiler and build the prebuilt shader header files.
Corbin Simpson [Wed, 10 Feb 2010 09:39:21 +0000 (01:39 -0800)]
r300g: Fix typo.
Thanks to Uros Bizjak for catching it.
Corbin Simpson [Wed, 10 Feb 2010 20:31:18 +0000 (12:31 -0800)]
r300/compiler: r500-fs: Properly set HW register swizzles.
Fixes fallout from
9a1bf52c.
José Fonseca [Wed, 10 Feb 2010 20:23:10 +0000 (20:23 +0000)]
scons: Only build progs when explicitly requested in the command line.
Roland Scheidegger [Wed, 10 Feb 2010 20:21:30 +0000 (21:21 +0100)]
svga: adapt to stencil ref changes
also add some disabled code to handle blendcolor?
Roland Scheidegger [Wed, 10 Feb 2010 20:20:24 +0000 (21:20 +0100)]
llvmpipe: adapt to stencil ref changes
since the driver doesn't actually handle stencil yet the code won't do much.
Roland Scheidegger [Wed, 10 Feb 2010 20:18:12 +0000 (21:18 +0100)]
identity: adapt to stencil ref changes
Roland Scheidegger [Wed, 10 Feb 2010 20:17:46 +0000 (21:17 +0100)]
fo: adapt to stencil ref changes
Roland Scheidegger [Wed, 10 Feb 2010 20:14:18 +0000 (21:14 +0100)]
st/mesa: more fixes for stencil ref change.
compiles. Might need to do something to make it possible to save/restore
stencil ref.
José Fonseca [Wed, 10 Feb 2010 19:59:53 +0000 (19:59 +0000)]
llvmpipe: Map rendertargets when bound as textures.
Fixes gearbox segfault (fdo 25960). Corruption remains.
Ian Romanick [Wed, 10 Feb 2010 19:43:02 +0000 (11:43 -0800)]
dri: Fix copy-and-paste brain damage in previous commit
A number of places in
3cce4a1e10361458630511543b7a8a6438544775 use
TRUE instead of GL_TRUE. This causes build failures in all of the
drivers that I don't typically build. Win.
Reported by sungami on IRC.
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Ian Romanick [Mon, 8 Feb 2010 21:11:58 +0000 (13:11 -0800)]
intel: Don't expose GLX_SWAP_EXCHANGE_OML
We can't always guarantee that the swap will happen by exchange, so we
can't expose this mode. GLX_SWAP_UNDEFINED_OML already covers the
case where the swap *might be* by exchange.
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Ian Romanick [Mon, 8 Feb 2010 19:21:29 +0000 (11:21 -0800)]
intel: Expose a minimal number of configs with accumulation buffer
Expose one config per color depth that includes accumulation buffer.
We could probably expose only one config with accumulation buffer, but
that would require figuring out the actual color depth. This is
easier and only exposes 2 useless configs.
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Ian Romanick [Mon, 8 Feb 2010 19:09:15 +0000 (11:09 -0800)]
dri: Allow selective generation of accum. buffer configs
Modify the interface to driCreateConfigs allowing drivers to not
expose configs with an accumuation buffer. All of the drivers calling
function have been updated to pass true for the accumulation
selector. This maintains the current behavior.
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Reviewed-by: Corbin Simpson <MostAwesomeDude@gmail.com>
Ian Romanick [Mon, 8 Feb 2010 18:34:52 +0000 (10:34 -0800)]
intel: Stop exposing useless 24 depth/0 stencil configs
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>