Ian Romanick [Tue, 20 Sep 2011 23:19:04 +0000 (16:19 -0700)]
mesa/es: Remove redundant logic op operand validation
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Brian Paul <brianp@vmware.com>
Ian Romanick [Tue, 20 Sep 2011 23:17:16 +0000 (16:17 -0700)]
mesa/es: Remove redundant alpha function validation
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Brian Paul <brianp@vmware.com>
Ian Romanick [Tue, 20 Sep 2011 23:15:22 +0000 (16:15 -0700)]
mesa/es: Remove redundant separate stencil mask face validation
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Brian Paul <brianp@vmware.com>
Ian Romanick [Tue, 20 Sep 2011 23:13:14 +0000 (16:13 -0700)]
mesa/es: Remove redundant front-face mode validation
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Brian Paul <brianp@vmware.com>
Ian Romanick [Tue, 20 Sep 2011 23:09:55 +0000 (16:09 -0700)]
mesa/es: Remove redundant face culling mode validation
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Brian Paul <brianp@vmware.com>
Ian Romanick [Tue, 20 Sep 2011 23:08:35 +0000 (16:08 -0700)]
mesa/es: Remove redundant blend equation mode validation
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Brian Paul <brianp@vmware.com>
Ian Romanick [Tue, 20 Sep 2011 23:07:04 +0000 (16:07 -0700)]
mesa/es: Remove redundant texture target validation
Mesa doesn't check the parameter passed to glMultiTexCoord*. It does,
however, mask the texture value to prevent out-of-bounds writes. This
patch will promote this non-conformant behavior to OpenGL ES 1. I don't
think anyone will care, and the gets some silly code out of a hot path.
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Brian Paul <brianp@vmware.com>
Ian Romanick [Mon, 3 Oct 2011 20:09:51 +0000 (13:09 -0700)]
mesa/es: Rearrange placement of GL_TEXTURE_MAX_ANISOTROPY_EXT in APIspec
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Brian Paul <brianp@vmware.com>
Ian Romanick [Mon, 3 Oct 2011 20:05:47 +0000 (13:05 -0700)]
mesa/es: Remove redundant min/mag filter validation
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Brian Paul <brianp@vmware.com>
Mathias Fröhlich [Sat, 18 Aug 2012 09:43:51 +0000 (11:43 +0200)]
radeon-llvm: Start multithreaded before using llvm.
This is required to make some of llvm's api calls
thread save. In particular the PassRegistry, which is
implicitly accessed while compiling shader programs.
The PassRegistry uses a mutex that is only active if
the llvm_is_multithreaded() returns true.
Calling llvm_start_multithreading() makes this happen
and by calling this function we try to make sure that
we can savely compile shaders in paralell.
Since there is also a call llvm_stop_multithreading()
in the llvm api, we cannot guarantee that this does
not get switched off while we are relying on this being
set, but for the easier use cases this fixes a race with
the radeon llvm compiler we have as of today.
Signed-off-by: Mathias Froehlich <Mathias.Froehlich@web.de>
Signed-off-by: Tom Stellard <thomas.stellard@amd.com>
archibald [Mon, 20 Aug 2012 14:44:39 +0000 (14:44 +0000)]
r600g: Move common compute/3D register init to its own function
Signed-off-by: Tom Stellard <thomas.stellard@amd.com>
Christoph Bumiller [Sat, 18 Aug 2012 15:34:44 +0000 (17:34 +0200)]
nv50/ir/tgsi: handle DP2 in tgsi Instruction srcMask
Solved by Tiziano Bacocco on IRC.
Christoph Bumiller [Sat, 18 Aug 2012 15:30:50 +0000 (17:30 +0200)]
nv50/ir/emit: don't forget saturation bit on f32 add immediate
Solved by Maxim Levitsky on IRC.
Tilman Sauerbeck [Sat, 18 Aug 2012 09:51:22 +0000 (11:51 +0200)]
mesa: use #if over #ifdef in the FEATURE_ES1 check to fix a build failure.
mfeatures.h will define FEATURE_ES1 to 0 if it's not defined yet.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=53664
Signed-off-by: Brian Paul <brianp@vmware.com>
Brian Paul [Fri, 17 Aug 2012 14:16:23 +0000 (08:16 -0600)]
st/mesa: fix sampler view counting
In the past, when we called pipe::set_sampler_views(n) the drivers set
samplers [n..MAX] to NULL. We no longer do that. The state tracker
code was already trying to set unused sampler views to NULL to cover
that case, but the logic was broken and unnoticed until now. This patch
fixes it.
Strictly speaking, this patch shouldn't be necessary. Drivers should simply
ignore unused samplers and sampler views. But some drivers like llvmpipe (and
others?) count those things and they figure into state validation. That could
be fixed in the future.
Fixes http://bugs.freedesktop.org/show_bug.cgi?id=53617
Reviewed-by: Marek Olšák <maraeo@gmail.com>
Brian Paul [Fri, 17 Aug 2012 21:20:07 +0000 (15:20 -0600)]
util: update and fix u_upload_mgr.h comments
Brian Paul [Fri, 17 Aug 2012 20:33:31 +0000 (14:33 -0600)]
st/mesa: use Elements() instead of hard-coded number
And add a comment about the velems_util_draw[] array.
Brian Paul [Fri, 17 Aug 2012 16:52:16 +0000 (10:52 -0600)]
mesa: remove unused params, add const qualifiers
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Brian Paul [Fri, 17 Aug 2012 16:50:39 +0000 (10:50 -0600)]
mesa: querying GL_TEXTURE_COMPRESSED_IMAGE_SIZE for a buffer obj is illegal
GL_INVALID_OPERATION is to be raised when querying a non-compressed
image/buffer. Since a buffer object can't have a compressed format this
query always generates an error.
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Ian Romanick [Fri, 27 Jul 2012 21:38:37 +0000 (14:38 -0700)]
mesa/es: Don't generate ES1 type conversion wrappers
These are gradually going to get whittled away and eventually folded into the
source files with the native type functions.
v2: Add (speculative) SConscript changes. These may be broken.
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Eric Anholt [Tue, 14 Aug 2012 01:08:56 +0000 (18:08 -0700)]
i965: Fix bug in the old FS backend's projtex() calculation.
In the old backend, we looked at any FS attribute's proj_attrib_mask bits, not
just texcoords. Now that we have _mesa_vert_result_to_frag_attrib(), we can
fill in the other FS inputs with correct proj_attrib_mask info.
NOTE: This is a candidate for stable branches.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=46644
Signed-off-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Kenneth Graunke [Mon, 11 Jun 2012 02:33:01 +0000 (19:33 -0700)]
mesa: Support GL_TEXTURE_BUFFER in GetTexLevelParameter[if]v in GL 3.1+.
The OpenGL 3.1 specification explicitly allows this. Oddly, the
ARB_texture_buffer_object spec's issues section claims this isn't
allowed, but proceeds to explain that the extension simply doesn't edit
the underlying spec to allow it, and thus it didn't appear in the list
of legal texture targets.
Thus, this patch legalizes it only in 3.1+ contexts, but still returns
INVALID_ENUM in earlier contexts that expose ARB_texture_buffer_object.
Unfortunately, the behavior of the call is horrendously undefined.
Fixes oglconform's tbo/negative.textureParams test.
v2: Require desktop OpenGL.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Signed-off-by: Eric Anholt <eric@anholt.net>
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Kenneth Graunke [Mon, 11 Jun 2012 08:27:38 +0000 (01:27 -0700)]
mesa: Split out part of glGetTexLevelParameter into a helper function.
Move the _mesa_GetTexLevelParameter[iv] functions below the helper
function so the prototype is available.
This will be useful in the next commit.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Kenneth Graunke [Mon, 11 Jun 2012 02:11:55 +0000 (19:11 -0700)]
mesa: Add GL_TEXTURE_CUBE_MAP to _mesa_max_texture_levels(). [v2]
For cube maps, _mesa_generate_mipmap() calls this with
GL_TEXTURE_CUBE_MAP (the gl_texture_object's Target) rather than one
of the faces. This caused _mesa_max_texture_levels() to return 0, which
resulted in maxLevels == -1 and the next line's assertion to fail.
This function is called from seven places:
- fbobject.c: framebuffer_texture()
- mipmap.c: _mesa_generate_mipmap()
- texgetimage.c:
- getteximage_error_check()
- getcompressedteximage_error_check()
- texparam.c: _mesa_GetTexLevelParameteriv()
- texstorage.c: tex_storage_error_check()
All of these (or their callers) now explicitly check for invalid targets
already, so this shouldn't cause invalid targets to slip through.
(Technically _mesa_generate_mipmap() doesn't check for invalid targets,
but the API-facing _mesa_GenerateMipmapEXT() function does.)
+2 oglconforms (float-texture/mipmap.automatic and mipmap.manual)
In addition to fixing the mipmap bug, it should also cause glTexStorage
to accept GL_TEXTURE_CUBE_MAP, which is explicitly allowed by the spec.
v2: Drop alterations to callers; this is now in a patch series that adds
explicit checking to API functions.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Kenneth Graunke [Mon, 11 Jun 2012 01:41:58 +0000 (18:41 -0700)]
mesa: Add explicit target checking to GetTexLevelParameter[if]v().
Previously, it relied on _mesa_max_texture_levels() for texture target
error checking. This was somewhat dodgy, as _mesa_max_texture_levels()
is called in seven diferent places, not all of which necessarily accept
the same list of targets.
I copied the list of legal targets from _mesa_max_texture_levels(), so
this patch should not introduce any change in behavior. Future patches
will cause the two to diverge.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Kenneth Graunke [Mon, 11 Jun 2012 02:06:23 +0000 (19:06 -0700)]
mesa: Add explicit target checking to Get[Compressed]TexImage().
Previously, they relied on _mesa_max_texture_levels() for texture target
error checking. This was somewhat dodgy, as _mesa_max_texture_levels()
is called in seven diferent places, not all of which necessarily accept
the same list of targets.
I copied the list of legal targets from _mesa_max_texture_levels() but
removed the proxy targets, as both functions explicitly rejected those
targets. This changes the order in which we check errors, which could
change whether we return INVALID_VALUE or INVALID_ENUM. However, it
shouldn't change the list of accepted targets.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Brian Paul [Fri, 17 Aug 2012 14:58:15 +0000 (08:58 -0600)]
llvmpipe: remove polygon stipple assertion
It's possible for us to have an unused sampler bound when the fragment
shader itself doesn't use any samplers. So the assertion isn't valid.
Fixes http://bugs.freedesktop.org/show_bug.cgi?id=53616
Brian Paul [Thu, 16 Aug 2012 21:27:39 +0000 (15:27 -0600)]
svga: minor code reformatting
To be consistent with other functions.
Matt Turner [Tue, 14 Aug 2012 19:36:01 +0000 (12:36 -0700)]
build: Remove -shared from OSMesa's LDFLAGS
Would break the static build.
Matt Turner [Tue, 14 Aug 2012 19:33:32 +0000 (12:33 -0700)]
build: Remove EXTRA_LIB_PATH
You can add extra library paths to LDFLAGS directly.
Matt Turner [Tue, 14 Aug 2012 19:24:31 +0000 (12:24 -0700)]
build: Require X11 pkg-config files
Marek Olšák [Thu, 16 Aug 2012 18:36:40 +0000 (20:36 +0200)]
r600g: disable tiling for 422 formats again
Marek Olšák [Tue, 7 Aug 2012 01:30:43 +0000 (03:30 +0200)]
r600g: fix blits of subsampled formats
Marek Olšák [Thu, 16 Aug 2012 17:05:22 +0000 (19:05 +0200)]
r600g: fix copying between NPOT mipmapped compressed textures
We aligned the dimensions to the blocksize, then divided by it
(in r600_blit.c), then minified, which was wrong.
The minification must be done first, not last.
This fixes piglit/fbo-generatemipmap-formats with S3TC and maybe
a bunch of other tests too. Tested on RV730.
Marek Olšák [Wed, 15 Aug 2012 17:06:04 +0000 (19:06 +0200)]
r600g: make F2U trans-only on r600-r700
This fixes a failing assertion in r600_asm.c.
Marek Olšák [Tue, 14 Aug 2012 20:10:35 +0000 (22:10 +0200)]
r600g: set CB_COLOR_INFO to INVALID for disabled colorbuffers on r600-r700
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Marek Olšák [Tue, 14 Aug 2012 00:29:17 +0000 (02:29 +0200)]
r600g: rename r600_resource_texture to r600_texture
Marek Olšák [Tue, 14 Aug 2012 00:22:13 +0000 (02:22 +0200)]
r600g: always put tiled textures in VRAM
Marek Olšák [Sun, 12 Aug 2012 16:40:33 +0000 (18:40 +0200)]
r600g: cleanup r600_resource_texture in favor of radeon_surface
Marek Olšák [Sun, 12 Aug 2012 16:36:52 +0000 (18:36 +0200)]
r600g: remove unused parameter in r600_texture_create_object
Marek Olšák [Sun, 12 Aug 2012 16:22:42 +0000 (18:22 +0200)]
r600g: fixup the usage flag for the flushed depth texture
Philipp Brüschweiler [Wed, 15 Aug 2012 16:14:44 +0000 (18:14 +0200)]
wayland-drm: close fd after the display is uninitialized
This fixes a "kernel rejected pushbuf: Bad file descriptor" error on
wl_drm display destruction.
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
José Fonseca [Wed, 15 Aug 2012 18:24:58 +0000 (19:24 +0100)]
scons: Fix MinGW cross compilation.
Compensate for the recent changes and assumptions added to
Makefiles.sources
Tom Stellard [Thu, 16 Aug 2012 15:49:31 +0000 (15:49 +0000)]
radeon/llvm: Lower implicit parameters before ISel
Brian Paul [Thu, 16 Aug 2012 15:16:42 +0000 (09:16 -0600)]
gallium/draw: move misplaced brace
Brian Paul [Fri, 10 Aug 2012 16:36:17 +0000 (10:36 -0600)]
mesa: raise GL_INVALID_OPERATION in glGenerateMipmap for missing base image
This seems to be expected by the WebGL texture-mips test. The error makes
sense, but I haven't found (yet) any OpenGL documentation specifying this
error condition.
See http://bugs.freedesktop.org/show_bug.cgi?id=44912
Note: This is a candidate for the 8.0 branch.
Reviewed-by: José Fonseca <jfonseca@vmware.com>
Brian Paul [Fri, 10 Aug 2012 02:59:44 +0000 (20:59 -0600)]
r600: update sampler, sampler_view code for the future
For when we have pipe->set_sampler_states(pipe, shader, start, num, samplers),
etc.
Reviewed-by: Marek Olšák <maraeo@gmail.com>
Brian Paul [Fri, 10 Aug 2012 02:59:44 +0000 (20:59 -0600)]
rbug: update data structures, functions for future changes
To support geom/compute/etc shaders, samplers, sampler views, etc.
To support pipe->bind_sampler_states() w/ start_slot.
Brian Paul [Fri, 10 Aug 2012 02:59:44 +0000 (20:59 -0600)]
gallium/trace: add 'start' parameter to bind_sampler_states/views()
Brian Paul [Fri, 10 Aug 2012 02:59:44 +0000 (20:59 -0600)]
gallium/identity: add 'start' parameter to bind_sampler_states/views()
Brian Paul [Fri, 10 Aug 2012 02:59:44 +0000 (20:59 -0600)]
galahad: add 'start' parameter to bind_sampler_states/views()
Brian Paul [Fri, 10 Aug 2012 02:59:44 +0000 (20:59 -0600)]
svga: add 'start' parameter to bind_sampler_states/views()
Brian Paul [Fri, 10 Aug 2012 02:59:44 +0000 (20:59 -0600)]
llvmpipe: add 'start' parameter to bind_sampler_states/views()
Brian Paul [Fri, 10 Aug 2012 02:59:44 +0000 (20:59 -0600)]
softpipe: add 'start' parameter to bind_sampler_states/views()
To support updating a sub-range of sampler states/views in the future.
Note that we always pass start=0 at this time.
Brian Paul [Fri, 10 Aug 2012 02:59:44 +0000 (20:59 -0600)]
gallium/trace: consolidate sampler, sampler_view code
Brian Paul [Fri, 10 Aug 2012 02:59:44 +0000 (20:59 -0600)]
gallium/identity: consolidate sampler, sampler_view code
This will simplify things when the pipe_context functions are consolidated.
Brian Paul [Fri, 10 Aug 2012 02:59:44 +0000 (20:59 -0600)]
st/mesa: add support for GS textures and samplers
Brian Paul [Fri, 10 Aug 2012 02:59:44 +0000 (20:59 -0600)]
st/mesa: combine vertex/fragment sampler state in arrays
As with other recent changes, put the vertex and fragment sampler state
into arrays indexed by the shader type. This will let us easily add
support for other types of shaders in the future.
Brian Paul [Fri, 10 Aug 2012 02:59:44 +0000 (20:59 -0600)]
gallium: remove PIPE_MAX_VERTEX/GEOMETRY_SAMPLERS #define
PIPE_MAX_SAMPLERS, PIPE_MAX_VERTEX_SAMPLERS and PIPE_MAX_GEOMETRY_SAMPLERS
were all defined to the same value (16).
In various places we're creating arrays such as
sampler_views[PIPE_SHADER_TYPES][PIPE_MAX_SAMPLERS] so we were assuming
the same number of max samplers for all shader stages anyway.
Of course, drivers are still free to advertise different numbers of max
samplers for different shaders.
Brian Paul [Fri, 10 Aug 2012 02:59:44 +0000 (20:59 -0600)]
draw: index samplers and sampler_view state by shader type
So that we can handle GS state and other types of shaders in the future.
Brian Paul [Fri, 10 Aug 2012 02:59:44 +0000 (20:59 -0600)]
draw: move tgsi-related state into a tgsi sub-struct
To better organize things a bit.
Brian Paul [Fri, 10 Aug 2012 02:59:43 +0000 (20:59 -0600)]
gallium: add a shader stage/type param to some draw functions
To prepare for geometry shader texture support in the draw module.
Note: we still only handle the vertex shader case.
Brian Paul [Thu, 16 Aug 2012 15:00:08 +0000 (09:00 -0600)]
st/mesa: silence signed/unsigned comparison warning
Brian Paul [Thu, 9 Aug 2012 16:14:51 +0000 (10:14 -0600)]
svga: move result->key expression after result != NULL check
Brian Paul [Thu, 9 Aug 2012 16:14:51 +0000 (10:14 -0600)]
svga: fix result==NULL logic in emit_fs_consts()
The previous test for result != NULL was kind of bogus since we dereferenced
the pointer earlier in the code. Now, check for result != NULL first, then
get the result->key info.
Also, remove the useless "offset +=" code at the end.
Brian Paul [Thu, 9 Aug 2012 16:14:51 +0000 (10:14 -0600)]
svga: update comment (s/SVGA_NEW_VS_RESULT/SVGA_NEW_VS_PRESCALE/)
Brian Paul [Thu, 9 Aug 2012 16:14:51 +0000 (10:14 -0600)]
svga: rename svga_hw_vs_parameters -> svga_hw_vs_constants
and similarly for svga_hw_fs_parameters
Niels Ole Salscheider [Sun, 12 Aug 2012 16:35:10 +0000 (18:35 +0200)]
st/mesa: index can be negative in the PROGRAM_CONSTANT case
NOTE: This is a candidate for the 8.0 branch.
Reviewed-by: Brian Paul <brianp@vmware.com>
Brian Paul [Thu, 16 Aug 2012 14:55:48 +0000 (08:55 -0600)]
mesa: add cast to silence warning in _mesa_pack_rgba_span_from_ints()
Brian Paul [Thu, 16 Aug 2012 14:53:55 +0000 (08:53 -0600)]
meta: remove unused variable
Michel Dänzer [Thu, 16 Aug 2012 08:37:44 +0000 (10:37 +0200)]
radeonsi: Fix symbol conflicts with r600g.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=50389
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Michel Dänzer [Wed, 15 Aug 2012 15:18:30 +0000 (17:18 +0200)]
radeonsi: Fix memory leaks if returning early from some state functions.
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Michel Dänzer [Wed, 15 Aug 2012 16:22:46 +0000 (18:22 +0200)]
radeonsi: Fix LLVM context leak.
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Michel Dänzer [Wed, 15 Aug 2012 13:59:07 +0000 (15:59 +0200)]
gallium/radeon: Don't assign virtual address space for BO that already has one.
We'd end up re-using the old one and throwing away the new one anyway, but only
after a roundtrip to the kernel.
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Michel Dänzer [Wed, 15 Aug 2012 13:48:34 +0000 (15:48 +0200)]
gallium/radeon: Create hole for waste when allocating from va_offset.
Otherwise, the wasted area could never be used for an allocation again.
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Michel Dänzer [Thu, 26 Apr 2012 10:10:02 +0000 (12:10 +0200)]
gallium/radeon: Fix potential address space loss in radeon_bomgr_force_va().
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Michel Dänzer [Thu, 26 Apr 2012 18:41:17 +0000 (20:41 +0200)]
gallium/radeon: Delete uppermost virtual address space hole if it's at the top.
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Michel Dänzer [Thu, 26 Apr 2012 09:44:11 +0000 (11:44 +0200)]
gallium/radeon: Fix losing holes when allocating virtual address space.
If a hole exactly matches the allocated size plus alignment, we would fail to
preserve the alignment as a hole. This would result in never being able to use
the alignment area for an allocation again.
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Michel Dänzer [Thu, 26 Apr 2012 13:26:01 +0000 (15:26 +0200)]
gallium/radeon: Merge holes when freeing virtual address space.
Otherwise we'll likely end up with an ever increasing amount of ever smaller
holes.
Requires keeping the list ordered wrt offsets.
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Michel Dänzer [Fri, 3 Aug 2012 09:31:55 +0000 (11:31 +0200)]
gallium/radeon: Make va_offset 64 bits wide.
Otherwise we'd wrap around after 32 bits. The kernel currently limits GPU
virtual address space to 4GB anyway, but that will probably change sooner or
later, and this would result in confusing error messages when running out of
virtual address space even now.
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Vinson Lee [Sun, 12 Aug 2012 07:20:21 +0000 (00:20 -0700)]
llvmpipe: Silence Coverity incorrect sizeof expression defect.
Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Reviewed-by: José Fonseca <jfonseca@vmware.com>
Vinson Lee [Sun, 12 Aug 2012 07:02:32 +0000 (00:02 -0700)]
scons: Add option to enable floating-point textures.
Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Reviewed-by: José Fonseca <jfonseca@vmware.com>
Dave Airlie [Fri, 12 Mar 2010 06:09:23 +0000 (16:09 +1000)]
glx/dri2: add dri2 prime support.
This adds support for having libGL pick a different driver for prime support.
DRI_PRIME env var is set to the value retrieved from the server randr
provider calls, by the calling process. (generally DRI_PRIME=1 will be
the right answer).
Signed-off-by: Dave Airlie <airlied@redhat.com>
Vincent Lejeune [Wed, 1 Aug 2012 20:49:44 +0000 (22:49 +0200)]
radeon/llvm: Enable if-cvt
Signed-off-by: Tom Stellard <thomas.stellard@amd.com>
Vincent Lejeune [Wed, 1 Aug 2012 20:49:43 +0000 (22:49 +0200)]
radeon/llvm: Add callbacks needed by if-cvt
Signed-off-by: Tom Stellard <thomas.stellard@amd.com>
Vincent Lejeune [Wed, 1 Aug 2012 20:49:42 +0000 (22:49 +0200)]
radeon/llvm: Lower branch/branch_cond into predicated jump
Signed-off-by: Tom Stellard <thomas.stellard@amd.com>
Vincent Lejeune [Wed, 1 Aug 2012 20:49:41 +0000 (22:49 +0200)]
radeon/llvm: Add a predicated JUMP instruction
Signed-off-by: Tom Stellard <thomas.stellard@amd.com>
Vincent Lejeune [Wed, 1 Aug 2012 20:49:40 +0000 (22:49 +0200)]
radeon/llvm: Support for predicate bit
Tom Stellard:
- A few changes to predicate register defs
Signed-off-by: Tom Stellard <thomas.stellard@amd.com>
Vincent Lejeune [Fri, 3 Aug 2012 22:15:51 +0000 (00:15 +0200)]
r600g: Glue to handle predicate aware output from llvm
Signed-off-by: Tom Stellard <thomas.stellard@amd.com>
Vincent Lejeune [Wed, 1 Aug 2012 20:52:54 +0000 (22:52 +0200)]
r600g: Fix instruction group merge when there are predicated insts.
Signed-off-by: Tom Stellard <thomas.stellard@amd.com>
Vincent Lejeune [Wed, 1 Aug 2012 20:52:53 +0000 (22:52 +0200)]
radeon/llvm: Do not use PV/PS if PRED_SEL does not match
Signed-off-by: Tom Stellard <thomas.stellard@amd.com>
Vincent Lejeune [Wed, 1 Aug 2012 20:52:52 +0000 (22:52 +0200)]
r600g: Add support for predicates
Signed-off-by: Tom Stellard <thomas.stellard@amd.com>
Christian König [Sun, 12 Aug 2012 17:42:38 +0000 (19:42 +0200)]
radeonsi: move ps sampler state into PM4 stream
Signed-off-by: Christian König <deathsimple@vodafone.de>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Sun, 12 Aug 2012 17:32:51 +0000 (19:32 +0200)]
radeonsi: move ps sampler views into PM4 stream
Signed-off-by: Christian König <deathsimple@vodafone.de>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Sun, 12 Aug 2012 17:26:24 +0000 (19:26 +0200)]
radeonsi: move vertex state descriptors into PM4 stream
Signed-off-by: Christian König <deathsimple@vodafone.de>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Sun, 12 Aug 2012 17:16:48 +0000 (19:16 +0200)]
radeonsi: add shader data infrastructure
With this we can embed data for the shaders (like resource
descriptors) into the PM4 stream.
Signed-off-by: Christian König <deathsimple@vodafone.de>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Sun, 12 Aug 2012 18:06:21 +0000 (20:06 +0200)]
radeon/llvm: add support to fetch temps as vectors
Necessary for texture fetches with temp regs as source on SI.
Signed-off-by: Christian König <deathsimple@vodafone.de>
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Tom Stellard [Thu, 26 Jul 2012 19:20:14 +0000 (19:20 +0000)]
radeon/llvm: Remove AMDGPUUtil.cpp
Apostolos Bartziokas [Thu, 26 Jul 2012 19:00:20 +0000 (15:00 -0400)]
radeon/llvm: Cleanup AMDGPUUtil.cpp
Tom Stellard [Thu, 26 Jul 2012 14:30:23 +0000 (14:30 +0000)]
radeon/llvm: Lower loads from USE_SGPR adddress space during DAG lowering