OSDN Git Service
Mathias Fröhlich [Sun, 21 Sep 2014 16:09:22 +0000 (18:09 +0200)]
mesa: Implement ARB_clip_control.
Implement the mesa parts of ARB_clip_control.
So far no driver enables this.
v3:
Restrict getting clip control state to the availability
of ARB_clip_control.
Move to transformation state.
Handle clip control state with the GL_TRANSFORM_BIT.
Move _FrontBit update into state.c.
Reviewed-by: Brian Paul <brianp@vmware.com>
Signed-off-by: Mathias Froehlich <Mathias.Froehlich@web.de>
Mathias Fröhlich [Sun, 21 Sep 2014 16:09:21 +0000 (18:09 +0200)]
mesa: Refactor viewport transform computation.
This is for preparation of ARB_clip_control.
v3:
Add comments.
Reviewed-by: Brian Paul <brianp@vmware.com>
Signed-off-by: Mathias Froehlich <Mathias.Froehlich@web.de>
Eric Anholt [Fri, 24 Oct 2014 16:16:59 +0000 (17:16 +0100)]
vc4: Reuse uniform_data/contents indices when making uniforms.
This allows vc4_opt_cse.c to CSE-away operations involving the same
uniform values.
total instructions in shared programs: 37341 -> 36906 (-1.16%)
instructions in affected programs: 10233 -> 9798 (-4.25%)
total uniforms in shared programs: 10523 -> 10320 (-1.93%)
uniforms in affected programs: 2467 -> 2264 (-8.23%)
Eric Anholt [Fri, 24 Oct 2014 15:50:37 +0000 (16:50 +0100)]
vc4: When asked to discard-map a whole resource, discard it.
This saves a bunch of extra flushes when texsubimaging a whole texture
that's been used for rendering, or subdataing a whole BO. In particular,
this massively reduces the runtime of piglit texture-packed-formats (when
the probes have been moved out of the inner loop).
Eric Anholt [Fri, 24 Oct 2014 15:45:04 +0000 (16:45 +0100)]
vc4: Refactor flushing before mapping a BO.
I'm going to want to make some other decisions here before flushing.
Eric Anholt [Fri, 24 Oct 2014 14:03:04 +0000 (15:03 +0100)]
vc4: Allow dead code elimination of unused varyings.
total instructions in shared programs: 39022 -> 37341 (-4.31%)
instructions in affected programs: 26979 -> 25298 (-6.23%)
total uniforms in shared programs: 11242 -> 10523 (-6.40%)
uniforms in affected programs: 5836 -> 5117 (-12.32%)
Eric Anholt [Wed, 22 Oct 2014 17:02:18 +0000 (18:02 +0100)]
vc4: Add debug output to match shaderdb info to program dumps.
I'm going to be using VC4_DEBUG=shaderdb,norast to do shaderdb stats, but
when debugging regressions, I want to match shaderdb output to shader
disassembly.
Andreas Boll [Thu, 23 Oct 2014 12:52:55 +0000 (14:52 +0200)]
radeon: enable Hyper-Z on r600g and radeonsi by default
This reverts commit
01e637114914453451becc0dc8afe60faff48d84.
Since then many Hyper-Z issues have been fixed or worked around.
Enable Hyper-Z by default so that we get enough feedback for the upcoming
mesa 10.4 release.
If you have issues with Hyper-Z try to disable Hyper-Z using the enviroment
variable R600_DEBUG=nohyperz and please report the issue on the bugtracker.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=75011
See also: https://bugs.freedesktop.org/show_bug.cgi?id=75112
Signed-off-by: Andreas Boll <andreas.boll.dev@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Matt Turner [Thu, 23 Oct 2014 22:45:35 +0000 (15:45 -0700)]
i965: Silence unused variable warning.
Matt Turner [Thu, 23 Oct 2014 22:45:15 +0000 (15:45 -0700)]
i965/fs: Silence uninitialized variable warning.
The compiler isn't privy to the knowledge that we're doing at least one
framebuffer write.
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
Matt Turner [Tue, 30 Sep 2014 23:24:39 +0000 (16:24 -0700)]
util: Add assume() macro.
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Jan Vesely [Wed, 13 Aug 2014 20:47:28 +0000 (16:47 -0400)]
glapi: Fix compiler warning and script name
Signed-off-by: Jan Vesely <jan.vesely@rutgers.edu>
Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
Rob Clark [Thu, 23 Oct 2014 12:50:50 +0000 (08:50 -0400)]
Revert "freedreno/a3xx: only emit dirty consts"
This reverts commit
94bb33617d1e8978dc52b8aaa4eb41bfb6703f79.
Which somehow broke gnome-shell.. and needs more investigation. For
now, revert..
Rob Clark [Wed, 22 Oct 2014 20:36:24 +0000 (16:36 -0400)]
freedreno: fix PIPE_TRANSFER_DISCARD_WHOLE_RESOURCE
fd_bo_cpu_prep() doesn't realize the bo is already referenced in
unflushed cmdstream. It could be made to do so (but would have to be
implemented twice, ie. both for msm and kgsl). But we still can't do
the expected thing if the caller isn't using _NOSYNC. Because of the
way the tiling works, we need to build quite a bit of cmdstream at flush
time, which is not possible to do at the libdrm level.
So rather than trying to make fd_bo_cpu_prep() smarter than it can
possibly be, just *always* discard and reallocate if the
PIPE_TRANSFER_DISCARD_WHOLE_RESOURCE flag is set.
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Jan Vesely [Tue, 21 Oct 2014 15:59:34 +0000 (11:59 -0400)]
clover: Require libelf
v2: test for libelf once, check in both radeon and clover
CC: Tom Stellard <tom@stellard.net>
CC: Emil Velikov <emil.l.velikov@gmail.com>
CC: Francisco Jerez <currojerez@riseup.net>
Signed-off-by: Jan Vesely <jan.vesely@rutgers.edu>
Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
Emil Velikov [Mon, 20 Oct 2014 08:53:39 +0000 (09:53 +0100)]
clover: use correct typenames for compat::pair's first/second
Seems to be a typo judging from the overall declaration of the
template.
Cc: EdB <edb+mesa@sigluy.net>
Cc: Francisco Jerez <currojerez@riseup.net>
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
Reviewed-by: Francisco Jerez <currojerez@riseup.net>
Emil Velikov [Sun, 19 Oct 2014 15:16:51 +0000 (16:16 +0100)]
auxiliary/os: get the mmap/munmap wrappers working with android
- Use macro for munmap under Android - the STATIC_ASSERT uses
a off_t which is not used under Android for mmap. As loff_t size
does not vary as does off_t just ignore the assert.
- Wrap the long lines to improve readability.
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
Mauro Rossi [Sun, 19 Oct 2014 15:16:49 +0000 (16:16 +0100)]
gallium/nouveau: fully build the driver under android
Fix the trivial typo in the variable name.
Cc: "10.2 10.3" <mesa-stable@lists.freedesktop.org>
Alon Levy [Tue, 22 Jul 2014 21:07:06 +0000 (00:07 +0300)]
mesa/shaderimage.c: fix inconsistent sign warning
Signed-off-by: Alon Levy <alevy@redhat.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Alon Levy [Tue, 22 Jul 2014 21:07:04 +0000 (00:07 +0300)]
wgl: stw_pixelformat_get_info: correct type for index variable
Signed-off-by: Alon Levy <alevy@redhat.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Alon Levy [Tue, 22 Jul 2014 21:07:03 +0000 (00:07 +0300)]
u_math.h: fix 64 to 32 bit truncation warning
Signed-off-by: Alon Levy <alevy@redhat.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
José Fonseca [Thu, 23 Oct 2014 09:42:12 +0000 (10:42 +0100)]
gallivm: Fix build with LLVM 3.3.
The setMCJITMemoryManager method doesn't exist in LLVM 3.3.
I thought I had tested the latest version of my earlier change with LLVM
3.3, but it looks I missed it.
Trivial.
José Fonseca [Wed, 22 Oct 2014 19:08:57 +0000 (20:08 +0100)]
gallivm: Properly update for removal of JITMemoryManager in LLVM 3.6.
JITMemoryManager was removed in LLVM 3.6, and replaced by its base class
RTDyldMemoryManager.
This change fixes our JIT memory managers specializations to derive from
RTDyldMemoryManager in LLVM 3.6 instead of JITMemoryManager.
This enables llvmpipe to run with LLVM 3.6.
However, lp_free_generated_code is basically a no-op because there are
not enough hook points in RTDyldMemoryManager to track and free the code
of a module. In other words, with MCJIT, code once created, stays
forever allocated until process destruction. This is not speicfic to
LLVM 3.6 -- it will happen whenever MCJIT is used regardless of version.
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
José Fonseca [Wed, 22 Oct 2014 14:40:53 +0000 (15:40 +0100)]
gallivm: Fix white-space.
Replace tabs with spaces.
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
José Fonseca [Wed, 22 Oct 2014 12:09:59 +0000 (13:09 +0100)]
gallivm,llvmpipe,clover: Bump required LLVM version to 3.3.
We'll need to update gallivm for the interface changes in LLVM 3.6, and
the fewer the number of older LLVM versions we support the less hairy that
will be.
As consequence HAVE_AVX define can disappear. (Note HAVE_AVX meant
whether LLVM version supports AVX or not. Runtime support for AVX is
always checked and enforced independently.)
Verified llvmpipe builds and runs with with LLVM 3.3, 3.4, and 3.5.
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Ilia Mirkin [Wed, 22 Oct 2014 02:20:50 +0000 (22:20 -0400)]
mesa: remove conditional render and rgtc from ES3 requirements
The functionality exposed by those extensions does not appear in ES3
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Brian Paul [Tue, 21 Oct 2014 16:30:22 +0000 (10:30 -0600)]
u_blitter: put a comment on util_blitter_cache_all_shaders()
Trivial.
Brian Paul [Tue, 21 Oct 2014 16:26:24 +0000 (10:26 -0600)]
u_blitter: use ctx->bind_fs_state(), not pipe->bind_fs_state()
Consistently use the function pointer we saved earlier.
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Brian Paul [Tue, 21 Oct 2014 18:14:06 +0000 (12:14 -0600)]
u_blitter: create basic fs shaders in util_blitter_cache_all_shaders()
We need to create all fs shaders in this function.
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Brian Paul [Tue, 21 Oct 2014 16:22:35 +0000 (10:22 -0600)]
u_blitter: do error checking assertions for shader caching
If the user calls util_blitter_cache_all_shaders() set a flag and assert
that we never try to create any new fragment shaders after that point.
If the assertions fails, it means we missed generating some shader in
util_blitter_cache_all_shaders().
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Anuj Phogat [Mon, 22 Sep 2014 22:10:28 +0000 (15:10 -0700)]
glsl: Use signed array index in update_max_array_access()
Avoids a crash in case of negative array index is used in a
shader program.
Cc: <mesa-stable@lists.freedesktop.org>
Signed-off-by: Anuj Phogat <anuj.phogat@gmail.com>
Reviewed-by: Chris Forbes <chrisf@ijw.co.nz>
Anuj Phogat [Thu, 18 Sep 2014 23:30:31 +0000 (16:30 -0700)]
glsl: Fix crash due to negative array index
Currently Mesa crashes with a shader like this:
[fragmnet shader]
float[5] array;
int idx = -2;
void main()
{
gl_FragColor = vec4(0.0, 1.0, 0.0, array[idx]);
}
Cc: <mesa-stable@lists.freedesktop.org>
Signed-off-by: Anuj Phogat <anuj.phogat@gmail.com>
Reviewed-by: Chris Forbes <chrisf@ijw.co.nz>
Marek Olšák [Wed, 22 Oct 2014 08:59:49 +0000 (10:59 +0200)]
radeonsi: implement pipe_rasterizer_state::clip_halfz
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Marek Olšák [Wed, 22 Oct 2014 08:59:49 +0000 (10:59 +0200)]
r600g: implement pipe_rasterizer_state::clip_halfz
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Marek Olšák [Wed, 22 Oct 2014 08:59:49 +0000 (10:59 +0200)]
r300g: implement pipe_rasterizer_state::clip_halfz
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Michel Dänzer [Tue, 21 Oct 2014 03:40:15 +0000 (12:40 +0900)]
r600g: Drop references to destroyed blend state
Fixes use-after-free when the currently bound blend state is destroyed.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=85267
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=84140
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
Cc: mesa-stable@lists.freedesktop.org
Kenneth Graunke [Thu, 16 Oct 2014 01:57:07 +0000 (18:57 -0700)]
i965/vec4: Generate better code for ir_triop_csel.
Previously, we generated an extra CMP instruction:
cmp.ge.f0(8) g6<1>D g1<0,4,1>F 0F
cmp.nz.f0(8) null g6<4,4,1>D 0D
(+f0) sel(8) g5<1>F g1.4<0,4,1>F g2<0,4,1>F
The first operand is always a boolean, and we want to predicate the SEL
on that. Rather than producing a boolean value and comparing it against
zero, we can just produce a condition code in the flag register.
Now we generate:
cmp.ge.f0(8) null g1<0,4,1>F 0F
(+f0) sel(8) g5<1>F g1.4<0,4,1>F g2<0,4,1>F
No difference in shader-db.
v2: Remember to delete the old code (thanks Matt).
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Kenneth Graunke [Thu, 16 Oct 2014 02:17:21 +0000 (19:17 -0700)]
i965/vec4: Simplify visit(ir_expression *)'s result_src/dst setup.
Using dst_reg(this, ir->type) automatically sets the writemask to the
proper size for the type; src_reg(dst_reg) preserves that. This should
be equivalent, but less code.
Note that src_reg(dst_reg) either uses SWIZZLE_XXXX or SWIZZLE_XYZW, so
the old code did need the manual writemask adjustment, since it
constructed the registers the other way around.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Kenneth Graunke [Thu, 16 Oct 2014 02:13:16 +0000 (19:13 -0700)]
i965/vec4: Delete some dead code in visit(ir_expression *).
Nothing uses the vector_elements temporary variable.
Setting this->result.file is dead because we overwrite this->result a
few lines later.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Kenneth Graunke [Thu, 16 Oct 2014 01:57:07 +0000 (18:57 -0700)]
i965/fs: Generate better code for ir_triop_csel.
Previously, we generated an extra CMP instruction:
cmp.ge.f0(8) g4<1>D g2<0,1,0>F 0F
cmp.nz.f0(8) null g4<8,8,1>D 0D
(+f0) sel(8) g120<1>F g2.4<0,1,0>F g3<0,1,0>F
The first operand is always a boolean, and we want to predicate the SEL
on that. Rather than producing a boolean value and comparing it against
zero, we can just produce a condition code in the flag register.
Now we generate:
cmp.ge.f0(8) null g2<0,1,0>F 0F
(+f0) sel(8) g124<1>F g2.4<0,1,0>F g3<0,1,0>F
total instructions in shared programs:
5473459 ->
5473253 (-0.00%)
instructions in affected programs: 6219 -> 6013 (-3.31%)
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Kenneth Graunke [Thu, 16 Oct 2014 16:28:42 +0000 (09:28 -0700)]
glsl: Delete unused gl_uniform_driver_format enum values.
A while back, Matt made the uniform upload functions simply upload
ctx->Const.UniformBooleanTrue for boolean values instead of 0/1, which
removed the need to convert it later. We also set UniformBooleanTrue to
1.0f for drivers which want to treat booleans as 0.0/1.0f.
Nothing ever sets these, so they are dead.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Rob Clark [Tue, 21 Oct 2014 21:08:10 +0000 (17:08 -0400)]
freedreno/a3xx: fix depth/stencil restore format
Also fix z16 restore format which was completely wrong.
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Rob Clark [Tue, 21 Oct 2014 16:25:28 +0000 (12:25 -0400)]
freedreno/a3xx: fix viewport state during clear
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Rob Clark [Tue, 21 Oct 2014 15:28:53 +0000 (11:28 -0400)]
freedreno: mark scissor state dirty when enable bit changes
We don't have a scissor enable bit in hw, so when a raster state change
results in scissor enable bit changing, we need to also mark scissor
state as dirty.
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Rob Clark [Tue, 21 Oct 2014 14:30:49 +0000 (10:30 -0400)]
freedreno: clear vs scissor
The optimization of avoiding restore (mem2gmem) if there was a clear
falls down a bit if you don't have a fullscreen scissor. We need to
make the decision logic a bit more clever to keep track of *what* was
cleared, so that we can (a) completely skip mem2gmem if entire buffer
was cleared, or (b) skip mem2gmem on a per-tile basis for tiles that
were completely cleared.
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Vinson Lee [Sun, 19 Oct 2014 07:13:33 +0000 (00:13 -0700)]
clover: Fix build error with LLVM 3.4.
DataLayoutPass was added in LLVM 3.5 r202168, commit
57edc9d4ff1648568a5dd7e9958649065b260dca "Make DataLayout a plain
object, not a pass.".
This patch fixes this build error with LLVM 3.4.
CXX llvm/libclllvm_la-invocation.lo
llvm/invocation.cpp: In function 'void {anonymous}::optimize(llvm::Module*, unsigned int, const std::vector<llvm::Function*>&)':
llvm/invocation.cpp:324:18: error: expected type-specifier
PM.add(new llvm::DataLayoutPass(mod));
^
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=85189
Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
Marek Olšák [Tue, 23 Sep 2014 15:17:01 +0000 (17:17 +0200)]
r600g,radeonsi: convert TGSI shader type to LLVM shader type
The values are hardcoded in the LLVM backend, but the TGSI definitions are
going to be changed with tessellation, e.g. TGSI_PROCESSOR_COMPUTE will be
increased by 2.
We'll use VS for LS and HS, because there's nothing special about them
from the LLVM backend point of view, even though the hardware side is
different. We do the same for ES.
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Marek Olšák [Sun, 5 Oct 2014 22:19:31 +0000 (00:19 +0200)]
radeonsi: add some missing register definitions
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Marek Olšák [Sun, 5 Oct 2014 11:33:40 +0000 (13:33 +0200)]
radeonsi: load ring resource descriptors only once
v2: document the new functions
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Marek Olšák [Fri, 26 Sep 2014 21:06:32 +0000 (23:06 +0200)]
radeonsi: clarify shader constant load functions
I'll need indexed loads without the meta data flag for tessellation later.
Also rename load_const to buffer_load_const to distinguish it from indexed
const loads.
v2: add comments
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Marek Olšák [Sun, 5 Oct 2014 10:38:54 +0000 (12:38 +0200)]
radeonsi: statically declare resource and sampler arrays
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Marek Olšák [Thu, 16 Oct 2014 14:20:26 +0000 (16:20 +0200)]
radeonsi: remove conversion of DX9 FACE input to GL
st/mesa and gallium expect the DX9 format, so this is useless.
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Marek Olšák [Tue, 14 Oct 2014 20:51:10 +0000 (22:51 +0200)]
radeonsi: revert hack for random failures in glsl-max-varyings
This reverts commit
032e5548b3d4b5efa52359218725cb8e31b622ad.
I've run glsl-max-varyings 30 times and it always passed.
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Marek Olšák [Tue, 14 Oct 2014 15:48:52 +0000 (17:48 +0200)]
radeonsi: generate shader pm4 states right after shader compilation
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Marek Olšák [Tue, 14 Oct 2014 15:36:30 +0000 (17:36 +0200)]
radeonsi: make pm4 state generation for shaders independent of the context
The si_pm4_delete_state calls became useless, because the pm4 state is
always generated only once.
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Marek Olšák [Tue, 14 Oct 2014 15:31:00 +0000 (17:31 +0200)]
radeonsi: inline si_pm4_alloc_state
It seemed like the function needed a context pointer. Let's remove it
to make it less confusing.
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Marek Olšák [Mon, 20 Oct 2014 13:41:42 +0000 (15:41 +0200)]
r300g: replace r300_get_num_samples with a util variant
Marek Olšák [Mon, 6 Oct 2014 19:12:14 +0000 (21:12 +0200)]
glsl_to_tgsi: use _mesa_copy_linked_program_data
This deduplicates some code.
Marek Olšák [Thu, 16 Oct 2014 14:21:54 +0000 (16:21 +0200)]
glsl_to_tgsi: fix the value of gl_FrontFacing with native integers
We must convert it to boolean from the DX9 float encoding that Gallium
specifies.
Later, we should probably define that FACE should be 0 or ~0 if native
integers are supported.
Cc: 10.2 10.3 <mesa-stable@lists.freedesktop.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
Marek Olšák [Sun, 5 Oct 2014 16:55:47 +0000 (18:55 +0200)]
st/mesa: add ST_DEBUG=wf option which enables wireframe rendering
Useful for tessellation.
Marek Olšák [Wed, 1 Oct 2014 18:28:17 +0000 (20:28 +0200)]
gallium: add PIPE_SHADER_CAP_MAX_OUTPUTS and use it in st/mesa
With 5 shader stages and various combinations of enabled and disabled shaders,
the maximum number of outputs in one shader doesn't have to be equal to
the maximum number of inputs in the following shader.
v2: return 32 for softpipe and llvmpipe
Eric Anholt [Tue, 21 Oct 2014 14:46:48 +0000 (15:46 +0100)]
vc4: Fix SRC_ALPHA_SATURATE blending.
Fixes glean blendFunc.
Eric Anholt [Mon, 20 Oct 2014 20:14:57 +0000 (21:14 +0100)]
vc4: Fix stencil writemask handling.
If the writemask doesn't compress, then we want to put in the uncompressed
writemask, not the compressed writemask failure value (all-on).
Fixes glean's stencil2 and fbo-clear-formats on stencil.
Eric Anholt [Mon, 20 Oct 2014 21:53:07 +0000 (22:53 +0100)]
vc4: Don't look at back stencil state unless two-sided stencil is enabled.
Fixes regressions in the next bugfix, because gallium util stuff leaves
the back stencil state as 0 if !back->enabled.
Rob Clark [Sun, 19 Oct 2014 18:55:32 +0000 (14:55 -0400)]
freedreno/ir3: add debug flag to disable cp
FD_MESA_DEBUG=nocp will disable copy propagation pass.
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Ilia Mirkin [Fri, 3 Oct 2014 20:23:19 +0000 (16:23 -0400)]
freedreno: positions come out as integers, not half-integers
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Rob Clark [Sat, 18 Oct 2014 20:52:44 +0000 (16:52 -0400)]
freedreno/a3xx: disable early-z when we have kill's
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Rob Clark [Sat, 18 Oct 2014 19:28:16 +0000 (15:28 -0400)]
freedreno/ir3: fix potential gpu lockup with kill
It seems like the hardware is unhappy if we execute a kill instruction
prior to last input (ei). Probably the shader thread stops executing
and the end-input flag is never set.
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Rob Clark [Sat, 18 Oct 2014 18:46:35 +0000 (14:46 -0400)]
freedreno/ir3: comment + better fxn name
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Rob Clark [Fri, 17 Oct 2014 12:57:16 +0000 (08:57 -0400)]
freedreno/a3xx: only emit dirty consts
If app only updates (for example) vertex uniforms, it would be nice to
only re-emit those and not also frag uniforms. Means we need to mark
the first frag shader const buffer dirty after a clear.
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Rob Clark [Wed, 15 Oct 2014 21:15:06 +0000 (17:15 -0400)]
freedreno/a3xx: more layer/level fixes
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Brian Paul [Mon, 20 Oct 2014 17:53:33 +0000 (11:53 -0600)]
mesa: fix 'feeedback' typo in comment
Trivial.
Brian Paul [Mon, 20 Oct 2014 17:49:17 +0000 (11:49 -0600)]
mesa: fix 'misalgned' typos in error messages
Trivial.
Brian Paul [Fri, 17 Oct 2014 19:31:53 +0000 (13:31 -0600)]
glsl: fix several use-after-free bugs
The get_variable_being_redeclared() function can free the 'var' argument.
Thereafter, we cannot assume that 'var' is a valid pointer. This patch
replaces 'var->name' with 'earlier->name' in two places and calls
is_gl_identifier(var->name) before 'var' might get freed.
This fixes several piglit GLSL crashes, including:
spec/glsl-1.50/execution/geometry/clip-distance-in-param
spec/glsl-1.50/execution/geometry/clip-distance-bulk-copy
spec/glsl-1.50/compiler/gs-redeclares-pervertex-out-before-global-redeclaration.geom
I'm not sure why these were not spotted sooner.
A similar bug was previously fixed by
f9cecca7a.
Cc: <mesa-stable@lists.freedesktop.org>
Reviewed-by: Chris Forbes <chrisf@ijw.co.nz>
Tapani Pälli [Tue, 14 Oct 2014 09:39:54 +0000 (12:39 +0300)]
mesa: validate sampler uniforms during gluniform calls
Patch fixes 'glsl-2types-of-textures-on-same-unit' in WebGL conformance
test suite. No Piglit regressions, fixes gl-2.0-active-sampler-conflict.
To avoid adding potentially heavy check during draw (valid_to_render),
check is done during uniform updates by inspecting TexturesUsed mask.
A new boolean variable is introduced to cache validation state.
v2: take into account case where 2 uniforms use same unit (curro)
also do the check only when SSO is not in use, SSO has own
path for sampler validation.
Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Francisco Jerez <currojerez@riseup.net>
EdB [Sat, 11 Oct 2014 16:01:36 +0000 (18:01 +0200)]
clover: Don't return CL_INVALID_VALUE if there is no header.
Reviewed-by: Francisco Jerez <currojerez@riseup.net>
EdB [Sat, 11 Oct 2014 22:58:39 +0000 (01:58 +0300)]
clover: Add allow_empty_tag.
To allow empty objs() list checks.
Reviewed-by: Francisco Jerez <currojerez@riseup.net>
EdB [Mon, 20 Oct 2014 07:34:17 +0000 (10:34 +0300)]
clover: Add initial implementation of clCompileProgram for CL 1.2.
[ Francisco Jerez: General clean-up. ]
Reviewed-by: Francisco Jerez <currojerez@riseup.net>
EdB [Wed, 8 Oct 2014 22:06:48 +0000 (01:06 +0300)]
clover: Add a simple compat::pair.
std::pair is not c++98/c++11 safe.
Reviewed-by: Francisco Jerez <currojerez@riseup.net>
Francisco Jerez [Wed, 8 Oct 2014 22:02:19 +0000 (01:02 +0300)]
clover/util: Allow using key_equals with pair-like objects other than std::pair.
Francisco Jerez [Wed, 8 Oct 2014 22:06:21 +0000 (01:06 +0300)]
clover/util: Define equality operators for a couple of compat classes.
Francisco Jerez [Wed, 8 Oct 2014 17:01:26 +0000 (20:01 +0300)]
clover/util: Fix construction of compat::vector with a general container as argument.
Tapani Pälli [Wed, 6 Aug 2014 06:46:54 +0000 (09:46 +0300)]
glsl: implement switch flow control using a loop
Patch removes old variable based logic for handling a break inside
switch. Switch is put inside a loop so that existing infrastructure
for loop flow control can be used for the switch, now also dead code
elimination works properly.
Possible 'continue' call inside a switch needs now special handling
which is taken care of by detecting continue, breaking out and calling
continue for the outside loop.
v2: remove one unnecessary ir_expression (Curro)
Fixes following Piglit tests:
fs-exec-after-break.shader_test
fs-conditional-break.shader_test
No Piglit or es3conform regressions.
Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Francisco Jerez <currojerez@riseup.net>
Eric Anholt [Sat, 18 Oct 2014 11:50:05 +0000 (12:50 +0100)]
vc4: Translate 4-byte index buffers to 2 bytes.
Fixes assertion failures in 14 piglit tests (half of which now pass).
Eric Anholt [Fri, 3 Oct 2014 05:14:03 +0000 (22:14 -0700)]
vc4: Add support for rebasing texture levels so firstlevel == 0.
GLES2 doesn't have GL_TEXTURE_BASE_LEVEL, so the hardware doesn't. Fixes
piglit levelclamp, tex-miplevel-selection, and texture-storage/2D mipmap
rendering.
Eric Anholt [Fri, 17 Oct 2014 14:28:02 +0000 (15:28 +0100)]
vc4: Apply a Newton-Raphson step to improve RSQ
Fixes all the piglit built-in-functions/*sqrt tests, among others.
Eric Anholt [Fri, 17 Oct 2014 13:01:15 +0000 (14:01 +0100)]
vc4: Apply a Newton-Raphson step to improve RCP.
Fixes all the piglit floating-point *-op-div tests, among others.
Eric Anholt [Fri, 17 Oct 2014 14:04:27 +0000 (15:04 +0100)]
vc4: Add a little bit more packet parsing to make dump reading easier.
Probably should have done this *before* staring at all those render lists
today.
Chris Forbes [Sat, 11 Oct 2014 05:19:17 +0000 (18:19 +1300)]
meta/msaa-blit: consider weird sample count case unreachable
Suppresses a bunch of warning noise about sample_map possibly being used
uninitialized.
Signed-off-by: Chris Forbes <chrisf@ijw.co.nz>
Reviewed-by: Anuj Phogat <anuj.phogat@gmail.com>
Jason Ekstrand [Thu, 16 Oct 2014 19:16:08 +0000 (12:16 -0700)]
i965/fs: Change the type of booleans to UD and emit correct immediates
Before, we used the a signed d-word for booleans and the immedates we
emitted varried between signed and unsigned. This commit changes the type
to unsigned (I think that makes more sense) and makes immediates more
consistent. This allows copy propagation to work better cleans up some
instructions.
total instructions in shared programs:
5473519 ->
5465864 (-0.14%)
instructions in affected programs: 432849 -> 425194 (-1.77%)
GAINED: 27
LOST: 0
Signed-off-by: Jason Ekstrand <jason.ekstrand@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Kenneth Graunke [Fri, 17 Oct 2014 19:59:18 +0000 (12:59 -0700)]
i965/fs: Don't pass ir_variable * to emit_sampleid_setup().
gl_SampleID is a built-in variable that always is of type "int".
Suggested by Connor Abbott.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Connor Abbott <cwabbott0@gmail.com>
Eric Anholt [Fri, 17 Oct 2014 08:43:54 +0000 (09:43 +0100)]
vc4: Make some assertions about how many flushes/EOFs the simulator sees.
This caught the previous commit's bug in the kernel validator.
Eric Anholt [Fri, 17 Oct 2014 11:14:11 +0000 (12:14 +0100)]
vc4: Fix accidental dropping of the low bits of the store tilebuffer packet.
Notably this included the EOF flag (the other bits are the full buffer
dump selection, but we don't do full dumps), which caused the kernel
checking for frame completion to trigger.
Eric Anholt [Thu, 16 Oct 2014 09:17:57 +0000 (10:17 +0100)]
vc4: Set the primitive list format at the start of rendering.
The other driver does this manually before calling into each tile, but we
can just let it get binned into the tiles (saving repeated kernel
validation on the packet).
Fixes simulator assertion failures on polygon-mode and non-auto texwrap.
Eric Anholt [Fri, 17 Oct 2014 08:42:35 +0000 (09:42 +0100)]
vc4: Replace the FLUSH_ALL with FLUSH.
We don't need to emit all of our current state at the end of each bin
list. We're going to be smashing it all at the start of the next tile's
bin list, anyway.
Eric Anholt [Fri, 17 Oct 2014 08:40:12 +0000 (09:40 +0100)]
vc4: Add some comments about state management.
Eric Anholt [Thu, 16 Oct 2014 09:42:04 +0000 (10:42 +0100)]
vc4: Make sure there's exactly 1 tile store per tile coords packet.
It's not documented that I can see, but the other driver does it (check
vg_hw_4.c), and one of the HW guys confirmed that you really do need to do
it.
Michel Dänzer [Thu, 16 Oct 2014 06:10:20 +0000 (15:10 +0900)]
winsys/radeon: Use a single buffer cache manager again
The trick is to generate a unique buffer usage value for each possible
combination of domains and flags, with only one bit set each for the
domains and flags. This ensures pb_check_usage() only returns TRUE when
the domains and flags the cached buffer was created for exactly match
the requested ones.
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Tom Stellard [Tue, 30 Sep 2014 14:32:33 +0000 (10:32 -0400)]
clover: Add environment variables for dumping kernel code v2
There are two debug variables:
CLOVER_DEBUG which you can set to any combination of llvm,clc,asm
(separated by commas) to dump llvm IR, OpenCL C, and native assembly.
CLOVER_DEBUG_FILE which you can set to a file name for dumping output
instead of stderr. If you set this variable, the output will be split
into three separate files with different suffixes: .cl for OpenCL C,
.ll for LLVM IR, and .asm for native assembly. Note that when data
is written, it is always appended to the files.
v2:
- Code cleanups
- Add CLOVER_DEBUG_FILE environment variable for dumping to a file.
Reviewed-by: Francisco Jerez <currojerez@riseup.net>
Tom Stellard [Fri, 26 Sep 2014 01:08:20 +0000 (18:08 -0700)]
clover: Register an llvm diagnostic handler v3
This will allow us to handle internal compiler errors.
v2:
- Code cleanups.
v3:
- More cleanups.
Reviewed-by: Francisco Jerez <currojerez@riseup.net>