Sergii Romantsov [Thu, 6 Sep 2018 09:59:47 +0000 (12:59 +0300)]
i965/tools: 32bit compilation with meson
Building of 32bit mesa with meson causes issue:
"implicit declaration of function ‘__builtin_ia32_clflush’".
Fixed by adding msse2 compilation flag.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=107843
Fixes: 314879f7fec0 (i965: Fix asynchronous mappings on !LLC platforms.)
Signed-off-by: Sergii Romantsov <sergii.romantsov@globallogic.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Timothy Arceri [Sat, 1 Sep 2018 13:57:38 +0000 (23:57 +1000)]
glsl: fixer lexer for unreachable defines
If we have something like:
#ifdef NOT_DEFINED
#define A_MACRO(x) \
if (x)
#endif
The # on the #define is not skipped but the define itself is so
this then gets recognised as #if.
Until
28a3731e3f this didn't happen because we ended up in
<HASH>{NONSPACE} where BEGIN INITIAL was called stopping the
problem from happening.
This change makes sure we never call RETURN_TOKEN_NEVER_SKIP for
if/else/endif when processing a define.
Cc: Ian Romanick <idr@freedesktop.org>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=107772
Tested-By: Eero Tamminen <eero.t.tamminen@intel.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Hyunjun Ko [Thu, 30 Aug 2018 02:58:54 +0000 (11:58 +0900)]
freedreno/ir3: insert mov if same instruction in the outputs.
For example,
result0 = texture(sampler[indexBase + 5], coords);
result1 = texture(sampler[indexBase + 0], coords);
result2 = texture(sampler[indexBase + 0], coords);
out_result0 = result0;
out_result1 = result1;
out_result2 = result2;
In this kind of case we need to insert an extra mov to the outputs
so that the result could be assigned to each register respectively.
Signed-off-by: Rob Clark <robdclark@gmail.com>
Hyunjun Ko [Thu, 30 Aug 2018 02:58:53 +0000 (11:58 +0900)]
freedreno/ir3: make immediates array dynamic
Since most shaders wouldn't need that large array of immediates, making
the array dynamic could save unnecessary spaces.
In addition, sometimes we can potentially have a much larger array
of immediates to be lowered, which might be more than 64.
Signed-off-by: Rob Clark <robdclark@gmail.com>
Rob Clark [Mon, 3 Sep 2018 20:33:59 +0000 (16:33 -0400)]
freedreno: allocate ctx's batch on demand
Don't fall over when app wants more than 32 contexts. Instead allocate
contexts on demand.
Signed-off-by: Rob Clark <robdclark@gmail.com>
Rob Clark [Mon, 3 Sep 2018 20:07:17 +0000 (16:07 -0400)]
freedreno: add fd_context_batch() accessor
For cases in which (after the following commit) ctx->batch may be null.
Prep work for following commit.
Signed-off-by: Rob Clark <robdclark@gmail.com>
Rob Clark [Sun, 2 Sep 2018 17:24:12 +0000 (13:24 -0400)]
freedreno/a6xx: fix mem2gmem for zsbuf
Signed-off-by: Rob Clark <robdclark@gmail.com>
Rob Clark [Wed, 29 Aug 2018 20:04:22 +0000 (16:04 -0400)]
freedreno/batch: fix crash in !reorder case
We aren't using the batch-cache if reorder==false.
Signed-off-by: Rob Clark <robdclark@gmail.com>
Rob Clark [Wed, 29 Aug 2018 19:31:13 +0000 (15:31 -0400)]
freedreno/ir3: better compile_error() printing
Try to show the error at the appropriate line of nir
Signed-off-by: Rob Clark <robdclark@gmail.com>
Rob Clark [Wed, 22 Aug 2018 13:33:06 +0000 (09:33 -0400)]
freedreno/a6xx: bordercolor fixes
Port fixes from a5xx (
f0715442)
TODO maybe this should move to shared code, since it seems to be the
same.
Signed-off-by: Rob Clark <robdclark@gmail.com>
Rob Clark [Wed, 22 Aug 2018 13:07:37 +0000 (09:07 -0400)]
freedreno: fix context teardown harder
The border_color_uploaders need to be torn down before the transfer_pool
is destroyed.
Fixes: e11e9d63943 freedreno: fix context teardown race
Signed-off-by: Rob Clark <robdclark@gmail.com>
Rob Clark [Tue, 21 Aug 2018 12:33:39 +0000 (08:33 -0400)]
freedreno/ir3: ignore unused inputs
We could end up w/ inputs larger than vec4, simply because unused inputs
are not split.
Fixes things like dEQP-GLES31.functional.separate_shader.random.77 (and
probably a handful of others)
Signed-off-by: Rob Clark <robdclark@gmail.com>
Rob Clark [Mon, 20 Aug 2018 14:55:37 +0000 (10:55 -0400)]
freedreno/a6xx: fix debug build crash
Porting
0c8d9e923aa9239e20f9bc969faf9caa0b85237f to a6xx.
Signed-off-by: Rob Clark <robdclark@gmail.com>
Dylan Baker [Tue, 4 Sep 2018 21:42:07 +0000 (14:42 -0700)]
meson: Print a message about why a libdrm version was selected
We require a single version of libdrm for all of our libdrm
dependencies (core and driver), but the way this is structured can make
the error message less than helpful, as one driver might be the one
setting the libdrm requirement, while another might be the one that
generates the version failure.
This adds a simple message to the output announcing which libdrm module
set the version, which might be more helpful.
v2: - Use message suggested by Eric Engstrom
Fixes: c445b1d56f47922206de55e557444aadb62e11f6
("meson: Use the same version for all libdrm checks")
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Charmaine Lee [Fri, 5 May 2017 19:16:33 +0000 (12:16 -0700)]
svga: rename face to layer_face
Reviewed-by: Neha Bhende <bhenden@vmware.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Brian Paul [Mon, 31 Jul 2017 12:40:12 +0000 (06:40 -0600)]
svga: encode sample count in resource declarations
No regressions before the corresponding host-side change.
Reviewed-by: Neha Bhende <bhenden@vmware.com>
Charmaine Lee [Thu, 1 Jun 2017 22:12:14 +0000 (15:12 -0700)]
svga: sync with upstream changes to surface flags
SVGA device now supports 64 bits surface flags. This patch
updates the winsys interface to allow 64 bits surface flags.
The linux winsys layer will for now only honor the lower 32 bits of
the surface flags.
Reviewed-by: Brian Paul <brianp@vmware.com>
Neha Bhende [Wed, 10 May 2017 19:17:44 +0000 (12:17 -0700)]
svga: avoid try_blit() for some depth formats on non vgpu10.
On non vgpu10, driver doesn't support util_blitter_blit for SVGA3D_Z_D16,
SVGA3D_Z_D24x8, SVGA3D_Z_D24S8. Patch fixes following piglit tests regression on hwv8 caused
by commit
27bf35caea5e:
spec@arb_depth_texture@fbo-depth-gl-depth-component16-blit
spec@arb_depth_texture@fbo-depth-gl-depth-component24-blit
spec@arb_depth_texture@fbo-depth-gl-depth-component32-blit
Tested with mtt-piglit on hw 8,9,10,11,13 and mtt-glretrace on windows and linux.
Reviewed-by: Charmaine Lee <charmainel@vmware.com>
Neha Bhende [Thu, 4 May 2017 18:25:04 +0000 (11:25 -0700)]
svga: convert dst format to linear when blending is enabled.
When blending is enabled, framebuffer colorspace has to be linear.
Previously, we never hit this case because we were not supporting sRGB
drawable. Previous patch added that support.
Tested with mtt glretrace, viewperf, piglit, conform.
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Charmaine Lee <charmainel@vmware.com>
Neha Bhende [Wed, 29 Mar 2017 17:35:12 +0000 (10:35 -0700)]
winsys/svga: Avoid cap2 code path for now
CAP2 functionality is not yet part of vmwgfx. This is causing unnecessary
dmesg error messages.
Reviewed-by: Charmaine Lee <charmainel@vmware.com>
Neha Bhende [Mon, 6 Mar 2017 23:55:04 +0000 (15:55 -0800)]
svga: start using SVGA3dCmdIntraSurfaceCopy command for svga_blit.
Basically, SVGA3dCmdIntraSurfaceCopy command allow copying when
source and destination are same.
Tested with MTT piglit, glretrace, viewperf, conform
v2: changes as per Charmaine's comment
v3: changes as per Charmaine's comment
Reviewed-by: Charmaine Lee <charmainel@vmware.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Neha Bhende [Mon, 6 Mar 2017 23:50:20 +0000 (15:50 -0800)]
svga/winsys: Add cap2 support in winsys
Reviewed-by: Charmaine Lee <charmainel@vmware.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Neha Bhende [Mon, 6 Mar 2017 23:53:58 +0000 (15:53 -0800)]
svga: Add SVGA3dCmdIntraSurfaceCopy command support in OpenGL driver
v2: changes as per Charmaine's comment
Reviewed-by: Charmaine Lee <charmainel@vmware.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Brian Paul [Wed, 5 Sep 2018 16:28:25 +0000 (10:28 -0600)]
svga: update device header files from upstream
This is a squash commit of several earlier patches.
Signed-off-by: Brian Paul <brianp@vmware.com>
Charmaine Lee [Thu, 9 Aug 2018 23:22:52 +0000 (16:22 -0700)]
winsys/drm: Fix assert when try to accumulate an invalid fd
This patch makes sure there is a valid fd before merging it
to the context's fd in vmw_svga_winsys_fence_server_sync().
This fixes the assert running webot.
No regression running kmscube.
Reviewed-by: Sinclair Yeh <syeh@vmware.com>
Eric Anholt [Thu, 30 Aug 2018 17:24:51 +0000 (10:24 -0700)]
loader: Drop unused argument from dri3_update_drawable().
The argument has never been used since the function was added.
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Alejandro Piñeiro [Tue, 4 Sep 2018 11:02:57 +0000 (13:02 +0200)]
i965/fs: include multisamplers on image_intrinsic_coord_components
This is the second patch needed to fix the following piglit tests:
tests/spec/arb_gl_spirv/linker/uniform/multisampler.shader_test
tests/spec/arb_gl_spirv/linker/uniform/multisampler-array.shader_test
Although in this case it doesn't affect so many borrowed tests, as
there aren't too many tests using multisamplers on Intel.
It is worth to note that this patch is also needed when those tests
are run on GLSL mode (using the --glsl option). Although most Intel
drivers would not be able to run/execute tests using multisamplers, as
GL_MAX_IMAGE_SAMPLES is zero, technically those tests are expected to
link correctly, so linking tests should pass.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Alejandro Piñeiro [Wed, 5 Sep 2018 10:39:35 +0000 (12:39 +0200)]
i965: move brw_nir_lower_gl_images call
At this moment that lowering is using info coming from the
UniformStorage, so for the ARB_gl_spirv codepath, it needs to be done
after calling gl_nir_link_uniforms. As for the GLSL codepath it can
also be called later, we just move the call on both cases, to avoid
adding several shader->spirv_data checks, and keep the patch as small
as possible.
This is the first patch needed to fix the following piglit tests:
tests/spec/arb_gl_spirv/linker/uniform/multisampler.shader_test
tests/spec/arb_gl_spirv/linker/uniform/multisampler-array.shader_test
but fixes thousands of tests when borrowing the tests from other specs
(that needs to be done manually right now).
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Alejandro Piñeiro [Wed, 5 Sep 2018 10:34:47 +0000 (12:34 +0200)]
intel/compiler: rename brw_nir_lower_glsl_images
To brw_nir_lower_gl_images, as it will be also used on the
ARB_gl_spirv codepath, that doesn't involves GLSL at all. So the
lowering is about images following the OpenGL semantics. In any case
"brw_nir_lower_opengl_images" seemed too long to me, so I just used
gl. That shortening is already used on other parts of the code.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Alejandro Piñeiro [Tue, 4 Sep 2018 10:43:07 +0000 (12:43 +0200)]
intel/compiler: remove unused variable num_images
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Gert Wollny [Tue, 4 Sep 2018 07:36:53 +0000 (09:36 +0200)]
winsys/virgl/vtest: Correct off-by-one error in resource allocation
The resource bo array must already extended when the target index is
equal to the current size of the array.
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Gert Wollny [Mon, 3 Sep 2018 08:14:07 +0000 (10:14 +0200)]
winsys/virgl: Initialize value to silence valgrind
Silences:
Conditional jump or move depends on uninitialised value(s)
at 0xB72F2C0: virgl_drm_winsys_create (virgl_drm_winsys.c:854)
by 0xB72F2C0: virgl_drm_screen_create (virgl_drm_winsys.c:926)
by 0xB21C885: pipe_virgl_create_screen (drm_helper.h:275)
by 0xB7201F0: pipe_loader_create_screen (pipe_loader.c:137)
by 0xB639C91: dri2_init_screen (dri2.c:2112)
by 0xB634F68: driCreateNewScreen2 (dri_util.c:153)
by 0x63023E6: dri3_create_screen (dri3_glx.c:893)
by 0x62D35BD: AllocAndFetchScreenConfigs (glxext.c:820)
by 0x62D35BD: __glXInitialize (glxext.c:946)
by 0x62CECB3: GetGLXPrivScreenConfig (glxcmds.c:174)
by 0x62CF69C: glXQueryExtensionsString (glxcmds.c:1304)
by 0x60AA7D9: ??? (in /usr/lib/x86_64-linux-gnu/libwaffle-1.so.0.5.2)
by 0x4F81450: wfl_checked_display_connect (piglit-util-waffle.h:74)
by 0x4F829E0: piglit_wfl_framework_init (piglit_wfl_framework.c:627)
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Gert Wollny [Mon, 3 Sep 2018 08:05:44 +0000 (10:05 +0200)]
winsys/virgl: correct resource and handle allocation (v2)
Fixes crash with
piglit/bin/map_buffer_range-invalidate CopyBufferSubData \
increment-offset -auto -fbo
* Resize the resource storage already when the count is equal to the
allocated size, fixes:
Invalid write of size 8
at 0xB72E4CF: virgl_drm_add_res (virgl_drm_winsys.c:629)
by 0xB72E4CF: virgl_drm_emit_res (virgl_drm_winsys.c:663)
by 0xB72A44A: virgl_encode_resource_copy_region (virgl_encode.c:776)
by 0xB40CD12: st_copy_buffer_subdata (st_cb_bufferobjects.c:585)
by 0xB244A3B: _mesa_CopyBufferSubData (bufferobj.c:2940)
by 0x109A1E: upload (invalidate.c:169)
by 0x109C2F: piglit_display (invalidate.c:215)
by 0x4F80FBE: run_test (piglit_fbo_framework.c:52)
by 0x4F66E5F: piglit_gl_test_run (piglit-framework-gl.c:229)
by 0x10949D: main (invalidate.c:47)
Address 0xbe07d30 is 0 bytes after a block of size 4,096 alloc'd
at 0x4C31B25: calloc (in
/usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
by 0xB72DAAF: virgl_drm_cmd_buf_create (virgl_drm_winsys.c:567)
* Also resize the space allocated for the handles, fixes:
Invalid write of size 4
at 0xB72E4F0: virgl_drm_add_res (virgl_drm_winsys.c:631)
by 0xB72E4F0: virgl_drm_emit_res (virgl_drm_winsys.c:663)
by 0xB72A44A: virgl_encode_resource_copy_region (virgl_encode.c:776)
by 0xB40CD12: st_copy_buffer_subdata (st_cb_bufferobjects.c:585)
by 0xB244A3B: _mesa_CopyBufferSubData (bufferobj.c:2940)
by 0x109A1E: upload (invalidate.c:169)
by 0x109C2F: piglit_display (invalidate.c:215)
by 0x4F80FBE: run_test (piglit_fbo_framework.c:52)
by 0x4F66E5F: piglit_gl_test_run (piglit-framework-gl.c:229)
by 0x10949D: main (invalidate.c:47)
Address 0xbe08570 is 0 bytes after a block of size 2,048 alloc'd
at 0x4C2FB0F: malloc (
in /usr/lib/valgrind/vgpreload_memcheck-amd64- linux.so)
by 0xB72DAC8: virgl_drm_cmd_buf_create (virgl_drm_winsys.c:572)
Fixes: 4b15b5e803e ("virgl: resize resource bo allocation if we need to.")
v2: - Use REALLOC macro and avoid memory leak when re-allocation fails
- add Fixes tag (both Emil Velikov)
- reorder commit message
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Tomeu Vizoso [Tue, 17 Jul 2018 11:13:21 +0000 (13:13 +0200)]
virgl: use hw-atomics instead of in-ssbo ones
Emulating atomics on top of ssbos can lead to too small max SSBO count,
so let's use the hw-atomics mechanism to expose atomic buffers instead.
Signed-off-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Gurchetan Singh <gurchetansingh@chromium.org>
Erik Faye-Lund [Wed, 29 Aug 2018 07:07:26 +0000 (09:07 +0200)]
virgl: update minor differences to upstream header
virgl_protocol.h is considered to have it's upstream in the
virglrenderer repository, and somehow these minor differences has
crept in.
Let's sync with the upstream to avoid this.
Signed-off-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Gurchetan Singh <gurchetansingh@chromium.org>
Erik Faye-Lund [Wed, 29 Aug 2018 14:11:14 +0000 (16:11 +0200)]
gallium: add PIPE_CAP_MAX_COMBINED_HW_ATOMIC_COUNTER{S,_BUFFERS}
This moves the evergreen-specific max-sizes out as a driver-cap, so
other drivers with less strict requirements also can use hw-atomics.
Remove ssbo_atomic as it's no longer needed.
We should now be able to use hw-atomics for some stages and not for
other, if needed.
Signed-off-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Gurchetan Singh <gurchetansingh@chromium.org>
Erik Faye-Lund [Thu, 30 Aug 2018 09:04:17 +0000 (11:04 +0200)]
gallium: add PIPE_CAP_MAX_COMBINED_SHADER_BUFFERS
This gets rid of a r600 specific hack in the state-tracker, and prepares
for other drivers to be able to use hw-atomics.
While we're at it, clean up some indentation in the various drivers.
Signed-off-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Gurchetan Singh <gurchetansingh@chromium.org>
Erik Faye-Lund [Wed, 29 Aug 2018 13:48:29 +0000 (15:48 +0200)]
st/mesa: simplify MaxAtomicBufferSize-logic
MaxAtomicCounters has already been assigned in the loop above in the
ssbo_atomic = true case, so this will calculate the same value as the
default.
While we're at it, fixup indentation on the MaxAtomicBufferBindings
assign.
Signed-off-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Gurchetan Singh <gurchetansingh@chromium.org>
Erik Faye-Lund [Wed, 29 Aug 2018 13:35:11 +0000 (15:35 +0200)]
st/mesa: clean up atomic vs ssbo code
This makes the code a bit easier to follow; we first set up
MaxShaderStorageBlocks, then we either set up a dedicated
MaxAtomicBuffers, or we split MaxShaderStorageBlocks in two.
While we're at it, also make the SSBO-splitting code tolerate the
hypothetical case of having an odd number of SSBOs without incorrectly
dropping the last SSBO.
This has the nice result that the SSBOs and atomic buffers are dealt
with almost completely orthogonally, easing some upcoming patches.
Signed-off-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Gurchetan Singh <gurchetansingh@chromium.org>
Erik Faye-Lund [Wed, 29 Aug 2018 13:24:26 +0000 (15:24 +0200)]
st/mesa: use real bool for can_ubo
We're doing full c99 now, so there's no point in using the old boolean
type.
Signed-off-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Gurchetan Singh <gurchetansingh@chromium.org>
Marek Olšák [Sat, 1 Sep 2018 07:10:27 +0000 (03:10 -0400)]
gallium/u_threaded: increase batch size to increase performance
This reduces mutex overhead.
radeonsi: +4.4% performance with piglit/drawoverhead, DrawElements, Ryzen X1700
iris_dri.so: +14% with piglit/drawoverhead, DrawArrays, i7 7700HQ.
Acked-by: Kenneth Graunke <kenneth@whitecape.org>
Marek Olšák [Mon, 20 Aug 2018 18:52:52 +0000 (14:52 -0400)]
st/vdpau: silence an unitialized-variable warning
Marek Olšák [Tue, 21 Aug 2018 01:33:24 +0000 (21:33 -0400)]
st/mesa: help fix stencil border color for GL_DEPTH_STENCIL textures
GL_STENCIL_INDEX uses GL_INTENSITY for the border color, which is nicer
to hardware that doesn't read the stencil border value from the X channel.
This fixes a bunch of dEQP tests on Vega & Raven.
Cc: 18.1 18.2 <mesa-stable@lists.freedesktop.org>
Ernestas Kulik [Thu, 30 Aug 2018 16:02:44 +0000 (19:02 +0300)]
glsl_to_tgsi: Fix potential leak
Reported by Coverity: arr_live_ranges is freed in a different branch
than the one in which it was allocated.
Signed-off-by: Ernestas Kulik <ernestas.kulik@gmail.com>
Signed-off-by: Marek Olšák <marek.olsak@amd.com>
Ernestas Kulik [Thu, 30 Aug 2018 16:02:45 +0000 (19:02 +0300)]
u_vbuf: Fix leak
Reported by Coverity: data is heap-allocated, but only freed in the
info->index_size != 0 branch.
Signed-off-by: Ernestas Kulik <ernestas.kulik@gmail.com>
Signed-off-by: Marek Olšák <marek.olsak@amd.com>
Cc: 18.2 <mesa-stable@lists.freedesktop.org>
Eric Anholt [Sat, 11 Aug 2018 01:15:45 +0000 (18:15 -0700)]
freedreno: Drop a bunch of duplicated gallium PIPE_CAP default code.
Now that we have the util function for the default values, we can get rid
of the boilerplate.
v2: Rebase on new gallium caps
Reviewed-by: Rob Clark <robdclark@gmail.com> (v1)
Eric Anholt [Sat, 11 Aug 2018 01:04:40 +0000 (18:04 -0700)]
v3d: Drop a bunch of duplicated gallium PIPE_CAP default code.
Now that we have the util function for the default values, we can get rid
of the boilerplate.
v2: Rebase on new gallium caps
Eric Anholt [Sat, 11 Aug 2018 01:02:02 +0000 (18:02 -0700)]
vc4: Drop a bunch of duplicated gallium PIPE_CAP default code.
Now that we have the util function for the default values, we can get rid
of the boilerplate.
v2: drop GLSL level in favor of defaults.
v3: Rebase on new gallium caps
Eric Anholt [Fri, 10 Aug 2018 23:57:31 +0000 (16:57 -0700)]
gallium: Add a helper for implementing PIPE_CAP_* default values.
One of the pains of implementing a gallium driver is filling in a million
pipe caps you don't know about yet when you're just starting out. One of
the pains of working on gallium is copy-and-pasting your new PIPE_CAP into
each driver. We can fix both of these by having each driver call into the
default helper from their default case, so that both sides can ignore each
other until they need to.
v2: fix i915g build, revert swr change to avoid breaking scons build
(https://travis-ci.org/anholt/mesa/jobs/
419739857)
v3: Rebase on 3 new gallium caps.
Reviewed-by: Marek Olšák <marek.olsak@amd.com> (v1)
Cc: Bruce Cherniak <bruce.cherniak@intel.com>
Cc: George Kyriazis <george.kyriazis@intel.com>
Cc: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Mon, 3 Sep 2018 18:20:54 +0000 (13:20 -0500)]
intel/compiler: Remove redundant nir_remove_dead_variables call
As of
07a2098a708a2, brw_nir_optimize calls nir_remove_dead_variables as
the last optimization. Doing it again is just pointless.
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Lionel Landwerlin [Thu, 23 Aug 2018 13:34:19 +0000 (14:34 +0100)]
intel: compiler: remove dead local variables at optimization pass
We're hitting an assert in gfxbench because one of the local variable
is a sampler (according to Jason this isn't valid) :
testfw_app: ../src/compiler/nir_types.cpp:551: void glsl_get_natural_size_align_bytes(const glsl_type*, unsigned int*, unsigned int*): Assertion `!"type does not have a natural size"' failed.
Since this particular variable isn't used, it can be eliminated by
removing unused local variables at the end of the optimization loop.
This makes sense also for valid local variables.
v2: Move additional local variable removal out of optimization loop,
but before large constant removal (Jason/Lionel)
v3: Move the removal at the end of brw_nir_optimize()
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=107806
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Andrii Simiklit [Mon, 20 Aug 2018 16:20:59 +0000 (19:20 +0300)]
intel/decoder: fix the possible out of bounds group_iter
The "gen_group_get_length" function can return a negative value
and it can lead to the out of bounds group_iter.
v2: printing of "unknown command type" was added
v3: just the asserts are added
Signed-off-by: Andrii Simiklit <andrii.simiklit@globallogic.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Bas Nieuwenhuizen [Mon, 3 Sep 2018 00:34:04 +0000 (02:34 +0200)]
radv: Fix CMASK dimensions.
Mirrors
1e40f694831 "ac/surface: fix CMASK fast clear for NPOT textures with mipmapping on SI/CI/VI"
CC: <mesa-stable@lists.freedesktop.org>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Bas Nieuwenhuizen [Mon, 3 Sep 2018 00:30:48 +0000 (02:30 +0200)]
radv: Use a lower max offchip buffer count.
No clue what gets fixed by this but both radeonsi and amdvlk do it.
CC: <mesa-stable@lists.freedesktop.org>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Bas Nieuwenhuizen [Mon, 3 Sep 2018 00:19:25 +0000 (02:19 +0200)]
radv: Add VEGA20 support.
Just mirror the radeonsi bits. Since this is just adding the extra
switch entries for new HW I think this should be fine for stable.
CC: <mesa-stable@lists.freedesktop.org>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Dave Airlie [Fri, 31 Aug 2018 05:55:15 +0000 (15:55 +1000)]
radv: don't expose linear depth surfaces on SI/CIK/VI either.
ac_surface.c: gfx6_compute_surface says
/* DB doesn't support linear layouts. */
Now if we expose linear depth and create a linear depth image
and use CmdCopyImage to copy into it, we can't map the underlying
memory and read it linearly which I think should work.
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Mauro Rossi [Wed, 15 Aug 2018 12:46:25 +0000 (14:46 +0200)]
egl/android: do not indent HAVE_DRM_GRALLOC preprocessor directive
Fixes: 3f7bca44d9 ("egl/android: #ifdef out flink name support")
Fixes: c7bb82136b ("egl/android: Add DRM node probing and filtering")
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Signed-off-by: Mauro Rossi <issor.oruam@gmail.com>
Jason Ekstrand [Sat, 1 Sep 2018 14:11:17 +0000 (09:11 -0500)]
anv/blorp: Fix a comment as per Nanley's review feedback
This accidentally didn't make it into
62378c5e9e5
Jason Ekstrand [Thu, 30 Aug 2018 17:05:06 +0000 (12:05 -0500)]
anv/blorp: Do more flushing around HiZ clears
We make the flush after a HiZ clear unconditional and add a flush/stall
before the clear as well.
Cc: mesa-stable@lists.freedesktop.org
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=107760
Reviewed-by: Chad Versace <chadversary@chromium.org>
Reviewed-by: Nanley Chery <nanley.g.chery@intel.com>
Ian Romanick [Tue, 19 Jun 2018 00:02:58 +0000 (17:02 -0700)]
i965/vec4: Clamp indirect tes input array reads with 0x0fffffff
Page 190 of "Volume 7: 3D Media GPGPU Engine (Haswell)" says the valid
range of the offset is [0, 0FFFFFFFh].
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Cc: mesa-stable@lists.freedesktop.org
Ian Romanick [Sat, 16 Jun 2018 02:39:56 +0000 (19:39 -0700)]
i965/vec4: Correctly handle uniform sources in generate_tes_add_indirect_urb_offset
Fixes failure in the new piglit test
tes-patch-input-array-vec2-index-invalid-rd.shader_test.
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Cc: mesa-stable@lists.freedesktop.org
Andres Gomez [Thu, 30 Aug 2018 15:03:04 +0000 (18:03 +0300)]
docs: update calendar to extended the 18.1 cycle by one more release
Due to having 2 additional RCs for 18.2.
Cc: Dylan Baker <dylan.c.baker@intel.com>
Cc: Juan A. Suarez <jasuarez@igalia.com>
Cc: Emil Velikov <emil.velikov@collabora.com>
Signed-off-by: Andres Gomez <agomez@igalia.com>
Acked-by: Dylan Baker <dylan@pnwbakers.com>
Acked-by: Juan A. Suarez <jasuarez@igalia.com>
Acked-by: Emil Velikov <emil.velikov@collabora.com>
Rodrigo Vivi [Thu, 30 Aug 2018 21:39:27 +0000 (14:39 -0700)]
intel: Introducing Amber Lake platform
Amber Lake uses the same gen graphics as Kaby Lake, including a id
that were previously marked as reserved on Kaby Lake, but that
now is moved to AML page.
This follows the ids and approach used on kernel's commit
e364672477a1 ("drm/i915/aml: Introducing Amber Lake platform")
Reported-by: Timo Aaltonen <timo.aaltonen@canonical.com>
Cc: José Roberto de Souza <jose.souza@intel.com>
Cc: Anuj Phogat <anuj.phogat@gmail.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Rodrigo Vivi [Thu, 30 Aug 2018 21:32:57 +0000 (14:32 -0700)]
intel: aubinator: Adding missed platforms to the error message.
Many new platforms got added to gen_device_name_to_pci_device_id()
but the error message inside aubinator didn't reflected those
changes. So syncing on the same order to be sure that we are not
missing any now.
Cc: Anuj Phogat <anuj.phogat@gmail.com>
Cc: Matt Turner <mattst88@gmail.com>
Cc: Jordan Justen <jordan.l.justen@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Nanley Chery [Wed, 22 Aug 2018 17:43:32 +0000 (10:43 -0700)]
i965/gen7_urb: Re-emit PUSH_CONSTANT_ALLOC on some gen9
According to internal docs, some gen9 platforms have a pixel shader push
constant synchronization issue. Although not listed among said
platforms, this issue seems to be present on the GeminiLake 2x6's we've
tested.
We consider the available workarounds to be too detrimental on
performance. Instead, we mitigate the issue by applying part of one of
the workarounds. Re-emit PUSH_CONSTANT_ALLOC at the top of every batch
(as suggested by Ken).
Fixes ext_framebuffer_multisample-accuracy piglit test failures with the
following options:
* 6 depth_draw small depthstencil
* 8 stencil_draw small depthstencil
* 6 stencil_draw small depthstencil
* 8 depth_resolve small
* 6 stencil_resolve small depthstencil
* 4 stencil_draw small depthstencil
* 16 stencil_draw small depthstencil
* 16 depth_draw small depthstencil
* 2 stencil_resolve small depthstencil
* 6 stencil_draw small
* all_samples stencil_draw small
* 2 depth_draw small depthstencil
* all_samples depth_draw small depthstencil
* all_samples stencil_resolve small
* 4 depth_draw small depthstencil
* all_samples depth_draw small
* all_samples stencil_draw small depthstencil
* 4 stencil_resolve small depthstencil
* 4 depth_resolve small depthstencil
* all_samples stencil_resolve small depthstencil
v2: Include more platforms in WA (Ken).
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=106865
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=93355
Cc: <mesa-stable@lists.freedesktop.org>
Tested-by: Mark Janes <mark.a.janes@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Christian Gmeiner [Thu, 9 Aug 2018 05:12:24 +0000 (07:12 +0200)]
imx: make use of loader_open_render_node(..) helper
Gets rid of hard-coded gpu device path.
Signed-off-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Christian Gmeiner [Thu, 2 Aug 2018 18:04:45 +0000 (20:04 +0200)]
tegra: make use loader_open_render_node(..) helper
Signed-off-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Christian Gmeiner [Thu, 9 Aug 2018 05:12:22 +0000 (07:12 +0200)]
loader: add loader_open_render_node(..)
This helper is almost a 1:1 copy of tegra_open_render_node().
Signed-off-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Christian Gmeiner [Fri, 10 Aug 2018 09:51:25 +0000 (11:51 +0200)]
tegra: fix memory leak
Fixes: 1755f608f52 ("tegra: Initial support")
Signed-off-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Daniel Stone [Fri, 31 Aug 2018 16:34:03 +0000 (17:34 +0100)]
st/dri: Don't expose sRGB formats to clients
Though the SARGB8888 format is used internally through its FourCC value,
it is not a real format as defined by drm_fourcc.h; it cannot be used
with KMS or other interfaces expecting drm_fourcc.h format codes.
Ensure we don't advertise it through the dmabuf format/modifier query
interfaces, preventing us from tripping over an assert.
Signed-off-by: Daniel Stone <daniels@collabora.com>
Reported-by: Michel Dänzer <michel.daenzer@amd.com>
Fixes: 8c1b9882b2e0 ("egl/dri2: Guard against invalid fourcc formats")
Acked-by: Jason Ekstrand <jason.ekstrand@intel.com>
Samuel Pitoiset [Thu, 30 Aug 2018 09:43:47 +0000 (11:43 +0200)]
radv: add missing support for protected memory properties
Fixes Vulkan CTS CL#2849. Similar to the ANV driver.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Samuel Pitoiset [Thu, 30 Aug 2018 08:33:37 +0000 (10:33 +0200)]
radv: remove dead code in scan_shader_output_decl()
Never used.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Samuel Pitoiset [Thu, 30 Aug 2018 08:33:00 +0000 (10:33 +0200)]
radv: remove radv_shader_context::num_output_{clips,culls}
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Samuel Pitoiset [Thu, 30 Aug 2018 08:30:54 +0000 (10:30 +0200)]
radv: adjust the cull dist mask in scan_shader_output_decl()
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Samuel Pitoiset [Thu, 30 Aug 2018 08:12:03 +0000 (10:12 +0200)]
radv: get length of the clip/cull distances array from usage mask
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Samuel Pitoiset [Thu, 30 Aug 2018 08:01:26 +0000 (10:01 +0200)]
radv: do not recompute the output usage mask for clipdist twice
The shader info pass takes care of this now.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Samuel Pitoiset [Thu, 30 Aug 2018 07:43:29 +0000 (09:43 +0200)]
radv: gather the output usage mask for clip/cull distances correctly
It's a special case because both are combined into a single array.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Samuel Pitoiset [Thu, 30 Aug 2018 07:35:41 +0000 (09:35 +0200)]
radv: add set_output_usage_mask() helper
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Samuel Pitoiset [Wed, 29 Aug 2018 20:13:52 +0000 (22:13 +0200)]
radv: fix passing clip/cull distances from VS to PS
CTS doesn't test input clip/cull distances for the fragment
shader stage, which explains why this was totally broken. I
wrote a simple test locally that works now.
This fixes a crash with GTA V and DXVK.
Note that we are exporting unused parameters from the vertex
shader now, but this can't be optimized easily because we don't
keep the fragment shader info...
Cc: mesa-stable@lists.freedesktop.org
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=107477
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Juan A. Suarez Romero [Thu, 30 Aug 2018 08:14:49 +0000 (10:14 +0200)]
egl/wayland: do not leak wl_buffer when it is locked
If color buffer is locked, do not set its wayland buffer to NULL;
otherwise it can not be freed later.
Rather, flag it in order to destroy it later on the release event.
v2: instruct release event to unlock only or free wl_buffer too (Daniel)
This also fixes dEQP-EGL.functional.swap_buffers_with_damage.* tests.
CC: Daniel Stone <daniel@fooishbar.org>
Reviewed-by: Daniel Stone <daniels@collabora.com>
Dave Airlie [Wed, 29 Aug 2018 03:52:15 +0000 (13:52 +1000)]
ac/radeonsi: fix CIK copy max size
While adding transfer queues to radv, I started writing some tests,
the first test I wrote fell over copying a buffer larger than this
limit.
Checked AMDVLK and found the correct limit.
Cc: <mesa-stable@lists.freedesktop.org>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Dave Airlie [Fri, 31 Aug 2018 00:12:06 +0000 (01:12 +0100)]
radeonsi: fix regression in indirect input swizzles.
This fixes:
tests/spec/arb_enhanced_layouts/execution/component-layout/vs-fs-array-dvec3.shader_test
since I reworked the 64-bit swizzles.
Fixes: bb17ae49ee2 (gallivm: allow to pass two swizzles into fetches.)
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Dave Airlie [Thu, 30 Aug 2018 23:27:44 +0000 (00:27 +0100)]
radeonsi: fix tess/gs fetchs for new swizzle.
I have piglit results from my machine, but I must have messed up,
and not built mesa in between properly.
Fixes: bb17ae49ee2 (gallivm: allow to pass two swizzles into fetches.)
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Marek Olšák [Thu, 30 Aug 2018 19:14:46 +0000 (15:14 -0400)]
mesa: ignore VAO IDs equal to 0 in glDeleteVertexArrays
This fixes a firefox crash.
Fixes: 781a78914c798dc64005b37c6ca1224ce06803fc
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Kenneth Graunke [Thu, 30 Aug 2018 18:19:51 +0000 (11:19 -0700)]
Revert "intel/tools/aubwrite: Always use physical addresses for traces."
This reverts commit
f8cfc7766016d0ff7d52953e7a992b1e77c521d0.
This appears to break intel_dump_gpu for Gen9 systems - I can load them
in the simulator, but nothing happens. Reverting the patch makes the
simulator properly execute our commands and shaders again.
Jason Ekstrand [Thu, 30 Aug 2018 17:50:31 +0000 (12:50 -0500)]
intel/nir: Lowering image loads and stores trashes all metadata
This fixes the GL_ARB_fragment_shader_interlock piglit test on gen8
platforms where the lack of metadata dirtying was causing another pass
to accidentally delete a much needed loop.
https://bugs.freedesktop.org/show_bug.cgi?id=107745
Fixes: 37f7983bcca1 "intel/compiler: Do image load/store lowering..."
Jason Ekstrand <jason@jlekstrand.net> writes:
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
Jason Ekstrand [Tue, 28 Aug 2018 20:25:23 +0000 (15:25 -0500)]
i965/screen: Allow modifiers on sRGB formats
This effectively reverts
a26693493570a9d0f0fba1be617e01ee7bfff4db which
was a misguided attempt at protecting intel_query_dma_buf_modifiers from
invalid formats. Unfortunately, in some internal EGL cases, we can get
an SRGB format validly in this function. Rejecting such formats caused
us to not allow CCS in some cases where we should have been allowing it.
This regressed the performance of some SynMark tests as well as GfxBench
ALU2, Tessellation and Manhattan 3.0 tests
There's some question of whether or not we really should be using SRGB
"fourcc" formats that aren't actually in drm_foucc.h but there's not
much harm in allowing them through here.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=107223
Fixes: a26693493570 "i965/screen: Return false for unsupported..."
Tested-By: Eero Tamminen <eero.t.tamminen@intel.com>
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Jason Ekstrand [Tue, 28 Aug 2018 21:43:57 +0000 (16:43 -0500)]
egl/dri2: Guard against invalid fourcc formats
We already reject attempts to import images with invalid fourcc formats
but don't really guard the queries all that well. This makes us error
out in any calls to eglQueryDmaBufModifiersEXT if the given format is
not a valid fourcc format. We also add an assert to ensure that drivers
don't advertise any non-fourcc formats.
Cc: mesa-stable@lists.freedesktop.org
Tested-By: Eero Tamminen <eero.t.tamminen@intel.com>
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Jason Ekstrand [Tue, 28 Aug 2018 21:31:22 +0000 (16:31 -0500)]
egl/dri2: Add a helper for the number of planes for a FOURCC format
This also serves as a convenient "is this a fourcc format" check as well
which we'll take advantage of in the next commit.
Cc: mesa-stable@lists.freedesktop.org
Tested-By: Eero Tamminen <eero.t.tamminen@intel.com>
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Jason Ekstrand [Thu, 30 Aug 2018 00:47:19 +0000 (19:47 -0500)]
radv/meta: Set num_components on image_store intrinsics
Now that image load/store intrinsics are variable-width, we need to set
num_components accordingly. In
15d39f474b890, both glsl_to_nir and
spirv_to_nir were updated to properly set num_components but radv meta
was left behind.
Fixes: 15d39f474b890 "nir: Make image load/store intrinsics..."
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Tested-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Vicki Pfau [Sun, 19 Aug 2018 21:17:01 +0000 (14:17 -0700)]
gallivm: Detect VSX separately from Altivec
Previously gallivm would attempt to use VSX instructions on all systems
where it detected that Altivec is supported; however, VSX was added to
POWER long after Altivec, causing lots of crashes on older POWER/PPC
hardware, e.g. PPC Macs. By detecting VSX separately from Altivec we can
automatically disable it on hardware that supports Altivec but not VSX
Signed-off-by: Vicki Pfau <vi@endrift.com>
Ilia Mirkin [Sun, 26 Aug 2018 21:47:12 +0000 (17:47 -0400)]
nv50: bump compat glsl level to same as core
Passes the compat piglits. I'm sure that there will be odd issues that
aren't caught by them, but at least it should basically work.
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Ilia Mirkin [Sun, 26 Aug 2018 18:21:39 +0000 (14:21 -0400)]
nvc0: bump compat GLSL version to match core
This passes the handful of tests in piglit.
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Ilia Mirkin [Sun, 26 Aug 2018 17:48:10 +0000 (13:48 -0400)]
glsl: avoid lowering texcoord array except in simple cases
With compat creeping up to geometry and tess shaders, lowering texcoord
accesses/writes becomes more complicated. Since it's an optimization
anyways, just avoid the complication for now.
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Andres Gomez [Thu, 30 Aug 2018 00:32:02 +0000 (03:32 +0300)]
docs: update calendar 18.2.0-rc5 is out, extend to 18.2.0-rc6
Signed-off-by: Andres Gomez <agomez@igalia.com>
Timothy Arceri [Wed, 29 Aug 2018 05:48:47 +0000 (15:48 +1000)]
st/mesa, gallium: add a workaround for No Mans Sky
The spec seems clear this is not allowed but the Nvidia binary
forces apps to add layout qualifiers so this works around the
issue for No Mans Sky until the CTS can be sorted out.
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Timothy Arceri [Wed, 29 Aug 2018 05:48:46 +0000 (15:48 +1000)]
glsl: add a mechanism to allow layout qualifiers on function params
The spec is quite clear this is not allowed:
From Section 4.4. (Layout Qualifiers) of the GLSL 4.60 spec:
"Layout qualifiers can appear in several forms of declaration.
They can appear as part of an interface block definition or
block member, as shown in the grammar in the previous section.
They can also appear with just an interface-qualifier to establish
layouts of other declarations made with that qualifier:
layout-qualifier interface-qualifier ;
Or, they can appear with an individual variable declared with
an interface qualifier:
layout-qualifier interface-qualifier declaration ;"
From Section 4.10 (Memory Qualifiers) of the GLSL 4.60 spec:
"Layout qualifiers cannot be used on formal function parameters,
and layout qualification is not included in parameter matching."
However on the Nvidia binary driver they actually fail to compile
if image function params don't have a layout qualifier. This results
in applications such as No Mans Sky using layout qualifiers on params.
I've submitted a CTS test to expose this problem in the Nvidia driver
but until that is resolved this patch will help Mesa drivers work
around the issue.
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Timothy Arceri [Wed, 29 Aug 2018 01:36:51 +0000 (11:36 +1000)]
glsl: skip stringification in preprocessor if in unreachable branch
This fixes compilation of some "No Mans Sky" shaders where the stringification
happens in branches intended for DX12.
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Bas Nieuwenhuizen [Wed, 29 Aug 2018 15:04:25 +0000 (17:04 +0200)]
radv: Add missing checks in radv_get_image_format_properties.
CC: <mesa-stable@lists.freedesktop.org>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Dave Airlie [Mon, 27 Aug 2018 01:03:41 +0000 (02:03 +0100)]
gallivm: allow to pass two swizzles into fetches.
This hijacks the top 16-bits of swizzle, to pass in the swizzle
for the second channel.
This fixes handling .yx swizzles of 64-bit values.
This should fixup radeonsi and llvmpipe.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=107524
Reviewed-by: Marek Olšák <marek.olsak@amd.com>