Marek Olšák [Tue, 27 Jun 2017 19:54:20 +0000 (21:54 +0200)]
winsys/amdgpu: use 128KB BOs for suballocations of up to 64KB BOs
This decreases the number of BOs, but might also increase memory usage.
It's better for small textures.
The gameplay is on the far right:
https://people.freedesktop.org/~mareko/suballoc.svg
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Marek Olšák [Tue, 27 Jun 2017 18:32:37 +0000 (20:32 +0200)]
gallium/radeon: allow suballocating textures
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Marek Olšák [Tue, 27 Jun 2017 18:12:25 +0000 (20:12 +0200)]
gallium/radeon: generalize the function for in-place texture reallocation
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Marek Olšák [Tue, 27 Jun 2017 18:27:29 +0000 (20:27 +0200)]
gallium/radeon: add radeon_winsys::buffer_is_suballocated
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Marek Olšák [Thu, 29 Jun 2017 16:49:48 +0000 (18:49 +0200)]
gallium/radeon: clean up pb_cache bucket/usage determination
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Marek Olšák [Thu, 29 Jun 2017 16:33:05 +0000 (18:33 +0200)]
gallium/radeon: enable suballocations for VRAM with no CPU access
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Marek Olšák [Thu, 29 Jun 2017 16:01:16 +0000 (18:01 +0200)]
gallium/radeon: clean up (domain, flags) <-> (slab heap) translations
This is cleaner, and we are down to 4 slabs.
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Marek Olšák [Thu, 29 Jun 2017 15:09:04 +0000 (17:09 +0200)]
gallium/radeon: remove RADEON_FLAG_CPU_ACCESS
https://lists.freedesktop.org/archives/amd-gfx/2017-June/010591.html
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Marek Olšák [Tue, 27 Jun 2017 17:32:48 +0000 (19:32 +0200)]
gallium/radeon: disallow exports of sparse and suballocated BOs
I think it's unsafe, because the slabs can reuse exported storage.
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Marek Olšák [Tue, 27 Jun 2017 17:40:09 +0000 (19:40 +0200)]
gallium/radeon: clean up r600_texture_get_handle
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Marek Olšák [Tue, 27 Jun 2017 17:37:43 +0000 (19:37 +0200)]
gallium/radeon: rename RADEON_FLAG_HANDLE -> RADEON_FLAG_NO_SUBALLOC
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Marek Olšák [Tue, 27 Jun 2017 17:24:20 +0000 (19:24 +0200)]
gallium/radeon: fix a possible crash for buffer exports
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Marek Olšák [Tue, 27 Jun 2017 17:21:29 +0000 (19:21 +0200)]
gallium/radeon: ignore PIPE_BIND_SHARED for buffers
BO exports can't be predicted this way.
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Marek Olšák [Thu, 29 Jun 2017 17:20:06 +0000 (19:20 +0200)]
radeonsi: add a HUD query for getting an average GFX BO list size
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Philipp Zabel [Fri, 30 Jun 2017 08:39:53 +0000 (10:39 +0200)]
st/mesa: release EGLImage on EGLImageTarget* error
The smapi->get_egl_image() call in st_egl_image_get_surface() stores a
reference to the EGLImage's texture in stimg.texture. That reference is
released via pipe_resource_reference(&stimg.texture, NULL) before stimg
goes out of scope at the end of the function, but not in the error path
if !is_format_supported().
Fixes:
83e9de25f325 ("st/mesa: EGLImageTarget* error handling")
Cc: mesa-stable@lists.freedesktop.org
Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Juan A. Suarez Romero [Mon, 3 Jul 2017 11:04:30 +0000 (11:04 +0000)]
vc4: automake: include vc4_cl_dump.h in
Ensure vc4_cl_dump.h and $(BROADCOM_FILES) are distributed in the
dist-file.
This fixes `make distcheck`
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Marek Olšák [Sat, 1 Jul 2017 14:05:48 +0000 (16:05 +0200)]
st/mesa: fix tessellation shaders with no support for shareable shaders
Broken by:
b43c887a9bf1e3fb99b0dc22bfea5db81375a06e
Reported by Gert Wollny.
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Dave Airlie [Wed, 15 Feb 2017 04:58:48 +0000 (04:58 +0000)]
radv: enable Int64 capability (v2)
I'm not 100% sure this is all wired up but it looks like it is.
v2: actually enable extension.
Signed-off-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Connor Abbott [Sat, 1 Jul 2017 02:10:43 +0000 (19:10 -0700)]
ac/nir: fix 64-bit shifts
NIR always makes the shift amount 32 bits, but LLVM asserts if the two
sources aren't the same type. Zero-extend the shift amount to make LLVM
happy.
Signed-off-by: Connor Abbott <cwabbott0@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Connor Abbott [Wed, 7 Jun 2017 21:16:31 +0000 (14:16 -0700)]
ac/nir: implement 64-bit packing and unpacking
We implement the split opcodes, and tell NIR to lower the original ones.
The lowering to LLVM is a little more complicated, but NIR can optimize
the split ones a little better, and some NIR lowering passes that we
might want to use (particularly for doubles) emit the split ones.
This should fix pack/unpackDouble2x32, which seems like a bug since when
we enabled the Float64 capability. It will also fix pack/unpackInt2x32
when we enable the Int64 capability.
Fixes:
798ae37c ("radv: Enable Float64 support.")
Signed-off-by: Connor Abbott <cwabbott0@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Connor Abbott [Wed, 7 Jun 2017 21:12:05 +0000 (14:12 -0700)]
spirv: fix OpBitcast when the src and dst bitsize are different (v3)
Before, we were just implementing it with a move, which is incorrect
when the source and destination have different bitsizes. To implement
it properly, we need to use the 64-bit pack/unpack opcodes. Since
glslang uses OpBitcast to implement packInt2x32 and unpackInt2x32, this
should fix them on anv (and radv once we enable the int64 capability).
v2: make supporting non-32/64 bit easier (Jason)
v3: add another assert (Jason)
Fixes:
b3135c3c ("anv: Advertise shaderInt64 on Broadwell and above")
Signed-off-by: Connor Abbott <cwabbott0@gmail.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Brian Paul [Fri, 30 Jun 2017 21:11:01 +0000 (14:11 -0700)]
svga: don't call svga_texture_device_format_has_alpha() for PIPE_BUFFER
svga_texture_device_format_has_alpha() is only intended to work for
texture resources, not buffer resources. This fixes a failed assertion
in the svga_texture() cast function when running texture buffer tests.
Also, add an assertion in svga_texture_device_format_has_alpha() to
catch the issue sooner.
Reviewed-by: Charmaine Lee <charmainel@vmware.com>
Brian Paul [Fri, 30 Jun 2017 21:08:49 +0000 (14:08 -0700)]
svga: fix texture buffer object regression
With change
8aba778fa2cd98a0b5a7429d3c5057778a0c808c we stopped binding
sampler objects for texture buffers. That broke our texture sample /
sampler view setup code.
Now, we loop over the max(num samplers, num sampler views) and handle
the sampler and view information separately. For texture buffers,
the sampler will be NULL but the sampler view non-null.
Reviewed-by: Charmaine Lee <charmainel@vmware.com>
Brian Paul [Fri, 30 Jun 2017 21:07:08 +0000 (14:07 -0700)]
svga: move assertion in draw_vgpu10()
The buffer binding flags aren't ensured until after the
svga_buffer_handle() call, so move the assertion after it.
Reviewed-by: Charmaine Lee <charmainel@vmware.com>
Brian Paul [Fri, 30 Jun 2017 21:02:20 +0000 (14:02 -0700)]
svga: fix buffer binding flags initialization
If a buffer is created/initialized with glNamedBufferData we will
have no target (GL_ARRAY_BUFFER, GL_UNIFORM_BUFFER, etc) so the
svga_buffer::bind_flags will be zero until we try to get the buffer
handle.
This patch initializes the svga_buffer::bind_flags field when it's
zero.
This fixes the Piglit arb_uniform_buffer_object-rendering-dsa test.
Note that there's still issues in this area that'll have to be
addressed in the future. For example, creating a buffer object
as GL_UNIFORM_BUFFER and later using it as a vertex buffer will
fail.
Reviewed-by: Charmaine Lee <charmainel@vmware.com>
Brian Paul [Sun, 2 Jul 2017 14:48:26 +0000 (08:48 -0600)]
docs: update bug reporting guidelines
Suggest attaching output of glxinfo/wglinfo. Suggest providing
an apitrace.
Nicolai Hähnle [Sun, 25 Jun 2017 16:10:32 +0000 (18:10 +0200)]
st/mesa: remove an obsolete comment
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Nicolai Hähnle [Sat, 24 Jun 2017 10:26:46 +0000 (12:26 +0200)]
mesa: remove unused parameter/member of add_uniform_to_shader
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Nicolai Hähnle [Sun, 14 May 2017 22:47:36 +0000 (00:47 +0200)]
util/disk_cache: fix a comment
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Nicolai Hähnle [Sun, 11 Jun 2017 10:48:52 +0000 (12:48 +0200)]
glsl: simplify disable_varying_optimizations_for_sso
We always have stage == first and stage == last when first == last, so
drop the special case. Also rephrase the comment to make the logic
clearer.
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Nicolai Hähnle [Wed, 14 Jun 2017 10:42:28 +0000 (12:42 +0200)]
glsl: always print non-zero var->data.location_frac
This is helpful in debugging varying assignments.
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Nicolai Hähnle [Tue, 27 Jun 2017 07:50:32 +0000 (09:50 +0200)]
winsys/radeon: only call pb_slabs_reclaim when slabs are actually used
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=100242
Fixes:
fb827c055cb1 ("winsys/radeon: enable buffer allocation from slabs")
Cc: mesa-stable@lists.freedesktop.org
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Samuel Iglesias Gonsálvez [Wed, 28 Jun 2017 13:52:34 +0000 (15:52 +0200)]
anv: check support for enabled features in vkCreateDevice()
From Vulkan spec, 4.2.1. "Device Creation":
"vkCreateDevice verifies that extensions and features requested in
the ppEnabledExtensionNames and pEnabledFeatures members of
pCreateInfo, respectively, are supported by the implementation."
Signed-off-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@gmail.com>
Samuel Iglesias Gonsálvez [Wed, 28 Jun 2017 07:39:55 +0000 (09:39 +0200)]
anv: merge tessellation's primitive mode in merge_tess_info()
SPIR-V tessellation shaders that were created from HLSL will have
the primitive generation domain set in tessellation control shader
(hull shader in HLSL) instead of the tessellation evaluation shader.
v2:
- Add assert (Kenneth)
Signed-off-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Bruce Cherniak [Sat, 1 Jul 2017 03:24:46 +0000 (22:24 -0500)]
swr: Limit memory held by defer deleted resources.
This patch limits the number of items on the fence work queue (the
deferred deletion list) by submitting a sync fence when the queue size
exceeds a threshold. This initiates deferred deletion of all resources
on the list and decreases the total amount of memory held waiting for
"deferred deletion".
This resolves bug 101467 filed against swr for the piglit
streaming-texture-leak test. For those running on smaller memory
(16GB?) systems, this will prevent oom-killer.
Thus far, we have not seen any real world applications that exhibit
behavior like the streaming-texture-leak test; as any form of pipeline
flush will trigger the defer queue and properly free any retained
allocations. But, this addresses those as well.
Cc: "17.1" <mesa-stable@lists.freedesktop.org>
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Lionel Landwerlin [Sat, 1 Jul 2017 23:23:29 +0000 (00:23 +0100)]
anv: fix reported timestampPeriod value
We lost some precision on a previous change due to switching to
integers. Since we report a float in timestampPeriod, we want the
division to happen in floats.
CID: 1413021
Fixes:
c77d98ef32 ("intel: common: express timestamps units in frequency")
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Lionel Landwerlin [Fri, 30 Jun 2017 16:25:43 +0000 (17:25 +0100)]
intel: genxml: make a couple of enums show up in aubinator
In particular Shader Channel Select & Texture Address Control Mode.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com>
Kenneth Graunke [Sat, 1 Jul 2017 04:16:00 +0000 (21:16 -0700)]
i965: Print access flags in INTEL_DEBUG=buf output.
Being able to see the access mode of various mappings is incredibly
useful for debugging. With this patch, INTEL_DEBUG=buf now shows
data such as:
bo_create: buf 7 (bufferobj) 640b
bo_map_gtt: 7 (bufferobj) -> 0x7fca1fae5000, WRITE ASYNC
brw_bo_map_cpu: 7 (bufferobj) -> 0x7fca1fae4000, READ
bo_map_gtt: 5 (bufferobj) -> 0x7fca1fad4000, WRITE ASYNC
brw_bo_map_cpu: 7 (bufferobj) -> 0x7fca1fae4000, READ
which makes it easy to see that there are async GTT writes with
intervening CPU reads.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Chris Wilson [Sat, 1 Jul 2017 10:14:00 +0000 (11:14 +0100)]
i965: Remove clearing of bo->map_gtt after failure
With the conversion to storing the result of drm_mmap to a local and not
directly to bo->map_gtt itself, we no longer should clear bo->map_gtt.
In the best the operation is redundant as we know bo->map_gtt to already
be NULL, but in the worst case we overwrite a concurrent thread that
successfully mmaped the GTT.
Fixes:
314647c4c206 ("i965: Drop global bufmgr lock from brw_bo_map_* functions.")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Kenneth Graunke [Sat, 1 Jul 2017 03:33:57 +0000 (20:33 -0700)]
i965: Add inline to brw_bo_unmap
I meant to do this in "i965: Make brw_bo_unmap a static inline."
but botched the commit fixup.
Chris Wilson [Fri, 30 Jun 2017 21:21:54 +0000 (14:21 -0700)]
i965: Drop global bufmgr lock from brw_bo_map_* functions.
After removing the unusuable debugging code in the previous commit, we
can also entirely remove the global mutex around mapping the buffer for
the first time and replace it with a single atomic operation to update
the cache once we retrieve the mmap.
v2 (Ken): Split out from Chris's original commit.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Kenneth Graunke [Fri, 30 Jun 2017 21:05:21 +0000 (14:05 -0700)]
i965: Make brw_bo_unmap a static inline.
With the broken debugging code gone, it doesn't do anything anymore.
We could technically eliminate it, but I'd like to keep it around in
case we want to add something there again someday. Otherwise we'd
have to go all over the codebase adding unmap calls back again.
Based on a patch by Chris Wilson.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Chris Wilson [Tue, 20 Jun 2017 10:57:34 +0000 (11:57 +0100)]
i965: Discard bo->map_count
Supposedly we were keeping a reference count for the number of users of
a mapping so that we could use valgrind to detect access to the map
outside of the valid section. However, we were incrementing the counter
only when first creating the cached mapping but decrementing on every
unmap. The bo->map_count tracking was wrong and so the debugging code
was completely useless.
v2 (Ken): Separate out atomic compare and swap optimization.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Kenneth Graunke [Fri, 30 Jun 2017 21:00:01 +0000 (14:00 -0700)]
i965: Add a comment about not needing VALGRIND_MALLOCLIKE_BLOCK.
At first glance this seems missing, since we handle it manually for CPU
and WC maps. Although a bit inconsistent, it's actually not necessary.
Thanks to Chris Wilson for explaining this to me.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Bas Nieuwenhuizen [Fri, 30 Jun 2017 19:03:40 +0000 (21:03 +0200)]
radv: Use v4i32 variant of llvm.SI.load.const.
We apparently still used v16i8 ....
As radeonsi doesn't use it with LLVM version checks I don't think
we need them either.
Signed-off-by: Bas Nieuwenhuizen <basni@google.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Brian Paul [Tue, 27 Jun 2017 15:51:25 +0000 (09:51 -0600)]
svga: add texture size/levels sanity check code in svga_texture_create()
The state tracker should never ask us to create a texture with invalid
dimensions / mipmap levels. Do some assertions to check that.
No Piglit regressions.
Reviewed-by: Charmaine Lee <charmainel@vmware.com>
Brian Paul [Tue, 27 Jun 2017 15:36:19 +0000 (09:36 -0600)]
st/mesa: fix texture image resource selection in st_render_texture()
If we're rendering to an incomplete/inconsistent (cube) texture, the
different faces/levels of the texture may be stored in different
resources. Before, we always used the texture object resource. Now,
we use the texture image resource. In normal circumstances, that's
the same resource. But in some cases, such as the Piglit
fbo-incomplete-texture-03 test, the cube faces are in different
resources and we need to render to the texture image resource.
Fixes fbo-incomplete-texture-03 with VMware driver.
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Brian Paul [Tue, 27 Jun 2017 15:29:18 +0000 (09:29 -0600)]
st/mesa: check for incomplete texture in st_finalize_texture()
Return early from st_finalize_texture() if we have an incomplete
texture. This avoids trying to create a texture resource with invalid
parameters (too many mipmap levels given the base dimension).
Specifically, the Piglit fbo-incomplete-texture-03 test winds up
calling pipe_screen::resource_create() with width0=32, height0=32 and
last_level=6 because the first five cube faces are 32x32 but the sixth
face is 64x64. Some drivers handle this, but others (like VMware svga)
do not (generates device errors).
Note that this code is on the path that's usually not taken (we normally
build consistent textures).
No Piglit regressions.
v2: only need to check for base-level completeness since that's what has to
be consistent in order to specify the dimensions for a new gallium texture.
Per Roland.
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Brian Paul [Fri, 30 Jun 2017 15:23:43 +0000 (09:23 -0600)]
gallium/docs: document that TXF is used with PIPE_BUFFER resources
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Brian Paul [Fri, 30 Jun 2017 15:23:17 +0000 (09:23 -0600)]
gallium/docs: clarify that samplers are not used with PIPE_BUFFER resources
Commit
8aba778fa2cd98a0b5a7429d3c5057778a0c808c "st/mesa: don't set
sampler states for TBOs" changed how texture buffer objects are handled.
Document the new convention.
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Eric Anholt [Fri, 16 Dec 2016 19:00:46 +0000 (11:00 -0800)]
vc4: Start using XML unpack functions in CL dump.
For now this is a no-op on the output, but it makes it clear that we've
had weird things going on with things like
V3D21_CLIPPER_Z_SCALE_AND_OFFSET.
Eric Anholt [Mon, 6 Feb 2017 22:29:30 +0000 (14:29 -0800)]
vc4: Replace a couple of magic numbers with #define usage.
Eric Anholt [Mon, 6 Feb 2017 22:06:12 +0000 (14:06 -0800)]
vc4: Move rasterizer state packing to CSO creation time.
This gets our vc4_emit.c size back down a bit:
before:
1020 0 0 1020 3fc src/gallium/drivers/vc4/.libs/vc4_emit.o
after:
968 0 0 968 3c8 src/gallium/drivers/vc4/.libs/vc4_emit.o
Eric Anholt [Sat, 10 Dec 2016 23:47:26 +0000 (15:47 -0800)]
vc4: Convert the driver to emitting the shader record using pack macros.
Eric Anholt [Sun, 11 Dec 2016 22:59:17 +0000 (14:59 -0800)]
vc4: Simplify pack header usage
Take the CL pointer in, which will be useful for enabling relocs.
However, our code expands a bit more:
before:
4449 0 0 4449 1161 src/gallium/drivers/vc4/.libs/vc4_draw.o
988 0 0 988 3dc src/gallium/drivers/vc4/.libs/vc4_emit.o
after:
4481 0 0 4481 1181 src/gallium/drivers/vc4/.libs/vc4_draw.o
1020 0 0 1020 3fc src/gallium/drivers/vc4/.libs/vc4_emit.o
Eric Anholt [Wed, 26 Oct 2016 19:46:58 +0000 (12:46 -0700)]
vc4: Start using the pack header.
This slightly inflates the size of the generated code, in exchange for
getting us some convenient tools.
before:
4389 0 0 4389 1125 src/gallium/drivers/vc4/.libs/vc4_draw.o
808 0 0 808 328 src/gallium/drivers/vc4/.libs/vc4_emit.o
after:
4449 0 0 4449 1161 src/gallium/drivers/vc4/.libs/vc4_draw.o
988 0 0 988 3dc src/gallium/drivers/vc4/.libs/vc4_emit.o
Eric Anholt [Wed, 26 Oct 2016 17:14:37 +0000 (10:14 -0700)]
vc4: Introduce XML-based packet header generation like Intel's.
I really liked this idea, as it should help with management of packet
parsing tools like the CL dump. The python script is forked off of theirs
because our packets are byte-based instead of dwords, and the changes to
do so while avoiding performance regressions due to unaligned accesses
were quite invasive.
v2: Fix Android.mk paths, drop shebang for python script, fix overlap
detection.
Acked-by: Jason Ekstrand <jason@jlekstrand.net>
Acked-by: Kenneth Graunke <kenneth@whitecape.org>
Tested-by: Rob Herring <robh@kernel.org>
Bruce Cherniak [Thu, 29 Jun 2017 19:41:31 +0000 (14:41 -0500)]
swr: Minor cleanup of variable usage, no functional change.
In swr_update_derived, for consistency, index buffer validation should
be using the p_draw_info copy "info" rather than referencing
p_draw_info.
No functional change.
Reviewed-by: Tim Rowley <timothy.o.rowley@intel.com>
Tim Rowley [Fri, 30 Jun 2017 16:50:22 +0000 (11:50 -0500)]
swr: use swr_query_result type instead of void
Tag pStat field in swr_draw_context structure so gen_llvm_types.py
can deal with the actual structure type instead of using void.
Code cleanup, no functional change.
Reviewed-by: Bruce Cherniak <bruce.cherniak@intel.com>
Tim Rowley [Wed, 21 Jun 2017 19:57:10 +0000 (14:57 -0500)]
swr/rast: increase number of possible draws in flight
Increases performance of some large workloads on KNL by ~30%.
Reviewed-by: Bruce Cherniak <bruce.cherniak at intel.com>
Tim Rowley [Wed, 21 Jun 2017 19:56:24 +0000 (14:56 -0500)]
swr/rast: move default split size from driver to rasterizer
Reviewed-by: Bruce Cherniak <bruce.cherniak at intel.com>
Tim Rowley [Mon, 19 Jun 2017 17:14:15 +0000 (12:14 -0500)]
swr/rast: Fix missing setup of psContext.pColorBuffer
Fixes render target read access from pixel shaders.
Reviewed-by: Bruce Cherniak <bruce.cherniak at intel.com>
Tim Rowley [Thu, 15 Jun 2017 20:24:07 +0000 (15:24 -0500)]
swr/rast: Switch intrinsic usage to SIMDLib
Switch from a macro-based simd intrinsics layer to a more C++
implementation, which also adds AVX512 optimizations to 128-bit
and 256-bit SIMD.
Reviewed-by: Bruce Cherniak <bruce.cherniak at intel.com>
Tim Rowley [Tue, 27 Jun 2017 15:47:28 +0000 (10:47 -0500)]
scons: allow .inl file extension
Intended for header files which are not meant to be included directly.
Reviewed-by: Bruce Cherniak <bruce.cherniak at intel.com>
Tim Rowley [Wed, 14 Jun 2017 19:40:28 +0000 (14:40 -0500)]
swr/rast: Fix unused variable warnings
Reviewed-by: Bruce Cherniak <bruce.cherniak at intel.com>
Tim Rowley [Mon, 26 Jun 2017 17:41:38 +0000 (12:41 -0500)]
swr/rast: Split rasterizer.cpp to improve compile time
Hardcode split to four files currently. Decreases swr build
time on KNL by over 50%.
Reviewed-by: Bruce Cherniak <bruce.cherniak at intel.com>
Tim Rowley [Mon, 26 Jun 2017 17:40:45 +0000 (12:40 -0500)]
swr/rast: gen_backends.py remove extraneous semicolon
Reviewed-by: Bruce Cherniak <bruce.cherniak at intel.com>
Tim Rowley [Mon, 12 Jun 2017 23:43:49 +0000 (18:43 -0500)]
swr/rast: Support dynamically sized vertex layout
Each shader stage state (VS, TS, GS, SO, BE/CLIP) now has a
vertexAttribOffset to specify the offset to the start of the
general attribute section of the incoming verts for that stage.
It is up to the driver to set this up correctly based on the
active stages. All the shader stages use this value instead of
VERTEX_ATTRIB_START_SLOT to offset to the incoming attributes.
Only the vertex shader stage supports dynamic layout output
currently. The other stages continue to expect the output to be
the fixed layout slots as before. Will be enabling GS next.
Reviewed-by: Bruce Cherniak <bruce.cherniak at intel.com>
Tim Rowley [Mon, 26 Jun 2017 18:00:27 +0000 (13:00 -0500)]
swr/rast: Split backend.cpp to improve compile time
Hardcode split to four files currently. Decreases swr build
time on a quad-core by ~10%.
Reviewed-by: Bruce Cherniak <bruce.cherniak at intel.com>
Tim Rowley [Mon, 26 Jun 2017 17:59:28 +0000 (12:59 -0500)]
swr/rast: gen_backends.py removal of commented debug prints
Reviewed-by: Bruce Cherniak <bruce.cherniak at intel.com>
Tim Rowley [Mon, 26 Jun 2017 17:43:49 +0000 (12:43 -0500)]
swr/rast: gen_backends.py quote cleanup
Reviewed-by: Bruce Cherniak <bruce.cherniak at intel.com>
Tim Rowley [Mon, 26 Jun 2017 17:34:24 +0000 (12:34 -0500)]
swr/rast: generators will create target directories
Reviewed-by: Bruce Cherniak <bruce.cherniak at intel.com>
Andres Gomez [Fri, 30 Jun 2017 17:37:52 +0000 (20:37 +0300)]
docs: update calendar, add news item and link release notes for 17.1.4
Signed-off-by: Andres Gomez <agomez@igalia.com>
Andres Gomez [Fri, 30 Jun 2017 17:31:44 +0000 (20:31 +0300)]
docs: add sha256 checksums for 17.1.4
Signed-off-by: Andres Gomez <agomez@igalia.com>
(cherry picked from commit
5a24aa8c55c2b7893d50a4a203f14300d4ecab87)
Andres Gomez [Fri, 30 Jun 2017 17:11:50 +0000 (20:11 +0300)]
docs: add release notes for 17.1.4
Signed-off-by: Andres Gomez <agomez@igalia.com>
(cherry picked from commit
e60d010ef41f049df84d683a6465d1e5bf0cbb3d)
Mauro Rossi [Fri, 30 Jun 2017 06:54:34 +0000 (08:54 +0200)]
Android: fix typo in symlink for driver loading and 32 bit builds
There is typo in the mkdir command path,
the correct one is $(TARGET_OUT)/$(l)/$(MESA_DRI_MODULE_REL_PATH)
The other issue is in 32bit builds, because lib64 does not exist there,
we can use TARGET_IS_64_BIT to refine the post install command.
Fixes:
a3d98ca62f ("Android: use symlinks for driver loading")
Signed-off-by: Rob Herring <robh@kernel.org>
Brian Paul [Wed, 28 Jun 2017 22:43:52 +0000 (16:43 -0600)]
svga: update a few surface format names
To sync with in-house changes.
Reviewed-by: Neha Bhende <bhenden@vmware.com>
Brian Paul [Tue, 27 Jun 2017 21:55:10 +0000 (15:55 -0600)]
svga: whitespace fixes in svga_resource_buffer_upload.c
Trivial.
Charmaine Lee [Tue, 27 Jun 2017 16:59:55 +0000 (10:59 -0600)]
svga: add mksstats for surface view emulation
Add mksstats for surface view emulation and also tighten the stat
CreateBackedView for the actual creation of backed view.
Reviewed-by: Brian Paul <brianp@vmware.com>
Brian Paul [Thu, 22 Jun 2017 20:45:07 +0000 (14:45 -0600)]
svga: change error handling convention for svga_set_stream_output()
In general, the functions which emit commands to the command buffer check
for failure and return a PIPE_ERROR_x code. It's up to the caller to
flush the buffer and retry the command.
But svga_set_stream_output() did its own flushing and the callers never
checked the return value (though, it would always be PIPE_OK) in practice.
This patch changes svga_set_stream_output() so that it does not call
svga_context_flush() when the buffer is full. And we update the callers
to check the return value as we do for other functions, like
svga_set_shader().
No Piglit regressions. Also tested w/ Nature demo.
Reviewed-by: Charmaine Lee <charmainel@vmware.com>
Charmaine Lee [Tue, 6 Jun 2017 18:52:50 +0000 (11:52 -0700)]
svga: fixed surface size to include array size
This patch fixes the total surface size in surface cache
to include array size as well.
Tested with MTT glretrace.
Reviewed-by: Brian Paul <brianp@vmware.com>
Neha Bhende [Thu, 27 Oct 2016 19:35:03 +0000 (12:35 -0700)]
svga: loop over box.depth for ReadBack_image on each slice
piglit test ext_texture_array-gen-mipmap is fixed with this patch.
Tested with mtt piglit, glretrace, viewperf and conform. No regression.
Reviewed-by: Brian Paul <brianp@vmware.com>
Charmaine Lee [Tue, 8 Nov 2016 19:54:56 +0000 (11:54 -0800)]
svga: add mksstats for context creation
Reviewed-by: Brian Paul <brianp@vmware.com>
Charmaine Lee [Tue, 27 Jun 2017 16:47:01 +0000 (10:47 -0600)]
svga: re-validate sampler view at draw time if needed
This patch validates those sampler views with backing copy
of texture whose original copy has been updated since the
view is last validated.
This is done here at draw time because the texture binding might not
have modified, hence validation is not triggered at state update time,
and yet the texture might have been updated in another context, so
we need to re-validate the sampler view in order to update the backing
copy of the updated texture.
This fixes a rendering flickering issue with Photoshop running in
Linux VM with HWversion 11. The problem is Photoshop renders to texture A
in context X, and then bind texture A to context Y. The first time
when texture A is bound to context Y, cso calls pipe->set_sampler_views().
Validation of sampler views is done, rendering is fine.
But when texture A is rendered to again in context X, and rebound in
context Y, cso skips pipe->set_sampler_views() because texture A is already
bound in context Y. SVGA driver is not given a chance to re-validate
the texture binding, the backing copy of the texture is not updated,
and hence causes black image.
Tested with Photoshop, MTT glretrace, piglit.
Fixes VMware bug 1769103.
Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
Juan A. Suarez Romero [Fri, 30 Jun 2017 11:23:57 +0000 (11:23 +0000)]
automake: include git_sha1_gen.sh into EXTRA_DIST
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Nicolai Hähnle [Sat, 10 Jun 2017 18:46:27 +0000 (20:46 +0200)]
Update Khronos-supplied headers
Taken from commit
676834dd529d620ee25090e738d2607dfde003d8
of https://github.com/KhronosGroup/OpenGL-Registry.git
v2:
- keep the BUILDING_MESA bits (Matt)
Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Johnson Lin [Fri, 16 Jun 2017 05:51:34 +0000 (13:51 +0800)]
i965/i915: Add UYVY as the supported format
Trigger the correct sampler options for it. Similar with YUYV
Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com>
Johnson Lin [Fri, 16 Jun 2017 05:40:31 +0000 (13:40 +0800)]
nir: Add a lowering pass for UYVY textures
Similar with support for YUYV but with byte order difference in sampler
Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com>
Johnson Lin [Fri, 16 Jun 2017 05:34:02 +0000 (13:34 +0800)]
dri: Add UYVY as available format
UYVY is diffrent with YUYV in byte order.
YUYV is already declared in dri_interface.h,
this CL add the difinitions for UYVY.
Drivers can add UYVY as supported format
Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com>
Rob Herring [Wed, 28 Jun 2017 21:52:29 +0000 (17:52 -0400)]
gbm: add XBGR8888 support for dumb buffers
Add GBM_FORMAT_XBGR8888 format support which is needed for Android.
Signed-off-by: Rob Herring <robh@kernel.org>
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Rob Herring [Wed, 28 Jun 2017 21:52:28 +0000 (17:52 -0400)]
gallium: os_process fixes for Android
The function getprogname() is available on Android, since it reuses
various BSD solutions C runtime.
Signed-off-by: Rob Herring <robh@kernel.org>
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Tomeu Vizoso [Wed, 28 Jun 2017 21:52:27 +0000 (17:52 -0400)]
etnaviv: Add unreachable statement to etna_amode to fix compilation warnings
Signed-off-by: Robert Foss <robert.foss@collabora.com>
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Bruce Cherniak [Wed, 28 Jun 2017 18:42:03 +0000 (13:42 -0500)]
swr: Remove need to allocate vertex buffer scratch space all in one go
Deferred deletion (via "fence_work") has obsoleted the need to allocate
all client vertex buffer scratch space in a single chunk. Scratch
allocations are now valid until the referenced fence is complete.
Reviewed-by: Tim Rowley <timothy.o.rowley@intel.com>
Bruce Cherniak [Tue, 27 Jun 2017 22:49:06 +0000 (17:49 -0500)]
swr: conditionally validate vertex buffer state
Vertex buffer state doesn't need to be validated on every call,
only on dirty _NEW_VERTEX or indexed draws.
Unconditional validation was introduced as part of patch
330d0607ed6,
"remove pipe_index_buffer and set_index_buffer", with the expectation
we'd optimize later.
Reviewed-by: Tim Rowley <timothy.o.rowley@intel.com>
Tim Rowley [Wed, 28 Jun 2017 14:54:32 +0000 (09:54 -0500)]
swr: set dynamic vertex size
Reduces the memory footprint of the frontend processing by packing
vertices.
Reviewed-by: Bruce Cherniak <bruce.cherniak@intel.com>
Eric Engestrom [Thu, 29 Jun 2017 16:21:44 +0000 (17:21 +0100)]
scons: wait on subprocess' completion
Windows doesn't allow you to move a file that's opened, and Popen()
doesn't wait on its subprocess' completion before returning, which leads
to broken Windows build.
Fixes:
3fd425aed764fb771f2f "build systems: uniformize git_sha1.h generation"
Suggested-by: Scott D Phillips <scott.d.phillips@intel.com>
Signed-off-by: Eric Engestrom <eric.engestrom@imgtec.com>
Eric Engestrom [Tue, 27 Jun 2017 11:08:41 +0000 (12:08 +0100)]
build systems: uniformize git_sha1.h generation
Signed-off-by: Eric Engestrom <eric.engestrom@imgtec.com>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Marek Olšák [Tue, 27 Jun 2017 15:33:54 +0000 (17:33 +0200)]
radeonsi: set COMPUTE_DISPATCH_INITIATOR.ORDER_MODE = 1
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Marek Olšák [Tue, 27 Jun 2017 15:25:28 +0000 (17:25 +0200)]
radeonsi: use the DISPATCH packets to force COMPUTE_START_X/Y/Z = 0
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Rob Herring [Tue, 23 May 2017 00:04:48 +0000 (19:04 -0500)]
Android: use symlinks for driver loading
Instead of having special driver loading logic for Android, create
symlinks to gallium_dri.so so we can use the standard loading logic.
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Signed-off-by: Rob Herring <robh@kernel.org>