OSDN Git Service
Neil Roberts [Tue, 24 Nov 2015 15:15:20 +0000 (16:15 +0100)]
i965/gen9/fast-clear: Handle linear→SRGB conversion
If GL_FRAMEBUFFER_SRGB is enabled when writing to an SRGB-capable
framebuffer then the color will be converted from linear to SRGB
before being written. There is no chance for the hardware to do this
itself because it can't modify the clear color that is programmed in
the surface state so it seems pretty clear that the driver should be
handling this itself.
Note that this wasn't a problem before Gen9 because previously we were
only able to do fast clears to 0 or 1 and those values are the same in
linear and SRGB space.
Reviewed-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
Jordan Justen [Sun, 27 Sep 2015 06:50:55 +0000 (23:50 -0700)]
docs: Add ARB_compute_shader to 11.2.0 release notes
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Sun, 27 Sep 2015 06:49:52 +0000 (23:49 -0700)]
docs: Mark ARB_compute_shader as done for i965
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Wed, 2 Sep 2015 22:47:33 +0000 (15:47 -0700)]
i965: Enable ARB_compute_shader extension on supported hardware
Enable ARB_compute_shader on gen7+, on hardware that supports the
OpenGL 4.3 requirements of a local group size of 1024.
With SIMD16 support, this is limited to Ivy Bridge and Haswell.
Broadwell will work with a local group size up to 896 on SIMD16
meaning programs that use this size or lower should run when setting
MESA_EXTENSION_OVERRIDE=GL_ARB_compute_shader.
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Sat, 10 Oct 2015 20:01:03 +0000 (13:01 -0700)]
i965/nir: Implement shared variable atomic operations
v3:
* Update based on latest SSBO code (Iago)
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Sat, 10 Oct 2015 19:25:39 +0000 (12:25 -0700)]
nir: Add nir intrinsics for shared variable atomic operations
v3:
* Update min/max based on latest SSBO code (Iago)
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Sat, 10 Oct 2015 18:30:33 +0000 (11:30 -0700)]
glsl: Disable several optimizations on shared variables
Shared variables can be accessed by other threads within the same
local workgroup. This prevents us from performing certain
optimizations with shared variables.
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Sat, 10 Oct 2015 16:36:22 +0000 (09:36 -0700)]
glsl: Buffer atomics are supported for compute shaders
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Sat, 10 Oct 2015 14:55:42 +0000 (07:55 -0700)]
glsl: Translate atomic intrinsic functions on shared variables
When an intrinsic atomic operation is used on a shared variable, we
translate it to a new 'shared variable' specific intrinsic function
call.
For example, a call to __intrinsic_atomic_add when used on a shared
variable will be translated to a call to
__intrinsic_atomic_add_shared.
v3:
* Fix stale comments copied from SSBOs (Iago)
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Tue, 17 Nov 2015 18:55:26 +0000 (10:55 -0800)]
glsl: Check for SSBO variable in check_for_ssbo_store
The compiler probably already blocks this earlier on, but we should be
checking for an SSBO here.
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Sat, 10 Oct 2015 00:43:53 +0000 (17:43 -0700)]
glsl: Check for SSBO variable in SSBO atomic lowering
When an atomic function is called, we need to check to see if it is
for an SSBO variable before lowering it to the SSBO specific intrinsic
function.
v2:
* is_in_buffer_block => is_in_shader_storage_block (Iago)
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Fri, 9 Oct 2015 22:55:34 +0000 (15:55 -0700)]
glsl: Replace atomic_ssbo and ssbo_atomic with atomic
The atomic functions can also be used with shared variables in compute
shaders.
When lowering the intrinsic in lower_ubo_reference, we still create an
SSBO specific intrinsic since SSBO accesses can be indirectly
addressed, whereas all compute shader shared variable live in a single
shared variable area.
v2:
* Also remove the _internal suffix from ssbo atomic intrinsic names (Iago)
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Fri, 9 Oct 2015 22:32:33 +0000 (15:32 -0700)]
glsl: Allow atomic functions to be used with shared variables
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Timothy Arceri <timothy.arceri@collabora.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Tue, 28 Jul 2015 21:55:00 +0000 (14:55 -0700)]
i965: Lower shared variable references to intrinsic calls
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Tue, 28 Jul 2015 22:30:30 +0000 (15:30 -0700)]
i965: Enable shared local memory for CS shared variables
v3:
* Check shared variable size at link time
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Tue, 28 Jul 2015 22:25:46 +0000 (15:25 -0700)]
i965/fs: Handle nir shared variable store intrinsic
v4:
* Apply similar optimization for shared variable stores as
0cb7d7b4b7c32246d4c4225a1d17d7ff79a7526d. This was causing a
OpenGLES 3.1 CTS failure, but
867c436ca841b4196b4dde4786f5086c76b20dd7 fixes that.
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Tue, 28 Jul 2015 22:24:13 +0000 (15:24 -0700)]
i965/fs: Handle nir shared variable load intrinsic
v3:
* Remove extra #includes (Iago)
* Use recently added GEN7_BTI_SLM instead of BRW_SLM_SURFACE_INDEX (curro)
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Fri, 16 Oct 2015 17:45:05 +0000 (10:45 -0700)]
i965: Disable vector splitting on shared variables
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Tue, 28 Jul 2015 22:17:34 +0000 (15:17 -0700)]
nir: Translate glsl shared var store intrinsic to nir intrinsic
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Tue, 28 Jul 2015 22:11:37 +0000 (15:11 -0700)]
nir: Translate glsl shared var load intrinsic to nir intrinsic
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Tue, 28 Jul 2015 21:45:56 +0000 (14:45 -0700)]
glsl: Add lowering pass for shared variable references
In this lowering pass, shared variables are decomposed into intrinsic
calls.
v2:
* Send mem_ctx as a parameter (Iago)
v3:
* Shared variables don't have an associated interface block (Iago)
* Always use 430 packing (Iago)
* Comment / whitespace cleanup (Iago)
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Iago Toral Quiroga [Tue, 1 Dec 2015 00:43:09 +0000 (16:43 -0800)]
glsl: Don't assert on shared variable matrices with 'inherited' layout
We use column-major for shared variable matrices.
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Tue, 21 Jul 2015 21:04:11 +0000 (14:04 -0700)]
glsl: Don't lower_variable_index_to_cond_assign for shared variables
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Tue, 17 Nov 2015 02:10:32 +0000 (18:10 -0800)]
glsl: Remove mem_ctx as member variable in lower_ubo_reference_visitor
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Tue, 17 Nov 2015 02:09:27 +0000 (18:09 -0800)]
glsl ubo/ssbo: Move common code into lower_buffer_access::setup_buffer_access
This code will also be usable by the pass to lower shared variables.
Note, that *const_offset is adjusted by setup_buffer_access so it must
be initialized before calling setup_buffer_access.
v2:
* Add comment for lower_buffer_access::setup_buffer_access
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Fri, 13 Nov 2015 20:08:26 +0000 (12:08 -0800)]
glsl ubo/ssbo: Move is_dereferenced_thing_row_major into lower_buffer_access
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Fri, 13 Nov 2015 07:43:04 +0000 (23:43 -0800)]
glsl ubo/ssbo: Add lower_buffer_access class
This class has code that will be shared by lower_ubo_reference and
lower_shared_reference. (lower_shared_reference will be used to
support compute shader shared variables.)
v2:
* Add lower_buffer_access.h to makefile (Emil)
* Remove static is_dereferenced_thing_row_major from
lower_buffer_access.cpp. This will become a lower_buffer_access
method in the next commit.
* Pass mem_ctx as parameter rather than using a member variable (Iago)
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Fri, 13 Nov 2015 18:47:06 +0000 (10:47 -0800)]
glsl ubo/ssbo: Split buffer access to insert_buffer_access
This allows the code in emit_access to be generic enough to also be
for lowering shared variables.
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Fri, 13 Nov 2015 09:49:43 +0000 (01:49 -0800)]
glsl ubo/ssbo: Use enum to track current buffer access type
v2:
* Rename ssbo_get_array_length to ssbo_unsized_array_length_access (Iago)
* Use always use this-> when referencing buffer_access_type (Iago)
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Tapani Pälli [Wed, 9 Dec 2015 07:48:57 +0000 (09:48 +0200)]
glsl: do not loose always_active_io when packing varyings
Otherwise packed and inactive varyings get optimized away. This needs
to be prevented when using separate shader objects where interface
needs to be preserved.
Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Timothy Arceri <timothy.arceri@collabora.com>
Tapani Pälli [Tue, 8 Dec 2015 17:02:14 +0000 (19:02 +0200)]
mesa: invalidate pipeline status after glUseProgramStages
This will cause validation to run during next draw, this is done
because possible changes in used stages and programs can cause
invalid pipeline state.
This fixes a subtest in following CTS test:
ES31-CTS.sepshaderobjs.StateInteraction
Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Timothy Arceri <timothy.arceri@collabora.com>
Dave Airlie [Thu, 10 Dec 2015 00:36:42 +0000 (10:36 +1000)]
mesa/varray: set double arrays to non-normalised.
Doesn't have any effect in practice I don't think, but
CTS reads back using GetVertexAttrib.
This fixes: GL41-CTS.vertex_attrib_64bit.get_vertex_attrib
Cc: "11.0 11.1" <mesa-stable@lists.freedesktop.org>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Michel Dänzer [Wed, 9 Dec 2015 06:46:46 +0000 (15:46 +0900)]
clover: Fix build against LLVM 3.8 SVN >= r255078
Reviewed-by: Francisco Jerez <currojerez@riseup.net>
Brian Paul [Wed, 9 Dec 2015 23:00:55 +0000 (16:00 -0700)]
mesa: fix ID usage for buffer warnings
We need a different ID pointer for each call site.
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Brian Paul [Wed, 9 Dec 2015 22:14:02 +0000 (15:14 -0700)]
docs: remove stray <ul> tag from 11.0.5.html file to fix indentation
Serge Martin [Sun, 6 Dec 2015 14:32:17 +0000 (15:32 +0100)]
freedreno: little clean up in fd_create_surface
in order to avoid returing invalid adress if CALLOC_STRUCT return NULL.
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Serge Martin [Sun, 6 Dec 2015 14:32:16 +0000 (15:32 +0100)]
freedreno: change to goto fail
in fd_resource_transfer_map, like the others error cases
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Serge Martin [Sun, 6 Dec 2015 14:32:15 +0000 (15:32 +0100)]
freedreno: fix bind_sampler_states when hwcso is NULL
src/gallium/tests/trivial/compute.c expects samplers to be cleaned
when the samplers list is NULL.
Like in radeon, the function behave like when the number of samplers
parameter is set to 0.
[small s/hwsco/hwcso/ typo fix]
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Edward O'Callaghan [Wed, 9 Dec 2015 09:07:57 +0000 (20:07 +1100)]
gallium/util: Make u_prims_for_vertices() safe
Let us avoid trapping in hardware from a SIGFPE and instead
assert on a zero divisor.
Hint: This can occur if a PIPE_PRIM_? is not handled in
u_prim_vertex_count() that results in ' info ' not
being initialized in the expected manner.
Further, we also fix a possibly NULL pointer dereference
from ' info ' being NULL from a u_prim_vertex_count() call.
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Signed-off-by: Marek Olšák <marek.olsak@amd.com>
Andreas Boll [Wed, 9 Dec 2015 21:44:52 +0000 (22:44 +0100)]
docs: add news item for mesa-demos 8.3.0 release
Signed-off-by: Andreas Boll <andreas.boll.dev@gmail.com>
Patrick Rudolph [Sun, 6 Dec 2015 09:11:59 +0000 (10:11 +0100)]
nv50,nvc0: fix use-after-free when vertex buffers are unbound
Always reset the vertex bufctx to make sure there's no pointer to
an already freed pipe_resource left after unbinding buffers.
Fixes use after free crash in nvc0_bufctx_fence().
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=93004
Signed-off-by: Patrick Rudolph <siro@das-labor.org>
[imirkin: simplify nvc0 fix, apply to nv50]
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Cc: "11.0 11.1" <mesa-stable@lists.freedesktop.org>
Andreas Boll [Wed, 9 Dec 2015 16:25:40 +0000 (17:25 +0100)]
mesa: Fix a typo in a comment
s/suports/supports/
Signed-off-by: Andreas Boll <andreas.boll.dev@gmail.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Andreas Boll [Wed, 9 Dec 2015 16:25:15 +0000 (17:25 +0100)]
glx: Fix a typo in a comment
s/suports/supports/
Signed-off-by: Andreas Boll <andreas.boll.dev@gmail.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Andreas Boll [Wed, 9 Dec 2015 16:12:45 +0000 (17:12 +0100)]
st/osmesa: Fix a typo in a comment
s/suport/support/
Signed-off-by: Andreas Boll <andreas.boll.dev@gmail.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Andreas Boll [Wed, 9 Dec 2015 16:11:29 +0000 (17:11 +0100)]
meta: Fix a typo in a print message
s/Unkown/Unknown/
Signed-off-by: Andreas Boll <andreas.boll.dev@gmail.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Andreas Boll [Wed, 9 Dec 2015 16:10:33 +0000 (17:10 +0100)]
mesa: Fix typos in print messages
s/inconsistant/inconsistent/
s/occurences/occurrences/
Signed-off-by: Andreas Boll <andreas.boll.dev@gmail.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Andreas Boll [Wed, 9 Dec 2015 16:09:25 +0000 (17:09 +0100)]
glsl: Fix a typo in a comment
s/suports/supports/
Signed-off-by: Andreas Boll <andreas.boll.dev@gmail.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Brian Paul [Tue, 8 Dec 2015 16:30:32 +0000 (09:30 -0700)]
svga: initialize pipe_driver_query_info entries with a macro
To be safe, set all the fields in case the enums ordering/values
ever change.
Reviewed-by: Charmaine Lee <charmainel@vmware.com>
Brian Paul [Tue, 8 Dec 2015 01:38:03 +0000 (18:38 -0700)]
mesa: detect inefficient buffer use and report through debug output
When a buffer is created with GL_STATIC_DRAW, its contents should not
be changed frequently. But that's exactly what one application I'm
debugging does. This patch adds code to try to detect inefficient
buffer use in a couple places. The GL_ARB_debug_output mechanism is
used to report the issue.
NVIDIA's driver detects these sort of things too.
Other types of inefficient buffer use could also be detected in the
future.
Reviewed-by: José Fonseca <jfonseca@vmware.com>
Emil Velikov [Wed, 9 Dec 2015 16:12:32 +0000 (16:12 +0000)]
docs: add news item and link release notes for 11.0.7
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Emil Velikov [Wed, 9 Dec 2015 16:09:37 +0000 (16:09 +0000)]
docs: add sha256 checksums for 11.0.7
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
(cherry picked from commit
f9715bc449d6b2cc5693e44bb82a9d5305a32ab5)
Emil Velikov [Wed, 9 Dec 2015 15:19:30 +0000 (15:19 +0000)]
docs: add release notes for 11.0.7
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
(cherry picked from commit
bec983b738a7f149478ee692ba0e1d26fcc9fd8e)
Francisco Jerez [Fri, 4 Sep 2015 12:30:41 +0000 (15:30 +0300)]
i965: Resolve color and flush for all active shader images in intel_update_state().
Fixes arb_shader_image_load_store/execution/load-from-cleared-image.shader_test.
Couldn't reproduce any significant FPS regression in CPU-bound
benchmarks from the Finnish benchmarking system on neither VLV nor BSW
after 30 runs with 95% confidence level.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=92849
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Jason Ekstrand <jason.ekstrand@intel.com>
Cc: "11.0 11.1" <mesa-stable@lists.freedesktop.org>
Tested-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Francisco Jerez [Wed, 25 Nov 2015 12:51:16 +0000 (14:51 +0200)]
i965: Document inconsistent units the URB size is represented in.
Every other gen the representation of the URB size was changed and
previous ones weren't updated. I'd be willing to write a series
normalizing this to be KB on all generations if anybody else cares.
Francisco Jerez [Tue, 8 Dec 2015 16:53:57 +0000 (18:53 +0200)]
i965: Hook up L3 partitioning state atom.
Reviewed-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Francisco Jerez [Thu, 3 Sep 2015 15:23:19 +0000 (18:23 +0300)]
i965: Work around L3 state leaks during context switches.
This is going to require some rather intrusive kernel changes to fix
properly, in the meantime (and forever on at least pre-v4.1 kernels)
we'll have to restore the hardware defaults at the end of every batch
in which the L3 configuration was changed to avoid interfering with
the DDX and GL clients that use an older non-L3-aware version of Mesa.
Reviewed-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
v2: Optimize look-up of the default configuration by assuming it's the
first entry of the L3 config array in order to avoid an FPS
regression in GpuTest Triangle and SynMark OglBatch2-7 on most
affected platforms.
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Francisco Jerez [Sat, 5 Sep 2015 12:05:48 +0000 (15:05 +0300)]
i965: Add debug flag to print out the new L3 state during transitions.
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Francisco Jerez [Tue, 8 Dec 2015 16:53:22 +0000 (18:53 +0200)]
i965: Implement L3 state atom.
The L3 state atom calculates the target L3 partition weights when the
program bound to some shader stage is modified, and in case they are
far enough from the current partitioning it makes sure that the L3
state is re-emitted.
v2: Fix for inconsistent units the context URB size is expressed in.
Clamp URB size to 1008 KB on SKL due to FF hardware limitation.
Reviewed-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Acked-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Francisco Jerez [Thu, 3 Sep 2015 15:15:19 +0000 (18:15 +0300)]
i965: Calculate appropriate L3 partition weights for the current pipeline state.
This calculates a rather conservative partitioning of the L3 cache
based on the shaders currently bound to the pipeline and whether they
use SLM, atomics, images or scratch space. The result is intended to
be fine-tuned later on based on other pipeline state.
Note that the L3 partitioning calculated for VLV in the non-SLM non-DC
case differs from the hardware defaults in that it doesn't include a
DC partition and has twice as much RO cache space -- This is an
intentional functional change that improves performance in several
bandwidth-bound benchmarks on VLV (5% significance): SynMark
OglTexFilterAniso by 14.18%, SynMark OglTexFilterTri by 7.15%, Unigine
Heaven by 4.91%, SynMark OglShMapPcf by 2.15%, GpuTest Fur by 1.83%,
SynMark OglDrvRes by 1.80%, SynMark OglVsTangent by 1.71%, and a few
other benchmarks from the Finnish system by less than 1%.
Reviewed-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Francisco Jerez [Thu, 3 Sep 2015 15:10:24 +0000 (18:10 +0300)]
i965: Implement selection of the closest L3 configuration based on a vector of weights.
The input of the L3 set-up code is a vector giving the approximate
desired relative size of each partition. This implements logic to
compare the input vector against the table of validated configurations
for the device and pick the closest compatible one.
Reviewed-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Francisco Jerez [Thu, 26 Nov 2015 14:42:43 +0000 (16:42 +0200)]
i965: Define and use REG_MASK macro to make masked MMIO writes slightly more readable.
Reviewed-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Francisco Jerez [Thu, 3 Sep 2015 15:08:59 +0000 (18:08 +0300)]
i965/hsw: Enable L3 atomics.
Improves performance of the arb_shader_image_load_store-atomicity
piglit test by over 25x (which isn't a real benchmark it's just heavy
on atomics -- the improvement in a microbenchmark I wrote a while ago
seemed to be even greater). The drawback is one needs to be
extra-careful not to hang the GPU (in fact the whole system). A DC
partition must have been allocated on L3, the "convert L3 cycle for DC
to UC" bit may not be set, the MOCS L3 cacheability bit must be set
for all surfaces accessed using DC atomics, and the SCRATCH1 and
ROW_CHICKEN3 bits must be kept in sync.
A fairly recent kernel is required for the command parser to allow
writes to these registers.
Reviewed-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Acked-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Francisco Jerez [Thu, 3 Sep 2015 15:07:44 +0000 (18:07 +0300)]
i965: Implement programming of the L3 configuration.
Reviewed-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Acked-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Francisco Jerez [Tue, 8 Dec 2015 16:52:56 +0000 (18:52 +0200)]
i965: Import tables enumerating the set of validated L3 configurations.
It should be possible to use additional L3 configurations other than
the ones listed in the tables of validated allocations ("BSpec »
3D-Media-GPGPU Engine » L3 Cache and URB [IVB+] » L3 Cache and URB [*]
» L3 Allocation and Programming"), but it seems sensible for now to
hard-code the tables in order to stick to the hardware docs. Instead
of setting up the arbitrary L3 partitioning given as input, the
closest validated L3 configuration will be looked up in these tables
and used to program the hardware.
The included tables should work for Gen7-9. Note that the quantities
are specified in ways rather than in KB, this is because the L3
control registers expect the value in ways, and because by doing that
we can re-use a single table for all GT variants of the same
generation (and in the case of IVB/HSW and CHV/SKL across different
generations) which generally have different L3 way sizes but allow the
same combinations of way allocations.
v2: Use slice count from the devinfo structure instead of the gt
number to implement get_l3_way_size().
Reviewed-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Acked-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Francisco Jerez [Wed, 25 Nov 2015 13:15:20 +0000 (15:15 +0200)]
i965: Add slice count to the brw_device_info structure.
Reviewed-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Francisco Jerez [Thu, 3 Sep 2015 14:19:10 +0000 (17:19 +0300)]
i965/gen8: Don't add workaround bits to PIPE_CONTROL stalls if DC flush is set.
According to the hardware docs a DC flush is sufficient to make
CS_STALL happy, there's no need to add STALL_AT_SCOREBOARD whenever
it's present.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Francisco Jerez [Thu, 3 Sep 2015 14:15:57 +0000 (17:15 +0300)]
i965: Define state flag to signal that the URB size has been altered.
This will make sure that we recalculate the URB layout anytime the URB
size is modified by the L3 partitioning code.
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Francisco Jerez [Thu, 3 Sep 2015 14:08:16 +0000 (17:08 +0300)]
i965: Keep track of whether LRI is allowed in the context struct.
This stores the result of can_do_pipelined_register_writes() in the
context struct so we can find out later whether LRI can be used to
program the L3 configuration.
v2:
* Split change of gen check in can_do_pipelined_register_writes (jljusten)
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Francisco Jerez [Thu, 3 Sep 2015 14:08:16 +0000 (17:08 +0300)]
i965: Adjust gen check in can_do_pipelined_register_writes
Allow for pipelined register writes for gen < 7.
v2:
* Split from another patch and adjust comment (jljusten)
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Francisco Jerez [Thu, 3 Sep 2015 14:02:48 +0000 (17:02 +0300)]
i965: Define symbolic constants for some useful L3 cache control registers.
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Dave Airlie [Wed, 9 Dec 2015 07:02:38 +0000 (17:02 +1000)]
radeonsi: handle loading doubles as geometry shader inputs.
This adds the double code to the geometry shader input handling.
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Cc: "11.0 11.1" <mesa-stable@lists.freedesktop.org>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Dave Airlie [Wed, 9 Dec 2015 07:02:01 +0000 (17:02 +1000)]
radeonsi: handle doubles in lds load path.
This handles loading doubles from LDS properly.
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Cc: "11.0 11.1" <mesa-stable@lists.fedoraproject.org>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Dave Airlie [Mon, 7 Dec 2015 05:38:00 +0000 (05:38 +0000)]
r600: handle geometry dynamic input array index
This fixes:
glsl-1.50/execution/geometry/dynamic_input_array_index.shader_test
my profanity.
We need to load the AR register with the value from the index reg
Cc: "11.0 11.1" <mesa-stable@lists.freedesktop.org>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Dave Airlie [Mon, 7 Dec 2015 03:49:59 +0000 (03:49 +0000)]
r600g: fix geom shader input indirect indexing.
This fixes:
gs-input-array-vec4-index-rd
The others run out of gprs unfortunately.
Cc: "11.0 11.1" <mesa-stable@lists.freedesktop.org>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Dave Airlie [Mon, 7 Dec 2015 01:36:21 +0000 (01:36 +0000)]
r600g: fix outputing to non-0 buffers for stream 0.
This fixes:
arb_transform_feedback3-ext_interleaved_two_bufs_gs
arb_transform_feedback3-ext_interleaved_two_bufs_gs_max
transform-feedback-builtins
If we are only emitting one ring, then emit all output
buffers on it.
Cc: "11.0 11.1" <mesa-stable@lists.freedesktop.org>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Edward O'Callaghan [Mon, 7 Dec 2015 09:35:55 +0000 (20:35 +1100)]
r600: Add ARB_copy_image support
[airlied: update relnotes]
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Edward O'Callaghan [Mon, 7 Dec 2015 13:36:26 +0000 (00:36 +1100)]
r600g: allow copying between compatible un/compressed formats
See: `commit
e82c527f1fc2f8ddc64954ecd06b0de3cea92e93`
which is where a block in src maps to a pixel in dst and vice versa.
e.g. DXT1 <-> R32G32_UINT
DXT5 <-> R32G32B32A32_UINT
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Ilia Mirkin [Wed, 9 Dec 2015 02:01:15 +0000 (21:01 -0500)]
nv50/ir: fix cutoff for using r63 vs r127 when replacing zero
The only effect here is a space savings - 822 programs in shader-db
affected with the following overall change:
total bytes used in shared programs :
44154976 ->
44139880 (-0.03%)
Fixes:
641eda0c (nv50/ir: r63 is only 0 if we are using less than 63 registers)
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Cc: "11.0 11.1" <mesa-stable@lists.freedesktop.org>
Ilia Mirkin [Tue, 8 Dec 2015 23:46:34 +0000 (18:46 -0500)]
nv50/ir: prefer to color mad def and src2 with the same color
This allows us to use the short encoding, and potentially fold
immediates in later on.
total instructions in shared programs :
6379731 ->
6367861 (-0.19%)
total gprs used in shared programs : 728502 -> 728683 (0.02%)
total local used in shared programs : 9904 -> 9904 (0.00%)
total bytes used in shared programs :
44661008 ->
44154976 (-1.13%)
local gpr inst bytes
helped 0 51 7267 20306
hurt 0 232 125 274
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Ilia Mirkin [Tue, 8 Dec 2015 23:21:26 +0000 (18:21 -0500)]
nv50/ir: reduce degree limit on ops that can't encode large reg dests
Operations that take immediates can only encode registers up to 64. This
fixes a shader in a "Powered by Unity" intro.
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Ilia Mirkin [Mon, 7 Dec 2015 04:25:54 +0000 (23:25 -0500)]
nv50/ir: only unspill once ahead of a group of instructions
We already semi-did this but the list of uses as unsorted, so it was
unreliable. Sort the uses by bb and serial, and don't unspill for each
instruction in a sequence. (And also don't unspill multiple times for a
single instruction that uses the value in question multiple times.)
This causes a minor reduction in generated instructions for shader-db
(as few programs spill) but more importantly it brings determinism to
each run's output.
On SM10:
total instructions in shared programs :
6387945 ->
6379359 (-0.13%)
total gprs used in shared programs : 728544 -> 728544 (0.00%)
total local used in shared programs : 9904 -> 9904 (0.00%)
local gpr inst bytes
helped 0 0 322 322
hurt 0 0 0 0
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Ilia Mirkin [Tue, 8 Dec 2015 21:48:06 +0000 (16:48 -0500)]
nv50/ir: check if the target supports the new offset before inlining
Fixes:
abd326e81b (nv50/ir: propagate indirect loads into instructions)
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=93300
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Dave Airlie [Mon, 12 Oct 2015 05:35:40 +0000 (15:35 +1000)]
llvmpipe: fix fp64 inputs to geom shader.
This fixes the fetching of fp64 inputs to the geometry shader,
this fixes the recently posted piglit's
arb_gpu_shader_fp64/execution/gs-fs-vs-double-array.shader_test
arb_vertex_attrib_64bit/execution/gs-fs-vs-attrib-double-array.shader_test
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Matt Turner [Wed, 2 Dec 2015 00:13:11 +0000 (16:13 -0800)]
nir: Optimize useless comparisons against true/false.
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com> [v1]
Reviewed-by: Eric Anholt <eric@anholt.net> [v1]
v2: Move new rule to Boolean simplification section
Add a a@bool != true simplification
Suggested-by: Neil Roberts <neil@linux.intel.com>
Matt Turner [Mon, 30 Nov 2015 18:24:53 +0000 (10:24 -0800)]
glsl: Switch opcode and avail parameters to binop().
To make it match unop().
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Matt Turner [Wed, 2 Dec 2015 01:09:55 +0000 (17:09 -0800)]
glsl_to_tgsi: Skip useless comparison instructions.
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Matt Turner [Mon, 7 Dec 2015 22:18:38 +0000 (14:18 -0800)]
glsl: Relax qualifier ordering restriction in ES 3.1.
... and allow the "binding" qualifier in ES 3.1 as well.
GLSL ES 3.1 incorporates only a few features from the extension
ARB_shading_language_420pack: the relaxed qualifier ordering
requirements and the binding qualifier.
Cc: "11.1" <mesa-stable@lists.freedesktop.org>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Matt Turner [Mon, 7 Dec 2015 22:11:01 +0000 (14:11 -0800)]
glsl: Use has_420pack().
These features would not have been enabled with #version 420 otherwise.
Cc: "11.1" <mesa-stable@lists.freedesktop.org>
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Matt Turner [Mon, 7 Dec 2015 19:14:56 +0000 (11:14 -0800)]
glsl: Allow binding of image variables with 420pack.
This interaction was missed in the addition of ARB_image_load_store.
Cc: "11.0 11.1" <mesa-stable@lists.freedesktop.org>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=93266
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Jose Fonseca [Tue, 8 Dec 2015 22:26:32 +0000 (22:26 +0000)]
appveyor: Cache winflexbison archive.
Unforunately the Appveyor -> SourceForge connection seems a bit
unreliable, causing frequent build failures while downloading
winflexbison (approx once every 2 days).
Fetching winflexbison archive into Appveyor's cache should eliminate
these.
Fetching Python modules from PyPI doesn't seem to be a problem, so they
are left alone for now, though they could eventually get the same
treatment.
Eric Anholt [Sat, 20 Jun 2015 22:21:40 +0000 (15:21 -0700)]
vc4: Enable MSAA.
We still have several failures in the newly enabled tests in simulation:
sRGB downsampling is done as if it was just linear, stencil blits are not
supported on MSAA either, and derivatives are still not supported
(breaking some MSAA simulation shaders). So, other than sRGB downsampling
quality, things seem to be in good shape.
Eric Anholt [Sun, 22 Nov 2015 22:31:51 +0000 (14:31 -0800)]
vc4: Add support for mapping of MSAA resources.
The pipe_transfer_map API requires that we do an implicit
downsample/upsample and return a mapping of that.
Eric Anholt [Tue, 23 Jun 2015 16:50:36 +0000 (09:50 -0700)]
vc4: Add support for texel fetches from MSAA resources.
This is the core of ARB_texture_multisample. Most of the piglit tests for
GL_ARB_texture_multisample require GL 3.0, but exposing support for this
lets us use the gallium blitter for multisample resolves. We can
sometimes multisample resolve using just the RCL, but that requires that
the blit is 1:1, unflipped, and aligned to tile boundaries.
Eric Anholt [Tue, 23 Jun 2015 20:11:55 +0000 (13:11 -0700)]
vc4: Add support for multisample framebuffer operations.
This includes GL_SAMPLE_COVERAGE, GL_SAMPLE_ALPHA_TO_ONE, and
GL_SAMPLE_ALPHA_TO_COVAGE.
I haven't implemented a dithering function yet, and gallium doesn't give
me a good chance to do so for GL_SAMPLE_COVERAGE.
Eric Anholt [Sat, 21 Nov 2015 04:25:46 +0000 (20:25 -0800)]
vc4: Add a workaround for HW-2905, and additional failure I saw with MSAA.
I only stumbled on this while experimenting due to reading about HW-2905.
I don't know if the EZ disable in the Z-clear is actually necessary, but
go with it for now.
Eric Anholt [Mon, 22 Jun 2015 20:12:27 +0000 (13:12 -0700)]
vc4: Add support for drawing in MSAA.
Eric Anholt [Sat, 20 Jun 2015 22:30:04 +0000 (15:30 -0700)]
vc4: Add kernel RCL support for MSAA rendering.
Eric Anholt [Mon, 22 Jun 2015 18:37:32 +0000 (11:37 -0700)]
vc4: Rename color_ms_write to color_write.
I was thinking this was the only MSAA resolve thing, so it should be noted
separately, but actually load/store general also do MSAA resolve.
Eric Anholt [Mon, 7 Dec 2015 23:18:59 +0000 (15:18 -0800)]
vc4: Allow RCL blits to the edge of the surface.
The recent unaligned fix successfully prevented RCL blits that weren't
aligned inside of the surface, but we also want to be able to do RCL blits
for the whole surface when the width or height of the surface aren't
aligned (we don't care what renders inside of the padding).
Eric Anholt [Tue, 8 Dec 2015 04:27:20 +0000 (20:27 -0800)]
vc4: Add disabled debug printf for describing blits.
I keep typing variants of this while debugging RCL blits for MSAA.