Commit graph

226 commits

Author SHA1 Message Date
Marek Olšák
802fcdc0d2 radeonsi: atomize L2 prefetches
to move the big conditional statement out of draw_vbo

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2017-01-30 13:27:14 +01:00
Marek Olšák
c99ba3eb47 radeonsi: unbind disabled shader stages to prevent useless L2 prefetches
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2017-01-30 13:27:14 +01:00
Marek Olšák
35cd7551a4 radeonsi: use the correct target machine when building shader variants
If the shader selector is created with a different context than
the shader variant, we should use the calling context's target machine
for the shader variant.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=99419

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2017-01-18 19:51:31 +01:00
Marek Olšák
3ae3be6dd4 radeonsi: move shader pipe context state into a separate structure
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2017-01-18 19:51:31 +01:00
Nicolai Hähnle
5e94e5bb9b radeonsi: fix R600_DEBUG=nooptvariant
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Vedran Miletić <vedran@miletic.net>
2017-01-16 20:16:18 +01:00
Marek Olšák
44e9b67229 radeonsi: make fix_fetch 64-bit
v2: add u_bit_consecutive64

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2017-01-16 18:07:08 +01:00
Marek Olšák
6f356d15be radeonsi: cleanly communicate whether si_shader_dump should check R600_DEBUG
Tested-by: Edmondo Tommasina <edmondo.tommasina@gmail.com>
Acked-by: Edward O'Callaghan <funfunctor@folklore1984.net>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2017-01-09 12:01:30 +01:00
Marek Olšák
4b93ba542c radeonsi: assume that a TES without POSITION precedes GS
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2017-01-06 21:05:48 +01:00
Marek Olšák
3753dc896d radeonsi: update clip_regs if clip_disable changes to fix a hang
This seems to fix the GPU hangs caused by:

commit ed3190b3f3
Author: Marek Olšák <marek.olsak@amd.com>
Date:   Sun Nov 13 18:41:43 2016 +0100

    radeonsi: don't export ClipVertex and ClipDistance[] if clipping is disabled

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=99219

Tested-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
2017-01-05 14:01:18 +01:00
Nicolai Hähnle
ec0a0a60cc radeonsi: shrink the GSVS ring to account for the reduced item sizes
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
2016-12-12 09:05:17 +01:00
Nicolai Hähnle
6fdef7d265 radeonsi: shrink each vertex stream to the actually required size
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
2016-12-12 09:05:13 +01:00
Nicolai Hähnle
2f2e941e2d radeonsi: use a single descriptor for the GSVS ring
We can hardcode all of the fields for swizzling in the geometry shader.

The advantage is that we use fewer descriptor slots and we no longer have to
update any of the (ring) descriptors when the geometry shader changes.

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
2016-12-12 09:05:05 +01:00
Nicolai Hähnle
7b5b3d63c5 radeonsi: update all GSVS ring descriptors for new buffer allocations
Fixes GL45-CTS.gtf40.GL3Tests.transform_feedback3.transform_feedback3_geometry_instanced.

Cc: mesa-stable@lists.freedesktop.org
Reviewed-by: Edward O'Callaghan <funfunctor@folklore1984.net>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
2016-12-12 09:04:06 +01:00
Marek Olšák
e5302ad936 radeonsi: add a debug flag that disables optimized shader variants
Reviewed-by: Edward O'Callaghan <funfunctor@folklore1984.net>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2016-11-23 18:49:10 +01:00
Marek Olšák
86514d84e0 util: import CRC32 implementation from gallium
Reviewed-by: Timothy Arceri <timothy.arceri@collabora.com>
2016-11-22 18:05:51 +01:00
Marek Olšák
bf75ef3f92 radeonsi: remove all varyings for depth-only rendering or rasterization off
Tested-by: Edmondo Tommasina <edmondo.tommasina@gmail.com>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2016-11-21 21:44:35 +01:00
Marek Olšák
ef6c84b301 radeonsi: eliminate VS outputs that aren't used by PS at runtime
A past commit added the ability to compile "optimized" shader variants
asynchronously (not stalling the app).

This commit builds upon that and adds what is basically a runtime shader
linker. If a VS output isn't used by the currently-bound PS, a new VS
compilation is started without that output. The new shader variant
is used when it's ready.

All apps using separate shader objects I've seen had unused VS outputs.

Eliminating unused/useless VS outputs also eliminates the corresponding
vertex attribute loads.

Tested-by: Edmondo Tommasina <edmondo.tommasina@gmail.com>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2016-11-21 21:44:35 +01:00
Marek Olšák
7e76f9a7a8 radeonsi: record information about all written and read varyings
It's just tgsi_shader_info with DEFAULT_VAL varyings removed.

Tested-by: Edmondo Tommasina <edmondo.tommasina@gmail.com>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2016-11-21 21:44:35 +01:00
Marek Olšák
c7f3e5c647 radeonsi: make si_shader_io_get_unique_index stricter
Tested-by: Edmondo Tommasina <edmondo.tommasina@gmail.com>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2016-11-21 21:44:35 +01:00
Marek Olšák
ed3190b3f3 radeonsi: don't export ClipVertex and ClipDistance[] if clipping is disabled
This is the first user of optimized monolithic shader variants.

Cull distances can't be disabled by states.

Tested-by: Edmondo Tommasina <edmondo.tommasina@gmail.com>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2016-11-21 21:44:35 +01:00
Marek Olšák
d984a324bf radeonsi: add infrastr. for compiling optimized shader variants asynchronously
Tested-by: Edmondo Tommasina <edmondo.tommasina@gmail.com>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2016-11-21 21:44:35 +01:00
Marek Olšák
d2a56985d7 radeonsi: don't set vs.epilog.export_prim_id if TES is bound
there is no VS epilog in this case

Tested-by: Edmondo Tommasina <edmondo.tommasina@gmail.com>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2016-11-21 21:44:35 +01:00
Marek Olšák
fee71fec25 radeonsi: simplify checking for monolithic compilation
Tested-by: Edmondo Tommasina <edmondo.tommasina@gmail.com>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2016-11-21 21:44:35 +01:00
Marek Olšák
6d5c2a8b5c radeonsi: split the shader key into 3 logical parts
key->part.*: prolog and epilog flags only
key->as_{ls,es}: special flags
key->mono.*: flags for monolithic compilation only

Tested-by: Edmondo Tommasina <edmondo.tommasina@gmail.com>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2016-11-21 21:44:35 +01:00
Marek Olšák
e59389d738 radeonsi: assume that a VS without POSITION is LS
Tested-by: Edmondo Tommasina <edmondo.tommasina@gmail.com>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2016-11-21 21:44:35 +01:00
Nicolai Hähnle
2c875158e2 radeonsi: fix vertex fetches for 2_10_10_10 formats
The hardware always treats the alpha channel as unsigned, so add a shader
workaround. This is rare enough that we'll just build a monolithic vertex
shader.

The SINT case cannot actually happen in OpenGL, but I've included it for
completeness since it's just a mix of the other cases.

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
2016-11-04 21:30:18 +01:00
Nicolai Hähnle
908f92ad1f radeonsi: generate GS prolog to (partially) fix triangle strip adjacency rotation
Fixes GL45-CTS.geometry_shader.adjacency.adjacency_indiced_triangle_strip and
others.

This leaves the case of triangle strips with adjacency and primitive restarts
open. It seems that the only thing that cares about that is a piglit test.
Fixing this efficiently would be really involved, and I don't want to use the
hammer of degrading to software handling of indices because there may well
be software that uses this draw mode (without caring about the precise
rotation of triangles).

v2:
- skip the GS prolog entirely if workaround is not needed
- only check for TES (TES is always non-null when tessellation is used)

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
2016-11-03 10:11:24 +01:00
Nicolai Hähnle
3b2516721b radeonsi: make the GS copy shader owned by the GS selector
The copy shader only depends on the selector. This change avoids creating
separate code paths for monolithic vs. non-monolithic geometry shaders.

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
2016-11-03 10:07:50 +01:00
Nicolai Hähnle
9c6f7d66dc radeonsi: si_shader_vs only depends on the GS selector
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
2016-11-03 10:07:48 +01:00
Nicolai Hähnle
693435d846 radeonsi: si_vgt_gs_mode only depends on the selector
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
2016-11-03 10:07:45 +01:00
Marek Olšák
d268b7f95e radeonsi: add a driver query for shader cache hits
This is an 8-month old patch.

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2016-11-01 22:33:13 +01:00
Marek Olšák
ad45dce4a2 radeonsi: remove si_resource_create_custom
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2016-10-26 13:02:58 +02:00
Marek Olšák
29144d0f34 gallium/radeon: stop using PIPE_BIND_CUSTOM
it has no effect whatsoever

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2016-10-26 13:02:58 +02:00
Marek Olšák
3ec9975555 radeonsi: eliminate trivial constant VS outputs
These constant value VS PARAM exports:
- 0,0,0,0
- 0,0,0,1
- 1,1,1,0
- 1,1,1,1
can be loaded into PS inputs using the DEFAULT_VAL field, and the VS exports
can be removed from the IR to save export & parameter memory.

After LLVM optimizations, analyze the IR to see which exports are equal to
the ones listed above (or undef) and remove them if they are.

Targeted use cases:
- All DX9 eON ports always clear 10 VS outputs to 0.0 even if most of them
  are unused by PS (such as Witcher 2 below).
- VS output arrays with unused elements that the GLSL compiler can't
  eliminate (such as Batman below).

The shader-db deltas are quite interesting:
(not from upstream si-report.py, it won't be upstreamed)

PERCENTAGE DELTAS    Shaders PARAM exports (affected only)
batman_arkham_origins    589  -67.17 %
bioshock-infinite       1769   -0.47 %
dirt-showdown            548   -2.68 %
dota2                   1747   -3.36 %
f1-2015                  776   -4.94 %
left_4_dead_2           1762   -0.07 %
metro_2033_redux        2670   -0.43 %
portal                   474   -0.22 %
talos_principle          324   -3.63 %
warsow                   176   -2.20 %
witcher2                1040  -73.78 %
----------------------------------------
All affected             991  -65.37 %  ... 9681 -> 3353
----------------------------------------
Total                  26725  -10.82 %  ... 58490 -> 52162

v2: treat Undef as both 0 and 1

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com> (v1)
Tested-by: Edmondo Tommasina <edmondo.tommasina@gmail.com> (v1)
2016-10-19 22:21:46 +02:00
Marek Olšák
a2ea653a49 radeonsi: remove cb0_is_integer handling
st/mesa does this for us.

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2016-10-19 19:26:30 +02:00
Marek Olšák
7dddf0b7ab radeonsi: adjust and clean up Z_ORDER and EXEC_ON_x settings
The table was copied from the Vulkan driver. The comment lines are as long
as the table for cosmetic reasons.

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2016-10-13 19:00:51 +02:00
Marek Olšák
e12c1cab5d radeonsi: disable ReZ
This is a serious performance fix. Discovered by luck.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=94354

Cc: 12.0 <mesa-stable@lists.freedesktop.org>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2016-10-13 19:00:51 +02:00
Marek Olšák
e4bbab9022 radeonsi: fix R600_DEBUG=precompile for shader-db
radeonsi no longer supports pixel shaders without interpolation optimizations,
which led to assertion failures in si_shader_ps when running shader-db.

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2016-10-12 18:29:40 +02:00
Marek Olšák
300a8221e9 radeonsi: add assertions to validate interpolation flags
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2016-10-05 21:03:23 +02:00
Marek Olšák
8c6ea5a6ff radeonsi: remove unnecessary #includes
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Reviewed-by: Edward O'Callaghan <funfunctor@folklore1984.net>
2016-10-04 16:12:07 +02:00
Marek Olšák
53d2c8f00f radeonsi: don't re-create shader PM4 states after scratch buffer update
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Reviewed-by: Edward O'Callaghan <funfunctor@folklore1984.net>
2016-10-04 16:12:05 +02:00
Marek Olšák
275c073c6a radeonsi: export SampleMask from pixel shaders at full rate
Heaven and Valley write gl_SampleMask and not Z.
Use 16_ABGR instead of 32_ABGR if Z isn't written.

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2016-09-13 20:38:25 +02:00
Marek Olšák
6c8b76263d radeonsi: also do VS_PARTIAL_FLUSH before updating VGT ring pointers
ported from Vulkan

Acked-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2016-09-05 18:01:15 +02:00
Marek Olšák
c3f716fe67 gallium/radeon: merge USER_SHADER and INTERNAL_SHADER priority flags
there's no reason to separate these

Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
2016-08-26 15:50:10 +02:00
Marek Olšák
e722b90bc9 radeonsi: eliminate PS OUT[1] if dual src blending is off and CB1 is not bound
All VP DX9 ports benefit from this.

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2016-08-17 12:24:35 +02:00
Marek Olšák
c15a9dec29 radeonsi: skip unnecessary si_update_shaders calls
Small decrease in draw call overhead.

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2016-08-03 17:46:46 +02:00
Marek Olšák
1e5f00f9d5 radeonsi: pre-generate shader logs for ddebug
This cuts down the overhead of si_dump_shader when ddebug is capturing
shader logs, which is done for every draw call unconditionally (that's
quite a lot of work for a draw call).

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2016-07-26 23:06:46 +02:00
Nicolai Hähnle
3d69357da9 radeonsi: ensure sample locations are set for line and polygon smoothing
Since commit d938b8c, the sample locations are no longer set unconditionally,
so we need to set the atom to dirty on all chips, not just Polaris.

Cc: 12.0 <mesa-stable@lists.freedesktop.org>
2016-07-23 15:36:39 +02:00
Rob Clark
44bbfedbd9 gallium/u_queue: add optional cleanup callback
Adds a second optional cleanup callback, called after the fence is
signaled.  This is needed if, for example, the queue has the last
reference to the object that embeds the util_queue_fence.  In this
case we cannot drop the ref in the main callback, since that would
result in the fence being destroyed before it is signaled.

Signed-off-by: Rob Clark <robdclark@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
2016-07-16 10:00:04 -04:00
Nicolai Hähnle
04d93ea619 radeonsi: disable multi-threading when shader dumps are enabled
Otherwise, shader dumps can become interleaved and unusable.

Reviewed-by: Edward O'Callaghan <funfunctor@folklore1984.net>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
2016-07-08 10:59:36 +02:00