mesa.git
5 years agoiris: drop XXX's about swizzling
Kenneth Graunke [Tue, 16 Oct 2018 02:25:24 +0000 (19:25 -0700)]
iris: drop XXX's about swizzling

pretty sure this is unnecessary on modern HW

5 years agoiris: drop dead format //'s
Kenneth Graunke [Tue, 16 Oct 2018 02:22:22 +0000 (19:22 -0700)]
iris: drop dead format //'s

these just aren't supported

5 years agoiris: yes
Kenneth Graunke [Tue, 16 Oct 2018 02:15:06 +0000 (19:15 -0700)]
iris: yes

5 years agoiris: initial compute caps
Kenneth Graunke [Fri, 27 Jul 2018 06:13:51 +0000 (23:13 -0700)]
iris: initial compute caps

RET macro borrowed from freedreno

5 years agoiris: Enable fb fetch
Kenneth Graunke [Mon, 15 Oct 2018 17:43:49 +0000 (10:43 -0700)]
iris: Enable fb fetch

needed for ES 3.2

5 years agoiris: advertise GL_ARB_shader_texture_image_samples
Kenneth Graunke [Mon, 15 Oct 2018 16:51:54 +0000 (09:51 -0700)]
iris: advertise GL_ARB_shader_texture_image_samples

5 years agoiris: Set num_uniforms in bytes
Jordan Justen [Tue, 18 Sep 2018 21:22:34 +0000 (14:22 -0700)]
iris: Set num_uniforms in bytes

Ref: brw_nir_lower_uniforms, type_size_scalar_bytes

Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
5 years agoiris: move images next to textures in binding table
Kenneth Graunke [Thu, 11 Oct 2018 04:44:43 +0000 (21:44 -0700)]
iris: move images next to textures in binding table

5 years agoiris: null for non-existent cbufs
Kenneth Graunke [Tue, 2 Oct 2018 17:21:57 +0000 (10:21 -0700)]
iris: null for non-existent cbufs

prevents BTs from being shifted down incorrectly

5 years agoiris: actually set image access
Kenneth Graunke [Tue, 2 Oct 2018 17:09:26 +0000 (10:09 -0700)]
iris: actually set image access

5 years agoiris: Don't lower image formats for write-only images
Jason Ekstrand [Tue, 2 Oct 2018 03:18:00 +0000 (22:18 -0500)]
iris: Don't lower image formats for write-only images

5 years agoiris: set image access correctly
Kenneth Graunke [Fri, 14 Sep 2018 07:49:13 +0000 (00:49 -0700)]
iris: set image access correctly

5 years agoiris: bother with BTIs
Kenneth Graunke [Thu, 30 Aug 2018 22:49:32 +0000 (15:49 -0700)]
iris: bother with BTIs

5 years agoiris: implement set_shader_images hook
Kenneth Graunke [Thu, 30 Aug 2018 22:45:36 +0000 (15:45 -0700)]
iris: implement set_shader_images hook

5 years agoiris: lower storage image derefs
Kenneth Graunke [Thu, 30 Aug 2018 22:20:12 +0000 (15:20 -0700)]
iris: lower storage image derefs

5 years agoiris: set the binding table size
Kenneth Graunke [Fri, 14 Sep 2018 21:18:13 +0000 (14:18 -0700)]
iris: set the binding table size

we weren't doing mark_surface_used on images (i965 does it while
uploading the unnecessary image uniforms), so our binding tables were
too small...

5 years agoiris: X32_S8X24 :/
Kenneth Graunke [Tue, 9 Oct 2018 23:47:54 +0000 (16:47 -0700)]
iris: X32_S8X24 :/

This can happen when faking Z32_S8X24 and setting StencilSampling = true

I guess we'll just turn it into S8_UINT...

Fixes KHR-GL45.texture_swizzle.functional

5 years agoiris: enable I/L formats
Kenneth Graunke [Mon, 8 Oct 2018 03:50:19 +0000 (20:50 -0700)]
iris: enable I/L formats

5 years agoiris: Use R/RG instead of I/L/A when sampling
Kenneth Graunke [Mon, 8 Oct 2018 03:47:21 +0000 (20:47 -0700)]
iris: Use R/RG instead of I/L/A when sampling

5 years agoiris: rework format translation apis
Kenneth Graunke [Mon, 8 Oct 2018 03:31:09 +0000 (20:31 -0700)]
iris: rework format translation apis

5 years agoiris: Allow PIPE_CONTROL with Stall at Scoreboard and RT flush
Kenneth Graunke [Mon, 8 Oct 2018 20:44:01 +0000 (13:44 -0700)]
iris: Allow PIPE_CONTROL with Stall at Scoreboard and RT flush

It's nonsensical, but not illegal, and mandatory on Icelake

5 years agoiris: add gen11 to genX_call
Kenneth Graunke [Mon, 8 Oct 2018 13:26:15 +0000 (06:26 -0700)]
iris: add gen11 to genX_call

5 years agoiris: inline stage_from_pipe to avoid unused warnings
Kenneth Graunke [Sun, 7 Oct 2018 05:54:38 +0000 (22:54 -0700)]
iris: inline stage_from_pipe to avoid unused warnings

5 years agoiris: pipe to scs -> iris_pipe.h
Kenneth Graunke [Sun, 7 Oct 2018 05:50:34 +0000 (22:50 -0700)]
iris: pipe to scs -> iris_pipe.h

5 years agoiris: force persample interp cap
Kenneth Graunke [Fri, 5 Oct 2018 22:25:53 +0000 (15:25 -0700)]
iris: force persample interp cap

5 years agoiris: stencil texturing
Kenneth Graunke [Fri, 5 Oct 2018 02:49:06 +0000 (19:49 -0700)]
iris: stencil texturing

5 years agoiris: fix Z32_S8 depth sampling
Kenneth Graunke [Fri, 5 Oct 2018 02:34:34 +0000 (19:34 -0700)]
iris: fix Z32_S8 depth sampling

We were accidentally using the ISL_FORMAT_R32_FLOAT_X8X24_TYPELESS
format, which is NOT what we use.  We just store R32_FLOAT depth.

fixes Piglit's texwrap GL_ARB_depth_buffer_float

5 years agoiris: don't mark contains_draw = false when chaining batches
Kenneth Graunke [Wed, 3 Oct 2018 03:09:01 +0000 (20:09 -0700)]
iris: don't mark contains_draw = false when chaining batches

chaining to a new batch reuses create_batch(), but we don't need to do
the work of pinning BOs we inherit from a previous batch...when that is
actually part of the same execbuf invocation.

instead, just flag it when setting primary_batch_size = 0, in
iris_batch_reset

5 years agoiris: vma_free bo->size, not bo_size
Kenneth Graunke [Wed, 3 Oct 2018 02:55:29 +0000 (19:55 -0700)]
iris: vma_free bo->size, not bo_size

this is more obviously correct.  I think the two end up being the same
in practice, since this is in the alloc_from_cache case, and presumably
bo from the bucket has bo->size == bucket->size, and bo_size also is
bucket->size...

still.  better to do the obvious thing.

brw_bufmgr already does it this way.

5 years agoiris: drop a bunch of pipe_sampler_state stuff we don't need
Kenneth Graunke [Tue, 2 Oct 2018 01:33:17 +0000 (18:33 -0700)]
iris: drop a bunch of pipe_sampler_state stuff we don't need

5 years agoiris: just mark snapshots_landed from the CPU
Kenneth Graunke [Sun, 30 Sep 2018 06:49:02 +0000 (08:49 +0200)]
iris: just mark snapshots_landed from the CPU

otherwise, get results may check q->map->snapshots_landed...before our
commands to initialize it to false have actually executed...so it'd get
some random garbage from the BO...

5 years agoiris: Enable ARB_shader_vote
Kenneth Graunke [Sat, 29 Sep 2018 09:04:48 +0000 (11:04 +0200)]
iris: Enable ARB_shader_vote

The easiest get out the vote campaign ever

5 years agoiris: magic number 36 -> #define
Kenneth Graunke [Sat, 29 Sep 2018 08:47:01 +0000 (10:47 +0200)]
iris: magic number 36 -> #define

5 years agoiris: better query file comment
Kenneth Graunke [Sat, 29 Sep 2018 08:43:00 +0000 (10:43 +0200)]
iris: better query file comment

5 years agoiris: early return properly
Kenneth Graunke [Sat, 29 Sep 2018 08:41:44 +0000 (10:41 +0200)]
iris: early return properly

5 years agoiris: 36-bit overflow fixes
Kenneth Graunke [Sat, 29 Sep 2018 08:39:39 +0000 (10:39 +0200)]
iris: 36-bit overflow fixes

5 years agoiris: Need to | 1 when asking for timestamps
Kenneth Graunke [Fri, 28 Sep 2018 14:45:04 +0000 (16:45 +0200)]
iris: Need to | 1 when asking for timestamps

5 years agoiris: glGet timestamps, more correct timestamps
Kenneth Graunke [Fri, 28 Sep 2018 13:42:51 +0000 (15:42 +0200)]
iris: glGet timestamps, more correct timestamps

5 years agoiris: ...and SO prims emitted queries
Kenneth Graunke [Fri, 28 Sep 2018 10:39:10 +0000 (12:39 +0200)]
iris: ...and SO prims emitted queries

looks like we have queries

some fails still due to races between snapshots_written and start/end
not being garbage...not sure what that's about

5 years agoiris: timestamps
Kenneth Graunke [Fri, 28 Sep 2018 10:25:36 +0000 (12:25 +0200)]
iris: timestamps

5 years agoiris: drop explicit pinning
Kenneth Graunke [Fri, 28 Sep 2018 10:09:19 +0000 (12:09 +0200)]
iris: drop explicit pinning

writes will already rw_bo or ro_bo that

5 years agoiris: primitives generated query support
Kenneth Graunke [Fri, 28 Sep 2018 10:07:54 +0000 (12:07 +0200)]
iris: primitives generated query support

5 years agoiris: pipeline stats
Kenneth Graunke [Thu, 27 Sep 2018 12:36:31 +0000 (14:36 +0200)]
iris: pipeline stats

5 years agoiris: play chicken with timer queries for now
Kenneth Graunke [Thu, 27 Sep 2018 12:27:41 +0000 (14:27 +0200)]
iris: play chicken with timer queries for now

they have been crashy in the past and I don't want to risk tanking my
laptop right before my XDC talk

5 years agoiris: gpr0 to bool
Kenneth Graunke [Thu, 27 Sep 2018 10:58:52 +0000 (12:58 +0200)]
iris: gpr0 to bool

I think OQ is basically working now.

5 years agoiris: fix random failures via CS stall...but why?
Kenneth Graunke [Thu, 27 Sep 2018 10:50:26 +0000 (12:50 +0200)]
iris: fix random failures via CS stall...but why?

5 years agoiris: flush batch when asking for result via QBO
Kenneth Graunke [Thu, 27 Sep 2018 10:14:26 +0000 (12:14 +0200)]
iris: flush batch when asking for result via QBO

5 years agoiris: results write
Kenneth Graunke [Thu, 27 Sep 2018 08:39:04 +0000 (10:39 +0200)]
iris: results write

5 years agoiris: gen10+ workarounds and break fix
Kenneth Graunke [Wed, 26 Sep 2018 10:51:44 +0000 (12:51 +0200)]
iris: gen10+ workarounds and break fix

5 years agoiris: initial query code
Kenneth Graunke [Mon, 17 Sep 2018 23:37:26 +0000 (16:37 -0700)]
iris: initial query code

5 years agoiris: LRM/SRM/SDI hooks
Kenneth Graunke [Mon, 17 Sep 2018 23:37:26 +0000 (16:37 -0700)]
iris: LRM/SRM/SDI hooks

5 years agoiris: rw_bo for pipe controls
Kenneth Graunke [Tue, 18 Sep 2018 04:15:48 +0000 (21:15 -0700)]
iris: rw_bo for pipe controls

this is used for WRITE IMMEDIATE...
but maybe we don't want to for the workaround BO?

5 years agoiris: use 0 for TCS passthrough program string ID
Kenneth Graunke [Fri, 21 Sep 2018 22:43:49 +0000 (15:43 -0700)]
iris: use 0 for TCS passthrough program string ID

the passthrough shader doesn't need a real program string ID - that's
basically used for ARB programs indicating total program source code
changes, or other pre-baked uniform changes, etc...none of which a
passthrough shader has...so we don't need a unique identifier to
distinguish them.  We want to use a consistent value so we find
existing passthrough shaders in the cache.

5 years agoiris: Add support for TCS passthrough
Caio Marcelo de Oliveira Filho [Fri, 21 Sep 2018 19:22:34 +0000 (12:22 -0700)]
iris: Add support for TCS passthrough

If no TCS is provided, create a "passthrough" TCS that will take the
default values set in the API as constants and pass to the TES, along
with any other inputs it expects.  The code to create the NIR shader
is the same as in i965.

Tested with

    ./piglit run -t 'tess' quick_shader r

and fixed a dozen crashes from that list.

5 years agoiris: inherit the index buffer properly
Kenneth Graunke [Fri, 21 Sep 2018 00:27:47 +0000 (17:27 -0700)]
iris: inherit the index buffer properly

5 years agoiris: delete bogus comment
Kenneth Graunke [Thu, 20 Sep 2018 23:46:58 +0000 (16:46 -0700)]
iris: delete bogus comment

Caio asked what was wrong.  There is nothing wrong.  :)

5 years agoiris: properly re-pin stencil buffers
Kenneth Graunke [Thu, 20 Sep 2018 22:35:14 +0000 (15:35 -0700)]
iris: properly re-pin stencil buffers

5 years agoiris: fix context restore of 3DSTATE_CONSTANT ranges
Kenneth Graunke [Thu, 20 Sep 2018 20:32:07 +0000 (13:32 -0700)]
iris: fix context restore of 3DSTATE_CONSTANT ranges

if clean we want to DO the pinning...not SKIP the pinning.

thanks to Jordan Justen for catching this!

5 years agoiris: silence const warning
Kenneth Graunke [Thu, 20 Sep 2018 18:12:27 +0000 (11:12 -0700)]
iris: silence const warning

not sure why this is labeled const, I'm pretty sure we are taking the
reference and owning this, so there's no particular reason we can't
change it.  it certainly seems to be working for non-compute.  and,
freedreno's ir3_shader.c seems to do this as well.  still...gross :/

5 years agoiris: refactor program CSO stuff
Kenneth Graunke [Fri, 27 Jul 2018 04:59:20 +0000 (21:59 -0700)]
iris: refactor program CSO stuff

5 years agoiris: Fix uses of gl_TessLevel*
Caio Marcelo de Oliveira Filho [Wed, 19 Sep 2018 17:17:23 +0000 (10:17 -0700)]
iris: Fix uses of gl_TessLevel*

The backend compiler expects the gl_TessLevel* variables to be mapped
as inputs instead of system values.  Use the new PIPE_CAP to get this
behavior from GLSL compiler.

Tested with:
tests/spec/arb_tessellation_shader/execution/vs-tcs-tes-tessinner-tessouter-inputs-quads.shader_test

5 years agoiris: totally untested icelake support
Kenneth Graunke [Tue, 18 Sep 2018 18:04:44 +0000 (11:04 -0700)]
iris: totally untested icelake support

5 years agoiris: initialize "don't suck" bits, as Ben likes to call them
Kenneth Graunke [Tue, 18 Sep 2018 18:07:16 +0000 (11:07 -0700)]
iris: initialize "don't suck" bits, as Ben likes to call them

5 years agoiris: refactor LRIs in context setup
Kenneth Graunke [Tue, 18 Sep 2018 18:04:24 +0000 (11:04 -0700)]
iris: refactor LRIs in context setup

we're going to have more of them, so reduce the boilerplate

5 years agoiris: enable ARB_enhanced_layouts
Kenneth Graunke [Sat, 15 Sep 2018 22:27:59 +0000 (15:27 -0700)]
iris: enable ARB_enhanced_layouts

5 years agoiris: re-pin binding table contents if we didn't re-emit them
Kenneth Graunke [Sat, 15 Sep 2018 21:50:26 +0000 (14:50 -0700)]
iris: re-pin binding table contents if we didn't re-emit them

fixes glsl-vs-loop and other regressions from multibinder.

5 years agoiris: move binder pinning outside the dirty == 0 check
Kenneth Graunke [Sat, 15 Sep 2018 21:35:47 +0000 (14:35 -0700)]
iris: move binder pinning outside the dirty == 0 check

This might be a new batch with back to back non-dirty calls, if so we
need to inherit the old binder...

5 years agoiris: fix memzone_for_address since multibinder changes
Chris Wilson [Fri, 14 Sep 2018 22:12:36 +0000 (15:12 -0700)]
iris: fix memzone_for_address since multibinder changes

5 years agoiris: update comments for multibinder
Kenneth Graunke [Thu, 13 Sep 2018 06:31:46 +0000 (23:31 -0700)]
iris: update comments for multibinder

5 years agoiris: fix SO offset writes for multiple streams
Kenneth Graunke [Thu, 13 Sep 2018 05:23:50 +0000 (22:23 -0700)]
iris: fix SO offset writes for multiple streams

5 years agoiris: Support multiple binder BOs, update Surface State Base Address
Kenneth Graunke [Sun, 9 Sep 2018 02:43:34 +0000 (19:43 -0700)]
iris: Support multiple binder BOs, update Surface State Base Address

5 years agoiris: fix null FB and unbound tex surface state addresses
Kenneth Graunke [Tue, 11 Sep 2018 08:09:27 +0000 (01:09 -0700)]
iris: fix null FB and unbound tex surface state addresses

5 years agoiris: set EXEC_OBJECT_CAPTURE on all driver internal buffers
Kenneth Graunke [Sat, 8 Sep 2018 06:39:30 +0000 (23:39 -0700)]
iris: set EXEC_OBJECT_CAPTURE on all driver internal buffers

5 years agoiris: fix constant buffer 0 to be absolute
Kenneth Graunke [Fri, 7 Sep 2018 19:26:55 +0000 (12:26 -0700)]
iris: fix constant buffer 0 to be absolute

thanks to Jason for catching this.  Fixes some va64 tests.  Surprisingly
not much else, as apparently getting to UBO range 4 is uncommon!

5 years agoiris: don't unconditionally emit 3DSTATE_VF / 3DSTATE_VF_TOPOLOGY
Kenneth Graunke [Sat, 1 Sep 2018 01:03:19 +0000 (18:03 -0700)]
iris: don't unconditionally emit 3DSTATE_VF / 3DSTATE_VF_TOPOLOGY

this was just laziness on my part

5 years agoiris: skip over whole function if dirty == 0
Kenneth Graunke [Sat, 1 Sep 2018 07:58:29 +0000 (00:58 -0700)]
iris: skip over whole function if dirty == 0

kinda pointless in non-pathological cases, but does boost our score in
the drawarrays case by 50%...

5 years agoiris: Allow inlining of require/get_command_space
Kenneth Graunke [Sun, 19 Aug 2018 17:31:29 +0000 (10:31 -0700)]
iris: Allow inlining of require/get_command_space

eliminates so many callqs for ptr++

5 years agoiris: use Eric's new caps helper
Kenneth Graunke [Fri, 7 Sep 2018 02:42:58 +0000 (19:42 -0700)]
iris: use Eric's new caps helper

this does change a couple caps...PRIMITIVE_RESTART_FOR_PATCHES...

5 years agoiris: new caps
Kenneth Graunke [Sat, 1 Sep 2018 04:22:31 +0000 (21:22 -0700)]
iris: new caps

5 years agoiris: fix blend state memcpy
Kenneth Graunke [Wed, 29 Aug 2018 22:24:45 +0000 (15:24 -0700)]
iris: fix blend state memcpy

thanks to Jason for noticing grumpy valgrind

5 years agoiris: Skip primitive ID overrides if the shader wrote a custom value
Kenneth Graunke [Sat, 25 Aug 2018 07:11:24 +0000 (00:11 -0700)]
iris: Skip primitive ID overrides if the shader wrote a custom value

Fixes glsl-1.50/execution/geometry/primitive-id-out

5 years agoiris: fix crash when binding optional shader for the first time
Kenneth Graunke [Fri, 24 Aug 2018 04:21:04 +0000 (21:21 -0700)]
iris: fix crash when binding optional shader for the first time

5 years agoiris: handle level/layer in direct maps
Kenneth Graunke [Thu, 23 Aug 2018 09:36:18 +0000 (02:36 -0700)]
iris: handle level/layer in direct maps

needed now that we do 1D linear

5 years agoiris: use linear for 1D textures
Kenneth Graunke [Thu, 23 Aug 2018 06:39:38 +0000 (23:39 -0700)]
iris: use linear for 1D textures

This gets us the gen9 compact linear storage

5 years agoiris: big old hack for tex-miplevel-selection
Kenneth Graunke [Thu, 23 Aug 2018 08:49:49 +0000 (01:49 -0700)]
iris: big old hack for tex-miplevel-selection

copied from ilo.  I don't understand this at all..

5 years agoiris: fix sampler state setting
Kenneth Graunke [Thu, 23 Aug 2018 09:15:12 +0000 (02:15 -0700)]
iris: fix sampler state setting

5 years agoiris: try to hack around binder issue
Kenneth Graunke [Tue, 21 Aug 2018 18:50:56 +0000 (11:50 -0700)]
iris: try to hack around binder issue

5 years agoiris: fix line-aa-width
Kenneth Graunke [Tue, 21 Aug 2018 18:30:09 +0000 (11:30 -0700)]
iris: fix line-aa-width

we should probably move the roundf to st_atom_raster

5 years agoiris: implement get_sample_position
Kenneth Graunke [Tue, 21 Aug 2018 06:37:13 +0000 (23:37 -0700)]
iris: implement get_sample_position

Fixes arb_sample_shading/builtin-gl-sample-position

5 years agoiris: z_res -> s_res
Kenneth Graunke [Tue, 21 Aug 2018 04:23:25 +0000 (21:23 -0700)]
iris: z_res -> s_res

fixes crashes introduced a few commits ago

5 years agoiris: reenable R32G32B32 texture buffers
Kenneth Graunke [Tue, 21 Aug 2018 03:55:17 +0000 (20:55 -0700)]
iris: reenable R32G32B32 texture buffers

This dropped us from GL 4.2 to GL 3.3 by mistake.  Thanks to Dave for
catching this!

5 years agoiris: Record reusability of bo on construction
Chris Wilson [Sun, 19 Aug 2018 09:32:43 +0000 (10:32 +0100)]
iris: Record reusability of bo on construction

We know that if the bufmgr->reuse is set to false or if the bo is too
large for a bucket, the same will be true when we come to free the bo.

5 years agoiris: Reduce binder alignment from 64 to 32
Kenneth Graunke [Sun, 19 Aug 2018 21:05:53 +0000 (14:05 -0700)]
iris: Reduce binder alignment from 64 to 32

3DSTATE_BINDING_TABLE_POINTER_XS's alignment requirement is only 32B.

Makes us waste less precious binder space.

5 years agoiris: precompute hashes for cache tracking
Kenneth Graunke [Sun, 19 Aug 2018 17:08:05 +0000 (10:08 -0700)]
iris: precompute hashes for cache tracking

saves a touch of cpu overhead in the new resolve tracking

5 years agoiris: AMD_pinned_memory
Chris Wilson [Tue, 31 Jul 2018 13:47:02 +0000 (14:47 +0100)]
iris: AMD_pinned_memory

(rebased by Ken, mainly set res->internal_format)

5 years agoiris: proper cache tracking
Kenneth Graunke [Sun, 19 Aug 2018 07:21:34 +0000 (00:21 -0700)]
iris: proper cache tracking

this is copied from the i965 aux resolve stuff...minus the aux resolves

5 years agoiris: Move cache tracking to iris_resolve.c
Kenneth Graunke [Sun, 19 Aug 2018 07:18:51 +0000 (00:18 -0700)]
iris: Move cache tracking to iris_resolve.c

5 years agoiris: use consistent copyright formatting
Kenneth Graunke [Sun, 19 Aug 2018 07:31:46 +0000 (00:31 -0700)]
iris: use consistent copyright formatting

some of them had typos, didn't say 'authors or copyright holders',
or other mistakes.  This is now https://opensource.org/licenses/MIT
text, formatted consistently.

5 years agoiris: track depth/stencil writes enabled
Kenneth Graunke [Sun, 19 Aug 2018 06:21:41 +0000 (23:21 -0700)]
iris: track depth/stencil writes enabled

5 years agoiris: Move iris_sampler_view declaration to iris_resource.h
Kenneth Graunke [Sun, 19 Aug 2018 06:58:54 +0000 (23:58 -0700)]
iris: Move iris_sampler_view declaration to iris_resource.h

We'll need this for resolve tracking.  There's also no genxml stuff here