mesa.git
7 years agoradeonsi: decompress DCC in set_sampler_view instead of create_sampler_view (v2)
Marek Olšák [Fri, 24 Mar 2017 02:16:41 +0000 (03:16 +0100)]
radeonsi: decompress DCC in set_sampler_view instead of create_sampler_view (v2)

v2: don't add a new decompress helper function

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi: decompress DCC in set_framebuffer_state instead of create_surface (v2)
Marek Olšák [Fri, 24 Mar 2017 02:02:53 +0000 (03:02 +0100)]
radeonsi: decompress DCC in set_framebuffer_state instead of create_surface (v2)

for threaded gallium, which can't use pipe_context in create_surface

v2: don't add a new decompress helper function

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agost/glsl_to_tgsi: fix 64-bit integer bit shifts
Nicolai Hähnle [Thu, 30 Mar 2017 15:07:34 +0000 (17:07 +0200)]
st/glsl_to_tgsi: fix 64-bit integer bit shifts

Fix a bug that was caused by a type mismatch in the shift count between
GLSL and TGSI. I briefly considered adjusting the TGSI semantics, but
since both LLVM and AMD GCN require both arguments to be of the same type,
it makes more sense to keep TGSI as-is -- it reflects the underlying
implementation better.

I'm also sending out piglit tests that expose this error.

v2: use the right number of components for the temporary register

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
7 years agotgsi: fix printing of 64-bit integer immediates
Nicolai Hähnle [Thu, 30 Mar 2017 12:11:12 +0000 (14:11 +0200)]
tgsi: fix printing of 64-bit integer immediates

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
7 years agointel: genxml: fix out of tree builds
Lionel Landwerlin [Fri, 31 Mar 2017 10:05:42 +0000 (11:05 +0100)]
intel: genxml: fix out of tree builds

v2: use Emil's recommendation
    change rule to closer to genxml/genX_bits.h

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
7 years agogbm/dri: Check dri extension version before flush after unmap
Thomas Hellstrom [Thu, 30 Mar 2017 14:33:47 +0000 (16:33 +0200)]
gbm/dri: Check dri extension version before flush after unmap

The commit mentioned below required the __DRI2FlushExtension to have
version 4 or above, for GBM functionality. That broke GBM with some
classic dri drivers. Relax that requirement so that we only flush
after unmap if we have version 4 or above. Drivers that require the flush
for correct functionality should implement the desired version.

Fixes: ba8df228 ("gbm/dri: Flush after unmap")
Cc: <mesa-stable@lists.freedesktop.org>
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Tested-by: Dylan Baker <dylan@pnwbakers.com>
7 years agoradeonsi: implement ARB_shader_group_vote
Nicolai Hähnle [Wed, 29 Mar 2017 18:29:37 +0000 (20:29 +0200)]
radeonsi: implement ARB_shader_group_vote

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
7 years agoradeonsi: enable ARB_shader_clock
Nicolai Hähnle [Thu, 30 Mar 2017 06:55:22 +0000 (08:55 +0200)]
radeonsi: enable ARB_shader_clock

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
7 years agoradeonsi: emit TGSI_OPCODE_CLOCK
Nicolai Hähnle [Thu, 30 Mar 2017 07:24:24 +0000 (09:24 +0200)]
radeonsi: emit TGSI_OPCODE_CLOCK

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
7 years agost/mesa: implement ARB_shader_clock
Nicolai Hähnle [Thu, 30 Mar 2017 06:55:05 +0000 (08:55 +0200)]
st/mesa: implement ARB_shader_clock

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
7 years agotgsi: add CLOCK opcode
Ilia Mirkin [Thu, 9 Feb 2017 23:37:48 +0000 (18:37 -0500)]
tgsi: add CLOCK opcode

Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
7 years agogallium: add PIPE_CAP_TGSI CLOCK
Nicolai Hähnle [Wed, 29 Mar 2017 18:44:57 +0000 (20:44 +0200)]
gallium: add PIPE_CAP_TGSI CLOCK

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
7 years agoglsl: fix clockARB builtin function
Nicolai Hähnle [Thu, 30 Mar 2017 06:54:46 +0000 (08:54 +0200)]
glsl: fix clockARB builtin function

The underlying intrinsic is defined to always have a uvec2 return type.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
7 years agoanv: change BLOCK_POOL_MEMFD_SIZE to 1GB
Tapani Pälli [Thu, 30 Mar 2017 10:56:25 +0000 (13:56 +0300)]
anv: change BLOCK_POOL_MEMFD_SIZE to 1GB

This allows us to run 32bit Vulkan apps on Android, ftruncate
call would fail on 2GB (max size being 2GB - 1).

Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
7 years agoandroid: add libmesa_genxml as dep to libmesa_isl
Tapani Pälli [Thu, 30 Mar 2017 05:51:25 +0000 (08:51 +0300)]
android: add libmesa_genxml as dep to libmesa_isl

This is to fix following compile error with libmesa_isl:
   mesa/src/intel/isl/isl.c:28:10: fatal error: 'genxml/genX_bits.h' file not found

Fixes: f0eaf38 ("genxml: New generated header genX_bits.h (v6)")
Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Emli Velikov <emil.velikov@collabora.com>
7 years agomesa: remove MESA_GLSL=opt
Timothy Arceri [Thu, 30 Mar 2017 11:21:32 +0000 (22:21 +1100)]
mesa: remove MESA_GLSL=opt

This is unused.

Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Reviewed-by: Emil Velikov <emli.velikov@collabora.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
7 years agomesa: remove MESA_GLSL=no_opts env option
Timothy Arceri [Thu, 30 Mar 2017 11:14:21 +0000 (22:14 +1100)]
mesa: remove MESA_GLSL=no_opts env option

This is confusing because is only applys to GL_ARB_vertex/fragment_program,
and because of that its also not very useful.

If someone requires this for debugging they can just make an ad-hoc
code change.

Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
7 years agomesa: move FLUSH_VERTICES() call to meta
Timothy Arceri [Thu, 30 Mar 2017 10:24:13 +0000 (21:24 +1100)]
mesa: move FLUSH_VERTICES() call to meta

There is no need for this to be in the common code.

Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
7 years agomesa/vbo: remove redundant _mesa_is_bufferobj() calls
Timothy Arceri [Thu, 30 Mar 2017 10:14:43 +0000 (21:14 +1100)]
mesa/vbo: remove redundant _mesa_is_bufferobj() calls

This is already called inside the vbo_exec_vtx_{unmap,map}()
functions.

Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
7 years agomesa/glthread: add async support to ARB_gpu_shader_int64 uniform functions
Timothy Arceri [Thu, 30 Mar 2017 09:43:08 +0000 (20:43 +1100)]
mesa/glthread: add async support to ARB_gpu_shader_int64 uniform functions

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agomesa/glthread: add async support to ARB_gpu_shader_fp64 uniform functions
Timothy Arceri [Thu, 30 Mar 2017 09:43:07 +0000 (20:43 +1100)]
mesa/glthread: add async support to ARB_gpu_shader_fp64 uniform functions

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoaubinator: enable snb/ilk through --gen
Lionel Landwerlin [Sat, 25 Mar 2017 13:59:52 +0000 (13:59 +0000)]
aubinator: enable snb/ilk through --gen

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
7 years agointel: genxml: compress all gen files into one
Lionel Landwerlin [Sat, 25 Mar 2017 02:52:33 +0000 (02:52 +0000)]
intel: genxml: compress all gen files into one

Combining all the files into a single string didn't make any
difference in the size of the aubinator binary.

With this change we now also embed gen4/4.5/5 descriptions, which
increases the aubinator size by ~16Kb.

v2 (Lionel): rebase makefiles

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
7 years agoradv: Use the guard band.
Bas Nieuwenhuizen [Wed, 1 Mar 2017 08:32:19 +0000 (09:32 +0100)]
radv: Use the guard band.

Signed-off-by: Bas Nieuwenhuizen <basni@google.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
7 years agoradv: Prepare for not using the guard band for lines & points.
Bas Nieuwenhuizen [Wed, 29 Mar 2017 20:58:10 +0000 (22:58 +0200)]
radv: Prepare for not using the guard band for lines & points.

Vulkan Clipping is defined in terms of vertices, the scissor based
clipping happens on pixels. There is a difference with points and
lines, as a vertex can be outside the viewport while some pixels are in.
On Vulkan thoise pixels shouldn't be drawn, while they would be with
the guardband.

Signed-off-by: Bas Nieuwenhuizen <basni@google.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
7 years agoradv: Drop the default viewport when 0 viewports are given.
Bas Nieuwenhuizen [Wed, 1 Mar 2017 01:14:08 +0000 (02:14 +0100)]
radv: Drop the default viewport when 0 viewports are given.

Signed-off-by: Bas Nieuwenhuizen <basni@google.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
7 years agoradv: Set proper viewport & scissor for meta draws.
Bas Nieuwenhuizen [Tue, 28 Feb 2017 23:39:58 +0000 (00:39 +0100)]
radv: Set proper viewport & scissor for meta draws.

Signed-off-by: Bas Nieuwenhuizen <basni@google.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
7 years agomesa: Fix trailing whitespace in polygon.c
Lyude [Thu, 23 Mar 2017 00:51:09 +0000 (20:51 -0400)]
mesa: Fix trailing whitespace in polygon.c

Signed-off-by: Lyude <lyude@redhat.com>
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
7 years agomesa: Fix gross indenting in _mesa_PolygonMode()
Lyude [Thu, 23 Mar 2017 00:51:08 +0000 (20:51 -0400)]
mesa: Fix gross indenting in _mesa_PolygonMode()

Signed-off-by: Lyude <lyude@redhat.com>
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
7 years agor300: Fix indenting in r300_get_param()
Lyude [Thu, 23 Mar 2017 00:51:07 +0000 (20:51 -0400)]
r300: Fix indenting in r300_get_param()

Signed-off-by: Lyude <lyude@redhat.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
7 years agovc4: Fix indenting in vc4_screen_get_param()
Lyude [Thu, 23 Mar 2017 00:51:06 +0000 (20:51 -0400)]
vc4: Fix indenting in vc4_screen_get_param()

Signed-off-by: Lyude <lyude@redhat.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
7 years agointel: Add INTEL_CFLAGS to aubinator CFLAGS.
Kenneth Graunke [Thu, 30 Mar 2017 18:55:33 +0000 (11:55 -0700)]
intel: Add INTEL_CFLAGS to aubinator CFLAGS.

It still needs intel_aub.h.  Fixes the build.

7 years agonir: Add support for 8 and 16-bit types
Jason Ekstrand [Thu, 9 Mar 2017 04:34:28 +0000 (20:34 -0800)]
nir: Add support for 8 and 16-bit types

Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Eduardo Lima Mitev <elima@igalia.com>
7 years agonir/constant_expressions: Don't switch on bit size when not needed
Jason Ekstrand [Tue, 14 Mar 2017 17:31:21 +0000 (10:31 -0700)]
nir/constant_expressions: Don't switch on bit size when not needed

For opcodes such as the nir_op_pack_64_2x32 for which all sources and
destinations have explicit sizes, the bit_size parameter to the evaluate
function is pointless and *should* do nothing.  Previously, we were
always switching on the bit_size and asserting if it isn't one of the
sizes in the list.  This generates way more code than needed and is a
bit cruel because it doesn't let us have a bit_size of zero on an ALU op
which shouldn't need a bit_size.

Reviewed-by: Eduardo Lima Mitev <elima@igalia.com>
7 years agonir/constant_expressions: Pull the guts out into a helper block
Jason Ekstrand [Tue, 14 Mar 2017 17:27:38 +0000 (10:27 -0700)]
nir/constant_expressions: Pull the guts out into a helper block

Reviewed-by: Eduardo Lima Mitev <elima@igalia.com>
7 years agoi965: Stop using legacy dri_bufmgr_* and intel_* names.
Kenneth Graunke [Tue, 21 Mar 2017 21:46:39 +0000 (14:46 -0700)]
i965: Stop using legacy dri_bufmgr_* and intel_* names.

Eric renamed these from dri_bufmgr_* and intel_bufmgr_* to drm_intel_*
in libdrm commit 4b9826408f65976a1a13387beda748b65e03ec52, circa 2008,
but we've been using the legacy names this whole time.

Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
7 years agointel: automake: move INTEL_CFLAGS as applicable
Emil Velikov [Fri, 17 Mar 2017 16:55:25 +0000 (16:55 +0000)]
intel: automake: move INTEL_CFLAGS as applicable

Only common/decoder.[ch] requires it [for intel_aub.h].

v2: The code was moved to from intel/tools to intel/common,
update accordingly.

Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
7 years agointel: android: remove libdrm_intel requirement
Emil Velikov [Fri, 17 Mar 2017 16:55:24 +0000 (16:55 +0000)]
intel: android: remove libdrm_intel requirement

The only part which requires libdrm_intel tools/aubinator is not built
on Android.

Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
7 years agoPartially revert "amd/addrlib: silence warnings" to fix builds with DEBUG
Marek Olšák [Thu, 30 Mar 2017 17:01:02 +0000 (19:01 +0200)]
Partially revert "amd/addrlib: silence warnings" to fix builds with DEBUG

This partially reverts commit 8a74140a21fe6b0d2e8a60b065b890f797f2db51.

7 years agoddebug: implement clear_texture
Marek Olšák [Tue, 28 Mar 2017 00:15:23 +0000 (02:15 +0200)]
ddebug: implement clear_texture

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi: fix an unused-variable warning in a release build
Marek Olšák [Thu, 30 Mar 2017 15:21:47 +0000 (17:21 +0200)]
radeonsi: fix an unused-variable warning in a release build

7 years agovdpau: fix a maybe-uninitialized warning
Marek Olšák [Thu, 30 Mar 2017 15:11:41 +0000 (17:11 +0200)]
vdpau: fix a maybe-uninitialized warning

7 years agosoftpipe: fix a maybe-uninitialized warning
Marek Olšák [Thu, 30 Mar 2017 15:11:41 +0000 (17:11 +0200)]
softpipe: fix a maybe-uninitialized warning

/home/marek/dev/mesa-main/src/gallium/drivers/softpipe/sp_compute.c:178:
 warning: 'grid_size' may be used uninitialized in this function
 [-Wmaybe-uninitialized]

7 years agogallivm: fix a maybe-uninitialized warning
Marek Olšák [Thu, 30 Mar 2017 15:11:41 +0000 (17:11 +0200)]
gallivm: fix a maybe-uninitialized warning

/home/marek/dev/mesa-main/src/gallium/auxiliary/gallivm/lp_bld_sample_soa.c:3598:
 warning: 'level' may be used uninitialized in this function [-Wmaybe-uninitialized]
       out1 = lp_build_cmp(&leveli_bld, PIPE_FUNC_GREATER, level, last_level);
            ^

7 years agogallium/radeon: s/dcc_disable/disable_dcc/
Marek Olšák [Wed, 29 Mar 2017 17:41:48 +0000 (19:41 +0200)]
gallium/radeon: s/dcc_disable/disable_dcc/

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Tested-by: Edmondo Tommasina <edmondo.tommasina@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
7 years agoradeonsi: handle incompatible DCC formats in resource_copy_region
Marek Olšák [Fri, 24 Mar 2017 11:21:20 +0000 (12:21 +0100)]
radeonsi: handle incompatible DCC formats in resource_copy_region

Required because of later commits.

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Tested-by: Edmondo Tommasina <edmondo.tommasina@gmail.com>
7 years agoradeonsi: remove a workaround for inexact *8_SNORM blits
Marek Olšák [Fri, 24 Mar 2017 11:31:34 +0000 (12:31 +0100)]
radeonsi: remove a workaround for inexact *8_SNORM blits

All tests pass on Fiji now. This prevents DCC disablement due to
incompatible DCC formats due to the fallback.

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Tested-by: Edmondo Tommasina <edmondo.tommasina@gmail.com>
7 years agogallium/radeon: add and use a new helper vi_dcc_enabled
Marek Olšák [Fri, 24 Mar 2017 01:58:54 +0000 (02:58 +0100)]
gallium/radeon: add and use a new helper vi_dcc_enabled

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Tested-by: Edmondo Tommasina <edmondo.tommasina@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
7 years agogallium/radeon: formalize that r600_query_hw_add_result doesn't need a context
Marek Olšák [Sun, 12 Mar 2017 20:26:22 +0000 (21:26 +0100)]
gallium/radeon: formalize that r600_query_hw_add_result doesn't need a context

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Tested-by: Edmondo Tommasina <edmondo.tommasina@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
7 years agoradeonsi: don't make a copy of pipe_index_buffer in draw_vbo
Marek Olšák [Tue, 28 Mar 2017 20:19:29 +0000 (22:19 +0200)]
radeonsi: don't make a copy of pipe_index_buffer in draw_vbo

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Tested-by: Edmondo Tommasina <edmondo.tommasina@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
7 years agogallium/util: use const in u_index_modify helpers
Marek Olšák [Tue, 28 Mar 2017 20:20:56 +0000 (22:20 +0200)]
gallium/util: use const in u_index_modify helpers

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Tested-by: Edmondo Tommasina <edmondo.tommasina@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
7 years agowinsys/amdgpu: remove AMDGPU_INFO_NUM_EVICTIONS
Samuel Pitoiset [Wed, 29 Mar 2017 19:06:38 +0000 (21:06 +0200)]
winsys/amdgpu: remove AMDGPU_INFO_NUM_EVICTIONS

This is now exposed with libdrm_amdgpu 2.4.76.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi: add Vega10 PCI IDs
Marek Olšák [Tue, 13 Dec 2016 17:35:35 +0000 (18:35 +0100)]
radeonsi: add Vega10 PCI IDs

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeon/uvd: set correct vega10 db pitch alignment
Boyuan Zhang [Thu, 16 Mar 2017 21:52:33 +0000 (17:52 -0400)]
radeon/uvd: set correct vega10 db pitch alignment

Create new function to get correct alignment based on Asics, and change
the corresponding decode message buffer and dpb buffer size calculations

Signed-off-by: Boyuan Zhang <boyuan.zhang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
7 years agoradeon/vce: add vce support for firmware 53.19.4
Leo Liu [Fri, 10 Feb 2017 15:44:05 +0000 (10:44 -0500)]
radeon/vce: add vce support for firmware 53.19.4

v2: squashed with other similar commits

Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
7 years agoradeon/vce: adapt gfx9 surface to vce
Leo Liu [Fri, 10 Feb 2017 15:41:31 +0000 (10:41 -0500)]
radeon/vce: adapt gfx9 surface to vce

Signed-off-by: Leo Liu <leo.liu@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
7 years agowinsys/surface: add height pitch for gfx9
Leo Liu [Fri, 10 Feb 2017 15:36:21 +0000 (10:36 -0500)]
winsys/surface: add height pitch for gfx9

Signed-off-by: Leo Liu <leo.liu@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
7 years agoradeon/uvd: clear message buffer when reuse
Leo Liu [Thu, 9 Feb 2017 15:30:21 +0000 (10:30 -0500)]
radeon/uvd: clear message buffer when reuse

As required by firmware

Signed-off-by: Leo Liu <leo.liu@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
7 years agoradeon/uvd: adapt gfx9 surface to uvd
Leo Liu [Thu, 9 Feb 2017 15:25:20 +0000 (10:25 -0500)]
radeon/uvd: adapt gfx9 surface to uvd

Signed-off-by: Leo Liu <leo.liu@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
7 years agoradeon/uvd: add uvd soc15 register
Leo Liu [Thu, 9 Feb 2017 15:16:06 +0000 (10:16 -0500)]
radeon/uvd: add uvd soc15 register

Signed-off-by: Leo Liu <leo.liu@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
7 years agoradeonsi/gfx9: disable features that don't work
Marek Olšák [Tue, 14 Mar 2017 22:26:30 +0000 (23:26 +0100)]
radeonsi/gfx9: disable features that don't work

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi/gfx9: only allow GL 3.1
Marek Olšák [Sat, 15 Oct 2016 14:07:58 +0000 (16:07 +0200)]
radeonsi/gfx9: only allow GL 3.1

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi/gfx9: add linear address computations for texture transfers
Marek Olšák [Wed, 26 Oct 2016 22:10:13 +0000 (00:10 +0200)]
radeonsi/gfx9: add linear address computations for texture transfers

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi/gfx9: don't generate LS and ES states
Marek Olšák [Wed, 23 Nov 2016 01:41:14 +0000 (02:41 +0100)]
radeonsi/gfx9: don't generate LS and ES states

these shaders don't exist on GFX9

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi/gfx9: SPI_SHADER_USER_DATA changes
Marek Olšák [Sat, 15 Oct 2016 12:38:59 +0000 (14:38 +0200)]
radeonsi/gfx9: SPI_SHADER_USER_DATA changes

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agowinsys/amdgpu: set/get BO tiling flags for GFX9
Marek Olšák [Wed, 23 Nov 2016 17:42:53 +0000 (18:42 +0100)]
winsys/amdgpu: set/get BO tiling flags for GFX9

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi/gfx9: handle pitch and offset overrides for texture_from_handle
Marek Olšák [Thu, 27 Oct 2016 15:33:42 +0000 (17:33 +0200)]
radeonsi/gfx9: handle pitch and offset overrides for texture_from_handle

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi/gfx9: set/validate GFX9 BO metadata
Marek Olšák [Sun, 6 Nov 2016 14:46:26 +0000 (15:46 +0100)]
radeonsi/gfx9: set/validate GFX9 BO metadata

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi/gfx9: add radeon_surf.gfx9.surf_offset
Marek Olšák [Wed, 15 Feb 2017 23:11:58 +0000 (00:11 +0100)]
radeonsi/gfx9: add radeon_surf.gfx9.surf_offset

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi/gfx9: don't write mipmap level offsets to BO metadata
Marek Olšák [Sat, 15 Oct 2016 13:27:57 +0000 (15:27 +0200)]
radeonsi/gfx9: don't write mipmap level offsets to BO metadata

GFX9 doesn't have (usable) mipmap offsets.

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi/gfx9: flush CB & DB caches with an EOP TS event
Marek Olšák [Sat, 15 Oct 2016 14:21:20 +0000 (16:21 +0200)]
radeonsi/gfx9: flush CB & DB caches with an EOP TS event

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi/gfx9: use ACQUIRE_MEM
Marek Olšák [Sat, 15 Oct 2016 14:09:26 +0000 (16:09 +0200)]
radeonsi/gfx9: use ACQUIRE_MEM

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi/gfx9: only use CE RAM for most-used descriptors
Marek Olšák [Sun, 6 Nov 2016 19:22:12 +0000 (20:22 +0100)]
radeonsi/gfx9: only use CE RAM for most-used descriptors

because the CE RAM size decreased to 4 KB.

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi/gfx9: emit FLUSH_DFSM where required
Marek Olšák [Sun, 6 Nov 2016 18:27:09 +0000 (19:27 +0100)]
radeonsi/gfx9: emit FLUSH_DFSM where required

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi/gfx9: emit BREAK_BATCH in emit_framebuffer_state
Marek Olšák [Sun, 6 Nov 2016 18:25:12 +0000 (19:25 +0100)]
radeonsi/gfx9: emit BREAK_BATCH in emit_framebuffer_state

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi/gfx9: fix MIP0_WIDTH & MIP0_HEIGHT for compressed texture blits
Marek Olšák [Mon, 30 Jan 2017 23:56:34 +0000 (00:56 +0100)]
radeonsi/gfx9: fix MIP0_WIDTH & MIP0_HEIGHT for compressed texture blits

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi/gfx9: fix textureSize/imageSize for 1D textures
Marek Olšák [Tue, 31 Jan 2017 20:02:19 +0000 (21:02 +0100)]
radeonsi/gfx9: fix textureSize/imageSize for 1D textures

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi/gfx9: add a workaround for 1D depth textures
Marek Olšák [Tue, 24 Jan 2017 20:39:42 +0000 (21:39 +0100)]
radeonsi/gfx9: add a workaround for 1D depth textures

The same workaround is used by Vulkan.

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi/gfx9: enable clamping for Z UNORM formats promoted to Z32F
Marek Olšák [Tue, 31 Jan 2017 21:56:38 +0000 (22:56 +0100)]
radeonsi/gfx9: enable clamping for Z UNORM formats promoted to Z32F

so that shaders don't have to do it.

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi/gfx9: image descriptor changes in mutable fields
Marek Olšák [Sat, 15 Oct 2016 13:27:18 +0000 (15:27 +0200)]
radeonsi/gfx9: image descriptor changes in mutable fields

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi/gfx9: FMASK image descriptor changes
Marek Olšák [Sat, 15 Oct 2016 13:25:44 +0000 (15:25 +0200)]
radeonsi/gfx9: FMASK image descriptor changes

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi/gfx9: image descriptor changes in immutable fields
Marek Olšák [Sat, 15 Oct 2016 13:24:45 +0000 (15:24 +0200)]
radeonsi/gfx9: image descriptor changes in immutable fields

The border color swizzle logic was copied from Vulkan. It doesn't make any
sense to me, but it passes all piglits except the stencil ones.

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi/gfx9: DB changes
Marek Olšák [Sat, 15 Oct 2016 13:22:34 +0000 (15:22 +0200)]
radeonsi/gfx9: DB changes

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi/gfx9: CB changes
Marek Olšák [Sat, 15 Oct 2016 13:09:47 +0000 (15:09 +0200)]
radeonsi/gfx9: CB changes

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi/gfx9: do DCC clears on non-mipmapped textures only
Marek Olšák [Thu, 27 Oct 2016 21:48:44 +0000 (23:48 +0200)]
radeonsi/gfx9: do DCC clears on non-mipmapped textures only

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi/gfx9: update can_sample_z/s flags
Marek Olšák [Thu, 27 Oct 2016 18:45:15 +0000 (20:45 +0200)]
radeonsi/gfx9: update can_sample_z/s flags

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi/gfx9: pass correct parameters to buffer_get_handle
Marek Olšák [Thu, 27 Oct 2016 18:25:37 +0000 (20:25 +0200)]
radeonsi/gfx9: pass correct parameters to buffer_get_handle

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi/gfx9: update si_set_optimal_micro_tile_mode
Marek Olšák [Wed, 26 Oct 2016 22:13:50 +0000 (00:13 +0200)]
radeonsi/gfx9: update si_set_optimal_micro_tile_mode

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi/gfx9: don't check array_mode for allowing TC-compatible HTILE
Marek Olšák [Sun, 6 Nov 2016 21:31:49 +0000 (22:31 +0100)]
radeonsi/gfx9: don't check array_mode for allowing TC-compatible HTILE

GFX9 supports this with all modes except linear.

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi/gfx9: update HTILE/CMASK/FMASK allocators
Marek Olšák [Mon, 24 Oct 2016 00:34:04 +0000 (02:34 +0200)]
radeonsi/gfx9: update HTILE/CMASK/FMASK allocators

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi/gfx9: stub testdma - array_mode_to_string
Marek Olšák [Wed, 26 Oct 2016 14:44:06 +0000 (16:44 +0200)]
radeonsi/gfx9: stub testdma - array_mode_to_string

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi/gfx9: update r600_print_texture_info
Marek Olšák [Sun, 6 Nov 2016 15:40:28 +0000 (16:40 +0100)]
radeonsi/gfx9: update r600_print_texture_info

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agogallium/radeon: move pre-GFX9 radeon_bo_metadata.* to u.legacy.*
Marek Olšák [Sun, 6 Nov 2016 13:51:57 +0000 (14:51 +0100)]
gallium/radeon: move pre-GFX9 radeon_bo_metadata.* to u.legacy.*

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agowinsys/amdgpu: set num_tile_pipes, pipe_interleave_bytes for GFX9
Marek Olšák [Thu, 12 Jan 2017 01:47:05 +0000 (02:47 +0100)]
winsys/amdgpu: set num_tile_pipes, pipe_interleave_bytes for GFX9

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agowinsys/amdgpu: wire up new addrlib for GFX9
Marek Olšák [Mon, 24 Oct 2016 10:30:17 +0000 (12:30 +0200)]
winsys/amdgpu: wire up new addrlib for GFX9

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agowinsys/amdgpu: update amdgpu_addr_create for GFX9
Marek Olšák [Fri, 21 Oct 2016 11:31:40 +0000 (13:31 +0200)]
winsys/amdgpu: update amdgpu_addr_create for GFX9

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agowinsys/amdgpu: rename GFX6 surface functions
Marek Olšák [Thu, 20 Oct 2016 20:14:04 +0000 (22:14 +0200)]
winsys/amdgpu: rename GFX6 surface functions

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agogallium/radeon: add GFX9 surface info to radeon_surf
Marek Olšák [Sun, 23 Oct 2016 14:45:14 +0000 (16:45 +0200)]
gallium/radeon: add GFX9 surface info to radeon_surf

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agogallium/radeon: move pre-GFX9 radeon_surf.* members to radeon_surf.u.legacy.*
Marek Olšák [Sun, 23 Oct 2016 11:08:46 +0000 (13:08 +0200)]
gallium/radeon: move pre-GFX9 radeon_surf.* members to radeon_surf.u.legacy.*

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
7 years agoradeonsi/gfx9: allow Z16_UNORM for TC-compatible HTILE
Marek Olšák [Sat, 15 Oct 2016 13:06:01 +0000 (15:06 +0200)]
radeonsi/gfx9: allow Z16_UNORM for TC-compatible HTILE

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>