mesa.git
6 years agodrirc: set allow_glsl_cross_stage_interpolation_mismatch for more games
Tapani Pälli [Wed, 20 Dec 2017 07:23:55 +0000 (09:23 +0200)]
drirc: set allow_glsl_cross_stage_interpolation_mismatch for more games

Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Suggested-by: Darius Spitznagel <d.spitznagel@goodbytez.de>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=104288
Acked-by: Kenneth Graunke <kenneth@whitecape.org>
6 years agoanv: disallow VK_REMAINING_ARRAY_LAYERS in vkCmdClearAttachments()
Samuel Iglesias Gonsálvez [Tue, 19 Dec 2017 07:59:36 +0000 (08:59 +0100)]
anv: disallow VK_REMAINING_ARRAY_LAYERS in vkCmdClearAttachments()

Vulkan spec doesn't specify that VK_REMAINING_ARRAY_LAYERS is allowed
in the passed VkClearRect struct.

Signed-off-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
6 years agonvc0/ir: change textureGrad to always use lane 0 as the tex origin
Ilia Mirkin [Wed, 16 Aug 2017 04:34:43 +0000 (00:34 -0400)]
nvc0/ir: change textureGrad to always use lane 0 as the tex origin

Thanks to Karol Herbst for the debugging / tracing work that led to this
change.

Move to using lane 0 as the "work" lane for the texture. It is unclear
why this helps, as that computation should be identical to doing it in
the "correct" lane with the properly adjusted quadops.

In order to be able to use the lane 0 result, we also have to ensure
that lane 0 contains the proper array/indirect/shadow values.

This applies to Fermi and Kepler. Maxwell+ may or may not need fixing,
but that lowering logic is separate.

Fixes KHR-GL45.texture_cube_map_array.sampling

Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
6 years agobroadcom/vc5: Add missing setting of the UIF XOR disable flag in textures.
Eric Anholt [Tue, 19 Dec 2017 22:23:06 +0000 (14:23 -0800)]
broadcom/vc5: Add missing setting of the UIF XOR disable flag in textures.

Most piglit textures happened to work out by RGBW not changing in that
bit, but it did cause failures in RGBA16F fbo-generatemipmap-formats.

6 years agobroadcom/vc5: Clean up the comment and code around level 0 UIF.
Eric Anholt [Tue, 19 Dec 2017 22:20:19 +0000 (14:20 -0800)]
broadcom/vc5: Clean up the comment and code around level 0 UIF.

I wrote this early in driver development, and our UIF handling is much
better now.

6 years agobroadcom/vc5: Simplify the tiling calculations.
Eric Anholt [Tue, 19 Dec 2017 22:08:18 +0000 (14:08 -0800)]
broadcom/vc5: Simplify the tiling calculations.

The mb_tile_layout table was just the utile_w/h times two, so reuse the
utile code instead.

6 years agobroadcom/vc5: Return the depth in all components of depth textures.
Eric Anholt [Thu, 16 Nov 2017 20:01:13 +0000 (12:01 -0800)]
broadcom/vc5: Return the depth in all components of depth textures.

Apparently gallium's u_blitter wants depth from at least the .z component,
and other swizzling appears to apply on top of that.  Fixes
fbo-generatemipmap-formats failures with depth formats.

6 years agobroadcom/vc5: Enable decompressing RGTC for desktop GL support.
Eric Anholt [Fri, 15 Dec 2017 22:40:43 +0000 (14:40 -0800)]
broadcom/vc5: Enable decompressing RGTC for desktop GL support.

This matches freedreno's behavior.

6 years agobroadcom/vc5: Use u_transfer_helper for MSAA mappings.
Eric Anholt [Fri, 15 Dec 2017 22:40:24 +0000 (14:40 -0800)]
broadcom/vc5: Use u_transfer_helper for MSAA mappings.

6 years agobroadcom/vc5: Start adding support for rendering to Z32F_S8X24_UINT.
Eric Anholt [Wed, 6 Dec 2017 02:58:41 +0000 (18:58 -0800)]
broadcom/vc5: Start adding support for rendering to Z32F_S8X24_UINT.

There may be some more RCL work to be done (I think I need to split my Z/S
stores when doing separate stencil), but this gets piglit's "texwrap
GL_ARB_depth_buffer_float" working.

v2: Unwrap the z32f_wrapper before calling the helper, rather than having
    the helper have a callback.
v3: Rebase on Rob Clark's u_transfer_helper instead

6 years agofreedreno: add debug flag to force high priority context
Rob Clark [Fri, 25 Aug 2017 11:58:40 +0000 (07:58 -0400)]
freedreno: add debug flag to force high priority context

Mainly for testing, FD_MESA_DEBUG=hiprio will force high priority
contexts.

Signed-off-by: Rob Clark <robdclark@gmail.com>
6 years agofreedreno: context priority support
Rob Clark [Thu, 24 Aug 2017 14:22:24 +0000 (10:22 -0400)]
freedreno: context priority support

For devices (and kernels) which support different priority ringbuffers,
expose context priority support.

Signed-off-by: Rob Clark <robdclark@gmail.com>
6 years agogallium: plumb context priority through to driver
Rob Clark [Wed, 23 Aug 2017 18:39:55 +0000 (14:39 -0400)]
gallium: plumb context priority through to driver

Signed-off-by: Rob Clark <robdclark@gmail.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Andres Rodriguez <andresx7@gmail.com>
Reviewed-by: Wladimir J. van der Laan <laanwj@gmail.com>
6 years agointel/compiler/gen10: Disable push constants.
Rafael Antognolli [Mon, 18 Dec 2017 23:23:11 +0000 (15:23 -0800)]
intel/compiler/gen10: Disable push constants.

We still have gpu hangs on Cannonlake when using push constants, so
disable them for now until we have a proper fix for these hangs.

v2: Add warning message when creating context too.

Signed-off-by: Rafael Antognolli <rafael.antognolli@intel.com>
Cc: Ben Widawsky <ben@bwidawsk.net>
Cc: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Ben Widawsky <ben@bwidawsk.net>
6 years agoradv: properly load unused gl_LocalInvocationID/gl_WorkGroupID components
Samuel Pitoiset [Mon, 18 Dec 2017 21:06:38 +0000 (22:06 +0100)]
radv: properly load unused gl_LocalInvocationID/gl_WorkGroupID components

F1 2017 looks good now.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
6 years agoradv: do not add extra SGPR when push constants are not used
Samuel Pitoiset [Mon, 18 Dec 2017 18:38:58 +0000 (19:38 +0100)]
radv: do not add extra SGPR when push constants are not used

This is not because the vertex stage needs some push constants
that other stages need them too. This should reduce the number
of loaded SGPRs in some situations.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
6 years agoradv: change the needs_push_constants logic
Samuel Pitoiset [Mon, 18 Dec 2017 18:38:57 +0000 (19:38 +0100)]
radv: change the needs_push_constants logic

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
6 years agoradv: store pipeline stages that need push constants
Samuel Pitoiset [Mon, 18 Dec 2017 18:38:56 +0000 (19:38 +0100)]
radv: store pipeline stages that need push constants

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
6 years agoradv: remove one useless check in ac_nir_shader_info_pass()
Samuel Pitoiset [Mon, 18 Dec 2017 18:38:55 +0000 (19:38 +0100)]
radv: remove one useless check in ac_nir_shader_info_pass()

pipeline->layout can't be NULL now.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
6 years agoradv: remove one useless check in radv_flush_constants()
Samuel Pitoiset [Mon, 18 Dec 2017 18:38:54 +0000 (19:38 +0100)]
radv: remove one useless check in radv_flush_constants()

pipeline->layout can't be NULL now.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
6 years agoradv: add assertions to make sure pipeline layout objects are valid
Samuel Pitoiset [Mon, 18 Dec 2017 18:38:53 +0000 (19:38 +0100)]
radv: add assertions to make sure pipeline layout objects are valid

The spec requires it.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
6 years agoradv: create pipeline layout objects for all meta operations
Samuel Pitoiset [Mon, 18 Dec 2017 18:38:52 +0000 (19:38 +0100)]
radv: create pipeline layout objects for all meta operations

They are dummy objects but the spec requires layout to not be
NULL, this just makes sure we are creating valid pipeline layout
objects. This will allow us to remove some useless checks.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
6 years agoradv: Use a sort for rebuilding the sparse buffer bo list.
Bas Nieuwenhuizen [Tue, 19 Dec 2017 08:01:32 +0000 (09:01 +0100)]
radv: Use a sort for rebuilding the sparse buffer bo list.

It uses slightly more memory (though still bounded by the number
of mapped ranges), but gives less quadratic behavior.

Cuts 4 minutes from the runtime of the CTS *.sparse.* tests.

Reviewed-by: Eric Engestrom <eric.engestrom@imgtec.com>
6 years agofreedreno/ir3: handle VTXID_BASE for indirect draws
Rob Clark [Mon, 18 Dec 2017 20:09:49 +0000 (15:09 -0500)]
freedreno/ir3: handle VTXID_BASE for indirect draws

Need to do some gymnastics to copy the parameter from the indirect
parameters buffer to uniform so shader sees the correct base-vertex-id.

Fixes ./bin/arb_draw_indirect-vertexid on a5xx and probably a4xx too.

Signed-off-by: Rob Clark <robdclark@gmail.com>
6 years agofreedreno/ir3: add ctx->mem_to_mem()
Rob Clark [Mon, 18 Dec 2017 20:06:37 +0000 (15:06 -0500)]
freedreno/ir3: add ctx->mem_to_mem()

For dealing with indirect-draw + gl_VertexID, we'll introduce another
case where we need to use CP_MEM_TO_MEM.  Rather than adding more
if(a5xx)/else make this a ctx vfunc.

Signed-off-by: Rob Clark <robdclark@gmail.com>
6 years agofreedreno/a5xx: use vertex_id_zero_base
Rob Clark [Mon, 18 Dec 2017 18:34:18 +0000 (13:34 -0500)]
freedreno/a5xx: use vertex_id_zero_base

Cmdstream traces from blob make it clear that the blob driver dev's
*think* a5xx has a real (non-zero-based) vtxid.  But reality claims
differently.

Fixes ./bin/gl-3.2-basevertex-vertexid and probably others.

This means draw-indirect is going to need some gymnastics to copy
base-vertex into uniform.  (a4xx probably needs that too.)

Signed-off-by: Rob Clark <robdclark@gmail.com>
6 years agor600: clear compressed flags in image state on unbind.
Dave Airlie [Tue, 19 Dec 2017 05:36:53 +0000 (05:36 +0000)]
r600: clear compressed flags in image state on unbind.

If we aren't binding an image, clear the compressed flags.

This fixes a segfault seen with an apitrace.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=104331
Signed-off-by: Dave Airlie <airlied@redhat.com>
6 years agoswr: Account for index_bias in offsets
George Kyriazis [Thu, 14 Dec 2017 18:01:53 +0000 (12:01 -0600)]
swr: Account for index_bias in offsets

When calculating buffer offsets for client buffers account for info.index_bias.

Fixes the follow piglit tests:
arb_draw_elements_base_vertex-drawelements-user_varrays
arb_draw_elements_base_vertex-negative-index-user_varrays

Reviewed-By: Bruce Cherniak <bruce.cherniak@intel.com>
6 years agor600: only reported tgsi ir compute support on evergreen+
Dave Airlie [Mon, 18 Dec 2017 21:38:09 +0000 (21:38 +0000)]
r600: only reported tgsi ir compute support on evergreen+

This fixes a crash on r600/r700.

Signed-off-by: Dave Airlie <airlied@redhat.com>
6 years agoradv: Advertise sync fd import and export.
Bas Nieuwenhuizen [Mon, 18 Dec 2017 20:09:19 +0000 (21:09 +0100)]
radv: Advertise sync fd import and export.

Passes dEQP-VK.*.sync_fd.*

Reviewed-by: Dave Airlie <airlied@redhat.com>
6 years agoradv: Implement sync file import/export for fences & semaphores.
Bas Nieuwenhuizen [Mon, 18 Dec 2017 20:02:05 +0000 (21:02 +0100)]
radv: Implement sync file import/export for fences & semaphores.

Reviewed-by: Dave Airlie <airlied@redhat.com>
6 years agoradv/amdgpu: wrap sync fd import/export.
Bas Nieuwenhuizen [Mon, 18 Dec 2017 19:33:07 +0000 (20:33 +0100)]
radv/amdgpu: wrap sync fd import/export.

Reviewed-by: Dave Airlie <airlied@redhat.com>
6 years agoac/nir: fix lds store for patch outputs.
Dave Airlie [Mon, 18 Dec 2017 06:53:44 +0000 (16:53 +1000)]
ac/nir: fix lds store for patch outputs.

This wasn't calculating the correct value, this along with
a nir patch fixes a regression in:
dEQP-VK.tessellation.shader_input_output.barrier

Fixes: 043d14db30a (ac/nir: don't write tcs outputs to LDS that aren't read back.)
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Signed-off-by: Dave Airlie <airlied@redhat.com>
6 years agonir/linking: always set the used_across_stages/outputs_read bits
Dave Airlie [Mon, 18 Dec 2017 06:49:43 +0000 (16:49 +1000)]
nir/linking: always set the used_across_stages/outputs_read bits

If we don't remap and output this code would trample the outputs
read bits.

This fixes a regression in
dEQP-VK.tessellation.shader_input_output.barrier

Fixes: 1c9c42d16b4c (nir: add varying component packing helpers)
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Signed-off-by: Dave Airlie <airlied@redhat.com>
6 years agospirv: Relax the validation conditions of OpSelect
Jason Ekstrand [Fri, 15 Dec 2017 03:53:05 +0000 (19:53 -0800)]
spirv: Relax the validation conditions of OpSelect

The Talos Principle contains shaders with an OpSelect between two
vectors where the condition is a scalar boolean.  This is technically
against the spec bout nir_builder gracefully handles it by splatting
out the condition to all the channels.  So long as the condition is a
boolean, just emit a warning instead of failing.

Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=104246

6 years agoradv: remove useless radv_cmask_info::base_address_reg
Samuel Pitoiset [Fri, 15 Dec 2017 17:54:00 +0000 (18:54 +0100)]
radv: remove useless radv_cmask_info::base_address_reg

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
6 years agoamd/common: add ac_vgt_gs_mode() helper
Samuel Pitoiset [Fri, 15 Dec 2017 14:37:19 +0000 (15:37 +0100)]
amd/common: add ac_vgt_gs_mode() helper

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
6 years agoamd/common: add ac_get_cb_shader_mask() helper
Samuel Pitoiset [Fri, 15 Dec 2017 14:37:18 +0000 (15:37 +0100)]
amd/common: add ac_get_cb_shader_mask() helper

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
6 years agoRevert "radv: do not load unused gl_LocalInvocationID/gl_WorkGroupID components"
Samuel Pitoiset [Fri, 15 Dec 2017 15:01:56 +0000 (16:01 +0100)]
Revert "radv: do not load unused gl_LocalInvocationID/gl_WorkGroupID components"

This reverts commit 2294d35b243dee15af15895e876a63b7d22e48cc.

We can't do this without adjusting the input SGPRs/VGPRs logic.
For now, just revert it. I will send a proper solution later.

It fixes a rendering issue in F1 2017 that CTS didn't catch up.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Tested-by: Alex Smith <asmith@feralinteractive.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
6 years agoradv: port merge tess info from anv
Dave Airlie [Mon, 18 Dec 2017 05:05:52 +0000 (15:05 +1000)]
radv: port merge tess info from anv

anv merges the tess info correctly, but radv wasn't doing this.

This fixes hangs in
dEQP-VK.tessellation.winding.default_domain.hlsl_triangles_ccw

Fixes: 60fc0544e0 (radv/pipeline: handle tessellation shader compilation)
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Signed-off-by: Dave Airlie <airlied@redhat.com>
6 years agoradv: Add external fence support.
Bas Nieuwenhuizen [Mon, 27 Nov 2017 23:28:14 +0000 (00:28 +0100)]
radv: Add external fence support.

Reviewed-by: Dave Airlie <airlied@redhat.com>
6 years agoradv: Implement VK_KHR_external_fence_fd.
Bas Nieuwenhuizen [Mon, 27 Nov 2017 23:21:12 +0000 (00:21 +0100)]
radv: Implement VK_KHR_external_fence_fd.

Reviewed-by: Dave Airlie <airlied@redhat.com>
6 years agoradv: Implement fences based on syncobjs.
Bas Nieuwenhuizen [Mon, 27 Nov 2017 22:58:35 +0000 (23:58 +0100)]
radv: Implement fences based on syncobjs.

Reviewed-by: Dave Airlie <airlied@redhat.com>
6 years agoamd/common: Add detection of the syncobj wait/signal/reset ioctls.
Bas Nieuwenhuizen [Mon, 27 Nov 2017 00:06:11 +0000 (01:06 +0100)]
amd/common: Add detection of the syncobj wait/signal/reset ioctls.

First amdgpu bump after inclusion was 20 (which was done for local BOs).

Reviewed-by: Dave Airlie <airlied@redhat.com>
6 years agoradv: Add syncobj signal/reset/wait to winsys.
Bas Nieuwenhuizen [Mon, 27 Nov 2017 00:02:42 +0000 (01:02 +0100)]
radv: Add syncobj signal/reset/wait to winsys.

Reviewed-by: Dave Airlie <airlied@redhat.com>
6 years agoconfigure/meson: Bump libdrm_amdgpu version requirement.
Bas Nieuwenhuizen [Sat, 16 Dec 2017 23:51:58 +0000 (00:51 +0100)]
configure/meson: Bump libdrm_amdgpu version requirement.

For the radv dependencies on syncobj signal/reset.

Reviewed-by: Dave Airlie <airlied@redhat.com>
6 years agoandroid: fix vulkan driver build
Tapani Pälli [Tue, 12 Dec 2017 08:01:57 +0000 (10:01 +0200)]
android: fix vulkan driver build

fixes undefined references by adding missing wsi common API

Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
6 years agoandroid: fix undefined references to futex API
Tapani Pälli [Tue, 12 Dec 2017 08:01:56 +0000 (10:01 +0200)]
android: fix undefined references to futex API

Fixes: f98a2768ca "mesa: Add new fast mtx_t mutex type for basic use cases"
Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
6 years agodocs: mark GL4.3 as finished for r600
Dave Airlie [Mon, 18 Dec 2017 04:28:07 +0000 (04:28 +0000)]
docs: mark GL4.3 as finished for r600

Still only on fp64 supported hw.

6 years agor600: export robust buffer access
Dave Airlie [Fri, 3 Nov 2017 01:53:36 +0000 (11:53 +1000)]
r600: export robust buffer access

Signed-off-by: Dave Airlie <airlied@redhat.com>
6 years agor600: export GLSL 430
Dave Airlie [Fri, 3 Nov 2017 01:52:26 +0000 (11:52 +1000)]
r600: export GLSL 430

Signed-off-by: Dave Airlie <airlied@redhat.com>
6 years agor600/cs: add compute support to caps
Dave Airlie [Fri, 3 Nov 2017 01:30:12 +0000 (11:30 +1000)]
r600/cs: add compute support to caps

Signed-off-by: Dave Airlie <airlied@redhat.com>
6 years agor600: always flush between gfx and compute
Dave Airlie [Fri, 24 Nov 2017 00:51:35 +0000 (10:51 +1000)]
r600: always flush between gfx and compute

This is in no way optimal, but there seems to be some problems
mixing at the moment, lots of hangs, it is possible, just need
to figure out more magic.

Signed-off-by: Dave Airlie <airlied@redhat.com>
6 years agor600: fix unused variable warning
Dave Airlie [Mon, 18 Dec 2017 04:29:19 +0000 (04:29 +0000)]
r600: fix unused variable warning

Signed-off-by: Dave Airlie <airlied@redhat.com>
6 years agoradv: Fix multi-layer blits.
Bas Nieuwenhuizen [Sun, 17 Dec 2017 22:53:37 +0000 (23:53 +0100)]
radv: Fix multi-layer blits.

We did not set the layer correctly for the dst, as we would keep
using the base layer. Same for the source image.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=102710
CC: <mesa-stable@lists.freedesktop.org>
Reviewed-by: Dave Airlie <airlied@redhat.com>
6 years agofreedreno/a5xx: add a5xx blitter
Rob Clark [Thu, 23 Nov 2017 16:58:31 +0000 (11:58 -0500)]
freedreno/a5xx: add a5xx blitter

FD_MESA_DEBUG=noblit to disable

Signed-off-by: Rob Clark <robdclark@gmail.com>
6 years agofreedreno: add generic blitter
Rob Clark [Wed, 22 Nov 2017 17:37:15 +0000 (12:37 -0500)]
freedreno: add generic blitter

Basically a clone of util_blitter_blit() but with special handling to
blit PIPE_BUFFER as a PIPE_TEXTURE_1D.

Signed-off-by: Rob Clark <robdclark@gmail.com>
6 years agofreedreno: add non-draw batches for compute/blit
Rob Clark [Fri, 24 Nov 2017 15:37:22 +0000 (10:37 -0500)]
freedreno: add non-draw batches for compute/blit

Get rid of "gmem" (ie. tiling) ringbuffer, and just emit setup commands
directly to "draw" ringbuffer for compute (and in future for blits not
using the 3d pipe).  This way we can have a simple flat cmdstream buffer
and bypass setup related to 3d pipe.

Signed-off-by: Rob Clark <robdclark@gmail.com>
6 years agofreedreno: track staging and shadow perf ctrs for the HUD
Rob Clark [Tue, 21 Nov 2017 18:20:53 +0000 (13:20 -0500)]
freedreno: track staging and shadow perf ctrs for the HUD

Signed-off-by: Rob Clark <robdclark@gmail.com>
6 years agofreedreno: staging upload transfers
Rob Clark [Mon, 20 Nov 2017 20:34:40 +0000 (15:34 -0500)]
freedreno: staging upload transfers

In the busy && !needs_flush case, we can support a DISCARD_RANGE upload
using a staging buffer.  This is a bit different from the case of mid-
batch uploads which require us to shadow the whole resource (because
later draws in an earlier tile happen before earlier draws in a later
tile).

Signed-off-by: Rob Clark <robdclark@gmail.com>
6 years agofreedreno: update generated headers
Rob Clark [Sat, 25 Nov 2017 19:10:34 +0000 (14:10 -0500)]
freedreno: update generated headers

Signed-off-by: Rob Clark <robdclark@gmail.com>
6 years agoanv: Remove unused variable.
Bas Nieuwenhuizen [Sat, 16 Dec 2017 21:02:11 +0000 (22:02 +0100)]
anv: Remove unused variable.

Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
6 years agoradeonsi: don't call force_dcc_off for buffers
Marek Olšák [Tue, 12 Dec 2017 21:21:13 +0000 (22:21 +0100)]
radeonsi: don't call force_dcc_off for buffers

This was undefined yet harmless behavior in LLVM.
Not anymore - it causes a hang now.

Cc: 17.3 <mesa-stable@lists.freedesktop.org>
Tested-by: Michel Dänzer <michel.daenzer@amd.com>
6 years agoisl: Don't require VALIGN_2 for R32G32B32_FLOAT on Haswell.
Kenneth Graunke [Fri, 15 Dec 2017 00:17:45 +0000 (16:17 -0800)]
isl: Don't require VALIGN_2 for R32G32B32_FLOAT on Haswell.

According to the RENDER_SURFACE_STATE internal documentation, the
R32G32B32_FLOAT restriction is marked "IVB" only.  We choose to apply
it to Ivybridge and Baytrail, but not Haswell.

Apparently fixes KHR-GL46.texture_size_promotion.functional on Haswell.

Changes these tests from crashing to skipping on Haswell:
- KHR-GL46.direct_state_access.textures_storage_multisample_2d_rgb32f
- KHR-GL46.direct_state_access.textures_storage_multisample_3d_rgb32f

Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
6 years agoradeon/uvd: add and manage render picture list
Boyuan Zhang [Fri, 15 Dec 2017 16:23:25 +0000 (11:23 -0500)]
radeon/uvd: add and manage render picture list

Create a list in decoder to store all render picture buffer pointers that
currently being used in reference picture lists.

During get message buffer call, check each pointer in render_pic_list[]
within given pic->ref[] list, remove pointer that no longer being used by
pic->ref[]. Then add current render surface pointer to the render_pic_list[]
and assign the associated index to result.curr_idx.

As a result, result.curr_idx will have the correct index to represent the
current render picture, instead of the previous increamenting values.

Signed-off-by: Boyuan Zhang <boyuan.zhang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
6 years agoradeon/vcn: add and manage render picture list
Boyuan Zhang [Fri, 15 Dec 2017 16:17:32 +0000 (11:17 -0500)]
radeon/vcn: add and manage render picture list

Create a list in decoder to store all render picture buffer pointers that
currently being used in reference picture lists.

During get message buffer call, check each pointer in render_pic_list[]
within given pic->ref[] list, remove pointer that no longer being used by
pic->ref[]. Then add current render surface pointer to the render_pic_list[]
and assign the associated index to result.curr_idx.

As a result, result.curr_idx will have the correct index to represent the
current render picture, instead of the previous increamenting values.

Signed-off-by: Boyuan Zhang <boyuan.zhang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
6 years agovl: remove is idr flag
Boyuan Zhang [Thu, 7 Dec 2017 21:13:51 +0000 (16:13 -0500)]
vl: remove is idr flag

Remove is_idr flag since not being used anymore.

Signed-off-by: Boyuan Zhang <boyuan.zhang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
6 years agost/va: directly use idr pic flag
Boyuan Zhang [Fri, 8 Dec 2017 23:22:25 +0000 (18:22 -0500)]
st/va: directly use idr pic flag

Remove is_idr flag, and use idr_pic_flag provided by vaapi directly

Signed-off-by: Boyuan Zhang <boyuan.zhang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
6 years agoradeon/vce: determine idr by pic type
Boyuan Zhang [Thu, 7 Dec 2017 21:10:13 +0000 (16:10 -0500)]
radeon/vce: determine idr by pic type

Vaapi encode interface provides idr frame flags, where omx interface doesn't.
Therefore, change to use picture type to determine idr frame, which will
work for both interfaces.

Signed-off-by: Boyuan Zhang <boyuan.zhang@amd.com>
Reviewed-by: Leo Liu <leo.liu@amd.com>
6 years agoradeon/vcn: determine idr by pic type
Boyuan Zhang [Thu, 30 Nov 2017 16:58:32 +0000 (11:58 -0500)]
radeon/vcn: determine idr by pic type

Vaapi encode interface provides idr frame flags, where omx interface doesn't.
Therefore, change to use picture type to determine idr frame, which will
work for both interfaces.

Signed-off-by: Boyuan Zhang <boyuan.zhang@amd.com>
Reviewed-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
6 years agoutil: scons: wire up the sha1 test
Emil Velikov [Thu, 14 Dec 2017 17:20:30 +0000 (17:20 +0000)]
util: scons: wire up the sha1 test

Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
Reviewed-by: Andres Gomez <agomez@igalia.com>
6 years agoswr/rast: Move more RTAI handling out of binner
Tim Rowley [Thu, 14 Dec 2017 19:49:56 +0000 (13:49 -0600)]
swr/rast: Move more RTAI handling out of binner

Reviewed-by: Bruce Cherniak <bruce.cherniak@intel.com>
6 years agoswr/rast: EXTRACT2 changed from vextract/vinsert to vshuffle
Tim Rowley [Thu, 14 Dec 2017 19:39:29 +0000 (13:39 -0600)]
swr/rast: EXTRACT2 changed from vextract/vinsert to vshuffle

Reviewed-by: Bruce Cherniak <bruce.cherniak@intel.com>
6 years agoswr/rast: Fix cache of API thread event manager
Tim Rowley [Wed, 13 Dec 2017 23:52:52 +0000 (17:52 -0600)]
swr/rast: Fix cache of API thread event manager

Reviewed-by: Bruce Cherniak <bruce.cherniak@intel.com>
6 years agoswr/rast: Replace VPSRL with LSHR
Tim Rowley [Tue, 12 Dec 2017 20:23:50 +0000 (14:23 -0600)]
swr/rast: Replace VPSRL with LSHR

Replace use of x86 intrinsic with general llvm IR instruction.

Generates the same final assembly.

Reviewed-by: Bruce Cherniak <bruce.cherniak@intel.com>
6 years agoswr/rast: Rework thread binding parameters for machine partitioning
Tim Rowley [Mon, 11 Dec 2017 23:45:58 +0000 (17:45 -0600)]
swr/rast: Rework thread binding parameters for machine partitioning

Add BASE_NUMA_NODE, BASE_CORE, BASE_THREAD parameters to
SwrCreateContext.

Add optional SWR_API_THREADING_INFO parameter to SwrCreateContext to
control reservation of API threads.

Add SwrBindApiThread() function to allow binding of API threads to
reserved HW threads.

Reviewed-by: Bruce Cherniak <bruce.cherniak@intel.com>
6 years agoswr/rast: Pull of RTAI gather & offset out of clip/bin code
Tim Rowley [Mon, 11 Dec 2017 21:51:46 +0000 (15:51 -0600)]
swr/rast: Pull of RTAI gather & offset out of clip/bin code

Reviewed-by: Bruce Cherniak <bruce.cherniak@intel.com>
6 years agoswr/rast: Remove no-op VBROADCAST of vID
Tim Rowley [Mon, 11 Dec 2017 14:38:46 +0000 (08:38 -0600)]
swr/rast: Remove no-op VBROADCAST of vID

Reviewed-by: Bruce Cherniak <bruce.cherniak@intel.com>
6 years agoswr/rast: SIMD16 Fetch - Fully widen 32-bit integer vertex components
Tim Rowley [Mon, 11 Dec 2017 05:54:30 +0000 (23:54 -0600)]
swr/rast: SIMD16 Fetch - Fully widen 32-bit integer vertex components

Also widen the 16-bit a 8-bit integer vertex component gathers to SIMD16.

Reviewed-by: Bruce Cherniak <bruce.cherniak@intel.com>
6 years agoswr/rast: Replace INSERT2 vextract/vinsert with JOIN2 vshuffle
Tim Rowley [Fri, 8 Dec 2017 23:33:23 +0000 (17:33 -0600)]
swr/rast: Replace INSERT2 vextract/vinsert with JOIN2 vshuffle

Reviewed-by: Bruce Cherniak <bruce.cherniak@intel.com>
6 years agoswr/rast: SIMD16 Fetch - Fully widen 16-bit float vertex components
Tim Rowley [Fri, 8 Dec 2017 19:59:19 +0000 (13:59 -0600)]
swr/rast: SIMD16 Fetch - Fully widen 16-bit float vertex components

Reviewed-by: Bruce Cherniak <bruce.cherniak@intel.com>
6 years agoswr/rast: SIMD16 Fetch - Fully widen 32-bit float vertex components
Tim Rowley [Fri, 8 Dec 2017 00:37:07 +0000 (18:37 -0600)]
swr/rast: SIMD16 Fetch - Fully widen 32-bit float vertex components

Reviewed-by: Bruce Cherniak <bruce.cherniak@intel.com>
6 years agoswr/rast: Pass prim to ClipSimd
Tim Rowley [Thu, 7 Dec 2017 23:54:40 +0000 (17:54 -0600)]
swr/rast: Pass prim to ClipSimd

Reviewed-by: Bruce Cherniak <bruce.cherniak@intel.com>
6 years agoswr/rast: Pull most of the VPAI manipulation out of the binner/clipper
Tim Rowley [Thu, 7 Dec 2017 17:59:45 +0000 (11:59 -0600)]
swr/rast: Pull most of the VPAI manipulation out of the binner/clipper

Move out of binner/clipper; hand them down from the frontend code instead.

Reviewed-by: Bruce Cherniak <bruce.cherniak@intel.com>
6 years agoswr/rast: Move GatherScissors to header
Tim Rowley [Wed, 6 Dec 2017 18:07:59 +0000 (12:07 -0600)]
swr/rast: Move GatherScissors to header

Reviewed-by: Bruce Cherniak <bruce.cherniak@intel.com>
6 years agoswr/rast: Rewrite Shuffle8bpcGatherd using shuffle
Tim Rowley [Wed, 6 Dec 2017 16:37:41 +0000 (10:37 -0600)]
swr/rast: Rewrite Shuffle8bpcGatherd using shuffle

Ease future code maintenance, prepare for folding simd8 and simd16 versions.

Reviewed-by: Bruce Cherniak <bruce.cherniak@intel.com>
6 years agoswr/rast: Convert gather masks to Nx1bit
Tim Rowley [Mon, 4 Dec 2017 21:16:13 +0000 (15:16 -0600)]
swr/rast: Convert gather masks to Nx1bit

Simplifies calling code, gets gather function interface closer to llvm's
masked_gather.

Reviewed-by: Bruce Cherniak <bruce.cherniak@intel.com>
6 years agoswr/rast: WIP - Widen fetch shader to SIMD16
Tim Rowley [Mon, 4 Dec 2017 00:49:29 +0000 (18:49 -0600)]
swr/rast: WIP - Widen fetch shader to SIMD16

Widen vertex gather/storage to SIMD16 for all component types.

Reviewed-by: Bruce Cherniak <bruce.cherniak@intel.com>
6 years agoswr/rast: Corrections to multi-scissor handling
Tim Rowley [Wed, 29 Nov 2017 21:14:20 +0000 (15:14 -0600)]
swr/rast: Corrections to multi-scissor handling

binner's GatherScissors() will be turned into a real gather in the not
too distant future.

Reviewed-by: Bruce Cherniak <bruce.cherniak@intel.com>
6 years agoswr/rast: Binner fixes for viewport index offset handling
Tim Rowley [Wed, 29 Nov 2017 16:46:49 +0000 (10:46 -0600)]
swr/rast: Binner fixes for viewport index offset handling

Reviewed-by: Bruce Cherniak <bruce.cherniak@intel.com>
6 years agoswr/rast: Remove unneeded copy of gather mask
Tim Rowley [Tue, 21 Nov 2017 17:05:08 +0000 (11:05 -0600)]
swr/rast: Remove unneeded copy of gather mask

Reviewed-by: Bruce Cherniak <bruce.cherniak@intel.com>
6 years agoi965: Allow old begin/end queryobj for gen4/5 with HW contexts
Chris Wilson [Thu, 23 Nov 2017 09:57:08 +0000 (09:57 +0000)]
i965: Allow old begin/end queryobj for gen4/5 with HW contexts

Since we have HW contexts on gen4/5, we could take advantage of them, as
done for gen6+ in commit e32cd5ffbb72 ("i965: Rely on hardware contexts
for query objects on Gen6+."), to only emit a pair of counters at
begin/end queryobj, rather than around every primitive. However, to keep
queryobj working in the meantime as we bringup support for HW ctx on
gen4/5, we can keep using the existing code.

References: e32cd5ffbb72 ("i965: Rely on hardware contexts for query objects on Gen6+.")
Cc: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
6 years agofreedreno: use u_transfer_helper
Rob Clark [Mon, 4 Dec 2017 14:15:27 +0000 (09:15 -0500)]
freedreno: use u_transfer_helper

Signed-off-by: Rob Clark <robdclark@gmail.com>
6 years agogallium/util: add u_transfer_helper
Rob Clark [Tue, 28 Nov 2017 15:47:06 +0000 (10:47 -0500)]
gallium/util: add u_transfer_helper

Add a new helper that drivers can use to emulate various things that
need special handling in particular in transfer_map:

 1) z32_s8x24.. gl/gallium treats this as a single buffer with depth
    and stencil interleaved but hardware frequently treats this as
    separate z32 and s8 buffers.  Special pack/unpack handling is
    needed in transfer_map/unmap to pack/unpack the exposed buffer

 2) fake RGTC.. GPUs designed with GLES in mind, but which can other-
    wise do GL3, if native RGTC is not supported it can be emulated
    by converting to uncompressed internally, but needs pack/unpack
    in transfer_map/unmap

 3) MSAA resolves in the transfer_map() case

v2: add MSAA resolve based on Eric's "gallium: Add helpers for MSAA
    resolves in pipe_transfer_map()/unmap()." patch; avoid wrapping
    pipe_resource, to make it possible for drivers to use both this
    and threaded_context.

Signed-off-by: Rob Clark <robdclark@gmail.com>
6 years agoi965: enable EXT_disjoint_timer_query extension
Tapani Pälli [Thu, 14 Dec 2017 11:53:10 +0000 (13:53 +0200)]
i965: enable EXT_disjoint_timer_query extension

Following dEQP cases pass:
   dEQP-EGL.functional.get_proc_address.extension.gl_ext_disjoint_timer_query
   dEQP-EGL.functional.client_extensions.disjoint

Piglit test 'ext_disjoint_timer_query-simple' passes with these changes.

No changes/regression observed in Intel CI.

Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
6 years agomesa: GL_EXT_disjoint_timer_query extension API bits
Tapani Pälli [Tue, 12 Dec 2017 12:46:13 +0000 (14:46 +0200)]
mesa: GL_EXT_disjoint_timer_query extension API bits

Patch adds GL_GPU_DISJOINT_EXT and enables to use timer queries when
EXT_disjoint_timer_query is enabled.

v2: enable extension only when EXT_disjoint_timer_query set

Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com> (v1)
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
6 years agoglapi: add GL_EXT_disjoint_timer_query
Tapani Pälli [Mon, 20 Nov 2017 06:36:52 +0000 (08:36 +0200)]
glapi: add GL_EXT_disjoint_timer_query

Most entrypoints already available via other extensions like
GL_EXT_occlusion_query_boolean, GL_EXT_timer_query.

Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
6 years agomesa: add DisjointOperation to gl_shared_state
Tapani Pälli [Mon, 20 Nov 2017 06:31:40 +0000 (08:31 +0200)]
mesa: add DisjointOperation to gl_shared_state

This state will be used by EXT_disjoint_timer_query. As first
usage, patch sets DisjointOperation true when gpu reset happens.

Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
6 years agobroadcom/vc5: Fix a typo in memcmp for sig unpack checking.
Eric Anholt [Thu, 14 Dec 2017 17:41:16 +0000 (09:41 -0800)]
broadcom/vc5: Fix a typo in memcmp for sig unpack checking.

This shockingly ended up working out, because only the first byte of *sig
is used and (sizeof(*sig) != 0) == 1.  Fixes a compiler warning.

Link: https://bugs.freedesktop.org/show_bug.cgi?id=104183
6 years agobroadcom/vc5: Enable NIR txd lowering on all txd instructions.
Eric Anholt [Wed, 22 Nov 2017 00:33:29 +0000 (16:33 -0800)]
broadcom/vc5: Enable NIR txd lowering on all txd instructions.

Fixes almost all of piglit's arb_shader_texture_lod grad tests, except for
the base -texgrad/texgradcube ones which fail on what appear to be
precision problems.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>