AlexIndustrial/mesa
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
141,245 Commits 1 Branch 0 Tags
3da170faaec2a57d76e668ab56781621c23c6703
Commit Graph

141245 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Marek Olšák c105c8ad2c winsys/amdgpu: remove another 8 bytes from amdgpu_winsys_bo by packing better 
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9809>
 2021-04-06 22:31:15 +00:00
Marek Olšák 815d98c22d gallium/pb: remove 8 bytes from pb_buffer by packing variables 
Only svga used usages beyond 16 bits. All other drivers are OK with 16 bits.

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9809>
 2021-04-06 22:31:15 +00:00
Marek Olšák 712b629abf gallium/pb: change pb_buffer::alignment to alignment_log2 
Alignments are always 2^n, so store n = log2(alignment). The next commit
will take advantage of the saved space.

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9809>
 2021-04-06 22:31:15 +00:00
Marek Olšák 3d0a4864ce winsys/amdgpu: add amdgpu_cs::ws to reduce dereferences 
Reviewed-by: Zoltán Böszörményi <zboszor@gmail.com>
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9809>
 2021-04-06 22:31:15 +00:00
Marek Olšák d5b877619d winsys/amdgpu: remove amdgpu_winsys_bo::ws 
Reviewed-by: Zoltán Böszörményi <zboszor@gmail.com>
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9809>
 2021-04-06 22:31:15 +00:00
Marek Olšák 65495e6caa radeon_winsys.h: add a winsys parameter to most winsys buffer functions 
This will allow removing the winsys pointer from buffers.

The amdgpu winsys adds dummy_ws to get radeon_winsys because there can be
no radeon_winsys around (e.g. while amdgpu_winsys is being destroyed), but
we still need some way to call buffer functions.

Reviewed-by: Zoltán Böszörményi <zboszor@gmail.com>
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9809>
 2021-04-06 22:31:15 +00:00
Marek Olšák aed8af5456 radeon_winsys.h: add a new function radeon_bo_reference that takes a winsys 
This will allow removing the winsys pointer from buffers.

Reviewed-by: Zoltán Böszörményi <zboszor@gmail.com>
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9809>
 2021-04-06 22:31:15 +00:00
Marek Olšák 522938f271 winsys/radeon: rename radeon_bo_reference -> radeon_ws_bo_reference 
radeon_bo_reference will be a new function in radeon_winsys.h.

Reviewed-by: Zoltán Böszörményi <zboszor@gmail.com>
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9809>
 2021-04-06 22:31:15 +00:00
Marek Olšák 6c6a39682e gallium/pb: pass an optional winsys pointer to the buffer destroy function 
This will allow removing the winsys pointer from buffers.

Reviewed-by: Zoltán Böszörményi <zboszor@gmail.com>
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9809>
 2021-04-06 22:31:15 +00:00
Lionel Landwerlin 33b2daab1a intel/compiler: lower bit sizes in NIR postprocessing 
It appears that between preprocess & postprocess some descriptor
lowering introduces 8bit types in the shader, so run the lower bit
size again to make sure we don't have any unsupported types in our
shader.

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: e06144a818 ("anv: Use 64bit_global_32bit_offset for SSBOs")
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/4478
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9705>
 2021-04-06 23:21:30 +03:00
Alyssa Rosenzweig d01628d24d pan/bi: Remove conversion lowering 
Inherited from Midgard, it's easier to just do at NIR->BIR time now that
we have a builder that doesn't suck. Impeded certain optimizations.

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10065>
 2021-04-06 19:35:33 +00:00
Alyssa Rosenzweig 266de379c2 pan/bi: Generalize f2i16, f2u16 
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10065>
 2021-04-06 19:35:33 +00:00
Alyssa Rosenzweig be74b84e6f pan/bi: Fill in some more conversions 
The trick is that downcasts are just swizzling out what you don't want,
so things like U32_TO_F16 can be synthesized as V2U16_TO_V2F16 with
src.h00

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10065>
 2021-04-06 19:35:33 +00:00
Alyssa Rosenzweig 435800e750 pan/bi: Lower swizzles 
Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10065>
 2021-04-06 19:35:33 +00:00
Alyssa Rosenzweig e3e2c5b594 pan/bi: Optimize MKVEC.v2i16 generation 
Prefer SWZ.v2i16 in cases where that's possible, it will be easier to
optimize later.

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10065>
 2021-04-06 19:35:33 +00:00
Alyssa Rosenzweig f9bc168bd3 pan/bi: Add bi_swz_16 helper 
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10065>
 2021-04-06 19:35:33 +00:00
Marek Olšák 6bc7a0b366 radeonsi: fix si_compute_copy_image if DCC decompression happens before a copy 
The buffer clear after DCC decompression overwrote user SGPRs that we set
here, leading to invalid image coordinates. Set them after we bind images.

This will actually be fixed by "radeonsi: don't decompress DCC for float
formats in si_compute_copy_image" as well.

Fixes: ad71ef9326 "radeonsi: don't use a constant buffer for the copy_image compute shader"
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/4583

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10061>
 2021-04-06 19:19:56 +00:00
Mike Blumenkrantz fc7ce9b636 zink: drop VK_PIPELINE_CREATE_DISABLE_OPTIMIZATION_BIT from compute path 
this was dropped a while ago from gfx but I failed to pick it up for compute

Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10062>
 2021-04-06 18:11:08 +00:00
Adam Jackson 84d6ee0920 mesa: Remove unused gl_config::mutableRenderBuffer 
This bit is only meaningful for EGL, where we in fact derive it from
whether EGL_KHR_mutable_render_buffer is enabled and don't actually look
at the DRI attribute.

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10045>
 2021-04-06 12:16:17 -04:00
Adam Jackson fa425511ed mesa: Make _mesa_initialize_visual return void 
The assertions here clearly never get hit, and if they did it'd mean we
were _trying_ to enable 64-bit Z or something. And checking only that
the accumulation has non-negative size is... curious. Just delete it and
clean up the callers.

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10045>
 2021-04-06 12:16:16 -04:00
Adam Jackson 9dd0d68d26 mesa: Remove unused _mesa_create_framebuffer 
One fewer gl_config in the API.

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10045>
 2021-04-06 12:16:16 -04:00
Adam Jackson a6f1d56b7b radeon: Exchange one curious idiom for another in radeonMakeCurrent 
radeonMakeCurrent is the only consumer of _mesa_create_framebuffer(),
and all we're doing here is making a dummy for the context to hang on to
when there is no current drawable. intelMakeCurrent uses
_mesa_get_incomplete_framebuffer() for this instead, which is
equivalent, and which matches the usage elsewhere in the tree.

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10045>
 2021-04-06 12:16:15 -04:00
Adam Jackson 6d81b9c972 gallium: Remove curious st_visual::no_config 
The gallium context creation API embeds storage for a whole st_visual
even if the context is to have no config. no_config was being used to
signal this, but we can just as easily check that the color format is
non-NONE.

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10045>
 2021-04-06 12:16:10 -04:00
Mike Blumenkrantz 421fbb5f5e zink: flag DYNAMIC resources as coherent 
the contents of these may change frequently, so we should skip all the
flush/invalidate

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9983>
 2021-04-06 12:29:56 +00:00
Mike Blumenkrantz 6ff6d01c37 zink: don't use cached mem for staging resources 
Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9983>
 2021-04-06 12:29:56 +00:00
Simon Ser 7499f84956 radv: fix format feature reporting for modifiers 
The format_feature_flags bitfield is derived from the modifier if
the tiling is set to VK_IMAGE_TILING_DRM_FORMAT_MODIFIER_EXT.
However radv will reset the tiling to either LINEAR or OPTIMAL if
the caller supplied a VkPhysicalDeviceImageDrmFormatModifierInfoEXT
in the chain.

Stop resetting the tiling, so that we can compute the correct feature
flags.

Signed-off-by: Simon Ser <contact@emersion.fr>
Fixes: 6c83e3ea98 ("radv: Add format modifier format queries.")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9978>
 2021-04-06 09:20:26 +00:00
Rhys Perry 292ac71a4a nir/lower_tex: handle deref casts 
A RDR2 shader has a undef->texture cast which is eventually optimized out.
Without handling NULL from nir_deref_instr_get_variable(), compiling this
shader will result in a crash.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Fixes: bc438c91d9 ("nir/lower_tex: ignore texture_index if tex_instr has deref src")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10038>
 2021-04-06 08:35:39 +00:00
Mike Blumenkrantz 00fc85a011 zink: reset queries when suspending if >50% of total pool is used 
this is a great time for resets since there's never a renderpass active,
so an early reset here may mean avoiding a renderpass split later

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10049>
 2021-04-06 02:01:07 +00:00
Mike Blumenkrantz f5dc4db35f zink: defer timestamp query pool resets to end_query 
we don't do a tc context unwrap in create_query(), so this isn't a good place
to be doing anything that needs the ctx

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10049>
 2021-04-06 02:01:07 +00:00
Mike Blumenkrantz 24923b4640 zink: check last_finished first in fence_finish early out case 
a fence's completed flag is irrelevant if we already know it has completed

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10048>
 2021-04-05 21:45:53 -04:00
Mike Blumenkrantz d3734ae2dd zink: relax maybe_flush batch count threshold 
10 batch states isn't actually that many if an app happens to be flushing
a lot, so let's multiply that to catch outlier cases

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10006>
 2021-04-05 22:45:52 +00:00
Mike Blumenkrantz 1ba6dcb40f zink: relax maybe_flush mem threshold 
10% of video mem ends up being way too aggressive and causes constant
stalls

really this is supposed to be a failsafe against ooming in a single batch, so
using 50% here ends up serving that purpose much more effectively

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10006>
 2021-04-05 22:45:52 +00:00
Mike Blumenkrantz e4426d7828 zink: split total_mem off to total_video_mem, use total_mem for tc 
tc needs the total host memory available, not the total video memory

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10006>
 2021-04-05 22:45:52 +00:00
Mike Blumenkrantz c624489968 zink: move cmdpool reset to batch state reset 
this is a more logical place for it

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9964>
 2021-04-05 22:34:16 +00:00
Mike Blumenkrantz a565ead64a zink: only use host mem for staging resources with linear tiling 
VK_EXT_4444_formats strikes again

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9985>
 2021-04-05 22:24:32 +00:00
Mike Blumenkrantz 6433661cda zink: use 2 variant to check image format props during create 
just to be pedantic

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9984>
 2021-04-05 22:14:26 +00:00
Mike Blumenkrantz 733e07565f zink: use vkGetPhysicalDeviceFormatProperties2 when available 
technically shouldn't matter but whatever

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9984>
 2021-04-05 22:14:26 +00:00
Mike Blumenkrantz 3fa7ce4d46 zink: use better usage flags for staging resources 
enforce linear bind more logically and mark with more accurate usage

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9982>
 2021-04-05 21:58:08 +00:00
Mike Blumenkrantz 12f93a7377 zink: hook up cs push constant for nir_intrinsic_load_work_dim 
Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9966>
 2021-04-05 17:38:16 -04:00
Mike Blumenkrantz 796e95d6a7 util/threaded_context: support pipe_context::set_sample_locations 
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10031>
 2021-04-05 20:52:45 +00:00
Alejandro Piñeiro 1e0a69afa7 vulkan: track number of bindings instead of max binding for CreateDescriptorSetLayout 
As that handles better, and more clear, the case of bindingCount being
zero. For the case of Anvil and Turnip, this avoids allocating a
non-needed binding when bindingCount is zero.

Inspired on radv, that was what it was doing so far.

Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/4526

Reviewed-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Hyunjun Ko <zzoon@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9905>
 2021-04-05 20:17:53 +00:00
Dave Airlie 5890c322e1 zink/ci: handle getting correct drisw driver. 
With zink not being autopicked, fixup the CI to pick the
correct drivers for zink testing, noting that xvfb really
doesn't want zink to be chosen.

Reviewed-by: Eric Anholt <eric@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9816>
 2021-04-05 18:46:16 +00:00
Dave Airlie 3e1698fe1b drisw: move zink down the list below the sw drivers. 
We don't ever want drisw path picking zink as the driver,
we can revisit this when the penny wrapper work gets further
along.

This selection causes systems with nvidia/intel dual-gpus
to try and pick the intel gpu for rendering in the nvidia
context if there is no nvidia GL driver or accel doesn't work.

This is a partial revert of the original commit.

Fixes: 4a3b42a717 ("drisw: Prefer hardware-layered sw-winsys drivers over pure sw")
Acked-by: Jesse Natalie <jenatali@microsoft.com>
Reviewed-by: Adam Jackson <ajax@redhat.com>
Acked-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9816>
 2021-04-05 18:46:16 +00:00
Bas Nieuwenhuizen 57511d1458 radv: Flush caches for shader read operations. 
As part of the fmask expand we very much read from the images as
well ...

Fixes: 8f8d72af55 ("radv: Use access helpers for flushing with meta operations.")
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10042>
 2021-04-05 18:34:53 +00:00
Danylo Piliaiev 0709a6b363 turnip: fix alignment of non-32b types in workgroup memory 
Fixes tests:
 dEQP-VK.spirv_assembly.instruction.compute.workgroup_memory.float16

Signed-off-by: Danylo Piliaiev <dpiliaiev@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10039>
 2021-04-05 17:31:11 +00:00
Samuel Pitoiset 8b80e8f832 radv: init CMASK/FMASK/DCC in parallel 
To remove bubbles during layout transitions from UNDEFINED, especially
with MSAA because we might have all.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10004>
 2021-04-05 16:21:12 +00:00
Jesse Natalie 91dbad7956 microsoft/compiler: Fix barrier flag for shared memory 
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10010>
 2021-04-05 15:58:03 +00:00
Juan A. Suarez Romero c1bd3d3afc ci/broadcom: update expected list 
Reviewed-by: Jose Maria Casanova Crespo <jmcasanova@igalia.com>
Signed-off-by: Juan A. Suarez Romero <jasuarez@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10041>
 2021-04-05 17:40:27 +02:00
Samuel Pitoiset 14b312b223 radv: handle implicit subpass dependencies per attachment 
From the Vulkan spec 1.2.172:

"If there is no subpass dependency from VK_SUBPASS_EXTERNAL to the
 first subpass that uses an attachment, then an implicit subpass
 dependency exists from VK_SUBPASS_EXTERNAL to the first subpass
 it is used in."

"Similarly, if there is no subpass dependency from the last subpass
 that uses an attachment to VK_SUBPASS_EXTERNAL, then an implicit
 subpass dependency exists from the last subpass it is used in to
 VK_SUBPASS_EXTERNAL."

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9714>
 2021-04-05 11:42:51 +00:00
Danylo Piliaiev 9b9ad8d66c docs: mark float_controls,float16_int8,16bit_storage as done on Turnip 
Signed-off-by: Danylo Piliaiev <dpiliaiev@igalia.com>
Acked-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9993>
 2021-04-05 09:51:46 +00:00