freedreno/drm: Remove bo_del_or_recycle()

In prep for the next patch, where locking is swapped around to cover the
whole bo_del() path, decouple handling of the recycle-to-BO-cache path.

Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20918>

freedreno/drm: Detect zombie BOs

When importing from a GEM name or dmabuf fd, we can race with the final
unref of the same BO, in which case we can get a hit in the handle
table for an fd_bo that another thread is about to free().  Detect and
handle this case.

Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20918>

freedreno/drm: Add some ref/unref debugging

Helpful to catch common refcnt issues, like resurrecting a zombie
object.

Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20918>

freedreno: Don't sync timestamps while perfetto isn't running.

This may help with the regression in trace perf testing since enabling
perfetto on the test drivers.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20916>

dzn: Report as a software device for non-Windows

Fixes: 5f1b8b3e6c9 ("dzn: Use DXGI swapchains")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20939>

dzn: Don't recursively lock the physical device enum mutex

Fixes: cfa260cd274 ("dzn: Use common physical device list/enumeration helpers")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20939>

dzn: Fix Windows WSI

This was a merge conflict from the Win32 WSI DXGI swapchain changes.
I missed moving a new line of code that was added when rearranging
things for using the common helpers.

Fixes: cfa260cd ("dzn: Use common physical device list/enumeration helpers")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20944>

docs: update calendar for 22.3.4

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20942>

docs/relnotes: add sha256sum for 22.3.4

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20942>

docs: add release notes for 22.3.4

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20942>

freedreno/registers: Add RBBM_GPR0_CNTL for non-GMU operation

Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20925>

freedreno/a6xx: Add a few kernel regs/etc

Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20925>

virgl: remove unused virgl_encoder_inline_write

The only user was removed with

be8eeb3b59abf18d6b706016c5fa05b746d88f65
  virgl: remove unused virgl_transfer_inline_write

so drop this code too.

Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18543>

ir3, isaspec: add raw instruction to assembler/disassembler.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20789>

frontends/va: revert commit 0b02db30

revert commit 0b02db30 as it is not a
proper way to fix av1 decoding 10bit issue.

this is corresponding to the fix in
https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20870

Signed-off-by: Ruijing Dong <ruijing.dong@amd.com>
Reviewed-by: Boyuan Zhang <Boyuan.Zhang@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20911>

kopper: Do not free the given screen in initScreen implementation

The given screen is already freed by the caller in case a NULL-pointer is
returned by the implementation.

Cc: mesa-stable
Signed-off-by: Corentin Noël <corentin.noel@collabora.com>
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20933>

util/tests/cache_test: Skip Cache.List if not supported

FOZ_DB_UTIL_DYNAMIC_LIST depends on inotify support

Fixes: 3b69b67545b ("util/fossilize_db: add runtime RO foz db loading via FOZ_DBS_DYNAMIC_LIST")

Signed-off-by: Juston Li <justonli@google.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20865>

util/fossilize_db: add ifdef for inotify header

FOZ_DB_UTIL_DYNAMIC_LIST is defined if the inotify header was
detected.

Fixes: 3b69b67545b ("util/fossilize_db: add runtime RO foz db loading via FOZ_DBS_DYNAMIC_LIST")

Signed-off-by: Juston Li <justonli@google.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20865>

nir: Add algebraic optimization for VKD3D-Proton fp32->fp16 conversion.

VKD3D-Proton DXBC f32 to f16 conversion implements a float conversion using PackHalf2x16.
Because the spec does not specify a rounding mode, it emits a sequence to ensure
D3D-like behaviour for infinity.

When we know the current backend has pack_half_2x16_rtz_split,
we can eliminate the extra sequence.

Fossil DB stats on GFX11:
Totals from 835 (0.62% of 134913) affected shaders:
VGPRs: 49368 -> 49224 (-0.29%)
CodeSize: 5341956 -> 5124564 (-4.07%)
Instrs: 1024062 -> 987041 (-3.62%)
Latency: 6530956 -> 6465120 (-1.01%); split: -1.01%, +0.00%
InvThroughput: 908189 -> 870253 (-4.18%)
VClause: 18704 -> 18702 (-0.01%); split: -0.02%, +0.01%
SClause: 33406 -> 33284 (-0.37%); split: -0.38%, +0.01%
Copies: 67440 -> 65992 (-2.15%); split: -2.15%, +0.00%
Branches: 18498 -> 18465 (-0.18%)
PreSGPRs: 38409 -> 38331 (-0.20%)
PreVGPRs: 44089 -> 43834 (-0.58%)

Note, some fossils are from before this pattern was added to VKD3D-Proton,
so the above may not reflect real-world impact.

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Acked-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Reviewed-by: Jason Ekstrand <jason.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/15838>

nir: Lower pack_half_2x16_split to RTZ if available.

Constant folding always uses RTNE for pack_half_2x16_split, but some
backends implement it with RTZ.

Lowering to RTZ when available ensures that the behaviour will be
consistent between constant folding and the backend.

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Acked-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Reviewed-by: Jason Ekstrand <jason.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/15838>

radv, aco, ac: Implement pack_half_2x16_rtz_split.

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Acked-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Reviewed-by: Jason Ekstrand <jason.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/15838>

nir: Add pack_half_2x16_rtz_split opcode.

Same as pack_half_2x16_rtz_split, but always uses RTZ mode.
Note that pack_half_2x16 rounding mode is unspecified.

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Acked-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Reviewed-by: Jason Ekstrand <jason.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/15838>

intel/fs: drop FS_OPCODE_UNIFORM_PULL_CONSTANT_LOAD_GFX7

We can lower FS_OPCODE_UNIFORM_PULL_CONSTANT_LOAD into other more
generic sends and drop this internal opcode.

The idea behind this change is to allow bindless surfaces to be used
for UBO pulls and why it's interesting to be able to reuse
setup_surface_descriptors(). But that will come in a later change.

No shader-db changes on TGL & DG2.

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20416>

lima/ci: Add more piglit unsupported tests to skip

It is not an exhaustive list but it helps by reducing the bulk of
"Failed to create waffle_context for OpenGL [34].x" errors in the logs
by thousands of occurrences and those are probably not going to be
needed.

Signed-off-by: Erico Nunes <nunes.erico@gmail.com>
Acked-by: Daniel Stone <daniels@collabora.com>
Reviewed-by: David Heidelberg <david.heidelberg@collabora.com>
Reviewed-by: Vasily Khoruzhick <anarsoul@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20909>

llvmpipe: Ensure floating point SSE state is reset regardless of the write mask.

The code emitted by lp_build_fpstate_set to reset the FP state could be
jumped over when the write mask was zero, leading to denormals not being
flushed to zero.

Spotted by Roland Scheidegger.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20901>

radv: fix ignoring graphics shader stages that don't need to be imported

If a shader stage is already imported from a library it should be
properly ignored.

Fixes recent CTS dEQP-VK.pipeline.fast_linked_library.misc.unused_shader_stages*.

Fixes: c8765c5244a ("radv: ignore shader stages that don't need to be imported with GPL")
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20899>

radv: fix creating libraries with PS epilog and all CB states as dynamic

It's legal to create a library with FRAGMENT_OUTPUT_INTERFACE and with
all CB states as dynamic, in this case the PS epilog should be dynamic.

This fixes a bunch of regressions while running Zink/RADV CTS with
RADV_PERFTEST=gpl.

Zink is the final boss.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20882>

v3dv: add a cl_advance_and_end helper

For the common case where we're emitting packet we don't need to
update the cl_out pointer and then store the result in cl->next,
we can directly update cl->next.

This shows a small improvement in vkoverhead's scores for basic
draw tests.

Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20897>

dzn: Support Vulkan 1.2

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20919>

dzn: Ensure we don't mix DSV+simultaneous-access

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20919>

dzn: Enhanced barriers fixes/workarounds

Fix: Acquire/release should have one valid access/sync and one set
to none.

Workaround: D3D doesn't like simultaneous access resources leaving
COMMON layout, nor does it like setting UAV/RTV access bits for the
COMMON layout.
Use UNDEFINED -> UNDEFINED layout transitions, where the access bits
just aren't validated.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20919>

dzn: Always do clears with copies on non-graphics queues

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20919>

dzn: Support float control

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20919>

dzn: Cache GPUVA for buffers

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20919>

dzn: Handle separate stencil usage

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20919>

dzn: Fix dynamic rendering clear load op for non-multiview

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20919>

microsoft/compiler: Support float controls

Float controls are emitted as function attributes on the entrypoint.
These function attributes are not the standard build-in LLVM kind, but
are strings, which the DXIL backend didn't know how to emit. So, this
change adds string attribute support and uses it for fp32 ftz/preserve.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20919>

aco: Remove dynamic VS input loads.

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20733>

radv: Lower dynamic VS inputs in NIR.

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20733>

aco: Enable constant exec mask based optimization on compute shaders.

We know for sure exec is initially -1 when the shader always has full subgroups.

Fossil DB stats on GFX11:
Totals from 3884 (2.88% of 134913) affected shaders:
SpillSGPRs: 1673 -> 1697 (+1.43%); split: -1.67%, +3.11%
SpillVGPRs: 2316 -> 2310 (-0.26%); split: -0.65%, +0.39%
CodeSize: 19584436 -> 19567156 (-0.09%); split: -0.13%, +0.04%
Scratch: 217088 -> 216832 (-0.12%)
Instrs: 3784596 -> 3780303 (-0.11%); split: -0.15%, +0.03%
Latency: 39971204 -> 39794967 (-0.44%); split: -0.47%, +0.03%
InvThroughput: 7885552 -> 7801247 (-1.07%); split: -1.14%, +0.07%
VClause: 74654 -> 74611 (-0.06%); split: -0.07%, +0.01%
SClause: 103139 -> 103043 (-0.09%); split: -0.13%, +0.04%
Copies: 279864 -> 281995 (+0.76%); split: -0.72%, +1.48%
Branches: 92082 -> 92084 (+0.00%); split: -0.03%, +0.03%
PreSGPRs: 155637 -> 149491 (-3.95%)

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20670>

radv, aco: Add uses_full_subgroups to compute shader info.

Allow the compiler to assume that the shader always has full subgroups,
meaning that the initial EXEC mask is -1 in all waves (all lanes enabled).
This assumption is incorrect for ray tracing and internal (meta) shaders
because they can use unaligned dispatch.

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20670>

radv: Get rid of app_shaders_internal.

This will make sure the internal field is set to true for internal
shaders which are initialized outside of radv_device_init_meta.

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Friedrich Vock <friedrich.vock@gmx.de>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20670>

vulkan/rmv: Remove delta parameter from dump helpers

It was unused.

Acked-by: Konstantin Seurer <konstantin.seurer@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20820>

vulkan/rmv: Use the timestamp divisor instead of a hardcoded value

Acked-by: Konstantin Seurer <konstantin.seurer@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20820>

radv/rmv: Correct timestamp shifting

The shifting was off-by-one compared to how it is done in the kernel. Also, excess_length needs to be casted to uint64_t to prevent zeroing everything except the 5 LSBs.

Fixes: abf3bcd6 ("radv: Add RMV resource tracking")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20820>

radv/rmv: Log bo destruction before freeing it

Fixes: abf3bcd6 ("radv: Add RMV resource tracking")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20820>

radv/rmv: Avoid more CPU unmap deadlocks

Fixes: 8d0e6c02 ("radv: Add RMV tracing utilities")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20820>

radv/rmv: Also check the other pid field

Sometimes it seems like this field contains the correct pid instead.

Fixes: 8d0e6c02 ("radv: Add RMV tracing utilities")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20820>

ci: Add manual rules variations to disable irrelevant driver jobs.

If you're only affecting one or a couple of drivers, it would be nice if
your pipeline buttons on the web UI weren't full of manual run buttons for
all the other drivers.

This is a bunch of duplicated lines, but less than it could have been now
that we have !references.

In some of these cases (i915g, nouveau, etnaviv), we have no non-manual
jobs for those drivers, so I could have just rewritten the original
"driver-rules" to "driver-manual-rules".  I decided to keep things
consistent between drivers, though, because this is all esoteric enough to
readers already without making different drivers' rules look different.

Fixes: #4891
Acked-by: David Heidelberg <david.heidelberg@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17445>

ci: Move the performance jobs' allow_failure:true to the gl rules.

This helps clarify what's going on with these magic manual jobs.

Acked-by: David Heidelberg <david.heidelberg@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17445>

vulkan/wsi/wayland: improve same gpu detection

Some compositor like KWin do not return the render node.

v2: Make sure we test if only drm_info.hasPrimary is true (Jason)

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: db42ed1e04 ("vulkan/wsi/wl: correctly find whether the compositor uses the same GPU")
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/8139
Reviewed-by: Simon Ser <contact@emersion.fr>
Reviewed-by: Daniel Stone <daniels@collabora.com>
Reviewed-by: Jason Ekstrand <jason.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20914>

mesa: fix out of bounds stack access on big endian

The texture format code relies on a python-generated atlas of structs
that describe a lookup table for texture swizzling. Many of these
texture formats contain the index "6" used for this lookup. The 6th
index just so happens to represent a "don't care" value, however the
out of bounds read is still best to be avoided. The address sanitizer
finds this issue pretty immediately but it only shows up on big endian
because the textures don't need this on little.

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20846>

zink: Fix non debug builds failing to compile on

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20917>

intel/fs: Fix src and dst types of LOAD_PAYLOAD ACP entries during copy propagation.

The ACP entries created by copy propagation to track the implied
copies of LOAD_PAYLOAD instructions don't model the behavior of
LOAD_PAYLOAD correctly, since (as of 41868bb6824c6106a55c844) header
moves are implicitly retyped to UD and the destination of non-header
copies implicitly uses the same type as the corresponding source, even
though the ACP entries created for such copies could incorrectly
represent a type conversion, which can lead to mis-optimization of the
program.

According to Marcin, this fixes the func.mesh.ext.workgroup_id.task.q0
crucible test.

Fixes: 41868bb6824c6106a55c844 ("i965/fs: Rework the fs_visitor LOAD_PAYLOAD instruction")
Reported-by: Marcin Ślusarz <marcin.slusarz@intel.com>
Tested-by: Marcin Ślusarz <marcin.slusarz@intel.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18980>

docs: Lower Python requirement to 3.6

Issue 7716 bumped the Python requirement from 3.6 to 3.7 for two
reasons. They are now non-issues.

  - Usage of f-strings. But f-strings are a 3.6 feature.
  - Usage of `from __future__ import annotations`. But the previous
    commit removes that usage.

See: https://gitlab.freedesktop.org/mesa/mesa/-/issues/7716
See: 5cf862fbb9 ("docs: update Python requirement to 3.7")
Acked-by: Eric Engestrom <eric@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20237>

util/glsl2spirv: Fix build with Python 3.6

ChromeOS still uses Python 3.6, but the glsl2spirv script uses module
'__future__.annotations', introduced in Python 3.7. Fix the build by
removing module, but otherwise preserve the type annotations.

Fixes: 949c3b55db ("util/glsl2spirv: add type annotations")
Acked-by: Eric Engestrom <eric@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20237>

zink: validation ci updates

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20891>

zink: force GENERAL layout for all fb attachments with image binds

this is incredibly stupid, but KHR-GL46.texture_view.coherency does all
kinds of rasterization discard draws with fb attachments bound as images,
and there's no other sane way to catch it dynamically

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20891>

zink: update sampler layout when detecting feedback loop for first time

ensure synchronization between descriptor layout and layout used here

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20891>

zink: return false for implicit feedback loop check with image binds

this can only be GENERAL layout so it's not a feedback loop ever

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20891>

zink: don't skip repeated handling feedback loops

these can persist across multiple draws, so return true if it's still
a feedback loop

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20891>

zink: move barrier jit to zink_context.c

this gives access to all the good descriptor functions and enshrinkens
the c++ file size

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20891>

zink: fix VK_DYNAMIC_STATE_LINE_WIDTH usage

add a special tracker here to set the state only when necessary

Fixes: 659c39fafbb ("zink: rework primitive rasterization type logic")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20886>

radv/winsys: prefix all error messages with RADV

RadeonSI prints the exact same message and it can be confusing.
All other error messages in the winsys are prefixed now.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20904>

radv: Move constant flushing check out to callers.

Approximately 10% improvement in CPU overhead score on 3900X.

Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20655>

intel/ds: Nuke intel_ds_queue::queue_id

queue_id is not used anywhere.

Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Reviewed-by: Marcin Ślusarz <marcin.slusarz@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20888>

hasvk: Nuke anv_queue:index_in_family

This value is also stored in vk_queue, so we can nuke from anv_queue.

Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Reviewed-by: Marcin Ślusarz <marcin.slusarz@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20888>

anv: Nuke anv_queue:index_in_family

This value is also stored in vk_queue, so we can nuke from anv_queue.

Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Reviewed-by: Marcin Ślusarz <marcin.slusarz@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20888>

intel: Add kmd_type parameter to necessary intel_gem.h functions

Here adding kmd_type parameter to
intel_gem_read_render_timestamp(), intel_gem_can_render_on_fd() and
intel_gem_supports_protected_context().
Those 3 functions will have Xe implementations, the other functions
in intel_gem.h will not be called by Xe code paths so not adding
kernel_driver_type to it.

Reviewed-by: Marcin Ślusarz <marcin.slusarz@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20773>

intel: Add intel_kmd_type parameter to intel_engine_get_info()

This will be needed to execute proper function to fetch engines from
kernel mode driver.

Reviewed-by: Marcin Ślusarz <marcin.slusarz@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20773>

intel/dev: Detect what is the kernel mode driver loaded

Intel GPUs will have a new kernel mode driver, support for some
platforms will overlap so we will need to detect in runtime what is
the kernel mode driver loaded for each device.

Reviewed-by: Marcin Ślusarz <marcin.slusarz@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20773>

radeonsi: correctly declare YUV420_10 RT Format support for AV1

This appears to have been an oversight. AV1 Main profile requires
support for both 8bit and 10bit, and so we should always declare
support for the YUV420_10 RT Format. This support then cascades
into supporting the appropriate surface formats and meets expectations
of vaapi clients (especially ffmpeg based) on how to detect support
for these formats.

Note that the commit [0b02db3007] was also made with the intention of
fixing this problem, but it does so in a non-idiomatic way. With that
change, there is still no declared YUV420_10 RT Format, and instead
the P010 surface format is reported under the YUV420 RT Format. This
is not going to work with all vaapi clients. I recommend that this
commit be reverted.

Signed-off-by: Philip Langdale <philipl@overt.org>
Reviewed-by: Ruijing Dong <ruijing.dong@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20870>

radv: Merge the leaf and internal converter

We have everything we need in the internal one already so we can just
encode leaf nodes there. Since this functionality isn't split anymore,
the shader was renamed to "encode".

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20768>

radv: skip creation of null TLAS for null winsys

This won't work because there are no memory types.

Fixes fossilize replay with NULL winsys.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Friedrich Vock <friedrich.vock@gmx.de>
Reviewed-by: Konstantin Seurer <konstantin.seurer@gmail.com>
Fixes: 31ca19589fe ("radv: Create a null TLAS as meta state")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20878>

ci/b2c: uprev to b2c v0.9.9

No functional change expected, just a minor reduction in size compared
to the previous version and updated components.

Signed-off-by: Martin Roukala (né Peres) <martin.roukala@mupuf.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20806>

ci/android: add android to the ci

Add android to the ci, so we can test mesa on it.

Add debian/x86_test-android and test-android jobs.
One build the container to run tests, and the other execute the tests.

Android is executed on top of cuttlefish VM with virgl.
Mesa libs are replaced on the cuttlefish image, and deqp and deqp-runner
are used to execute tests.

Co-developed-by: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Signed-off-by: Helen Koike <helen.koike@collabora.com>
Reviewed-by: Sergi Blanch Torné <sergi.blanch.torne@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20019>

ci: export artifacts from debian-android

export mesa build from this job, so it can be picked up by an android
testing job.

Co-developed-by: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Signed-off-by: Helen Koike <helen.koike@collabora.com>
Reviewed-by: Sergi Blanch Torné <sergi.blanch.torne@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20019>

ci: debian-android compile virgl

compile virgl so we can test android on a vm.

Signed-off-by: Helen Koike <helen.koike@collabora.com>
Reviewed-by: Sergi Blanch Torné <sergi.blanch.torne@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20019>

ci: compile deqp-runner for android

To allow testing mesa on android, we need deqp-runner compiled for it,
so we can add android on the ci.

Signed-off-by: Helen Koike <helen.koike@collabora.com>
Reviewed-by: Sergi Blanch Torné <sergi.blanch.torne@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20019>

ci: compile deqp for android

To allow testing mesa on android, we need deqp compiled for it, so we
can add android on the ci.

Add patch that redirect logs to stdout instead of logcat.
Compile deqp to allow running it directly from command line instead of
Android activity.

Co-developed-by: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Signed-off-by: Helen Koike <helen.koike@collabora.com>
Reviewed-by: Sergi Blanch Torné <sergi.blanch.torne@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20019>

ci: Use NDK 25b to build for the Android ABI level 33

25b is the currently supported NDK, as per the below:

https://github.com/android/ndk/wiki

Signed-off-by: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Signed-off-by: Helen Koike <helen.koike@collabora.com>
Reviewed-by: Sergi Blanch Torné <sergi.blanch.torne@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20019>

ci/android: move sdk version and ndk to a job variable

make it easier to update sdk version and ndk without the need to make
changes all over the code.

Suggested-by: David Heidelberg <david.heidelberg@collabora.com>
Signed-off-by: Helen Koike <helen.koike@collabora.com>
Reviewed-by: Sergi Blanch Torné <sergi.blanch.torne@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20019>

ci: Build for Android with libbacktrace=false

This is needed for using this build inside Android systems with VNDK
enabled, as in recent versions of it libbacktrace isn't available to
libraries in /vendor.

Signed-off-by: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Signed-off-by: Helen Koike <helen.koike@collabora.com>
Reviewed-by: Sergi Blanch Torné <sergi.blanch.torne@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20019>

android: Cleanup unneeded headers from the sync stub

These were probably copy pasted from the previous unified stub.

Signed-off-by: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Signed-off-by: Helen Koike <helen.koike@collabora.com>
Reviewed-by: Sergi Blanch Torné <sergi.blanch.torne@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20019>

android: Make libbacktrace optional again

For the same reason why we used to have USE_LIBBACKTRACE with the old
Android makefiles, allow to build Mesa without linking to it.

In recent VNDK versions, libbacktrace isn't available.

When building without linking libbacktrace, for some reason some symbols
related to C++ exception handling are exposed. Allow them in the symbols
check script.

Signed-off-by: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Signed-off-by: Helen Koike <helen.koike@collabora.com>
Reviewed-by: Sergi Blanch Torné <sergi.blanch.torne@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20019>

android: allow system = 'android' on cross file

Allow building with system = 'android' on the cross file.
This avoids creating symlinks for the libs.

Suggested-by: Roman Stratiienko <r.stratiienko@gmail.com>
Suggested-by: David Heidelberg <david.heidelberg@collabora.com>
Signed-off-by: Helen Koike <helen.koike@collabora.com>
Reviewed-by: Sergi Blanch Torné <sergi.blanch.torne@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20019>

ci: move patches to patches directory

Group all patches in a directory.
Improve organization to organize the current and future patches.

Signed-off-by: Helen Koike <helen.koike@collabora.com>
Reviewed-by: Sergi Blanch Torné <sergi.blanch.torne@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20019>

ci/debian-android: move pkgconfig paths to the cross file

Move pkgconfig paths out of the job script into the cross file.
Use pkg_config_libdir on the cross file instead.
Which makes more sense and cleaner code.

Suggested-by: David Heidelberg <david.heidelberg@collabora.com>
Suggested-by: Roman Stratiienko <r.stratiienko@gmail.com>
Signed-off-by: Helen Koike <helen.koike@collabora.com>
Reviewed-by: Sergi Blanch Torné <sergi.blanch.torne@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20019>

panfrost: fix off-by-one when exporting format modifiers

`count` should not be incremented before the check, because it causes
the modifiers array to be filled starting from position 1 instead of 0.

This bug causes one less format modifier to be available than would
otherwise be expected, which could then lead to a dmabuf query failing
in situations where a supported modifier wouldn't be advertised.

It also causes garbage data to be advertised as a modifier in position 0
of the array, although this is not very likely to cause issues.

Fixes: 2a1217513 ("panfrost: Implement panfrost_query_dmabuf_modifiers")
Cc: mesa-stable
Signed-off-by: Italo Nicola <italonicola@collabora.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20879>

radv: fix RADV_DEBUG=hang with multiple cmdbuffer per submission

With RADV_DEBUG=hang, there is only one cmdbuffer per submission and
this has been broken recently.

This fixes a segfault when generating GPU hang reports.

Fixes: 76deaa1b1a9 ("radv: Refactor command buffer handling in radv_queue_submit_normal.")
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20860>

radv: Reduce descriptor pool allocation for alignment.

Since we can now rely on this due to the stricter layout code.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20847>

radv: Strictly limit alignment needed within a descriptor set.

By doing two passes we limit the number of times we need to have a gap
after a 16-byte descriptor to align for an image descriptor.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20847>

radeonsi/gfx11: fix ge_cntl programming

gfx11 renamed PRIM_GRP_SIZE to VERTS_PER_SUBGRP but another change was
was missed.

Update our code based on PAL's UniversalCmdBuffer::CalcGeCntl function
(especially useVgtOnchipCntlForTess being false for gfx11).

Fixes: 25a66477d02 ("radeonsi/gfx11: register changes")
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20728>

radeonsi/gfx11: clamp PRIM_GRP_SIZE

Legal range of values is [1, 256].

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20728>

winsys/amdgpu: use DMA_BUF_SET_NAME_B if available

Give a name to dma-buf. This name appears in /sys/kernel/debug/dma_buf/bufinfo
and could be useful to debug dma-buf:

Dma-buf Objects:
size     flags    mode     count    exp_name ino      name
00606208 00000002 00080007 00000003 drm 00192014 2321705-glxgears

The name is only added to non-shared buffer, to avoid overwriting
an existing name when exporting an imported buffer (otherwise all
dma-buf will pretend to be created by XWayland).

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20728>

drm-uapi/dma-buf.h: use __u32/__u64 types

Otherwise we might get build errors: https://gitlab.freedesktop.org/mesa/mesa/-/jobs/34886940

Reviewed-by: Jason Ekstrand <jason.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20728>

iris: Add missing untyped data port flush on PIPELINE_SELECT

This is needed when switching away from GPGPU mode.  See the previous
commit for anv.  This is not likely to make a practical difference for
iris because it never switches back and forth between modes like anv.

Fixes: 172e0b0ebff ("iris: Update PIPELINE_CONTROL flush when switching pipeline mode in TGL+")
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: José Roberto de Souza <jose.souza@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20774>

anv: Add missing untyped data port flush on PIPELINE_SELECT

See the comments in emit_apply_pipe_flushes().  Flushing HDC is not
sufficient in GPGPU mode, and we need to set the untyped data port flush
bit as well.

Fixes many dEQP-VK failures with INTEL_COMPUTE_CLASS=1 on Alchemist.

Fixes: 1067ec90a59 ("anv: Update PIPELINE_CONTROL flush when switching pipeline mode in TGL+")
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: José Roberto de Souza <jose.souza@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20774>

radeonsi: remove some shadow reg optimization for bf1 game

This patch removes below shadow reg optimization. This is done for
Vega64 battlefield 1 crash when shadow regs enabled.

  + reset only dirty states with buffers in si_pm4_reset_emitted()
  + various draw states in si_begin_new_gfx_cs()

v2: remove first_cs parameter from si_pm4_reset_emitted() (Marek Olšák)

Signed-off-by: Yogesh Mohan Marimuthu <yogesh.mohanmarimuthu@amd.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18301>

radv: allow NULL initial_preamble_cs in radv_amdgpu_winsys_cs_submit_sysmem()

In case of mcbp, shadowed_regs is initialized early in radv_queue_init()
function by submitting the command buffer. The command buffer is submitted in
radv_init_shadowed_regs_buffer_state() function. When RADV_DEBUG=noibs is used
radv_amdgpu_winsys_cs_submit_sysmem() function is used to submit command buffer.
radv_amdgpu_winsys_cs_submit_sysmem() crashes here because initial_preamble_cs
is NULL. This patch fixes the radv_amdgpu_winsys_cs_submit_sysmem() function
to support NULL initial_preamble_cs.

Signed-off-by: Yogesh Mohan Marimuthu <yogesh.mohanmarimuthu@amd.com>
Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18301>