The implementation only takes the ownership after a successful import.
On import failure, the caller is going to handle the fd. Meanwhile,
amend a missing error code on an error path.
Fixes: 895d3399f7 ("lavapipe: add support for KHR_external_memory_fd")
Reviewed-by: Lucas Fryzek <lfryzek@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36204>
lvp hasn't used common device memory obj, and it allocates and imports
ahb on its own. Thus it has to implement the AHB export api itself.
- before: total 116, skip 66, pass 24, fail 26
- after: total 116, skip 66, pass 36, fail 14
Fixes: cebb2bf266 ("lavapipe: Add AHB extension")
Reviewed-by: Lucas Fryzek <lfryzek@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36204>
Currently tile swizzle can only be non zero for single plane
formats, for multi plane formats we always set PIPE_BIND_SHARED.
Luma only (Y400) JPG decode and encode with RGB input surface (EFC)
are the only two cases where we can get surface with tile swizzle
and ignoring it would result in corrupted output.
Cc: mesa-stable
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/13346
Acked-by: Ruijing Dong <ruijing.dong@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35647>
There is no need to have an own copy
Signed-off-by: Christian Gmeiner <cgmeiner@igalia.com>
Reviewed-by: Karmjit Mahil <karmjit.mahil@igalia.com>
Reviewed-by: Jose Maria Casanova Crespo <jmcasanova@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36218>
While the API spec does describe which flags _may_ be passed in, the
overall CL working group agreement is, that implementations should expect
random flags to be passed in as other implementations _may_ use them to
further restrict or allow image formats.
Also fix validation for importing GL objects while at it.
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36216>
This unblocks the main worker thread to keep submitting work to the driver
while we still have something waiting on the completion of batches sent to
the hardware and to signal completion to the attached events.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36158>
The runtime vk_android.h header has proper android detection inside, so
no need to wrap it with redundant android detection. Meanwhile, the enum
VK_EXTERNAL_MEMORY_HANDLE_TYPE_ANDROID_HARDWARE_BUFFER_BIT_ANDROID is
defined in the vulkan_core.h, so no need Android wrap either.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Acked-by: Rob Clark <robclark@freedesktop.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36151>
Otherwise we can have a case where binning VS uses more consts than
full VS (when safe variant is used for full VS), that will result in
a rendering issue because SP_VS_CONST_CONFIG.CONSTLEN is shared between
full and binning VS in PROGRAM_CONFIG state and gets the value from the
full VS.
There are two alternative solutions that can allow binning VS to always
use maximum constlen:
- Move constlen emission to per-XS config. This interferes
PROGRAM_CONFIG state which uploads consts and does SP_UPDATE_CNTL.
Consts would need to be uploaded after constlen is defined, while
SP_UPDATE_CNTL must be done before per-XS state is emitted.
Also having SP_UPDATE_CNTL in a draw state that is always DIRTY
isn't great.
Something didn't work out on A6XX, so this idea was dropped.
- Emit constlen again in VS_BINNING draw state. This seem to work
but also likely an undefined behaviour since constlen is changed
after some consts are uploaded.
Cc: mesa-stable
Signed-off-by: Danylo Piliaiev <dpiliaiev@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36203>
After commit 545727f97c (ci/android: Move ANDROID_CTS_MODULES to build
script, 2025-06-24) the comment about ANDROID_CTS_MODULES in
lvp-android-angle-android-cts-include.txt has become inaccurate.
Update the instructions to reflect the latest status.
Fixes: 545727f97c
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36187>
Thanks to the optimizations in the previous commits and further
increasing tests_per_group, we can now run more tests in a618-vk
while also reducing its parallelism.
Dropping `DEQP_FRACTION: 2` and only increasing the `fraction` of
the basic test set to 3 results in broader overall coverage.
Signed-off-by: Valentine Burley <valentine.burley@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36027>
The a618 jobs run on two different 8 thread devices, which is the optimal
value for FDO_CI_CONCURRENT on most of their jobs.
Signed-off-by: Valentine Burley <valentine.burley@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36027>
For import, requested size might be smaller than the whole blob mem
size. Currently, venus virtgpu backend tracks the initial import size
and sanity checks with later re-import size. So let's add a way to
request whole blob mem size
No need vtest backend changes since dma_buf import is not supported.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36128>
Prepare for adopting common runtime so that we don't leave unused vars
after dropping codes from vn_AllocateMemory. It's trying to minimize
code changes in the venus adoption since that'd be huge.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36128>
