If someone were to remove the libraries that are needed for these,
`default` would simply not enable these tests, and the only thing we
could notice is that test jobs would suddenly take less time to run.
Instead, let's have a check to make sure dEQP's cmake has detected
everything and enabled these platforms.
Reviewed-by: David Heidelberg <david.heidelberg@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27041>
In the case both sample locations and rasterization samples is supported by a
driver as dynamic state, there is a case vk_multisample_sample_locations_state_init()
does not fill ms->sample_locations at all. In this case we need to check
this pointer when dereferencing it in vk_dynamic_graphics_state_init_ms().
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Reviewed-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27025>
Today, the only thing that this really affects is descriptor sets and
dynamic state as everything else is re-emitted almost every time.
However, as we add more dirtying, we'll need to be more and more careful
about stale state leaking across secondary command buffer executions.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27049>
For VK_EXT_graphics_pipeline_library, we need to be able to handle
missing descriptor sets. THis screws up dynamic buffers because we're
no longer guaranteed to be able to statically compute the index into
nvk_root_descriptor_table::dynamic_buffers while lowering descriptors.
Instead, we may need to look up the start index for the set in
nvk_root_descriptor_table::set_dynamic_buffer_start and compute from
there.
We also rework nvk_nir_lower_descriptors to take an array of (possibly
NULL) descriptor sets. This ensures that we don't make any assumptions
about the pipeline layout. It's also the interface we'll want for
implementing VK_EXT_shader_object.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27048>
This prepares us for VK_EXT_graphics_pipeline_library by allowing null
descriptor sets to be bound and handling holes in pipeline layouts. We
also add a set_dynamic_buffer_start map to the root descriptor table
which says where in dynamic_buffers each set starts. This can be used
by the pipeline layout lowering in the case where we can't statically
the dynamic buffer index for a binding.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27048>
There are paths were a v3d screen can be created instead of vc4 screen,
in a real rpi4 hardware. For instance, in the CI itself.
So assuming that it was only possible with the simulator was wrong.
This reverts commit c31be1f4ba.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27047>
Prior to 06b526de, the mesa format was used for these completeness checks.
That was to address the case where a *different* internal format selected
the *same* mesa format, and the texture shouldn't be considered compatible.
But this didn't address the case where the *same* internal format selected
a *different* mesa format, e.g. because the type passed to the TexImage
API was different.
An old WGL demo app called TexFilter.exe tries to redefine a mipped RGBA16
texture as RGBA8. This incorrect logic caused Mesa to try to copy the RGBA16
data from the smaller mips into the newly created RGBA8 data, because it
thought that the texture was still mip-complete, despite the format changing.
Cc: mesa-stable
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27023>
These use different tables but map to the same bits, so it is just
a matter of picking the right tables for the instruction.
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/26860>
ring_seqno_valid indicates a successful ring cmd submission, and can be
used to avoid invalid reply decoding due to failed submit alloc.
Otherwise, the garbled VkResult will mislead into initialization failure
instead of oom.
Below cts failure is fixed:
dEQP-VK.api.device_init.create_instance_device_intentional_alloc_fail.basic
Fixes: ec131c6e55 ("venus: use instance allocator for ring allocs")
Signed-off-by: Yiwei Zhang <zzyiwei@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27026>
While it technically was already only using an ir_instruction field, better make
sure that it is effectively an ir_variable using the dedicated method instead of
relying on an undefined behavior.
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Signed-off-by: Corentin Noël <corentin.noel@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/26144>
The parameter_lists_match_exact function was wrongly assuming that all the
elements were ir_variable when there can also be ir_dereference_variable elements.
Add case taking this into account.
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Signed-off-by: Corentin Noël <corentin.noel@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/26144>
The vdrm_execbuf() missed to set the seqno field for requests sent to host.
This causes vdrm_host_sync() to lock up due to the unset seqno in a case
where two or more threads are using vdrm_execbuf() and vdrm_send_req()
concurrently, like in this scenario:
thread1: vdrm_send_req() shmem->seqno=1 req->seqno=2
thread2: vdrm_execbuf() shmem->seqno=1 req->seqno=0
thread1: vdrm_host_sync() shmem->seqno=0 req->seqno=2
Fix the lockup by setting the seqno in vdrm_execbuf().
Signed-off-by: Dmitry Osipenko <dmitry.osipenko@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27021>
If we have to wait for a buffer to be released in swrast_update_buffers(),
we can leak it.
Take care to reuse the existing buffer if one is still available.
Signed-off-by: Derek Foreman <derek.foreman@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/26761>
