The extension allows returning 0 if a given rate is unknown, which
allows us to support this on all GPUs, but since the extension depends
on Vulkan 1.1, we only expose it on v10+ for now.
Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Lars-Ivar Hesselberg Simonsen <lars-ivar.simonsen@arm.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35155>
Adds pixel rate, texel rate, and fma rate to the model table for v9+
GPUs. This will be exposed via VK_ARM_shader_core_properties and
GL_ARM_shader_core_properties later.
Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Lars-Ivar Hesselberg Simonsen <lars-ivar.simonsen@arm.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35155>
Rather than splitting the GPU ID in two, let the GPU ID users do that
when they need.
We also rework the model detection to use a mask so we can customize
the mask if the version major/minor fields are required to differentiate
two GPUs with the same arch major/minor and product major.
Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Lars-Ivar Hesselberg Simonsen <lars-ivar.simonsen@arm.com>
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35155>
Similar to nir_lower_alu_width(), the callback can return the
desired number of components for a phi, or 0 for no lowering.
The previous behavior of nir_lower_phis_to_scalar() with lower_all=true
can be elicited via nir_lower_all_phis_to_scalar() while the previous
behavior with lower_all=false now corresponds to nir_lower_phis_to_scalar()
with NULL callback.
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Mel Henning <mhenning@darkrefraction.com>
Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35783>
The GLSL compiler always lowers inputs to temps for VS and GS, so exclude
them from driver support because the GLSL compiler will no longer do that
unconditionally. Thus, indirect VS and GS inputs are completely untested
and broken in a lot of drivers.
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35945>
Before doing register allocation, use information available from
the SSA representation to determine register pressure and to
spill registers. This spilling doesn't have to be perfect (the
register allocator is still allowed to spill) but it will be
much faster to do the SSA spilling than RA spilling. In general
this should vastly improve the performance of register allocation.
Reviewed-by: Mary Guillemard <mary.guillemard@collabora.com>
Acked-by: Boris Brezillon <boris.brezillon@collabora.com>
Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34446>
We're currently not using image layouts in PanVK so we can advertise
this extension without additional changes.
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: John Anthony <john.anthony@arm.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35911>
We have enough DUTs to increase the job parallelism to 8. At this level,
the runtime reported by deqp-runner averages about 8:30, which is below
the 10-minute target recommended by the docs.
Signed-off-by: Valentine Burley <valentine.burley@collabora.com>
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35913>
The JOB_OFFSET only actually affect the global id and not wg id.
In NIR common, we assume that if base wg isn't supported, it means that
global and wg id already contains it.
To follow the convention around, we remove the offset to assume
global id will need an offset added.
Alone, this doesn't change functionality as we always lower away global
id so far.
Signed-off-by: Mary Guillemard <mary.guillemard@collabora.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35909>
FEX is a 64-bit process potentially running x86 (32-bit) binary, in
which case the automatic user MMIO offset detection doesn't work, so
let's explicitly set the user MMIO offset when we can.
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Acked-by: Mary Guillemard <mary.guillemard@collabora.com>
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34573>
Enforce strictness while we still can. We can't enforce strictness
on AFBC(RGB) because the driver might be used by a compositor that
imports buffers from clients that were relying on the old behavior.
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Lars-Ivar Hesselberg Simonsen <lars-ivar.simonsen@arm.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35762>
Add all newly defined tracepoints, including meta, render,
dispatch/dispatch_indirect, barrier, and sync_wait.
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
Reviewed-by: Christoph Pillmayer <christoph.pillmayer@arm.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32693>
Define tracepoints for these higl-level api calls
- meta
- render
- dispatch and dispatch_indirect
- barrier
and these low-level cs cmds
- sync_wait
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
Reviewed-by: Christoph Pillmayer <christoph.pillmayer@arm.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32693>
We can rely on panvk_per_arch(queue_check_status) to detect device lost.
Because we no longer emit cs_sync32_add from finish_cs to increment
debug syncobj, if an instr between the last draw/dispatch and
end-of-stream causes a CS error, the CS error is ignored. This is fine
because the instr should have no side effect and the kernel emits
ERROR_BARRIER to recover from the CS error.
If that is undesirable, we can restore the old behavior by emitting
cs_sync64_add from finish_cs to increment regular syncobj (and fix
cs_progress_seqno_reg) when PANVK_DEBUG is set.
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
Reviewed-by: Lars-Ivar Hesselberg Simonsen <lars-ivar.simonsen@arm.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35606>
If panvk_cs_sync64::error is non-zero at the end of a primary cmdbuf,
copy the error to panvk_cs_subqueue_context::last_error.
Update panvk_per_arch(queue_check_status) to check the last_error and
treat it as device lost if non-zero.
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
Reviewed-by: Lars-Ivar Hesselberg Simonsen <lars-ivar.simonsen@arm.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35606>
Versions prior to the r1p0 revision of Mali-T760 has a bug that prevents
16x MSAA from working properly. Let's detect this GPU, and disable the
functionality for it.
It's a bit awkward that we can't easily do this as part of the pan_model
infrastructure, but we'd otherwise have to introduce a new quirk that
depends on the GPU revision, which isn't something the current pan_model
infrastructure is prepared for. So let's just do it in get_max_msaa()
instead.
And while we're at it, let's also make the max_4x_msaa quirk do the
MIN2() properly. This can currently only happen on V4 GPUs when the
pan_allow_128bit_rts_v4 driconf is enabled, but it should be fixed in
either case.
Fixes: 423f3fd485 ("panfrost: enable 8x and 16x msaa modes when supported")
Reviewed-by: Daniel Stone <daniels@collabora.com>
Reviewed-by: Lars-Ivar Hesselberg Simonsen <lars-ivar.simonsen@arm.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35871>
When we encode inline uniform data into the descriptor table, we should
skip the descriptors that are actually data instead of real descriptors.
Let's do that and also print the inlined data for convenience.
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Lars-Ivar Hesselberg Simonsen <lars-ivar.simonsen@arm.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35831>
Timestamps are made up of multiple results from the different subqueues,
decided by the target stage for the timestamp. Afterwards, when the
timestamp query is copied, those individual results will be combined
into the final result.
For the computation of the final result, each timestamp query also has
a separate info field, containing the info required to perform the copy.
Timestamps inside a renderpass which cover the fragment subqueue need
to be deferred until after the RUN_FRAGMENT job. This happens via a linked
list of timestamps in the subqueue context.
Once in each primary command buffer, the syncobjs of the queries completed
so far are signalled. The list of those queries is part of the subqueue
context as well.
Reviewed-by: Lars-Ivar Hesselberg Simonsen <lars-ivar.simonsen@arm.com>
Acked-by: Boris Brezillon <boris.brezillon@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34932>
This makes it possible to choose, at runtime, to skip the wait in the
dst subqueue of the barrier.
This is needed to remove unnecessary dependencies between subqueues when
no timestamp queries are used.
Reviewed-by: Lars-Ivar Hesselberg Simonsen <lars-ivar.simonsen@arm.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34932>
This makes it easier to implement internal syncs between subqueues later
in this MR.
Reviewed-by: Lars-Ivar Hesselberg Simonsen <lars-ivar.simonsen@arm.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34932>
When we ExecuteCommands, we can have the following situation:
secondary {
beginQuery
endQuery
}
primary {
beginRendering
executeCommands(secondary)
endRendering
}
See dEQP-VK.query_pool.concurrent_queries.secondary_command_buffer.
For the logic in EndRendering to correctly detect that an occlusion
query has ended inside this render pass, the oq state has to be
propagated back to the primary from the secondary.
Reviewed-by: Lars-Ivar Hesselberg Simonsen <lars-ivar.simonsen@arm.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35738>
If an occlusion query is started inside a renderpass with multiview, we
need to signal as many queries as there are bits set in the view mask.
It is enough to signal them, we don't have to change how we write the
occlusion query. All but the first of the n queries will return 0.
Reviewed-by: Lars-Ivar Hesselberg Simonsen <lars-ivar.simonsen@arm.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35738>
Recent nvidia hardware has a native instruction for
nir_intrinsic_vote_ieq but not for nir_intrinsic_vote_feq. So, split
this boolean into two so we can contol the lowering separately for each
instruction.
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35778>
Since we have already dropped the refcounting there, replacing with a
back pointer to the device memory makes it clear we don't hold any ref.
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35779>
This is mainly useful for enabling validation layers, but might also
be useful for a few apps that use very basic atomic operations. In
general these operations do not work properly in Bifrost, but they
work well enough to e.g. pass some piglit tests.
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35710>
I should really have performed a full run here rather than just copying
the failures from a recent run... poorly.
Fixes: 8e646739baf ("panfrost/ci: add new 8x/16x msaa fails on t760")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35716>
This test regularly takes close to an hour, and sometimes exceeds that.
Let's bump the timeout for it, to give it a better chance to finish.
This job is only run manually, so it shouldn't affect the time it takes
to merge changes.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35716>
We have more bits to encode the resource size there, so let's increase
the max limits.
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Lars-Ivar Hesselberg Simonsen <lars-ivar.simonsen@arm.com>
Reviewed-by: Christoph Pillmayer <christoph.pillmayer@arm.com>
Reviewed-by: Mary Guillemard <mary.guillemard@collabora.com>
Tested-by: Mary Guillemard <mary.guillemard@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35731>
v11 started extending the size/surface stride of a few descriptors to
allow bigger images. Add the necessary code to take that into account.
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Lars-Ivar Hesselberg Simonsen <lars-ivar.simonsen@arm.com>
Reviewed-by: Christoph Pillmayer <christoph.pillmayer@arm.com>
Reviewed-by: Mary Guillemard <mary.guillemard@collabora.com>
Tested-by: Mary Guillemard <mary.guillemard@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35731>
For the ZS emission logic, we split the descriptor in 3 parts: ZS, S and
CRC, so we can easily mix and match each variant with pan_merge(). For
RTs, we just define a layout per variant instead of overlays, which makes
dumps less verbose, and avoid the situation where overlapping fields
get accidentally overwritten in the desc emission logic.
While at it, add the `Surface stride hi` fiels to the ZS/Color target
descriptors so we're all set to bump the image size limit.
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Lars-Ivar Hesselberg Simonsen <lars-ivar.simonsen@arm.com>
Reviewed-by: Christoph Pillmayer <christoph.pillmayer@arm.com>
Reviewed-by: Mary Guillemard <mary.guillemard@collabora.com>
Tested-by: Mary Guillemard <mary.guillemard@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35731>
We currently don't use this, but it's good to have it defined anyway.
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Lars-Ivar Hesselberg Simonsen <lars-ivar.simonsen@arm.com>
Reviewed-by: Christoph Pillmayer <christoph.pillmayer@arm.com>
Reviewed-by: Mary Guillemard <mary.guillemard@collabora.com>
Tested-by: Mary Guillemard <mary.guillemard@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35731>
The enum has more than 3 values now, and the field is 3-bit.
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Lars-Ivar Hesselberg Simonsen <lars-ivar.simonsen@arm.com>
Reviewed-by: Christoph Pillmayer <christoph.pillmayer@arm.com>
Reviewed-by: Mary Guillemard <mary.guillemard@collabora.com>
Tested-by: Mary Guillemard <mary.guillemard@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35731>
