When the TS is shared all sharing instances must see the same status
information about the resource TS buffer. Add this information to the
shared TS metadata and make it take precedence over the internal
status tracking when the TS is shared.
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Reviewed-by: Christian Gmeiner <cgmeiner@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19964>
Handle imports/exports always deal with one specific level of
a resource, so the shared TS metadata should also be per level.
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Reviewed-by: Christian Gmeiner <cgmeiner@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19964>
Wrap the setting of the resource level TS validity into a
helper function to allow the implementation to change later.
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Reviewed-by: Christian Gmeiner <cgmeiner@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19964>
Add a small helper to get the validity of the TS buffer for
a resource level. We can drop the ts_size check in several
places, as we never set ts_valid to true if there is no TS
buffer.
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Reviewed-by: Christian Gmeiner <cgmeiner@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19964>
Add a small helper to transfer the age (seqno) from one resource level
to another.
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Reviewed-by: Christian Gmeiner <cgmeiner@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19964>
Add a small helper to mark a resource level as changed so the
seqno handling is hidden.
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Reviewed-by: Christian Gmeiner <cgmeiner@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19964>
Add a small helper to mark a resource level as flushed so the
seqno handling is hidden.
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Reviewed-by: Christian Gmeiner <cgmeiner@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19964>
If we sync/flush a full resource we can skip any level where the
target is of the same age as the source.
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Reviewed-by: Christian Gmeiner <cgmeiner@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19964>
A blit into a render target may destroy valid TS information, as the
destination TS isn't updated. Flush the blit destination when necessary
to make sure that all pending TS is resolved into the destination before
the blit is executed.
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Reviewed-by: Christian Gmeiner <cgmeiner@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19964>
Resource maps, blits and surfaces all target a specific level of a
resource, so they can have different ages. Move the seqnos tracking
the age to etna_resource_level.
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Reviewed-by: Christian Gmeiner <cgmeiner@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19964>
Thanks to our SSA-based RA, we only use nir_register for arrays, and we only
access array registers with dedicated moves anyway. So there's no reason to need
any fancy coalescing... we can just switch to register access intrinsics and
translate them to moves exactly like we would've done when getting srcs/dests
before.
This addresses the ir3 portion of #9051.
No shader-db changes with a (significant subset of) Rob's shader-db. (Some
shaders are affected by this change but not in any way that shows up in the
stats.)
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Connor Abbott <cwabbott0@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24126>
We expect that these will be lowered in NIR now.
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24005>
This is beneficial for intrinsics that do an algebraic
instruction such as bitfield extract on shader arguments,
because it allows NIR to be aware of these instructions and
optimize them together with other algebraic instructions in
the shader.
Currently, just handle subgroup_id and num_subgroups intrinsics.
More will be added here in the future.
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24005>
This is so we can just use the same function when it's zero.
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24005>
This commit adds a partial render command to job submission.
For geom only jobs we must always submit a pr command in case we
enter SPM. For now, for geom+frag jobs, we'll also always submit
a pr command event.
Signed-off-by: Karmjit Mahil <Karmjit.Mahil@imgtec.com>
Reviewed-by: Frank Binns <frank.binns@imgtec.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24138>
Now things are structured in sections, like the other xml files.
And elements within a section are sorted alphabetically.
Signed-off-by: Karmjit Mahil <Karmjit.Mahil@imgtec.com>
Reviewed-by: Frank Binns <frank.binns@imgtec.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24138>
- Update and add csbgen definitions to make the content of the
geom and frag km stream more obvious.
- Replace some of the hard coded constants with defines.
- Adds some static assert to make the provenance of definitions
more clear as well as making sure things fit properly.
Signed-off-by: Karmjit Mahil <Karmjit.Mahil@imgtec.com>
Reviewed-by: Frank Binns <frank.binns@imgtec.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24138>
Remove the mrt setup stuff since the EOT program only support
output registers for now. When implementing the tile buffer
support this change can be reverted, or things could be changed
to better fit with how the compiler wants things.
Signed-off-by: Karmjit Mahil <Karmjit.Mahil@imgtec.com>
Reviewed-by: Frank Binns <frank.binns@imgtec.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24138>
If the *first* BO is not marked as "written", but the same BO is marked
as "written" later, then it will be added twice, because the first
instance will have index_for_handle equal to 0.
Reviewed-by: José Roberto de Souza <jose.souza@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24108>
This is a temporary measure until the zeroed shaders are replaced with the real
ones. This avoids a VK_ERROR_OUT_OF_DEVICE_MEMORY error due to a zero sized
allocation.
Signed-off-by: Frank Binns <frank.binns@imgtec.com>
Reviewed-by: Karmjit Mahil <Karmjit.Mahil@imgtec.com>
Fixes: 1dfd535124 ("pvr: Setup SPM background object")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24139>
Use all the nir_legacy.h features to transition away from the deprecated
structures. shader-db is quite happy. I assume that's a mix of more aggressive
source mod usage and better coalescing (nir-to-tgsi doesn't copyprop).
total instructions in shared programs: 900179 -> 887156 (-1.45%)
instructions in affected programs: 562077 -> 549054 (-2.32%)
helped: 5198
HURT: 470
Instructions are helped.
total temps in shared programs: 91165 -> 91162 (<.01%)
temps in affected programs: 398 -> 395 (-0.75%)
helped: 21
HURT: 18
Inconclusive result (value mean confidence interval includes 0).
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Tested-by: Pavel Ondračka <pavel.ondracka@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24116>
Consider code like:
32x4 %2 = @load_interpolated_input (%1, %0 (0x0)) (base=0, component=0, dest_type=float32, io location=VARYING_SLOT_VAR0 slots=1 mediump) // Color
32x4 %3 = fabs %2
32x4 %4 = fsat %3
32x4 %5 = fsin %4
The existing logic would incorrectly tell the backend that both fabs and fsat
could be folded, and then half the shader disappears. Whoops. Fix by stopping
the folding in this case. I choose to do this check in the fsat rather than the
fabs because it's more straightforward (1 source vs N uses) but it's somewhat
arbitrary.
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Pavel Ondračka <pavel.ondracka@gmail.com>
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24116>
Consider the IR:
%0 = load_reg
%1 = fneg %0
%2 = ffloor %1
%3 = bcsel .., .., %1
Because the fneg has both foldable and non-foldable users, nir/legacy does not
fold the fneg into the load_reg. This ensures that the backend correctly emits a
dedicated fneg instruction (with the load_reg folded in) for the bcsel to use.
However, because the chasing helpers did not previously take other uses of a
modifier into account, the helpers would fuse in the fneg to the ffloor. Except
that doesn't work, because the load_reg instruction is supposed to be
eliminated. So we end up with broken chased IR:
1 = fneg r0
2 = ffloor -NULL
3 = bcsel, ..., 1
The fix is easy: only fold modifiers into ALU instructions if the modifiers can
be folded away. If we can't eliminate the modifier instruction altogether, it's
not necessarily beneficial to fold it anyway from a register pressure
perspective. So this is probably ok. With that check in place we get correct IR
1 = fneg r0
2 = ffloor 1
3 = bcsel, ..., 1
Fixes carchase/230.shader_test under softpipe.
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24116>
This removes all the users of the compiler enums, and is a lot more natural now
that nir_lower_blend speaks PIPE_BLEND enums.
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Italo Nicola <italonicola@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24076>
This avoids the silly compiler versions. Some bits are slightly more
complicated, because they have to account for inverted enum values (rather than
a separate invert bit), but this is a LOT friendlier to drivers using the pass
and it makes the pass itself more readable.
The conversion functions in panfrost/panvk will go away momentarily.
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Italo Nicola <italonicola@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24076>
Vulkan drivers that use nir_lower_blend need to translate Vulkan enums to the
common (non-Vulkan) versions used in nir_lower_blend. We don't need to duplicate
that boilerplate in every VK driver that uses nir_lower_blend, move panvk's
versions to common code so we can use them in agxv. I suspect powervr wants this
too.
It might be useful to also share the logic to translate vk_color_blend_state
to nir_lower_blend_options wholesale, but panvk wouldn't use it and agxv is
downstream so it wouldn't have any in-tree users. So I'll keep that part
vendored (for now). For now, let's share the easy win of the enum translation.
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Italo Nicola <italonicola@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24076>
PanVK will use this too.
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Italo Nicola <italonicola@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24076>
For sharing across the tree.
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Italo Nicola <italonicola@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24076>
We're going to do some code motions out of p_state.h so the one file assumption
will fail. relax it.
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Yonggang Luo <luoyonggang@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Italo Nicola <italonicola@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24076>
This hasn't been updated since the a5xx days, and we've learned much
more since then. I've tried to expand it from a random collection of
notes to a more complete guide to explaining how to read the firmware
and understand the various tricks it uses to make code more compact.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24125>
This is because we still may want to be able to expose vulkan 1.3
on some devices that technically do not support it, for instance
the adreno 610 has everything required except multiview, which is
not essential for most games.
Signed-off-by: Amber Amber <amber@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20991>
