This also allows us to disable a workaround in ACO for GFX9, but it
can't be removed completely because RadeonSI needs to use it.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31385>
Add firmware for Gen10 Arm Mali GPUs directly to rootfs since
it is not available from debian package. For the panthor kernel
driver to be able to execute jobs on the GPU it needs this firmware.
Update the kernel, which contains a fix to build the panthor driver
as module and use HWCI_KERNEL_MODULES to load panthor module.
Signed-off-by: Vignesh Raman <vignesh.raman@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31342>
This reverts commit 36c81b5e88.
With
nir/opt_algebraic: Allow two-step lowering of ftrunc@64 to use ffract@64
it is no longer necessary to call nir_lower_doubles explicitely in r600/sfn.
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/29281>
If ftrunc@64 is lowered by nir_lower_doubles it is turned into a
comparable long series of 32 bit operations. If the hardware
supports ffract@64 then nir_opt_algebraic can first lower ftrunc@64
to use some combinations with ffloor@64. They can then be turned
into a combination of fsub@64 and ffract@64 resulting in less
all-over instructions.
Fixes: 5218cff34b
nir/algebraic: avoid double lowering of some fp64 operations
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/29281>
The workaround is already implemented by
batch_emit_pipe_control_write(), we don't need to do it here as well.
This was spotted by Lionel Landwerlin. The credits go to him, I just
wrote the patch.
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31412>
Commit a603cc0633 ("anv: move some pc was to
batch_emit_pipe_control_write") moved some WAs from
emit_apply_pipe_flushes() to batch_emit_pipe_control_write(), but it
turns out one of them was already there since cf7e1f3817 ("anv,
iris: add missing CS_STALL bit for GPGPU texture invalidation").
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31412>
Doesn't work with half registers. For 16-bit operations, there is mad.x16.
Cc: mesa-stable
Reviewed-by: Connor Abbott <cwabbott0@gmail.com>
Signed-off-by: David Heidelberg <david@ixit.cz>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31334>
When alpha_to_coverage is enabled, the zs_update_operation field
must be set to force_late, according to (some of) the documentation.
Actually the docs are ambiguous; the main thrust is that late coverage
updates are only required when Z or S is written, or when occlusion
queries are enabled. But there is a side note in a table that indicates
force_late should be used for coverage updates even if Z or S is not
written.
Logically this shouldn't be necessary and the note is probably
just lazily written. But it turns out that we do seem to need the
force_late setting on valhall. It's currently unclear whether
there's a hardware issue on valhall, or some other issue.
Fixes piglit ext_framebuffer_multisample-*alpha-to-coverage* tests
on valhall.
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/31181>
There's good reasons to emulate compressed formats; mobile and desktop
HW generally have different ideas of what formats are worth the extra
silicon, so they support different formats based on the legacy they were
designed to support. This means we can get better application
compatibility by emulating these formats.
But, that comes at a price, namely that applications that supports
multiple formats don't really have a way to detect which ones are
natively supported and which ones are not. This means they might make
the wrong choice, and end up wasting memory and bandwidth when that
could have been avoided.
So let's make the emulation optional, by introducing a DRIconf variable
to disable it.
This intentionally still leaves in place transcoding support, because
that's still *probably* a win. Besides, those needs to be opted into
anyway.
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Acked-by: David Heidelberg <david@ixit.cz>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31420>
We were unlocking the bo handles hash table right after removing the bo handle
and afterward closing it. This leads to race conditions where the handle could
have been re-acquired on another thread. As the kernel would return the same
bo handle and do not reference count them, this leads to it being used after
being closed.
Signed-off-by: Corentin Noël <corentin.noel@collabora.com>
Reviewed-by: Dmitry Osipenko <dmitry.osipenko@collabora.com>
cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31421>
Fixes `spec@glsl-1.10@execution@vs-loop-partial-unroll-multiple-breaks`.
Reviewed-by: Jose Maria Casanova Crespo <jmcasanova@igalia.com>
Signed-off-by: Juan A. Suarez Romero <jasuarez@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31427>
Changes required after commit 514df444e
dri_gbm module is installed as /vendor/lib{64}/dri_gbm.so
to avoid changes in system/linkerconfig AOSP project
gbm-backends-path option is set as /vendor/$(MESA3D_LIB_DIR)
MESA3D_GBM_BINS variable simplifies dependencies declarations
Test results: gbm_gralloc works again
Fixes: 514df444e ("gbm: split the DRI backend into a separate library and unify backend handling")
Reviewed-by: Eric Engestrom <eric@igalia.com>
Reviewed-by: Roman Stratiienko <r.stratiienko@gmail.com>
Tested-by: Mauro Rossi <issor.oruam@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31428>
Only update cached adjustment when it is valid.
Reviewed-by: Roy Chan <roy.chan@amd.com>
Acked-by: Alan Liu <haoping.liu@amd.com>
Signed-off-by: Tomson Chang <tomson.chang@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31274>
When color adjustment changed from custom values to default values, the
useColorAdj will be false and the code to update new stream ctx color adj
will be skipped, thus when the color adj changes to the same custom values
later, it will show as "not dirty", causing custom color adj not taking effect
unless its value changed.
We should always update the new steam ctx color adj no matter it is default or not.
Reviewed-by: Roy Chan <roy.chan@amd.com>
Acked-by: Alan Liu <haoping.liu@amd.com>
Signed-off-by: Tomson Chang <tomson.chang@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31274>
1. Implement a vector like data structure to expand
command list memory dynamically.
2. The size will be doubled once we reach its capacity.
3. Make vector a generic util under utils folder.
Reviewed-by: Roy Chan <roy.chan@amd.com>
Reviewed-by: Jesse Agate <jesse.agate@amd.com>
Acked-by: Alan Liu <haoping.liu@amd.com>
Signed-off-by: Jude Shih <shenshih@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31274>
[why]
Fix in 0 input stream, it would crash in accessing null param->streams.
[how]
- Only use stream_ctx->stream instead of param->stream in color
handling.
- Revise the geometric downscaling support.
Reviewed-by: Tomson Chang <tomson.chang@amd.com>
Acked-by: Alan Liu <haoping.liu@amd.com>
Signed-off-by: Roy Chan <roy.chan@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31274>
[why]
The fix point conversion takes quite a bit of CPU time.
And if there are no changes, we don't need to convert all sw points
into hw points and generate the corresponding configs.
[how]
Introduce a config cache header so that all config caching handlings
can be done in the same way.
Luts won't be cached if it is in bypass mode, only cache when it is
non bypass and some dirty flag is set by the upper layer.
The config cache handling will re-apply the cached config if not
dirty and not in bypass mode.
Reviewed-by: Krunoslav Kovac <krunoslav.kovac@amd.com>
Acked-by: Alan Liu <haoping.liu@amd.com>
Signed-off-by: Roy Chan <roy.chan@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31274>
If the stride we're adding to our loop counter is larger than the total
amount of shared local memory we're trying to initialize, we know the
loop will run at most one time. So we can skip emitting a loop.
Loop unrolling appears to be unable to detect this currently.
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31312>
This is also encouraged by another wa, Wa_14018813551.
Both workarounds state that StackIDControlOverride_RTGlobals should
always be set to 0 (i.e. 2k).
Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30937>
NIR discard_if expects bools, therefore we end with extr CMP in
the end, however we can handle floats just fine (only with the
reversed logic, so it triggers for src < 0).
Shader-db RV530:
total instructions in shared programs: 120787 -> 120687 (-0.08%)
instructions in affected programs: 5559 -> 5459 (-1.80%)
helped: 102
HURT: 8
total temps in shared programs: 16770 -> 16764 (-0.04%)
temps in affected programs: 151 -> 145 (-3.97%)
helped: 8
HURT: 5
total cycles in shared programs: 181520 -> 181444 (-0.04%)
cycles in affected programs: 10932 -> 10856 (-0.70%)
helped: 101
HURT: 12
Signed-off-by: Pavel Ondračka <pavel.ondracka@gmail.com>
Reviewed-by: Filip Gawin <filip@gawin.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30967>
