I haven't seen this happening anywhere, but let's have it for
correctness.
Reviewed-by: José Roberto de Souza <jose.souza@intel.com>
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30252>
In this series we're going to further change this function, adding a
lot more lines, so this patch should make the next diffs a little
easier to comprehend and review.
Reviewed-by: José Roberto de Souza <jose.souza@intel.com>
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30252>
The offset variable declaration at the beginning of the function was left over
after the variable was moved inside the if statement.
Fixes: 17c12a9924 ("turnip/kgsl: Support external memory via ION/DMABUF buffers")
Signed-off-by: Valentine Burley <valentine.burley@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30300>
I'm not sure if two otherwise equal texture instructions ever have sources
in different orders, but they should be considered equal.
ministat of nir_opt_cse:
N Min Max Median Avg Stddev
x 9 6.586801 6.718673 6.682875 6.6621411 0.047817119
+ 9 6.519098 6.609235 6.552997 6.5605604 0.028879587
Difference at 95.0% confidence
-0.101581 +/- 0.0394755
-1.52475% +/- 0.585928%
(Student's t, pooled s = 0.0395)
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Tatsuyuki Ishi <ishitatsuyuki@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30145>
This is faster.
ministat of nir_opt_cse:
N Min Max Median Avg Stddev
x 9 6.724212 6.84511 6.788336 6.7873378 0.034363882
+ 9 6.586801 6.718673 6.682875 6.6621411 0.047817119
Difference at 95.0% confidence
-0.125197 +/- 0.0416115
-1.84456% +/- 0.609248%
(Student's t, pooled s = 0.0416374)
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Tatsuyuki Ishi <ishitatsuyuki@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30145>
ministat of nir_opt_cse:
N Min Max Median Avg Stddev
x 9 7.393408 7.490593 7.434056 7.4338972 0.028150325
+ 9 6.724212 6.84511 6.788336 6.7873378 0.034363882
Difference at 95.0% confidence
-0.646559 +/- 0.0313916
-8.69745% +/- 0.407925%
(Student's t, pooled s = 0.0314111)
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Tatsuyuki Ishi <ishitatsuyuki@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30145>
Based on the other similar logs we only want to log when extensions
is NULL.
Use this opportunity to indicate the source of the log and remove
the extra ')' at the end of each line.
Fixes: 50fc7cc290 ("glx: directly link to gallium")
Reviewed-by: Eric Engestrom <eric@igalia.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30257>
Otherwise AMD_DEBUG=noexporteddcc will be ignored when modifier are
used.
Similarly to AMD_DEBUG=nodcc handling, this makes the application
unable to import buffers with DCC as well - the alternative would be
to implement the filtering only in the texture creation path, so in
the si_modifier_supports_resource function.
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30257>
If max_dw is smaller than the pm4 array the allocation size would be
smaller than sizeof(ac_pm4_state).
Fixes: 428601095c ("ac,radeonsi import PM4 state from RadeonSI")
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30257>
A specific clear_surface blitter operation is required, because in this
case we need to save framebuffer information, but not in standard clear,
as we are currently doing.
This fixes a leak in depthstencil surface, which happens because we were
storing saving it as part for the framebuffer information, but the
blitter clear wasn't restoring it because it wasn't required (only it
is required in clearing a surface).
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Signed-off-by: Juan A. Suarez Romero <jasuarez@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30240>
So far, in order to know what we need to save before using the blitter
utility, we use a boolean to know if we are going to do a blit or not,
and if we need to take in account conditional rendering or not.
In other to allow to specify more operations than blit or not, use an
enum to define what operation we would like to do, and based on that
what information we need to store.
This also merge the conditional rendering as part of these operations.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Signed-off-by: Juan A. Suarez Romero <jasuarez@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30240>
Valgrind will report some memory possibly lost because of this singleton
(it's dynamically allocated when it is first accessed).
Use atexit() to register a handler that releases this singleton.
Signed-off-by: Icenowy Zheng <uwu@icenowy.me>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30216>
The ownership of the TargetMachine object is released when LPJit
singleton is constructed, leads to a slight memory loss detectable.
Keep the ownership by saving the unique pointer as another class member
named tm_unique.
Signed-off-by: Icenowy Zheng <uwu@icenowy.me>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30216>
Pre-Volta, OpAtom works like OpSuAtom where the comparison value and the
data are packed into a vector. We need some way of expressing this in
the IR. We could have a separate OpAtom instruction but that seems
unnecessary.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30281>
On MTL ChromeOS boards, during AI based video conference, we were
observing a lot of overhead from invalidations. Upon debug, it was found
that we were using clflush in this function and that isn't efficient.
With this change, while executing compute workloads like zoo models, we
are getting ~25% performance improvements in a best case scenario.
Rework:
* Jordan: Call intel_clflushopt_range() rather than
__builtin_ia32_clflushopt() because intel_mem.c is not compiled
with -mclflushopt.
Backport-to: 24.1 24.2
Signed-off-by: Sushma Venkatesh Reddy <sushma.venkatesh.reddy@intel.com>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30238>
On some low-end GPUs (e.g. BXE/BXM series), the Imagination proprietary
Vulkan driver do not implement geometryShader feature, which is required
by Zink currently to smoothen lines. In addition, these vulkan drivers
shipped are usually not robust enough to run Zink at all, and
frequently fail with crashes.
Reject these drivers.
Signed-off-by: Icenowy Zheng <uwu@icenowy.me>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30280>
