we need to predicate the store, since we can't do a hardware demote after
running tests. this is similar to what the blob does.
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/29607>
this is a hack that fixes faults with eMRT. I do not understand what's going on
here. hopefully we can root cause this at some point :-/
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/29607>
Farm variable is added for devices in collabora farm.
So add support in LAVA job submitter to use this in
structured log files.
Signed-off-by: Vignesh Raman <vignesh.raman@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/29583>
This makes the function easier to use when radeon_info is not available.
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/29243>
For the same reason we set it for image stores. GFX6 has a caching bug
which requires this.
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/29243>
There doesn't seem to be a restriction on the mentioned data types on
LNL anymore. Default to a maximum exec size of SIMD16.
This patch fixes dEQP-VK.subgroups.shuffle.framebuffer.* on LNL
Signed-off-by: Rohan Garg <rohan.garg@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/29504>
When nextStage is 0, the driver needs to assume that a stage might be
used with any valid next stages.
Fixes new dEQP-VK.shader_object.binding.*_no_next_stage.
Cc: mesa-stable
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/29567>
This register stores the depth format of the underlying depth
buffer, it seemingly doesn't change anything about the LRZ buffer
itself and has no behavioral changes over setting it to 0.
However, it's possible that there's some case where it does matter
so matching the proprietary driver's behavior is safer.
Signed-off-by: Mark Collins <mark@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/29453>
I found this while reverse engineering A7XX LRZ, it's going to be
relevant when we implement concurrent binning so I've added the
register definition.
Signed-off-by: Mark Collins <mark@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/29453>
A7XX allows setting the FC depth to an arbitrary F32 value rather
than being limited to 0.0/1.0, we use this to match the depth clear
value.
Signed-off-by: Mark Collins <mark@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/29453>
The layout of the LRZ FC section has changed substantially between
A6XX and A7XX so the best way to express the layout was determined
to be a templated structure.
Signed-off-by: Mark Collins <mark@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/29453>
This can be seen emitted in traces related to FDM on A740, it's set
to zero to ensure there's no side effects from prop writing to it.
Signed-off-by: Mark Collins <mark@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/29453>
A lot of CHIP template parameters were hardcoded to A6XX rather than
the actual chip which would lead to an incorrect command stream being
generated.
Signed-off-by: Mark Collins <mark@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/29453>
The functionality of GRAS_LRZ_CNTL on A6XX was split into GRAS_LRZ_CNTL
and GRAS_LRZ_CNTL2 on A7XX. The only new field is for the Z function to
be specified.
Signed-off-by: Mark Collins <mark@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/29453>
This is an exceptional case where any writes to gl_Depth should be
ignored, it means we can use LRZ in this case and don't need to
disable it.
Signed-off-by: Mark Collins <mark@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/29453>
