Modify the ALU width callback to return 0 for dot product operations,
preventing nir_lower_alu_width from decomposing them into multiply-add
chains. This preserves fdot2, fdot3, and fdot4 as single instructions
rather than expanding them into multiple fmul+fadd operations.
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/13531
Signed-off-by: Christian Gmeiner <cgmeiner@igalia.com>
Reviewed-by: Lucas Stach <l.stach@pengutronix.de>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36130>
Add a custom ALU width callback that returns 4 to enable vectorization
of scalar operations up to 4 components. Without this callback, the
nir_lower_alu_width pass generates excessive scalar code instead of
utilizing etnaviv's vector capabilities.
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/13531
Signed-off-by: Christian Gmeiner <cgmeiner@igalia.com>
Reviewed-by: Lucas Stach <l.stach@pengutronix.de>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36130>
Add additional 16 and 32b float formats, and the missing BGR161616.
For the dri2_format_table, just use the pipe formats twice, rather than
introducing new __DRI_IMAGE_FORMAT_x in this day and age (they are the
same thing).
Signed-off-by: Rob Clark <rob.clark@oss.qualcomm.com>
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36081>
If there is a separate stencil in use, the resource invalidation
flag was not being removed for the depth buffer as rsc was assigned
to the separate stencil.
Fixes: 6ff509593c ("v3d: Only apply TLB load invalidation on first job after FB state update")
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36030>
This decouples Kopper and swrast. With this change, dri/kopper.c no
longer references DRIswrastLoaderEsxtension at all.
Reviewed-by: Adam Jackson <ajax@redhat.com>
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36014>
The border color of the rv770 gpu behaves the same way as
the evergreen border color. This change updates the software
accordingly.
This change is enabled for all the pre-evergreen gpus.
This change fixes 120 piglit tests. The rv770 ci is updated
as well.
Cc: mesa-stable
Signed-off-by: Patrick Lerda <patrick9876@free.fr>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34502>
This is the backport of 0c0b978938 "radeonsi: set NEVER as
the depth compare func if depth compare is disabled".
The function r600_tex_compare arguments are updated with the "const"
keyword.
This change fixes the test below which was broken after 0c6e56c391:
khr-gl4[5-6]/incomplete_texture_access/sampler: fail pass
Fixes: 0c6e56c391 ("mesa: (more) correctly handle incomplete depth textures")
Signed-off-by: Patrick Lerda <patrick9876@free.fr>
Acked-by: Vitaliy Triang3l Kuzmin <triang3l@yandex.ru>
Reviewed-by: Gert Wollny <gert.wollny@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35968>
Add a new `zink-anv-adl-traces-performance` job, which runs the same set
of traces as the `iris-anv-adl-traces-performance` job.
Signed-off-by: Valentine Burley <valentine.burley@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36074>
Add support for disabling the VRT (Variable Register Thread) feature.
The strategy here is to force the old BRW_MAX_GRF limit for the
register allocator (locks the upper limit) and make sure
ptl_register_blocks() always return that amount of blocks (locks
the lower limit).
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35781>
This is already done for a6xx. For everyone else move it out of the
common code path and into fd_layout_resource_for_handle().
Signed-off-by: Rob Clark <rob.clark@oss.qualcomm.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36075>
Explicitly layout the resource in all cases, rather than depending on
the partial/incomplete layout that is done in fd_resource_from_handle()
(which will be going away).
Now that we are not YOLOing the layout, it turns up some questionable
assumptions in piglit. Just mark those xfails for now.
Signed-off-by: Rob Clark <rob.clark@oss.qualcomm.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36075>
And rename it to layout_resource_for_handle. The handle contains
additional information such as the pitch/offset for the imported
buffer.
Signed-off-by: Rob Clark <rob.clark@oss.qualcomm.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36075>
Export the existing helper, and call it from the legacy backends which
are not using fdl.
This moves all the layout initialization to the backends (and eventually
to fdl for the backends using that).
Signed-off-by: Rob Clark <rob.clark@oss.qualcomm.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36075>
Pass the layout type to setup_slices, renamed to layout_resource, to
move some of the partial layout initialization to the gen specific
backend.
Signed-off-by: Rob Clark <rob.clark@oss.qualcomm.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36075>
We already use fdl_layout_buffer() when the resource is allocated. But
it was missed to convert fd_resource_resize(). Convert it so there is
no path into ->setup_slices() with something that is a buffer.
Signed-off-by: Rob Clark <rob.clark@oss.qualcomm.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36075>
Previously, we would only offset register ids for LValues that are
directly used in a merge/split instruction, but this is incorrect.
We instead need to apply the offset to all LValues that compMask
has been propagated to. By calcuating this from compMask instead
of figuring it out a second time, we fix that issue and also manage
to simplify the code a bit.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24541>
