I have been running into crashes in this function when using blender.
Some of the entries in ice->state.framebuffer.base.cbufs[0] can
apparently have the texture field be null, which was causing a segfault
in this loop.
In my case, nr_cbufs was 3, and the first two cbufs entries had a null
texture and format set to PIPE_FORMAT_NONE. The last entry had format of
PIPE_FORMAT_R16G16_FLOAT and a non-null texture.
Adding this null check before attempting to dereference the texture
fixes the crash for me and allows blender to work normally.
Fixes: ca96f8517c ("iris: remove uses of pipe_surface as a pointer")
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/40688>
Abusing RADV_PERFTEST for experimental features doesn't make real
sense, and I think we should stop doing that.
The existing RADV_PERFTEST options like RADV_PERFTEST=transfer_queue
still exists but they are marked as deprecated, they will be removed
in future Mesa releases.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/40646>
As with vk_shader_object, vk_nir_lower_descriptor_heaps() is called
right after the driver preprocess step. The resulting mapping and
embedded samplers are then baked into the pre-compile shader.
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/40649>
The original version of the pass relied on derefs and passing bindings
and indices to drivers through the callback. While this works, it's
much more convenient with heaps to just take an index into the embedded
descriptor table.
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/40649>
Lowers all mappings and embedded samplers to descriptor heaps without
mappings. This was based on a pass written by Konstantin Seurer and
Mike Blumenkrantz but was basically entirely rewritten and uses
different NIR intrinsics.
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/40649>
So we can get through all the casting inserted by heaps.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/40649>
We also add a new boolean which indicates that the texture op uses an
embedded sampler.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/40649>
Since 31af989270 ("nir/lower_continue_constructs: Simplify loops before
lowering continue constructs"), we never ingest loops with continues. That lets
us delete a bunch of now dead code (and outdated comments) around control flow.
This patch is part of the treewide effort to improve loops in NIR. I already
sent the Intel patch earlier this week and this weekend hit delete here too.
https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/40609
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenz.ca>
Reviewed-by: Mary Guillemard <mary@mary.zone>
Tested-by: Mary Guillemard <mary@mary.zone>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/40690>
struct nu_handle is hashed and deduplicated using struct nu_handle_key, which ignored
parent_deref. That means all instructions will use the first parent_deref when rewriting
the sources.
Avoid this by not including the parent deref in the struct, and instead querying it
when needed.
Fixes: 4d09cd7fa5 ("nir/lower_non_uniform_access: Group accesses using the same resource")
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/work_items/15173
Reviewed-by: Konstantin Seurer <konstantin.seurer@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/40654>
Initially I'd added the offset to make things match up to blob driver on
x2-85/a840. But this gets in the way on parts with smaller GMEM.
Signed-off-by: Rob Clark <rob.clark@oss.qualcomm.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/40655>
It didn't save states properly. The only correct place to save them is
si_blitter_begin. Unfortunately, we can't skip saving and restoring
those states because we don't know in advance whether the rectangle path
will be used.
Cc: mesa-stable
Reviewed-by: Pierre-Eric
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/40634>
This should be faster because 2 triangles are inefficient on the diagonal,
generating helper invocations and potentially extra memory loads from dst
because tiles aren't fully covered.
Reviewed-by: Pierre-Eric
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/40633>
