The motivating factor is: this lowering may cause
nir_intrinsic_load_local_group_size intrinsics to be added to the
shader, and by moving this around we make possible for the drivers to
lower that intrinsic by themselves.
Iris will do just that in a later patch for implementing variable
group size.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4794>
Intrinsic to get the SIMD width, which not always the same as subgroup
size. Starting with a small scope (Intel), but we can rename it later
to generalize if this turns out useful for other drivers.
Change brw_nir_lower_cs_intrinsics() to use this intrinsic instead of
a width will be passed as argument. The pass also used to optimized
load_subgroup_id for the case that the workgroup fitted into a single
thread (it will be constant zero). This optimization moved together
with lowering of the SIMD.
This is a preparation for letting the drivers call it before the
brw_compile_cs() step.
No shader-db changes in BDW, SKL, ICL and TGL.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4794>
Adding this since Iris will handle variable group size parameters by
itself.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4794>
(Co-authored with Chris Wilson.)
Frequently, games create fences and later check them with a timeout of
0 to see if that work has completed yet. They do not want the work to
be flushed immediately upon fence creation.
This is what PIPE_FLUSH_DEFERRED does - it inhibits the flush at fence
creation time, but still guarantees that a flush will occur later on
once fence_finish() is called.
Since syncpts can only occur at batch boundaries, when deferring a
flush, we have to wait for the syncpt at the end of the batch being
constructed. This is later than desired, but safe if blocking. To
avoid extra delays, we additionally insert a PIPE_CONTROL to write an
availability bit at the exact point of the fence. We can poll this
on the CPU, allowing us to check whether the fence has gone by, even
if the batch hasn't completed. It can also let us skip kernel calls.
Improves performance in Bioshock Infinite by 10% on Icelake GT2 on
-ForceCompatLevel=5 settings. Thanks to Felix Degrood and Mark Janes
for helping notice the extraneous stalls and batches, Marek Olšák for
adding deferred flush support to Gallium to solve this issue, and
Chris Wilson for reworking a lot of the internals of this work.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/3802>
Receiving a fence_server_sync (iris_fence_await) means that any future
work needs to wait for the fence. But previous work doesn't need to.
So flush it now, to avoid delaying it arbitrarily.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/3802>
In the next patch, we will introduce deferred fences where we will need
to flush a fence later. To do this, we need to know which batch requires
flushing, so keep a 1:1 mapping between seqno[] and the associated
batch.
It's also substantially less confusing to have a 1:1 mapping.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/3802>
By using the breadcrumbs we inject into the batch, we can build a
lightweight fence - that can be evaluated in userspace without having to
check in the kernel. In order to pass the fences between processes, and
to wait efficiently, we continue to track the syncobj for each batch and
use that as a terminator for the fence, and for passing coarse
scheduling decisions to the kernel on execbuf.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/3802>
We can use seqno as a basic for fast userspace fences: where we can
check a value directly to test for fence completion without having to
query using the kernel. To do so we need to write a breadcrumb from the
batch and track those writes as the basis for our lightweight fences.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/3802>
Batches are going to have an uploader in the next commit, so destroying
batches will destroy uploaders, which will unmap transfers, which will
return things to the slab allocator. So we need to reorder destroying
the slab allocator to the end to avoid crashing.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/3802>
We're going to need it to create a uploader in the batch soon. We still
avoid storing it, to maintain the charade of separation, and make people
think twice about fetching random fields from there and intertwining
things even worse.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/3802>
This is just a refcounted wrapper around a drm_syncobj. There is
enough terminology going on in the area of synchronization (sync
objects, sync files, ...) that I'd rather not invent our own.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/3802>
Lets us drop some cut and pasted kernel header contents.
Linux 4.7 came out 4 years before we the first officially supported
release of this driver; iris won't run on kernels older than 4.16,
and 4.18.11+ is strongly recommended. So I suspect it's safe to
assume that a kernel header from 4.7 will exist at build time.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/3802>
Avoids surprises where you set an OVERRIDE but it gets ignored and the
system PCI ID is used.
Also fixes the bug that the error of invalid platform name being
printed too early, even when the passed platform was a PCI ID (which
is also supported).
For the case where euid != uid, a warning was added but the behavior
wasn't changed: it is still going to fallback to system PCI ID.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4841>
gen12 does away with the single patch dispatch mode for tcs, and
increases some limits so that 8_patch mode can always work. Make the
necessary changes so we don't try to fall back to single patch mode.
Fixes KHR-GL46.tessellation_shader.single.max_patch_vertices and others
Fixes: 44754279ac ("intel/fs/gen12: Use TCS 8_PATCH mode.")
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Acked-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4843>
If you're going out of your way to do per-sample interpolation, you are
almost surely going to be doing so to an MSAA framebuffer. Should reduce
recompiles with MSAA enabled.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4562>
Now that we normalize our keys fairly well, build a variant at shader
state creation time so that hopefully you don't have to call the compiler
at draw time (as is now the case with glmark2 ES and most of the humus GL
demos).
Fixes: #2782
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4562>
We can remove a bunch of conditional code at key comparison time by
computing a bitmask of used key bits at ir3_shader creation time. This
also gives us a nice place to put additional key simplification to reduce
how many variants we create (like skipping rastflat if we don't read
colors in the FS, or skipping vclamp_color if we don't write colors).
It does mean walking the whole key to AND it, but the key is just 28 bytes
so far so that seems pretty fine.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4562>
We weren't filling in the tess mode of the key, or setting has_gs on GS
shaders, resulting in assertion failures when NIR intrinsics didn't get
lowered.
We have to make a guess at prim mode for TCS, but it should be better to
have some shader-db coverage than none, and it will avoid these failures
happening when we start precompiling shaders.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4562>
Some of the negative API tests make shaders for tess stages that don't do
all the stores they need to. Once we start precompiling (or doing
shader-db of tess), we need to at least not segfault when generating them.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4562>
We were failing to tell the allocator about the restriction that scalar
texture instructions (allocated as scalar regs) couldn't be allocated such
that the start of the full unwritemasked vector started before r0. There
was a patch in select_reg_callback on a6xx that tried to work around that,
but you could still end up backed into a corner you shouldn't be because
we didn't tell the RA what it needed.
Fixes compiler assertion failures on a300-a400's blit_z shader, used for
Z32F gmem blits.
Looks like as a result we get tighter register allocation but more nops:
instructions in affected programs: 757945 -> 760356 (0.32%)
nops in affected programs: 317983 -> 320468 (0.78%)
non-nops in affected programs: 27525 -> 27451 (-0.27%)
mov in affected programs: 3098 -> 3023 (-2.42%)
dwords in affected programs: 109664 -> 110656 (0.90%)
last-baryf in affected programs: 112701 -> 112847 (0.13%)
full in affected programs: 4326 -> 4011 (-7.28%)
sstall in affected programs: 120550 -> 120836 (0.24%)
(ss) in affected programs: 13939 -> 13918 (-0.15%)
(sy) in affected programs: 3006 -> 2786 (-7.32%)
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4562>
When the GS lowering was working on store_output intrinsics, we had to
clean up the split vars to avoid getting confused. Now that we shadow
the output vars instead, there's no confusion and we can drop this
hack.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4562>
We mostly got away with replacing a store_output with a store_var, but
for complex types like structs, that doesn't work. Once the IO has
been lowered from vars to intrinsic, we've lost the deref chains and
can't properly shadow the outputs.
This commits moves the GS lowering up so we do it before the output
variables get lowered to store_output. This way the pass works much
like nir_lower_io_to_temporaries() and cleanly shadows the outputs.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4562>
This pass lowers per-vertex input intrinsics to load_shared_ir3. This
was open coded in the TCS and GS lowering passes before - this way we
can share it. Furthermore, we'll need to run the rest of the GS
lowering earlier (before lowering IO) so we need to split off this
part that operates on the IO intrinsics first.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4562>
Render Target Array Index has moved from R0.0[26:16] to
R1.1[26:16] on gen12.
Fixes dEQP-VK.multiview.input_attachments.*
Cc: <mesa-stable@lists.freedesktop.org>
Reviewed-by: Francisco Jerez <currojerez@riseup.net>
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4836>