Tracking if we changed the shader is somewhat complicated, so assume
we always do.
Reviewed-by: Emma Anholt <emma@anholt.net>
Reviewed-by: Konstantin Seurer <konstantin.seurer@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36291>
See the comment at the top file :-)
The ideas in this pass are based on LLVM. The implementation itself is from
scratch because have you /tried/ to read that thing?
Because LLVM and therefore prop drivers do these optimizations, this should help
narrow Mesa's performance gap with the blobs.
This probably needs some tuning for best results on other ISAs, but the stats
for AGX speak for themselves, see next commits.
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Acked-by: Adam Jackson <ajax@redhat.com>
Reviewed-by: Mel Henning <mhenning@darkrefraction.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36147>
this is a little annoying.
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Mel Henning <mhenning@darkrefraction.com>
Acked-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36147>
For example, divergence analysis can call nir_print_instr with an
instruction that doesn't have a block set. When that happens,
print_state::shader will be NULL.
I stumbled on this while testing !36147.
v2: Use nir_instr::has_debug_info instead. Suggested by Konstantin.
Reviewed-by: Konstantin Seurer <konstantin.seurer@gmail.com>
Fixes: ce0f30b230 ("nir: Add variable debug info to instructions")
Fixes: 3aeab4ce40 ("nir/print: Do not print debug information when gathering it")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36267>
for intrinsics, we have these really nice builders using designated initializers
+ macros to specify optional indices. texture instrs have even more craziness
involved, but we can do the same trick. this commit takes the existing "fixed
form" deref-centric tex builders and generalizes them to work with non-deref
textures, making it useful also for GL and late VK passes, while providing an
API that strives to be ergonomic and consistent.
this series only implements a subset of possible texture operations for now, but
more generalizing could be added as people have need.
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Konstantin Seurer <konstantin.seurer@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36050>
mathemtically, associativity is only defined for binary operations. I have no
idea what "associativity" would even mean for imad. I can kinda see the idea for
iadd3 but iadd3 should not be formed until after reassociating adds so the point
is moot. Unmark the
"associative" ternary operations, and assert that associativity implies binary.
nothing uses associativity yet, so this doesn't cause any functional change.
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Mel Henning <mhenning@darkrefraction.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36257>
nothing currently uses the associative flag, but they will change soon. we need
to stop incorrectly marking fmul/fadd/etc as associative, because they're not,
but they almost are. distinguish these properties so we can correctly
handle floating point rules without any opcode-based special casing.
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Mel Henning <mhenning@darkrefraction.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36257>
this chain of fmul is deliberately chosen for floating point precision
reasons, it needs to be exact, or else we might try to reassociate it
and break subnormal handling.
avoids regressing dEQP-VK.glsl.builtin.precision.ldexp_subnormals.*
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Mel Henning <mhenning@darkrefraction.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36257>
We were missing a couple bits from hash and a bunch of stuff from the
comparison. This puts most of nir_tex_instr into a single pack_tex
helper that's used by both and grabs everything we were missing.
Cc: mesa-stable
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36234>
I keep reaching for this helper but it doesn't exist. So I fixed that.
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Reviewed-by: Eric Engestrom <eric@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36142>
Mary has written the initial code, I've documented my changes below.
v2: support cmat_convert (Karol)
fix cross matmul (Karol)
rework matrix layout clasasification (Karol)
add support for saturated cmat_muladd (Karol)
Co-authored-by: Mary Guillemard <mary.guillemard@collabora.com>
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32777>
e.g. load X; load W; ==> load XYZW. Verified with a shader test.
This will be used by AMD drivers. See the code comments.
Reviewed-by: Simon Perretta <simon.perretta@imgtec.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36098>
When we detect that the source is a conversion generated by the pass,
try to get the real source instead of doing a round-trip conversion.
Make sure that the nir_alu_type and the bit_size is the same between what
we need and what's before the detected conversion.
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35744>
The only last users of nir_link_opt_varyings are Vulkan drivers.
One linker error thrown by the optimizations is reimplemented
at the call site.
No interesting shader-db changes (other than random noise).
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36091>
A TF2 shader propagates 0 to the consumer, which eliminates 1 input
if we run algebraic opts and DCE before compaction.
This is a prerequisite for removing all IO var optimizations from the GLSL
linker that are redundant with nir_opt_varyings.
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36091>
Will be used by etnaviv too.
Signed-off-by: Christian Gmeiner <cgmeiner@igalia.com>
Reviewed-by: Lucas Stach <l.stach@pengutronix.de>
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35753>
It's a sysval in mesh shader, but it share the same
slot number with VARYING_SLOT_TESS_LEVEL_INNER.
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35931>
this is more explicit than vec2's and hence has fewer footguns. in particular
it's easier to handle with preambles in a sane way.
modelled on what ir3 does.
there's probably room for more clean up but for now this unblocks what I want to
do.
stats don't seem concerning.
Totals from 692 (1.29% of 53701) affected shaders:
MaxWaves: 441920 -> 442112 (+0.04%)
Instrs: 1588748 -> 1589304 (+0.03%); split: -0.05%, +0.08%
CodeSize: 11487976 -> 11491620 (+0.03%); split: -0.04%, +0.07%
ALU: 1234867 -> 1235407 (+0.04%); split: -0.06%, +0.10%
FSCIB: 1234707 -> 1235249 (+0.04%); split: -0.06%, +0.10%
IC: 380514 -> 380518 (+0.00%)
GPRs: 117292 -> 117332 (+0.03%); split: -0.08%, +0.11%
Preamble instrs: 314064 -> 313948 (-0.04%); split: -0.05%, +0.01%
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35949>
it is sometimes useful to turn lowered bindless intrinsics into bound or vice
versa, and it is annoying to do so without this helper, so generalize the
helper.
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Job Noorman <job@noorman.info>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35949>
Class represents an indexed "ideal" register class, where non-general classes
only allow defs that choose that class in the def_size callback.
nir_opt_preamble will try to assign specialized classes where possible, falling
back to the general class once the special-purpose classes are exhausted.
AGX will use this mechanism to promote bindless texture handles to bound texture
registers where possible, falling back to pushing the handle as a uniform where
not possible. Supporting multiple classes in nir_opt_preamble allows this
multi-level hoisting to work in a single nir_opt_preamble call with proper
global behaviour.
Add this concept to nir_opt_preamble so we can use it in AGX later in this MR.
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Job Noorman <job@noorman.info>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35949>
