spirv_to_nir sometimes wraps derefs in vec2 or mov instructions as part of
its texture handling. These get in the way of
nir_rematerialize_derefs_in_use_blocks_impl. Running copy propagation
should get rid of the extra move instructions and get us back to intact
deref chains for everything except variable pointer use-cases.
fossil-db (Sienna Cichlid):
Totals from 6 (0.00% of 134572) affected shaders:
CodeSize: 92656 -> 93088 (+0.47%)
Instrs: 17060 -> 17138 (+0.46%)
Latency: 224408 -> 227539 (+1.40%)
InvThroughput: 37402 -> 37924 (+1.40%)
VClause: 408 -> 402 (-1.47%)
Copies: 1065 -> 1107 (+3.94%)
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/5668
Fixes: 14a12b771d ("spirv: Rework our handling of images and samplers")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13924>
Sometimes you might want to find a constant source without going through
all the copy prop and constant folding to make your source be a
load_const.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11613>
This commit replaces the new_src parameter of nir_ssa_def_rewrite_uses()
with an SSA def, removes nir_ssa_def_rewrite_uses_ssa(), and rewrites
all the users as needed.
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Acked-by: Alyssa Rosenzweig <alyssa@collabora.com>
Acked-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9383>
There are less copy instructions than sources, so instead of visiting each
source and rewriting it if it's uses a copy instruction, visit each copy
instruction and rewrite it's users.
Besides improving compile time, this also has a side effect of fixing a
rare situation where copy-propagation does not happen:
loop {
a = phi ..., b
c = vec ...
b = mov c.y
}
It might have been the case that a phi source could not be rewritten until
the copy was visited later.
Compile-time (nir_copy_prop):
Difference at 95.0% confidence
-2613.13 +/- 15.2094
-27.4333% +/- 0.150247%
(Student's t, pooled s = 17.963)
Comple-time (overall):
Difference at 95.0% confidence
-2627.89 +/- 201.557
-2.05404% +/- 0.156221%
(Student's t, pooled s = 238.048)
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8784>
These were hurting performance of other passes.
Compile-time (overall):
Difference at 95.0% confidence
-5496.3 +/- 219.752
-4.11912% +/- 0.160285%
(Student's t, pooled s = 259.538)
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Connor Abbott <cwabbott0@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8784>
This moves the fi_types to a new mesa_private.h and removes the
imports.c file. The vast majority of this patch is just removing
pound includes of imports.h and fixing up the recursive includes.
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/3024>
The difference between imov and fmov has been a constant source of
confusion in NIR for years. No one really knows why we have two or when
to use one vs. the other. The real reason is that they do different
things in the presence of source and destination modifiers. However,
without modifiers (which many back-ends don't have), they are identical.
Now that we've reworked nir_lower_to_source_mods to leave one abs/neg
instruction in place rather than replacing them with imov or fmov
instructions, we don't need two different instructions at all anymore.
Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com>
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Vasily Khoruzhick <anarsoul@gmail.com>
Acked-by: Rob Clark <robdclark@chromium.org>
If we increase the vector size in the future it would be good
to not have to fix these up, this should change nothing at present.
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
Even in a very basic shader this reduces the time spent in
nir_copy_prop() by ~17%.
No shader-db changes for radeonsi NIR or i965.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
NIR metadata validation verifies that the debug bit was unset (by a call
to nir_metadata_preserve) if a NIR optimization pass made progress on
the shader. With the expectation that the NIR shader consists of only a
single main function, it has been safe to call nir_metadata_preserve()
iff progress was made.
However, most optimization passes calculate progress per-function and
then return the union of those calculations. In the case that an
optimization pass makes progress only on a subset of the functions in
the shader metadata validation will detect the debug bit is still set on
any unchanged functions resulting in a failed assertion.
This patch offers a quick solution (short of a larger scale refactoring
which I do not wish to undertake as part of this series) that simply
unsets the debug bit on unchanged functions.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
These correspond directly to SPIR-V's OpPtrAccessChain. As such, they
treat whatever their parent gives them as if it's the first element in
some array and dereferences that array. If the parent is, itself, an
array deref, then the two indices can just be added together to get the
final array deref. However, it can also be used in cases where what you
have is a dereference to some random vec2 value somewhere. In this
case, we require a cast before the ptr_as_array and use the ptr_stride
field in the cast to provide a stride for the ptr_as_array derefs.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
We're going to want to do more deref optimizations going forward and
this gives us a central place to do them. Also, cast propagation will
get a bit more complicated with the addition of ptr_as_array derefs.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
Acked-by: Rob Clark <robdclark@gmail.com>
Acked-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Acked-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
This commit adds a new instruction type to NIR for handling derefs.
Nothing uses it yet but this adds the data structure as well as all of
the code to validate, print, clone, and [de]serialize them.
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
Acked-by: Rob Clark <robdclark@gmail.com>
Acked-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Acked-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Add helpers to get the number of src/dest components for an intrinsic,
and update spots that were open-coding this logic to use the helpers
instead.
Signed-off-by: Rob Clark <robdclark@gmail.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
In the near future we are going to require that the num_components in a
src dereference match the num_components of the SSA value being
dereferenced. To do that, we need copy_prop to not remove our MOVs from
a larger SSA value into an instruction that uses fewer channels.
Because we suddenly have to know how many components each source has,
this makes the pass a bit more complicated. Fortunately, copy
propagation is the only pass that cares about the number of components
are read by any given source so it's fairly contained.
Shader-db results on Sky Lake:
total instructions in shared programs: 13318947 -> 13320265 (0.01%)
instructions in affected programs: 260633 -> 261951 (0.51%)
helped: 324
HURT: 1027
Looking through the hurt programs, about a dozen are hurt by 3
instructions and the rest are all hurt by 2 instructions. From a
spot-check of the shaders, the story is always the same: They get a
vec4 from somewhere (frequently an input) and use the first two or three
components as a texture coordinate. Because of the vector component
mismatch, we have a mov or, more likely, a vecN sitting between the
texture instruction and the input. This means that the back-end inserts
a bunch of MOVs and split_virtual_grfs() goes to town. Because the
texture coordinate is also used by some other calculation, register
coalesce can't combine them back together and we end up with an extra 2
MOV instructions in our shader.
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Connor Abbott <cwabbott0@gmail.com>
The intent was to continue down the indirect chain, not to call ourselves
with unchanged input arguments. Found by code inspection, and comparison
to copy_prop_alu_src().
We haven't hit this because callers of NIR's copy prop are doing so in
SSA, before indirect variable dereferences have been lowered to registers.
Reviewed-by: Rob Clark <robdclark@gmail.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
This matches the "foreach x in container" pattern found in many other
programming languages. Generated by the following regular expression:
s/nir_foreach_function(\([^,]*\),\s*\([^,]*\))/nir_foreach_function(\2, \1)/
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
This matches the "foreach x in container" pattern found in many other
programming languages. Generated by the following regular expression:
s/nir_foreach_instr(\([^,]*\),\s*\([^,]*\))/nir_foreach_instr(\2, \1)/
and similar expressions for nir_foreach_instr_safe etc.
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>