If the only user is a trivial bcsel which in a second step
can be turned into a phi, this conversion is also worth it
even if the previous result is not undefined or constant.
Allows for some more loop unrolling or saves a few instructions.
Totals from 62 (0.04% of 139391) affected shaders (NAVI10):
SGPRs: 4976 -> 4992 (+0.32%)
VGPRs: 4408 -> 4472 (+1.45%); split: -0.45%, +1.91%
CodeSize: 453632 -> 464000 (+2.29%); split: -0.32%, +2.60%
MaxWaves: 527 -> 511 (-3.04%); split: +0.38%, -3.42%
Instrs: 84940 -> 86681 (+2.05%); split: -0.36%, +2.41%
Cycles: 11946844 -> 11783708 (-1.37%); split: -1.40%, +0.04%
VMEM: 9403 -> 10357 (+10.15%); split: +11.59%, -1.45%
SMEM: 3003 -> 3025 (+0.73%); split: +1.07%, -0.33%
VClause: 1756 -> 1997 (+13.72%); split: -0.11%, +13.84%
SClause: 2914 -> 2915 (+0.03%); split: -0.10%, +0.14%
Copies: 6426 -> 6768 (+5.32%); split: -4.14%, +9.46%
Branches: 2105 -> 2102 (-0.14%); split: -1.66%, +1.52%
PreSGPRs: 2921 -> 2909 (-0.41%); split: -0.55%, +0.14%
PreVGPRs: 4151 -> 4179 (+0.67%); split: -0.24%, +0.92%
cc: mesa-stable
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8123>
Fix defect reported by Coverity Scan.
Uninitialized pointer field (UNINIT_CTOR)
uninit_member: Non-static class member name is not initialized in
this constructor nor in any functions that it calls.
Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7766>
With the block's end_ip accidentally being the ip of the next instruction,
contrary to the comment, you would end up doing end-of-block freeing early
and have the value missing when it came time to emit the next instruction.
Just expand the ips to have separate ones for start and end of block --
while it means that nir_instr->index is no longer incremented by 1 per
instruction, it makes sense for use in liveness because a backend is
likely to need to do other things at block boundaries (like emit the if
statement's code), and having an ip to identify that stuff is useful.
Fixes: a206b58157 ("nir: Add a block start/end ip to live instr index metadata.")
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7658>
init_liveness_block(..) only gets called without using the
always true return value. Switch from bool to void.
Signed-off-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8051>
This removes the memory of some old OES extensions from the table, which
seems to be harmless.
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8037>
This involves determining the variables referenced by intrinsics, setting and
using the access qualifier correctly and considering that images and buffers
can alias.
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Connor Abbott <cwabbott0@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6483>
ACCESS_NON_WRITEABLE means that the memory is read-only, not the variable,
so we don't have to check for aliasing.
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Connor Abbott <cwabbott0@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6483>
ACCESS_NON_WRITEABLE means that the memory is read-only, not the variable,
so we have to check for aliasing first.
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Connor Abbott <cwabbott0@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6483>
It shouldn't matter whether an access/variable is coherent or not, just
that it's not written. The coherent qualifier doesn't mean anything with
read-only memory.
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Connor Abbott <cwabbott0@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6483>
among all Android gen rules '::' was used only here to declare dependencies;
mesa development and stable branch are worth receiving the fix
Fixes the following building errors with Android 7:
obj/STATIC_LIBRARIES/libmesa_nir_intermediates/spirv/gl_spirv.P:184: *** target file
gen/STATIC_LIBRARIES/libmesa_nir_intermediates/spirv/vtn_generator_ids.h' has both : and :: entries. Stop.
Cc: "20.3" <mesa-stable@lists.freedesktop.org>
Fixes: 1070bba19e ("android: fix SPIR-V -> NIR build")
Reported-by: youling257 <youling257@gmail.com>
If ACCESS_NON_UNIFORM is not specified, we can assume the resource is
uniform. This requires nir_lower_non_uniform_access to remove that flag.
A few Detroit: Become Human shaders use a index sourced from a fragment
input which is expected to be uniform.
shader-db (Navi):
Totals from 8 (0.01% of 127638) affected shaders:
SGPRs: 224 -> 384 (+71.43%)
VGPRs: 208 -> 112 (-46.15%)
CodeSize: 5360 -> 5344 (-0.30%); split: -1.49%, +1.19%
Instrs: 1036 -> 1028 (-0.77%); split: -1.93%, +1.16%
VMEM: 1320 -> 608 (-53.94%)
SMEM: 384 -> 336 (-12.50%); split: +14.58%, -27.08%
VClause: 24 -> 16 (-33.33%)
SClause: 48 -> 56 (+16.67%)
PreSGPRs: 124 -> 216 (+74.19%)
PreVGPRs: 168 -> 88 (-47.62%)
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5201>
- xxhash is faster than sha1.
- remove superfluous calls to strlen
Using SPECviewperf13 snx-03 first subtest and "perf -e cycles -g", perf report says:
Before | After | Function
---------|--------|---------------
47.39% | 47.36% | _mesa_CallList
5.00% | 3.03% | _mesa_program_resource_location
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7078>
In particular, if we have an index or bindless handle we were passing
the original handle which, technically, is uniform within the context of
the if. However, we can save the back-end compiler some effort if we
pass it the result of the read_first_invocation().
(Rebased by Kenneth Graunke and Rhys Perry.)
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7592>
In theory, I don't think this is a functional change. We should
generate the same code before and after.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7592>
This introduces a new flag in shader_info to know if a fragment
shader uses sample shading, even if there is no inputs.
During NIR linking, constants varyings are optimized and the
per-sample interpolation info (ie. the sample qualifier) might
be removed if nir_shader_gather_info() is called again.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7876>
Drivers should upload only UniformBytes of uniforms and constants,
and then use _mesa_upload_state_parameters to upload state parameters.
This allows removing one copy of state parameters.
Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6946>
Instead of having $matrix and $modifier as separate enums, combine them
to 1 enum $matrix_$modifier.
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6946>
As we now reuse the enums to remain within 64 values, we need to get
the proper name using the stage.
v2: Use enum type for parameter (Jason)
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7795>
v2: Fixup comment about bits in nir_intrinsics.py
v3: Use varying for primitive shading rate builtin (samuel)
v4: Reoder switch alphabetically
Make divergence of frag_shading_rate an option
v5: Remove stage check for frag_shading_rate in divergence (Samuel)
v6: s/frag_shading_rate_per_subgroup/single_frag_shading_rate_per_subgroup/ (Jason)
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7795>
In commit 00b28a50b2, Marek extended
a number of optimizations that had been 32-bit specific to work on
other bit-sizes.
Most optimizations preserve the data type across the transformation.
In other words, an optimization which generates e.g. fp64 operations
only does so when the source expression also contains fp64 operations.
These transformations are fine with respect to lowering, because we
will lower away all expressions that would trigger the search portion
of the expression, and so we'd never apply those rules.
However, a few of the rules create new operations that run afoul of
lowering passes. For example,
('bcsel', a, 1.0, 0.0) => ('b2f', a)
where the result is a double would simply be a selection between two
different 64-bit constants. The replacement expression, on the other
hand, involves a nir_op_b2f64 ALU operation. If we're run after
nir_lower_doubles, then it may not be legal to generate such an
expression anymore (at least without running lowering again, which we
don't do today).
Regressions due to this are blocking the 20.3 release, so for now, we
take the easy route and simply disallow those few rules when doing full
softfp64 lowering, which fixes the immediate problem. But it doesn't
solve the long-term problem in an extensible manner.
In the future, we may want to add a `lowered_alu_ops` bitfield to the
NIR shader, and as lowering passes are run, mark them as taboo. Then,
we could have each algebraic transformation track which operations it
creates in the replacement expression. With both of those in place,
nir_replace_instr could compare the transformation's list of ALU ops
against `lowered_alu_ops` and implicitly skip rules that generate
forbidden ALU operations.
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/3504
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7841>
