Replace calls to create hash tables and sets that use
_mesa_hash_pointer/_mesa_key_pointer_equal with the helpers
_mesa_pointer_hash_table_create() and _mesa_pointer_set_create().
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Acked-by: Eric Engestrom <eric@engestrom.ch>
Even though no one's been brave enough to ever use this pass, I like to
keep it functionally working.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: 393b59e077
('nir: Rework nir_lower_constant_initializers() to handle functions')
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
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>
Will be used to communicate that a shader uses 64-bit operations to the
concerned lowering passes.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
We're going to have multiple functions, so nir_shader_get_entrypoint()
needs to do something a little smarter.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Previously it assumed that only a single function (the entrypoint)
existed and attempted to lower constant initializers of shader outputs
for each function, for instance.
I think this was copy-and-paste mistake -- nir_opt_large_constants was
passing in glsl_get_natural_size_align_bytes() given brw_nir.c's arguments
to the opt pass.
I wanted to reuse this function for handling constant offsets of arrays of
images in V3D.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Rob Clark <robdclark@gmail.com>
V3D returns the texels in a different order in the resulting vec4 from
what GLSL wants, so we need to put in a swizzle. Fixes
dEQP-GLES31.functional.texture.gather.basic.2d.rgba8.base_level.level_1
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Initialize the variable with NULL. Fixes the following
In file included from ../src/compiler/nir/nir_lower_io.c:34:
../src/compiler/nir/nir_lower_io.c: In function ‘nir_lower_explicit_io’:
../src/compiler/nir/nir.h:668:11: warning: ‘addr’ may be used uninitialized in this function [-Wmaybe-uninitialized]
return src;
^~~
../src/compiler/nir/nir_lower_io.c:735:17: note: ‘addr’ was declared here
nir_ssa_def *addr;
^~~~
v2: Avoid using a 'default' case so we get help from the compiler when
new deref types are added. (Lionel)
Acked-by: Jason Ekstrand <jason@jlekstrand.net>
the naming is a bit confusing no matter how you look at it. Within SPIR-V
"global" memory is memory accessible from all threads. glsl "global" memory
normally refers to shader thread private memory declared at global scope. As
we already use "shared" for memory shared across all thrads of a work group
the solution where everybody could be happy with is to rename "global" to
"private" and use "global" later for memory usually stored within system
accessible memory (be it VRAM or system RAM if keeping SVM in mind).
glsl "local" memory is memory only accessible within a function, while SPIR-V
"local" memory is memory accessible within the same workgroup.
v2: rename local to function as well
v3: rename vtn_variable_mode_local as well
Signed-off-by: Karol Herbst <kherbst@redhat.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
For now, it's hidden behind a cap. Hopefully, we can eventually drop
that along with all the manual offset code in spirv_to_nir.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
Tested-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
This new pass is for lowering explicitly laid out memory coming in from
SPIR-V or a similar source. It's quite a bit more complicated than the
normal lower_io because we have to be able to handle matrices. The
way the stride information is stored for matrices is awkward and dealing
with row-major matrices is especially painful.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
This commit adds a new num_components value for intrinsic sources of -1
which means that it consumes everything and the number of components
effectively isn't validated. This is useful for deref sources which
just take the result of the deref and we leave it up to the driver to
decide what that size should be.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
We added this assert when first moving derefs over to instructions to
ensure that deref chains could go all the way back to the variables.
Now that we're going to start using derefs for things that we can do
variable pointers on such as UBOs and SSBOs, we need to be able to run
derefs through phi nodes, selects, and basically anything else.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
We already detect any incomplete deref chains (where the deref is used
for something other than another deref or a load/store) and flag the
variable as used thanks to deref_used_for_not_store. All that's left to
do is to properly skip casts when cleaning up.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
This pass is used when, for instance, we lazily change the mode of
variables rather than replacing the variable with a new one. Since we
only do this in cases where we know we have full deref chains, it's ok
to just skip them in fixup_deref_modes.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
The code which constructs deref paths already gives you the path
starting at the nearest deref_cast or deref_var. All we need to do for
casts is handle the case where the start of the path isn't a deref_var.
For ptr_as_array derefs, we just bail if we have any after the
divergence point between the two derefs. We may be able to do better in
the future but this works for now.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
When handling casts, we can't blindly propagate the parent of a cast
into a ptr_as_array deref because doing so might loose the stride
information from the cast. Instead, before we can propagate into
ptr_as_array derefs, we need to check that the cast is a cast of an
array deref and that the stride matches. For other types of derefs, we
can continue to propagate casts as normal because they don't need the
stride. We also add an optimization which can combine a ptr_as_array
deref with it parent if it is also an array deref of some form.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
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>
SPIR-V allows for matrix and array types to be decorated with explicit
byte stride decorations and matrix types to be decorated row- or
column-major. This commit adds support to glsl_type to encode this
information. Because this doesn't work nicely with std430 and std140
alignments, we add asserts to ensure that we don't use any of the std430
or std140 layout functions with explicitly laid out types.
In SPIR-V, the layout information for matrices is applied to the parent
struct member instead of to the matrix type itself. However, this is
gets rather clumsy when you're walking derefs trying to compute offsets
because, the moment you hit a matrix, you have to crawl back the deref
chain and find the struct. Instead, we take the same path here as we've
taken in spirv_to_nir and put the decorations on the matrix type itself.
This also subtly adds support for strided vector types. These don't
come up in SPIR-V directly but you can get one as the result of taking a
column from a row-major matrix or a row from a column-major matrix.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
It was added in bce6f99875 even though it's completely redundant with
glsl_array_type().
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
Previously, NIR had a single nir_var_uniform mode used for atomic
counters, UBOs, samplers, images, and normal uniforms. This commit
splits this into nir_var_uniform and nir_var_ubo where nir_var_uniform
is still a bit of a catch-all but the nir_var_ubo is specific to UBOs.
While we're at it, we also rename shader_storage to ssbo to follow the
convention.
We need this so that we can distinguish between normal uniforms and UBO
access at the deref level without going all the way back variable and
seeing if it has an interface type.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
I have no idea how shader_storage made it into the list of banned
variable modes for stores but it clearly should be allowed. This only
doesn't cause us a problem today because we never actually use derefs on
shader_storage variables.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
This doesn't currently change anything because array indices are
required to be 32 bits and all derefs are also 32 bits. However, we
will one day have 64-bit derefs for OpenCL.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
I've been doing this in the nir-to-vir and nir-to-qir backends of v3d and
vc4, but nir could potentially do some useful stuff for us (like avoiding
unpack/repacks) if we give it the information.
v2: Skip lowering for txs/query_levels
v3: Fix a crash on old-style shadow
v4: Rename to tex_packing, use nir_format_unpack_sint/uint helpers, pack
the enum.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
When copy_prop_vars also took care of dead write handling, intrin was
used as part of store_to_entry. Now it isn't, so this assignment
isn't used really used. Add a comment clarifying what happens to
intrin.
Fixes: 4dfa7adc10 "nir: Remove handling of dead writes from copy_prop_vars"
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
After trying multiple times to merge if-statements with phis
between them I've come to the conclusion that it cannot be done
without regressions. The problem is for some shaders we end up
with a whole bunch of phis for the merged ifs resulting in
increased register pressure.
So this patch just merges ifs that have no phis between them.
This seems to be consistent with what LLVM does so for radeonsi
we only see a change (although its a large change) in a single
shader.
Shader-db results i965 (SKL):
total instructions in shared programs: 13098176 -> 13098152 (<.01%)
instructions in affected programs: 1326 -> 1302 (-1.81%)
helped: 4
HURT: 0
total cycles in shared programs: 332032989 -> 332037583 (<.01%)
cycles in affected programs: 60665 -> 65259 (7.57%)
helped: 0
HURT: 4
The cycles estimates reported by shader-db for i965 seem inaccurate
as the only difference in the final code is the removal of the
redundent condition evaluations and jumps.
Also the biggest code reduction (~7%) for radeonsi was in a tomb
raider tressfx shader but for some reason this does not get merged
for i965.
Shader-db results radeonsi (VEGA):
Totals from affected shaders:
SGPRS: 232 -> 232 (0.00 %)
VGPRS: 164 -> 164 (0.00 %)
Spilled SGPRs: 59 -> 59 (0.00 %)
Spilled VGPRs: 0 -> 0 (0.00 %)
Private memory VGPRs: 0 -> 0 (0.00 %)
Scratch size: 0 -> 0 (0.00 %) dwords per thread
Code Size: 14584 -> 13520 (-7.30 %) bytes
LDS: 0 -> 0 (0.00 %) blocks
Max Waves: 13 -> 13 (0.00 %)
Wait states: 0 -> 0 (0.00 %)
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Makes debugging easier when we care about the deref chain and not the
deref instruction itself. To make it take a const pointer, constify
some of the static functions in nir_print.c.
Reviewed-by: Eric Anholt <eric@anholt.net>
This just cleans things up a little and make things more safe for
derefs.
Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
Reviewed-by: Eric Anholt <eric@anholt.net>
This will be reused by the following patch.
Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
The following patches will add support for an additional
optimisation so this function will no longer just optimise varying
constants.
Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
The former expects to see SSA-only things, but the latter injects registers.
The assertions in the lowering where not seeing this because they asserted
on the bit_size values only, not on the is_ssa field, so add that assertion
too.
Fixes: 11dc130779 "nir: Add a bool to int32 lowering pass"
CC: mesa-stable@lists.freedesktop.org
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
When copy propagation handles a store/copy, it iterates the current
copy entries to remove aliases, but keeps the "equal" entry (if
exists) to be updated.
The removal step may swap the entries around (to ensure there are no
holes), invalidating previous iteration pointers. The bug was saving
such pointer to use later. Change the code to first perform the
removals and then find the remaining right entry.
This was causing updates to be lost since they were being made to an
entry that was not part of the current copies.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=108624
Fixes: b3c6146925 "nir: Copy propagation between blocks"
Cc: mesa-stable@lists.freedesktop.org
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
When updating a copy entry source value from a "non-SSA" (the data
come from a copy instruction) to a "SSA" (the data or parts of it come
from SSA values), it was possible to hold invalid data in ssa[0]
depending on the writemask. Because the union, ssa[0] could contain a
pointer to a nir_deref_instr left-over from previous non-SSA usage.
Change code to clean up the array before use to avoid invalid data
around.
Fixes: 62332d139c "nir: Add a local variable-based copy propagation pass"
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
The quotation marks around 1.0 cause it to be treated as a string
instead of a floating point value. The generator then treats it as an
arbitrary variable replacement, so any iand involving a ('ineg', ('b2i',
a)) matches.
v2: Remove misleading comment about sized literals (suggested by
Timothy). Add assertion that the name of a varible is entierly
alphabetic (suggested by Jason).
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Tested-by: Timothy Arceri <tarceri@itsqueeze.com> [v1]
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com> [v1]
Fixes: 6bcd2af086 ("nir/algebraic: Add some optimizations for D3D-style Booleans")
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=109075
Tested on gen9.
v2: Rename lower_txd_3d_surafaces flag to lower_txd_3d (Jason Ekstrand)
Signed-off-by: Sagar Ghuge <sagar.ghuge@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Instead of going all the way back to the variable, just look at the
deref. The modes are guaranteed to be the same by nir_validate whenever
the variable can be found. This fixes clear_unused_for_modes for
derefs that don't have an accessible variable.
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Instead of going all the way back to the variable, just look at the
deref. The modes are guaranteed to be the same by nir_validate whenever
the variable can be found. This fixes apply_barrier_for_modes for
derefs that don't have an accessible variable.
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
This is instead of looking all the way back to the variable which may
not exist for all derefs. This makes this code properly ignore casts
with modes other than the mode[s] we care about (where casts aren't
allowed).
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
This is instead of looking all the way back to the variable which may
not exist for all derefs. This makes this code properly ignore casts
with modes other than the mode[s] we care about (where casts aren't
allowed).
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
This is instead of looking all the way back to the variable which may
not exist for all derefs. This makes this code properly ignore casts
with modes other than the mode[s] we care about (where casts aren't
allowed).
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
This is instead of looking all the way back to the variable which may
not exist for all derefs. This makes this code properly ignore casts
with modes other than the mode[s] we care about (where casts aren't
allowed).
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
If we can't find the variable from the deref, just assume it isn't
invariant and continue on. This can happen if, for instance, we're
writing to a deref that points into an SSBO.
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
"There's no point in walking the program if we're never going to
actually lower anything."
Except we might lower compacted local arrays. In that case, modes will
be 0, but there is still lowering to be done.
This reverts commit 7f75cf2a94.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=109081
Suggested-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Tested-by: Clayton Craft <clayton.a.craft@intel.com>
Cc: Kenneth Graunke <kenneth@whitecape.org>
On some GPUs, especially older Intel GPUs, some math instructions are
very expensive. On those architectures, don't reduce flow control to a
csel if one of the branches contains one of these expensive math
instructions.
This prevents a bunch of cycle count regressions on pre-Gen6 platforms
with a later patch (intel/compiler: More peephole select for pre-Gen6).
v2: Remove stray #if block. Noticed by Thomas.
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Thomas Helland <thomashelland90@gmail.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
That flow control may be trying to avoid invalid loads. On at least
some platforms, those loads can also be expensive.
No shader-db changes on any Intel platform (even with the later patch
"intel/compiler: More peephole select").
v2: Add a 'indirect_load_ok' flag to nir_opt_peephole_select. Suggested
by Rob. See also the big comment in src/intel/compiler/brw_nir.c.
v3: Use nir_deref_instr_has_indirect instead of deref_has_indirect (from
nir_lower_io_arrays_to_elements.c).
v4: Fix inverted condition in brw_nir.c. Noticed by Lionel.
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
I botched some copy-and-paste and clamped to signed int max instead of
uint max. Fixes KHR-GL46.shader_image_load_store.multiple-uniforms on
skl.
Fixes: d3e046e76c ("nir: Pull some of intel's image load/store format
conversion to nir_format.h")
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
nir_sweep already marks all metadata invalid, so it is safe to release
the memory here too.
mean soft fp64 using uint64: 1,342,759,331 => 1,010,670,475
gfxbench5 aztec ruins high 11: 63,555,571 => 61,889,811
deus ex mankind divided 148: 62,845,304 => 62,829,640
deus ex mankind divided 2890: 71,922,686 => 71,922,686
dirt showdown 676: 69,238,607 => 69,238,607
dolphin ubershaders 210: 77,822,072 => 77,822,072
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Found using pahole.
Changes in peak memory usage according to Valgrind massif:
mean soft fp64 using uint64: 1,343,991,403 => 1,342,759,331
gfxbench5 aztec ruins high 11: 63,619,971 => 63,555,571
deus ex mankind divided 148: 62,887,728 => 62,845,304
deus ex mankind divided 2890: 72,399,750 => 71,922,686
dirt showdown 676: 69,464,023 => 69,238,607
dolphin ubershaders 210: 78,359,728 => 77,822,072
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Replace the old array in each value with a hash table in each value.
Changes in peak memory usage according to Valgrind massif:
mean soft fp64 using uint64: 5,499,875,082 => 1,343,991,403
gfxbench5 aztec ruins high 11: 63,619,971 => 63,619,971
deus ex mankind divided 148: 62,887,728 => 62,887,728
deus ex mankind divided 2890: 72,402,222 => 72,399,750
dirt showdown 676: 74,466,431 => 69,464,023
dolphin ubershaders 210: 109,630,376 => 78,359,728
Run-time change for a full run on shader-db on my Haswell desktop (with
-march=native) is 1.22245% +/- 0.463879% (n=11). This is about +2.9
seconds on a 237 second run. The first time I sent this version of this
patch out, the run-time data was quite different. I had misconfigured
the script that ran the test, and none of the tests from higher GLSL
versions were run. These are generally more complex shaders, and they
are more affected by this change.
The previous version of this patch used a single hash table for the
whole phi builder. The mapping was from [value, block] -> def, so a
separate allocation was needed for each [value, block] tuple. There was
quite a bit of per-allocation overhead (due to ralloc), so the patch was
followed by a patch that added the use of the slab allocator. The
results of those two patches was not quite as good:
mean soft fp64 using uint64: 5,499,875,082 => 1,343,991,403
gfxbench5 aztec ruins high 11: 63,619,971 => 63,619,971
deus ex mankind divided 148: 62,887,728 => 62,887,728
deus ex mankind divided 2890: 72,402,222 => 72,402,222 *
dirt showdown 676: 74,466,431 => 72,443,591 *
dolphin ubershaders 210: 109,630,376 => 81,034,320 *
The * denote tests that are better now. In the tests that are the same
in both patches, the "after" peak memory usage was at a different
location. I did not check the local peaks.
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Suggested-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
D3D Booleans use a 32-bit 0/-1 representation. Because this previously
matched NIR exactly, we didn't have to really optimize for it. Now that
we have 1-bit Booleans, we need some specific optimizations to chew
through the D3D12-style Booleans.
Shader-db results on Kaby Lake:
total instructions in shared programs: 15136811 -> 14967944 (-1.12%)
instructions in affected programs: 2457021 -> 2288154 (-6.87%)
helped: 8318
HURT: 10
total cycles in shared programs: 373544524 -> 359701825 (-3.71%)
cycles in affected programs: 151029683 -> 137186984 (-9.17%)
helped: 7749
HURT: 682
total loops in shared programs: 4431 -> 4399 (-0.72%)
loops in affected programs: 32 -> 0
helped: 21
HURT: 0
total spills in shared programs: 10290 -> 10051 (-2.32%)
spills in affected programs: 2532 -> 2293 (-9.44%)
helped: 18
HURT: 18
total fills in shared programs: 22203 -> 21732 (-2.12%)
fills in affected programs: 3319 -> 2848 (-14.19%)
helped: 18
HURT: 18
Note that a large chunk of the improvement fixing regressions caused by
switching to 1-bit Booleans. Previously, our ability to optimize D3D
booleans was improved by using the D3D representation directly in NIR.
Now that NIR does 1-bit bools, we need a few more optimizations.
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Eric Anholt <eric@anholt.net>
Tested-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Tested-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
This is a squash of a few distinct changes:
glsl,spirv: Generate 1-bit Booleans
Revert "Use 32-bit opcodes in the NIR producers and optimizations"
Revert "nir/builder: Generate 32-bit bool opcodes transparently"
nir/builder: Generate 1-bit Booleans in nir_build_imm_bool
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Tested-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
We also enable it in all of the NIR drivers.
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Tested-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
We also have to add support for 1-bit integers while we're here so we
get 1-bit variants of iand, ior, and inot.
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Tested-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
This just makes it nicely scale across bit sizes.
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Tested-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Tested-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
This commit adds support for 1-bit Booleans and integers. Booleans
obviously take a value of true or false. Because we have to define the
semantics of 1-bit signed and unsigned integers, we define uint1_t to
take values of 0 and 1 and int1_t to take values of 0 and -1. 1-bit
arithmetic is then well-defined in the usual way, just with fewer bits.
The definition of int1_t and uint1_t doesn't usually matter but we do
need something for purposes of constant folding.
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Tested-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
This commit contains three related changes. First, we define boolN_t
for N = 8, 16, and 64 and move the definition of boolN_vec to the loop
with the other vec definitions. Second, there's no reason why we need
the != 0 on the source because that happens implicitly when it's
converted to bool. Third, for destinations, we use a signed integer
type and just do -(int)bool_val which will give us the 0/-1 behavior we
want and neatly scales to all bit widths.
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Tested-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
This is a squash of a bunch of individual changes:
nir/builder: Generate 32-bit bool opcodes transparently
nir/algebraic: Remap Boolean opcodes to the 32-bit variant
Use 32-bit opcodes in the NIR producers and optimizations
Generated with a little hand-editing and the following sed commands:
sed -i 's/nir_op_ball_fequal/nir_op_b32all_fequal/g' **/*.c
sed -i 's/nir_op_bany_fnequal/nir_op_b32any_fnequal/g' **/*.c
sed -i 's/nir_op_ball_iequal/nir_op_b32all_iequal/g' **/*.c
sed -i 's/nir_op_bany_inequal/nir_op_b32any_inequal/g' **/*.c
sed -i 's/nir_op_\([fiu]lt\)/nir_op_\132/g' **/*.c
sed -i 's/nir_op_\([fiu]ge\)/nir_op_\132/g' **/*.c
sed -i 's/nir_op_\([fiu]ne\)/nir_op_\132/g' **/*.c
sed -i 's/nir_op_\([fiu]eq\)/nir_op_\132/g' **/*.c
sed -i 's/nir_op_\([fi]\)ne32g/nir_op_\1neg/g' **/*.c
sed -i 's/nir_op_bcsel/nir_op_b32csel/g' **/*.c
Use 32-bit opcodes in the NIR back-ends
Generated with a little hand-editing and the following sed commands:
sed -i 's/nir_op_ball_fequal/nir_op_b32all_fequal/g' **/*.c
sed -i 's/nir_op_bany_fnequal/nir_op_b32any_fnequal/g' **/*.c
sed -i 's/nir_op_ball_iequal/nir_op_b32all_iequal/g' **/*.c
sed -i 's/nir_op_bany_inequal/nir_op_b32any_inequal/g' **/*.c
sed -i 's/nir_op_\([fiu]lt\)/nir_op_\132/g' **/*.c
sed -i 's/nir_op_\([fiu]ge\)/nir_op_\132/g' **/*.c
sed -i 's/nir_op_\([fiu]ne\)/nir_op_\132/g' **/*.c
sed -i 's/nir_op_\([fiu]eq\)/nir_op_\132/g' **/*.c
sed -i 's/nir_op_\([fi]\)ne32g/nir_op_\1neg/g' **/*.c
sed -i 's/nir_op_bcsel/nir_op_b32csel/g' **/*.c
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Tested-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Later in this series, bool is not going to imply 32-bit.
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Tested-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
This was originally added for the out-of-tree Mali driver but I think
we've all agreed it's easy enough for them to just do in their back-end.
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Tested-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Shader-db results on Kaby Lake:
total instructions in shared programs: 15072525 -> 15072525 (0.00%)
instructions in affected programs: 0 -> 0
helped: 0
HURT: 0
This helps prevent regressions in later commits.
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Tested-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Instead of looking at input_sizes[i] which contains the number of
components for each source, we look at the bit size of input_types[i].
This fixes a regression in the 1-bit boolean series though I have no
idea how we haven't seen it before now.
Fixes: 35baee5dce "nir/constant_folding: fix incorrect bit-size check"
Fixes: 9076c4e289 "nir: update opcode definitions for different bit sizes"
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Tested-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
The previous code was creating a boolean by doing an arithmetic right-
shift by 31 which produces a boolean which is true if the argument is
negative. This is the same as the expression r < 0 which is much
simpler and doesn't depend on NIR's representation of booleans.
Reviewed-by: Eric Anholt <eric@anholt.net>
The pass did not correctly handle loops ending in:
if ssa_7 {
block block_8:
/* preds: block_7 */
continue
/* succs: block_1 */
} else {
block block_9:
/* preds: block_7 */
break
/* succs: block_11 */
}
The break will get eliminated by another opt but if this pass gets
called first (as it does on RADV) we ended up inserting
instructions after the break.
Fixes: 5921a19d4b ("nir: add if opt opt_if_loop_last_continue()")
Reviewed-by: Dave Airlie <airlied@redhat.com>
I needed the same function for v3d. This was originally in d3e046e76c
("nir: Pull some of intel's image load/store format conversion to
nir_format.h") before we made am istake about simplifying the function.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
This helps a lot when debugging image load/store lowering on large
testcases. Unfortunately the Mesa enum name stuff is under src/mesa and
we can't get at it from the compiler.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
It's a reasonably well-known fact in the world of compilers that integer
divisions by constants can be replaced by a multiply, an add, and some
shifts. This commit adds such an optimization to NIR for easiest case
of udiv. Other division operations will be added in following commits.
In order to provide some additional driver control, the pass takes a
minimum bit size to optimize.
Reviewed-by: Ian Romanick ian.d.romanick@intel.com
I needed the same functions for v3d. Note that the color value in the
Intel lowering has already been cut down to image.chans num_components.
v2: Drop the half float one, since it was a 1-liner after cleanup.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Most of the bits were constant, but a few were missed. Avoids warnings
from v3d's upcoming static const bits declarations.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Here we rework force_unroll_array_access() so that we can reuse
the induction variable detection in a following patch.
Reviewed-by: Thomas Helland <thomashelland90@gmail.com>
This has thrown a few people off recently and it's good to have the
process and all the rational for it documented somewhere. A comment at
the top of nir_inline_functions seems as good a place as any.
Acked-by: Karol Herbst <kherbst@redhat.com>
Reviewed-by: Matt Turner <mattst88@gmail.com>
I don't know if one is better than the other or not but this approach
has the advantage that we never forget to copy information over and
we're not hard-coding quite as many assumptions. It's also a lot
simpler and much less code.
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Instead of having to call two different lower_gradient functions based
on whether or not it's a cube, just make lower_gradient handle cubes.
This significantly simplifies some of the logic.
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Just to make it easier to run a nir tests together.
Fixes: a0ae12ca91
("nir/algebraic: Add unit tests for bitsize validation")
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Following commits will introduce additional fields such as
guessed_trip_count. Renaming these will help avoid confusion
as our unrolling feature set grows.
Reviewed-by: Thomas Helland <thomashelland90@gmail.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
b2i can now take any size boolean in preparation for 1-bit booleans, so
the error message printed is slightly different.
Fixes: dca6cd9ce6 ("nir: Make boolean conversions sized just like the others")
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=108961
Cc: Jason Ekstrand <jason@jlekstrand.net>
Acked-by: Jason Ekstrand <jason@jlekstrand.net>
Instead of a single i2b and b2i, we now have i2b32 and b2iN where N is
one if 8, 16, 32, or 64. This leads to having a few more opcodes but
now everything is consistent and booleans aren't a weird special case
anymore.
Reviewed-by: Connor Abbott <cwabbott0@gmail.com>
Suffixes are dropped from a bunch of conversion opcodes when it makes
sense to do so. Others are kept if we really do want the bit-size
restriction.
Reviewed-by: Connor Abbott <cwabbott0@gmail.com>
All conversion opcodes require a destination size but this makes
constructing certain algebraic expressions rather cumbersome. This
commit adds support to nir_search and nir_algebraic for writing
conversion opcodes without a size. These meta-opcodes match any
conversion of that type regardless of destination size and the size gets
inferred from the sizes of the things being matched or from other
opcodes in the expression.
Reviewed-by: Connor Abbott <cwabbott0@gmail.com>
Instead of using an OrderedDict, just have a (necessarily sorted) array
of transforms and a set of opcodes.
Reviewed-by: Connor Abbott <cwabbott0@gmail.com>
Both of these things are already handled in the Value base class so we
don't need to handle them explicitly in Constant.
Reviewed-by: Connor Abbott <cwabbott0@gmail.com>
The non-failure path can be tested by just compiling mesa and then
testing it, but the failure paths won't be hit unless you make a mistake,
so it's best to test them with some unit tests.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
Before this commit, there were two copies of the algorithm: one in C,
that we would use to figure out what bit-size to give the replacement
expression, and one in Python, that emulated the C one and tried to
prove that the C algorithm would never fail to correctly assign
bit-sizes. That seemed pretty fragile, and likely to fall over if we
make any changes. Furthermore, the C code was really just recomputing
more-or-less the same thing as the Python code every time. Instead, we
can just store the results of the Python algorithm in the C
datastructure, and consult it to compute the bitsize of each value,
moving the "brains" entirely into Python. Since the Python algorithm no
longer has to match C, it's also a lot easier to change it to something
more closely approximating an actual type-inference algorithm. The
algorithm used is based on Hindley-Milner, although deliberately
weakened a little. It's a few more lines than the old one, judging by
the diffstat, but I think it's easier to verify that it's correct while
being as general as possible.
We could split this up into two changes, first making the C code use the
results of the Python code and then rewriting the Python algorithm, but
since the old algorithm never tracked which variable each equivalence
class, it would mean we'd have to add some non-trivial code which would
then get thrown away. I think it's better to see the final state all at
once, although I could also try splitting it up.
v2:
- Replace instances of "== None" and "!= None" with "is None" and
"is not None".
- Rename first_src to first_unsized_src
- Only merge the destination with the first unsized source, since the
sources have already been merged.
- Add a comment explaining what nir_search_value::bit_size now means.
v3:
- Fix one last instance to use "is not" instead of !=
- Don't try to be so clever when choosing which error message to print
based on whether we're in the search or replace expression.
- Fix trailing whitespace.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
This extension is not properly tested (testing for
GL_ARB_fragment_shader_interlock is not sufficient), and since this was
noted in review on August 28th no tests have been sent.
Revert "i965: Add INTEL_fragment_shader_ordering support."
Revert "mesa: Add GL/GLSL plumbing for INTEL_fragment_shader_ordering"
This reverts commit 03ecec9ed2.
This reverts commit 119435c877.
Cc: mesa-stable@lists.freedesktop.org
Acked-by: Jason Ekstrand <jason@jlekstrand.net>
Acked-by: Eric Anholt <eric@anholt.net>
We normally call with stderr which is unbuffered, so this won't affect
that, but it does let me call nir_print_shader(nir, fopen("log", "w+"))
from gdb and actually get the whole shader in my file.
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
lowers ceil(x) as -floor(-x)
Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Signed-off-by: Rob Clark <robdclark@gmail.com>
This reverts commit 1f29f4db1e.
For this to work the compiler must ensure that it never puts
the values that arrive to this helper into unsigned variables
at any point in its processing, since that would not apply sign
extension to the value and it would break the expectations here.
Unfortunately, we use uint64_t extensively to pass and copy
things around, so some times we get to this helper with values
that are not properly sign extended to 64-bit. Here is an example
for an 8-bit value that comes from a switch case:
(gdb) p /x x
$1 = 0xffffffd6
The value seems to have been sign extended to 32-bit at some point
getting proper sign extension, but then copied into a uint64_t
which wont' apply sign extension, breaking the expectations of
the assertion.
Reviewed-by: Juan A. Suarez <jasuarez@igalia.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Meson test has a concepts of suites, which allow tests to be grouped
together. This allows for a subtest of tests to be run only (say only
the tests for nir). A test can be added to more than one suite, but for
the most part I've only added a test to a single suite, though I've
added a compiler group that includes nir, glsl, and glcpp tests.
To use this you'll need to invoke meson test directly, instead of ninja
test (which always runs all targets). it can be invoked as:
`meson test -C builddir --suite $suitename` (meson test has addition
options that are pretty useful).
Tested-By: Gert Wollny <gert.wollny@collabora.com>
Acked-by: Eric Engestrom <eric.engestrom@intel.com>
The way nir_lower_clip_vs() works with store_output intrinsics makes a
ton of assumptions about the driver_location field.
In i965 and iris, I'd rather do this lowering early and work with
variables. v3d may want to switch to that as well, and ir3 could too,
but I'm not sure exactly what would need updating. For now, handle
both methods.
Reviewed-by: Eric Anholt <eric@anholt.net>
I posted a load of hacks before to do this, Jason suggested this,
just check the deref mode, not the variable mode and delay getting
the variable until we know the type.
avoids crashes when derefing shared memory pointers.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
It's not at all intel-specific; the formula is dictated by OpenGL and
Vulkan. The only intel-specific thing is that we need the lowering. As
a nice side-effect, the new version is variable-group-size ready.
Reviewed-by: Plamena Manolova <plamena.manolova@intel.com>
This also changes spirv_to_nir and glsl_to_nir to set them. The one
place that doesn't set them is shared memory access lowering in
nir_lower_io. That will have to be updated before any consumers of it
can effectively use these new alignments.
Reviewed-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Acked-by: Karol Herbst <kherbst@redhat.com>
The new helpers can generate any pack/unpack operation including those
for which we do not have specific opcodes and they express a bitcast in
terms of these pack/unpack operations. In particular, the new helpers
properly handle 8-bit types.
Reviewed-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
The pattern of adding or multiplying an integer by an immediate is
fairly common especially in deref chain handling. This adds a helper
for it and uses it a few places. The advantage to the helper is that
it automatically handles bit sizes for you.
Reviewed-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Reviewed-by: Karol Herbst <kherbst@redhat.com>
This assert won't catch all mistakes with this helper but it will at
least ensure that the top bits are all zero or all one which should help
catch bugs.
Reviewed-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
1. nir/nir_lower_vars_to_ssa.c:691:21: warning:
unused variable ‘var’
nir_variable *var = path->path[0]->var;
v2: Changes for some part of 'may be used uninitialized'
warnings were removed, seems like it is a compiler issue.
( Eric Engestrom <eric.engestrom@intel.com> )
Possible like this one:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=46684
This issue is flagged as duplicate but an
original one is not closed yet.
Signed-off-by: Andrii Simiklit <andrii.simiklit@globallogic.com>
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Some hardware supports source mods only for float operations. Make it
possible to skip lowering to source mods in these cases.
v2: use option flags instead of a boolean (Jason Ekstrand)
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
this helps reduce the overall code changes when a bit_size parameter is
added to nir_load_system_value
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Eric Anholt <eric@anholt.net>
Signed-off-by: Karol Herbst <kherbst@redhat.com>
this allows to replace some nir_load_system_value calls with the specific
system value constructor
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Eric Anholt <eric@anholt.net>
Signed-off-by: Karol Herbst <kherbst@redhat.com>
If the local work group size is variable it won't be available
at compile time so we can't lower it in nir_lower_system_values().
Signed-off-by: Plamena Manolova <plamena.n.manolova@gmail.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Karol Herbst <kherbst@redhat.com>
For example the following type of thing is seen in TCS from
a number of Vulkan and DXVK games:
vec1 32 ssa_557 = deref_var &oPatch (shader_out float)
vec1 32 ssa_558 = intrinsic load_deref (ssa_557) ()
vec1 32 ssa_559 = deref_var &oPatch@42 (shader_out float)
vec1 32 ssa_560 = intrinsic load_deref (ssa_559) ()
vec1 32 ssa_561 = deref_var &oPatch@43 (shader_out float)
vec1 32 ssa_562 = intrinsic load_deref (ssa_561) ()
intrinsic store_deref (ssa_557, ssa_558) (1) /* wrmask=x */
intrinsic store_deref (ssa_559, ssa_560) (1) /* wrmask=x */
intrinsic store_deref (ssa_561, ssa_562) (1) /* wrmask=x */
No shader-db changes on i965 (SKL).
vkpipeline-db results RADV (VEGA):
Totals from affected shaders:
SGPRS: 7832 -> 7728 (-1.33 %)
VGPRS: 6476 -> 6740 (4.08 %)
Spilled SGPRs: 0 -> 0 (0.00 %)
Spilled VGPRs: 0 -> 0 (0.00 %)
Private memory VGPRs: 0 -> 0 (0.00 %)
Scratch size: 0 -> 0 (0.00 %) dwords per thread
Code Size: 469572 -> 456596 (-2.76 %) bytes
LDS: 0 -> 0 (0.00 %) blocks
Max Waves: 989 -> 960 (-2.93 %)
Wait states: 0 -> 0 (0.00 %)
The Max Waves and VGPRS changes here are misleading. What is
happening is a bunch of TCS outputs are being optimised away as
they are now recognised as unused. This results in more varyings
being compacted via nir_compact_varyings() which can result in
more register pressure when they are not packed in an optimal way.
This is an existing problem independent of this patch. I've run
some benchmarks and haven't noticed any performance regressions
in affected games.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Byte ordering is :
0: V
1: U
2: Y
3: A
v2: Split refactoring of alpha channel (Lionel)
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Tapani Pälli <tapani.palli@intel.com> (v1)
Acked-by: Eric Engestrom <eric.engestrom@intel.com> (v2)
We're about to introduce AYUV support which provides its own alpha
channel. So give alpha as a parameter and set it to 1 on exising
formats.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
nir_alu_type_get_type_size takes a type as parameter and we were
passing a bit-size instead, which did what we wanted by accident,
since a bit-size of zero matches nir_type_invalid, which has a
size of 0 too.
Reviewed-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
We cannot use nir_build_alu() to create the new alu as it has no
way to know how many components of the src we will use. This
results in it guessing the max number of components from one of
its inputs.
Fixes the following CTS tests:
dEQP-VK.spirv_assembly.instruction.graphics.selection_block_order.out_of_order_frag
dEQP-VK.spirv_assembly.instruction.graphics.selection_block_order.out_of_order_geom
dEQP-VK.spirv_assembly.instruction.graphics.selection_block_order.out_of_order_tessc
dEQP-VK.spirv_assembly.instruction.graphics.selection_block_order.out_of_order_vert
Fixes: 2975422ceb ("nir: propagates if condition evaluation down some alu chains")
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
We need to update the cursor before we check if the alu use is
dominated by the if condition. Previously we were checking if
the current location of the alu instruction was dominated by
the if condition which would miss some optimisation opportunities.
Fixes: a3b4cb3458 ("nir/opt_if: Rework condition propagation")
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
This will be used on V3D to cut down the size of the VS inputs in the VPM
(memory area for sharing data between shader stages).
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
This is different from the GL_ARB_spirv pass because it generates a much
simpler data structure that isn't tied to OpenGL and mtypes.h.
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
This isn't a great solution for bit-sizes but we don't have a
particularly convenient way to get a bit size from the system value enum
and this keeps the lowering pass from changing it.
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Instead of doing our own constant folding, we just emit instructions and
let constant folding happen. This is substantially simpler and lets us
use the nir_imm_bool helper instead of dealing with the const_value's
ourselves.
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
This requires that we rework the interface a bit to use nir_builder but
that's a nice little modernization anyway.
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Missed one while converting to the nir_src_as_* helpers.
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Use nir_src_comp_as_uint() to read the proper second component, as
nir_src_as_uint() returns the first one.
v2: Use nir_src_comp_as_uint() [Jason]
Fixes: 16870de8a0 ("nir: Use nir_src_is_const and nir_src_as_* in core
code")
Signed-off-by: Juan A. Suarez Romero <jasuarez@igalia.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=108532
Tested-by: Michel Dänzer <michel.daenzer@amd.com>
Tested-by: Vinson Lee <vlee@freedesktop.org>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
The linking opts shouldn't try removing or compacting XFB varyings
in the consumer. To avoid this we copy the always_active_io flag
from the producer.
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
The conon_bit_class and canon_var_class variables got switched.
Fixes: 932c650e0b "nir/algebraic: Loosen a restriction on variables"
Reported-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Previously, we would fail if a variable had an assigned but unknown bit
size X and we tried to assign it an actual bit size. However, this is
ok because, at the time we do the search, the variable does have an
actual bit size and it will match X because of the NIR rules.
Reviewed-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
We rename some local variables in validate() to be more readable and
plumb the var through to get/set_var_bit_class instead of the var index.
Reviewed-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
This not only makes them safe for more bit sizes but it also fixes a bug
in is_zero_to_one where it would return true for constant NaN.
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Extend the pass to propagate the copies information along the control
flow graph. It performs two walks, first it collects the vars
that were written inside each node. Then it walks applying the copy
propagation using a list of copies previously available. At each node
the list is invalidated according to results from the first walk.
This approach is simpler than a full data-flow analysis, but covers
various cases. If derefs are used for operating on more memory
resources (e.g. SSBOs), the difference from a regular pass is expected
to be more visible -- as the SSA copy propagation pass won't apply to
those.
A full data-flow analysis would handle more scenarios: conditional
breaks in the control flow and merge equivalent effects from multiple
branches (e.g. using a phi node to merge the source for writes to the
same deref). However, as previous commentary in the code stated, its
complexity 'rapidly get out of hand'. The current patch is a good
intermediate step towards more complex analysis.
The 'copies' linked list was modified to use util_dynarray to make it
more convenient to clone it (to handle ifs/loops).
Annotated shader-db results for Skylake:
total instructions in shared programs: 15105796 -> 15105451 (<.01%)
instructions in affected programs: 152293 -> 151948 (-0.23%)
helped: 96
HURT: 17
All the HURTs and many HELPs are one instruction. Looking
at pass by pass outputs, the copy prop kicks in removing a
bunch of loads correctly, which ends up altering what other
other optimizations kick. In those cases the copies would be
propagated after lowering to SSA.
In few HELPs we are actually helping doing more than was
possible previously, e.g. consolidating load_uniforms from
different blocks. Most of those are from
shaders/dolphin/ubershaders/.
total cycles in shared programs: 566048861 -> 565954876 (-0.02%)
cycles in affected programs: 151461830 -> 151367845 (-0.06%)
helped: 2933
HURT: 2950
A lot of noise on both sides.
total loops in shared programs: 4603 -> 4603 (0.00%)
loops in affected programs: 0 -> 0
helped: 0
HURT: 0
total spills in shared programs: 11085 -> 11073 (-0.11%)
spills in affected programs: 23 -> 11 (-52.17%)
helped: 1
HURT: 0
The shaders/dolphin/ubershaders/12.shader_test was able to
pull a couple of loads from inside if statements and reuse
them.
total fills in shared programs: 23143 -> 23089 (-0.23%)
fills in affected programs: 2718 -> 2664 (-1.99%)
helped: 27
HURT: 0
All from shaders/dolphin/ubershaders/.
LOST: 0
GAINED: 0
The other generations follow the same overall shape. The spills and
fills HURTs are all from the same game.
shader-db results for Broadwell.
total instructions in shared programs: 15402037 -> 15401841 (<.01%)
instructions in affected programs: 144386 -> 144190 (-0.14%)
helped: 86
HURT: 9
total cycles in shared programs: 600912755 -> 600902486 (<.01%)
cycles in affected programs: 185662820 -> 185652551 (<.01%)
helped: 2598
HURT: 3053
total loops in shared programs: 4579 -> 4579 (0.00%)
loops in affected programs: 0 -> 0
helped: 0
HURT: 0
total spills in shared programs: 80929 -> 80924 (<.01%)
spills in affected programs: 720 -> 715 (-0.69%)
helped: 1
HURT: 5
total fills in shared programs: 93057 -> 93013 (-0.05%)
fills in affected programs: 3398 -> 3354 (-1.29%)
helped: 27
HURT: 5
LOST: 0
GAINED: 2
shader-db results for Haswell:
total instructions in shared programs: 9231975 -> 9230357 (-0.02%)
instructions in affected programs: 44992 -> 43374 (-3.60%)
helped: 27
HURT: 69
total cycles in shared programs: 87760587 -> 87727502 (-0.04%)
cycles in affected programs: 7720673 -> 7687588 (-0.43%)
helped: 1609
HURT: 1416
total loops in shared programs: 1830 -> 1830 (0.00%)
loops in affected programs: 0 -> 0
helped: 0
HURT: 0
total spills in shared programs: 1988 -> 1692 (-14.89%)
spills in affected programs: 296 -> 0
helped: 1
HURT: 0
total fills in shared programs: 2103 -> 1668 (-20.68%)
fills in affected programs: 438 -> 3 (-99.32%)
helped: 4
HURT: 0
LOST: 0
GAINED: 1
v2: Remove the DISABLE prefix from tests we now pass.
v3: Add comments about missing write_mask handling. (Caio)
Add unreachable when switching on cf_node type. (Jason)
Properly merge the component information in written map
instead of replacing. (Jason)
Explain how removal from written arrays works. (Jason)
Use mode directly from deref instead of getting the var. (Jason)
v4: Register the local written mode for calls. (Jason)
Prefer cf_node instead of node. (Jason)
Clarify that remove inside iteration only works in backward
iterations. (Jason)
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Calls are not used yet (functions are inlined), but since new code is
already taking them into account, do it here too. The convention here
and in other places is that no writable memory is assumed to remain
unchanged, as well as global variables.
Also, explicitly state the modes affected (instead of using the
reverse logic) in one of the apply_for_barrier_modes calls.
Suggested by Jason.
v2: Consider local vars used by a call to be conservative, SPIR-V has
such cases. (Jason)
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Also tests for removal of redundant loads, that we currently handle as
part of the copy propagation.
Note some tests involve multiple blocks and are currently DISABLED
because they (expectedly) fail.
v2: Add missing DISABLED prefix to "multi block" tests. (Jason)
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Instead of doing this as part of the existing copy_prop_vars pass.
Separation makes easier to expand the scope of both passes to be more
than per-block. For copy propagation, the information about valid
copies comes from previous instructions; while the dead write removal
depends on information from later instructions ("have any instruction
used this deref before overwrite it?").
Also change the tests to use this pass (instead of copy prop vars).
Note that the disabled tests continue to fail, since the standalone
pass is still per-block.
v2: Remove entries from dynarray instead of marking items as
deleted. Use foreach_reverse. (Caio)
(all from Jason)
Do not cache nir_deref_path. Not worthy for this patch.
Clear unused writes when hitting a call instruction.
Clean up enumeration of modes for barriers.
Move metadata calls to the inner function.
v3: For copies, use the vector length to calculate the mask.
(all from Jason)
Use nir_component_mask_t when applicable.
Rename functions for clarity.
Consider local vars used by a call to be conservative (SPIR-V has
such cases).
Comment and assert the assumption that stores and copies are
always to a deref that ends with a vector or scalar.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Note at the moment the pass called is nir_opt_copy_prop_vars, because
dead write elimination is implemented there.
Also added tests that involve identifying dead writes in multiple
blocks (e.g. the overwrite happens in another block). Those currently
fail as expected, so are marked to be skipped.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Add basic helpers for doing tests on the vars related optimization
passes. The main goal is to lower the barrier to create tests during
development and debugging of the passes. Full coverage is not a
requirement.
v2: Make find_next_intrinsic() skip blocks before 'after'. (Jason)
Move nir_imm_ivec2() to nir_builder.h. (Jason)
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
For gallium drivers where you want to do some linking at variant compile
time, you don't have the other producer/consumer shader on hand to modify.
By exposing the inner function, the driver can have the used varyings in
the compiled shader cache key and still do linking.
This is also useful for V3D, where the binning shader wants to only output
position and TF varyings. We've been removing those after nir_lower_io,
but this will be less driver-specific code and let more of the shader get
DCEed early in NIR.
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
This is needed for nir_gather_info to actually count the new textures,
since it operates solely on variables.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Dave Airlie <airlied@redhat.com>
v2: - change how the access qualifiers are accumulated
v3: - duplicate members in struct_member_decoration_cb()
- handle access qualifiers on variables
- remove access qualifiers handling in _vtn_variable_load_store()
- fix setting access qualifiers on type->array_element
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net
The b2f and b2i conversions always produce zero or one which are both
representable in every type and size. Since b2i and b2f support all bit
sizes, we can just get rid of the conversion opcode.
total instructions in shared programs: 15089335 -> 15084368 (-0.03%)
instructions in affected programs: 212564 -> 207597 (-2.34%)
helped: 896
HURT: 0
total cycles in shared programs: 369831123 -> 369826267 (<.01%)
cycles in affected programs: 2008647 -> 2003791 (-0.24%)
helped: 693
HURT: 216
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
These allows us to not support fsign.sat in the Intel compiler backend,
and that will simplify some later changes.
No shader-db changes on any Intel platform.
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Thomas Helland <thomashelland90@gmail.com>
The ssa_for_alu_src helper will correctly handle swizzles and other
source modifiers for you. The expansions for unpack_half_2x16,
pack_uvec2_to_uint, and pack_uvec4_to_uint were all broken with regards
to swizzles. The brokenness of unpack_half_2x16 was causing rendering
errors in Rise of the Tomb Raider on Intel ever since c11833ab24
which added an extra copy propagation to the optimization pipeline and
caused us to start seeing swizzles where we hadn't seen any before.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=107926
Fixes: 9ce901058f "nir: Add lowering of nir_op_unpack_half_2x16."
Fixes: 9b8786eba9 "nir: Add lowering support for packing opcodes."
Tested-by: Alex Smith <asmith@feralinteractive.com>
Tested-by: Józef Kucia <joseph.kucia@gmail.com>
Reviewed-by: Matt Turner <mattst88@gmail.com>
We already call nir_rematerialize_derefs_in_use_blocks_impl prior to
calling nir_lower_ssa_defs_to_regs_block so the assertion that all deref
uses in the block should hold. This fixes the following CTS test when
SPIR-V optimization recipe 1:
dEQP-VK.glsl.struct.local.loop_nested_struct_array_vertex
Fixes: 606eb56ab9 "intel/nir: Only lower load/store derefs"
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
If the block in which the jump is inserted is the predecessor of a phi
then we need to remove phi sources otherwise the phi may end up with
things improperly connected. This fixes the following CTS test when
dEQP is run with SPIR-V optimization recipe 1:
dEQP-VK.glsl.functions.control_flow.return_in_nested_loop_vertex
Cc: mesa-stable@lists.freedesktop.org
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
CC: Jason Ekstrand <jason@jlekstrand.net>
Fixes: 82799a5d1b8 ("nir: Add a small pass to rematerialize derefs
per-block")
Reviewed-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
The lcssa and phis_to_regs passes are used by various NIR optimizations
that modify the CFG. Putting a couple of asserts will help ensure that
we don't accidentally put derefs in phis as part of an optimization
pass.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
When we're about to re-arrange a bunch of blocks, it's a good idea to
make sure that we don't have deref uses crossing block boundaries.
Otherwise we may end up with a deref going through a phi and that would
be bad.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Cc: "18.2" <mesa-stable@lists.freedesktop.org>
This pass re-materializes deref instructions on a per-block basis to
ensure that every use of a deref occurs in the same block as the
instruction which uses it.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Cc: "18.2" <mesa-stable@lists.freedesktop.org>
In GLSL IR we cheat with switch statements and simply convert them
into loops with a single iteration. This allowed us to make use of
the existing jump instruction handling provided by the loop handing
code, it also allows dead code to be cleaned up once we have
wrapped the code in a loop.
However using loops in this way created previously unrollable loops
which limits further optimisations. Here we provide a way to unroll
loops that end in a break and have multiple other exits.
All shader-db changes are from the dolphin uber shaders. There is a
small amount of HURT shaders but in general the improvements far
exceed the HURT.
shader-db results IVB:
total instructions in shared programs: 10018187 -> 10016468 (-0.02%)
instructions in affected programs: 104080 -> 102361 (-1.65%)
helped: 36
HURT: 15
total cycles in shared programs: 220065064 -> 154529655 (-29.78%)
cycles in affected programs: 126063017 -> 60527608 (-51.99%)
helped: 51
HURT: 0
total loops in shared programs: 2515 -> 2308 (-8.23%)
loops in affected programs: 903 -> 696 (-22.92%)
helped: 51
HURT: 0
total spills in shared programs: 4370 -> 4124 (-5.63%)
spills in affected programs: 1397 -> 1151 (-17.61%)
helped: 9
HURT: 12
total fills in shared programs: 4581 -> 4419 (-3.54%)
fills in affected programs: 2201 -> 2039 (-7.36%)
helped: 9
HURT: 15
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
v2:
- only allow nir_op_inot or nir_op_b2i when alu input is 1.
- use some helpers as suggested by Jason.
v3:
- evaluate alu op for single input alu ops
- add helper function to decide if to propagate through alu
- make use of nir_before_src in another spot
shader-db IVB results:
total instructions in shared programs: 9993483 -> 9993472 (-0.00%)
instructions in affected programs: 1300 -> 1289 (-0.85%)
helped: 11
HURT: 0
total cycles in shared programs: 219476091 -> 219476059 (-0.00%)
cycles in affected programs: 7675 -> 7643 (-0.42%)
helped: 10
HURT: 1
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Since we know what side of the branch we ended up on we can just
replace the use with a constant.
All the spill changes in shader-db are from Dolphin uber shaders,
despite some small regressions the change is clearly positive.
V2: insert new constant after any phis in the
use->parent_instr->type == nir_instr_type_phi path.
v3:
- use nir_after_block_before_jump() for inserting const
- check dominance of phi uses correctly
v4:
- create some helpers as suggested by Jason.
v5 (Jason Ekstrand):
- Use LIST_ENTRY to get the phi src
shader-db results IVB:
total instructions in shared programs: 9999201 -> 9993483 (-0.06%)
instructions in affected programs: 163235 -> 157517 (-3.50%)
helped: 132
HURT: 2
total cycles in shared programs: 231670754 -> 219476091 (-5.26%)
cycles in affected programs: 143424120 -> 131229457 (-8.50%)
helped: 115
HURT: 24
total spills in shared programs: 4383 -> 4370 (-0.30%)
spills in affected programs: 1656 -> 1643 (-0.79%)
helped: 9
HURT: 18
total fills in shared programs: 4610 -> 4581 (-0.63%)
fills in affected programs: 374 -> 345 (-7.75%)
helped: 6
HURT: 0
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
and _mesa_bitcount_64 with util_bitcount_64. This fixes a build problem
in nir for platforms that don't have popcount or popcountll, such as
32bit msvc.
v2: - Fix additional uses of _mesa_bitcount added after this was
originally written
Acked-by: Eric Engestrom <eric.engestrom@intel.com> (v1)
Acked-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
It was very inconsistently handled; the only things that made use of it
were glsl_to_nir, glspirv, and nir_gather_info. In particular,
nir_lower_io completely ignored it so anyone using nir_lower_io on
64-bit vertex attributes was going to be in for a shock. Also, as of
the previous commit, it's set by every driver that supports 64-bit
vertex attributes. There's no longer any reason to have it be an option
so let's just delete it.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
We were going out of our way to disable dual-location re-mapping in NIR
only to then do the remapping in st_glsl_to_nir.cpp. Presumably, this
was so that double_inputs would be correct for the core state tracker.
However, now that we've it to gl_program::DualSlotInputs which is
unaffected by NIR lowering, we can let NIR lower things for us. The one
tricky bit here is that we have to remap the inputs_read bitfield back
to the single-slot convention for the gallium state tracker to use.
Since radeonsi is the only NIR-capable gallium driver that also supports
GL_ARB_vertex_attrib_64bit, we only have to worry about radeonsi when
making core gallium state tracker changes.
Acked-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Previously, we had two field in shader_info: double_inputs_read and
double_inputs. Presumably, the one was for all double inputs that are
read and the other is all that exist. However, because nir_gather_info
regenerates these two values, there is a possibility, if a variable gets
deleted, that the value of double_inputs could change over time. This
is a problem because double_inputs is used to remap the input locations
to a two-slot-per-dvec3/4 scheme for i965. If that mapping were to
change between glsl_to_nir and back-end state setup, we would fall over
when trying to map the NIR outputs back onto the GL location space.
This commit changes the way slot re-mapping works. Instead of the
double_inputs field in shader_info, it adds a DualSlotInputs bitfield to
gl_program. By having it in gl_program, we more easily guarantee that
NIR passes won't touch it after it's been set. It also makes more sense
to put it in a GL data structure since it's really a mapping from GL
slots to back-end and/or NIR slots and not really a NIR shader thing.
Tested-by: Alejandro Piñeiro <apinheiro@igalia.com> (ARB_gl_spirv tests)
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Previously, the back-end compiler turn image access into magic uniform
reads and there was a complex contract between back-end compiler and
driver about setting up and filling out those params. As of this
commit, both drivers now lower image_deref_load_param_intel intrinsics
to load_uniform intrinsics controlled by the driver and lower the other
image_deref_* intrinsics to image_* intrinsics which take an actual
binding table index. There are still "magic" uniforms but they are now
added and controlled entirely by the driver and that contract no longer
spans components.
This also has the side-effect of making most image use compile-time
binding table indices. Previously, all image access pulled the binding
table index from a uniform. Part of the reason for this was that the
magic uniforms made it difficult to decouple binding table indices from
the uniforms and, since they are indexed completely differently
(especially in Vulkan), it was hard to pull them apart. Now that the
driver is handling both, it's trivial to decouple the two and provide
actual binding table indices.
Shader-db results on Kaby Lake:
total instructions in shared programs: 15166872 -> 15164293 (-0.02%)
instructions in affected programs: 115834 -> 113255 (-2.23%)
helped: 191
HURT: 0
total cycles in shared programs: 571311495 -> 571196465 (-0.02%)
cycles in affected programs: 4757115 -> 4642085 (-2.42%)
helped: 73
HURT: 67
total spills in shared programs: 10951 -> 10926 (-0.23%)
spills in affected programs: 742 -> 717 (-3.37%)
helped: 7
HURT: 0
total fills in shared programs: 22226 -> 22201 (-0.11%)
fills in affected programs: 1146 -> 1121 (-2.18%)
helped: 7
HURT: 0
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
This commit expands the current memory access enum to contain the extra
two bits provided for images. We choose to follow the SPIR-V convention
of NonReadable and NonWriteable because readonly implies that you *can*
read so readonly + writeonly doesn't make as much sense as NonReadable +
NonWriteable.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
This commit moves our storage image format conversion codegen into NIR
instead of doing it in the back-end. This has the advantage of letting
us run it through NIR's optimizer which is pretty effective at shrinking
things down. In the common case of rgba8, the number of instructions
emitted after NIR is done with it is half of what it was with the
lowering happening in the back-end. On the downside, the back-end's
lowering is able to directly use predicates and the NIR lowering has to
use IFs.
Shader-db results on Kaby Lake:
total instructions in shared programs: 15166910 -> 15166872 (<.01%)
instructions in affected programs: 5895 -> 5857 (-0.64%)
helped: 15
HURT: 0
Clearly, we don't have that much image_load_store happening in the
shaders in shader-db....
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Instead of requiring 4 components, this allows them to potentially use
fewer. Both the SPIR-V and GLSL paths still generate vec4 intrinsics so
drivers which assume 4 components should be safe. However, we want to
be able to shrink them for i965.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
We have a name for that, it's called a uvec. This just makes the
function name a bit shorter. While we're here, we also add an assert
for one of the assumptions this function makes.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
There is nothing inherent about these opcodes that requires them to only
take scalars. It's very convenient if we let them take vectors as well.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Found by inspection. This doesn't help much now but we'll see this
pattern with images if you load UNORM and then store UNORM.
Shader-db results on Kaby Lake:
total instructions in shared programs: 15166916 -> 15166910 (<.01%)
instructions in affected programs: 761 -> 755 (-0.79%)
helped: 6
HURT: 0
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
This adds the "(a << N) >> M" family of mask or sign-extensions. Not a
huge win right now but this pattern will soon be generated by NIR format
lowering code.
Shader-db results on Kaby Lake:
total instructions in shared programs: 15166918 -> 15166916 (<.01%)
instructions in affected programs: 36 -> 34 (-5.56%)
helped: 2
HURT: 0
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
If it's not the right bit-size, it may not actually be the correct
extraction. For now, we'll only worry about 32-bit versions.
Fixes: 905ff86198 "nir: Recognize open-coded extract_u16"
Fixes: 76289fbfa8 "nir: Recognize open-coded extract_u8"
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
This adds support for unrolling the classic
do {
// ...
} while (false)
that is used to wrap multi-line macros. GLSL IR also wraps switch
statements in a loop like this.
shader-db results IVB:
total loops in shared programs: 2515 -> 2512 (-0.12%)
loops in affected programs: 33 -> 30 (-9.09%)
helped: 3
HURT: 0
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Now that SSA values can be derefs and they have special rules, we have
to be a bit more careful about our LCSSA phis. In particular, we need
to clean up in case LCSSA ended up creating a phi node for a deref.
This avoids validation issues with some CTS tests with the following
patch, but its possible this we could also see the same problem with
the existing unrolling passes.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
In order to be sure loop_terminator_list is an accurate
representation of all the jumps in the loop we need to be sure we
didn't encounter any other complex behaviour such as continues,
nested breaks, etc during analysis.
This will be used in the following patch.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
This will help later patches with unrolling loops that end with a
break i.e. loops the always exit on their first interation.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
This extension provides new GLSL built-in function
beginFragmentShaderOrderingIntel() that guarantees
(taking wording of GL_INTEL_fragment_shader_ordering
extension) that any memory transactions issued by
shader invocations from previous primitives mapped to
same xy window coordinates (and same sample when
per-sample shading is active), complete and are visible
to the shader invocation that called
beginFragmentShaderOrderingINTEL().
One advantage of INTEL_fragment_shader_ordering over
ARB_fragment_shader_interlock is that it provides a
function that operates as a memory barrie (instead
of a defining a critcial section) that can be called
under arbitary control flow from any function (in
contrast the begin/end of ARB_fragment_shader_interlock
may only be called once, from main(), under no control
flow.
Signed-off-by: Kevin Rogovin <kevin.rogovin@intel.com>
Reviewed-by: Plamena Manolova <plamena.manolova@intel.com>
This peephole optimization looks for a series of load/store_deref or
copy_deref instructions that copy an array from one variable to another
and turns it into a copy_deref that copies the entire array. The
pattern it looks for is extremely specific but it's good enough to pick
up on the input array copies in DXVK and should also be able to pick up
the sequence generated by spirv_to_nir for a OpLoad of a large composite
followed by OpStore. It can always be improved later if needed.
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
This pass looks for variables with vector or array-of-vector types and
narrows the type to only the components used.
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
This pass looks for array variables where at least one level of the
array is never indirected and splits it into multiple smaller variables.
This pass doesn't really do much now because nir_lower_vars_to_ssa can
already see through arrays of arrays and can detect indirects on just
one level or even see that arr[i][0][5] does not alias arr[i][1][j].
This pass exists to help other passes more easily see through arrays of
arrays. If a back-end does implement arrays using scratch or indirects
on registers, having more smaller arrays is likely to have better memory
efficiency.
v2 (Jason Ekstrand):
- Better comments and naming (some from Caio)
- Rework to use one hash map instead of two
v2.1 (Jason Ekstrand):
- Fix a couple of bugs that were added in the rework including one
which basically prevented it from running
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
This pass doesn't really do much now because nir_lower_vars_to_ssa can
already see through structures and considers them to be "split". This
pass exists to help other passes more easily see through structure
variables. If a back-end does implement arrays using scratch or
indirects on registers, having more smaller arrays is likely to have
better memory efficiency.
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
Since there's no particular reason for the index to be 0, choose an
index that is not used by other block. This is convenient when we
store "per-block" data in an array AND look for the successors
data (e.g. any kind of backwards data-flow analysis).
v2: Add a note about end_block's index. (Jason)
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Deref paths may share the same deref instructions in their chains,
e.g.
ssa_100 = deref_var A
ssa_101 = deref_struct "array_field" of ssa_100
ssa_102 = deref_array "[1]" of ssa_101
ssa_103 = deref_struct "field_a" of ssa_102
ssa_104 = deref_struct "field_a" of ssa_103
when comparing the two last deref instructions, their paths will share
a common sequence ssa_100, ssa_101, ssa_102. This patch skips to next
iteration if the deref instructions are the same. Path[0] (the var)
is still handled specially, so in the case above, only ssa_101 and
ssa_102 will be skipped.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Only used, when asserts are enabled.
Fixes an unused-variable warning with gcc-8:
../../../src/compiler/nir/nir_opt_if.c: In function 'opt_peel_loop_initial_if':
../../../src/compiler/nir/nir_opt_if.c:109:15: warning: unused variable 'prev_block' [-Wunused-variable]
nir_block *prev_block =
^~~~~~~~~~
Signed-off-by: Kai Wasserbäch <kai@dev.carbon-project.org>
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
The innermost check was added to stop us from unrolling multiple
loops in a single pass, and to stop outer loops from unrolling.
When we successfully unroll a loop we need to run the analysis
pass again before deciding if we want to go ahead an unroll a
second loop.
However the logic was flawed because it never tried to unroll any
nested loops other than the first innermost loop it found.
If this innermost loop is not unrolled we end up skipping all
other nested loops.
This unrolls a loop in a Deus Ex: MD shader on ultra settings and
also unrolls a loop in a shader from the game Prey when running
on DXVK.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Equivalent to the already existing how_declared at GLSL IR. The only
difference is that we are not adding all the declaration_type
available on GLSL, only the one that we will use on the short term. We
would add more mode if needed on the future.
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Now that all the build scripts are compatible with both Python 2 and 3,
we can flip the switch and tell Meson to use the latter.
Since Meson already depends on Python 3 anyway, this means we don't need
two different Python stacks to build Mesa.
Signed-off-by: Mathieu Bridon <bochecha@daitauha.fr>
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
Python 3 lost the long type: now everything is an int, with the right
size.
This commit makes the script compatible with Python 2 (where we check
for both int and long) and Python 3 (where we only check for int).
Signed-off-by: Mathieu Bridon <bochecha@daitauha.fr>
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
Mixing the two is a long-standing recipe for errors in Python 2, so much
so that Python 3 now completely separates them.
This commit stops treating both as if they were the same, and in the
process makes the script compatible with both Python 2 and 3.
Signed-off-by: Mathieu Bridon <bochecha@daitauha.fr>
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
The code was just reimplementing itertools.combinations_with_replacement
in a less efficient way.
This does change the order of the results slightly, but it should be ok.
Signed-off-by: Mathieu Bridon <bochecha@daitauha.fr>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
We're trying to write a unicode string (i.e decoded) to a file opened
in binary (i.e encoded) mode.
In Python 2 this works, because of the automatic conversion between
byte and unicode strings.
In Python 3 this fails though, as no automatic conversion is attempted.
This change makes the scripts compatible with both versions of Python.
Signed-off-by: Mathieu Bridon <bochecha@daitauha.fr>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
All Gen platforms had pretty similar results. (Skylake shown)
total instructions in shared programs: 14276892 -> 14276886 (<.01%)
instructions in affected programs: 484 -> 478 (-1.24%)
helped: 2
HURT: 0
total cycles in shared programs: 532578397 -> 532578395 (<.01%)
cycles in affected programs: 3522 -> 3520 (-0.06%)
helped: 1
HURT: 0
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Thomas Helland <thomashelland90@gmail.com>
No changes on any Gen platform.
v2: s/fmax/fmin/. Noticed by Thomas Helland.
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Thomas Helland <thomashelland90@gmail.com>
All Gen platforms had pretty similar results. (Skylake shown)
total instructions in shared programs: 14277220 -> 14277216 (<.01%)
instructions in affected programs: 422 -> 418 (-0.95%)
helped: 2
HURT: 0
total cycles in shared programs: 532577908 -> 532577848 (<.01%)
cycles in affected programs: 2800 -> 2740 (-2.14%)
helped: 2
HURT: 0
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Thomas Helland <thomashelland90@gmail.com>
Unlike the much older -abs(a) >= 0.0 transformation, this is not
precise. The behavior changes if the source is NaN.
No shader-db changes on any platform.
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Thomas Helland <thomashelland90@gmail.com>
Python 2 had two integer types: int and long. Python 3 dropped the
latter, as it made the int type automatically support bigger numbers.
As a result, Python 3 lost the 'L' suffix on integer litterals.
This probably doesn't make much difference when compiling the generated
C code, but adding it explicitly means that both Python 2 and 3 generate
the exact same C code anyway, which makes it easier to compare and check
for discrepencies when moving to Python 3.
Signed-off-by: Mathieu Bridon <bochecha@daitauha.fr>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
The hex() builtin returns a string containing the hexa-decimal
representation of an integer.
When the argument is not an integer, then the function calls that
object's __hex__() method, if one is defined. That method is supposed to
return a string.
While that's not explicitly documented, that string is supposed to be a
valid hexa-decimal representation for a number. Python 2 doesn't enforce
this though, which is why we got away with returning things like
'NIR_TRUE' which are not numbers.
In Python 3, the hex() builtin instead calls an object's __index__()
method, which itself must return an integer. That integer is then
automatically converted to a string with its hexa-decimal representation
by the rest of the hex() function.
As a result, we really can't make this compatible with Python 3 as it
is.
The solution is to stop using the hex() builtin, and instead use a hex()
object method, which can return whatever we want, in Python 2 and 3.
Signed-off-by: Mathieu Bridon <bochecha@daitauha.fr>
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
Whenever a non-zero stream is written to it now sets uses_streams to
true. This reflects the code in validate_geometry_shader_emissions for
GLSL.
v2: set uses_streams at gather_info instead that at spirv to nir
(Jason Ekstrand)
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
These are copied from the from the corresponding values in
ir_variable. The intention is to eventually use them in a pure-NIR
linker.
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
We weren't returning at the end of the nir_isntr_type_deref case in
nir_instrs_equal and it was falling through to the default of false.
While we're at it, make the default unreachable because all statements
in the switch now have their own returns. Had we done that before, we
would have caught this bug a long time ago.
Fixes: 19a4662a54 "nir: Add a deref instruction type"
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
Reviewed-by: Thomas Helland<thomashelland90@gmail.com>
v2: use nir_metadata_preserve
preserve metadata in case of !progress
Fixes: 074f5ba0b5
"nir: Add a simple int64 lowering pass"
Signed-off-by: Karol Herbst <kherbst@redhat.com>
Until now, we had separate passes for lowering gl_PatchVerticesIn to
a statically known constant (for TES inputs when linked against a TCS),
and a uniform in the other cases. Annoyingly, one had to be run before
nir_lower_system_values, and the other afterward. This simplified the
passes, but made life painful for the callers.
This patch combines both into a single pass. If you give it a non-zero
static count, it uses that. If you give it Mesa state slots, it turns
it back into a built-in uniform. Otherwise, it does nothing.
This also moves the i965 uniform lowering out to shared code.
v2: Make token arrays const.
Reviewed-by: Eric Anholt <eric@anholt.net>
This is controlled by a new nir_shader_compiler_options flag, and fixes
dEQP-GLES3.functional.shaders.builtin_variable.pointcoord on V3D.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
This breaks printing input/output variables with more than
4 components like mat4.
Fixes: 1beef89ad8 ("nir: prepare for bumping up max components to 16")
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
nir_sweep assumes that constants area always allocated off the variable
to which they belong. Violating this assumption causes them to get
freed early and leads to use-after-free bugs.
Fixes: 120da00975 "nir: add serialization and deserialization"
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=107366
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Tested-by: Mark Janes <mark.a.janes@intel.com>
we need rounding modes on other conversions involving floats and it is easier
to rename f2f16_undef than renaming all the other ones.
v2: rebased on master
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Acked-by: Rob Clark <robdclark@gmail.com>
Signed-off-by: Karol Herbst <kherbst@redhat.com>
also move some of the GLSL builtins over we will need for implementing
some OpenCL builtins
v2: replace NIR_IMM_FP by nir_imm_floatN_t in ported code
fix up changes caused by swizzle rework
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Signed-off-by: Karol Herbst <kherbst@redhat.com>
Python 2 has a range() function which returns a list, and an xrange()
one which returns an iterator.
Python 3 lost the function returning a list, and renamed the function
returning an iterator as range().
As a result, using range() makes the scripts compatible with both Python
versions 2 and 3.
Signed-off-by: Mathieu Bridon <bochecha@daitauha.fr>
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
In Python 2, iterators had a .next() method.
In Python 3, instead they have a .__next__() method, which is
automatically called by the next() builtin.
In addition, it is better to use the iter() builtin to create an
iterator, rather than calling its __iter__() method.
These were also introduced in Python 2.6, so using it makes the script
compatible with Python 2 and 3.
Signed-off-by: Mathieu Bridon <bochecha@daitauha.fr>
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
In Python 2, dictionaries have 2 sets of methods to iterate over their
keys and values: keys()/values()/items() and iterkeys()/itervalues()/iteritems().
The former return lists while the latter return iterators.
Python 3 dropped the method which return lists, and renamed the methods
returning iterators to keys()/values()/items().
Using those names makes the scripts compatible with both Python 2 and 3.
Signed-off-by: Mathieu Bridon <bochecha@daitauha.fr>
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>