The GL_ARB_shader_ballot spec says that gl_SubGroupSizeARB is declared
as a uniform. This means that it cannot change across an invocation
such as a draw call or a compute dispatch. For compute shaders, we're
ok because we only ever use one dispatch size. For fragment, however,
the hardware dynamically chooses between SIMD8 and SIMD16 which violates
the spec. Instead, let's just pick a subgroup size based on the shader
stage. The fixed size we choose for compute shaders is a bit higher
than strictly needed but there's no real harm in that. The advantage is
that, if they do anything interesting with the value, NIR will see it as
an immediate and can optimize better.
Acked-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Ballot intrinsics return a bitfield of subgroups. In GLSL and some
SPIR-V extensions, they return a uint64_t. In SPV_KHR_shader_ballot,
they return a uvec4. Also, some back-ends would rather pass around
32-bit values because it's easier than messing with 64-bit all the time.
To solve this mess, we make nir_lower_subgroups take a new parameter
called ballot_bit_size and it lowers whichever thing it gets in from the
source language (uint64_t or uvec4) to a scalar with the specified
number of bits. This replaces a chunk of the old lowering code.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
This lets you easily build integer immediates of arbitrary bit size.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
The SUBGROUP_*_MASK system values are uint64_t when coming in from GLSL
but uvec4 when coming in from SPIR-V. Lowering based on type allows us
to nicely handle both.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
This way they can return either a uvec4 or a uint64_t. At the moment,
this is a no-op since we still always return a uint64_t.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
This would be useful a number of places
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
This commit pulls nir_lower_read_invocations_to_scalar along with most
of the guts of nir_opt_intrinsics (which mostly does subgroup lowering)
into a new nir_lower_subgroups pass. There are various other bits of
subgroup lowering that we're going to want to do so it makes a bit more
sense to keep it all together in one pass. We also move it in i965 to
happen after nir_lower_system_values to ensure that because we want to
handle the subgroup mask system value intrinsics here.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
We're going to want subgroup ID for SPIR-V subgroups eventually anyway.
We really only want to push one and calculate the other from it. It
makes a bit more sense to push the subgroup ID because it's simpler to
calculate and because it's a real API thing. The only advantage to
pushing the base thread ID is to avoid a single SHL in the shader.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Previously, brw_nir_lower_intrinsics added the param and then emitted a
load_uniform intrinsic to load it directly. This commit switches things
over to use a specific NIR intrinsic for the thread id. The one thing I
don't like about this approach is that we have to copy thread_local_id
over to the new visitor in import_uniforms.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
v2 (Jason Ekstrand):
- Various whitespace cleanups
- Add helpers for reading/writing objects
- Rework derefs
- [de]serialize nir_shader::num_*
- Fix uses of blob_reserve_bytes
- Use a bitfield struct for packing tex_instr data
v3:
- Zero nir_variable struct on deserialization. (Jordan)
- Allow nir_serialize.h to be included in C++. (Jordan)
- Handle NULL info.name. (Jason)
- Set info.name to NULL when name is NULL. (Jordan)
Acked-by: Timothy Arceri <tarceri@itsqueeze.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Acked-by: Jason Ekstrand <jason@jlekstrand.net>
Previously the values were calculated by just shifting ~0 by the
invocation ID. This would end up including bits that are higher than
gl_SubGroupSizeARB. The corresponding CTS test effectively requires that
these high bits be zero so it was failing. There is a Piglit test as
well but this appears to checking the wrong values so it passes.
For the two greater-than bitmasks, this patch adds an extra mask with
(~0>>(64-gl_SubGroupSizeARB)) to force these bits to zero.
Fixes: KHR-GL45.shader_ballot_tests.ShaderBallotBitmasks
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=102680#c3
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Cc: mesa-stable@lists.freedesktop.org
Signed-off-by: Neil Roberts <nroberts@igalia.com>
i965 turns fddx/fddy into their coarse/fine variants based on the
ctx->Hint.FragmentShaderDerivative setting. It needs to know whether
this can impact a shader in order to better guess NOS settings.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
This commit fixes two issues: First, we were returning false regardless
of whether or not the function made progress. Second, we were calling
nir_metadata_preserve far more often than needed; we only need to call
it once per impl.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Also needed in freedreno/ir3.
Signed-off-by: Rob Clark <robdclark@gmail.com>
Reviewed-by: Eric Engestrom <eric.engestrom@imgtec.com>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
The pass only looks at var load/store intrinsics, not input load/store
intrinsics, so assert that we don't see the other type.
v2: Adjust comment indentation.
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
v2:
- Use helper to add a new source to the texture instruction.
v3:
- Use nir_tex_instr_src_index() to simplify the patch (Jason).
Signed-off-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
This is intended to be called before nir_lower_io() so that we
can do some linking optimisations with the results. It can also
be used with drivers that don't use nir_lower_io() at all such
as RADV.
v2: pass mode mask rather than first and last stage integer.
Reviewed-by: Eric Anholt <eric@anholt.net>
This looks like a copy+paste error. They don't actually write into that
variable as would be implied by putting the return there.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Cc: mesa-stable@lists.freedesktop.org
We didn't fold correctly in the case of 0x1 because we never let the
loop counter hit 0. Switching it to bit >= 0 solves this problem.
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Connor Abbott <cwabbott0@gmail.com>
Cc: mesa-stable@lists.freedesktop.org
For TGSI-based drivers, st_glsl_to_tgsi records this information.
For NIR-based drivers, nir_shader_gather_info() will do so.
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
With the ssao demo from Vulkan demos:
radv/rx480: 440->440fps
anv/haswell: 24->34 fps
The demo does a 0->32 loop across a ubo with 32 members.
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
I've been doing this inside of vc4, but vc5 wants it as well and it may be
useful for other drivers (Intel has a related path for pre-gen6 with MRT,
and freedreno had a TGSI path for it at one point).
This required defining a common enum for the standard comparison
functions, but other lowering passes are likely to also want that enum.
v2: Add to meson.build as well.
Acked-by: Rob Clark <robdclark@gmail.com>
This was missed in a rebase, and doesn't affect radv or anv, only i965.
Signed-off-by: Dylan Baker <dylanx.c.baker@intel.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
In truth gtest is an external dependency that upstream expects you to
"vendor" into your own tree. As such, it makes sense to treat it more
like a dependency than an internal library, and collect it's
requirements together in a dependency object.
v2: - include with -isystem instead of setting compiler args (Eric)
Signed-off-by: Dylan Baker <dylanx.c.baker@intel.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
This allows building and installing the Intel "anv" Vulkan driver using
meson and ninja, the driver has been tested against the CTS and has
seems to pass the same series of tests (they both segfault when the CTS
tries to run wayland wsi tests).
There are still a mess of TODO, XXX, and FIXME comments in here. Those
are mostly for meson bugs I'm trying to fix, or for additional things to
implement for other drivers/features.
I have configured all intermediate libraries and optional tools to not
build by default, meaning they will only be built if they're pulled in
as a dependency of a target that will actually be installed) this allows
us to avoid massive if chains, while ensuring that only the bits that
need to be built are.
v2: - enable anv, x11, and wayland by default
- add configure option to disable valgrind
v3: - fix typo in meson_options (Nicholas)
v4: - Remove dead code (Eric)
- Remove change to generator that was from v0 (Eric)
- replace if chain with loop (Eric)
- Fix typos (Eric)
- define HAVE_DLOPEN for both libdl and builtin dl cases (Eric)
v5: - rebase on util string buffer implementation
Signed-off-by: Dylan Baker <dylanx.c.baker@intel.com>
Reviewed-by: Eric Anholt <eric@anholt.net> (v4)
The initial helpers add support for removing unused varyings between
stages.
V2:
- Moved the io mask helper function into this file rather than
nir.h so it's not used elsewhere considering it doesn't handle
all corner cases.
- Use bitmask rather than hash table to handle tcs outputs (Ken)
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Will be used in nir link pass to decided if we can remove a varying
or not.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eduardo Lima Mitev <elima@igalia.com>
This being declared bool means it won't get merged with the previous
bitfields, this seems like an oversight rather than deliberate.
Noticed when running pahole.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Clang warns:
warning: absolute value function 'fabsf' given an argument of type
'const float64_t' (aka 'const double') but has parameter of type 'float'
which may cause truncation of value [-Wabsolute-value]
float64_t dst = bit_size == 64 ? fabs(src0) : fabsf(src0);
The type of the ternary expression will be the common type of fabs() and
fabsf(): double. So fabsf(src0) will be implicitly converted to double.
We may as well just convert src0 to double before a call to fabs() and
remove the needless complexity, à la
float64_t dst = fabs(src0);
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Eric Engestrom <eric.engestrom@imgtec.com>
The optimizations are only valid for 32-bit integers. They were
mistakenly firing for 64-bit integers as well.
Cc: mesa-stable@lists.freedesktop.org
Reviewed-by: Matt Turner <mattst88@gmail.com>
This is a further lowering of default-block uniform loads that transforms
load_uniform intrinsics into load_ubo intrinsics. This simplifies the rest
of the backend.
v2: transform from load_uniform instead of straight from variables
Reviewed-by: Eric Anholt <eric@anholt.net>
This pass is a replacement for the nir_lower_samplers pass, which has the
advantage of keeping sampler references as derefs. This allows a unified
treatment of texture instructions and image intrinsics in the backend.
Some drivers prefer to treat gl_FragCoord as a system value rather than
a fragment shader input, see Const.GLSLFragCoordIsSysVal.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
