This can be used by both etnaviv and freedreno/a2xx as they are both vec4
architectures with some instructions being scalar-only.
Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
I don't know why I thought NIR_PASS always set the progress variable.
Derp.
Fixes: d41cdef2a5 ("nir: Use the flrp lowering pass instead of nir_opt_algebraic")
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Coverity CID: 1444996
Coverity CID: 1444995
Coverity CID: 1444994
Coverity CID: 1444993
Coverity CID: 1444991
Coverity CID: 1444989
I tried to be very careful while updating all the various drivers, but I
don't have any of that hardware for testing. :(
i965 is the only platform that sets always_precise = true, and it is
only set true for fragment shaders. Gen4 and Gen5 both set lower_flrp32
only for vertex shaders. For fragment shaders, nir_op_flrp is lowered
during code generation as a(1-c)+bc. On all other platforms 64-bit
nir_op_flrp and on Gen11 32-bit nir_op_flrp are lowered using the old
nir_opt_algebraic method.
No changes on any other Intel platforms.
v2: Add panfrost changes.
Iron Lake and GM45 had similar results. (Iron Lake shown)
total cycles in shared programs: 188647754 -> 188647748 (<.01%)
cycles in affected programs: 5096 -> 5090 (-0.12%)
helped: 3
HURT: 0
helped stats (abs) min: 2 max: 2 x̄: 2.00 x̃: 2
helped stats (rel) min: 0.12% max: 0.12% x̄: 0.12% x̃: 0.12%
Reviewed-by: Matt Turner <mattst88@gmail.com>
PIPE_CAP_PACKED_UNIFORMS conflates several things: Lowering uniforms i/o
at the st level instead of the backend, packing uniforms with no padding
at all, and lowering to UBOs.
Requiring backends to lower uniforms i/o for !PIPE_CAP_PACKED_UNIFORMS
leads to the driver needing to either link against the type size function
in mesa/st, or duplicating it in the backend. Given that all backends
want this lower-io as far as I can tell, just move it to mesa/st to
resolve the link issue and avoid the driver author needing to understand
st's uniforms layout.
Incidentally, fixes uniform layout failures in nouveau in:
dEQP-GLES2.functional.shaders.struct.uniform.sampler_nested_fragment
dEQP-GLES2.functional.shaders.struct.uniform.sampler_nested_vertex
dEQP-GLES2.functional.shaders.struct.uniform.sampler_array_fragment
dEQP-GLES2.functional.shaders.struct.uniform.sampler_array_vertex
and I think in Lima as well.
v2: fix indents
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
When I implemented opt_if_loop_last_continue() I had restricted
this pass from moving other if-statements inside the branch opposite
the continue. At the time it was causing a bunch of spilling in
shader-db for i965.
However Samuel Pitoiset noticed that making this pass more aggressive
significantly improved the performance of Doom on RADV. Below are
the statistics he gathered.
28717 shaders in 14931 tests
Totals:
SGPRS: 1267317 -> 1267549 (0.02 %)
VGPRS: 896876 -> 895920 (-0.11 %)
Spilled SGPRs: 24701 -> 26367 (6.74 %)
Code Size: 48379452 -> 48507880 (0.27 %) bytes
Max Waves: 241159 -> 241190 (0.01 %)
Totals from affected shaders:
SGPRS: 23584 -> 23816 (0.98 %)
VGPRS: 25908 -> 24952 (-3.69 %)
Spilled SGPRs: 503 -> 2169 (331.21 %)
Code Size: 2471392 -> 2599820 (5.20 %) bytes
Max Waves: 586 -> 617 (5.29 %)
The codesize increases is related to Wolfenstein II it seems largely
due to an increase in phis rather than the existing jumps.
This gives +10% FPS with Doom on my Vega56.
Rhys Perry also benchmarked Doom on his VEGA64:
Before: 72.53 FPS
After: 80.77 FPS
v2: disable pass on non-AMD drivers
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> (v1)
Acked-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
i965 does this, and st's tgsi path does this. st/nir did not.
Cuts 138MB of memory from a DiRT Rally trace, which is about 44%
of the total GLSL IR memory.
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
We don't need to determine the number of uniform slots here, it's
already available as prog->Parameters->NumParameterValues. The way we
previously determined the number of slots was also broken for
PackedDriverUniformStorage, where we would add loc (in dwords) and
type_size() (in vec4s).
Signed-off-by: Kristian H. Kristensen <hoegsberg@chromium.org>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Rob Clark <robdclark@gmail.com>
This fixes a segfault when we try to access the array using a
-1 when the array wasn't allocated in the first place.
Before 7536af670b we would just access a pre-allocated array
that was also load/stored to/from the shader cache. But now the
cache will no longer allocate these arrays if they are empty.
The change resulted in tests such as the following segfaulting
when run with a warm shader cache.
tests/spec/arb_arrays_of_arrays/execution/sampler/fs-struct-const-index.shader_test
glsl_to_nir() is still missing support for converting certain
functions to NIR, so for those we use the GLSL IR optimisations
to remove the functions.
Reviewed-by: Eric Anholt <eric@anholt.net>
Now that we have a loop unrolling cost function and loop unrolling isn't
going to kill us the moment we have a 64-bit op in a loop, we can go
ahead and move 64-bit lowering later. This gives us the opportunity to
do more optimizations and actually let the full optimizer run even on
64-bit ops rather than hoping one round of opt_algebraic will fix
everything. This substantially reduces both fp64 shader compile times
and the resulting code size.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Instead of trusting the caller to already have created a softfp64
function shader and added all its functions to our shader, we simply
take the softfp64 shader as an argument and do the function inlining
ouselves. This means that there's no more nasty functions lying around
that the caller needs to worry about cleaning up.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Note that locations can be set in different units, and the multiplier
argument caters to supporting these different units. For example,
st_glsl_to_nir uses dwords (4 bytes) so the multiplier should be 4,
while tgsi_to_nir uses bytes, so the multiplier should be 16.
Signed-Off-By: Timur Kristóf <timur.kristof@gmail.com>
Tested-by: Andre Heider <a.heider@gmail.com>
Tested-by: Rob Clark <robdclark@gmail.com>
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
The nir_lower_uniforms_to_ubo function is useful outside of
mesa/state_tracker, and in fact is needed to produce NIR for
drivers that have the PIPE_CAP_PACKED_UNIFORMS capability.
Signed-Off-By: Timur Kristóf <timur.kristof@gmail.com>
Tested-by: Andre Heider <a.heider@gmail.com>
Tested-by: Rob Clark <robdclark@gmail.com>
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
This might enough for iris and possible r600 (when it gets NIR)
This appears to work for iris.
v2:
* change cap return so DOUBLES == 2 means sw emu
v3:
* Refactor using int64/doubles lowering options which were added
into nir options
* Remove DOUBLES == 2 added in v2
[jordan: Remove "2" value on PIPE_CAP_DOUBLES]
[jordan: Use lowering options added to nir options]
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Acked-by: Jason Ekstrand <jason@jlekstrand.net>
Eric and I would like a bitmask of which samplers are used, similar to
prog->SamplersUsed, but available in NIR. The linker uses SamplersUsed
for resource limit checking, but later optimizations may eliminate more
samplers. So instead of propagating it through, we gather a new one.
While there, we also gather the existing textures_used_by_txf bitmask.
Gathering these bitfields in nir_shader_gather_info is awkward at best.
The main reason is that it introduces an ordering dependency between the
two passes. If gathering runs before lower_samplers_as_deref, it can't
look at var->data.binding. If the driver doesn't use the full lowering
to texture_index/texture_array_size (like radeonsi), then the gathering
can't use those fields. Gathering might be run early /and/ late, first
to get varying info, and later to update it after variant lowering. At
this point, should gathering work on pre-lowered or post-lowered code?
Pre-lowered is also harder due to the presence of structure types.
Just doing the gathering when we do the lowering alleviates these
ordering problems. This fixes ordering issues in i965 and makes the
txf info gathering work for radeonsi (though they don't use it).
Reviewed-by: Eric Anholt <eric@anholt.net>
Today, st always sets LowerCombinedClipCullDistance, causing the GLSL IR
lowering to run, giving us vec4[2] arrays. I would like to disable this
and instead run the NIR lowering so that we get compact float[] arrays
instead.
Calling the new pass is a noop if the GLSL IR pass has already run, so
it's safe to call the pass unconditionally.
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
I want to reuse this for built-in shaders.
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Tested-by: Rob Clark <robdclark@gmail.com>
Tested-by: Eric Anholt <eric@anholt.net>
Without this we do not end up with a deterministic NIR because
temporary register variables are added in random order. NIR must
be deterministic because we use it to produce a sha for the
radeonsi backends disk cache.
This fixes the shader cache for a bunch of shaders.
Another positive is that this results in a large reduction in the
size of the NIR that the state tracker stores to the disk cache.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Signed-off-by: Karol Herbst <kherbst@redhat.com>
Acked-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Signed-off-by: Karol Herbst <kherbst@redhat.com>
Acked-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Signed-off-by: Karol Herbst <kherbst@redhat.com>
Acked-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Signed-off-by: Karol Herbst <kherbst@redhat.com>
Acked-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
If the TCS and TES are linked together, we can simply replace the TES's
gl_PatchVerticesIn system value with a constant, possibly allowing extra
optimization or letting the driver avoid uploading a special value.
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
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>
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>
The functional change here is moving the nir_lower_io_to_scalar_early()
calls inside st_nir_link_shaders() and moving the st_nir_opts() call
after the call to nir_lower_io_arrays_to_elements().
This fixes a bug with the following piglit test due to the current code
not cleaning up dead code after we lower arrays. This was causing an
assert in the new duplicate varyings link time opt introduced in
70be9afccb.
tests/spec/glsl-1.10/execution/vsfs-unused-array-member.shader_test
Moving the nir_lower_io_to_scalar_early() calls also allows us to tidy
up the code a little and merge some loops.
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>
This will help the new opt introduced in the following patches
allowing us to remove extra duplicate varyings.
Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
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>
LinkedTransformFeedback is normally populated, which had nerf'd varying
packing since the check was introduced.
Fixes: dbd52585fa st/nir: Disable varying packing when doing transform feedback.
Signed-off-by: Rob Clark <robdclark@gmail.com>
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
ffs() just returns the bit that is set, we need to know what
stage that bit represents so use u_bit_scan() instead.
Fixes: 2ca5d9548f ("st/glsl_to_nir: gather next_stage in shader_info")
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
On scalar ISAs, nir_lower_io_to_scalar_early enables significant
optimizations. However, on vector ISAs, it is counterproductive and
impedes optimal codegen. This patch only calls
nir_lower_io_to_scalar_early for scalar ISAs. It appears that at present
there are no upstreamed drivers using Gallium, NIR, and a vector ISA, so
for existing code, this should be a no-op. However, this patch is
necessary for the upcoming Panfrost (Midgard) and Lima (Utgard)
compilers, which are vector.
With this patch, Panfrost is able to consume NIR directly, rather than
TGSI with the TGSI->NIR conversion.
For how this affects Lima, see
https://www.mail-archive.com/mesa-dev@lists.freedesktop.org/msg189216.html
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
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>
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>