Right now if the shader indirects on some large constant array, we see NIR
load_consts (usually from the const file) of its contents into general
registers, then indirection on the GPRs. This often results in register
allocation failures, as it's easy to go beyond the ~256 dwords of
registers per invocation.
By moving the large constants to a UBO, we can load an arbitrary number of
them. They also can be theoretically moved to the constant reg file (~2k
dwords), though you're unlikely to hit this path without an indirect load
on your large constant, and we don't yet let UBO indirect loads get moved
to constant regs.
This possibly won't work out right if we have 16-bit load_constants, but
without other MRs in flight we won't see 16-bit temps to be lowered to
this.
This allows 2 kerbal-space-program shaders to compile that previously
would fail, and fixes the new dEQP-VK and -GLES2 tests I wrote that
dynamically index a 40-element temporary array of float/vec2/vec3/vec4
with constant element initializers.
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/2789
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5810>
With indirect load_uniform, we can only encode 10b of constant base
offset. This pass detects problematic cases and peels out the high
bits of the base offset.
Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7612>
Clip & cull distances, which are compact arrays, exposed a lot of holes
because they can take up multiple slots and partially overlap.
I wanted to eliminate our dependence on knowing the layout of the
variables, as this can get complicated with things like partially
overlapping arrays, which can happen with ARB_enhanced_layouts or with
clip/cull distance arrays. This means no longer changing the layout
based on whether the i/o is part of an array or not, and no longer
matching producer <-> consumer based on the variables. At the end of the
day we have to match things based on the user-specified location, so for
simplicity this switches the entire i/o handling to be based off the
user location rather than the driver location. This means that the
primitive map may be a little bigger, but it reduces the complexity
because we never have to build a table mapping user location to driver
location, and it reduces the amount of work done at link time in the SSO
case. It also brings us closer to what the other drivers do.
While here, I also fixed the handling of component qualifiers, which was
another thing broken with clip/cull distances.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6959>
This will merge loads of UBO components together into vec4 loads. At the
same time, it improves the alignment information on our loads, fixing the
regression from the vec3 loads fix.
shader-db results:
total instructions in shared programs: 12829370 -> 8755851 (-31.75%)
total cat6 in shared programs: 145840 -> 97027 (-33.47%)
Overall results from before the vec3 fix:
total instructions in shared programs: 8019997 -> 8755851 (9.18%)
total cat6 in shared programs: 87683 -> 97027 (10.66%)
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6612>
The lowering pass may introduce vector bcsels that we need to scalarize
back out. It's unusual to have UBOs and not get any lowered to push
constants, so the flag was usually set anyway.
Fixes: 2b25240993 ("freedreno/ir3: Replace our custom vec4 UBO intrinsic
with the shared lowering.")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6612>
This reverts commit 939ddf3f67.
Intel has a separate pass for fusing FFMAs selectively. We split
these flags in commit 1b72c31e1f and
the reasoning still stands. The patch being reverted was just a
cleanup, so there should be no issue with reverting it.
Acked-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6849>
This makes it explicit that this intrinsic is only for SSBOs. For the
v3dv driver, we'll be adding a get_ubo_size intrinsic and we want to be
able to distinguish between the two.
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6812>
This also fixes the inverted last parameter of nir_lower_flrp in most drivers.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6599>
This gets us fewer comparisons in the shaders that we need to optimize
back out, and reduces backend code.
total instructions in shared programs: 11547270 -> 7219930 (-37.48%)
total full in shared programs: 334268 -> 319602 (-4.39%)
Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com>
Reviewed-by: Connor Abbott <cwabbott0@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6378>
We're trying to get rid of the options argument passed to
nir_lower_int64() and use the nir_options.lower_int64_options instead.
But before we can do that we must patch nir_lower_int64() callers
that don't have this field properly set.
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Rob Clark <robdclark@chromium.org>
Acked-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5588>
We theoretically could push these sizes to the const file
opportunistically, which appears to be what the blob does, but the
maximum number of SSBO's is way too big to do that unconditionally. Just
use resinfo to get the size for now.
Fixes on turnip: dEQP-VK.ssbo.unsized_array_length.*
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6012>
This is needed for handling drivers that use an input for loading the
face, for example Panfrost with Midgard GPUs.
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Reviewed-by: Rob Clark <robdclark@chromium.org>
Tested-by: Urja Rannikko <urjaman@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5915>
GL driver was relying on this being done by gallium, but there might be
new loops to unroll during optimizations and turnip needs it.
Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5818>
For example only some UCPs may be used by the shader, triggering asserts
that too many consts are being uploaded.
While we're at it, also fix the const size when loading UCPs, since
otherwise it doesn't correspond to what the shader is actually using.
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5752>
Gallium drivers should never see nir_var_uniform because gallium lowers
regular uniforms to a UBO. No GL driver should ever see either
nir_var_mem_shared because that's lowered in GLSL IR.
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Connor Abbott <cwabbott0@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5418>
The next step is to hook this into pscreen->finalize_nir() so it can
come before the state tracker's shader-caching.
Unfortunately we still need to do lower_io after mesa/st, so that is
split out into a post-finalize pass.
Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5372>
This is part of adding VK_KHR_shader_draw_parameters for turnip.
IR3_DP_VTXID_BASE/IR3_DP_VTXCNT_MAX offsets are changed to match what
CP_DRAW_INDIRECT_MULTI requires.
Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5635>
We don't really need the varying remapping, and it seems to somehow
happen twice when shader-cache comes into the picture. But we can
just choose not to have this problem.
Now that everything is using the ir3_point_sprite() helper, we can
flip this pipe cap without it being a massive flag-day.
Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5595>
Since binning pass variants share the same const_state with their
draw-pass counterpart, we should re-use the draw-pass variant's ubo
range analysis. So split the two functions of the existing pass
into two parts.
Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5526>
For shader-cache, we want to not have anything important in `ir3_shader`.
And to have shader variants with lower const size limits (to properly
handle cross-stage limits), we also want variants to be able to have
their own const_state.
But we still need binning pass shaders to align with their draw pass
counterpart so that the same const emit can be used for both passes.
Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5508>
Make it an rzalloc'd ptr instead of embedded struct, so it can serve as
the mem ctx for immediates. This gets rid of needing to explicitly free
the immediates, so one less thing to deal with when moving const_state.
(Also, after we move const_state to the shader variant, we won't need
one for binning pass variants)
Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5508>
It seems a lot of the lowerings being run the second time were
unnecessary. In addition, when const_state is moved to the variant,
then it will become impossible to know ahead of time whether a variant
needs additional optimizing, which means that ir3_key_lowers_nir() needs
to go away. The new approach should have the same effect, since it skips
running lowerings that are unnecessary and then skips the opt loop if no
optimizations made progress, but it will work better when we move
ir3_nir_analyze_ubo_ranges() to be after variant creation.
The one maybe controversial thing I did is to make
nir_opt_algebraic_late() always happen during variant lowering. I wanted
to avoid code duplication, and it seems to me that we should push the
_late variants as far back as possible so that later opt_algebraic runs
don't miss out on optimization opportunities.
Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5508>
The only difference between this and `const_state->num_ubos` was that
the latter is counting # of ubos loaded via `ldg` (based on UBO addrs
in push-consts). But turns out there isn't really any reason to care.
Instead just add an early return in the one code-path that cares about
the number of `ldg` UBOs.
This gets rid of one more thing we need to move from `ir3_shader` to
`ir3_shader_variant`.
Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5508>
As with const_state, this will also need to move into the variant. To
simplify that, just move it into the const_state itself, since after all
it is related.
Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5508>
This allows us to do API specific checks before removing variable
without filling nir_remove_dead_variables() with API specific code.
In the following patches we will use this to support the removal
of dead uniforms in GLSL.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4797>
Signed-off-by: Rob Clark <robdclark@chromium.org>
Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
It saves addressing math, but may cause multiple loads to be done and
bcseled due to NIR not giving us good address alignment information
currently. I don't have any workloads I know of using non-const-uploaded
UBOs, so I don't have perf numbers for it
This makes us match the GLES blob's behavior, and turnip (other than being
bindful).
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4858>
With the upcoming LDC usage in the GL driver, we don't want to be
uploading descriptors for every UBO when they aren't actually in use.
Trimming NIR's num_ubos will avoid that, and cleans up num_ubo handling
elsewhere right now.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4858>
We mostly got away with replacing a store_output with a store_var, but
for complex types like structs, that doesn't work. Once the IO has
been lowered from vars to intrinsic, we've lost the deref chains and
can't properly shadow the outputs.
This commits moves the GS lowering up so we do it before the output
variables get lowered to store_output. This way the pass works much
like nir_lower_io_to_temporaries() and cleanly shadows the outputs.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4562>
This pass lowers per-vertex input intrinsics to load_shared_ir3. This
was open coded in the TCS and GS lowering passes before - this way we
can share it. Furthermore, we'll need to run the rest of the GS
lowering earlier (before lowering IO) so we need to split off this
part that operates on the IO intrinsics first.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4562>
The new option replaces the two other _split lowering options, since
there's no need for separate options.
Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Rob Clark <robdclark@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4738>