The result is about +5-ish fps in Doom Eternal.
It turns out that the location of position exports matters more
than we thought, and it's actually better to keep them at the bottom
for culling shaders rather than schedule it up to the top.
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10525>
This makes culling shaders more efficient because they split the
shader in two parts. It is better to optimize before this split
happens.
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10525>
Currently we don't enable it on any chip by default, but
we plan to enable it soon on GFX10.3 when we are comfortable
with its performance.
RADV_PERFTEST=nggc environment variable enables it on GFX10+ GPUs.
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10525>
These are very straightforward as they just copy data from
the newly added shader arguments.
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10525>
Add new shader arguments in RADV for:
- NGG culling settings
- Viewport transform
These will be used by NGG culling shaders.
Additionally, some tweaks are made to some config registers
in order to make culling shaders more efficient.
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10525>
We need to emit viewport transform information for culling shaders.
This is used for small primitive culling.
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10525>
Culling is traditionally done by the rasterizer, but that
can be a bottleneck when an app creates a large number
of primitives. Eg. a lot of tiny triangles reduce the
rasterziation efficiency.
NGG makes it possible for the shader to check primitives
and delete those that it can prove are not needed.
After this is done, we have to repack the surviving invocations
so they remain compact. This also saves bandwidth, because
some memory loads are only executed by those vertices that
survived the culling.
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10525>
The algorithms were originally implemented by Marek Olšák,
hence the copyright to AMD.
This commit just ports the LLVM based implementation to NIR,
using the new intrinsics added earlier.
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10525>
This migration was done with libclang-based automatic tooling, which
performed these replacements:
* Operand(uint8_t) -> Operand::c8
* Operand(uint16_t) -> Operand::c16
* Operand(uint32_t, false) -> Operand::c32
* Operand(uint32_t, bool) -> Operand::c32_or_c64
* Operand(uint64_t) -> Operand::c64
* Operand(0) -> Operand::zero(num_bytes)
Casts that were previously used for constructor selection have automatically
been removed (e.g. Operand((uint16_t)1) -> Operand::c16(1)).
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11653>
If the allocated_vec map contains a different RegType
for the elements, ensure that the size matches exactly.
Otherwise, it could happen that extracting a dword
element matched with a subdword element.
No fossil-db changes.
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11823>
Splitting self-intersecting register swaps into
3 sections was unnecessary and only worked because
the middle section was always empty for full dword
swaps.
No fossil-db changes.
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11823>
This allows more potential compiler optimizations if the value is a
constant or from a scalar load.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11579>
Might fix spurious failures on GFX6 and some GFX7 chips.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11675>
Android.mk files haven't really been supported by Mesa devs for a long
time. Most of us have been willing to update Makefile.sources if we
remember and sometimes we try to blind code some Android.mk for a new
generator. However, the reality is that it breaks regularly and ends up
being maintained by the Android community. To address this problem
another approach was implemented in !10183 utilizing the maintained
meson build system. The old Android.mk files are no longer required.
This commit was created with the following commands:
git rm **/Android.mk
git rm **/Android.*.mk
git rm **/Makefile.sources
git rm CleanSpec.mk
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/4487
Acked-by: Roman Stratiienko <r.stratiienko@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9728>
We should always prevent 1 CU from executing VS and GS waves
to prevent a deadlock.
Fixes: c377f45c18 "radeonsi/gfx10: rewrite late alloc computation"
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11754>
Always execute the bbox code regardless of negative W, and then simply
use || to discard the result if any W is negative. This is expected to be
rare. (it only happens when a primitive intersects the near plane)
This allows us to eliminate the else statement, which is no longer
executed for accepted primitives with negative W, which are the only
primitives that needed the else branch.
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11754>
Qualcomm has a mode with a subgroup size of 128, so just emitting larger
integer operations and then lowering them later isn't an option. This
makes the pass able to handle the lowering itself, so that we don't have
to go down to 64-thread wavefronts when ballots are used.
(The GLSL and legacy SPIR-V extensions only support a maximum of 64
threads, but I guess we'll cross that bridge when we come to it...)
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6752>
Lower it to a vote instead of a ballot. This was only used for AMD, and
in that case they're pretty much the same. However Qualcomm has a vote
builtin, which we want to use instead of ballots.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Acked-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6752>
