simd: riscv: implement RVV intrinsics - #9731
Merged
Merged
Conversation
cosmo0920
force-pushed
the
cosmo0920-implement-rvv-intrinsics
branch
from
December 18, 2024 02:12
6c44d3e to
7528308
Compare
cosmo0920
force-pushed
the
cosmo0920-implement-rvv-intrinsics
branch
from
December 18, 2024 02:17
7528308 to
7c5c3bc
Compare
cosmo0920
force-pushed
the
cosmo0920-implement-rvv-intrinsics
branch
from
December 18, 2024 05:06
7c5c3bc to
2b094f5
Compare
Contributor
|
I think we need to run the package tests once #9751 is merged to confirm all builds are good. |
cosmo0920
force-pushed
the
cosmo0920-implement-rvv-intrinsics
branch
from
December 20, 2024 02:31
1bd4543 to
ce4d6cb
Compare
cosmo0920
force-pushed
the
cosmo0920-implement-rvv-intrinsics
branch
from
December 24, 2024 05:31
ce4d6cb to
62f340d
Compare
cosmo0920
force-pushed
the
cosmo0920-implement-rvv-intrinsics
branch
from
December 31, 2024 15:22
62f340d to
aedc153
Compare
edsiper
requested changes
Jan 8, 2025
Signed-off-by: Hiroshi Hatake <hiroshi@chronosphere.io>
Signed-off-by: Hiroshi Hatake <hiroshi@chronosphere.io>
Signed-off-by: Hiroshi Hatake <hiroshi@chronosphere.io>
Signed-off-by: Hiroshi Hatake <hiroshi@chronosphere.io>
Signed-off-by: Hiroshi Hatake <hiroshi@chronosphere.io>
Signed-off-by: Hiroshi Hatake <hiroshi@chronosphere.io>
Signed-off-by: Hiroshi Hatake <hiroshi@chronosphere.io>
This is because RVV's instruction set is not having a fixed length. Instead, RVV needs to specify the VLEN for the width of vectorized calcucaltion. In this case, we assumed as 128 for VLEN. Signed-off-by: Hiroshi Hatake <hiroshi@chronosphere.io>
Signed-off-by: Hiroshi Hatake <hiroshi@chronosphere.io>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
In RISC-V intrinsics, RVV (RISC-V "Vector") extensions are existing.
In this PR, I experimented to implement RVV extensions and got succeeded to pass internal and runtime tests with turning on this RVV extensions.
This PR uses RVV v0.11 intrinsics.
Some of the RVV extensions are not corresponding one-by-one to NEON or SSE2.
Plus,
vuint8m1_tandvuint32m1_ttypes do not have fixed size.So, I assumed the fixed lengths for them is 16 like as the result of
sizeof(__m128i)orsizeof(uint8x16_t).Also, we define that
FLB_SIMD_VEC8_INST_LENmacro as an emulation for aligned behavior against SSE2 or NEON.For non-RVV, it just replaces with sizeof(flb_vector8). However, in RVV, gcc complains that there is no sized attributes for
vuint8m1_ttype.So, we define the alternative fixed size to determine the length of RVV's instruction set.
Enter
[N/A]in the box, if an item is not applicable to your change.Testing
Before we can approve your change; please submit the following in a comment:
If this is a change to packaging of containers or native binaries then please confirm it works for all targets.
ok-package-testlabel to test for all targets (requires maintainer to do).Documentation
Backporting
Fluent Bit is licensed under Apache 2.0, by submitting this pull request I understand that this code will be released under the terms of that license.