[
https://issues.apache.org/jira/browse/HADOOP-19724?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18039897#comment-18039897
]
ASF GitHub Bot commented on HADOOP-19724:
-----------------------------------------
steveloughran commented on PR #8031:
URL: https://github.com/apache/hadoop/pull/8031#issuecomment-3563411435
@PeterPtroc as noted, @leiwen2025 can help here.
@leiwen2025 -can you look at this PR as is and review it. Ideally: check it
out and do a -Pnative build running the native tests.
If you two are using different instructions, how do they differ.
Having just looked at what clmul/clmulh does, I can see why it offers
benefits
* how common is the instruction?
* @leiwen2025 how does your vectorized compare? This opcode is intended to
be pipelined and the opcode is designed for these kind of encryption/checksum
algorithms.
Looking at #7912 it's calling vclmul.vv -this is generally going to be
faster, isn't it?
Which means that while the code is more complex, ultimately it's going be
the best option on cores with the right feature flaggs.
This makes me think that this one can go in but the vector one goes in as
the followup, with the choice of operation dependent on feature, with priority
of: vclmul, cmul, classic.
> [RISC-V] Add rv bulk CRC32 (non-CRC32C) optimized path
> -------------------------------------------------------
>
> Key: HADOOP-19724
> URL: https://issues.apache.org/jira/browse/HADOOP-19724
> Project: Hadoop Common
> Issue Type: Sub-task
> Components: hadoop-common
> Affects Versions: 3.5.0
> Reporter: Ptroc
> Priority: Major
> Labels: native, pull-request-available, risc-v
>
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
