impala-reviews mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jim Apple (Code Review)" <ger...@cloudera.org>
Subject [Impala-ASF-CR] Experiment: glibc strncmp/memcmp appears much faster than SSE4.2
Date Mon, 01 May 2017 15:34:02 GMT
Jim Apple has posted comments on this change.

Change subject: Experiment: glibc strncmp/memcmp appears much faster than SSE4.2
......................................................................


Patch Set 1:

(1 comment)

http://gerrit.cloudera.org:8080/#/c/6768/1//COMMIT_MSG
Commit Message:

PS1, Line 17: memcmp is sse4.1-based
> does it fall back to non-SSE4.1, if the CPU doesn't have SSE4.1? (maybe usi
Yes, it looks like it:

https://github.com/bminor/glibc/blob/ee19f1de0d0da24114be554fdf94243c0ec6b86c/sysdeps/x86_64/multiarch/memcmp.S

Any suggestions on how to check the codegen case? Do you think the benchmark showing 5x improvement
in running a particular query in the shell in covers that?


-- 
To view, visit http://gerrit.cloudera.org:8080/6768
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-MessageType: comment
Gerrit-Change-Id: Ie4786a4a75fdaffedd6e17cf076b5368ba4b4e3e
Gerrit-PatchSet: 1
Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-Owner: Jim Apple <jbapple-impala@apache.org>
Gerrit-Reviewer: Dan Hecht <dhecht@cloudera.com>
Gerrit-Reviewer: Jim Apple <jbapple-impala@apache.org>
Gerrit-Reviewer: Mostafa Mokhtar <mmokhtar@cloudera.com>
Gerrit-HasComments: Yes

Mime
View raw message