impala-dev 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] IMPALA-4058: benchmark byteswap on misaligned memory
Date Thu, 01 Sep 2016 21:25:00 GMT
Jim Apple has posted comments on this change.

Change subject: IMPALA-4058: benchmark byteswap on misaligned memory
......................................................................


Patch Set 1:

(2 comments)

http://gerrit.cloudera.org:8080/#/c/4290/1/be/src/benchmarks/bswap-benchmark.cc
File be/src/benchmarks/bswap-benchmark.cc:

PS1, Line 120:  perfectly aligned of SIMD 
I'm not sure these are guaranteed to be aligned. I assume the array that vector allocates
on the heap is only aligned as much as the memory manager and the underlying data type require.

You can force it by using posix_memalign, or just picking the starting byte for data.inbuffer
based on the alignment of the first byte in the vector.


PS1, Line 122: vector
const, here and in the range-based for loop


-- 
To view, visit http://gerrit.cloudera.org:8080/4290
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-MessageType: comment
Gerrit-Change-Id: Ia9dfcca718b2456488d0fb4407d3817c796bfad0
Gerrit-PatchSet: 1
Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-Owner: Tim Armstrong <tarmstrong@cloudera.com>
Gerrit-Reviewer: Jim Apple <jbapple@cloudera.com>
Gerrit-HasComments: Yes

Mime
View raw message