hadoop-common-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Colin Patrick McCabe (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HADOOP-11887) Introduce Intel ISA-L erasure coding library for the native support
Date Wed, 02 Sep 2015 00:08:46 GMT

    [ https://issues.apache.org/jira/browse/HADOOP-11887?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14726452#comment-14726452
] 

Colin Patrick McCabe commented on HADOOP-11887:
-----------------------------------------------

bq. Thanks Colin for this another round of review and very good comments. I addressed most
of them and the left one is about how to invoke the unit test. I agree with you that current
way isn't good as it links the test codes into the production binary. So I will wrap the test
codes into an executable program as I previously did.

OK.

bq. The questions in this way are: 1) how to tell the situation in the POM file that ISA-L
isn't available so the test shouldn't be invoked..... How about: we always run the test but
in the test program we check if ISA-L is available or not, if not available it will exit (0)
in the beginning, which wouldn't be regarded as failure.

That would be fine.  I suppose you would need to use dlopen / dlsym to load the symbols in
that case.  You would also need to build the test program even when ISA-L was not installed.

If that's tricky, then perhaps you could add a shell code snippet in the pom.xml that checks
to see if the test binary exists before it runs it.

bq. The other problem is, how to come up the test program on Windows platform? It looks pretty
heavy to write up a new native (test) program for Windows. Is it acceptable if we only support
this test program for *nix platforms? Please help confirm these, thanks!

It should be OK as long as it doesn't break the Windows build.

Thanks, [~drankye].

> Introduce Intel ISA-L erasure coding library for the native support
> -------------------------------------------------------------------
>
>                 Key: HADOOP-11887
>                 URL: https://issues.apache.org/jira/browse/HADOOP-11887
>             Project: Hadoop Common
>          Issue Type: Sub-task
>          Components: io
>            Reporter: Kai Zheng
>            Assignee: Kai Zheng
>         Attachments: HADOOP-11887-v1.patch, HADOOP-11887-v2.patch, HADOOP-11887-v3.patch,
HADOOP-11887-v4.patch, HADOOP-11887-v5.patch, HADOOP-11887-v5.patch
>
>
> This is to introduce Intel ISA-L erasure coding library for the native support, via dynamic
loading mechanism (dynamic module, like *.so in *nix and *.dll on Windows).



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Mime
View raw message