[
https://issues.apache.org/jira/browse/HADOOP-11505?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15047265#comment-15047265
]
Chris Nauroth commented on HADOOP-11505:
----------------------------------------
bq. Is there any way of telling maven about the ordering requirement?
The last time I looked at the hadoop-maven-plugins problem, it really looked
like a Maven bug. Maven is supposed to be able to understand the dependency
relationships of modules in the current build (what they call the "reactor").
It seems it doesn't identify the relationship though if one of those modules is
a Maven plugin needed by a subsequent module. I believe {{mvn compile}} at the
root of the source tree wasn't working if the local Maven cache didn't have
hadoop-maven-plugins, and people were forced to do {{mvn install}} first.
The case pointed out by Allen is a little different from that though.
Pre-commit will build only hadoop-mapreduce-client-nativetask if a patch only
contains code in that module, and Maven won't have any reason to trigger an
automatic build of hadoop-common first. The Yetus Hadoop personality already
handles some of our special cases like this by forcing a hadoop-common build.
I think the only solutions I see for this would be to make another change to
the Yetus Hadoop personality, or duplicate the code. I'd prefer not to do the
latter. The code is non-trivial enough that I'd prefer not to duplicate.
> Various native parts use bswap incorrectly and unportably
> ---------------------------------------------------------
>
> Key: HADOOP-11505
> URL: https://issues.apache.org/jira/browse/HADOOP-11505
> Project: Hadoop Common
> Issue Type: Bug
> Affects Versions: 3.0.0
> Reporter: Colin Patrick McCabe
> Assignee: Alan Burlison
> Fix For: 3.0.0
>
> Attachments: HADOOP-11505.001.patch, HADOOP-11505.003.patch,
> HADOOP-11505.004.patch, HADOOP-11505.005.patch, HADOOP-11505.006.patch,
> HADOOP-11505.007.patch
>
>
> hadoop-mapreduce-client-nativetask fails to use x86 optimizations in some
> cases. Also, on some alternate, non-x86, non-ARM architectures the generated
> code is incorrect. Thanks to Steve Loughran and Edward Nevill for finding
> this.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
