[ 
https://issues.apache.org/jira/browse/HADOOP-11505?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15281356#comment-15281356
 ] 

Steve Loughran commented on HADOOP-11505:
-----------------------------------------

the ASF build infra is shared across all projects; if there is a power machine 
already there then the assistance would be welcome in setting up the build 
process for the scheduled builts


A harder problem is: getting people to care that they've broken it. We get that 
today with <1 JVM, some of the build machines having enough network differences 
for some tests to fail only on them (TestDNS)... the instability means that 
people don't immediately panic when a build fails (or: assume that they just 
broke it).  There's also the fact that the JVM will be the IBM JVM, which isn't 
used for dev and test by most people. Accordingly, what could appear to be a 
PPC failure could actually be a JVM failure. I'll particularly call out 
Kerberos here: someone needs to test IBM JVM + PPC in kerberos mode, and I'm 
glad to say it won't be me.

What would be good would if someone in your org got involved in helping to test 
ASF builds first to make sure they are all happy, and get involved in helping 
to make sure those nightly builds —and even more so, any release candidates— 
are all working.

> 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
>            Priority: Critical
>             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-11505.008.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)

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to