RFR: 8013917 Kitchensink crashed with SIGSEGV in BaselineReporter::diff_callsites

[Daniel D. Daugherty]

 > http://cr.openjdk.java.net/~zgu/8013917/webrev.00/

Thumbs up! RT_Baseline is currently closed, please hold the
HSX-25 version of this fix until it reopens.

src/share/vm/services/memReporter.cpp
     No comments.

test/runtime/NMT/JcmdDiffCallsite.java
     Nice test.

You might want to add a comment to the bug report that
shows what a failing run of the test looks like. It'll
help the engineer that tries to verify the fix.

Dan


On 6/6/13 5:48 AM, Zhengyu Gu wrote:
> A simple bug fix on NMT reporting code. When memory allocated from a 
> callsite has been freed, the callsite should use baselined record to 
> report callsite and the memory should now be 0.
>
> External: Not available
> JBS: https://jbs.oracle.com/bugs/browse/JDK-8013917
> Webrev: http://cr.openjdk.java.net/~zgu/8013917/webrev.00/ 
> <http://cr.openjdk.java.net/%7Ezgu/8013917/webrev.00/>
>
> Tests:
>   - JPRT
>   - Kitchensink on reported platform Solaris-x64
>   - New jtreg testcase:  failed without patch, passed with the patch.
>
> Thanks,
>
> -Zhengyu
>