PING^3: RFD: C2 bug: 8157306 random infrequent null pointer exceptions in javac

Vladimir Kozlov vladimir.kozlov at oracle.com
Mon Jul 25 21:20:35 UTC 2016


Hi Andrew,

I finally got time to verify that changes in lcm without the assert do not introduce regression - tests failed without lcm changes. I am pushing only lcm.cpp change as you suggested. And I filed a 
separate bug for failing cases with the assert.

https://bugs.openjdk.java.net/browse/JDK-8162496

Regards,
Vladimir

On 7/22/16 10:46 AM, Andrew Haley wrote:
> [re-sending]
>
> http://cr.openjdk.java.net/~aph/8157306.1/
>
> Hi Vladimir,
>
> The current status:
>
> This patch fixes a serious bug in AArch64 and perhaps other
> targets.  It's blocked because you discovered that an assertion
> (added by this patch) was triggered by some tests.  I do not believe
> that anything this patch does could cause whatever problem the
> assertion reveals.
>
> I have not been able to reproduce the problem that you saw.  Therefore
> there is nothing that I can do to fix it.  May I simply take the
> assertion out and resubmit the patch?
>
> This bug in C2 is real, and serious.  This patch is a blocker for
> AArch64.
>
> Thanks,
>
> Andrew.
>


More information about the hotspot-dev mailing list