[11u] RFR: 8248411: [aarch64] Insufficient error handling when CodeBuffer is exhausted

[Langer, Christoph]

Hi Christoph,

that's cool, thanks for tackling this backport.

We're seeing the same issue with test compiler/codecache/TestStressCodeBuffers.java  and once in a while the same pattern also for compiler/startup/SmallCodeCacheStartup.java and we also think that this should be fixed by JDK- 8248411. I'll run it through our tests tonight and if nothing breaks and you don't hear from me by tomorrow you can consider this reviewed.

Best regards
Christoph

> -----Original Message-----
> From: jdk-updates-dev <jdk-updates-dev-retn at openjdk.java.net> On
> Behalf Of Christoph Göttschkes
> Sent: Mittwoch, 3. März 2021 11:27
> To: jdk-updates-dev at openjdk.java.net
> Subject: [11u] RFR: 8248411: [aarch64] Insufficient error handling when
> CodeBuffer is exhausted
> 
> Hi,
> 
> please review this backport of JDK-8248411 from 15u to 11u.
> 
> Bug: https://bugs.openjdk.java.net/browse/JDK-8248411
> Original commit: https://github.com/openjdk/jdk/commit/f279ddfa
> 15u backport webrev:
> https://cr.openjdk.java.net/~phedlin/tr8248411.15u/webrev/index.html
> 
> Webrev: https://cr.openjdk.java.net/~cgo/8248411/backport-11u.00/
> 
> we are seeing intermittent test failure in jdk11u on aarch64 for the
> hotspot tier1 jtreg tests
> "compiler/codecache/TestStressCodeBuffers.java". I backported
> JDK-8248411, which resolves the issues.
> I executed the test case several times. Without the patch, 10 of 25 runs
> failed. With the patch applied, 0 of 25 runs failed.
> 
> I chose the 15u backport as a basis for this 11u backport. Everything
> except the changes to "src/hotspot/share/opto/output.cpp" applies
> cleanly. The changes in output.cpp are only the addition of one assert
> call and some formatting.
> 
> Tests executed:
>   * Hotspot tier1 on linux aarch64
>   * Tier1 on linux x86_64
> 
> Thanks,
> Christoph