RFR(XXS): 8028052: compiler/startup/SmallCodeCacheStartup.java fails there is no 'no space to run compiler' in the output

[Vladimir Kozlov]

Albert,

Since you don't need to analyze the output the test should be converted to normal jtreg test without forking process:

@run main/othervm -XX:ReservedCodeCacheSize=3m -XX:CICompilerCount=64 -version

Thanks,
Vladimir

On 12/19/13 3:52 AM, Albert Noll wrote:
> Hi,
>
> could I get reviews for this small patch?
>
> bug: https://bugs.openjdk.java.net/browse/JDK-8028052
> webrev: http://cr.openjdk.java.net/~anoll/8028052/webrev.00/
>
> Problem: The test assumes the existence of a C1 compiler. Not all VM configurations
> have a C1 compiler. E.g., ARM server VM has only C2 - that is why the test fails.
>
> Solution: Weaken test. Instead of checking for the expected warning that is printed
> by the VM ('no space to run compiler'), the test passes if the VM does not crash. I think
> weakening the test is fine since it still tests the cause of the original crash (8023014).
>
> Best,
> Albert
>
>