RFR [XS]: 8244500: jtreg test error in test/hotspot/jtreg/containers/docker/TestMemoryAwareness.java

[David Holmes]

On 14/05/2020 12:28 am, Baesken, Matthias wrote:
> Hello, here is a new webrev removing the comment :
> 
> http://cr.openjdk.java.net/~mbaesken/webrevs/8244500.1/
> 
> It removes the comment , and it also  handles  0  for getFreeSwapSpaceSize
> (  this can occur  after   JDK-8242480 : Negative value may be returned by getFreeSwapSpaceSize() in the docker   was pushed  recently ).

Sorry Matthias but I still see this partly as a bug in the underlying 
container support code. If you get the warning then what does 
getTotalSwapSpaceSize() return? Surely it must be specified to return 0 
or some such sentinel in such a case?

Other comments:

!         try {
! 
out.shouldMatch("OperatingSystemMXBean\\.getFreeSwapSpaceSize: 
[1-9][0-9]+");
!         } catch(RuntimeException ex) { // allow 0, see 8242480
! 
out.shouldContain("OperatingSystemMXBean.getFreeSwapSpaceSize: 0");
!         }

can't you just change the original regex to match >= 0?

Style nit: avoid historical comments like "see 8242480" and "since 
8231111 ...".

Thanks,
David

> Best regards, Matthias
> 
> -----Original Message-----
> From: Baesken, Matthias
> Sent: Mittwoch, 13. Mai 2020 10:11
> To: 'David Holmes' <david.holmes at oracle.com>; 'hotspot-dev at openjdk.java.net' <hotspot-dev at openjdk.java.net>
> Subject: RE: RFR [XS]: 8244500: jtreg test error in test/hotspot/jtreg/containers/docker/TestMemoryAwareness.java
> 
> 
>> But according to the comment immediately before your changes:
>>
>> // in case of warnings like : "Your kernel does not support swap limit
>> capabilities or the cgroup is not mounted. Memory limited without swap."
>> // the getTotalSwapSpaceSize does not return the expected result, but 0
>>
>> we should be returning zero. So this seems to be a bug in the
>> implementation not in the test.
> 
>   
> Hi David , I think  in case of  "Your kernel does not support swap limit  capabilities   ... "    we just do not get the values
>   that are normally expected.
> Previously we indeed got "0" .
> Probably the comment should be deleted .
> 
> Best regards, Matthias
>