RFR(T) : 8244142 : some hotspot/runtime tests don't check exit code of forked JVM :
igor.ignatyev at oracle.com
Wed Apr 29 21:58:47 UTC 2020
> On Apr 29, 2020, at 1:22 PM, gerard ziemski <gerard.ziemski at oracle.com> wrote:
> hi Igor,
> On 4/29/20 2:05 PM, Igor Ignatyev wrote:
>>> 7 lines changed: 3 ins; 0 del; 4 mod;
>> Hi all,
>> could you please review this trivial patch for 3 runtime tests?
>> from JBS:
>>> some of the hotspot/runtime tests which fork JVMs just check the content of their output/error and don't check exit code, b/c of this some defects might get unnoticed.
>> testing: the changed tests
>> webrev: http://cr.openjdk.java.net/~iignatyev/8244142/webrev.00/
>> JBS: https://bugs.openjdk.java.net/browse/JDK-8244142
> Those checks for exit!=0 seem reasonable.
> Did we actually hit such cases in real life, or are we just being super cautious?
well, we don't know ;) , and yes it's just to be on a safe side.
More information about the hotspot-runtime-dev