[DTrace-devel] [PATCH REVIEW 4/6] tests: reduce timeout of very slow test

[Kris Van Hees]

On Thu, Sep 09, 2021 at 02:46:05PM -0400, Eugene Loh wrote:
> Reviewed-by: Eugene Loh <eugene.loh at oracle.com>

I disagree - not an acceptable solution.

Instead, please mark this test as @@skip.  Since we do not support buffering
policies other than ring (and possibly never will in this version), there is
no point in running the test.  But I rather not change the test in case we do
actually implement the fill buffering policy in the future.

> but how about adding a note to the xfail message explaining why we 
> expect the failure.  I assume it's because we don't have "bufpolicy=fill".
> 
> On 9/9/21 7:13 AM, Nick Alcock wrote:
> > This test sets the timeout to a high value to fill a buffer that no
> > longer applies to v2, and is marked @@xfail: dtv2.  Drop the timeout so
> > we don't have to wait ten minutes on every test run for something we
> > know will fail.
> >
> > We can raise it again later if it seems necessary.
> >
> > Signed-off-by: Nick Alcock <nick.alcock at oracle.com>
> > ---
> >   test/unittest/printa/tst.many.d | 2 +-
> >   1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > Not really a fix, but saving nine minutes of test run time is a worthwhile
> > thing anyway!
> >
> > diff --git a/test/unittest/printa/tst.many.d b/test/unittest/printa/tst.many.d
> > index 96b5f17dad3a..6b7462dfffa8 100644
> > --- a/test/unittest/printa/tst.many.d
> > +++ b/test/unittest/printa/tst.many.d
> > @@ -10,7 +10,7 @@
> >    * set the timeout to a high value to, in effect, disable it.
> >    */
> >   /* @@xfail: dtv2 */
> > -/* @@timeout: 600 */
> > +/* @@timeout: 60 */
> >   
> >   #pragma D option bufsize=32m
> >   #pragma D option bufpolicy=fill
> 
> _______________________________________________
> DTrace-devel mailing list
> DTrace-devel at oss.oracle.com
> https://oss.oracle.com/mailman/listinfo/dtrace-devel