On 21 mars 2016, at 09:47, Dmitry Samersoff <dmitry.samers...@oracle.com> wrote: > > Staffan, > > Looks good for me.
Thanks! > > You may also consider to add -XDstringConcat=inline to javac options in > ShellScaffold I have considered it, but I’m reluctant to do it since it would mean that we are testing code paths that are different from what our users will see. /Staffan > > -Dmitry > > > On 2016-03-21 11:22, Staffan Larsen wrote: >> Hi, >> >> See the JIRA issue or my recent review request of JDK-7153107 for a >> discussion about the cause of this timeout. >> >> In this change I am changing the ShellScaffold to take the jtreg >> timeoutfactor into account. This allow more time for these tests on >> slower platforms. >> >> The timeoutfactor is a floating point number, but I could not find a way >> to do floating point calculations in bash without adding >> more dependencies. Instead I opted to truncate the resulting value into >> an integer. At most, this decreases the timeout by almost a second and >> should not matter. >> >> webrev: >> http://cr.openjdk.java.net/~sla/8151887/webrev.00/test/com/sun/jdi/ShellScaffold.sh.sdiff.html >> bug: https://bugs.openjdk.java.net/browse/JDK-8151887 >> >> Thanks, >> /Staffan >> > > > -- > Dmitry Samersoff > Oracle Java development team, Saint Petersburg, Russia > * I would love to change the world, but they won't give me the sources.
