Its better if we can make this test run only in Linux as GTK is specific to it. Other changes are fine.
Thanks, Jay > On 20-Dec-2019, at 10:31 PM, Phil Race <philip.r...@oracle.com> wrote: > > Looks fine. Please use the amended and corrected bug synopsis in the commit > message > > -phil. > > On 12/19/19 11:36 PM, Pankaj Bansal wrote: >> Hi All, >> >> Please review the following fix for jdk15. >> >> Bug: >> https://bugs.openjdk.java.net/browse/JDK-8222759 >> <https://bugs.openjdk.java.net/browse/JDK-8222759> >> >> webrev: >> http://cr.openjdk.java.net/~pbansal/8222759/webrev00/ >> <http://cr.openjdk.java.net/~pbansal/8222759/webrev00/> >> >> Issue: >> There are some unnecessary casts of javax.swing.plaf.synth.SynthStyleFactory >> to com.sun.java.swing.plaf.gtk.GTKStyleFactory in GTKLookAndFeel class. >> There is only one function “getStyle” being called after this cast, but that >> is declared in SynthStyleFactory as abstract and defined in GTKStyleFactory. >> So this function can be called without this cast too. >> Because of this cast, if someone is trying to use a custom SynthStyleFactory >> with GTKLookAndFeel, there is a ClassCastException. >> >> Fix: >> Removed the cast as it is not required here. >> >> Testing: >> I have run SwingSet2 with GTKLookAndFeel after this change. All works fine. >> Added testcase fails without fix and passes after the fix. >> >> Regards, >> Pankaj Bansal >> > >
