Wow, that hidden unboxing and primitive widening in the conditional operator is sinister. I had to write a small test program and use javap on it to see it with my own eyes. Surely this isn’t a javac bug?
Attila. > On 2017. Sep 13., at 11:18, Hannes Wallnöfer <hannes.wallnoe...@oracle.com> > wrote: > > +1 > > Hannes > >> Am 13.09.2017 um 07:21 schrieb Sundararajan Athijegannathan >> <sundararajan.athijegannat...@oracle.com>: >> >> +1 >> >> PS. You may want to add explanation to the bug (if not done already). We may >> have to revisit conditional expressions elsewhere as well.... >> >> -Sundar >> >> On 13/09/17, 10:37 AM, Priya Lakshmi Muthuswamy wrote: >>> Updated patch : http://cr.openjdk.java.net/~pmuthuswamy/8187362/webrev.01/ >>> corrected the alignment. >>> >>> Thanks, >>> Priya >>> >>> On 9/13/2017 9:58 AM, Priya Lakshmi Muthuswamy wrote: >>>> Hi, >>>> >>>> Please review JDK-8187362 : Nashorn unsigned right shift operator >>>> unexpectedly returns floating-point >>>> JBS: https://bugs.openjdk.java.net/browse/JDK-8187362 >>>> webrev: http://cr.openjdk.java.net/~pmuthuswamy/8187362/webrev.00/ >>>> >>>> Thanks, >>>> Priya >>> >
