Re: RFR (XS): 8151064: com/sun/jdi/RedefineAddPrivateMethod.sh fails intermittently
Andreas, Looks good for me. -Dmitry On 2016-03-03 17:05, Andreas Eriksson wrote: > Hi, > > Can I please have a review of this fix for > 8151064: com/sun/jdi/RedefineAddPrivateMethod.sh fails intermittently > https://bugs.openjdk.java.net/browse/JDK-8151064 > > Webrev: http://cr.openjdk.java.net/~aeriksso/8151064/webrev/ > > Still not sure why it only fails sometimes, but after this change the > test has not failed once after a couple of hours of testing. Before the > change it would fail after ~5 minutes of running it in a loop. > > Thanks, > Andreas -- Dmitry Samersoff Oracle Java development team, Saint Petersburg, Russia * I would love to change the world, but they won't give me the sources.
Re: RFR (XS): 8151064: com/sun/jdi/RedefineAddPrivateMethod.sh fails intermittently
Thanks. - Andreas On 2016-03-03 15:35, Dmitry Samersoff wrote: Andreas, Looks good for me. -Dmitry On 2016-03-03 17:05, Andreas Eriksson wrote: Hi, Can I please have a review of this fix for 8151064: com/sun/jdi/RedefineAddPrivateMethod.sh fails intermittently https://bugs.openjdk.java.net/browse/JDK-8151064 Webrev: http://cr.openjdk.java.net/~aeriksso/8151064/webrev/ Still not sure why it only fails sometimes, but after this change the test has not failed once after a couple of hours of testing. Before the change it would fail after ~5 minutes of running it in a loop. Thanks, Andreas
Re: RFR (XS): 8151064: com/sun/jdi/RedefineAddPrivateMethod.sh fails intermittently
Hi Andreas, Good++ Thanks, Serguei On 3/3/16 06:35, Dmitry Samersoff wrote: Andreas, Looks good for me. -Dmitry On 2016-03-03 17:05, Andreas Eriksson wrote: Hi, Can I please have a review of this fix for 8151064: com/sun/jdi/RedefineAddPrivateMethod.sh fails intermittently https://bugs.openjdk.java.net/browse/JDK-8151064 Webrev: http://cr.openjdk.java.net/~aeriksso/8151064/webrev/ Still not sure why it only fails sometimes, but after this change the test has not failed once after a couple of hours of testing. Before the change it would fail after ~5 minutes of running it in a loop. Thanks, Andreas
Re: RFR (XS): 8151064: com/sun/jdi/RedefineAddPrivateMethod.sh fails intermittently
Thanks Serguei. I'll go ahead and push this now, since I believe this change is small enough. - Andreas On 2016-03-03 17:54, serguei.spit...@oracle.com wrote: Hi Andreas, Good++ Thanks, Serguei On 3/3/16 06:35, Dmitry Samersoff wrote: Andreas, Looks good for me. -Dmitry On 2016-03-03 17:05, Andreas Eriksson wrote: Hi, Can I please have a review of this fix for 8151064: com/sun/jdi/RedefineAddPrivateMethod.sh fails intermittently https://bugs.openjdk.java.net/browse/JDK-8151064 Webrev: http://cr.openjdk.java.net/~aeriksso/8151064/webrev/ Still not sure why it only fails sometimes, but after this change the test has not failed once after a couple of hours of testing. Before the change it would fail after ~5 minutes of running it in a loop. Thanks, Andreas
Re: RFR (XS): 8151064: com/sun/jdi/RedefineAddPrivateMethod.sh fails intermittently
> Webrev: http://cr.openjdk.java.net/~aeriksso/8151064/webrev/ test/com/sun/jdi/RedefineAddPrivateMethod.sh L38: System.out.println("@1 breakpoint"); L39: System.out.println("@2 breakpoint"); With these tests, this is usually done like this: System.out.println("stop here for breakpoint 1"); // @1 breakpoint System.out.println("stop here for breakpoint 2"); // @2 breakpoint I _think_ the scaffold stuff can handle the way you did it on all platforms, but I'm not sure. Sorry, my memory is a bit rusty on this stuff. Thumbs up on the original version or you can change it to the above. Your choice. Dan P.S. It occurs to me that in the original code the main() method is empty, i.e., just comment lines. I have a vague memory about empty methods being treated differently in HotSpot, but I don't remember the exact details... On 3/3/16 10:05 AM, Andreas Eriksson wrote: Thanks Serguei. I'll go ahead and push this now, since I believe this change is small enough. - Andreas On 2016-03-03 17:54, serguei.spit...@oracle.com wrote: Hi Andreas, Good++ Thanks, Serguei On 3/3/16 06:35, Dmitry Samersoff wrote: Andreas, Looks good for me. -Dmitry On 2016-03-03 17:05, Andreas Eriksson wrote: Hi, Can I please have a review of this fix for 8151064: com/sun/jdi/RedefineAddPrivateMethod.sh fails intermittently https://bugs.openjdk.java.net/browse/JDK-8151064 Webrev: http://cr.openjdk.java.net/~aeriksso/8151064/webrev/ Still not sure why it only fails sometimes, but after this change the test has not failed once after a couple of hours of testing. Before the change it would fail after ~5 minutes of running it in a loop. Thanks, Andreas
Re: RFR (XS): 8151064: com/sun/jdi/RedefineAddPrivateMethod.sh fails intermittently
On 2016-03-03 18:29, Daniel D. Daugherty wrote: > Webrev: http://cr.openjdk.java.net/~aeriksso/8151064/webrev/ test/com/sun/jdi/RedefineAddPrivateMethod.sh L38: System.out.println("@1 breakpoint"); L39: System.out.println("@2 breakpoint"); With these tests, this is usually done like this: System.out.println("stop here for breakpoint 1"); // @1 breakpoint System.out.println("stop here for breakpoint 2"); // @2 breakpoint I _think_ the scaffold stuff can handle the way you did it on all platforms, but I'm not sure. Sorry, my memory is a bit rusty on this stuff. Thumbs up on the original version or you can change it to the above. Your choice. Sorry, saw this after I had already pushed with the original version. But other tests did similar things: RedefineException.sh - line 83: System.out.println("a3: @1 breakpoint here a3"); And it ran fine through RBT, so hopefully it should be no problem. Dan P.S. It occurs to me that in the original code the main() method is empty, i.e., just comment lines. I have a vague memory about empty methods being treated differently in HotSpot, but I don't remember the exact details... What was weird is that it only failed sometimes. I'll take a look tomorrow to see if I can find out why, but I probably wont spend too much time on it. Thanks, Andreas On 3/3/16 10:05 AM, Andreas Eriksson wrote: Thanks Serguei. I'll go ahead and push this now, since I believe this change is small enough. - Andreas On 2016-03-03 17:54, serguei.spit...@oracle.com wrote: Hi Andreas, Good++ Thanks, Serguei On 3/3/16 06:35, Dmitry Samersoff wrote: Andreas, Looks good for me. -Dmitry On 2016-03-03 17:05, Andreas Eriksson wrote: Hi, Can I please have a review of this fix for 8151064: com/sun/jdi/RedefineAddPrivateMethod.sh fails intermittently https://bugs.openjdk.java.net/browse/JDK-8151064 Webrev: http://cr.openjdk.java.net/~aeriksso/8151064/webrev/ Still not sure why it only fails sometimes, but after this change the test has not failed once after a couple of hours of testing. Before the change it would fail after ~5 minutes of running it in a loop. Thanks, Andreas
Re: RFR (XS): 8151064: com/sun/jdi/RedefineAddPrivateMethod.sh fails intermittently
On 3/3/16 12:56 PM, Andreas Eriksson wrote: On 2016-03-03 18:29, Daniel D. Daugherty wrote: > Webrev: http://cr.openjdk.java.net/~aeriksso/8151064/webrev/ test/com/sun/jdi/RedefineAddPrivateMethod.sh L38: System.out.println("@1 breakpoint"); L39: System.out.println("@2 breakpoint"); With these tests, this is usually done like this: System.out.println("stop here for breakpoint 1"); // @1 breakpoint System.out.println("stop here for breakpoint 2"); // @2 breakpoint I _think_ the scaffold stuff can handle the way you did it on all platforms, but I'm not sure. Sorry, my memory is a bit rusty on this stuff. Thumbs up on the original version or you can change it to the above. Your choice. Sorry, saw this after I had already pushed with the original version. But other tests did similar things: RedefineException.sh - line 83: System.out.println("a3: @1 breakpoint here a3"); And it ran fine through RBT, so hopefully it should be no problem. Sounds good to me... Dan P.S. It occurs to me that in the original code the main() method is empty, i.e., just comment lines. I have a vague memory about empty methods being treated differently in HotSpot, but I don't remember the exact details... What was weird is that it only failed sometimes. I'll take a look tomorrow to see if I can find out why, but I probably wont spend too much time on it. In the original .jtr file: Java HotSpot(TM) 64-Bit Server VM (fastdebug build 9-internal+0-2016-03-02-004742.mgronlun.upstream, mixed mode) Since we're in -Xmixed mode, the compiler could kick in, but on such a short-lived test? I'm dubious... Good hunting tomorrow! Dan Thanks, Andreas On 3/3/16 10:05 AM, Andreas Eriksson wrote: Thanks Serguei. I'll go ahead and push this now, since I believe this change is small enough. - Andreas On 2016-03-03 17:54, serguei.spit...@oracle.com wrote: Hi Andreas, Good++ Thanks, Serguei On 3/3/16 06:35, Dmitry Samersoff wrote: Andreas, Looks good for me. -Dmitry On 2016-03-03 17:05, Andreas Eriksson wrote: Hi, Can I please have a review of this fix for 8151064: com/sun/jdi/RedefineAddPrivateMethod.sh fails intermittently https://bugs.openjdk.java.net/browse/JDK-8151064 Webrev: http://cr.openjdk.java.net/~aeriksso/8151064/webrev/ Still not sure why it only fails sometimes, but after this change the test has not failed once after a couple of hours of testing. Before the change it would fail after ~5 minutes of running it in a loop. Thanks, Andreas