Yes, Rocky, I guess you better change these exits to RuntimeExceptions
and send the webrev to me. I'll post it as
http://cr.openjdk.java.net/~yan/8137113/webrev.01, and you'll send
a message to this list. That is the procedure here.
Thank you!
-yan
On 10/30/2015 01:32 AM, Rocky Sloan wrote:
Hi Sergey,
Thanks for the feedback.
I changed it to catch the filesystem error(s) and throw a
RuntimeException() and tested it.
The new version of the webrev is attached.
I used System.exit() because it was used in other files included in this
test.
Example in other files:
Task.print(working + " out of " + alive + " threads are working");
if ((working == 0) && (++alarm == 10)) {
Task.print("DEADLOCK DETECTED");
System.exit(100);
}
Should I change the other files to throw a RuntimeException() in a
similar fashion?
Regards,
Rocky
On 10/27/2015 7:30 AM, Sergey Bylokhov wrote:
Hi, Rocky.
Please do not use System.exit() in the tests, just throw runtime
exception if some unexpected errors occurred.
On 23.10.15 23:05, Rocky Sloan wrote:
Hello,
Please review the test fix for JDK9:
bug: https://bugs.openjdk.java.net/browse/JDK-8137113
webrev: http://cr.openjdk.java.net/~yan/8137113/webrev.00
Updated to use the new run-time modular filesystem in JDK9. Removed
JDK8 compatible code which accesses rt.jar.
The test originally used the set of classes in rt.jar. These classes
are now spread throughout the new filesystem in JDK9, so the updated
code will access a representation of these classes from their new
locations.
Thanks,
Rocky
