Hi Christopher, > > The compiler must have gotten smarter than the last time I checked. >
I'm sorry: I compiled that code with the Eclipse compiler. When compiling it with the Oracle Java compiler (javac), it prints out the line number of the "System.out.println". Maybe if Tomcat was compiled with the standard Java compiler, it could be that on the NPE at StandardWrapperValve.invoke(), also "context" was null, like in the NPE at StandardContextValve.invoke(). > > You could modify the source and add another dump of the original > exception (say, to stdout). That way, if it does happen again, you'll > at > least see the root cause. > OK, I may try that. But I have never seen that NPE in StandardWrapperValve before I started this thread, so I don't think it is worth it (Tomcat 7 releases are every month, and I don't want to modify the code in each version). :) > > No, it sounds like that might be a rarely-seen bug in Tomcat since it > happens only during context reload while a long-running request is in > progress. If you can build a simple test case, attach it to a new BZ > entry and we'll see about getting it fixed. > OK, Thanks. Regards, Konstantin Preißer --------------------------------------------------------------------- To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org For additional commands, e-mail: users-h...@tomcat.apache.org
