On Wed, Dec 21, 2011 at 9:25 AM, Hyrum K Wright <hyrum.wri...@wandisco.com> wrote: > On Tue, Dec 20, 2011 at 1:50 PM, Mark Phippard <markp...@gmail.com> wrote: >> On Tue, Dec 20, 2011 at 2:42 PM, Philip Martin >> <philip.mar...@wandisco.com> wrote: >> >>>> env->ReleaseLongArrayElements(jrevs, jrevArray, 0); >>> >>> Do we need to consider ReleaseLongArrayElements raising an exception? >> >> This is what we see on the Eclipse side: >> >> !ENTRY org.eclipse.core.jobs 4 2 2011-12-14 10:51:00.312 >> !MESSAGE An internal error occurred during: "SVN Update". >> !STACK 0 >> java.lang.ArrayIndexOutOfBoundsException: 68 >> at org.apache.subversion.javahl.SVNClient.update(Native Method) >> at >> org.tigris.subversion.svnclientadapter.javahl.AbstractJhlClientAdapter.update(AbstractJhlClientAdapter.java:1113) >> >> I will keep looking, but the error seems like it is happening in the >> native code. >> >> I will also point out that this has NOT come up a lot and I do not >> know how to reproduce it. > > For what it's worth, I *think* this means that the exception is > originating in the bindings, and not in the underlying libraries. If > the error were being thrown by the underlying libraries, it would > likely be an svn_error_t, which then gets added to the Java exception > stack frame. Since that's not happening, and since the exception > isn't a SubversionException, that would suggest this is a bug in the > bindings themselves. > > But, you all probably already knew that. ;)
I figured out the problem and am committing a fix. Once we had a recipe it was easier to see it. There is a new action added on the end of svn_wc_notify_action_t - svn_wc_notify_skip_conflicted. This is the 68th item in enum and we did not have a value for it in the Java file. -- Thanks Mark Phippard http://markphip.blogspot.com/
