Hi Marcus!

Sorry for being off-topic here, but I'm experiencing the same problem (luckily I do not need to profile a script which needs the bridge) - anyways - you mentioned in http://xdebug.org/archives/xdebug-general/1229.html that the usage of java.so is deprecated and that the development on the bridge is stalled. Since we've just started with a new project which uses the bridge, I'm a bit concerned about this comment actually - can you pinpoint me to a discussion or other ressource where you got these information?

Thanks in advance,
Thomas.

I can't fix this as I can't find the source code for the java bridge, and thus not reproduce this problem.

Sorry for answering this late, Thomas.
Just to document this in case someone stumbles over this thread: the php-java-bridge is still under very active development as of today as one can easily see at the bridge's site at http://php-java-bridge.sourceforge.net/. I was just refering to the bridge's java.so which by now is no longer available.

@Derick: I remember at the time we exchanged emails you said you wanted to fix this bug independently of the bridge, so here are all the source files i could dig up at a sourceforge mirror: http://mirror.optus.net/sourceforge/p/ph/php-java-bridge/. It was version 4.3.2 ff. that caused the segfault.

Please note that imho there is no need to reopen this thread. The bridge now handles all php-related stuff via a native Java.inc script which can easily be cached by opcode caches.

I just had a chat with some of the other people that know the Zend Engine. Running zend_execute() from RINIT is never going to work reliably. As this is fixed in the java bridge, and you shouldn't do this anyway, I'm going to close this issue again.