On Mar 24, 7:26 pm, Boris Zbarsky <bzbar...@mit.edu> wrote: > > In gdb, "set print object on" and then see what it claims the mRawPtr to be. >
It claims it's a nsISupports pointer. > Does the crash happen at this call, or somewhere inside CreateElement? All I can see in the backtrace is: #0 0x00007f85cf0d0015 in raise () from /lib/libc.so.6 #1 0x00007f85cf0d1b83 in abort () from /lib/libc.so.6 #2 0x00007f85cf1110c8 in ?? () from /lib/libc.so.6 #3 0x00007f85cf116a58 in ?? () from /lib/libc.so.6 #4 0x00007f85cf1190a6 in free () from /lib/libc.so.6 #5 0x00007f85ca0b60e2 in ?? () from /usr/lib/xulrunner-1.9.0.7/ libxul.so #6 0x00000000004021fe in WebListener::DocumentLoaded (this=0x24214e0) at src/Test.cpp:46 #7 0x00007f85d21dc9f0 in WebBrowserChrome::OnStateChange (this=0x2301d80, aStateFlags=131088) at ../../common/ WebBrowserChrome.cpp:191 ... > > Right. For what it's worth, looking at where DocumentLoaded is called, > it really should be equivalent to the JS onload event. > Right. This is what's confusing the #&%* out of me: things are clearly hooked up correctly enough for the page to display and this callback to fire at the right place, but the DOM information just doesn't seem to be there. Out of curiosity, does my little sample app work correctly for anyone else? I mean, it really *should* be the most minimal GTK example that hits the DOM, right? _______________________________________________ dev-embedding mailing list dev-embedding@lists.mozilla.org https://lists.mozilla.org/listinfo/dev-embedding
