Re: Creating and using GCJ ARM cross compiler

[Andrew Haley <aph at redhat dot com>]

Matthijs van de Water wrote:
> On Tue, Apr 22, 2008 at 2:02 PM, Andrew Haley <aph@redhat.com> wrote:
>>  That's almost certainly exception handling.  Put your bp on _Jv_Throw.
> Looks like you're right:
> 
> _Jv_Throw (value=0x4212efa0) at /data/gcc-4_2-branch/libjava/exception.cc:99
> 99      _Jv_Throw (jthrowable value)
> (gdb) n
> 104       if (value == NULL)
> (gdb) inspect value
> $4 = (jthrowable) 0x4212efa0
> (gdb) next
> 102         = static_cast<java_exception_header *>(_Jv_AllocRawObj
> (sizeof (*xh)));
> (gdb)
> 99      _Jv_Throw (jthrowable value)
> (gdb)
> 104       if (value == NULL)
> (gdb)
> 109               sizeof xh->unwindHeader.exception_class);
> (gdb) n
> 106       xh->value = value;
> (gdb) n
> 109               sizeof xh->unwindHeader.exception_class);
> (gdb) n
> 110       xh->unwindHeader.exception_cleanup = NULL;
> (gdb) n
> 118       code = _Unwind_RaiseException (&xh->unwindHeader);
> (gdb) s
> 110       xh->unwindHeader.exception_cleanup = NULL;
> (gdb) s
> 118       code = _Unwind_RaiseException (&xh->unwindHeader);
> (gdb) s
> terminate called without an active exception
> 
> Program received signal SIGABRT, Aborted.
> 0x41fb38c4 in raise () from /lib/libc.so.6
> (gdb)

OK, good.  If this were me I'd install gcc debuginfo and step through
_Unwind_RaiseException, but we have, I think, established the most likely
culprit.

I can't see anything very much wrong with the merged Debian source,
though.

Andrew.