bogoutil segfaults

[David Relson]

At 11:19 AM 7/23/03, Matthias Andree wrote:
>Clint Adams <schizo at debian.org> writes:
>
> >> Does SPARC support unaligned access?
> >
> > Not as far as I know.
>
>So that's not it. SunPro compiled code currently fails when optimized
>with -fast, in memcpy. Very strange, looks like an optimizer bug.
>
> > <<< FULL BACKTRACE <<<
> > #0  0x0001603c in db_foreach (vhandle=0x29098, hook=0x28096
> > <__JCR_LIST__+6>, userdata=0x0) at
> > ../../bogofilter-0.14.0/src/datastore_db.c:528
>
>The arguments look suspiciously close together.

Meaning vhandle=0x29098 and hook=0x28096??  FWIW, vhandle is xmalloc'd and 
hook is &db_dump_hook.  On intel at least, they're in data and code 
segments.  Don't know about hppa.


> >         i = 0
> >         ret = 0
> > #1  0x00011730 in db_oper (path=0xc33c4022 <Address 0xc33c4022 out of 
> bounds>, open_mode=DB_READ, funct=0x28096 <__JCR_LIST__+6>, userdata=0x0) 
> at ../../bogofilter-0.14.0/src/bogoutil.c:163
> >         dbh = (void *) 0x29098
>
>Gna, I've had these intermittently on i386 as well, but I haven't been
>able to track them down. I cannot reproduce these at the moment :-(

What, exactly, have you had "intermittently on i386"???