Oliver Graf <ograf(a)rhein-zeitung.de> writes in xemacs-beta(a)xemacs.org:
Hrvoje Niksic <hniksic(a)srce.hr> writes:
> SL Baur <steve(a)xemacs.org> writes:
>
> > Fatal error: assertion failed, file insdel.c, line 364, VALID_CHARPTR_P (ptr)
> >
> > Fatal error (6).
> > ...
>
> A gdb/dbx backtrace might be useful here.
Things are fucked up here. Both of the binaries I just crashed for
the backtraces below were compiled and linked with -g and full XEmacs
debugging/error-checking. Maybe I need a new gdb. I don't know. I
haven't upgraded from the gdb/egcs pair that Martin installed which
worked great in March, but don't seem to work now.
This backtrace is with the current 21.2 CVS revision
(almost-almost-21.2.15).
#1 0x3ff8019ca74 in tis_lock_global () from /usr/shlib/libc.so
#2 0x3ff8011451c in tis_raise () from /usr/shlib/libc.so
#3 0x3ff8015eaac in raise () from /usr/shlib/libc.so
#4 0x3ff801752a4 in abort () from /usr/shlib/libc.so
#5 0x1200ca0a8 in assert_failed (
file=0x7 <Error reading address 0x7: Invalid argument>, line=1084176448,
expr=0x0) at emacs.c:2663
#6 0x1201d66b0 in bytecount_to_charcount (
ptr=0x5d76 <Error reading address 0x5d76: Invalid argument>, len=6)
at insdel.c:364
#7 0x12006e900 in make_string (contents=0x0, length=536858720) at alloc.c:2111
#8 0x1202a3100 in x_cde_transfer_callback (widget=0x5d76, clientData=0x5d76,
callData=0x11fffd060) at frame-x.c:1268
#9 0x3ff8116e788 in _DtDndCallCallbackList () from /usr/dt/lib/libDtSvc.so
#10 0x3ff81170c9c in DtDndDropUnregister () from /usr/dt/lib/libDtSvc.so
warning: Hit beginning of text section without finding
warning: enclosing function for address 0x11fffd148
This warning occurs if you are debugging a function without any symbols
(for example, in a stripped executable). In that case, you may wish to
increase the size of the search with the `set heuristic-fence-post' command.
Otherwise, you told GDB there was a function where there isn't one, or
(more likely) you have encountered a bug in GDB.
This is with 21.1.2.
Fatal error: assertion failed, file insdel.c, line 364, VALID_CHARPTR_P (ptr)
Program received signal SIGABRT, Aborted.
0x3ff800ec508 in kill () from /usr/shlib/libc.so
(gdb) where
#0 0x3ff800ec508 in kill () from /usr/shlib/libc.so
#1 0x3ff8019ca74 in tis_lock_global () from /usr/shlib/libc.so
#2 0x3ff8011451c in tis_raise () from /usr/shlib/libc.so
#3 0x3ff8015eaac in raise () from /usr/shlib/libc.so
#4 0x3ff801752a4 in abort () from /usr/shlib/libc.so
#5 0x1200c95ec in assert_failed (file=0x3ff8116e788 "", line=1080495104,
expr=0x11fffdd68 "\003") at emacs.c:2609
warning: Hit heuristic-fence-post without finding
warning: enclosing function for address 0x1409691c8
This warning occurs if you are debugging a function without any symbols
(for example, in a stripped executable). In that case, you may wish to
increase the size of the search with the `set heuristic-fence-post' command.
Otherwise, you told GDB there was a function where there isn't one, or
(more likely) you have encountered a bug in GDB.
(gdb)
> I suspect the DND functions are _far_ from Mule-safety.
I suspect you are correct. Um, for 21.1 you are permitted to put
mojibake into the buffer, but you are not permitted to crash. We need
either a fix or to disable it as a default. 21.2 is allowed to crash
until we get a real fix.
Yap, correct. They are far away from being MULE save.
Do I need a special CDE version to test japanese script in dtterm? I
will look at this problem tommorrow...
I have no idea. If you need it, I suspect I can get you login access
if I ask nicely enough.
--
I protest the NATO war in Yugoslavia.