I get the same bt, I installed the -dbg version of glibc and debuilded
a part of the source (to get the patches applied):
(gdb) bt
#0 0x1b94e58f in __gconv_release_step (step=0x1bae2ccc) at gconv_db.c:198
#1 0x1b94f14d in __gconv_close_transform (steps=0x1bae2c90, nsteps=2)
at gconv_db.c:751
#2 0x1b94e256 in __gconv_close (cd=0x1bae3868) at gconv_close.c:64
#3 0x1b95c54d in _nl_free_domain_conv (domain=0x1baba698) at loadmsgcat.c:873
#4 0x1b95d0b4 in _nl_unload_domain (domain=0x1baba698) at loadmsgcat.c:1289
#5 0x1ba42afd in free_mem () at finddomain.c:179
#6 0x1ba42c45 in *__GI___libc_freeres () at set-freeres.c:49
#7 0x1b8fec51 in _vgw__freeres () at vg_intercept.c:117
#8 0x1b962b18 in *__GI_exit (status=0) at exit.c:82
I think the problem is in iconv/gconv_close.c, in __gconv_close() at
the very end of the file:
while (!((drunp++)->__flags & __GCONV_IS_LAST));
/* Free the data allocated for the descriptor. */
free (cd);
/* Close the participating modules. */
return __gconv_close_transform (srunp, nsteps);
I think "srunp" uses a reference in "cd".
I already tried in the past to build glibc on my system, and this was
too long a task for my laptop, could someone try swapping the free
after the __gconv_close_transform() or hand me a build if he can't
reproduce the bug?
Thanks,
--
Loïc Minier <email address hidden>
"Neutral President: I have no strong feelings one way or the other."
Hi,
This is a followup for Debian bug <http:// bugs.debian. org/279722>.
wim delvaux <email address hidden> - Thu, Nov 04, 2004:
> Valgrind shows the following backtrace ... release_ step (gconv_db.c:198) close_transform (gconv_db.c:751)
> ==7105== Invalid read of size 4
> ==7105== at 0x1C22857E: __gconv_
> ==7105== by 0x1C22914C: __gconv_
I can get that one fairly easily with:
% valgrind --db-attach=yes ls -l
(% locale LANG=fr_FR@euro LC_CTYPE=fr_FR@euro LC_NUMERIC= fr_FR@euro fr_FR@euro LC_MONETARY= fr_FR@euro fr_FR@euro LC_PAPER=fr_FR@euro LC_NAME=fr_FR@euro fr_FR@euro LC_TELEPHONE= fr_FR@euro LC_MEASUREMENT= fr_FR@euro ION=fr_ FR@euro LC_ALL=)
LC_TIME=fr_FR@euro LC_COLLATE=
LC_MESSAGES=
LC_ADDRESS=
LC_IDENTIFICAT
I get the same bt, I installed the -dbg version of glibc and debuilded
a part of the source (to get the patches applied):
(gdb) bt release_ step (step=0x1bae2ccc) at gconv_db.c:198 close_transform (steps=0x1bae2c90, nsteps=2) domain_ conv (domain=0x1baba698) at loadmsgcat.c:873 _libc_freeres () at set-freeres.c:49
#0 0x1b94e58f in __gconv_
#1 0x1b94f14d in __gconv_
at gconv_db.c:751
#2 0x1b94e256 in __gconv_close (cd=0x1bae3868) at gconv_close.c:64
#3 0x1b95c54d in _nl_free_
#4 0x1b95d0b4 in _nl_unload_domain (domain=0x1baba698) at loadmsgcat.c:1289
#5 0x1ba42afd in free_mem () at finddomain.c:179
#6 0x1ba42c45 in *__GI__
#7 0x1b8fec51 in _vgw__freeres () at vg_intercept.c:117
#8 0x1b962b18 in *__GI_exit (status=0) at exit.c:82
I think the problem is in iconv/gconv_ close.c, in __gconv_close() at +)->__flags & __GCONV_IS_LAST));
the very end of the file:
while (!((drunp+
/* Free the data allocated for the descriptor. */
free (cd);
/* Close the participating modules. */ close_transform (srunp, nsteps);
return __gconv_
I think "srunp" uses a reference in "cd".
I already tried in the past to build glibc on my system, and this was close_transform () or hand me a build if he can't
too long a task for my laptop, could someone try swapping the free
after the __gconv_
reproduce the bug?
Thanks,
--
Loïc Minier <email address hidden>
"Neutral President: I have no strong feelings one way or the other."