Arch Linux

Please read this before reporting a bug:
https://wiki.archlinux.org/title/Bug_reporting_guidelines

Do NOT report bugs when a package is just outdated, or it is in the AUR. Use the 'flag out of date' link on the package page, or the Mailing List.

REPEAT: Do NOT report bugs for outdated packages!
Tasklist

FS#26575 - util-linux 2.20-2+ mkfs.ext(2|3|4) fails with munmap_chunk(): invalid pointer

Attached to Project: Arch Linux
Opened by Jason Begley (jayray) - Saturday, 22 October 2011, 19:38 GMT
Last edited by Tom Gundersen (tomegun) - Tuesday, 25 October 2011, 18:20 GMT
Task Type Bug Report
Category Packages: Core
Status Closed
Assigned To Dave Reisner (falconindy)
Tom Gundersen (tomegun)
Architecture i686
Severity High
Priority Normal
Reported Version
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 0
Private No

Details

Description:

when running /sbin/mkfs.ext3 /dev/ram0 I get the following error:

mke2fs 1.41.14 (22-Dec-2010)
*** glibc detected *** /sbin/mkfs.ext3: munmap_chunk(): invalid pointer: 0xb7de4fee ***
======= Backtrace: =========
/lib/libc.so.6(+0x6e6e2)[0xb7de06e2]
/lib/libc.so.6(+0x6e928)[0xb7de0928]
/lib/libblkid.so.1(+0xfc38)[0xb7ef5c38]
/lib/libblkid.so.1(+0x19f0c)[0xb7efff0c]
/lib/libblkid.so.1(+0x19989)[0xb7eff989]
/lib/libblkid.so.1(+0x745a)[0xb7eed45a]
/lib/libblkid.so.1(blkid_probe_get_topology+0x22)[0xb7effb82]
/sbin/mkfs.ext3[0x804be4a]
/sbin/mkfs.ext3[0x804d2fd]
/lib/libc.so.6(__libc_start_main+0xf3)[0xb7d8b223]
======= Memory map: ========
08048000-08055000 r-xp 00000000 08:02 4046864 /sbin/mkfs.ext3
08055000-08056000 rw-p 0000c000 08:02 4046864 /sbin/mkfs.ext3
08056000-08077000 rw-p 00000000 00:00 0 [heap]
b7b54000-b7b70000 r-xp 00000000 08:02 2269255 /usr/lib/libgcc_s.so.1
b7b70000-b7b71000 rw-p 0001b000 08:02 2269255 /usr/lib/libgcc_s.so.1
b7b90000-b7d55000 r--p 00000000 08:02 2654317 /usr/lib/locale/locale-archive
b7d55000-b7d57000 rw-p 00000000 00:00 0
b7d57000-b7d6e000 r-xp 00000000 08:02 1327299 /lib/libpthread-2.14.so
b7d6e000-b7d6f000 r--p 00016000 08:02 1327299 /lib/libpthread-2.14.so
b7d6f000-b7d70000 rw-p 00017000 08:02 1327299 /lib/libpthread-2.14.so
b7d70000-b7d72000 rw-p 00000000 00:00 0
b7d72000-b7ed3000 r-xp 00000000 08:02 1327322 /lib/libc-2.14.so
b7ed3000-b7ed5000 r--p 00161000 08:02 1327322 /lib/libc-2.14.so
b7ed5000-b7ed6000 rw-p 00163000 08:02 1327322 /lib/libc-2.14.so
b7ed6000-b7ed9000 rw-p 00000000 00:00 0
b7ed9000-b7ede000 r-xp 00000000 08:02 1327286 /lib/libe2p.so.2.3
b7ede000-b7edf000 rw-p 00005000 08:02 1327286 /lib/libe2p.so.2.3
b7edf000-b7ee0000 rw-p 00000000 00:00 0
b7ee0000-b7ee4000 r-xp 00000000 08:02 1327284 /lib/libuuid.so.1.3.0
b7ee4000-b7ee5000 r--p 00003000 08:02 1327284 /lib/libuuid.so.1.3.0
b7ee5000-b7ee6000 rw-p 00004000 08:02 1327284 /lib/libuuid.so.1.3.0
b7ee6000-b7f0d000 r-xp 00000000 08:02 1327285 /lib/libblkid.so.1.1.0
b7f0d000-b7f0f000 r--p 00026000 08:02 1327285 /lib/libblkid.so.1.1.0
b7f0f000-b7f10000 rw-p 00028000 08:02 1327285 /lib/libblkid.so.1.1.0
b7f10000-b7f12000 r-xp 00000000 08:02 1327113 /lib/libcom_err.so.2.1
b7f12000-b7f13000 rw-p 00001000 08:02 1327113 /lib/libcom_err.so.2.1
b7f13000-b7f3c000 r-xp 00000000 08:02 1327293 /lib/libext2fs.so.2.4
b7f3c000-b7f3d000 rw-p 00029000 08:02 1327293 /lib/libext2fs.so.2.4
b7f5b000-b7f5d000 rw-p 00000000 00:00 0
b7f5d000-b7f5e000 r-xp 00000000 00:00 0 [vdso]
b7f5e000-b7f7d000 r-xp 00000000 08:02 1327326 /lib/ld-2.14.so
b7f7d000-b7f7e000 r--p 0001e000 08:02 1327326 /lib/ld-2.14.so
b7f7e000-b7f7f000 rw-p 0001f000 08:02 1327326 /lib/ld-2.14.so
bfc85000-bfc9a000 rw-p 00000000 00:00 0 [stack]
Aborted

This was noticed after upgrade to util-linux-2.20.1-1-i686 and rebooted (I run scripts at boot to generate/mount ram disks for use with HAVP)
I downgraded to util-linux-2.20-2-i686 and issue followed. The error did not exist when downgraded to util-linux-2.19.1-3. I reproduced this error on another similar box.

The following are exceptions to being update with current pacman:
manual compiled kernel 2.6.30.4; and the following packages excluded from pacman:
IgnorePkg = kernel26 kernel26-scsi kernel24 kernel24-scsi tightvnc freeradius dansguardian syslog-ng php-apache net-snmp rrdtool php ffmpeg x264


This task depends upon

Closed by  Tom Gundersen (tomegun)
Tuesday, 25 October 2011, 18:20 GMT
Reason for closing:  Upstream
Additional comments about closing:  We don't support this old kernels, please contact upstream if you want to help sort this out.
Comment by Dave Reisner (falconindy) - Sunday, 23 October 2011, 02:54 GMT
I can't reproduce this on i686 or x86_64. Would it be possible for you to rebuild util-linux with debug symbols so that we can get a backtrace with useful info from libblkid?
Comment by Dave Reisner (falconindy) - Sunday, 23 October 2011, 03:44 GMT
FYI, your kernel is older than what we generally support. I'm a little surprised that udev works properly. What's the need for it?
Comment by Jason Begley (jayray) - Sunday, 23 October 2011, 03:59 GMT
I will try the rebuild, and try on a workstation running current kernel. I haven't updated the kernel as I dont use the arch built kernel package due to some old hardware requirements.
Comment by Tom Gundersen (tomegun) - Sunday, 23 October 2011, 09:56 GMT
Just wanted to point out (to anyone who might read this) that 2.20-2 did not have libmount enabled, whereas 2.20.1-1 does, so it appears that this is not libmount related.

Also, if there are hardware issues with the stock Arch kernel, you should file bug reports so we can sort it out (in case you have not already done so).

Looking forward to the backtrace.
Comment by Jason Begley (jayray) - Sunday, 23 October 2011, 13:32 GMT
Here is the backtrace, I don't think it was done correctly based on my output... Just to confirm, I loaded current kernel and this issue did not exist. I may be transitioning to the new kernel.


#0 0xb7e6942e in free () from /lib/libc.so.6
#1 0xb7f787b3 in sysfs_deinit (cxt=0xbffff428) at ../../lib/sysfs.c:171
#2 0xb7f82dd4 in probe_sysfs_tp (pr=0x80592e8, mag=0x0) at sysfs.c:106
#3 0xb7f82961 in topology_probe (pr=0x80592e8, chn=0x8059354) at topology.c:186
#4 0xb7f71254 in blkid_probe_get_binary_data (pr=0x80592e8, chn=0x8059354) at probe.c:355
#5 0xb7f82a94 in blkid_probe_get_topology (pr=0x80592e8) at topology.c:138
#6 0x0804be4a in ?? ()
#7 0x0804d2fd in ?? ()
#8 0xb7e10223 in __libc_start_main () from /lib/libc.so.6
#9 0x08049fe1 in ?? ()
Backtrace stopped: Not enough registers or memory available to unwind further
Comment by Dave Reisner (falconindy) - Tuesday, 25 October 2011, 17:23 GMT
Unless Tom wants to push this further, I think we should blame the older kernel and close this.
Comment by Tom Gundersen (tomegun) - Tuesday, 25 October 2011, 18:19 GMT
Yeah, this does look like a real bug, but it would have to be taken upstream I think.

Loading...