when i run ltsp-update-image, it is aborted

Bug #134586 reported by Pavel Mlčoch
2
Affects Status Importance Assigned to Milestone
ltsp (Ubuntu)
Fix Released
Undecided
Scott Balneaves
squashfs (Ubuntu)
Invalid
Undecided
Unassigned

Bug Description

root@pavka:/home/pavka/commands# ltsp-update-image 1>output.txt 2>error.txt
*** glibc detected *** mksquashfs: malloc(): memory corruption: 0x0d48dde8 ***
======= Backtrace: =========
/lib/tls/i686/cmov/libc.so.6[0xb7e3d646]
/lib/tls/i686/cmov/libc.so.6(__libc_malloc+0x90)[0xb7e3efd0]
/lib/tls/i686/cmov/libc.so.6(getpwuid+0x11b)[0xb7e6496b]
mksquashfs[0x805a4bf]
/lib/tls/i686/cmov/libc.so.6(__libc_start_main+0xe0)[0xb7de9050]
mksquashfs[0x80494b1]
======= Memory map: ========
08048000-08067000 r-xp 00000000 fe:02 951672 /usr/sbin/mksquashfs
08067000-08068000 rw-p 0001e000 fe:02 951672 /usr/sbin/mksquashfs
08068000-0dc8e000 rw-p 08068000 00:00 0 [heap]
b0700000-b0762000 rw-p b0700000 00:00 0
b0762000-b0800000 ---p b0762000 00:00 0
b4800000-b48f1000 rw-p b4800000 00:00 0
b48f1000-b4900000 ---p b48f1000 00:00 0
b4a00000-b4af1000 rw-p b4a00000 00:00 0
b4af1000-b4b00000 ---p b4af1000 00:00 0
b4c00000-b4cf1000 rw-p b4c00000 00:00 0
b4cf1000-b4d00000 ---p b4cf1000 00:00 0
b4d4b000-b4dcc000 rw-p b4d4b000 00:00 0
b4dcc000-b4dcd000 ---p b4dcc000 00:00 0
b4dcd000-b55cd000 rw-p b4dcd000 00:00 0
b55cd000-b55ce000 ---p b55cd000 00:00 0
b55ce000-b5dce000 rw-p b55ce000 00:00 0
b5dce000-b5dcf000 ---p b5dce000 00:00 0
b5dcf000-b65cf000 rw-p b5dcf000 00:00 0
b65cf000-b65d0000 ---p b65cf000 00:00 0
b65d0000-b6dd0000 rw-p b65d0000 00:00 0
b6dd0000-b6dd1000 ---p b6dd0000 00:00 0
b6dd1000-b75d1000 rw-p b6dd1000 00:00 0
b75d1000-b75d2000 ---p b75d1000 00:00 0
b75d2000-b7dd3000 rw-p b75d2000 00:00 0
b7dd3000-b7f17000 r-xp 00000000 fe:02 1884247 /lib/tls/i686/cmov/libc-2.6.1.so
b7f17000-b7f18000 r--p 00143000 fe:02 1884247 /lib/tls/i686/cmov/libc-2.6.1.so
b7f18000-b7f1a000 rw-p 00144000 fe:02 1884247 /lib/tls/i686/cmov/libc-2.6.1.so
b7f1a000-b7f1d000 rw-p b7f1a000 00:00 0
b7f1d000-b7f40000 r-xp 00000000 fe:02 311619 /lib/tls/i686/cmov/libm-2.6.1.so
b7f40000-b7f42000 rw-p 00023000 fe:02 311619 /lib/tls/i686/cmov/libm-2.6.1.so
b7f42000-b7f43000 rw-p b7f42000 00:00 0
b7f43000-b7f57000 r-xp 00000000 fe:02 311708 /lib/tls/i686/cmov/libpthread-2.6.1.so
b7f57000-b7f59000 rw-p 00013000 fe:02 311708 /lib/tls/i686/cmov/libpthread-2.6.1.so
b7f59000-b7f5b000 rw-p b7f59000 00:00 0
b7f5b000-b7f6f000 r-xp 00000000 fe:02 2326607 /usr/lib/libz.so.1.2.3.3
b7f6f000-b7f70000 rw-p 00013000 fe:02 2326607 /usr/lib/libz.so.1.2.3.3
b7f9d000-b7fa7000 r-xp 00000000 fe:02 1180205 /lib/libgcc_s.so.1
b7fa7000-b7fa8000 rw-p 0000a000 fe:02 1180205 /lib/libgcc_s.so.1
b7fa8000-b7fab000 rw-p b7fa8000 00:00 0
b7fab000-b7fc5000 r-xp 00000000 fe:02 99541 /lib/ld-2.6.1.so
b7fc5000-b7fc7000 rw-p 00019000 fe:02 99541 /lib/ld-2.6.1.so
bfb0b000-bfb52000 rw-p bfb0b000 00:00 0 [stack]
ffffe000-fffff000 r-xp 00000000 00:00 0 [vdso]
root@pavka:/home/pavka/commands#

Revision history for this message
Pavel Mlčoch (pavkamlc) wrote :
Revision history for this message
Pavel Mlčoch (pavkamlc) wrote :
Revision history for this message
Oliver Grawert (ogra) wrote :

could it be that proc was still mounted in the chroot while you ran the command ?

Revision history for this message
Pavel Mlčoch (pavkamlc) wrote :

Yes, You are right. Error is there only if the proc is mounted in the chroot

Revision history for this message
Scott Balneaves (sbalneav) wrote :

I've just added a check for a mounted /proc in my tree. Hopefully that should address this.

Scott

Changed in ltsp:
assignee: nobody → sbalneav
status: New → Fix Committed
Revision history for this message
Oliver Grawert (ogra) wrote :

lets change:
+ echo "/proc mounted in chroot ${CHROOT}, Exiting." && \
+ exit 1

to be

+ echo "/proc mounted in chroot ${CHROOT}, Unmounting." && \
+ umount ${CHROOT}/proc

;)

Revision history for this message
Oliver Grawert (ogra) wrote :

fixed in 5.0.25

Changed in ltsp:
status: Fix Committed → Fix Released
Revision history for this message
Oliver Grawert (ogra) wrote :

mksquashfs has a switch to disable using proc, marking as invalid

Changed in squashfs:
status: New → Invalid
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.