2

I was installing glibc v2.18 and suddenly the install (make install) halts with an error. Afterwards, any basic terminal commands (ls, sudo, etc.) and attempting to perform system operations (shutdown, restart, etc.) resulted in segmentation faults.

I performed a hard restart of my system, and upon attempting to boot Ubuntu, my machine displays the following:

[    2.297365] init[1]: segfault at 3cd ip 00000000000003cd sp 00007fffa0a7f468 error 14 in libc-2.15.so[7fecf93a4000+1b5000]
[    2.297640] Kernel panic - not syncing: Attempted to kill init! exitcode=0x0000000b
[    2.297640] 
[    2.297707] Pid: 1, comm: init Not tainted 3.5.0-39-generic #60~precise1-Ubuntu
[    2.297766] Call Trace:
[    2.297814]  [<ffffffff8168500e>] panic+0xc1/0x1d7
[    2.297867]  [<ffffffff811198f0>] ? perf_adjust_period+0xb0/0xb0
[    2.297923]  [<ffffffff81057582>] find_new_reaper+0x132/0x140
[    2.297977]  [<ffffffff810589e5>] forget_original_parent+0x45/0x1a0
[    2.298033]  [<ffffffff81058b57>] exit_notify+0x17/0x110
[    2.298086]  [<ffffffff810593e5>] do_exit+0x1f5/0x480
[    2.298138]  [<ffffffff81089814>] do_group_exit+0x44/0xa0
[    2.298192]  [<ffffffff810698cb>] get_signal_to_deliver+0x22b/0x440
[    2.298249]  [<ffffffff810147e9>] do_signal+0x29/0x130
[    2.298301]  [<ffffffff81685185>] ? printk+0x61/0x63
[    2.298353]  [<ffffffff81144966>] ? vm_mmap_pgoff+0x96/0xb0
[    2.298407]  [<ffffffff81155c63>] ? sys_mmap_pgoff+0x103/0x1f0
[    2.298462]  [<ffffffff810149a0>] do_notify_resume+0x90/0xd0
[    2.298516]  [<ffffffff8169e32c>] retint_signal+0x48/0x8c

I am unsure how to resolve the kernel panic and successfully boot Ubuntu.

Alex Brooks
  • 177

1 Answers1

3

glibc is a crucial part of Ubuntu, so unless you really need a different version it is better to let it stay.

Regarding your problem, just re-installing libc6 should fix your issue. As you cannot use that system, you can try to recover it using a live cd.

After booting from a live cd/usb, mount your root partition (you can use Nautilus of the mount command to do that).

Then you should grab a copy of libc6 .deb. You can get it in http://packages.ubuntu.com/dist/libc6, where dist is your Ubuntu version (Precise, Quantal, Raring, etc). Double check you download the version that matches your architecture.

After that just extract that .deb to your root partition. This is not equal to use apt-get install, but as you cannot use any command on that system...

Supposing your root partition is mounted in /mnt, you can extract the .deb using

$ sudo dpkg -x /path/to/libc6*.deb /mnt

If everything went well, you can unmount your root partition and reboot to check if your problem is solved.

Edit

As pointed by fkraiem, after you got a working system you should run sudo apt-get install --reinstall libc6 to make sure libc gets correctly installed.

Community
  • 1
Salem
  • 19,744
  • 1
    Just to be sure, I would also cleanly reinstall the package with sudo apt-get install --reinstall libc6 in case the installation process does something more than just extracting the files. – fkraiem Oct 09 '14 at 21:19
  • Is there a way to obtain error dump during booting, as I am facing same issue and have kernel panic and cannot login, in my case I can see only the lower part of the error and not upper part which tells precisely which shared object of glibc is throwing the error. My problem is listed here – user0193 Jan 14 '23 at 13:46