nash sigfault (receiving SIGSEGV!) on boot

elekt · #1 18th October 2008, 06:42 PM

Hey guys, recently I updated my kernel to 2.6.26.5-45.fc9.x86_64 and now when I try to boot I run into two problems.
1. My LUKS passphrase is not accepted......this is really bad.
2. nash seg faults right when it is initialized, halting everything.

Firstly my swap space (/dev/sda2) is LUKS encrypted as is my /home partition (/dev/sdb3). Before anything else, right after the "Red Hat nash version 6.0.52 starting" message, it asks for the LUKS passphrase for the swap space (/dev/sda2). Usually I enter this and it continues booting, eventually asking for the LUKS passphrase for /home (/dev/sdb3). However, now no matter what I enter, it keeps prompting me for the passphrase eventually saying

Code:

Command failed: No key available with this passphrase.

Since it can't mount the swap space it continues booting without it, this is where nash craps a brick.

Here is everything that's displayed at boot (with the exception of the 'root (hd0,0) kernel /boot/vmlinuz........' part).
I typed everything out, so if something seems wrong/there are typos let me know and I'll see if I mistyped something.

Quote:

Originally Posted by at boot

Red Hat nash version 6.0.52 starting
Loading /lib/kbd/keymaps/i386/qwerty/us.map

Enter LUKS passphrase for /dev/sda2:
Enter LUKS passphrase for /dev/sda2:
Enter LUKS passphrase for /dev/sda2:
Command failed: No key available with this passphrase.

device-mapper: table ioctl failed: No such device or address
device-mapper: deps ioctl failed: No such device or address
init[1]: segfault at 10 ip 380e80e004 sp 7fff1b550c68 error 4 in libdevmapper.so.1.02[380e800000+15000]
nash received SIGSEGV! Backtrace (16):
/bin/nash[0x40d093]
/lib64/libc.so.6[0x38058322a0]
/lib64/libdevmapper.so.1.02(dm_task_get_deps+0x4)[0x380e80e004]
/usr/lib64/libnash.so.6.0.52[0x3dfa40ec1b]
/usr/lib64/libnash.so.6.0.52(nashDmDevGetName+0x3d)[0x3dfa40fab0]
/usr/lib64/libnash.so.6.0.52[0x3dfa40c2e3]
/usr/lib64/libnash.so.6.0.52[0x3dfa40c3fc]
/usr/lib64/libnash.so.6.0.52(nashBdevIterNext+0x120)[0x3dfa40c871]
/usr/lib64/libnash.so.6.0.52[0x3dfa40cabb]
/usr/lib64/libnash.so.6.0.52(nashFindFsByName+0x60)[0x3dfa40cb86]
/usr/lib64/libnash.so.6.0.52(nashAGetPathBySpec+0x86)[0x3dfa40cc73]
/bin/nash[0x408b24]
/bin/nash[0x40cf49]
/bin/nash[0x40d576]
/lib64/libc.so.6(__libc_start_main+0xfa)[0x380581e32a]
/bin/nash[0x404509]

Can anyone figure out what's wrong and/or how to get around this?

I should also mention that my root partition is unencrypted, so I can access that.

leigh123linux · #2 18th October 2008, 07:27 PM

Have you seen this post ?

http://forums.fedoraforum.org/showthread.php?t=201990

your error output is the same as this bug report

https://bugzilla.redhat.com/show_bug.cgi?id=443332

elekt · #3 18th October 2008, 07:40 PM

I have, however I have been booting with a 16gb Corsair Flash Voyager plugged in, unplugging it and everything except for my Saitek Eclipse keyboard doesn't solve the problem. It doesn't change anything as far as I can see.

That's why I posted here, I thought maybe someone else had another idea :/

leigh123linux · #4 18th October 2008, 08:31 PM

I wasn't referring to that usb rubbish , I was pointing out the patch for mkinitrd .

I have applied the patch and attached some 64 bit rpm's

elekt · #5 18th October 2008, 09:02 PM

Ah, okay, thank you!

A quick question though. How is the patch applied? Seeing as my Fedora won't boot, can I apply it from a live distro?

leigh123linux · #6 18th October 2008, 09:13 PM

I edited the spec file then rebuilt the srpm .

Code:

%{!?python_sitelib: %define python_sitelib %(%{__python} -c "from distutils.sysconfig import get_python_lib; print get_python_lib(1)[1:]")}

Summary: Creates an initial ramdisk image for preloading modules.
Name: mkinitrd
Version: 6.0.52
Release: 2.1%{?dist}
License: GPLv2+
Group: System Environment/Base
Source0: mkinitrd-%{version}.tar.bz2
Patch1:  buffer_overrun.patch
ExclusiveOs: Linux
Prereq: dev
Requires: /bin/sh, /sbin/insmod.static, /sbin/losetup
Requires: fileutils, grep, mount, gzip, tar, mktemp >= 1.5-5, findutils

%prep
%setup -q -n mkinitrd-%{version}
%patch1 -p1 -b .buffer_overrun
find . -name "Makefile*" -exec sed -i 's|-Werror||g' {} \;

%build
make LIB=%{_lib}
make LIB=%{_lib} test

Quote:

Seeing as my Fedora won't boot, can I apply it from a live distro?

Yes you can use a livecd ( 64 bit only , a livecd version of fedora would be a good choice ) , then mount the root partition and chroot into it , then use rpm to install the new rpm's

elekt · #7 19th October 2008, 12:00 AM

Alright thanks so much, I'll get to it right away.

...now I just gotta find my FC9 dvd

Acajoe · #8 27th November 2008, 03:10 PM

Hello,
I am brand new to Fedora. I have some experience installing FreeBSD and Gentoo Linux, but am pretty much a noob.
I just installed Fedora 10 and am also having this nash segfault described above.
Just need to know...what exactly do I do with this code?

Code:

%{!?python_sitelib: %define python_sitelib %(%{__python} -c "from distutils.sysconfig import get_python_lib; print get_python_lib(1)[1:]")}

Summary: Creates an initial ramdisk image for preloading modules.
Name: mkinitrd
Version: 6.0.52
Release: 2.1%{?dist}
License: GPLv2+
Group: System Environment/Base
Source0: mkinitrd-%{version}.tar.bz2
Patch1:  buffer_overrun.patch
ExclusiveOs: Linux
Prereq: dev
Requires: /bin/sh, /sbin/insmod.static, /sbin/losetup
Requires: fileutils, grep, mount, gzip, tar, mktemp >= 1.5-5, findutils

%prep
%setup -q -n mkinitrd-%{version}
%patch1 -p1 -b .buffer_overrun
find . -name "Makefile*" -exec sed -i 's|-Werror||g' {} \;

%build
make LIB=%{_lib}
make LIB=%{_lib} test

Perhaps you could explain:

Quote:

use rpm to install the new rpm's

Thank you much.

firesteel · #9 1st December 2008, 03:10 PM

I just installed Fedora 10 and I have the problem described here. Would I apply the same patch to it?

Tquilha · #10 5th February 2009, 07:33 PM

Hello, I installed FC10 yesterday, and had the same problem discussed here. From another site, I saw that having RAID capable hardware, but not actually using RAID could cause some problems. So I re-installed FC 10 to the same hardware, only using the nodmraid option.
Now, it works fine

Here's my setup :

AMD Athlon64 3700+
1 GB RAM DDR 400
Asus A8N motherboard
2 IDE HDs
1 SATA HD
GeForce 7800GTX
Creative Labs Audigy 2 ZS

My linux partitions are all on the 2nd IDE drive.

elekt · #11 5th February 2009, 08:05 PM

Great tip Tquilha! Thanks!!

However I already fixed this problem when I upgraded my F9 install with the F10 DVD. Now it boots and works fine! But thats a great tip to keep in mind, thanks again!

coyoteboy · #12 17th May 2009, 12:40 AM

God how damned annoying is this. Just installed 10 (32 bit) with encryption and enjoyed using it for several days, only to boot this morning with no apparent changes and it does exactly as seen above. Can't get to boot options and the encryption passphrase cant be entered correctly. Looks like I'm going to have to waste the last few days work and give in with Fedora, I like it but there's just so many show-stopping faults I've come across with 8 9 and 10 (possibly partially down to my own incompetence) that I just cant justify wasting time fixing it constantly.

Depressed