Bug 156909 - system halts when booting new kernel-2.6.11-1.1284_FC4
system halts when booting new kernel-2.6.11-1.1284_FC4
Status: CLOSED RAWHIDE
Product: Fedora
Classification: Fedora
Component: mkinitrd (Show other bugs)
4
i686 Linux
medium Severity medium
: ---
: ---
Assigned To: Peter Jones
David Lawrence
:
Depends On:
Blocks:
  Show dependency treegraph
 
Reported: 2005-05-05 01:17 EDT by Sean Godsell
Modified: 2007-11-30 17:11 EST (History)
5 users (show)

See Also:
Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of:
Environment:
Last Closed: 2005-05-16 15:07:17 EDT
Type: ---
Regression: ---
Mount Type: ---
Documentation: ---
CRM:
Verified Versions:
Category: ---
oVirt Team: ---
RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: ---


Attachments (Terms of Use)

  None (edit)
Description Sean Godsell 2005-05-05 01:17:37 EDT
Description of problem:
Install the latest updates for FC4 test 2, and the kernel does not boot now.

Version-Release number of selected component (if applicable):
kernel-2.6.11-1.1284_FC4.i686

How reproducible:
Just boot that kernel version on a toshiba A70 notebook

Steps to Reproduce:
1. Installed FC4 test 2 and all the packages
2. Install all the package updates including kernel-2.6.11-1.1284_FC4
3.
  
Actual results:
stops at Initializing hardware
or right after udev (it gets an error MAKEDEV: dir found)

Expected results:
boot up

Additional info:
Also on this notebook the mouse never works with the FC4 test 2 kernels (any of
them).  So I just download from kernel.org and run my own kernel with ps/2
support and the mouse works fine.  But point being your kernel should work with
my built in mouse.
Comment 1 Tom Wood 2005-05-05 08:23:46 EDT
I can confirm this hanging for both the UP and SMP kernels on a dual PIII, Via
686 chipset (Abit VP6).  Power cycling is required to reboot (three finger
salute doesn't work).
Comment 2 Frank Swasey 2005-05-05 15:48:51 EDT
I'm seeing the same problem on a Dell Lattitude C800 laptop with the
2.6.11-1.1284_FC4 and 2.6.11-1.1286_FC4 UP kernels.  2.6.11-1.1276_FC4 still
boots and works.
Comment 3 Sean Godsell 2005-05-05 16:07:03 EDT
I installed 1286_FC4 and it still dies at Initializing hardware...  But I notice
above is something that I believe is causing the problem.  When udevd is
started, there is a bad message  MAKEDEV: mkdir: File exists.  This should not
be.  Also this started occuring only after 1284_FC4.
Comment 4 Sean Godsell 2005-05-05 16:18:31 EDT
The initrd-xxxx.img is buggered up since 1284_FC4.  This needs to be fixed!
Comment 5 Sean Godsell 2005-05-05 19:12:36 EDT
I found the problem!  Starting in 2.6.11-1.1284_FC4 and above then initrd was
changed.  The init inside the initrd-2.6.11-1.128X_FC4.img was modified.  Lines
7 and 9 were changed from
7:   mount -t sysfs none /sys
to
7:   mount -t sysfs /sys /sys
and line 11 from
11:  mount -o mode=0755 -t tmpfs none /dev
to
11:  mount -o mode=0755 -t tmpfs /dev /dev

This is why there is an error when udev starts up.  The /dev is not mounted and
udev fails.
Note: to self (person making these images) test before release.  Say it again!
Test before release.  :-)
PS. Also get the mouse working please.  I would like to use a stock kernel where
the mouse would actually work.
Thank you.
Comment 6 djh 2005-05-05 19:26:20 EDT
This is a dupe of bug #156862.
Comment 7 Jonathan S. Shapiro 2005-05-06 12:31:01 EDT
I'm seeing the same problem in 2.6.11-1.1286.

It looks like line 3 is similarly buggered. It should read:

    from:
      mount -t proc /proc /proc
    to:
      mount -t proc none /proc

And in case sgodsell's post wasn't entirely clear, the correct /sys and /dev
lines are:

      mount -t sysfs none /sys
      mount -o mode=0755 -t tmpfs none /dev

Given that this bug isn't an obscure machine-dependent issue, one wonders how it
is possible to push an update RPM that obviously hasn't been tested?
Comment 8 Jonathan S. Shapiro 2005-05-06 12:32:41 EDT
Priority of this should be bumped to "high", because outsider testing is
substantially impeded until this is resolved.
Comment 9 Peter Jones 2005-05-16 15:07:17 EDT
This should be fixed with an updated udev.

Note You need to log in before you can comment on or make changes to this bug.