Red Hat Bugzilla – Bug 403671
Kernel 2.6.24-0.43.rc3.git1.fc9 hangs during boot during isapnp scan for PnP cards
Last modified: 2007-12-12 16:52:23 EST
Description of problem:
kernel 2.6.24-0.43.rc3.git1.fc9 hangs during boot
Version-Release number of selected component (if applicable):
Steps to Reproduce:
1.Boot the system
Kernel hangs during boot after message:
isipnp: Scanning for PnP cards...
Previous kernel (2.6.24-0.42.rc3.git1.fc9) continues booting after outputting
the following message:
isapnp: No Plug & Play device found
Same problem exists in 2.6.24-0.45-rc3.git1.fc9 downloaded from koji.
Comment #1 of bug 402061 appears to be the same problem (although original
report appears to be a different problem), but is reported against
2.6.24-0.42.rc3.git1.fc9, but this is the latest version that works for me.
Seeing same on 2.6.24-0.55.rc3.git3.fc9.
2.6.24-0.42.rc3.git1.fc9 is also the last one that works for me.
Seeing same problem with kernel-2.6.24-0.57.rc3.git4.fc9
Last one to boot was 0.42.
using kernel parm noisapnp shows additional msgs before freeze (hand transcribed):
PCI_hotplug: PCI Hot Plug PCI Core Version: 0.5
isapnp: ISA Plug & Play Support disabled
Real Time Clock Driver v1.12ac
Linux agpgart interface V0.103
agpgart: Detected an Intel i875 Chipset.
agpgart: AGP aperture is 64M @ 0xf8000000
Serial: 8250/16550 driver $ Revision: 1.90 $ 4 ports, IRQ sharing enabled
(hard freeze at this point)
Last good kernel on this x86 system was 2.6.24-0.42.rc3.git1.fc9.
How do you find changes made after 2.6.24-0.42.rc3.git1.fc9 to try to
exactly the same here as per Clyde comment
I have experienced a same probelm.
But, I think that gcc-4.1.2-34 is broken.
I get back to gcc-4.1.2-33 and generate 2.6.24-0.57.rc3.git4.
Then Booting is complete.
kernels 43 -> 61 will not boot. All stop very early at
Time: hpet clocksource has been installed
Thinkpad T61 2.4Ghz Duocore. This sysyem has never needed
any kernel params
Could you try booting with "initcall_debug" turned on? It should help us narrow
down which code is problematic.
I assume this is 32-bit? (My t61 is booting fine in x86_64 mode with these kernels.)
I can confirm Masao Takahashi: gcc-4.1.2-34 breaks the i686 kernel packages.
I downloaded 2.6.24-0.61.rc3.git5.fc9.i686, it fails to boot.
I reverted gcc to gcc-4.1.2-33 and built i686 packages. It boots fine.
I'll check BZ for gcc, and if nothing is filed, I'll 'cross file' this.
This seems to be BZ'ed here: https://bugzilla.redhat.com/show_bug.cgi?id=407281
Comment #7 From Kyle McMartin
""initcall_debug" turned on" How do I do that?
Yes, it is 32bit rawhide, updated daily since F7. If it helps the mseg just
before the one I posted (hpet) states somethin about 3 64bit timers.
I forgot to mention selinux is disabled.
Comment #7 From Kyle McMartin (again)
There was a new kernel in this mornings updates so I
SWAG'd an edited grub.conf to
title Fedora (2.6.24-0.62.rc3.git5.fc9)
kernel /vmlinuz-2.6.24-0.62.rc3.git5.fc9 ro root=LABEL=/1 initcall_debug
Same failure, but more verbose
initcall 0xc0789b6e: init_pipe_fs+0x0/0x41()<6> Time: hpet clocksource has been
<hang> Did I do it right, does this help?
This kernel boots for me:
You might have hit bug 407281 here. See if the above kernel boots for you
Comment #12, #13 From Josh Boyer
Todays 32bit rawhid update, 2.6.24-0.73.rc4.git1.fc9 boots normally.
Seems to fixed in kernel-2.6.24-0.73.rc4.git1.fc9!