Description of problem:
most of my attempts to hibernate my ThinkPad X60 (type: 1706-GMG) fail
Version-Release number of selected component (if applicable):
Steps to Reproduce:
1. swapoff -a
2. run badblocks (in write-mode with -p 2) to make sure swap is actually OK
4. swapon -a
5. select 'hibernate' from the gnome power applet
machine tries to hibernate but end up getting I/O errors
100% successful hibernate and wake-up
A serial console log of a failed atempt will be attached.
This X60 has a Core2Duo and is running x86_64 RHEL5.
Created attachment 151414 [details]
serial console output of a failed suspend, the log ends with me powering off the machine
Created attachment 152991 [details]
serial capture when trying to suspend with pci=nomsi
[pcfe@x60-pcfe ~]$ uname -r
[pcfe@x60-pcfe ~]$ cat /proc/cmdline
ro root=/dev/VG_x60_internal/LV_root console=ttyS0,115200 console=tty0
[pcfe@x60-pcfe ~]$ cat /proc/interrupts
0: 136602 72151 IO-APIC-edge timer
1: 136 23 IO-APIC-edge i8042
7: 1 0 IO-APIC-edge parport0
8: 1 0 IO-APIC-edge rtc
9: 562 38 IO-APIC-level acpi
12: 153 0 IO-APIC-edge i8042
201: 23557 0 IO-APIC-level yenta, uhci_hcd:usb1, libata, eth0,
209: 3618 2023 IO-APIC-level uhci_hcd:usb2, ipw3945, HDA Intel
217: 7 0 IO-APIC-level uhci_hcd:usb3, sdhci:slot0
225: 8425 6202 IO-APIC-level uhci_hcd:usb4, ehci_hcd:usb5
NMI: 220 93
LOC: 208546 208652
I don't like the look of those I/O errors. Are you getting filesystem corruption
in response to Comment #5
don't seem to, just did a full fsck (shutdown -F) and it did not nag at me. Also
/lost+found is empty.
I will now move to FC devel (because of another bug) and report here with a
short note if suspend works with that kernel or not.
additional response to Comment #5
2.6.21-1.3116.fc7 handles this better
It puts the machine to sleep without any I/O errors.
(machine fails to fully wake, but that will be another bug)
FWIW: failure to wake under Fedora development has been filed as Bug #238407
Looks like this can be closed. Is that right?
As I have moved the box in question from RHEL to Fedora, yes this can be closed
as I am unable to provide further testing.