Description of problem: laptop hard disk inaccessible after a suspend-to-ram-resume cycle. kernel prints "journal commit I/O error" and "sd 0:0:0:0: [sda] Result: hostbyte=DID_BAD_TARGET driverbyte=DRIVER_OK,SUGGEST_OK" (or something to that effect, I had to write it down). Version-Release number of selected component (if applicable): kernel-2.6.23.1-23.fc8 known-bad, kernel-2.6.22.4-65.fc7 known-good. Sorry, that's a wide window, I know. How reproducible: Every time. Steps to Reproduce: 1. pm-suspend (or trigger the same by gnome-power-manager) 2. Resume 3. try to write to disk (or read outside of cache, I think)
Created attachment 233651 [details] lspci -v -nn output
What kind of laptop?
Acer TravelMate 6292, Core 2 Duo, Santa Rosa chipset. The attachment has the hardware details. I've been suspending and resuming this quite (but not 100%) reliably through gnome-power-manager and hal since I got it this summer under the F7 kernels. What instability I've had with it has been related to X and the Intel driver, but these days that's quite good as well. I believe this is the first kernel-related issue I've seen with the machine.
Tighter window -- the last F7 update kernel kernel-2.6.22.9-91.fc7 works, and the first F8 test kernel kernel-2.6.23-0.164.rc5.fc8 does not.
I found kernel-2.6.23.1-10.fc7 from koji. This kernel works. On the other hand, kernel-2.6.23.1-30.fc8 does not. I hope these are close enough to each other to see the difference...
Still continue to have the problem with kernel-2.6.23.1-42.fc8. Certainly seems to be some systematic difference between F7 and F8 kernel profiles.
Which kernel: i686 or x86_64?
Oh, sorry, neglected to mention that. i686.
kernel-2.6.23.8-63.fc8.i686 from updates-testing works