firewire not detected on resume from suspend

Bug #48317 reported by Stoic Jed
18
Affects Status Importance Assigned to Milestone
linux-source-2.6.15 (Ubuntu)
Fix Released
Medium
Unassigned

Bug Description

After resuming from suspend, firewire devices are no longer detected. Unplugging and replugging the firewire cable has no effect.

The following messages were found in /var/log/kern.log after suspending and resuming with a mounted firewire mass storage device.

ohci1394: fw-host0: Running dma failed because Node ID is not valid
ieee1394: sbp2: aborting sbp2 command
ohci1394: fw-host0: Running dma failed because Node ID is not valid
ohci1394: fw-host0: Running dma failed because Node ID is not valid
ieee1394: sbp2: aborting sbp2 command
ohci1394: fw-host0: Running dma failed because Node ID is not valid
ieee1394: sbp2: reset requested
ieee1394: sbp2: Generating sbp2 fetch agent reset
ohci1394: fw-host0: Running dma failed because Node ID is not valid
ohci1394: fw-host0: Running dma failed because Node ID is not valid
ieee1394: sbp2: aborting sbp2 command
ohci1394: fw-host0: Running dma failed because Node ID is not valid
ieee1394: sbp2: reset requested
ieee1394: sbp2: Generating sbp2 fetch agent reset
ohci1394: fw-host0: Running dma failed because Node ID is not valid

Note: the original reporter indicated the bug was in package 'linux-image-2.6.15-21-686'; however, that package was not published in Ubuntu.

Revision history for this message
Stefan Richter (stefan-r-ubz) wrote :

This is tracked as an upstream bug at http://bugzilla.kernel.org/show_bug.cgi?id=7072 . You could add your e-mail address there if you are interested in latest news. Don't expect a fix soon.

As a workaround, mark ohci1394 (and perhaps the other IEEE 1394 drivers) to be unloaded before suspend. I don't know though how to do this in Ubuntu.

Revision history for this message
Stefan Richter (stefan-r-ubz) wrote :

PS: Of course you should unmount any FireWire storage devices before unloading ohci1394.

Revision history for this message
Stefan Richter (stefan-r-ubz) wrote :

fixed in upstream sources (kernel.org's sources) in linux-2.6.21

Revision history for this message
Steven Ellis (steven-openmedia) wrote :

I can replicate the problem on Ubuntu Feisty with Kernel 2.6.20-16-generic.

Currently running linux-image-2.6.20-16.28

System can system/resume but the error means attempts to suspend are reported as failures even though the work.

The following is reported in the kernel log

[ 1606.058603] ohci1394 0000:00:14.0: suspend
[ 1606.058610] ohci1394 does not fully support suspend and resume yet

Revision history for this message
Stefan Richter (stefan-r-ubz) wrote :
Revision history for this message
Stefan Richter (stefan-r-ubz) wrote :

PS: Even after the fix, the sbp2 driver may still have problems to transparently reconnect to the disk after resume. It could happen that the SCSI device is destroyed (during suspend or right after resume) and then re-created as a different SCSI device. It may even be necessary to unplug and replug the disk. Therefore it's still advisable to umount before suspend. I have vague ideas how to improve this but I haven't had time to get hacking on this.

Revision history for this message
Gareth Fitzworthington (mapping-gp-deactivatedaccount) wrote :

Upstream appears to have committed a patch for this. This should have made it into Gutsy.
Can anyone verify this?

Changed in linux-source-2.6.15:
status: New → Incomplete
Revision history for this message
Connor Imes (ckimes) wrote :

Thank you again for taking the time to report this bug. This bug report is being closed due to comments regarding this being fixed with an update. If you find the bug was not fixed, please re-open it by changing the status to New.

Changed in linux-source-2.6.15:
status: Incomplete → Fix Released
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Duplicates of this bug

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.