You can't always trust the timestamps in the messages around suspend time, the last messages during suspend are stored in a buffer and are timestamped and written out after resume. So typically all the suspend and the earliest resume messages will have the same timestamp.
For now I hope the upstream experts will chime in. In fact, with the 2.6.30 kernels I see similar issues on my X700 Mobility card (no shared memory BTW)...
You can't always trust the timestamps in the messages around suspend time, the last messages during suspend are stored in a buffer and are timestamped and written out after resume. So typically all the suspend and the earliest resume messages will have the same timestamp.
For now I hope the upstream experts will chime in. In fact, with the 2.6.30 kernels I see similar issues on my X700 Mobility card (no shared memory BTW)...