First, the problem re-occured.
Second, *sometimes* the MB fails to boot, so I guess it is a HW problem.
Or, maybe a HP problem
Third, I found a non-deterministic way to test it sometimes.
The bug reoccured twice, both time when syncing the file systems with a
Java tool (but there were ten other cases, when syncing with the same tool
was fine). When occurs, the bug causes FAT corruption (nasty ones -
invalid dirs, different files using the same sector, lost chains, etc. ).
I can not reproduce it at will. With fsck.msdos /dev/sda1 -rtV I made a
few tests but it seems to be a problem that "goes away" after a while.
Would usb debugging or SCSI debugging turned on in the kernel help?
The error messages, de-duplicated:
reset high speed USB device using ehci_hcd and address X
device descriptor read/64, error -110
device not accepting address X, error -110
USB disconnect, address X
sd 1:0:0:0: scsi: Device offlined - not ready after error recovery
sd 1:0:0:0: [sda] Result: hostbyte=0x01 driverbyte=0x00
end_request: I/O error, dev sda, sector 1968432
Buffer I/O error on device sda1, logical block 1968XXX
[sda] Result: hostbyte=0x01 driverbyte=0x00
end_request: I/O error, dev sda, sector 1968XXX
scsi 1:0:0:0: rejecting I/O to dead device
FAT: Filesystem panic (dev sda1)
fat_get_cluster: invalid cluster chain (i_pos 0)
File system has been set read-only
FAT: Corrupted directory (i_pos XXXXXXXX)
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact email@example.com