When using copydir to copy to my optical drive, it hung and the process went into D mode. This doesn't always happen. Usually it works OK. I suppose the likelihood is that it is a harware problem, but I should be grateful for confirmation of this. /var/log/messages gives: kernel: Assertion failure in do_get_write_access() at transaction.c:725: "(((jh2bh(jh))->b_state & (1UL << BH_Uptodate)) != 0)" kernel: invalid operand: 0000 kernel: CPU: 0 kernel: EIP: 0010:[do_get_write_access+1196/1296] kernel: EIP: 0010:[<c0159adc>] kernel: EFLAGS: 00010292 kernel: eax: 0000007b ebx: 00000000 ecx: 00000005 edx: cefb7f64 kernel: esi: 00000001 edi: c728c000 ebp: c8aef190 esp: cab5bdc8 kernel: ds: 0018 es: 0018 ss: 0018 kernel: Process copydir (pid: 10208, stackpage=cab5b000) kernel: Stack: c0225420 c022582e c0225400 000002d5 c02259e0 00000001 00000000 c44d8400 kernel: ce980720 c44d8494 c44d8400 c91e79c0 c8aef190 c0159b76 c91e79c0 c8aef190 kernel: 00000000 c217ca40 c91e79c0 cf74f400 c95ea900 c0155477 c91e79c0 c217ca40 kernel: Call Trace: [journal_get_write_access+54/96] [ext3_orphan_add+135/656] [ext3_mark_iloc_dirty+53/80] [__jbd_kmalloc+43/192] [start_this_handle+452/640] kernel: Call Trace: [<c0159b76>] [<c0155477>] [<c0153a65>] [<c015f37b>] [<c0158e24>] kernel: [journal_start+174/224] [ext3_setattr+132/272] [notify_change+79/208] [do_truncate+107/160] [lookup_hash+66/144] [open_namei+1033/1328] kernel: [<c0158f8e>] [<c0153924>] [<c01423af>] [<c012dfdb>] [<c0139032>] [<c0139609>] kernel: [dentry_open+195/320] [filp_open+54/96] [sys_open+52/144] [system_call+51/56] kernel: [<c012ef23>] [<c012ee36>] [<c012f124>] [<c0106edb>] kernel: kernel: Code: 0f 0b 8b 45 00 83 c4 14 8b 50 38 8b 70 34 8b 7d 0c 0f b7 40 kernel: SCSI disk error : host 0 channel 0 id 4 lun 0 return code = 28000002 kernel: Current sd08:15: sense key Data Protect kernel: Additional sense indicates Medium format corrupted scsi id 4 lun 0 is the optical disk (Panasonic PD drive) but Ext3 fs reports on mounting the drive that the disk is sd08:21 and not sd08:15 as above Using copydir is the only time I ever have problems with this disk/drive in practice. There have also been a couple of, perhaps unrelated, oddities: kernel: spurious 8259A interrupt: IRQ7. I also have frequently seen this: kernel: end_request: I/O error, dev 02:00 (floppy), sector 0 It seems odd that the kernel should generate an error when there is no disk in the drive and it isn't intended that there should be. All sorts of other things don't happen that don't generate errors :) I'm using Mandrake 8.1, with a recompiled 2.4.8-34.1 - Richard. -- Richard Kimber Political Science Resources: http://www.psr.keele.ac.uk/
