On RHEL5 (kernel-2.6.18-8.1.1.el5) I've got this error message: ata7: translated ATA stat/err 0x51/0c to SCSI SK/ASC/ASCQ 0xb/00/00 ata7: status=0x51 { DriveReady SeekComplete Error } ata7: error=0x0c { DriveStatusError } BUG: warning at drivers/scsi/sata_mv.c:651/mv_start_dma() (Not tainted) [<f8889dcd>] mv_qc_issue+0xd7/0xfe [sata_mv] [<f88fc799>] ata_qc_issue+0x416/0x46c [libata] [<f88d94f7>] scsi_done+0x0/0x16 [scsi_mod] [<f8900a12>] ata_scsi_translate+0xaf/0xeb [libata] [<f88d94f7>] scsi_done+0x0/0x16 [scsi_mod] [<f8901cea>] ata_scsi_queuecmd+0xec/0x109 [libata] [<f890145d>] ata_scsi_rw_xlat+0x0/0x3cd [libata] [<f88d9a80>] scsi_dispatch_cmd+0x208/0x281 [scsi_mod] [<f88dde71>] scsi_request_fn+0x247/0x2f9 [scsi_mod] [<c04d1ec3>] blk_run_queue+0x37/0x63 [<f88dd605>] scsi_queue_insert+0x8b/0x93 [scsi_mod] [<c04d289e>] blk_done_softirq+0x4d/0x58 [<c04281cf>] __do_softirq+0x5a/0xbb [<c0406461>] do_softirq+0x52/0x9d [<c0406406>] do_IRQ+0xa5/0xae [<c040492e>] common_interrupt+0x1a/0x20 [<c0402ccf>] mwait_idle+0x25/0x38 [<c0402c90>] cpu_idle+0x9f/0xb9 This seems like disc problem so this may be only kernel noise.
There is no EH (Error Handling) in sata_mv driver and I had HW issues with one of my WD4000KD. A message from Tejun Heo (see http://lwn.net/Articles/226256/ about incorporating new EH to the latest kernel): Okay, it's sata_mv. sata_mv is just now receiving new error handler. The old EH path is almost non-existent and bound to hang the machine when IO error occurs. So, you can... 1. Wait for sata_mv to get new EH. I dunno how long it would take to propagate through distribution but maybe in several months from now. 2. Go out and grab a SATA controller with working EH. So I wish new libata in upcoming RHEL 5.1... (is it planned?) Please close this bug as DEFFERED as I'm unable to do so even I'm a submiter.
Closing since error handling has been updated in RHEL5.