Re: libata patches and log spam?

[Chris Boot]

Chris Boot wrote:

> Hi all,
>
> I've started using Jeff's libata patches from 
> http://www.kernel.org/pub/linux/kernel/people/jgarzik/libata/, mostly 
> so I can check the SMART status of a few drives that seem to be on 
> their way out. I've only just noticed that recently, my kernel has 
> been spewing out messages like the following every minute or so:
>
> Sep 18 19:34:55 [kernel] [4376428.164000] ata1: translated ATA 
> stat/err 0xb0/00 to SCSI SK/ASC/ASCQ 0xb/47/00
> Sep 18 19:34:55 [kernel] [4376428.216000] ata1: translated ATA 
> stat/err 0xb0/00 to SCSI SK/ASC/ASCQ 0xb/47/00
> Sep 18 19:34:55 [kernel] [4376428.216000] ata2: translated ATA 
> stat/err 0xb0/00 to SCSI SK/ASC/ASCQ 0xb/47/00
> Sep 18 19:34:55 [kernel] [4376428.268000] ata2: translated ATA 
> stat/err 0xb0/00 to SCSI SK/ASC/ASCQ 0xb/47/00
> Sep 18 19:34:55 [kernel] [4376428.268000] ata3: translated ATA 
> stat/err 0xb0/00 to SCSI SK/ASC/ASCQ 0xb/47/00
> Sep 18 19:34:55 [kernel] [4376428.319000] ata3: translated ATA 
> stat/err 0xb0/00 to SCSI SK/ASC/ASCQ 0xb/47/00
> Sep 18 19:34:55 [kernel] [4376428.320000] ata4: translated ATA 
> stat/err 0xb0/00 to SCSI SK/ASC/ASCQ 0xb/47/00
> Sep 18 19:34:55 [kernel] [4376428.374000] ata4: translated ATA 
> stat/err 0xb0/00 to SCSI SK/ASC/ASCQ 0xb/47/00
> Sep 18 19:36:00 [kernel] [4376493.196000] ata1: translated ATA 
> stat/err 0xb0/00 to SCSI SK/ASC/ASCQ 0xb/47/00
> Sep 18 19:36:00 [kernel] [4376493.266000] ata1: translated ATA 
> stat/err 0xb0/00 to SCSI SK/ASC/ASCQ 0xb/47/00
> Sep 18 19:36:00 [kernel] [4376493.273000] ata2: translated ATA 
> stat/err 0xb0/00 to SCSI SK/ASC/ASCQ 0xb/47/00
> Sep 18 19:36:00 [kernel] [4376493.324000] ata2: translated ATA 
> stat/err 0xb0/00 to SCSI SK/ASC/ASCQ 0xb/47/00
> Sep 18 19:36:00 [kernel] [4376493.327000] ata3: translated ATA 
> stat/err 0xb0/00 to SCSI SK/ASC/ASCQ 0xb/47/00
> Sep 18 19:36:00 [kernel] [4376493.378000] ata3: translated ATA 
> stat/err 0xb0/00 to SCSI SK/ASC/ASCQ 0xb/47/00
> Sep 18 19:36:00 [kernel] [4376493.388000] ata4: translated ATA 
> stat/err 0xb0/00 to SCSI SK/ASC/ASCQ 0xb/47/00
> Sep 18 19:36:00 [kernel] [4376493.440000] ata4: translated ATA 
> stat/err 0xb0/00 to SCSI SK/ASC/ASCQ 0xb/47/00
>
> These messages don't seem to cause any trouble and I don't get any 
> other errors. Is this just because of increased verbosity due to the 
> patches? Can these messages be safely ignored / nuked?
>
> I'm using 2.6.13-ck5 + 2.6.13-rc7-libata1.patch + 
> reiser4-for-2.6.12-3.patch.gz + 
> vesafb-tng-0.9-rc7-r1-2.6.13-rc6.patch. ata{1,2} are on a sata_sil 
> controller, ata{3,4} are on sata_via.
>
> It also has to be noted that I get similar errors when I try to 
> enable/disable SMART automatic offline tests (and other similar 
> settings) on a drive:
>
> [4377001.595000] ata1: PIO error, drv_stat 0x51
> [4377001.595000] ata1: translated ATA stat/err 0xb0/00 to SCSI 
> SK/ASC/ASCQ 0xb/47/00
> [4377001.595000] ata1: status=0xb0 { Busy }
> [4377001.596000] ata1: PIO error, drv_stat 0x51
> [4377001.596000] ata1: translated ATA stat/err 0xb0/00 to SCSI 
> SK/ASC/ASCQ 0xb/47/00
> [4377001.596000] ata1: status=0xb0 { Busy }
> [4377001.596000] ata1: PIO error, drv_stat 0x51
> [4377001.596000] ata1: translated ATA stat/err 0xb0/00 to SCSI 
> SK/ASC/ASCQ 0xb/47/00
> [4377001.596000] ata1: status=0xb0 { Busy }
> [4377001.597000] ata1: PIO error, drv_stat 0x51
> [4377001.597000] ata1: translated ATA stat/err 0xb0/00 to SCSI 
> SK/ASC/ASCQ 0xb/47/00
> [4377001.597000] ata1: status=0xb0 { Busy }
> [4377001.597000] ata1: PIO error, drv_stat 0x51
> [4377001.597000] ata1: translated ATA stat/err 0xb0/00 to SCSI 
> SK/ASC/ASCQ 0xb/47/00
> [4377001.597000] ata1: status=0xb0 { Busy }
>
> I'm getting similar messages when playing with SMART on another 
> machine with vanilla 2.6.13.1 + 2.6.13-rc7-libata1.patch, but not the 
> regular, minute-by-minute messages on that machine. This one has two 
> drives on a sata_piix controller.
>
> Other than that, no trouble. Thanks for the patches, they're really 
> handy! :-)
>
> PS: I'm willing to apply patches to, and reboot at will, the first 
> machine, but not the second since it's my production web server.
Aha! That would have been hddtemp scanning the temperature every minute 
or so. Now, how come these messages appear in the first place? Needless 
to say the temperature is retrieved correctly.

Thanks,
Chris

--
Chris Boot
bootc@xxxxxxxxx
http://www.bootc.net/

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/