SATA oddities about linux-sunxi HOT 6 CLOSED

I haven't figured out how to tag the bug yet, but this is all with the v1 kernel.

from linux-sunxi.

can you try in 3.0-v2? .... and a different sata cable?

from linux-sunxi.

I've been running an overnight kernel build on v2, using a SATA root filesystem, and nothing's appeared. (Total uptime 14 hours.) I haven't tried with a different cable yet.

Could PSU issues be the culprit for this?

from linux-sunxi.

no errors so far?

from linux-sunxi.

Seems solid on v2. I'll soak test it for a while and if nothing shows up close the bug. shrug

from linux-sunxi.

After running debsums for a bunch of hours with the v2 kernel I've had a total of four errors, all looking like this:

[42600.830000] ata1.00: exception Emask 0x10 SAct 0x1 SErr 0x200100 action 0x6 frozen
[42600.840000] ata1.00: irq_stat 0x08000000, interface fatal error
[42600.840000] ata1: SError: { UnrecovData BadCRC }
[42600.850000] ata1.00: failed command: READ FPDMA QUEUED
[42600.850000] ata1.00: cmd 60/00:00:e0:e5:6a/01:00:00:00:00/40 tag 0 ncq 131072 in
[42600.850000] res 40/00:04:e0:e5:6a/00:00:00:00:00/40 Emask 0x10 (ATA bus error)
[42600.870000] ata1.00: status: { DRDY }
[42600.870000] ata1: hard resetting link
[42601.470000] ata1: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
[42601.810000] ata1.00: configured for UDMA/100
[42601.810000] ata1: EH complete

This is apparently an absolute dead-on classic bad cable/power message, as you suggested:

http://lime-technology.com/wiki/index.php/The_Analysis_of_Drive_Issues#Drive_interface_issue_.231

This is sounding much less like a kernel issue, so I'll close. Thanks very much.

from linux-sunxi.