Use log directory to check for old error/self-test log support

[Christian Franke]

Smartctl and smartd assume that the old SMART error log (log address 0x01) is supported if an error-logging capability bit in IDENTIFY or SMART data is set. Same applies to old SMART self-test log (log address 0x06). These logs were introduced in ATA-5 and are now outdated because 48-bit LBA addresses are not supported. Recent devices may only support the newer 48-bit capable logs (extended comprehensive error-log, extended self-test log) introduced in ATA-6.

smartctl -a ... prints misleading messages then:

Error SMART Error Log Read failed: Input/output error
Smartctl: SMART Error Log Read Failed
Error SMART Error Self-Test Log Read failed: Input/output error
Smartctl: SMART Self Test Log Read Failed

Misleading kernel messages may appear during smartd startup:

... kernel: hda: drive_cmd: status=0x51 { DriveReady SeekComplete Error }
... kernel: hda: drive_cmd: error=0x04 { DriveStatusError }
... kernel: ide: failed opcode was: 0xb0
... smartd[...]: Device: /dev/hda, Read SMART Self Test Log Failed
... smartd[...]: Device: /dev/hda, no SMART Self-Test log; remove -l selftest Directive from smartd.conf

(smartctl -x ... is not affected)

Smartctl and smartd should read the log directory (log address 0x00) to check whether the old logs are still supported and print an "unsupported" message if not.

[Christian Franke]

r3239.