[SCSI] isci: fix needless ata reset escalations

isci is needlessly tying libata's hands by returning
SAM_STAT_CHECK_CONDITION to some ata errors.  Instead, prefer
SAS_PROTO_RESPONSE to let libata (via sas_ata_task_done()) disposition
the device-to-host fis.

For example isci is triggering an HSM Violation where AHCI is showing a
simple media error for the same bus condition:

isci:
ata7.00: failed command: READ VERIFY SECTOR(S)
ata7.00: cmd 40/00:01:00:00:00/00:00:00:00:00/e0 tag 0
         res 01/04:00:00:00:00/00:00:00:00:00/e0 Emask 0x3 (HSM violation)

ahci:
ata6.00: failed command: READ VERIFY SECTOR(S)
ata6.00: cmd 40/00:01:00:00:00/00:00:00:00:00/e0 tag 0
         res 51/40:01:00:00:00/00:00:00:00:00/e0 Emask 0x9 (media error)

Note that the isci response matches this from sas_ata_task_done():
	/* We saw a SAS error. Send a vague error. */
	[..]
	dev->sata_dev.fis[3] = 0x04; /* status err */
	dev->sata_dev.fis[2] = ATA_ERR;

The end effect is that isci is needlessly triggering hard resets when
they are not necessary.

Reported-by: Xun Ni <xun.ni@intel.com>
Tested-by: Nelson Cheng <nelson.cheng@intel.com>
Acked-by: Lukasz Dorau <lukasz.dorau@intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
This commit is contained in:
Dan Williams 2014-02-06 12:23:15 -08:00 committed by James Bottomley
parent 3af74a3c07
commit 2193b1b16c

View file

@ -2723,13 +2723,9 @@ static void isci_process_stp_response(struct sas_task *task, struct dev_to_host_
memcpy(resp->ending_fis, fis, sizeof(*fis));
ts->buf_valid_size = sizeof(*resp);
/* If the device fault bit is set in the status register, then
* set the sense data and return.
*/
if (fis->status & ATA_DF)
/* If an error is flagged let libata decode the fis */
if (ac_err_mask(fis->status))
ts->stat = SAS_PROTO_RESPONSE;
else if (fis->status & ATA_ERR)
ts->stat = SAM_STAT_CHECK_CONDITION;
else
ts->stat = SAM_STAT_GOOD;