[PATCH] libata: Delegate nonrot flag setting to SCSI

[PATCH] libata: Delegate nonrot flag setting to SCSI

[Martin K. Petersen]

Now that the SCSI disk driver correctly handles non-rotational devices
we can move setting the queue flag to SCSI.

Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

---

diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
--- a/drivers/ata/libata-scsi.c
+++ b/drivers/ata/libata-scsi.c
@@ -1119,10 +1119,6 @@ static int ata_scsi_dev_config(struct sc
 
 		blk_queue_dma_drain(q, atapi_drain_needed, buf, ATAPI_MAX_DRAIN);
 	} else {
-		if (ata_id_is_ssd(dev->id))
-			queue_flag_set_unlocked(QUEUE_FLAG_NONROT,
-						sdev->request_queue);
-
 		/* ATA devices must be sector aligned */
 		blk_queue_update_dma_alignment(sdev->request_queue,
 					       ATA_SECT_SIZE - 1);

Re: [PATCH] libata: Delegate nonrot flag setting to SCSI

[Jeff Garzik]

Martin K. Petersen wrote:
> Now that the SCSI disk driver correctly handles non-rotational devices
> we can move setting the queue flag to SCSI.
> 
> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
> 
> ---
> 
> diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
> --- a/drivers/ata/libata-scsi.c
> +++ b/drivers/ata/libata-scsi.c
> @@ -1119,10 +1119,6 @@ static int ata_scsi_dev_config(struct sc
>  
>  		blk_queue_dma_drain(q, atapi_drain_needed, buf, ATAPI_MAX_DRAIN);
>  	} else {
> -		if (ata_id_is_ssd(dev->id))
> -			queue_flag_set_unlocked(QUEUE_FLAG_NONROT,
> -						sdev->request_queue);
> -

Does the existence of this code cause any problems or regressions today?

It looks correct vis a vis current drivers/scsi/sd.c, but given that the 
merge window just closed, I lean towards classifying this as a cleanup 
(thus, 2.6.32).

	Jeff

Re: [PATCH] libata: Delegate nonrot flag setting to SCSI

[Martin K. Petersen]

>>>>> "Jeff" == Jeff Garzik <jeff@garzik.org> writes:

Jeff> Does the existence of this code cause any problems or regressions
Jeff> today?

Nope.  We set the flag twice, that's all.


Jeff> It looks correct vis a vis current drivers/scsi/sd.c, but given
Jeff> that the merge window just closed, I lean towards classifying this
Jeff> as a cleanup (thus, 2.6.32).

Up to you.  I obviously couldn't submit this patch until the relevant
SCSI bits hit Linus' tree and that didn't happen until yesterday.

-- 
Martin K. Petersen	Oracle Linux Engineering

Re: [PATCH] libata: Delegate nonrot flag setting to SCSI

[Jeff Garzik]

Martin K. Petersen wrote:
> Now that the SCSI disk driver correctly handles non-rotational devices
> we can move setting the queue flag to SCSI.
> 
> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
> 
> ---
> 
> diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
> --- a/drivers/ata/libata-scsi.c
> +++ b/drivers/ata/libata-scsi.c
> @@ -1119,10 +1119,6 @@ static int ata_scsi_dev_config(struct sc
>  
>  		blk_queue_dma_drain(q, atapi_drain_needed, buf, ATAPI_MAX_DRAIN);
>  	} else {
> -		if (ata_id_is_ssd(dev->id))
> -			queue_flag_set_unlocked(QUEUE_FLAG_NONROT,
> -						sdev->request_queue);
> -
>  		/* ATA devices must be sector aligned */
>  		blk_queue_update_dma_alignment(sdev->request_queue,

applied