[PATCH 0/2] Fixes for ATA device queue depth control

[PATCH 0/2] Fixes for ATA device queue depth control

[Damien Le Moal]

A couple of patches to fix maximum queue depth initialization and
control for ATA devices used through libsas.

Damien Le Moal (2):
  ata: libata-scsi: Fix initialization of device queue depth
  ata: libata-sata: Fix device queue depth control

 drivers/ata/libata-sata.c           | 24 ++++++++++++------------
 drivers/ata/libata-scsi.c           | 13 ++++++++-----
 drivers/scsi/libsas/sas_scsi_host.c |  3 ++-
 include/linux/libata.h              |  4 ++--
 4 files changed, 24 insertions(+), 20 deletions(-)

-- 
2.37.3

[PATCH 1/2] ata: libata-scsi: Fix initialization of device queue depth

[Damien Le Moal]

For SATA devices supporting NCQ, drivers using libsas first initialize a
scsi device queue depth based on the controller and device capabilities,
leading to the scsi device queue_depth field being 32 (ATA maximum queue
depth) for most setup. However, if libata was loaded using the
force=[ID]]noncq argument, the default queue depth should be set to 1 to
reflect the fact that queuable commands will never be used. This is
consistent with manually setting a device queue depth to 1 through sysfs
as that disables NCQ use for the device.

Fix ata_scsi_dev_config() to honor the noncq parameter by sertting the
device queue depth to 1 for devices that do not have the ATA_DFLAG_NCQ
flag set.

Signed-off-by: Damien Le Moal <damien.lemoal@opensource.wdc.com>
---
 drivers/ata/libata-scsi.c | 10 ++++------
 1 file changed, 4 insertions(+), 6 deletions(-)

diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
index 29e2f55c6faa..ff9602a0e54e 100644
--- a/drivers/ata/libata-scsi.c
+++ b/drivers/ata/libata-scsi.c
@@ -1055,6 +1055,7 @@ EXPORT_SYMBOL_GPL(ata_scsi_dma_need_drain);
 int ata_scsi_dev_config(struct scsi_device *sdev, struct ata_device *dev)
 {
 	struct request_queue *q = sdev->request_queue;
+	int depth = 1;
 
 	if (!ata_id_has_unload(dev->id))
 		dev->flags |= ATA_DFLAG_NO_UNLOAD;
@@ -1100,13 +1101,10 @@ int ata_scsi_dev_config(struct scsi_device *sdev, struct ata_device *dev)
 	if (dev->flags & ATA_DFLAG_AN)
 		set_bit(SDEV_EVT_MEDIA_CHANGE, sdev->supported_events);
 
-	if (dev->flags & ATA_DFLAG_NCQ) {
-		int depth;
-
+	if (dev->flags & ATA_DFLAG_NCQ)
 		depth = min(sdev->host->can_queue, ata_id_queue_depth(dev->id));
-		depth = min(ATA_MAX_QUEUE, depth);
-		scsi_change_queue_depth(sdev, depth);
-	}
+	depth = min(ATA_MAX_QUEUE, depth);
+	scsi_change_queue_depth(sdev, depth);
 
 	if (dev->flags & ATA_DFLAG_TRUSTED)
 		sdev->security_supported = 1;
-- 
2.37.3

[PATCH 2/2] ata: libata-sata: Fix device queue depth control

[Damien Le Moal]

The function __ata_change_queue_depth() uses the helper
ata_scsi_find_dev() to get the ata_device structure of a scsi device and
set that device maximum queue depth. However, when the ata device is
managed by libsas, ata_scsi_find_dev() return NULL, turning
__ata_change_queue_depth() into a nop, which prevents the user from
setting the maximum queue depth of ATA devices used with libsas based
HBAs.

Fix this by renaming __ata_change_queue_depth() to
ata_change_queue_depth() and adding a pointer to the ata_device
structure of the target device as argument. This pointer is provided by
ata_scsi_change_queue_depth() using ata_scsi_find_dev() in the case ofi
a libata managed device and by sas_change_queue_depth() using
sas_to_ata_dev() in the case of a libsas managed ata device.

Signed-off-by: Damien Le Moal <damien.lemoal@opensource.wdc.com>
---
 drivers/ata/libata-sata.c           | 24 ++++++++++++------------
 drivers/ata/libata-scsi.c           |  5 +++++
 drivers/scsi/libsas/sas_scsi_host.c |  3 ++-
 include/linux/libata.h              |  4 ++--
 4 files changed, 21 insertions(+), 15 deletions(-)

diff --git a/drivers/ata/libata-sata.c b/drivers/ata/libata-sata.c
index 7a5fe41aa5ae..13b9d0fdd42c 100644
--- a/drivers/ata/libata-sata.c
+++ b/drivers/ata/libata-sata.c
@@ -1018,26 +1018,25 @@ DEVICE_ATTR(sw_activity, S_IWUSR | S_IRUGO, ata_scsi_activity_show,
 EXPORT_SYMBOL_GPL(dev_attr_sw_activity);
 
 /**
- *	__ata_change_queue_depth - helper for ata_scsi_change_queue_depth
- *	@ap: ATA port to which the device change the queue depth
+ *	ata_change_queue_depth - Set a device maximum queue depth
+ *	@ap: ATA port of the target device
+ *	@dev: target ATA device
  *	@sdev: SCSI device to configure queue depth for
  *	@queue_depth: new queue depth
  *
- *	libsas and libata have different approaches for associating a sdev to
- *	its ata_port.
+ *	Helper to set a device maximum queue depth, usable with both libsas
+ *	and libata.
  *
  */
-int __ata_change_queue_depth(struct ata_port *ap, struct scsi_device *sdev,
-			     int queue_depth)
+int ata_change_queue_depth(struct ata_port *ap, struct ata_device *dev,
+			   struct scsi_device *sdev, int queue_depth)
 {
-	struct ata_device *dev;
 	unsigned long flags;
 
-	if (queue_depth < 1 || queue_depth == sdev->queue_depth)
+	if (!dev || !ata_dev_enabled(dev))
 		return sdev->queue_depth;
 
-	dev = ata_scsi_find_dev(ap, sdev);
-	if (!dev || !ata_dev_enabled(dev))
+	if (queue_depth < 1 || queue_depth == sdev->queue_depth)
 		return sdev->queue_depth;
 
 	/* NCQ enabled? */
@@ -1059,7 +1058,7 @@ int __ata_change_queue_depth(struct ata_port *ap, struct scsi_device *sdev,
 
 	return scsi_change_queue_depth(sdev, queue_depth);
 }
-EXPORT_SYMBOL_GPL(__ata_change_queue_depth);
+EXPORT_SYMBOL_GPL(ata_change_queue_depth);
 
 /**
  *	ata_scsi_change_queue_depth - SCSI callback for queue depth config
@@ -1080,7 +1079,8 @@ int ata_scsi_change_queue_depth(struct scsi_device *sdev, int queue_depth)
 {
 	struct ata_port *ap = ata_shost_to_port(sdev->host);
 
-	return __ata_change_queue_depth(ap, sdev, queue_depth);
+	return ata_change_queue_depth(ap, ata_scsi_find_dev(ap, sdev),
+				      sdev, queue_depth);
 }
 EXPORT_SYMBOL_GPL(ata_scsi_change_queue_depth);
 
diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
index ff9602a0e54e..c63bb50323c1 100644
--- a/drivers/ata/libata-scsi.c
+++ b/drivers/ata/libata-scsi.c
@@ -1101,6 +1101,11 @@ int ata_scsi_dev_config(struct scsi_device *sdev, struct ata_device *dev)
 	if (dev->flags & ATA_DFLAG_AN)
 		set_bit(SDEV_EVT_MEDIA_CHANGE, sdev->supported_events);
 
+	pr_info("#### can queue %d, ata qd %d, scsi qd %d\n",
+		sdev->host->can_queue,
+		ata_id_queue_depth(dev->id),
+		sdev->queue_depth);
+
 	if (dev->flags & ATA_DFLAG_NCQ)
 		depth = min(sdev->host->can_queue, ata_id_queue_depth(dev->id));
 	depth = min(ATA_MAX_QUEUE, depth);
diff --git a/drivers/scsi/libsas/sas_scsi_host.c b/drivers/scsi/libsas/sas_scsi_host.c
index 9c82e5dc4fcc..a36fa1c128a8 100644
--- a/drivers/scsi/libsas/sas_scsi_host.c
+++ b/drivers/scsi/libsas/sas_scsi_host.c
@@ -872,7 +872,8 @@ int sas_change_queue_depth(struct scsi_device *sdev, int depth)
 	struct domain_device *dev = sdev_to_domain_dev(sdev);
 
 	if (dev_is_sata(dev))
-		return __ata_change_queue_depth(dev->sata_dev.ap, sdev, depth);
+		return ata_change_queue_depth(dev->sata_dev.ap,
+					      sas_to_ata_dev(dev), sdev, depth);
 
 	if (!sdev->tagged_supported)
 		depth = 1;
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 698032e5ef2d..20765d1c5f80 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -1136,8 +1136,8 @@ extern int ata_scsi_slave_config(struct scsi_device *sdev);
 extern void ata_scsi_slave_destroy(struct scsi_device *sdev);
 extern int ata_scsi_change_queue_depth(struct scsi_device *sdev,
 				       int queue_depth);
-extern int __ata_change_queue_depth(struct ata_port *ap, struct scsi_device *sdev,
-				    int queue_depth);
+extern int ata_change_queue_depth(struct ata_port *ap, struct ata_device *dev,
+				  struct scsi_device *sdev, int queue_depth);
 extern struct ata_device *ata_dev_pair(struct ata_device *adev);
 extern int ata_do_set_mode(struct ata_link *link, struct ata_device **r_failed_dev);
 extern void ata_scsi_port_error_handler(struct Scsi_Host *host, struct ata_port *ap);
-- 
2.37.3

Re: [PATCH 2/2] ata: libata-sata: Fix device queue depth control

[Sergey Shtylyov]

Hello!

On 9/24/22 9:29 AM, Damien Le Moal wrote:

> The function __ata_change_queue_depth() uses the helper
> ata_scsi_find_dev() to get the ata_device structure of a scsi device and
> set that device maximum queue depth. However, when the ata device is
> managed by libsas, ata_scsi_find_dev() return NULL, turning

   Returns?

> __ata_change_queue_depth() into a nop, which prevents the user from
> setting the maximum queue depth of ATA devices used with libsas based
> HBAs.
> 
> Fix this by renaming __ata_change_queue_depth() to
> ata_change_queue_depth() and adding a pointer to the ata_device
> structure of the target device as argument. This pointer is provided by
> ata_scsi_change_queue_depth() using ata_scsi_find_dev() in the case ofi

   Of?

> a libata managed device and by sas_change_queue_depth() using
> sas_to_ata_dev() in the case of a libsas managed ata device.
> 
> Signed-off-by: Damien Le Moal <damien.lemoal@opensource.wdc.com>
[...]

> diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
> index ff9602a0e54e..c63bb50323c1 100644
> --- a/drivers/ata/libata-scsi.c
> +++ b/drivers/ata/libata-scsi.c
> @@ -1101,6 +1101,11 @@ int ata_scsi_dev_config(struct scsi_device *sdev, struct ata_device *dev)
>  	if (dev->flags & ATA_DFLAG_AN)
>  		set_bit(SDEV_EVT_MEDIA_CHANGE, sdev->supported_events);
>  
> +	pr_info("#### can queue %d, ata qd %d, scsi qd %d\n",

    Hm... isn't this a debugging pr_info() call you forgot to remove?

> +		sdev->host->can_queue,
> +		ata_id_queue_depth(dev->id),
> +		sdev->queue_depth);
> +
>  	if (dev->flags & ATA_DFLAG_NCQ)
>  		depth = min(sdev->host->can_queue, ata_id_queue_depth(dev->id));
>  	depth = min(ATA_MAX_QUEUE, depth);
[...]

MBR, Sergey

Re: [PATCH 2/2] ata: libata-sata: Fix device queue depth control

[Damien Le Moal]

On 9/24/22 18:49, Sergey Shtylyov wrote:
> Hello!
> 
> On 9/24/22 9:29 AM, Damien Le Moal wrote:
> 
>> The function __ata_change_queue_depth() uses the helper
>> ata_scsi_find_dev() to get the ata_device structure of a scsi device and
>> set that device maximum queue depth. However, when the ata device is
>> managed by libsas, ata_scsi_find_dev() return NULL, turning
> 
>    Returns?
> 
>> __ata_change_queue_depth() into a nop, which prevents the user from
>> setting the maximum queue depth of ATA devices used with libsas based
>> HBAs.
>>
>> Fix this by renaming __ata_change_queue_depth() to
>> ata_change_queue_depth() and adding a pointer to the ata_device
>> structure of the target device as argument. This pointer is provided by
>> ata_scsi_change_queue_depth() using ata_scsi_find_dev() in the case ofi
> 
>    Of?
> 
>> a libata managed device and by sas_change_queue_depth() using
>> sas_to_ata_dev() in the case of a libsas managed ata device.
>>
>> Signed-off-by: Damien Le Moal <damien.lemoal@opensource.wdc.com>
> [...]
> 
>> diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
>> index ff9602a0e54e..c63bb50323c1 100644
>> --- a/drivers/ata/libata-scsi.c
>> +++ b/drivers/ata/libata-scsi.c
>> @@ -1101,6 +1101,11 @@ int ata_scsi_dev_config(struct scsi_device *sdev, struct ata_device *dev)
>>  	if (dev->flags & ATA_DFLAG_AN)
>>  		set_bit(SDEV_EVT_MEDIA_CHANGE, sdev->supported_events);
>>  
>> +	pr_info("#### can queue %d, ata qd %d, scsi qd %d\n",
> 
>     Hm... isn't this a debugging pr_info() call you forgot to remove?

Oops. Yes. Sending an update. Thanks for the review.

> 
>> +		sdev->host->can_queue,
>> +		ata_id_queue_depth(dev->id),
>> +		sdev->queue_depth);
>> +
>>  	if (dev->flags & ATA_DFLAG_NCQ)
>>  		depth = min(sdev->host->can_queue, ata_id_queue_depth(dev->id));
>>  	depth = min(ATA_MAX_QUEUE, depth);
> [...]
> 
> MBR, Sergey

-- 
Damien Le Moal
Western Digital Research