[PATCH 0/6] scsi: libsas: Some coding style fixes and cleanups

[PATCH 0/6] scsi: libsas: Some coding style fixes and cleanups

[Jason Yan]

A few coding style fixes and cleanups. There should be no functional
changes in this series besides the debug log prints.

Jason Yan (6):
  scsi: libsas: move sas_get_ata_command_set() up to save the
    declaration
  scsi: libsas: delete wrapper function sas_discover_end_dev()
  scsi: libsas: rename sas_discover_sata() and related refactors
  scsi: libsas: remove useless dev_list delete in
    sas_ex_discover_end_dev()
  scsi: libsas: factor out sas_ata_add_dev()
  scsi: libsas: factor out sas_ex_add_dev()

 drivers/scsi/libsas/sas_ata.c      |  94 ++++++++++++++++++----
 drivers/scsi/libsas/sas_discover.c |  21 +----
 drivers/scsi/libsas/sas_expander.c | 125 ++++++++++-------------------
 include/scsi/libsas.h              |   3 -
 include/scsi/sas_ata.h             |  15 ++++
 5 files changed, 135 insertions(+), 123 deletions(-)

-- 
2.31.1

[PATCH 1/6] scsi: libsas: move sas_get_ata_command_set() up to save the declaration

[Jason Yan]

There is a sas_get_ata_command_set() declaration above sas_get_ata_info()
to make it compile ok. However this function is defined in the same file
below. So move it up to save the declaration.

Cc: John Garry <john.g.garry@oracle.com>
Signed-off-by: Jason Yan <yanaijie@huawei.com>
---
 drivers/scsi/libsas/sas_ata.c | 28 +++++++++++++---------------
 1 file changed, 13 insertions(+), 15 deletions(-)

diff --git a/drivers/scsi/libsas/sas_ata.c b/drivers/scsi/libsas/sas_ata.c
index f7439bf9cdc6..34009c330eb2 100644
--- a/drivers/scsi/libsas/sas_ata.c
+++ b/drivers/scsi/libsas/sas_ata.c
@@ -239,7 +239,19 @@ static struct sas_internal *dev_to_sas_internal(struct domain_device *dev)
 	return to_sas_internal(dev->port->ha->core.shost->transportt);
 }
 
-static int sas_get_ata_command_set(struct domain_device *dev);
+static int sas_get_ata_command_set(struct domain_device *dev)
+{
+	struct dev_to_host_fis *fis =
+		(struct dev_to_host_fis *) dev->frame_rcvd;
+	struct ata_taskfile tf;
+
+	if (dev->dev_type == SAS_SATA_PENDING)
+		return ATA_DEV_UNKNOWN;
+
+	ata_tf_from_fis((const u8 *)fis, &tf);
+
+	return ata_dev_classify(&tf);
+}
 
 int sas_get_ata_info(struct domain_device *dev, struct ex_phy *phy)
 {
@@ -637,20 +649,6 @@ void sas_ata_task_abort(struct sas_task *task)
 	complete(waiting);
 }
 
-static int sas_get_ata_command_set(struct domain_device *dev)
-{
-	struct dev_to_host_fis *fis =
-		(struct dev_to_host_fis *) dev->frame_rcvd;
-	struct ata_taskfile tf;
-
-	if (dev->dev_type == SAS_SATA_PENDING)
-		return ATA_DEV_UNKNOWN;
-
-	ata_tf_from_fis((const u8 *)fis, &tf);
-
-	return ata_dev_classify(&tf);
-}
-
 void sas_probe_sata(struct asd_sas_port *port)
 {
 	struct domain_device *dev, *n;
-- 
2.31.1

Re: [PATCH 1/6] scsi: libsas: move sas_get_ata_command_set() up to save the declaration

[John Garry]

On 04/12/2022 08:16, Jason Yan wrote:
> There is a sas_get_ata_command_set() declaration above sas_get_ata_info()
> to make it compile ok. However this function is defined in the same file
> below. So move it up to save the declaration.
> 
> Cc: John Garry <john.g.garry@oracle.com>
> Signed-off-by: Jason Yan <yanaijie@huawei.com>

Apart from comments, below:
Reviewed-by: John Garry <john.g.garry@oracle.com>

> ---
>   drivers/scsi/libsas/sas_ata.c | 28 +++++++++++++---------------
>   1 file changed, 13 insertions(+), 15 deletions(-)
> 
> diff --git a/drivers/scsi/libsas/sas_ata.c b/drivers/scsi/libsas/sas_ata.c
> index f7439bf9cdc6..34009c330eb2 100644
> --- a/drivers/scsi/libsas/sas_ata.c
> +++ b/drivers/scsi/libsas/sas_ata.c
> @@ -239,7 +239,19 @@ static struct sas_internal *dev_to_sas_internal(struct domain_device *dev)
>   	return to_sas_internal(dev->port->ha->core.shost->transportt);
>   }
>   
> -static int sas_get_ata_command_set(struct domain_device *dev);
> +static int sas_get_ata_command_set(struct domain_device *dev)
> +{
> +	struct dev_to_host_fis *fis =
> +		(struct dev_to_host_fis *) dev->frame_rcvd;

nit: I did not think that we add a whitespace before casting. And I 
think that it would be neater if fis was assigned separately, avoiding 
overflowing the line in this way.

Having said that, it does seem odd to cast from u8 [] to struct 
dev_to_host_fis * and then back to (u8 *) in the ata_tf_from_fis() call, 
below.


> +	struct ata_taskfile tf;
> +
> +	if (dev->dev_type == SAS_SATA_PENDING)
> +		return ATA_DEV_UNKNOWN;
> +
> +	ata_tf_from_fis((const u8 *)fis, &tf);
> +
> +	return ata_dev_classify(&tf);
> +}
>   
>   int sas_get_ata_info(struct domain_device *dev, struct ex_phy *phy)
>   {
> @@ -637,20 +649,6 @@ void sas_ata_task_abort(struct sas_task *task)
>   	complete(waiting);
>   }
>   
> -static int sas_get_ata_command_set(struct domain_device *dev)
> -{
> -	struct dev_to_host_fis *fis =
> -		(struct dev_to_host_fis *) dev->frame_rcvd;
> -	struct ata_taskfile tf;
> -
> -	if (dev->dev_type == SAS_SATA_PENDING)
> -		return ATA_DEV_UNKNOWN;
> -
> -	ata_tf_from_fis((const u8 *)fis, &tf);
> -
> -	return ata_dev_classify(&tf);
> -}
> -
>   void sas_probe_sata(struct asd_sas_port *port)
>   {
>   	struct domain_device *dev, *n;

Re: [PATCH 1/6] scsi: libsas: move sas_get_ata_command_set() up to save the declaration

[Jason Yan]

On 2022/12/5 16:45, John Garry wrote:
> On 04/12/2022 08:16, Jason Yan wrote:
>> There is a sas_get_ata_command_set() declaration above sas_get_ata_info()
>> to make it compile ok. However this function is defined in the same file
>> below. So move it up to save the declaration.
>>
>> Cc: John Garry <john.g.garry@oracle.com>
>> Signed-off-by: Jason Yan <yanaijie@huawei.com>
> 
> Apart from comments, below:
> Reviewed-by: John Garry <john.g.garry@oracle.com>
> 

Thank you John.

>> ---
>>   drivers/scsi/libsas/sas_ata.c | 28 +++++++++++++---------------
>>   1 file changed, 13 insertions(+), 15 deletions(-)
>>
>> diff --git a/drivers/scsi/libsas/sas_ata.c 
>> b/drivers/scsi/libsas/sas_ata.c
>> index f7439bf9cdc6..34009c330eb2 100644
>> --- a/drivers/scsi/libsas/sas_ata.c
>> +++ b/drivers/scsi/libsas/sas_ata.c
>> @@ -239,7 +239,19 @@ static struct sas_internal 
>> *dev_to_sas_internal(struct domain_device *dev)
>>       return to_sas_internal(dev->port->ha->core.shost->transportt);
>>   }
>> -static int sas_get_ata_command_set(struct domain_device *dev);
>> +static int sas_get_ata_command_set(struct domain_device *dev)
>> +{
>> +    struct dev_to_host_fis *fis =
>> +        (struct dev_to_host_fis *) dev->frame_rcvd;
> 
> nit: I did not think that we add a whitespace before casting. And I 
> think that it would be neater if fis was assigned separately, avoiding 
> overflowing the line in this way.
> 
> Having said that, it does seem odd to cast from u8 [] to struct 
> dev_to_host_fis * and then back to (u8 *) in the ata_tf_from_fis() call, 
> below.

Yeah, odd. I'd prefer remove 'fis' and then do:

ata_tf_from_fis(dev->frame_rcvd, &tf).

Thanks,
Jason

[PATCH 2/6] scsi: libsas: delete wrapper function sas_discover_end_dev()

[Jason Yan]

After commit 0558f33c06bb ("scsi: libsas: direct call probe and destruct")
this function is only a wrapper of sas_notify_lldd_dev_found(). And the
function name does not reflect the real purpose of this function now.
Remove it and call sas_notify_lldd_dev_found() directly. The log is also
changed accordingly.

Cc: John Garry <john.g.garry@oracle.com>
Signed-off-by: Jason Yan <yanaijie@huawei.com>
---
 drivers/scsi/libsas/sas_discover.c | 13 +------------
 drivers/scsi/libsas/sas_expander.c |  4 ++--
 include/scsi/libsas.h              |  1 -
 3 files changed, 3 insertions(+), 15 deletions(-)

diff --git a/drivers/scsi/libsas/sas_discover.c b/drivers/scsi/libsas/sas_discover.c
index d5bc1314c341..efc6bf95bb67 100644
--- a/drivers/scsi/libsas/sas_discover.c
+++ b/drivers/scsi/libsas/sas_discover.c
@@ -269,17 +269,6 @@ static void sas_resume_devices(struct work_struct *work)
 	sas_resume_sata(port);
 }
 
-/**
- * sas_discover_end_dev - discover an end device (SSP, etc)
- * @dev: pointer to domain device of interest
- *
- * See comment in sas_discover_sata().
- */
-int sas_discover_end_dev(struct domain_device *dev)
-{
-	return sas_notify_lldd_dev_found(dev);
-}
-
 /* ---------- Device registration and unregistration ---------- */
 
 void sas_free_device(struct kref *kref)
@@ -447,7 +436,7 @@ static void sas_discover_domain(struct work_struct *work)
 
 	switch (dev->dev_type) {
 	case SAS_END_DEVICE:
-		error = sas_discover_end_dev(dev);
+		error = sas_notify_lldd_dev_found(dev);
 		break;
 	case SAS_EDGE_EXPANDER_DEVICE:
 	case SAS_FANOUT_EXPANDER_DEVICE:
diff --git a/drivers/scsi/libsas/sas_expander.c b/drivers/scsi/libsas/sas_expander.c
index a04cad620e93..aa8ea3b1f2e4 100644
--- a/drivers/scsi/libsas/sas_expander.c
+++ b/drivers/scsi/libsas/sas_expander.c
@@ -855,9 +855,9 @@ static struct domain_device *sas_ex_discover_end_dev(
 
 		list_add_tail(&child->disco_list_node, &parent->port->disco_list);
 
-		res = sas_discover_end_dev(child);
+		res = sas_notify_lldd_dev_found(child);
 		if (res) {
-			pr_notice("sas_discover_end_dev() for device %016llx at %016llx:%02d returned 0x%x\n",
+			pr_notice("notify lldd for device %016llx at %016llx:%02d returned 0x%x\n",
 				  SAS_ADDR(child->sas_addr),
 				  SAS_ADDR(parent->sas_addr), phy_id, res);
 			goto out_list_del;
diff --git a/include/scsi/libsas.h b/include/scsi/libsas.h
index 1aee3d0ebbb2..87682390fb76 100644
--- a/include/scsi/libsas.h
+++ b/include/scsi/libsas.h
@@ -736,7 +736,6 @@ void sas_init_disc(struct sas_discovery *disc, struct asd_sas_port *);
 void sas_discover_event(struct asd_sas_port *, enum discover_event ev);
 
 int  sas_discover_sata(struct domain_device *);
-int  sas_discover_end_dev(struct domain_device *);
 
 void sas_unregister_dev(struct asd_sas_port *port, struct domain_device *);
 
-- 
2.31.1

Re: [PATCH 2/6] scsi: libsas: delete wrapper function sas_discover_end_dev()

[John Garry]

On 04/12/2022 08:16, Jason Yan wrote:
> After commit 0558f33c06bb ("scsi: libsas: direct call probe and destruct")
> this function is only a wrapper of sas_notify_lldd_dev_found(). And the
> function name does not reflect the real purpose of this function now.

Why is this? Maybe add "dev_found" to the name could help.

> Remove it and call sas_notify_lldd_dev_found() directly. The log is also
> changed accordingly.
> 
> Cc: John Garry <john.g.garry@oracle.com>
> Signed-off-by: Jason Yan <yanaijie@huawei.com>
> ---
>   drivers/scsi/libsas/sas_discover.c | 13 +------------
>   drivers/scsi/libsas/sas_expander.c |  4 ++--
>   include/scsi/libsas.h              |  1 -
>   3 files changed, 3 insertions(+), 15 deletions(-)
> 
> diff --git a/drivers/scsi/libsas/sas_discover.c b/drivers/scsi/libsas/sas_discover.c
> index d5bc1314c341..efc6bf95bb67 100644
> --- a/drivers/scsi/libsas/sas_discover.c
> +++ b/drivers/scsi/libsas/sas_discover.c
> @@ -269,17 +269,6 @@ static void sas_resume_devices(struct work_struct *work)
>   	sas_resume_sata(port);
>   }
>   
> -/**
> - * sas_discover_end_dev - discover an end device (SSP, etc)
> - * @dev: pointer to domain device of interest
> - *
> - * See comment in sas_discover_sata().
> - */
> -int sas_discover_end_dev(struct domain_device *dev)
> -{
> -	return sas_notify_lldd_dev_found(dev);
> -}
> -
>   /* ---------- Device registration and unregistration ---------- */
>   
>   void sas_free_device(struct kref *kref)
> @@ -447,7 +436,7 @@ static void sas_discover_domain(struct work_struct *work)
>   
>   	switch (dev->dev_type) {
>   	case SAS_END_DEVICE:
> -		error = sas_discover_end_dev(dev);
> +		error = sas_notify_lldd_dev_found(dev);

For me, personally, I prefer consistent API name, like 
sas_discover_end_dev() and sas_discover_sata(), even if 
sas_discover_end_dev() is just a wrapper.

>   		break;
>   	case SAS_EDGE_EXPANDER_DEVICE:
>   	case SAS_FANOUT_EXPANDER_DEVICE:
> diff --git a/drivers/scsi/libsas/sas_expander.c b/drivers/scsi/libsas/sas_expander.c
> index a04cad620e93..aa8ea3b1f2e4 100644
> --- a/drivers/scsi/libsas/sas_expander.c
> +++ b/drivers/scsi/libsas/sas_expander.c
> @@ -855,9 +855,9 @@ static struct domain_device *sas_ex_discover_end_dev(
>   
>   		list_add_tail(&child->disco_list_node, &parent->port->disco_list);
>   
> -		res = sas_discover_end_dev(child);
> +		res = sas_notify_lldd_dev_found(child);
>   		if (res) {
> -			pr_notice("sas_discover_end_dev() for device %016llx at %016llx:%02d returned 0x%x\n",
> +			pr_notice("notify lldd for device %016llx at %016llx:%02d returned 0x%x\n",
>   				  SAS_ADDR(child->sas_addr),
>   				  SAS_ADDR(parent->sas_addr), phy_id, res);
>   			goto out_list_del;
> diff --git a/include/scsi/libsas.h b/include/scsi/libsas.h
> index 1aee3d0ebbb2..87682390fb76 100644
> --- a/include/scsi/libsas.h
> +++ b/include/scsi/libsas.h
> @@ -736,7 +736,6 @@ void sas_init_disc(struct sas_discovery *disc, struct asd_sas_port *);
>   void sas_discover_event(struct asd_sas_port *, enum discover_event ev);
>   
>   int  sas_discover_sata(struct domain_device *);
> -int  sas_discover_end_dev(struct domain_device *);
>   
>   void sas_unregister_dev(struct asd_sas_port *port, struct domain_device *);
>   

Re: [PATCH 2/6] scsi: libsas: delete wrapper function sas_discover_end_dev()

[Jason Yan]

On 2022/12/5 16:57, John Garry wrote:
> On 04/12/2022 08:16, Jason Yan wrote:
>> After commit 0558f33c06bb ("scsi: libsas: direct call probe and 
>> destruct")
>> this function is only a wrapper of sas_notify_lldd_dev_found(). And the
>> function name does not reflect the real purpose of this function now.
> 
> Why is this? Maybe add "dev_found" to the name could help.
> 
>> Remove it and call sas_notify_lldd_dev_found() directly. The log is also
>> changed accordingly.
>>
>> Cc: John Garry <john.g.garry@oracle.com>
>> Signed-off-by: Jason Yan <yanaijie@huawei.com>
>> ---
>>   drivers/scsi/libsas/sas_discover.c | 13 +------------
>>   drivers/scsi/libsas/sas_expander.c |  4 ++--
>>   include/scsi/libsas.h              |  1 -
>>   3 files changed, 3 insertions(+), 15 deletions(-)
>>
>> diff --git a/drivers/scsi/libsas/sas_discover.c 
>> b/drivers/scsi/libsas/sas_discover.c
>> index d5bc1314c341..efc6bf95bb67 100644
>> --- a/drivers/scsi/libsas/sas_discover.c
>> +++ b/drivers/scsi/libsas/sas_discover.c
>> @@ -269,17 +269,6 @@ static void sas_resume_devices(struct work_struct 
>> *work)
>>       sas_resume_sata(port);
>>   }
>> -/**
>> - * sas_discover_end_dev - discover an end device (SSP, etc)
>> - * @dev: pointer to domain device of interest
>> - *
>> - * See comment in sas_discover_sata().
>> - */
>> -int sas_discover_end_dev(struct domain_device *dev)
>> -{
>> -    return sas_notify_lldd_dev_found(dev);
>> -}
>> -
>>   /* ---------- Device registration and unregistration ---------- */
>>   void sas_free_device(struct kref *kref)
>> @@ -447,7 +436,7 @@ static void sas_discover_domain(struct work_struct 
>> *work)
>>       switch (dev->dev_type) {
>>       case SAS_END_DEVICE:
>> -        error = sas_discover_end_dev(dev);
>> +        error = sas_notify_lldd_dev_found(dev);
> 
> For me, personally, I prefer consistent API name, like 
> sas_discover_end_dev() and sas_discover_sata(), even if 
> sas_discover_end_dev() is just a wrapper.

Fair enough. I was just thinking that this API name is not proper now
because it is only notifying the lldd.

I will drop this patch if you insist.

Thanks,
Jason

[PATCH 3/6] scsi: libsas: rename sas_discover_sata() and related refactors

[Jason Yan]

After commit 0558f33c06bb ("scsi: libsas: direct call probe and destruct")
this name does not reflect the real purpose of this function. So rename
it to sas_ata_notify_lldd_dev_found().

On the other hand, the coding style where calling this interface is
inconsistent with other interfaces for sata devices. The standard style
for other sata interfaces is like:

    #ifdefine CONFIG_SCSI_SAS_ATA
    void sas_ata_task_abort(struct sas_task *task);
    #else
    static inline void sas_ata_task_abort(struct sas_task *task)
    {
    }
    #endif

And the callers does not have to do things like "#ifdefine CONFIG_SCSI_SAS_ATA"
and may call the interface directly. So follow the standard style here.

Cc: John Garry <john.g.garry@oracle.com>
Signed-off-by: Jason Yan <yanaijie@huawei.com>
---
 drivers/scsi/libsas/sas_ata.c      | 4 ++--
 drivers/scsi/libsas/sas_discover.c | 8 +-------
 drivers/scsi/libsas/sas_expander.c | 4 ++--
 include/scsi/libsas.h              | 2 --
 include/scsi/sas_ata.h             | 6 ++++++
 5 files changed, 11 insertions(+), 13 deletions(-)

diff --git a/drivers/scsi/libsas/sas_ata.c b/drivers/scsi/libsas/sas_ata.c
index 34009c330eb2..01b6175a8960 100644
--- a/drivers/scsi/libsas/sas_ata.c
+++ b/drivers/scsi/libsas/sas_ata.c
@@ -738,14 +738,14 @@ void sas_resume_sata(struct asd_sas_port *port)
 }
 
 /**
- * sas_discover_sata - discover an STP/SATA domain device
+ * sas_ata_notify_lldd_dev_found - notify an STP/SATA domain device found
  * @dev: pointer to struct domain_device of interest
  *
  * Devices directly attached to a HA port, have no parents.  All other
  * devices do, and should have their "parent" pointer set appropriately
  * before calling this function.
  */
-int sas_discover_sata(struct domain_device *dev)
+int sas_ata_notify_lldd_dev_found(struct domain_device *dev)
 {
 	if (dev->dev_type == SAS_SATA_PM)
 		return -ENODEV;
diff --git a/drivers/scsi/libsas/sas_discover.c b/drivers/scsi/libsas/sas_discover.c
index efc6bf95bb67..e7e3e230e8d1 100644
--- a/drivers/scsi/libsas/sas_discover.c
+++ b/drivers/scsi/libsas/sas_discover.c
@@ -444,14 +444,8 @@ static void sas_discover_domain(struct work_struct *work)
 		break;
 	case SAS_SATA_DEV:
 	case SAS_SATA_PM:
-#ifdef CONFIG_SCSI_SAS_ATA
-		error = sas_discover_sata(dev);
+		error = sas_ata_notify_lldd_dev_found(dev);
 		break;
-#else
-		pr_notice("ATA device seen but CONFIG_SCSI_SAS_ATA=N so cannot attach\n");
-		fallthrough;
-#endif
-		/* Fall through - only for the #else condition above. */
 	default:
 		error = -ENXIO;
 		pr_err("unhandled device %d\n", dev->dev_type);
diff --git a/drivers/scsi/libsas/sas_expander.c b/drivers/scsi/libsas/sas_expander.c
index aa8ea3b1f2e4..a8af723fab3c 100644
--- a/drivers/scsi/libsas/sas_expander.c
+++ b/drivers/scsi/libsas/sas_expander.c
@@ -830,9 +830,9 @@ static struct domain_device *sas_ex_discover_end_dev(
 
 		list_add_tail(&child->disco_list_node, &parent->port->disco_list);
 
-		res = sas_discover_sata(child);
+		res = sas_ata_notify_lldd_dev_found(child);
 		if (res) {
-			pr_notice("sas_discover_sata() for device %16llx at %016llx:%02d returned 0x%x\n",
+			pr_notice("notify lldd for device %16llx at %016llx:%02d returned 0x%x\n",
 				  SAS_ADDR(child->sas_addr),
 				  SAS_ADDR(parent->sas_addr), phy_id, res);
 			goto out_list_del;
diff --git a/include/scsi/libsas.h b/include/scsi/libsas.h
index 87682390fb76..b8b05f653a96 100644
--- a/include/scsi/libsas.h
+++ b/include/scsi/libsas.h
@@ -735,8 +735,6 @@ void sas_unregister_domain_devices(struct asd_sas_port *port, int gone);
 void sas_init_disc(struct sas_discovery *disc, struct asd_sas_port *);
 void sas_discover_event(struct asd_sas_port *, enum discover_event ev);
 
-int  sas_discover_sata(struct domain_device *);
-
 void sas_unregister_dev(struct asd_sas_port *port, struct domain_device *);
 
 void sas_init_dev(struct domain_device *);
diff --git a/include/scsi/sas_ata.h b/include/scsi/sas_ata.h
index 9c927d46f136..a3857595ba55 100644
--- a/include/scsi/sas_ata.h
+++ b/include/scsi/sas_ata.h
@@ -36,6 +36,7 @@ void sas_ata_device_link_abort(struct domain_device *dev, bool force_reset);
 int sas_execute_ata_cmd(struct domain_device *device, u8 *fis,
 			int force_phy_id);
 int smp_ata_check_ready_type(struct ata_link *link);
+int sas_ata_notify_lldd_dev_found(struct domain_device *dev);
 #else
 
 
@@ -103,6 +104,11 @@ static inline int smp_ata_check_ready_type(struct ata_link *link)
 {
 	return 0;
 }
+static inline int sas_ata_notify_lldd_dev_found(struct domain_device *dev)
+{
+	pr_notice("ATA device seen but CONFIG_SCSI_SAS_ATA=N so cannot attach\n");
+	return -ENXIO;
+}
 #endif
 
 #endif /* _SAS_ATA_H_ */
-- 
2.31.1

[PATCH 4/6] scsi: libsas: remove useless dev_list delete in sas_ex_discover_end_dev()

[Jason Yan]

The domain device 'child' is allocated in sas_ex_discover_end_dev() and
never been added to dev_list. So remove the useless list_del() and
related locks.

Cc: John Garry <john.g.garry@oracle.com>
Signed-off-by: Jason Yan <yanaijie@huawei.com>
---
 drivers/scsi/libsas/sas_expander.c | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/drivers/scsi/libsas/sas_expander.c b/drivers/scsi/libsas/sas_expander.c
index a8af723fab3c..82ea7560a888 100644
--- a/drivers/scsi/libsas/sas_expander.c
+++ b/drivers/scsi/libsas/sas_expander.c
@@ -875,9 +875,6 @@ static struct domain_device *sas_ex_discover_end_dev(
  out_list_del:
 	sas_rphy_free(child->rphy);
 	list_del(&child->disco_list_node);
-	spin_lock_irq(&parent->port->dev_list_lock);
-	list_del(&child->dev_list_node);
-	spin_unlock_irq(&parent->port->dev_list_lock);
  out_free:
 	sas_port_delete(phy->port);
  out_err:
-- 
2.31.1

Re: [PATCH 4/6] scsi: libsas: remove useless dev_list delete in sas_ex_discover_end_dev()

[John Garry]

On 04/12/2022 08:16, Jason Yan wrote:
> The domain device 'child' is allocated in sas_ex_discover_end_dev() and
> never been added to dev_list. So remove the useless list_del() and
> related locks.
> 
> Cc: John Garry <john.g.garry@oracle.com>
> Signed-off-by: Jason Yan <yanaijie@huawei.com>
> ---
>   drivers/scsi/libsas/sas_expander.c | 3 ---
>   1 file changed, 3 deletions(-)
> 
> diff --git a/drivers/scsi/libsas/sas_expander.c b/drivers/scsi/libsas/sas_expander.c
> index a8af723fab3c..82ea7560a888 100644
> --- a/drivers/scsi/libsas/sas_expander.c
> +++ b/drivers/scsi/libsas/sas_expander.c
> @@ -875,9 +875,6 @@ static struct domain_device *sas_ex_discover_end_dev(
>    out_list_del:
>   	sas_rphy_free(child->rphy);
>   	list_del(&child->disco_list_node);
> -	spin_lock_irq(&parent->port->dev_list_lock);
> -	list_del(&child->dev_list_node);
> -	spin_unlock_irq(&parent->port->dev_list_lock);

Since we have the spin lock'ing, this seems to be have been 
intentionally added (and not some simple typo or similar) - any idea of 
the origin?

Thanks,
John

>    out_free:
>   	sas_port_delete(phy->port);
>    out_err:

Re: [PATCH 4/6] scsi: libsas: remove useless dev_list delete in sas_ex_discover_end_dev()

[Jason Yan]

On 2022/12/5 17:14, John Garry wrote:
> On 04/12/2022 08:16, Jason Yan wrote:
>> The domain device 'child' is allocated in sas_ex_discover_end_dev() and
>> never been added to dev_list. So remove the useless list_del() and
>> related locks.
>>
>> Cc: John Garry <john.g.garry@oracle.com>
>> Signed-off-by: Jason Yan <yanaijie@huawei.com>
>> ---
>>   drivers/scsi/libsas/sas_expander.c | 3 ---
>>   1 file changed, 3 deletions(-)
>>
>> diff --git a/drivers/scsi/libsas/sas_expander.c 
>> b/drivers/scsi/libsas/sas_expander.c
>> index a8af723fab3c..82ea7560a888 100644
>> --- a/drivers/scsi/libsas/sas_expander.c
>> +++ b/drivers/scsi/libsas/sas_expander.c
>> @@ -875,9 +875,6 @@ static struct domain_device *sas_ex_discover_end_dev(
>>    out_list_del:
>>       sas_rphy_free(child->rphy);
>>       list_del(&child->disco_list_node);
>> -    spin_lock_irq(&parent->port->dev_list_lock);
>> -    list_del(&child->dev_list_node);
>> -    spin_unlock_irq(&parent->port->dev_list_lock);
> 
> Since we have the spin lock'ing, this seems to be have been 
> intentionally added (and not some simple typo or similar) - any idea of 
> the origin?

The new device used to be added to the dev_list in this function. But 
after 92625f9bff38 ("[SCSI] libsas: restore scan order") and 
87c8331fcf72 ("[SCSI] libsas: prevent domain rediscovery competing with 
ata error handling") it is added to the disco_list instead. But the 
list_del() and locking is forgot to be removed.

Thanks,
Jason

Re: [PATCH 4/6] scsi: libsas: remove useless dev_list delete in sas_ex_discover_end_dev()

[John Garry]

On 08/12/2022 07:21, Jason Yan wrote:
>>> t_list_del:
>>>       sas_rphy_free(child->rphy);
>>>       list_del(&child->disco_list_node);
>>> -    spin_lock_irq(&parent->port->dev_list_lock);
>>> -    list_del(&child->dev_list_node);
>>> -    spin_unlock_irq(&parent->port->dev_list_lock);
>>
>> Since we have the spin lock'ing, this seems to be have been 
>> intentionally added (and not some simple typo or similar) - any idea 
>> of the origin?
> 
> The new device used to be added to the dev_list in this function. But 
> after 92625f9bff38 ("[SCSI] libsas: restore scan order") and 
> 87c8331fcf72 ("[SCSI] libsas: prevent domain rediscovery competing with 
> ata error handling") it is added to the disco_list instead. But the 
> list_del() and locking is forgot to be removed.

OK, so can we have a fixes tag then? That even helps review, as I can 
then quickly see where we went wrong.

Thanks,
John

Re: [PATCH 4/6] scsi: libsas: remove useless dev_list delete in sas_ex_discover_end_dev()

[Jason Yan]

On 2022/12/8 18:40, John Garry wrote:
> On 08/12/2022 07:21, Jason Yan wrote:
>>>> t_list_del:
>>>>       sas_rphy_free(child->rphy);
>>>>       list_del(&child->disco_list_node);
>>>> -    spin_lock_irq(&parent->port->dev_list_lock);
>>>> -    list_del(&child->dev_list_node);
>>>> -    spin_unlock_irq(&parent->port->dev_list_lock);
>>>
>>> Since we have the spin lock'ing, this seems to be have been 
>>> intentionally added (and not some simple typo or similar) - any idea 
>>> of the origin?
>>
>> The new device used to be added to the dev_list in this function. But 
>> after 92625f9bff38 ("[SCSI] libsas: restore scan order") and 
>> 87c8331fcf72 ("[SCSI] libsas: prevent domain rediscovery competing 
>> with ata error handling") it is added to the disco_list instead. But 
>> the list_del() and locking is forgot to be removed.
> 
> OK, so can we have a fixes tag then? That even helps review, as I can 
> then quickly see where we went wrong.
> 

Sure.

Thanks,
Jason

[PATCH 5/6] scsi: libsas: factor out sas_ata_add_dev()

[Jason Yan]

Factor out sas_ata_add_dev() and put it in sas_ata.c since it is a sata
related interface. Also follow the standard coding style to define an
inline empty function when CONFIG_SCSI_SAS_ATA is not enabled.

Cc: John Garry <john.g.garry@oracle.com>
Signed-off-by: Jason Yan <yanaijie@huawei.com>
---
 drivers/scsi/libsas/sas_ata.c      | 62 ++++++++++++++++++++++++++++++
 drivers/scsi/libsas/sas_expander.c | 54 +-------------------------
 include/scsi/sas_ata.h             |  9 +++++
 3 files changed, 73 insertions(+), 52 deletions(-)

diff --git a/drivers/scsi/libsas/sas_ata.c b/drivers/scsi/libsas/sas_ata.c
index 01b6175a8960..f440203eb280 100644
--- a/drivers/scsi/libsas/sas_ata.c
+++ b/drivers/scsi/libsas/sas_ata.c
@@ -677,6 +677,68 @@ void sas_probe_sata(struct asd_sas_port *port)
 
 }
 
+int sas_ata_add_dev(struct domain_device *parent, struct ex_phy *phy,
+		    struct domain_device *child, int phy_id)
+{
+	struct sas_rphy *rphy;
+	int ret;
+
+	if (child->linkrate > parent->min_linkrate) {
+		struct sas_phy *cphy = child->phy;
+		enum sas_linkrate min_prate = cphy->minimum_linkrate,
+			parent_min_lrate = parent->min_linkrate,
+			min_linkrate = (min_prate > parent_min_lrate) ?
+					parent_min_lrate : 0;
+		struct sas_phy_linkrates rates = {
+			.maximum_linkrate = parent->min_linkrate,
+			.minimum_linkrate = min_linkrate,
+		};
+
+		pr_notice("ex %016llx phy%02d SATA device linkrate > min pathway connection rate, attempting to lower device linkrate\n",
+			  SAS_ADDR(child->sas_addr), phy_id);
+		ret = sas_smp_phy_control(parent, phy_id,
+					  PHY_FUNC_LINK_RESET, &rates);
+		if (ret) {
+			pr_err("ex %016llx phy%02d SATA device could not set linkrate (%d)\n",
+			       SAS_ADDR(child->sas_addr), phy_id, ret);
+			return ret;
+		}
+		pr_notice("ex %016llx phy%02d SATA device set linkrate successfully\n",
+			  SAS_ADDR(child->sas_addr), phy_id);
+		child->linkrate = child->min_linkrate;
+	}
+	ret = sas_get_ata_info(child, phy);
+	if (ret)
+		return ret;
+
+	sas_init_dev(child);
+	ret = sas_ata_init(child);
+	if (ret)
+		return ret;
+
+	rphy = sas_end_device_alloc(phy->port);
+	if (!rphy)
+		return ret;
+
+	rphy->identify.phy_identifier = phy_id;
+	child->rphy = rphy;
+	get_device(&rphy->dev);
+
+	list_add_tail(&child->disco_list_node, &parent->port->disco_list);
+
+	ret = sas_ata_notify_lldd_dev_found(child);
+	if (ret) {
+		pr_notice("notify lldd for device %16llx at %016llx:%02d returned 0x%x\n",
+			  SAS_ADDR(child->sas_addr),
+			  SAS_ADDR(parent->sas_addr), phy_id, ret);
+		sas_rphy_free(child->rphy);
+		list_del(&child->disco_list_node);
+		return ret;
+	}
+
+	return 0;
+}
+
 static void sas_ata_flush_pm_eh(struct asd_sas_port *port, const char *func)
 {
 	struct domain_device *dev, *n;
diff --git a/drivers/scsi/libsas/sas_expander.c b/drivers/scsi/libsas/sas_expander.c
index 82ea7560a888..747f4fc795f4 100644
--- a/drivers/scsi/libsas/sas_expander.c
+++ b/drivers/scsi/libsas/sas_expander.c
@@ -785,61 +785,11 @@ static struct domain_device *sas_ex_discover_end_dev(
 	sas_ex_get_linkrate(parent, child, phy);
 	sas_device_set_phy(child, phy->port);
 
-#ifdef CONFIG_SCSI_SAS_ATA
 	if ((phy->attached_tproto & SAS_PROTOCOL_STP) || phy->attached_sata_dev) {
-		if (child->linkrate > parent->min_linkrate) {
-			struct sas_phy *cphy = child->phy;
-			enum sas_linkrate min_prate = cphy->minimum_linkrate,
-				parent_min_lrate = parent->min_linkrate,
-				min_linkrate = (min_prate > parent_min_lrate) ?
-					       parent_min_lrate : 0;
-			struct sas_phy_linkrates rates = {
-				.maximum_linkrate = parent->min_linkrate,
-				.minimum_linkrate = min_linkrate,
-			};
-			int ret;
-
-			pr_notice("ex %016llx phy%02d SATA device linkrate > min pathway connection rate, attempting to lower device linkrate\n",
-				   SAS_ADDR(child->sas_addr), phy_id);
-			ret = sas_smp_phy_control(parent, phy_id,
-						  PHY_FUNC_LINK_RESET, &rates);
-			if (ret) {
-				pr_err("ex %016llx phy%02d SATA device could not set linkrate (%d)\n",
-				       SAS_ADDR(child->sas_addr), phy_id, ret);
-				goto out_free;
-			}
-			pr_notice("ex %016llx phy%02d SATA device set linkrate successfully\n",
-				  SAS_ADDR(child->sas_addr), phy_id);
-			child->linkrate = child->min_linkrate;
-		}
-		res = sas_get_ata_info(child, phy);
-		if (res)
-			goto out_free;
-
-		sas_init_dev(child);
-		res = sas_ata_init(child);
+		res = sas_ata_add_dev(parent, phy, child, phy_id);
 		if (res)
 			goto out_free;
-		rphy = sas_end_device_alloc(phy->port);
-		if (!rphy)
-			goto out_free;
-		rphy->identify.phy_identifier = phy_id;
-
-		child->rphy = rphy;
-		get_device(&rphy->dev);
-
-		list_add_tail(&child->disco_list_node, &parent->port->disco_list);
-
-		res = sas_ata_notify_lldd_dev_found(child);
-		if (res) {
-			pr_notice("notify lldd for device %16llx at %016llx:%02d returned 0x%x\n",
-				  SAS_ADDR(child->sas_addr),
-				  SAS_ADDR(parent->sas_addr), phy_id, res);
-			goto out_list_del;
-		}
-	} else
-#endif
-	  if (phy->attached_tproto & SAS_PROTOCOL_SSP) {
+	} else if (phy->attached_tproto & SAS_PROTOCOL_SSP) {
 		child->dev_type = SAS_END_DEVICE;
 		rphy = sas_end_device_alloc(phy->port);
 		/* FIXME: error handling */
diff --git a/include/scsi/sas_ata.h b/include/scsi/sas_ata.h
index a3857595ba55..23a0853a5a0a 100644
--- a/include/scsi/sas_ata.h
+++ b/include/scsi/sas_ata.h
@@ -37,6 +37,8 @@ int sas_execute_ata_cmd(struct domain_device *device, u8 *fis,
 			int force_phy_id);
 int smp_ata_check_ready_type(struct ata_link *link);
 int sas_ata_notify_lldd_dev_found(struct domain_device *dev);
+int sas_ata_add_dev(struct domain_device *parent, struct ex_phy *phy,
+		    struct domain_device *child, int phy_id);
 #else
 
 
@@ -109,6 +111,13 @@ static inline int sas_ata_notify_lldd_dev_found(struct domain_device *dev)
 	pr_notice("ATA device seen but CONFIG_SCSI_SAS_ATA=N so cannot attach\n");
 	return -ENXIO;
 }
+static inline int sas_ata_add_dev(struct domain_device *parent, struct ex_phy *phy,
+				  struct domain_device *child, int phy_id)
+{
+	pr_notice("ATA is not enabled, target proto 0x%x at %016llx:0x%x\n",
+		  phy->attached_tproto, SAS_ADDR(parent->sas_addr), phy_id);
+	return -ENODEV;
+}
 #endif
 
 #endif /* _SAS_ATA_H_ */
-- 
2.31.1

Re: [PATCH 5/6] scsi: libsas: factor out sas_ata_add_dev()

[John Garry]

On 04/12/2022 08:16, Jason Yan wrote:
> +static inline int sas_ata_add_dev(struct domain_device *parent, struct ex_phy *phy,
> +				  struct domain_device *child, int phy_id)
> +{
> +	pr_notice("ATA is not enabled, target proto 0x%x at %016llx:0x%x\n",

nit: how about add "SCSI_SAS_ATA is not enabled ..." or use similar to 
the log in sas_discover_domain()?

> +		  phy->attached_tproto, SAS_ADDR(parent->sas_addr), phy_id);
> +	return -ENODEV;

Re: [PATCH 5/6] scsi: libsas: factor out sas_ata_add_dev()

[Jason Yan]

On 2022/12/5 17:24, John Garry wrote:
> On 04/12/2022 08:16, Jason Yan wrote:
>> +static inline int sas_ata_add_dev(struct domain_device *parent, 
>> struct ex_phy *phy,
>> +                  struct domain_device *child, int phy_id)
>> +{
>> +    pr_notice("ATA is not enabled, target proto 0x%x at %016llx:0x%x\n",
> 
> nit: how about add "SCSI_SAS_ATA is not enabled ..." or use similar to 
> the log in sas_discover_domain()?

Yes, make sense. Will update.

Thanks,
Jason

> 
>> +          phy->attached_tproto, SAS_ADDR(parent->sas_addr), phy_id);
>> +    return -ENODEV;
> 
> .

[PATCH 6/6] scsi: libsas: factor out sas_ex_add_dev()

[Jason Yan]

Factor out sas_ex_add_dev() to be consistent with sas_ata_add_dev() and
unify the error handling.

Cc: John Garry <john.g.garry@oracle.com>
Signed-off-by: Jason Yan <yanaijie@huawei.com>
---
 drivers/scsi/libsas/sas_expander.c | 68 +++++++++++++++++-------------
 1 file changed, 39 insertions(+), 29 deletions(-)

diff --git a/drivers/scsi/libsas/sas_expander.c b/drivers/scsi/libsas/sas_expander.c
index 747f4fc795f4..3c72b167d43a 100644
--- a/drivers/scsi/libsas/sas_expander.c
+++ b/drivers/scsi/libsas/sas_expander.c
@@ -751,13 +751,46 @@ static void sas_ex_get_linkrate(struct domain_device *parent,
 	child->pathways = min(child->pathways, parent->pathways);
 }
 
+static int sas_ex_add_dev(struct domain_device *parent, struct ex_phy *phy,
+			  struct domain_device *child, int phy_id)
+{
+	struct sas_rphy *rphy;
+	int res;
+
+	child->dev_type = SAS_END_DEVICE;
+	rphy = sas_end_device_alloc(phy->port);
+	if (unlikely(!rphy))
+		return -ENOMEM;
+
+	child->tproto = phy->attached_tproto;
+	sas_init_dev(child);
+
+	child->rphy = rphy;
+	get_device(&rphy->dev);
+	rphy->identify.phy_identifier = phy_id;
+	sas_fill_in_rphy(child, rphy);
+
+	list_add_tail(&child->disco_list_node, &parent->port->disco_list);
+
+	res = sas_notify_lldd_dev_found(child);
+	if (res) {
+		pr_notice("notify lldd for device %016llx at %016llx:%02d returned 0x%x\n",
+				SAS_ADDR(child->sas_addr),
+				SAS_ADDR(parent->sas_addr), phy_id, res);
+		sas_rphy_free(child->rphy);
+		list_del(&child->disco_list_node);
+		return res;
+	}
+
+	return 0;
+}
+
 static struct domain_device *sas_ex_discover_end_dev(
 	struct domain_device *parent, int phy_id)
 {
 	struct expander_device *parent_ex = &parent->ex_dev;
 	struct ex_phy *phy = &parent_ex->ex_phy[phy_id];
 	struct domain_device *child = NULL;
-	struct sas_rphy *rphy;
 	int res;
 
 	if (phy->attached_sata_host || phy->attached_sata_ps)
@@ -787,44 +820,21 @@ static struct domain_device *sas_ex_discover_end_dev(
 
 	if ((phy->attached_tproto & SAS_PROTOCOL_STP) || phy->attached_sata_dev) {
 		res = sas_ata_add_dev(parent, phy, child, phy_id);
-		if (res)
-			goto out_free;
 	} else if (phy->attached_tproto & SAS_PROTOCOL_SSP) {
-		child->dev_type = SAS_END_DEVICE;
-		rphy = sas_end_device_alloc(phy->port);
-		/* FIXME: error handling */
-		if (unlikely(!rphy))
-			goto out_free;
-		child->tproto = phy->attached_tproto;
-		sas_init_dev(child);
-
-		child->rphy = rphy;
-		get_device(&rphy->dev);
-		rphy->identify.phy_identifier = phy_id;
-		sas_fill_in_rphy(child, rphy);
-
-		list_add_tail(&child->disco_list_node, &parent->port->disco_list);
-
-		res = sas_notify_lldd_dev_found(child);
-		if (res) {
-			pr_notice("notify lldd for device %016llx at %016llx:%02d returned 0x%x\n",
-				  SAS_ADDR(child->sas_addr),
-				  SAS_ADDR(parent->sas_addr), phy_id, res);
-			goto out_list_del;
-		}
+		res = sas_ex_add_dev(parent, phy, child, phy_id);
 	} else {
 		pr_notice("target proto 0x%x at %016llx:0x%x not handled\n",
 			  phy->attached_tproto, SAS_ADDR(parent->sas_addr),
 			  phy_id);
-		goto out_free;
+		res = -ENODEV;
 	}
 
+	if (res)
+		goto out_free;
+
 	list_add_tail(&child->siblings, &parent_ex->children);
 	return child;
 
- out_list_del:
-	sas_rphy_free(child->rphy);
-	list_del(&child->disco_list_node);
  out_free:
 	sas_port_delete(phy->port);
  out_err:
-- 
2.31.1

Re: [PATCH 6/6] scsi: libsas: factor out sas_ex_add_dev()

[John Garry]

On 04/12/2022 08:16, Jason Yan wrote:
> Factor out sas_ex_add_dev() to be consistent with sas_ata_add_dev() and
> unify the error handling.
> 
> Cc: John Garry <john.g.garry@oracle.com>
> Signed-off-by: Jason Yan <yanaijie@huawei.com>
> ---
>   drivers/scsi/libsas/sas_expander.c | 68 +++++++++++++++++-------------
>   1 file changed, 39 insertions(+), 29 deletions(-)
> 
> diff --git a/drivers/scsi/libsas/sas_expander.c b/drivers/scsi/libsas/sas_expander.c
> index 747f4fc795f4..3c72b167d43a 100644
> --- a/drivers/scsi/libsas/sas_expander.c
> +++ b/drivers/scsi/libsas/sas_expander.c
> @@ -751,13 +751,46 @@ static void sas_ex_get_linkrate(struct domain_device *parent,
>   	child->pathways = min(child->pathways, parent->pathways);
>   }
>   
> +static int sas_ex_add_dev(struct domain_device *parent, struct ex_phy *phy,
> +			  struct domain_device *child, int phy_id)
> +{
> +	struct sas_rphy *rphy;
> +	int res;
> +
> +	child->dev_type = SAS_END_DEVICE;
> +	rphy = sas_end_device_alloc(phy->port);
> +	if (unlikely(!rphy))

nit: this is not fastpath so unlikely can be avoided

> +		return -ENOMEM;
> +
> +	child->tproto = phy->attached_tproto;
> +	sas_init_dev(child);
> +
> +	child->rphy = rphy;
> +	get_device(&rphy->dev);
> +	rphy->identify.phy_identifier = phy_id;
> +	sas_fill_in_rphy(child, rphy);
> +
> +	list_add_tail(&child->disco_list_node, &parent->port->disco_list);
> +
> +	res = sas_notify_lldd_dev_found(child);
> +	if (res) {
> +		pr_notice("notify lldd for device %016llx at %016llx:%02d returned 0x%x\n",
> +				SAS_ADDR(child->sas_addr),
> +				SAS_ADDR(parent->sas_addr), phy_id, res);

nit: these lines could be aligned with (, as it was before

> +		sas_rphy_free(child->rphy);
> +		list_del(&child->disco_list_node);
> +		return res;
> +	}
> +
> +	return 0;
> +}
> +
>   static struct domain_device *sas_ex_discover_end_dev(
>   	struct domain_device *parent, int phy_id)
>   {
>   	struct expander_device *parent_ex = &parent->ex_dev;
>   	struct ex_phy *phy = &parent_ex->ex_phy[phy_id];
>   	struct domain_device *child = NULL;
> -	struct sas_rphy *rphy;
>   	int res;
>   
>   	if (phy->attached_sata_host || phy->attached_sata_ps)
> @@ -787,44 +820,21 @@ static struct domain_device *sas_ex_discover_end_dev(
>   
>   	if ((phy->attached_tproto & SAS_PROTOCOL_STP) || phy->attached_sata_dev) {
>   		res = sas_ata_add_dev(parent, phy, child, phy_id);
> -		if (res)
> -			goto out_free;
>   	} else if (phy->attached_tproto & SAS_PROTOCOL_SSP) {
> -		child->dev_type = SAS_END_DEVICE;
> -		rphy = sas_end_device_alloc(phy->port);
> -		/* FIXME: error handling */

so has the error handling been fixed now?

> -		if (unlikely(!rphy))
> -			goto out_free;
> -		child->tproto = phy->attached_tproto;
> -		sas_init_dev(child);
> -
> -		child->rphy = rphy;
> -		get_device(&rphy->dev);
> -		rphy->identify.phy_identifier = phy_id;
> -		sas_fill_in_rphy(child, rphy);
> -
> -		list_add_tail(&child->disco_list_node, &parent->port->disco_list);
> -
> -		res = sas_notify_lldd_dev_found(child);
> -		if (res) {
> -			pr_notice("notify lldd for device %016llx at %016llx:%02d returned 0x%x\n",
> -				  SAS_ADDR(child->sas_addr),
> -				  SAS_ADDR(parent->sas_addr), phy_id, res);
> -			goto out_list_del;
> -		}
> +		res = sas_ex_add_dev(parent, phy, child, phy_id);
>   	} else {
>   		pr_notice("target proto 0x%x at %016llx:0x%x not handled\n",
>   			  phy->attached_tproto, SAS_ADDR(parent->sas_addr),
>   			  phy_id);
> -		goto out_free;
> +		res = -ENODEV;
>   	}
>   
> +	if (res)
> +		goto out_free;
> +
>   	list_add_tail(&child->siblings, &parent_ex->children);
>   	return child;
>   
> - out_list_del:
> -	sas_rphy_free(child->rphy);
> -	list_del(&child->disco_list_node);
>    out_free:
>   	sas_port_delete(phy->port);
>    out_err:

Re: [PATCH 6/6] scsi: libsas: factor out sas_ex_add_dev()

[Jason Yan]

On 2022/12/5 17:31, John Garry wrote:
> On 04/12/2022 08:16, Jason Yan wrote:
>> Factor out sas_ex_add_dev() to be consistent with sas_ata_add_dev() and
>> unify the error handling.
>>
>> Cc: John Garry <john.g.garry@oracle.com>
>> Signed-off-by: Jason Yan <yanaijie@huawei.com>
>> ---
>>   drivers/scsi/libsas/sas_expander.c | 68 +++++++++++++++++-------------
>>   1 file changed, 39 insertions(+), 29 deletions(-)
>>
>> diff --git a/drivers/scsi/libsas/sas_expander.c 
>> b/drivers/scsi/libsas/sas_expander.c
>> index 747f4fc795f4..3c72b167d43a 100644
>> --- a/drivers/scsi/libsas/sas_expander.c
>> +++ b/drivers/scsi/libsas/sas_expander.c
>> @@ -751,13 +751,46 @@ static void sas_ex_get_linkrate(struct 
>> domain_device *parent,
>>       child->pathways = min(child->pathways, parent->pathways);
>>   }
>> +static int sas_ex_add_dev(struct domain_device *parent, struct ex_phy 
>> *phy,
>> +              struct domain_device *child, int phy_id)
>> +{
>> +    struct sas_rphy *rphy;
>> +    int res;
>> +
>> +    child->dev_type = SAS_END_DEVICE;
>> +    rphy = sas_end_device_alloc(phy->port);
>> +    if (unlikely(!rphy))
> 
> nit: this is not fastpath so unlikely can be avoided
> 

ok. It is only a hint for the compiler so not a big deal. I can delete it.

>> +        return -ENOMEM;
>> +
>> +    child->tproto = phy->attached_tproto;
>> +    sas_init_dev(child);
>> +
>> +    child->rphy = rphy;
>> +    get_device(&rphy->dev);
>> +    rphy->identify.phy_identifier = phy_id;
>> +    sas_fill_in_rphy(child, rphy);
>> +
>> +    list_add_tail(&child->disco_list_node, &parent->port->disco_list);
>> +
>> +    res = sas_notify_lldd_dev_found(child);
>> +    if (res) {
>> +        pr_notice("notify lldd for device %016llx at %016llx:%02d 
>> returned 0x%x\n",
>> +                SAS_ADDR(child->sas_addr),
>> +                SAS_ADDR(parent->sas_addr), phy_id, res);
> 
> nit: these lines could be aligned with (, as it was before

Nice catch. Will fix.

> 
>> +        sas_rphy_free(child->rphy);
>> +        list_del(&child->disco_list_node);
>> +        return res;
>> +    }
>> +
>> +    return 0;
>> +}
>> +
>>   static struct domain_device *sas_ex_discover_end_dev(
>>       struct domain_device *parent, int phy_id)
>>   {
>>       struct expander_device *parent_ex = &parent->ex_dev;
>>       struct ex_phy *phy = &parent_ex->ex_phy[phy_id];
>>       struct domain_device *child = NULL;
>> -    struct sas_rphy *rphy;
>>       int res;
>>       if (phy->attached_sata_host || phy->attached_sata_ps)
>> @@ -787,44 +820,21 @@ static struct domain_device 
>> *sas_ex_discover_end_dev(
>>       if ((phy->attached_tproto & SAS_PROTOCOL_STP) || 
>> phy->attached_sata_dev) {
>>           res = sas_ata_add_dev(parent, phy, child, phy_id);
>> -        if (res)
>> -            goto out_free;
>>       } else if (phy->attached_tproto & SAS_PROTOCOL_SSP) {
>> -        child->dev_type = SAS_END_DEVICE;
>> -        rphy = sas_end_device_alloc(phy->port);
>> -        /* FIXME: error handling */
> 
> so has the error handling been fixed now?

IIUC, the error handling is ok so this comment can be deleted.

Thanks,
Jason