[PATCH net-next 04/13] sunhme: Return an ERR_PTR from quattro_pci_find

Sean Anderson posted 13 patches 3 years, 6 months ago
There is a newer version of this series
[PATCH net-next 04/13] sunhme: Return an ERR_PTR from quattro_pci_find
Posted by Sean Anderson 3 years, 6 months ago
In order to differentiate between a missing bridge and an OOM condition,
return ERR_PTRs from quattro_pci_find. This also does some general linting
in the area.

Signed-off-by: Sean Anderson <seanga2@gmail.com>
---

 drivers/net/ethernet/sun/sunhme.c | 33 +++++++++++++++++++------------
 1 file changed, 20 insertions(+), 13 deletions(-)

diff --git a/drivers/net/ethernet/sun/sunhme.c b/drivers/net/ethernet/sun/sunhme.c
index 1fc16801f520..52247505d08e 100644
--- a/drivers/net/ethernet/sun/sunhme.c
+++ b/drivers/net/ethernet/sun/sunhme.c
@@ -2569,30 +2569,33 @@ static void quattro_sbus_free_irqs(void)
 #ifdef CONFIG_PCI
 static struct quattro *quattro_pci_find(struct pci_dev *pdev)
 {
+	int i;
 	struct pci_dev *bdev = pdev->bus->self;
 	struct quattro *qp;
 
-	if (!bdev) return NULL;
+	if (!bdev)
+		return ERR_PTR(-ENODEV);
+
 	for (qp = qfe_pci_list; qp != NULL; qp = qp->next) {
 		struct pci_dev *qpdev = qp->quattro_dev;
 
 		if (qpdev == bdev)
 			return qp;
 	}
+
 	qp = kmalloc(sizeof(struct quattro), GFP_KERNEL);
-	if (qp != NULL) {
-		int i;
+	if (!qp)
+		return ERR_PTR(-ENOMEM);
 
-		for (i = 0; i < 4; i++)
-			qp->happy_meals[i] = NULL;
+	for (i = 0; i < 4; i++)
+		qp->happy_meals[i] = NULL;
 
-		qp->quattro_dev = bdev;
-		qp->next = qfe_pci_list;
-		qfe_pci_list = qp;
+	qp->quattro_dev = bdev;
+	qp->next = qfe_pci_list;
+	qfe_pci_list = qp;
 
-		/* No range tricks necessary on PCI. */
-		qp->nranges = 0;
-	}
+	/* No range tricks necessary on PCI. */
+	qp->nranges = 0;
 	return qp;
 }
 #endif /* CONFIG_PCI */
@@ -2949,11 +2952,15 @@ static int happy_meal_pci_probe(struct pci_dev *pdev,
 
 	if (!strcmp(prom_name, "SUNW,qfe") || !strcmp(prom_name, "qfe")) {
 		qp = quattro_pci_find(pdev);
-		if (qp == NULL)
+		if (IS_ERR(qp)) {
+			err = PTR_ERR(qp);
 			goto err_out;
+		}
+
 		for (qfe_slot = 0; qfe_slot < 4; qfe_slot++)
-			if (qp->happy_meals[qfe_slot] == NULL)
+			if (!qp->happy_meals[qfe_slot])
 				break;
+
 		if (qfe_slot == 4)
 			goto err_out;
 	}
-- 
2.37.1
Re: [PATCH net-next 04/13] sunhme: Return an ERR_PTR from quattro_pci_find
Posted by Jakub Kicinski 3 years, 6 months ago
On Sun, 18 Sep 2022 19:26:17 -0400 Sean Anderson wrote:
> +	int i;
>  	struct pci_dev *bdev = pdev->bus->self;
>  	struct quattro *qp;

reverse xmas tree variable ordering please
Re: [PATCH net-next 04/13] sunhme: Return an ERR_PTR from quattro_pci_find
Posted by Rolf Eike Beer 3 years, 6 months ago
Am Montag, 19. September 2022, 01:26:17 CEST schrieb Sean Anderson:
> In order to differentiate between a missing bridge and an OOM condition,
> return ERR_PTRs from quattro_pci_find. This also does some general linting
> in the area.
> 
> Signed-off-by: Sean Anderson <seanga2@gmail.com>
> ---
> 
>  drivers/net/ethernet/sun/sunhme.c | 33 +++++++++++++++++++------------
>  1 file changed, 20 insertions(+), 13 deletions(-)
> 
> diff --git a/drivers/net/ethernet/sun/sunhme.c
> b/drivers/net/ethernet/sun/sunhme.c index 1fc16801f520..52247505d08e 100644
> --- a/drivers/net/ethernet/sun/sunhme.c
> +++ b/drivers/net/ethernet/sun/sunhme.c
> @@ -2569,30 +2569,33 @@ static void quattro_sbus_free_irqs(void)
>  #ifdef CONFIG_PCI
>  static struct quattro *quattro_pci_find(struct pci_dev *pdev)
>  {
> +	int i;
>  	struct pci_dev *bdev = pdev->bus->self;
>  	struct quattro *qp;
> 
> -	if (!bdev) return NULL;
> +	if (!bdev)
> +		return ERR_PTR(-ENODEV);
> +
>  	for (qp = qfe_pci_list; qp != NULL; qp = qp->next) {
>  		struct pci_dev *qpdev = qp->quattro_dev;
> 
>  		if (qpdev == bdev)
>  			return qp;
>  	}
> +
>  	qp = kmalloc(sizeof(struct quattro), GFP_KERNEL);
> -	if (qp != NULL) {
> -		int i;
> +	if (!qp)
> +		return ERR_PTR(-ENOMEM);
> 
> -		for (i = 0; i < 4; i++)
> -			qp->happy_meals[i] = NULL;
> +	for (i = 0; i < 4; i++)
> +		qp->happy_meals[i] = NULL;

I know you are only reindenting it, but I dislike moving the variable up to 
the top of the function. Since the kernel is C99 meanwhile the variable could 
be declared just in the for loop. And when touching this anyway I think we 
could get rid of the magic "4" by using ARRAY_SIZE(qp->happy_meals). Or just 
replace the whole thing with memset(qp->happy_meals, 0, sizeof(qp-
>happy_meals)).

Eike
Re: [PATCH net-next 04/13] sunhme: Return an ERR_PTR from quattro_pci_find
Posted by Sean Anderson 3 years, 6 months ago
On 9/19/22 09:11, Rolf Eike Beer wrote:
> Am Montag, 19. September 2022, 01:26:17 CEST schrieb Sean Anderson:
>> In order to differentiate between a missing bridge and an OOM condition,
>> return ERR_PTRs from quattro_pci_find. This also does some general linting
>> in the area.
>>
>> Signed-off-by: Sean Anderson <seanga2@gmail.com>
>> ---
>>
>>   drivers/net/ethernet/sun/sunhme.c | 33 +++++++++++++++++++------------
>>   1 file changed, 20 insertions(+), 13 deletions(-)
>>
>> diff --git a/drivers/net/ethernet/sun/sunhme.c
>> b/drivers/net/ethernet/sun/sunhme.c index 1fc16801f520..52247505d08e 100644
>> --- a/drivers/net/ethernet/sun/sunhme.c
>> +++ b/drivers/net/ethernet/sun/sunhme.c
>> @@ -2569,30 +2569,33 @@ static void quattro_sbus_free_irqs(void)
>>   #ifdef CONFIG_PCI
>>   static struct quattro *quattro_pci_find(struct pci_dev *pdev)
>>   {
>> +	int i;
>>   	struct pci_dev *bdev = pdev->bus->self;
>>   	struct quattro *qp;
>>
>> -	if (!bdev) return NULL;
>> +	if (!bdev)
>> +		return ERR_PTR(-ENODEV);
>> +
>>   	for (qp = qfe_pci_list; qp != NULL; qp = qp->next) {
>>   		struct pci_dev *qpdev = qp->quattro_dev;
>>
>>   		if (qpdev == bdev)
>>   			return qp;
>>   	}
>> +
>>   	qp = kmalloc(sizeof(struct quattro), GFP_KERNEL);
>> -	if (qp != NULL) {
>> -		int i;
>> +	if (!qp)
>> +		return ERR_PTR(-ENOMEM);
>>
>> -		for (i = 0; i < 4; i++)
>> -			qp->happy_meals[i] = NULL;
>> +	for (i = 0; i < 4; i++)
>> +		qp->happy_meals[i] = NULL;
> 
> I know you are only reindenting it, but I dislike moving the variable up to
> the top of the function. Since the kernel is C99 meanwhile the variable could
> be declared just in the for loop. 

Hm, I thought this style was discouraged.

> And when touching this anyway I think we
> could get rid of the magic "4" by using ARRAY_SIZE(qp->happy_meals). Or just
> replace the whole thing with memset(qp->happy_meals, 0, sizeof(qp-
>> happy_meals)).

Yeah, that avoids the whole problem.

--Sean