[PATCH] octeon_ep: Add missing check for ioremap

[PATCH] octeon_ep: Add missing check for ioremap

[Jiasheng Jiang]

Add check for ioremap() and return the error if it fails in order to
guarantee the success of ioremap().

Fixes: 862cd659a6fb ("octeon_ep: Add driver framework and device initialization")
Signed-off-by: Jiasheng Jiang <jiasheng@iscas.ac.cn>
---
 drivers/net/ethernet/marvell/octeon_ep/octep_main.c | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)

diff --git a/drivers/net/ethernet/marvell/octeon_ep/octep_main.c b/drivers/net/ethernet/marvell/octeon_ep/octep_main.c
index e1853da280f9..06816d2baef8 100644
--- a/drivers/net/ethernet/marvell/octeon_ep/octep_main.c
+++ b/drivers/net/ethernet/marvell/octeon_ep/octep_main.c
@@ -969,7 +969,7 @@ static const char *octep_devid_to_str(struct octep_device *oct)
 int octep_device_setup(struct octep_device *oct)
 {
 	struct pci_dev *pdev = oct->pdev;
-	int i, ret;
+	int i, j, ret;
 
 	/* allocate memory for oct->conf */
 	oct->conf = kzalloc(sizeof(*oct->conf), GFP_KERNEL);
@@ -981,6 +981,9 @@ int octep_device_setup(struct octep_device *oct)
 		oct->mmio[i].hw_addr =
 			ioremap(pci_resource_start(oct->pdev, i * 2),
 				pci_resource_len(oct->pdev, i * 2));
+		if (!oct->mmio[i].hw_addr)
+			goto unsupported_dev;
+
 		oct->mmio[i].mapped = 1;
 	}
 
@@ -1015,8 +1018,8 @@ int octep_device_setup(struct octep_device *oct)
 	return 0;
 
 unsupported_dev:
-	for (i = 0; i < OCTEP_MMIO_REGIONS; i++)
-		iounmap(oct->mmio[i].hw_addr);
+	for (j = 0; j < i; j++)
+		iounmap(oct->mmio[j].hw_addr);
 
 	kfree(oct->conf);
 	return -1;
-- 
2.25.1

Re: [PATCH] octeon_ep: Add missing check for ioremap

[Jakub Kicinski]

On Wed, 14 Jun 2023 11:33:17 +0530 Kalesh Anakkur Purayil wrote:
> >         kfree(oct->conf);
> >         return -1;
> >  
> [Kalesh]: fix to return -ENOMEM instead of -1.

The author is not touching the return code, seems unrelated.

Re: [PATCH] octeon_ep: Add missing check for ioremap

[Jakub Kicinski]

On Wed, 14 Jun 2023 11:23:47 +0800 Jiasheng Jiang wrote:
> @@ -981,6 +981,9 @@ int octep_device_setup(struct octep_device *oct)
>  		oct->mmio[i].hw_addr =
>  			ioremap(pci_resource_start(oct->pdev, i * 2),
>  				pci_resource_len(oct->pdev, i * 2));
> +		if (!oct->mmio[i].hw_addr)
> +			goto unsupported_dev;
> +
>  		oct->mmio[i].mapped = 1;
>  	}
>  
> @@ -1015,8 +1018,8 @@ int octep_device_setup(struct octep_device *oct)
>  	return 0;
>  
>  unsupported_dev:
> -	for (i = 0; i < OCTEP_MMIO_REGIONS; i++)
> -		iounmap(oct->mmio[i].hw_addr);
> +	for (j = 0; j < i; j++)
> +		iounmap(oct->mmio[j].hw_addr);

Assuming @i is not changed by the rest of the function is a bit fragile.

Better way of handling this situation is:

unsupported_dev:
	i = OCTEP_MMIO_REGIONS;
unmap_prev:
	while (i--)
		iounmap(oct->mmio[i].hw_addr);

and jump to unmap_prev
-- 
pw-bot: cr

Re: [PATCH] octeon_ep: Add missing check for ioremap

[Jiasheng Jiang]

On Thu, 15 Jun 2023 03:53:04 +0800 Jakub Kicinski wrote:
> On Wed, 14 Jun 2023 11:23:47 +0800 Jiasheng Jiang wrote:
>> @@ -981,6 +981,9 @@ int octep_device_setup(struct octep_device *oct)
>>  		oct->mmio[i].hw_addr =
>>  			ioremap(pci_resource_start(oct->pdev, i * 2),
>>  				pci_resource_len(oct->pdev, i * 2));
>> +		if (!oct->mmio[i].hw_addr)
>> +			goto unsupported_dev;
>> +
>>  		oct->mmio[i].mapped = 1;
>>  	}
>>  
>> @@ -1015,8 +1018,8 @@ int octep_device_setup(struct octep_device *oct)
>>  	return 0;
>>  
>>  unsupported_dev:
>> -	for (i = 0; i < OCTEP_MMIO_REGIONS; i++)
>> -		iounmap(oct->mmio[i].hw_addr);
>> +	for (j = 0; j < i; j++)
>> +		iounmap(oct->mmio[j].hw_addr);
> 
> Assuming @i is not changed by the rest of the function is a bit fragile.
> 
> Better way of handling this situation is:
> 
> unsupported_dev:
> 	i = OCTEP_MMIO_REGIONS;
> unmap_prev:
> 	while (i--)
> 		iounmap(oct->mmio[i].hw_addr);
> 
> and jump to unmap_prev

Fine, I will submit a v2 to fix it.

Thanks,
Jiasheng