[PATCH] alx: Don't check netif_running() on resume

Tim Gardner tim.gardner at canonical.com
Wed Apr 24 21:07:08 EDT 2013


On 04/24/2013 02:09 PM, Huang, Xiong wrote:
>> Sent: Thursday, April 25, 2013 2:30 AM
>> To: unified-drivers at lists.infradead.org
>> Cc: Tim Gardner
>> Subject: [PATCH] alx: Don't check netif_running() on resume
>>
>> The resume function must power up the device regardless of the state of the
>> network stack.
>>
>> Signed-off-by: Tim Gardner <tim.gardner at canonical.com>
>> ---
>>  src/alx_main.c |    3 ---
>>  1 file changed, 3 deletions(-)
>>
>> diff --git a/src/alx_main.c b/src/alx_main.c index 7fe3f4b..c2606eb 100644
>> --- a/src/alx_main.c
>> +++ b/src/alx_main.c
>> @@ -1693,9 +1693,6 @@ static int alx_resume(struct device *dev)
>>  	struct alx_hw *hw = &adpt->hw;
>>  	int err;
>>
>> -	if (!netif_running(netdev))
>> -		return 0;
>> -
> 
> Hi Tim
> 
>      Alx_open will do same thing when this interface is 'up'. So just leave it as it, if this interface status is 'down' when resume.
> 
> Best Regards,
> Xiong
> 

Try this under Ubuntu and see how well it works for you. pm-suspend
works fine, but once you get the network stack involved via network
manager it doesn't work so well.

I looked at e1000 as an example.

rtg
-- 
Tim Gardner tim.gardner at canonical.com



More information about the unified-drivers mailing list