[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <49660774.6060603@hartkopp.net>
Date: Thu, 08 Jan 2009 15:02:28 +0100
From: Oliver Hartkopp <oliver@...tkopp.net>
To: Ben Hutchings <bhutchings@...arflare.com>
CC: Stephen Hemminger <shemminger@...tta.com>,
Linux Netdev List <netdev@...r.kernel.org>
Subject: Re: dev->get_stats(dev) in COMPAT_NET_DEV_OPS case
Ben Hutchings wrote:
> On Thu, 2009-01-08 at 14:12 +0100, Oliver Hartkopp wrote:
>
>> Hello Stephen,
>>
>> in commit eeda3fd64f75bcbfaa70ce946513abaf3f23b8e0 ("netdev: introduce dev_get_stats()")
>> you remove the setting of dev->get_stats :
>>
>> @@ -4370,7 +4384,6 @@ struct net_device *alloc_netdev_mq(int sizeof_priv, const char *name,
>>
>> netdev_init_queues(dev);
>>
>> - dev->get_stats = internal_stats;
>> netpoll_netdev_init(dev);
>> setup(dev);
>> strcpy(dev->name, name);
>>
>> I really appreciate your introduction of COMPAT_NET_DEV_OPS but when i
>> try to compile and run a non-converted driver, i get a NULL pointer
>> dereference. I tracked it down to be the call of dev->get_stats(dev).
>>
>> Did you miss something or is this behaviour intended?
>>
>
> You should be calling dev_get_stats() now.
>
>
I noticed that myself and besides a warning (about the const / non-const
definition) i got it working again.
But the question remains, if it still makes sense to provide
dev->get_stats() (referencing to dev_get_stats()) in the
COMPAT_NET_DEV_OPS case, as i would surely not be the only one tapping
into this problem ...
Regards,
Oliver
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Powered by blists - more mailing lists