Re: [PATCH net-next 3/3] net: allow to leave the buffer fragmented in skb_cow_data()

[Mathias Krause <mathias.krause@secunet.com>]

On 06.11.2013 17:39, Eric Dumazet wrote:
> On Wed, 2013-11-06 at 20:48 +0800, Herbert Xu wrote:
>> On Wed, Nov 06, 2013 at 01:42:03PM +0100, Mathias Krause wrote:
>>> Well, skb_cow_data() will only copy, i.e. call __pskb_pull_tail(), in
>>> case the skb is either cloned or fragmented. As you already said it
>>> won't be cloned in your case. Does it contain fragments, i.e. is
>>> skb_shinfo(skb)->nr_frags != 0? If not, we won't copy with the current
>>> code either.
>> Whenever we say page it means nr_frags != 0.  So currently as
>> long as we have pages in our skb we will copy.  With your patch
>> we will no longer copy in the case where we have pages but the
>> skb isn't cloned.  In fact that is the whole point of your patch.
>>
>>> Can you please explain why this would be needed? I still don't get the
>>> reasoning behind "pages are considered not writable at the moment even
>>> if they are anonymous".
>> As I said you don't know where the page in the skb came from.  It
>> may point to read-only memory or memory that's shared with another
>> task that isn't expecting things to change underneath it.
> 
> Note that we might have now a per skb flag telling is all page frags are
> owned.
> 
> This is SKBTX_SHARED_FRAG
> 
> If SKBTX_SHARED_FRAG is set on shared_info->tx_flags, then at least
> one frag is not safe and we must copy all frags.
> 
> 
> For example, this flag is set in TCP sendfile() path (vmsplice()...),
> and zero copy paths in general (macvtap, tun)
> 
> I am not 100% sure, but this could be a hint.

Thanks for the hint, Eric!


Mathias