Re: [PATCH net 1/1] net: devmem: prevent net-iov / page mixing

[Bobby Eshleman <bobbyeshleman@gmail.com>]

On Wed, Jul 22, 2026 at 07:58:27PM +0100, Pavel Begunkov wrote:
> On 7/22/26 18:59, Bobby Eshleman wrote:
> > On Wed, Jul 22, 2026 at 11:58:46AM +0100, Pavel Begunkov wrote:
> > > We should either have net_iov or page backed frags in a single skb,
> > > otherwise it blows up down the stack. Don't allow mixing in
> > > zerocopy_fill_skb_from_devmem().
> > > 
> > > Fixes: bd61848900bff ("net: devmem: Implement TX path")
> > > Cc: stable@vger.kernel.org
> > > Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
> > > ---
> > >   net/core/datagram.c | 3 +++
> > >   1 file changed, 3 insertions(+)
> > > 
> > > diff --git a/net/core/datagram.c b/net/core/datagram.c
> > > index c285c6465923..35febc1c25fa 100644
> > > --- a/net/core/datagram.c
> > > +++ b/net/core/datagram.c
> > > @@ -712,6 +712,9 @@ zerocopy_fill_skb_from_devmem(struct sk_buff *skb, struct iov_iter *from,
> > >   	size_t virt_addr, size, off;
> > >   	struct net_iov *niov;
> > > +	if (i && skb_frags_readable(skb))
> > > +		return -EEXIST;
> > 
> > Are we trying to hit the -EEXIST handler in tcp_sendmsg_locked() so that
> > we start a new skb? If so, I think we might need to plumb this -EEXIST
> > case through skb_zerocopy_iter_stream() too?
> 
> Easier to EFAULT. It'd more consistent, and I don't care how tcp takes
> it specifically.
> 
> -- 
> Pavel Begunkov
> 

One socket is allowed to queue non-devmem followed by devmem, and EFAULT
will break this case. I think EEXIST is right, just needs to propagate
up.

Best,
Bobby