Re: [PATCH net-next] net: fix up truesize of cloned skb in skb_prepare_for_shift()

[Marco Elver]

On Tue, 2 Feb 2021 at 18:59, Eric Dumazet <edumazet@xxxxxxxxxx> wrote:
>
> On Mon, Feb 1, 2021 at 5:04 PM Marco Elver <elver@xxxxxxxxxx> wrote:
> >
> > Avoid the assumption that ksize(kmalloc(S)) == ksize(kmalloc(S)): when
> > cloning an skb, save and restore truesize after pskb_expand_head(). This
> > can occur if the allocator decides to service an allocation of the same
> > size differently (e.g. use a different size class, or pass the
> > allocation on to KFENCE).
> >
> > Because truesize is used for bookkeeping (such as sk_wmem_queued), a
> > modified truesize of a cloned skb may result in corrupt bookkeeping and
> > relevant warnings (such as in sk_stream_kill_queues()).
> >
> > Link: https://lkml.kernel.org/r/X9JR/J6dMMOy1obu@xxxxxxxxxxxxxxxx
> > Reported-by: syzbot+7b99aafdcc2eedea6178@xxxxxxxxxxxxxxxxxxxxxxxxx
> > Suggested-by: Eric Dumazet <edumazet@xxxxxxxxxx>
> > Signed-off-by: Marco Elver <elver@xxxxxxxxxx>
>
> Signed-off-by: Eric Dumazet <edumazet@xxxxxxxxxx>

Thank you!