On Fri, Oct 4, 2024 at 11:19 AM Eric Dumazet <edumazet@xxxxxxxxxx> wrote:
>
> On Fri, Oct 4, 2024 at 12:05 PM Ignat Korchagin <ignat@xxxxxxxxxxxxxx> wrote:
> >
> > On Fri, Oct 4, 2024 at 9:55 AM Eric Dumazet <edumazet@xxxxxxxxxx> wrote:
> > >
> > > On Thu, Oct 3, 2024 at 11:50 PM Kuniyuki Iwashima <kuniyu@xxxxxxxxxx> wrote:
> > > >
> > > > From: Ignat Korchagin <ignat@xxxxxxxxxxxxxx>
> > > > Date: Thu, 3 Oct 2024 18:01:51 +0100
> > > > > We have recently noticed the exact same KASAN splat as in commit
> > > > > 6cd4a78d962b ("net: do not leave a dangling sk pointer, when socket
> > > > > creation fails"). The problem is that commit did not fully address the
> > > > > problem, as some pf->create implementations do not use sk_common_release
> > > > > in their error paths.
> > > > >
> > > > > For example, we can use the same reproducer as in the above commit, but
> > > > > changing ping to arping. arping uses AF_PACKET socket and if packet_create
> > > > > fails, it will just sk_free the allocated sk object.
> > > > >
> > > > > While we could chase all the pf->create implementations and make sure they
> > > > > NULL the freed sk object on error from the socket, we can't guarantee
> > > > > future protocols will not make the same mistake.
> > > > >
> > > > > So it is easier to just explicitly NULL the sk pointer upon return from
> > > > > pf->create in __sock_create. We do know that pf->create always releases the
> > > > > allocated sk object on error, so if the pointer is not NULL, it is
> > > > > definitely dangling.
> > > >
> > > > Sounds good to me.
> > > >
> > > > Let's remove the change by 6cd4a78d962b that should be unnecessary
> > > > with this patch.
> > > >
> > >
> > > Reviewed-by: Eric Dumazet <edumazet@xxxxxxxxxx>
> > >
> > > Even if not strictly needed we also could fix af_packet to avoid a
> > > dangling pointer.
> >
> > af_packet was just one example - I reviewed every pf->create function
> > and there are others. It would not be fair to fix this, but not the
> > others, right?
>
> I have not said your patch was not correct, I gave a +2 on it.
>
> In general, leaving pointers to a freed piece of memory (and possibly reused)
> can confuse things like kmemleak.
That's a good point actually.
> I have not said _you_ had to review all pf->create() functions.
Ah, NP. I reviewed them before your comment, before submitting the
patch - basically to decide whether I should go with the current
approach or just go and fix them.
