netdev - Re: [PATCH] net: core: set skb useful vars in __bpf_tx

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <CAMDZJNUvdk+hLvciYdOAsdpCzRG=j-7HrR5sDv0agTS+ehVkeQ@mail.gmail.com>
Date:   Thu, 4 Nov 2021 09:31:01 +0800
From:   Tonghao Zhang <xiangxia.m.yue@...il.com>
To:     Daniel Borkmann <daniel@...earbox.net>
Cc:     Linux Kernel Network Developers <netdev@...r.kernel.org>,
        Jakub Kicinski <kuba@...nel.org>
Subject: Re: [PATCH] net: core: set skb useful vars in __bpf_tx_skb

On Tue, Nov 2, 2021 at 9:58 AM Tonghao Zhang <xiangxia.m.yue@...il.com> wrote:
>
> On Tue, Nov 2, 2021 at 5:46 AM Daniel Borkmann <daniel@...earbox.net> wrote:
> >
> > On 10/29/21 3:54 AM, xiangxia.m.yue@...il.com wrote:
> > [...]
> > > diff --git a/net/core/filter.c b/net/core/filter.c
> > > index 4bace37a6a44..2dbff0944768 100644
> > > --- a/net/core/filter.c
> > > +++ b/net/core/filter.c
> > > @@ -2107,9 +2107,19 @@ static inline int __bpf_tx_skb(struct net_device *dev, struct sk_buff *skb)
> > >               return -ENETDOWN;
> > >       }
> > >
> > > -     skb->dev = dev;
> > > +     /* The target netdevice (e.g. ifb) may use the:
> > > +      * - skb_iif
> > > +      * - redirected
> > > +      * - from_ingress
> > > +      */
> > > +     skb->skb_iif = skb->dev->ifindex;
> >
> > This doesn't look right to me to set it unconditionally in tx path, isn't ifb_ri_tasklet()
> > setting skb->skb_iif in this case (or __netif_receive_skb_core() in main rx path)?
> Hi
> the act_mirred set the skb->skb_iif, redirected and from_ingress. and
> __netif_receive_skb_core also set skb->skb_iif.
> so we can use the act_mirred to ifb in ingress or egress path.
> For ingress, when we use the bpf_redirct to ifb, we should set
> redirected, and from_ingress.
> For egress, when we use the bpf_redirct to ifb, we should skb_iif ,
> set redirected, and from_ingress.
Hi Daniel,
because we don't know bpf_redirct invoked in tx path, or rx path.
can we set the  skb->skb_iif unconditionally in bpf_redirct? any thoughts?

> > Also, I would suggest to add a proper BPF selftest which outlines the issue you're solving
> > in here.
> Ok, thanks.
> > > +#ifdef CONFIG_NET_CLS_ACT
> > > +     skb_set_redirected(skb, skb->tc_at_ingress);
> > > +#else
> > >       skb->tstamp = 0;
> > > +#endif
> > >
> > > +     skb->dev = dev;
> > >       dev_xmit_recursion_inc();
> > >       ret = dev_queue_xmit(skb);
> > >       dev_xmit_recursion_dec();
> > >
> >
>
>
> --
> Best regards, Tonghao



-- 
Best regards, Tonghao