netdev - Re: [PATCH net-next 2/3] bpf: Add new cgroups prog type to enable sock modifications

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

Message-ID: <1477450103.7065.175.camel@edumazet-glaptop3.roam.corp.google.com>
Date:   Tue, 25 Oct 2016 19:48:23 -0700
From:   Eric Dumazet <eric.dumazet@...il.com>
To:     David Ahern <dsa@...ulusnetworks.com>
Cc:     netdev@...r.kernel.org, daniel@...que.org, ast@...com,
        daniel@...earbox.net
Subject: Re: [PATCH net-next 2/3] bpf: Add new cgroups prog type to enable
 sock modifications

On Tue, 2016-10-25 at 20:21 -0600, David Ahern wrote:
> On 10/25/16 5:39 PM, Eric Dumazet wrote:
> > On Tue, 2016-10-25 at 15:30 -0700, David Ahern wrote:
> >> Add new cgroup based program type, BPF_PROG_TYPE_CGROUP_SOCK. Similar to
> >> BPF_PROG_TYPE_CGROUP_SKB programs can be attached to a cgroup and run
> >> any time a process in the cgroup opens an AF_INET or AF_INET6 socket.
> >> Currently only sk_bound_dev_if is exported to userspace for modification
> >> by a bpf program.
> >>
> >> This allows a cgroup to be configured such that AF_INET{6} sockets opened
> >> by processes are automatically bound to a specific device. In turn, this
> >> enables the running of programs that do not support SO_BINDTODEVICE in a
> >> specific VRF context / L3 domain.
> > 
> > Does this mean that these programs no longer can use loopback ?
> 
> I am probably misunderstanding your question, so I'll ramble a bit and
> see if I cover it.
> 
> This patch set generically allows sk_bound_dev_if to be set to any
> value. It does not check that an index corresponds to a device at that
> moment (either bpf prog install or execution of the filter), and even
> if it did the device can be deleted at any moment. That seems to be
> standard operating procedure with bpf filters (user mistakes mean
> packets go no where and in this case a socket is bound to a
> non-existent device).
> 
> The index can be any interface (e.g., eth0) or an L3 device (e.g., a
> VRF). Loopback and index=1 is allowed.
> 
> The VRF device is the loopback device for the domain, so binding to it
> covers addresses on the VRF device as well as interfaces enslaved to
> it.
> 
> Did you mean something else?

Maybe I do not understand how you plan to use this.

Let say you want a filter to force a BIND_TO_DEVICE xxx because a task
runs in a cgroup yyy

Then a program doing a socket() + connect (127.0.0.1)  will fail ?

I do not see how a BPF filter at socket() time can be selective.