netdev - Re: [PATCH net-next 1/8] bpf: Recursively apply cgroup sock filters

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]

Date:   Tue, 22 Aug 2017 18:40:07 -0700
From:   Alexei Starovoitov <alexei.starovoitov@...il.com>
To:     David Ahern <dsahern@...il.com>
Cc:     netdev@...r.kernel.org, daniel@...earbox.net, ast@...nel.org,
        tj@...nel.org, davem@...emloft.net
Subject: Re: [PATCH net-next 1/8] bpf: Recursively apply cgroup sock filters

On Tue, Aug 22, 2017 at 05:20:14PM -0700, David Ahern wrote:
> Recursively apply sock filters attached to a cgroup. For now, start
> with the inner cgroup attached to the socket and work back to the
> root. If desired the inverse can be done use an attach flag (start
> with parent cgroup and go in).
> 
> Signed-off-by: David Ahern <dsahern@...il.com>
> ---
>  include/linux/bpf-cgroup.h |  5 +++--
>  kernel/bpf/cgroup.c        |  4 +---
>  kernel/cgroup/cgroup.c     | 18 ++++++++++++++++++
>  3 files changed, 22 insertions(+), 5 deletions(-)
> 
> diff --git a/include/linux/bpf-cgroup.h b/include/linux/bpf-cgroup.h
> index d41d40ac3efd..d95e44ccd549 100644
> --- a/include/linux/bpf-cgroup.h
> +++ b/include/linux/bpf-cgroup.h
> @@ -40,8 +40,9 @@ int __cgroup_bpf_run_filter_skb(struct sock *sk,
>  				struct sk_buff *skb,
>  				enum bpf_attach_type type);
>  
> -int __cgroup_bpf_run_filter_sk(struct sock *sk,
> +int __cgroup_bpf_run_filter_sk(struct cgroup *cgrp, struct sock *sk,
>  			       enum bpf_attach_type type);
> +int cgroup_bpf_run_filter_sk(struct sock *sk, enum bpf_attach_type type);
>  
>  int __cgroup_bpf_run_filter_sock_ops(struct sock *sk,
>  				     struct bpf_sock_ops_kern *sock_ops,
> @@ -74,7 +75,7 @@ int __cgroup_bpf_run_filter_sock_ops(struct sock *sk,
>  ({									       \
>  	int __ret = 0;							       \
>  	if (cgroup_bpf_enabled && sk) {					       \
> -		__ret = __cgroup_bpf_run_filter_sk(sk,			       \
> +		__ret = cgroup_bpf_run_filter_sk(sk,			       \
>  						 BPF_CGROUP_INET_SOCK_CREATE); \
>  	}								       \
>  	__ret;								       \
> diff --git a/kernel/bpf/cgroup.c b/kernel/bpf/cgroup.c
> index 546113430049..0480610bda83 100644
> --- a/kernel/bpf/cgroup.c
> +++ b/kernel/bpf/cgroup.c
> @@ -217,14 +217,12 @@ EXPORT_SYMBOL(__cgroup_bpf_run_filter_skb);
>   * This function will return %-EPERM if any if an attached program was found
>   * and if it returned != 1 during execution. In all other cases, 0 is returned.
>   */
> -int __cgroup_bpf_run_filter_sk(struct sock *sk,
> +int __cgroup_bpf_run_filter_sk(struct cgroup *cgrp, struct sock *sk,
>  			       enum bpf_attach_type type)
>  {
> -	struct cgroup *cgrp = sock_cgroup_ptr(&sk->sk_cgrp_data);
>  	struct bpf_prog *prog;
>  	int ret = 0;
>  
> -
>  	rcu_read_lock();
>  
>  	prog = rcu_dereference(cgrp->bpf.effective[type]);
> diff --git a/kernel/cgroup/cgroup.c b/kernel/cgroup/cgroup.c
> index df2e0f14a95d..7480cebab073 100644
> --- a/kernel/cgroup/cgroup.c
> +++ b/kernel/cgroup/cgroup.c
> @@ -5186,4 +5186,22 @@ int cgroup_bpf_update(struct cgroup *cgrp, struct bpf_prog *prog,
>  	mutex_unlock(&cgroup_mutex);
>  	return ret;
>  }
> +
> +int cgroup_bpf_run_filter_sk(struct sock *sk,
> +			     enum bpf_attach_type type)
> +{
> +	struct cgroup *cgrp = sock_cgroup_ptr(&sk->sk_cgrp_data);
> +	int ret = 0;
> +
> +	while (cgrp) {
> +		ret = __cgroup_bpf_run_filter_sk(cgrp, sk, type);
> +		if (ret < 0)
> +			break;
> +
> +		cgrp = cgroup_parent(cgrp);
> +	}

I think this walk changes semantics for existing setups, so we cannot do it
by default and have to add new attach flag.
Also why break on (ret < 0) ?
The caller of this does:
  err = BPF_CGROUP_RUN_PROG_INET_SOCK(sk);
  if (err) {
          sk_common_release(sk);
so we should probably break out of the loop on if (ret) too.