[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <5b228baf-078c-0053-e99b-6d6969b304b1@iogearbox.net>
Date: Thu, 11 Oct 2018 10:22:20 +0200
From: Daniel Borkmann <daniel@...earbox.net>
To: Björn Töpel <bjorn.topel@...il.com>,
ast@...nel.org, netdev@...r.kernel.org, eric.dumazet@...il.com
Cc: Björn Töpel <bjorn.topel@...el.com>,
magnus.karlsson@...el.com, magnus.karlsson@...il.com
Subject: Re: [PATCH bpf] xsk: do not call synchronize_net() under RCU read
lock
On 10/08/2018 07:40 PM, Björn Töpel wrote:
> From: Björn Töpel <bjorn.topel@...el.com>
>
> The XSKMAP update and delete functions called synchronize_net(), which
> can sleep. It is not allowed to sleep during an RCU read section.
>
> Instead we need to make sure that the sock sk_destruct (xsk_destruct)
> function is asynchronously called after an RCU grace period. Setting
> the SOCK_RCU_FREE flag for XDP sockets takes care of this.
>
> Fixes: fbfc504a24f5 ("bpf: introduce new bpf AF_XDP map type BPF_MAP_TYPE_XSKMAP")
> Reported-by: Eric Dumazet <eric.dumazet@...il.com>
> Signed-off-by: Björn Töpel <bjorn.topel@...el.com>
Applied to bpf, thanks everyone!
Powered by blists - more mailing lists