netdev - Re: [PATCH net-next] tuntap: introduce tx skb ring

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:	Wed, 18 May 2016 07:52:06 -0400
From:	Steven Rostedt <rostedt@...dmis.org>
To:	Jason Wang <jasowang@...hat.com>
Cc:	Jesper Dangaard Brouer <brouer@...hat.com>,
	Eric Dumazet <eric.dumazet@...il.com>, davem@...emloft.net,
	mst@...hat.com, netdev@...r.kernel.org,
	linux-kernel@...r.kernel.org, Peter Zijlstra <peterz@...radead.org>
Subject: Re: [PATCH net-next] tuntap: introduce tx skb ring

On Wed, 18 May 2016 18:23:48 +0800
Jason Wang <jasowang@...hat.com> wrote:

>
> >
> >
> > Maybe Steven Rostedt have an even better ring queue implementation
> > already avail in the kernel?
> >  
> 
> You mean ring buffer in tracing? Not sure, but it looks rather complex 
> at first glance.

Yes it is, and I'm not sure it would be appropriate here or not. The
tracing ring buffer is highly tuned to be lockless and per cpu
(allocates a buffer per cpu), it also does not need to disable
interrupts and can be used in NMI context.

The complexity comes from being able to locklessly swap out pages from
the ring buffer to send across the network or to disk while a record is
happening.

Perf has a simpler ring buffer made for mmapping, but it's still rather
coupled with the perf core. It would be nice to make that ring buffer a
bit more generic and not so tied to perf itself.

-- Steve