[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20170209173441.1f1eb96e@cakuba.netronome.com>
Date: Thu, 9 Feb 2017 17:34:41 -0800
From: Jakub Kicinski <jakub.kicinski@...ronome.com>
To: Or Gerlitz <ogerlitz@...lanox.com>
Cc: "David S. Miller" <davem@...emloft.net>,
Amir Vadai <amir@...ai.me>, Ido Schimmel <idosch@...lanox.com>,
Daniel Borkmann <daniel@...earbox.net>,
Roi Dayan <roid@...lanox.com>, netdev@...r.kernel.org
Subject: Re: [PATCH RFC net] net/mlx5e: Add preemption enable/disable around
TC statistics upcall
On Thu, 9 Feb 2017 17:38:43 +0200, Or Gerlitz wrote:
> Running with CONFIG_PREEMPT set, I get a
>
> BUG: using smp_processor_id() in preemptible [00000000] code: tc/3793
>
> asserion from the TC action (mirred) stats_update callback, when the do
>
> _bstats_cpu_update(this_cpu_ptr(a->cpu_bstats), bytes, packets)
>
> As done by commit 66860be "nfp: bpf: allow offloaded filters to update stats",
> disabling/enabling preemption around the TC upcall solves that.
>
> Fixes: aad7e08d39bd ('net/mlx5e: Hardware offloaded flower filter statistics support')
> Signed-off-by: Or Gerlitz <ogerlitz@...lanox.com>
> ---
>
> I marked it as RFC, since I wasn't fully sure on the nature of the
> problem, nor if this is the direction we should take to the fix.
I think it's the right fix, for net-next we could perhaps redo the
tcf_action_stats_update() helper so that it takes care of preemption and
the iteration so more people don't trip over this?
Powered by blists - more mailing lists