[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Mon, 10 Oct 2016 06:33:05 +0000
From: "Mintz, Yuval" <Yuval.Mintz@...ium.com>
To: Eric Dumazet <eric.dumazet@...il.com>,
"davem@...emloft.net" <davem@...emloft.net>
CC: "netdev@...r.kernel.org" <netdev@...r.kernel.org>
Subject: Re: [PATCH net-next 5/6] qed: Allow chance for fast ramrod
completions
> > + while (iter_cnt--) {
> > + /* Validate we receive completion update */
> > smp_rmb();
> > if (comp_done->done == 1) {
> > if (p_fw_ret)
> > *p_fw_ret = comp_done->fw_return_code;
> > return 0;
> > }
> Note that this smp_rmb() and accesses to ->done and ->fw_return_code are
> racy.
> fw_return_code needs to be written _before_ done.
Thanks for catching this up.
Dave - do you want me to re-spin for this?
I believe it's a day-1 issue [not introduced by this series],
and merits its own patch [and not incorporated into this one].
Still, if you'd like me to re-spin and include Eric's fix, I'll do it.
Powered by blists - more mailing lists