[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAHrpEqQdBqUwfrHL=dHGszEeU-pRo9mmOyuTy6aaw3QMBdkqUg@mail.gmail.com>
Date: Tue, 1 Mar 2016 16:26:15 -0600
From: Zhi Li <lznuaa@...il.com>
To: Troy Kisky <troy.kisky@...ndarydevices.com>
Cc: "netdev@...r.kernel.org" <netdev@...r.kernel.org>,
David Miller <davem@...emloft.net>,
Duan Fugang-B38611 <B38611@...escale.com>,
Fabio Estevam <fabio.estevam@...escale.com>,
Lucas Stach <l.stach@...gutronix.de>,
Andrew Lunn <andrew@...n.ch>,
Philippe Reynes <tremyfr@...il.com>,
Russell King - ARM Linux <linux@....linux.org.uk>,
"linux-arm-kernel@...ts.infradead.org"
<linux-arm-kernel@...ts.infradead.org>,
Laci Tele <laci@...ndarydevices.com>,
Shawn Guo <shawnguo@...nel.org>,
Johannes Berg <johannes@...solutions.net>,
Joshua Clayton <stillcompiling@...il.com>,
Sergei Shtylyov <sergei.shtylyov@...entembedded.com>,
Arnd Bergmann <arnd@...db.de>
Subject: Re: [PATCH net-next V2 03/16] net: fec: pass txq to fec_enet_tx_queue
instead of queue_id
On Tue, Mar 1, 2016 at 3:51 PM, Troy Kisky
<troy.kisky@...ndarydevices.com> wrote:
> True, but fec_txq/fec_rxq is called in a loop.
netdev_priv(ndev) is that pointer move.
dev + ALIGN(sizeof(struct net_device), NETDEV_ALIGN)
Modem compiler can handle it greatly.
You can't get any valuable performance gain by that.
The main time of CPU is call dma_map_xxx.
best regards
Frank Li
Powered by blists - more mailing lists