[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <0193456e-0acb-75d6-8c6f-be0917990708@nbd.name>
Date: Tue, 22 Nov 2022 10:41:28 +0100
From: Felix Fietkau <nbd@....name>
To: Jakub Kicinski <kuba@...nel.org>,
Lorenzo Bianconi <lorenzo@...nel.org>
Cc: netdev@...r.kernel.org, john@...ozen.org, sean.wang@...iatek.com,
Mark-MC.Lee@...iatek.com, davem@...emloft.net, edumazet@...gle.com,
pabeni@...hat.com, lorenzo.bianconi@...hat.com,
sujuan.chen@...iatek.com, linux-mediatek@...ts.infradead.org
Subject: Re: [PATCH net-next 5/5] net: ethernet: mtk_wed: add reset to
tx_ring_setup callback
On 22.11.22 05:19, Jakub Kicinski wrote:
> On Mon, 21 Nov 2022 22:18:33 +0100 Lorenzo Bianconi wrote:
>> > On Mon, 21 Nov 2022 09:59:25 +0100 Lorenzo Bianconi wrote:
>> > > +#define mtk_wed_device_tx_ring_setup(_dev, _ring, _regs, _reset) \
>> > > + (_dev)->ops->tx_ring_setup(_dev, _ring, _regs, _reset)
>> >
>> > FWIW I find the "op macros" quite painful when trying to read a driver
>> > I'm not familiar with. stmmac does this, too. Just letting you know,
>> > it is what it is.
>>
>> ack, fine. I maintained the approach currently used in the driver.
>> Do you prefer to run the function pointer directly?
>
> That's a tiny bit better, yes, saves the reader one lookup.
>
> Are the ops here serving as a HAL or a way of breaking the dependency
> between the SoC/Eth and the WiFi drivers?
The latter. For a multi-platform kernel it's important that the wifi
driver does not depend on mtk_eth_soc directly, even when support for
WED is enabled.
- Felix
Powered by blists - more mailing lists