[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <63385120-46C9-4D9E-81C7-9E72C3371C2C@net-swift.com>
Date: Sun, 11 Dec 2022 19:53:41 +0800
From: "mengyuanlou@...-swift.com" <mengyuanlou@...-swift.com>
To: Andrew Lunn <andrew@...n.ch>
Cc: netdev@...r.kernel.org, jiawenwu@...stnetic.com
Subject: Re: [PATCH net-next v2] net: ngbe: Add ngbe mdio bus driver.
> 2022年12月11日 19:28,Andrew Lunn <andrew@...n.ch> 写道:
>
>>>> +static void ngbe_up(struct ngbe_adapter *adapter)
>>>> +{
>>>> + struct ngbe_hw *hw = &adapter->hw;
>>>> +
>>>> + pci_set_master(adapter->pdev);
>>>> + if (hw->gpio_ctrl)
>>>> + /* gpio0 is used to power on control*/
>>>> + wr32(&hw->wxhw, NGBE_GPIO_DR, 0);
>>>
>>> Control of what?
>> Control for sfp modules power down/up.
>> The chip has not i2c interface, so I do not use phylink.
>
> Please give this a better name. Is this connected to the TX_DISABLE
> pin of the SFP, or the transmit and receiver power pins?
>
The transmit and receiver power pins.
> If you don't have the I2C bus, i'm wondering how you can actually
> driver the SFP and the MAC. You have no idea what has been
> inserted. Is it actually an SFF, not an SFP? Do you have any of the
> GPIOs normally associated with an SFP? TX_DISABLE, LOS, TX_FAULT?
>
1、Yeah. We can't know what module is inserted
2、There are not enough GPIOs to be associated with other pins.
>
> Andrew
>
Powered by blists - more mailing lists