Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <1c3facf9-1ecf-f090-e800-803d83a89e4b@ti.com>
Date:   Thu, 15 Sep 2022 14:29:00 +0530
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:91.0) Gecko/20100101
 Thunderbird/91.11.0
CC:     <davem@davemloft.net>, <edumazet@google.com>, <kuba@kernel.org>,
        <pabeni@redhat.com>, <robh+dt@kernel.org>,
        <krzysztof.kozlowski@linaro.org>,
        <krzysztof.kozlowski+dt@linaro.org>, <vladimir.oltean@nxp.com>,
        <grygorii.strashko@ti.com>, <vigneshr@ti.com>, <nsekhar@ti.com>,
        <netdev@vger.kernel.org>, <devicetree@vger.kernel.org>,
        <linux-kernel@vger.kernel.org>,
        <linux-arm-kernel@lists.infradead.org>, <kishon@ti.com>,
        <s-vadapalli@ti.com>
Subject: Re: [PATCH 5/8] net: ethernet: ti: am65-cpsw: Add support for
 fixed-link configuration
Content-Language: en-US
To:     "Russell King (Oracle)" <linux@armlinux.org.uk>
References: <20220914095053.189851-1-s-vadapalli@ti.com>
 <20220914095053.189851-6-s-vadapalli@ti.com>
 <YyH2GcLCAN+9GAn8@shell.armlinux.org.uk>
From:   Siddharth Vadapalli <s-vadapalli@ti.com>
In-Reply-To: <YyH2GcLCAN+9GAn8@shell.armlinux.org.uk>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 7bit
Precedence: bulk

Hello Russell,

On 14/09/22 21:11, Russell King (Oracle) wrote:
> On Wed, Sep 14, 2022 at 03:20:50PM +0530, Siddharth Vadapalli wrote:
>> diff --git a/drivers/net/ethernet/ti/am65-cpsw-nuss.c b/drivers/net/ethernet/ti/am65-cpsw-nuss.c
>> index 72b1df12f320..1739c389af20 100644
>> --- a/drivers/net/ethernet/ti/am65-cpsw-nuss.c
>> +++ b/drivers/net/ethernet/ti/am65-cpsw-nuss.c
>> @@ -1494,10 +1494,50 @@ static void am65_cpsw_nuss_mac_config(struct phylink_config *config, unsigned in
>>  							  phylink_config);
>>  	struct am65_cpsw_port *port = container_of(slave, struct am65_cpsw_port, slave);
>>  	struct am65_cpsw_common *common = port->common;
>> +	struct fwnode_handle *fwnode;
>> +	bool fixed_link = false;
>>  
>>  	if (common->pdata.extra_modes & BIT(state->interface))
>>  		writel(AM65_CPSW_SGMII_CONTROL_MR_AN_ENABLE,
>>  		       port->sgmii_base + AM65_CPSW_SGMII_CONTROL_REG);
>> +
>> +	/* Detecting fixed-link */
>> +	fwnode = of_node_to_fwnode(port->slave.phy_node);
>> +	if (fwnode)
>> +		fixed_link = !!fwnode_get_named_child_node(fwnode, "fixed-link");
>> +
>> +	if (fixed_link) {
>> +		/* In fixed-link mode, mac_link_up is not invoked.
>> +		 * Therefore, the relevant mac_link_up operations
>> +		 * have to be moved to mac_config.
>> +		 */
> 
> This seems very wrong. Why is mac_link_up() not invoked? Have you
> debugged this? It works for other people.
> 
> Please debug rather than adding hacks to drivers when you find
> things that don't seem to work.

I will debug and find out. I had assumed that mac_link_up() is not
invoked in fixed-link mode. Thank you for clarifying.

Regards,
Siddharth.