netdev - Re: [PATCH 3/3] net: ethernet: bgmac: driver power manangement

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <74efc661-f07d-93d5-613d-7f563c0fe732@gmail.com>
Date:   Fri, 3 Feb 2017 18:16:43 -0800
From:   Florian Fainelli <f.fainelli@...il.com>
To:     Jon Mason <jon.mason@...adcom.com>,
        David Miller <davem@...emloft.net>
Cc:     rafal@...ecki.pl, bcm-kernel-feedback-list@...adcom.com,
        netdev@...r.kernel.org, linux-kernel@...r.kernel.org,
        Joey Zhong <zhongx@...adcom.com>
Subject: Re: [PATCH 3/3] net: ethernet: bgmac: driver power manangement

On 02/03/2017 01:39 PM, Jon Mason wrote:
> From: Joey Zhong <zhongx@...adcom.com>
> 
> Implements suspend/resume, external phy 54810 is assumed
> to remain powered up during deep-sleep for wake-on-lane.

s/wake-on-lane/Wake-on-LAN, are you positive phy_stop() is not
suspending the PHY and issuing BMCR_PWRDOWN write?

This also seems incomplete in that, if the device is really configured
for Wake-on-LAN (through ethtool) you should call
device_set_wakeup_capable() and then check for device_may_wakeup()
during suspend or resume to know which part of the suspend/resume
portion should be done. You could refer to bcmgenet for an example.

>  
> +int bgmac_enet_suspend(struct bgmac *bgmac)
> +{
> +	netdev_info(bgmac->net_dev, "Suspending\n");

remove that message

> +
> +	if (netif_running(bgmac->net_dev)) {
> +		netif_stop_queue(bgmac->net_dev);
> +
> +		napi_disable(&bgmac->napi);
> +
> +		netif_tx_lock(bgmac->net_dev);
> +		netif_device_detach(bgmac->net_dev);
> +		netif_tx_unlock(bgmac->net_dev);
> +
> +		bgmac_chip_intrs_off(bgmac);
> +		bgmac_chip_reset(bgmac);
> +		bgmac_dma_cleanup(bgmac);
> +	}

Can you change the indentation to test for netiff_running() first and
return 0 in that case?

> +
> +	phy_stop(bgmac->net_dev->phydev);
> +
> +	return 0;
> +}
> +EXPORT_SYMBOL_GPL(bgmac_enet_suspend);
> +
> +int bgmac_enet_resume(struct bgmac *bgmac)
> +{
> +	int rc;
> +
> +	netdev_info(bgmac->net_dev, "Resuming\n");

Same here, this needs to be removed.

> +
> +	phy_start(bgmac->net_dev->phydev);
> +
> +	if (netif_running(bgmac->net_dev)) {
> +		rc = bgmac_dma_init(bgmac);
> +		if (rc)
> +			return rc;
> +
> +		bgmac_chip_init(bgmac);
> +
> +		napi_enable(&bgmac->napi);
> +
> +		netif_tx_lock(bgmac->net_dev);
> +		netif_device_attach(bgmac->net_dev);
> +		netif_tx_unlock(bgmac->net_dev);
> +
> +		netif_start_queue(bgmac->net_dev);
> +	}
-- 
Florian