All of lore.kernel.org
 help / color / mirror / Atom feed
From: Jakub Kicinski <kuba@kernel.org>
To: Joakim Zhang <qiangqing.zhang@nxp.com>
Cc: peppe.cavallaro@st.com, alexandre.torgue@st.com,
	joabreu@synopsys.com, davem@davemloft.net, f.fainelli@gmail.com,
	andrew@lunn.ch, Jisheng.Zhang@synaptics.com,
	netdev@vger.kernel.org, linux-imx@nxp.com
Subject: Re: [PATCH V4 net] net: stmmac: Fix MAC WoL not working if PHY does not support WoL
Date: Thu, 6 May 2021 17:55:22 -0700	[thread overview]
Message-ID: <20210506175522.49a2ad5b@kicinski-fedora-pc1c0hjn.dhcp.thefacebook.com> (raw)
In-Reply-To: <20210506050658.9624-1-qiangqing.zhang@nxp.com>

On Thu,  6 May 2021 13:06:58 +0800 Joakim Zhang wrote:
> Both get and set WoL will check device_can_wakeup(), if MAC supports PMT,
> it will set device wakeup capability. After commit 1d8e5b0f3f2c ("net:
> stmmac: Support WOL with phy"), device wakeup capability will be overwrite
> in stmmac_init_phy() according to phy's Wol feature. If phy doesn't support
> WoL, then MAC will lose wakeup capability.

Let's take a step back. Can we get a minimal fix for losing the config
in stmmac_init_phy(), and then extend the support for WoL for devices
which do support wake up themselves?

> This patch combines WoL capabilities both MAC and PHY from stmmac_get_wol(),
> set wakeup capability and give WoL priority to the PHY in stmmac_set_wol()
> when enable WoL. What PHYs do implement is WAKE_MAGIC, WAKE_UCAST, WAKE_MAGICSECURE
> and WAKE_BCAST.
> 
> Fixes: commit 1d8e5b0f3f2c ("net: stmmac: Support WOL with phy")

Please remove "commit" from the fixes tag.

> Suggested-by: Andrew Lunn <andrew@lunn.ch>
> Signed-off-by: Joakim Zhang <qiangqing.zhang@nxp.com>

> diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c
> index c5642985ef95..6d09908dec1f 100644
> --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c
> +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c
> @@ -629,35 +629,49 @@ static void stmmac_get_strings(struct net_device *dev, u32 stringset, u8 *data)
>  /* Currently only support WOL through Magic packet. */
>  static void stmmac_get_wol(struct net_device *dev, struct ethtool_wolinfo *wol)
>  {
> +	struct ethtool_wolinfo wol_phy = { .cmd = ETHTOOL_GWOL };
>  	struct stmmac_priv *priv = netdev_priv(dev);
>  
> -	if (!priv->plat->pmt)
> -		return phylink_ethtool_get_wol(priv->phylink, wol);
> -
>  	mutex_lock(&priv->lock);
> -	if (device_can_wakeup(priv->device)) {

Why remove the device_can_wakeup() check?

> +	if (priv->plat->pmt) {
>  		wol->supported = WAKE_MAGIC | WAKE_UCAST;
>  		if (priv->hw_cap_support && !priv->dma_cap.pmt_magic_frame)
>  			wol->supported &= ~WAKE_MAGIC;
> -		wol->wolopts = priv->wolopts;
>  	}
> +
> +	phylink_ethtool_get_wol(priv->phylink, &wol_phy);
> +
> +	/* Combine WoL capabilities both PHY and MAC */
> +	wol->supported |= wol_phy.supported;
> +	wol->wolopts = priv->wolopts;
> +
>  	mutex_unlock(&priv->lock);
>  }
>  
>  static int stmmac_set_wol(struct net_device *dev, struct ethtool_wolinfo *wol)
>  {
> +	u32 support = WAKE_MAGIC | WAKE_UCAST | WAKE_MAGICSECURE | WAKE_BCAST;

Why this list?

> +	struct ethtool_wolinfo wol_phy = { .cmd = ETHTOOL_GWOL };
>  	struct stmmac_priv *priv = netdev_priv(dev);
> -	u32 support = WAKE_MAGIC | WAKE_UCAST;

This list was the list of what the MAC supported, right?

> +	int ret;
>  
> -	if (!device_can_wakeup(priv->device))

Why remove this check?

> +	if (wol->wolopts & ~support)
>  		return -EOPNOTSUPP;
>  
> -	if (!priv->plat->pmt) {
> -		int ret = phylink_ethtool_set_wol(priv->phylink, wol);
> -
> -		if (!ret)
> -			device_set_wakeup_enable(priv->device, !!wol->wolopts);
> -		return ret;
> +	/* First check if can WoL from PHY */
> +	phylink_ethtool_get_wol(priv->phylink, &wol_phy);
> +	if (wol->wolopts & wol_phy.supported) {

So if _any_ requests match PHY capabilities we'd use PHY?
I think the check should be:

	if ((wol->woltops & wol_phy.supported) == wol->woltops)

That said I'm not 100% sure what the semantics for WoL are.

> +		wol->wolopts &= wol_phy.supported;
> +		ret = phylink_ethtool_set_wol(priv->phylink, wol);
> +
> +		if (!ret) {
> +			pr_info("stmmac: phy wakeup enable\n");
> +			device_set_wakeup_capable(priv->device, 1);
> +			device_set_wakeup_enable(priv->device, 1);
> +			goto wolopts_update;
> +		} else {
> +			return ret;
> +		}
>  	}
>  
>  	/* By default almost all GMAC devices support the WoL via
> @@ -666,18 +680,21 @@ static int stmmac_set_wol(struct net_device *dev, struct ethtool_wolinfo *wol)
>  	if ((priv->hw_cap_support) && (!priv->dma_cap.pmt_magic_frame))
>  		wol->wolopts &= ~WAKE_MAGIC;
>  
> -	if (wol->wolopts & ~support)
> -		return -EINVAL;

Now you seem to not validate against MAC capabilities anywhere.

> -	if (wol->wolopts) {
> -		pr_info("stmmac: wakeup enable\n");
> +	if (priv->plat->pmt && wol->wolopts) {
> +		pr_info("stmmac: mac wakeup enable\n");
> +		device_set_wakeup_capable(priv->device, 1);
>  		device_set_wakeup_enable(priv->device, 1);
>  		enable_irq_wake(priv->wol_irq);
> -	} else {
> +		goto wolopts_update;
> +	}
> +
> +	if (!wol->wolopts) {
> +		device_set_wakeup_capable(priv->device, 0);
>  		device_set_wakeup_enable(priv->device, 0);
>  		disable_irq_wake(priv->wol_irq);
>  	}
>  
> +wolopts_update:
>  	mutex_lock(&priv->lock);
>  	priv->wolopts = wol->wolopts;
>  	mutex_unlock(&priv->lock);


  reply	other threads:[~2021-05-07  0:55 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-05-06  5:06 [PATCH V4 net] net: stmmac: Fix MAC WoL not working if PHY does not support WoL Joakim Zhang
2021-05-07  0:55 ` Jakub Kicinski [this message]
2021-05-07 10:59   ` Joakim Zhang
2021-05-07 22:46     ` Jakub Kicinski
2021-05-08  9:05       ` Jisheng Zhang
2021-05-08 18:14       ` Andrew Lunn

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20210506175522.49a2ad5b@kicinski-fedora-pc1c0hjn.dhcp.thefacebook.com \
    --to=kuba@kernel.org \
    --cc=Jisheng.Zhang@synaptics.com \
    --cc=alexandre.torgue@st.com \
    --cc=andrew@lunn.ch \
    --cc=davem@davemloft.net \
    --cc=f.fainelli@gmail.com \
    --cc=joabreu@synopsys.com \
    --cc=linux-imx@nxp.com \
    --cc=netdev@vger.kernel.org \
    --cc=peppe.cavallaro@st.com \
    --cc=qiangqing.zhang@nxp.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.