[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20210903201210.GF1350@shell.armlinux.org.uk>
Date: Fri, 3 Sep 2021 21:12:10 +0100
From: Russell King - ARM Linux admin <linux@...linux.org.uk>
To: Joakim Zhang <qiangqing.zhang@....com>
Cc: Andrew Lunn <andrew@...n.ch>, Vladimir Oltean <olteanv@...il.com>,
"peppe.cavallaro@...com" <peppe.cavallaro@...com>,
"alexandre.torgue@...s.st.com" <alexandre.torgue@...s.st.com>,
"joabreu@...opsys.com" <joabreu@...opsys.com>,
"davem@...emloft.net" <davem@...emloft.net>,
"kuba@...nel.org" <kuba@...nel.org>,
"mcoquelin.stm32@...il.com" <mcoquelin.stm32@...il.com>,
"netdev@...r.kernel.org" <netdev@...r.kernel.org>,
"f.fainelli@...il.com" <f.fainelli@...il.com>,
"hkallweit1@...il.com" <hkallweit1@...il.com>,
dl-linux-imx <linux-imx@....com>
Subject: Re: [PATCH] net: stmmac: fix MAC not working when system resume back
with WoL enabled
Hi,
Here's a patch to try - you'll need to integrate the new calls into
stmmac's suspend and resume hooks. Obviously, given my previous
comments, this isn't tested!
I didn't need to repeat the mac_wol boolean to phylink_resume as we
can record the state internally - mac_wol should not change between
a call to phylink_suspend() and subsequent phylink_resume() anyway.
mac_wol should only be true if the MAC is involved in processing
packets for WoL, false otherwise.
Please let me know if this resolves your stmmac WoL issue.
Thanks.
diff --git a/drivers/net/phy/phylink.c b/drivers/net/phy/phylink.c
index f0c769027145..c4d0de04416a 100644
--- a/drivers/net/phy/phylink.c
+++ b/drivers/net/phy/phylink.c
@@ -33,6 +33,7 @@
enum {
PHYLINK_DISABLE_STOPPED,
PHYLINK_DISABLE_LINK,
+ PHYLINK_DISABLE_MAC_WOL,
};
/**
@@ -1313,6 +1314,9 @@ EXPORT_SYMBOL_GPL(phylink_start);
* network device driver's &struct net_device_ops ndo_stop() method. The
* network device's carrier state should not be changed prior to calling this
* function.
+ *
+ * This will synchronously bring down the link if the link is not already
+ * down (in other words, it will trigger a mac_link_down() method call.)
*/
void phylink_stop(struct phylink *pl)
{
@@ -1338,6 +1342,81 @@ void phylink_stop(struct phylink *pl)
}
EXPORT_SYMBOL_GPL(phylink_stop);
+
+/**
+ * phylink_suspend() - handle a network device suspend event
+ * @pl: a pointer to a &struct phylink returned from phylink_create()
+ * @mac_wol: true if the MAC needs to receive packets for Wake-on-Lan
+ *
+ * Handle a network device suspend event. There are several cases:
+ * - If Wake-on-Lan is not active, we can bring down the link between
+ * the MAC and PHY by calling phylink_stop().
+ * - If Wake-on-Lan is active, and being handled only by the PHY, we
+ * can also bring down the link between the MAC and PHY.
+ * - If Wake-on-Lan is active, but being handled by the MAC, the MAC
+ * still needs to receive packets, so we can not bring the link down.
+ */
+void phylink_suspend(struct phylink *pl, bool mac_wol)
+{
+ ASSERT_RTNL();
+
+ if (mac_wol && (!pl->netdev || pl->netdev->wol_enabled)) {
+ /* Wake-on-Lan enabled, MAC handling */
+ mutex_lock(&pl->state_mutex);
+
+ /* Stop the resolver bringing the link up */
+ __set_bit(PHYLINK_DISABLE_MAC_WOL, &pl->phylink_disable_state);
+
+ /* Disable the carrier, to prevent transmit timeouts,
+ * but one would hope all packets have been sent.
+ */
+ netif_carrier_off(pl->netdev);
+
+ /* We do not call mac_link_down() here as we want the
+ * link to remain up to receive the WoL packets.
+ */
+ mutex_unlock(&pl->state_mutex);
+ } else {
+ phylink_stop(pl);
+ }
+}
+EXPORT_SYMBOL_GPL(phylink_suspend);
+
+/**
+ * phylink_resume() - handle a network device resume event
+ * @pl: a pointer to a &struct phylink returned from phylink_create()
+ *
+ * Undo the effects of phylink_suspend(), returning the link to an
+ * operational state.
+ */
+void phylink_resume(struct phylink *pl)
+{
+ ASSERT_RTNL();
+
+ if (test_bit(PHYLINK_DISABLE_MAC_WOL, &pl->phylink_disable_state)) {
+ /* Wake-on-Lan enabled, MAC handling */
+
+ /* Call mac_link_down() so we keep the overall state balanced.
+ * Do this under the state_mutex lock for consistency. This
+ * will cause a "Link Down" message to be printed during
+ * resume, which is harmless - the true link state will be
+ * printed when we run a resolve.
+ */
+ mutex_lock(&pl->state_mutex);
+ phylink_link_down(pl);
+ mutex_unlock(&pl->state_mutex);
+
+ /* Re-apply the link parameters so that all the settings get
+ * restored to the MAC.
+ */
+ phylink_mac_initial_config(pl, true);
+ phylink_enable_and_run_resolve(pl, PHYLINK_DISABLE_MAC_WOL);
+ } else {
+ phylink_start(pl);
+ }
+}
+EXPORT_SYMBOL_GPL(phylink_resume);
+
/**
* phylink_ethtool_get_wol() - get the wake on lan parameters for the PHY
* @pl: a pointer to a &struct phylink returned from phylink_create()
diff --git a/include/linux/phylink.h b/include/linux/phylink.h
index bdeec800da5c..ba0ab7126b96 100644
--- a/include/linux/phylink.h
+++ b/include/linux/phylink.h
@@ -462,6 +462,9 @@ void phylink_mac_change(struct phylink *, bool up);
void phylink_start(struct phylink *);
void phylink_stop(struct phylink *);
+void phylink_suspend(struct phylink *pl, bool mac_wol);
+void phylink_resume(struct phylink *pl);
+
void phylink_ethtool_get_wol(struct phylink *, struct ethtool_wolinfo *);
int phylink_ethtool_set_wol(struct phylink *, struct ethtool_wolinfo *);
--
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 40Mbps down 10Mbps up. Decent connectivity at last!
Powered by blists - more mailing lists