netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH] net: dsa: mv88e6xxx: don't use PHY_DETECT on internal PHY's
@ 2021-10-11 14:27 Maarten Zanders
  2021-10-17 21:07 ` Russell King (Oracle)
  0 siblings, 1 reply; 2+ messages in thread
From: Maarten Zanders @ 2021-10-11 14:27 UTC (permalink / raw)
  Cc: maarten.zanders, Maarten Zanders, Maxime Chevallier, Andrew Lunn,
	Vivien Didelot, Florian Fainelli, Vladimir Oltean,
	David S. Miller, Jakub Kicinski, Russell King, netdev,
	linux-kernel

mv88e6xxx_port_ppu_updates() interpretes data in the PORT_STS
register incorrectly for internal ports (ie no PPU). In these
cases, the PHY_DETECT bit indicates link status. This results
in forcing the MAC state whenever the PHY link goes down which
is not intended. As a side effect, LED's configured to show
link status stay lit even though the physical link is down.

Add a check in mac_link_down and mac_link_up to see if it
concerns an external port and only then, look at PPU status.

Fixes: 5d5b231da7ac (net: dsa: mv88e6xxx: use PHY_DETECT in mac_link_up/mac_link_down)
Reported-by: Maarten Zanders <m.zanders@televic.com>
Reviewed-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: Maarten Zanders <maarten.zanders@mind.be>
---
 drivers/net/dsa/mv88e6xxx/chip.c | 15 ++++++++++++---
 1 file changed, 12 insertions(+), 3 deletions(-)

diff --git a/drivers/net/dsa/mv88e6xxx/chip.c b/drivers/net/dsa/mv88e6xxx/chip.c
index ee3f32d1cf46..f5ce05d78e11 100644
--- a/drivers/net/dsa/mv88e6xxx/chip.c
+++ b/drivers/net/dsa/mv88e6xxx/chip.c
@@ -726,9 +726,13 @@ static void mv88e6xxx_mac_link_down(struct dsa_switch *ds, int port,
 
 	ops = chip->info->ops;
 
 	mv88e6xxx_reg_lock(chip);
-	if ((!mv88e6xxx_port_ppu_updates(chip, port) ||
-	     mode == MLO_AN_FIXED) && ops->port_sync_link)
+	/* Internal PHYs propagate their configuration directly to the MAC.
+	 * External PHYs depend on whether the PPU is enabled for this port.
+	 */
+	if (((!mv88e6xxx_phy_is_internal(ds, port) &&
+	      !mv88e6xxx_port_ppu_updates(chip, port)) ||
+	     mode == MLO_AN_FIXED) && ops->port_sync_link)
 		err = ops->port_sync_link(chip, port, mode, false);
 	mv88e6xxx_reg_unlock(chip);
 
@@ -750,7 +754,12 @@ static void mv88e6xxx_mac_link_up(struct dsa_switch *ds, int port,
 	ops = chip->info->ops;
 
 	mv88e6xxx_reg_lock(chip);
-	if (!mv88e6xxx_port_ppu_updates(chip, port) || mode == MLO_AN_FIXED) {
+	/* Internal PHYs propagate their configuration directly to the MAC.
+	 * External PHYs depend on whether the PPU is enabled for this port.
+	 */
+	if ((!mv88e6xxx_phy_is_internal(ds, port) &&
+	     !mv88e6xxx_port_ppu_updates(chip, port)) ||
+	    mode == MLO_AN_FIXED) {
 		/* FIXME: for an automedia port, should we force the link
 		 * down here - what if the link comes up due to "other" media
 		 * while we're bringing the port up, how is the exclusivity
-- 
2.25.1


^ permalink raw reply related	[flat|nested] 2+ messages in thread

* Re: [PATCH] net: dsa: mv88e6xxx: don't use PHY_DETECT on internal PHY's
  2021-10-11 14:27 [PATCH] net: dsa: mv88e6xxx: don't use PHY_DETECT on internal PHY's Maarten Zanders
@ 2021-10-17 21:07 ` Russell King (Oracle)
  0 siblings, 0 replies; 2+ messages in thread
From: Russell King (Oracle) @ 2021-10-17 21:07 UTC (permalink / raw)
  To: Maarten Zanders
  Cc: Maarten Zanders, Maxime Chevallier, Andrew Lunn, Vivien Didelot,
	Florian Fainelli, Vladimir Oltean, David S. Miller,
	Jakub Kicinski, netdev, linux-kernel

On Mon, Oct 11, 2021 at 04:27:20PM +0200, Maarten Zanders wrote:
> mv88e6xxx_port_ppu_updates() interpretes data in the PORT_STS
> register incorrectly for internal ports (ie no PPU). In these
> cases, the PHY_DETECT bit indicates link status. This results
> in forcing the MAC state whenever the PHY link goes down which
> is not intended. As a side effect, LED's configured to show
> link status stay lit even though the physical link is down.

I know this patch has been merged, but I'm going to say this anyway
for the record.

The description is not entirely correct. It is not true that internal
ports do not have the PHY_DETECT bit. 88E6176 and friends are documented
that bit 12 is always the PHY_DETECT bit even for internal ports. Bit 11
there is Link status.

Looking at the definitions in port.h, some switches are different
(88E6250 family) and do indeed use bit 12 as link status.

The point I'm making is that the commit description is not universally
true and only applies to a subset of mv88e6xxx supported switches. It is
a shame it wasn't better reviewed before merging.

At least the patch is harmless; we leave the PHY_DETECT bit set on the
internal ports, which means the PPU fetches the configuration and
configures the port appropriately. The change merely helps to ensure
that we don't force link status on the internal ports.

-- 
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 40Mbps down 10Mbps up. Decent connectivity at last!

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2021-10-17 21:07 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2021-10-11 14:27 [PATCH] net: dsa: mv88e6xxx: don't use PHY_DETECT on internal PHY's Maarten Zanders
2021-10-17 21:07 ` Russell King (Oracle)

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).