[PATCH net] vlan: also check phy_driver ts_info for vlan's real device

[PATCH net] vlan: also check phy_driver ts_info for vlan's real device

[Hangbin Liu]

Just like function ethtool_get_ts_info(), we should also consider the
phy_driver ts_info call back. For example, driver dp83640.

Fixes: 37dd9255b2f6 ("vlan: Pass ethtool get_ts_info queries to real device.")
Acked-by: Richard Cochran <richardcochran@gmail.com>
Signed-off-by: Hangbin Liu <liuhangbin@gmail.com>
---
 net/8021q/vlan_dev.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/net/8021q/vlan_dev.c b/net/8021q/vlan_dev.c
index f7e83f6..236452e 100644
--- a/net/8021q/vlan_dev.c
+++ b/net/8021q/vlan_dev.c
@@ -29,6 +29,7 @@
 #include <linux/net_tstamp.h>
 #include <linux/etherdevice.h>
 #include <linux/ethtool.h>
+#include <linux/phy.h>
 #include <net/arp.h>
 #include <net/switchdev.h>
 
@@ -665,8 +666,11 @@ static int vlan_ethtool_get_ts_info(struct net_device *dev,
 {
 	const struct vlan_dev_priv *vlan = vlan_dev_priv(dev);
 	const struct ethtool_ops *ops = vlan->real_dev->ethtool_ops;
+	struct phy_device *phydev = vlan->real_dev->phydev;
 
-	if (ops->get_ts_info) {
+	if (phydev && phydev->drv && phydev->drv->ts_info) {
+		 return phydev->drv->ts_info(phydev, info);
+	} else if (ops->get_ts_info) {
 		return ops->get_ts_info(vlan->real_dev, info);
 	} else {
 		info->so_timestamping = SOF_TIMESTAMPING_RX_SOFTWARE |
-- 
2.5.5

Re: [PATCH net] vlan: also check phy_driver ts_info for vlan's real device

[Andrew Lunn]

On Fri, Mar 30, 2018 at 09:44:00AM +0800, Hangbin Liu wrote:
> Just like function ethtool_get_ts_info(), we should also consider the
> phy_driver ts_info call back. For example, driver dp83640.

Hi Hangbin

Would it not be better to just call ethtool_get_ts_info() on the real
device? That would then also deal with the case that the 'real' device
is another virtual device stacked on top of a real device.

   Andrew

Re: [PATCH net] vlan: also check phy_driver ts_info for vlan's real device

[Richard Cochran]

On Fri, Mar 30, 2018 at 05:02:14PM +0200, Andrew Lunn wrote:
> Would it not be better to just call ethtool_get_ts_info() on the real
> device? That would then also deal with the case that the 'real' device
> is another virtual device stacked on top of a real device.

That won't work.

The returned 'ethtool_ts_info' is on the stack of
ethtool_get_ts_info().  Both the top level and the inner call to
will call copy_to_user(), but the top level will clobber the correct
result with zeros.

Thanks,
Richard

Re: [PATCH net] vlan: also check phy_driver ts_info for vlan's real device

[Andrew Lunn]

On Fri, Mar 30, 2018 at 08:21:21AM -0700, Richard Cochran wrote:
> On Fri, Mar 30, 2018 at 05:02:14PM +0200, Andrew Lunn wrote:
> > Would it not be better to just call ethtool_get_ts_info() on the real
> > device? That would then also deal with the case that the 'real' device
> > is another virtual device stacked on top of a real device.
> 
> That won't work.
> 
> The returned 'ethtool_ts_info' is on the stack of
> ethtool_get_ts_info().  Both the top level and the inner call to
> will call copy_to_user(), but the top level will clobber the correct
> result with zeros.

Ah, O.K.

But it still seems like there should be one central place to handle a
device. Maybe factor out the part which deals with getting the
information from the device, from the part that copies it to
userspace?

	Andrew

Re: [PATCH net] vlan: also check phy_driver ts_info for vlan's real device

[David Miller]

From: Hangbin Liu <liuhangbin@gmail.com>
Date: Fri, 30 Mar 2018 09:44:00 +0800

> Just like function ethtool_get_ts_info(), we should also consider the
> phy_driver ts_info call back. For example, driver dp83640.
> 
> Fixes: 37dd9255b2f6 ("vlan: Pass ethtool get_ts_info queries to real device.")
> Acked-by: Richard Cochran <richardcochran@gmail.com>
> Signed-off-by: Hangbin Liu <liuhangbin@gmail.com>

Applied and queued up for -stable.