2017-03-04 10:18:57

by Philippe Reynes

[permalink] [raw]
Subject: [PATCH] net: smsc: smc91x: use new api ethtool_{get|set}_link_ksettings

The ethtool api {get|set}_settings is deprecated.
We move this driver to new api {get|set}_link_ksettings.

As I don't have the hardware, I'd be very pleased if
someone may test this patch.

Signed-off-by: Philippe Reynes <[email protected]>
---
drivers/net/ethernet/smsc/smc91x.c | 45 +++++++++++++++++++----------------
1 files changed, 24 insertions(+), 21 deletions(-)

diff --git a/drivers/net/ethernet/smsc/smc91x.c b/drivers/net/ethernet/smsc/smc91x.c
index 65077c7..23bac47 100644
--- a/drivers/net/ethernet/smsc/smc91x.c
+++ b/drivers/net/ethernet/smsc/smc91x.c
@@ -1535,32 +1535,33 @@ static int smc_close(struct net_device *dev)
* Ethtool support
*/
static int
-smc_ethtool_getsettings(struct net_device *dev, struct ethtool_cmd *cmd)
+smc_ethtool_get_link_ksettings(struct net_device *dev,
+ struct ethtool_link_ksettings *cmd)
{
struct smc_local *lp = netdev_priv(dev);
int ret;

- cmd->maxtxpkt = 1;
- cmd->maxrxpkt = 1;
-
if (lp->phy_type != 0) {
spin_lock_irq(&lp->lock);
- ret = mii_ethtool_gset(&lp->mii, cmd);
+ ret = mii_ethtool_get_link_ksettings(&lp->mii, cmd);
spin_unlock_irq(&lp->lock);
} else {
- cmd->supported = SUPPORTED_10baseT_Half |
+ u32 supported = SUPPORTED_10baseT_Half |
SUPPORTED_10baseT_Full |
SUPPORTED_TP | SUPPORTED_AUI;

if (lp->ctl_rspeed == 10)
- ethtool_cmd_speed_set(cmd, SPEED_10);
+ cmd->base.speed = SPEED_10;
else if (lp->ctl_rspeed == 100)
- ethtool_cmd_speed_set(cmd, SPEED_100);
+ cmd->base.speed = SPEED_100;
+
+ cmd->base.autoneg = AUTONEG_DISABLE;
+ cmd->base.port = 0;
+ cmd->base.duplex = lp->tcr_cur_mode & TCR_SWFDUP ?
+ DUPLEX_FULL : DUPLEX_HALF;

- cmd->autoneg = AUTONEG_DISABLE;
- cmd->transceiver = XCVR_INTERNAL;
- cmd->port = 0;
- cmd->duplex = lp->tcr_cur_mode & TCR_SWFDUP ? DUPLEX_FULL : DUPLEX_HALF;
+ ethtool_convert_legacy_u32_to_link_mode(
+ cmd->link_modes.supported, supported);

ret = 0;
}
@@ -1569,24 +1570,26 @@ static int smc_close(struct net_device *dev)
}

static int
-smc_ethtool_setsettings(struct net_device *dev, struct ethtool_cmd *cmd)
+smc_ethtool_set_link_ksettings(struct net_device *dev,
+ const struct ethtool_link_ksettings *cmd)
{
struct smc_local *lp = netdev_priv(dev);
int ret;

if (lp->phy_type != 0) {
spin_lock_irq(&lp->lock);
- ret = mii_ethtool_sset(&lp->mii, cmd);
+ ret = mii_ethtool_set_link_ksettings(&lp->mii, cmd);
spin_unlock_irq(&lp->lock);
} else {
- if (cmd->autoneg != AUTONEG_DISABLE ||
- cmd->speed != SPEED_10 ||
- (cmd->duplex != DUPLEX_HALF && cmd->duplex != DUPLEX_FULL) ||
- (cmd->port != PORT_TP && cmd->port != PORT_AUI))
+ if (cmd->base.autoneg != AUTONEG_DISABLE ||
+ cmd->base.speed != SPEED_10 ||
+ (cmd->base.duplex != DUPLEX_HALF &&
+ cmd->base.duplex != DUPLEX_FULL) ||
+ (cmd->base.port != PORT_TP && cmd->base.port != PORT_AUI))
return -EINVAL;

// lp->port = cmd->port;
- lp->ctl_rfduplx = cmd->duplex == DUPLEX_FULL;
+ lp->ctl_rfduplx = cmd->base.duplex == DUPLEX_FULL;

// if (netif_running(dev))
// smc_set_port(dev);
@@ -1744,8 +1747,6 @@ static int smc_ethtool_seteeprom(struct net_device *dev,


static const struct ethtool_ops smc_ethtool_ops = {
- .get_settings = smc_ethtool_getsettings,
- .set_settings = smc_ethtool_setsettings,
.get_drvinfo = smc_ethtool_getdrvinfo,

.get_msglevel = smc_ethtool_getmsglevel,
@@ -1755,6 +1756,8 @@ static int smc_ethtool_seteeprom(struct net_device *dev,
.get_eeprom_len = smc_ethtool_geteeprom_len,
.get_eeprom = smc_ethtool_geteeprom,
.set_eeprom = smc_ethtool_seteeprom,
+ .get_link_ksettings = smc_ethtool_get_link_ksettings,
+ .set_link_ksettings = smc_ethtool_set_link_ksettings,
};

static const struct net_device_ops smc_netdev_ops = {
--
1.7.4.4


2017-03-04 10:20:54

by Russell King (Oracle)

[permalink] [raw]
Subject: Re: [PATCH] net: smsc: smc91x: use new api ethtool_{get|set}_link_ksettings

On Sat, Mar 04, 2017 at 11:11:32AM +0100, Philippe Reynes wrote:
>
> // lp->port = cmd->port;

Hi,

Review looks good. However, please also update the above comment as
well.

Thanks.

--
RMK's Patch system: http://www.armlinux.org.uk/developer/patches/
FTTC broadband for 0.8mile line: currently at 9.6Mbps down 400kbps up
according to speedtest.net.

2017-03-04 20:28:03

by Robert Jarzmik

[permalink] [raw]
Subject: Re: [PATCH] net: smsc: smc91x: use new api ethtool_{get|set}_link_ksettings

Philippe Reynes <[email protected]> writes:

> The ethtool api {get|set}_settings is deprecated.
> We move this driver to new api {get|set}_link_ksettings.
>
> As I don't have the hardware, I'd be very pleased if
> someone may test this patch.
I have the hardware.

Do you want anything specific to be tested, like setting full/half duplex or
autoneg through ethtool, or just a driver probe/usage ?

Cheers.

--
Robert

2017-03-04 22:47:49

by Philippe Reynes

[permalink] [raw]
Subject: Re: [PATCH] net: smsc: smc91x: use new api ethtool_{get|set}_link_ksettings

Hi Robert,

On 3/4/17, Robert Jarzmik <[email protected]> wrote:
> Philippe Reynes <[email protected]> writes:
>
>> The ethtool api {get|set}_settings is deprecated.
>> We move this driver to new api {get|set}_link_ksettings.
>>
>> As I don't have the hardware, I'd be very pleased if
>> someone may test this patch.
> I have the hardware.
>
> Do you want anything specific to be tested, like setting full/half duplex
> or
> autoneg through ethtool, or just a driver probe/usage ?

I would be very nice if you could test the setting of half/full
duplex, speed, ...


> Cheers.
>
> --
> Robert
>

Thanks a lot for this test,

Regards,
Philippe