| Message ID | 20230713095743.30517-1-fercerpav@gmail.com (mailing list archive) |
|---|---|
| State | Accepted |
| Commit | 2cee73cef253a333c7d48a0f8edda53e3cca990c |
| Delegated to: | Netdev Maintainers |
| Headers | show |
| Series | net: ftgmac100: support getting MAC address from NVMEM | expand |
On Thu, Jul 13, 2023 at 3:28 PM Paul Fertser <fercerpav@gmail.com> wrote: > > Make use of of_get_ethdev_address() to support reading MAC address not > only from the usual DT nodes but also from an NVMEM provider (e.g. using > a dedicated area in an FRU EEPROM). > Looks like earlier ftgmac100_probe() would move on with self generated (random) MAC addr if getting it from the device failed. Now you will fail the probe in a failure case. Is that OK? > Signed-off-by: Paul Fertser <fercerpav@gmail.com> > --- > drivers/net/ethernet/faraday/ftgmac100.c | 16 ++++++++++++---- > 1 file changed, 12 insertions(+), 4 deletions(-) > > diff --git a/drivers/net/ethernet/faraday/ftgmac100.c b/drivers/net/ethernet/faraday/ftgmac100.c > index a03879a27b04..9135b918dd49 100644 > --- a/drivers/net/ethernet/faraday/ftgmac100.c > +++ b/drivers/net/ethernet/faraday/ftgmac100.c > @@ -177,16 +177,20 @@ static void ftgmac100_write_mac_addr(struct ftgmac100 *priv, const u8 *mac) > iowrite32(laddr, priv->base + FTGMAC100_OFFSET_MAC_LADR); > } > > -static void ftgmac100_initial_mac(struct ftgmac100 *priv) > +static int ftgmac100_initial_mac(struct ftgmac100 *priv) > { > u8 mac[ETH_ALEN]; > unsigned int m; > unsigned int l; > + int err; > > - if (!device_get_ethdev_address(priv->dev, priv->netdev)) { > + err = of_get_ethdev_address(priv->dev->of_node, priv->netdev); > + if (err == -EPROBE_DEFER) > + return err; > + if (!err) { > dev_info(priv->dev, "Read MAC address %pM from device tree\n", > priv->netdev->dev_addr); > - return; > + return 0; > } > > m = ioread32(priv->base + FTGMAC100_OFFSET_MAC_MADR); > @@ -207,6 +211,8 @@ static void ftgmac100_initial_mac(struct ftgmac100 *priv) > dev_info(priv->dev, "Generated random MAC address %pM\n", > priv->netdev->dev_addr); > } > + > + return 0; > } > > static int ftgmac100_set_mac_addr(struct net_device *dev, void *p) > @@ -1843,7 +1849,9 @@ static int ftgmac100_probe(struct platform_device *pdev) > priv->aneg_pause = true; > > /* MAC address from chip or random one */ > - ftgmac100_initial_mac(priv); > + err = ftgmac100_initial_mac(priv); > + if (err) > + goto err_phy_connect; > > np = pdev->dev.of_node; > if (np && (of_device_is_compatible(np, "aspeed,ast2400-mac") || > -- > 2.34.1 > >
Hello Pavan, On Fri, Jul 14, 2023 at 10:14:02AM +0530, Pavan Chebbi wrote: > On Thu, Jul 13, 2023 at 3:28 PM Paul Fertser <fercerpav@gmail.com> wrote: > > Make use of of_get_ethdev_address() to support reading MAC address not > > only from the usual DT nodes but also from an NVMEM provider (e.g. using > > a dedicated area in an FRU EEPROM). > > Looks like earlier ftgmac100_probe() would move on with self generated > (random) MAC addr if getting it from the device failed. > Now you will fail the probe in a failure case. Is that OK? I think the previous behaviour is preserved with this patch in all the cases other than of_get_ethdev_address returning -EPROBE_DEFER. Can you please explain what failure case you have in mind and how the probe is going to be failed in that case?
On Fri, 2023-07-14 at 08:18 +0300, Paul Fertser wrote: > Hello Pavan, > > On Fri, Jul 14, 2023 at 10:14:02AM +0530, Pavan Chebbi wrote: > > On Thu, Jul 13, 2023 at 3:28 PM Paul Fertser <fercerpav@gmail.com> wrote: > > > Make use of of_get_ethdev_address() to support reading MAC address not > > > only from the usual DT nodes but also from an NVMEM provider (e.g. using > > > a dedicated area in an FRU EEPROM). > > > > Looks like earlier ftgmac100_probe() would move on with self generated > > (random) MAC addr if getting it from the device failed. > > Now you will fail the probe in a failure case. Is that OK? > > I think the previous behaviour is preserved with this patch in all the > cases other than of_get_ethdev_address returning -EPROBE_DEFER. Can > you please explain what failure case you have in mind and how the > probe is going to be failed in that case? FTR, I agree with the above: it looks like the old behavior is preserved. The patch LGTM, thanks! Paolo
Hello: This patch was applied to netdev/net-next.git (main) by Paolo Abeni <pabeni@redhat.com>: On Thu, 13 Jul 2023 12:57:43 +0300 you wrote: > Make use of of_get_ethdev_address() to support reading MAC address not > only from the usual DT nodes but also from an NVMEM provider (e.g. using > a dedicated area in an FRU EEPROM). > > Signed-off-by: Paul Fertser <fercerpav@gmail.com> > --- > drivers/net/ethernet/faraday/ftgmac100.c | 16 ++++++++++++---- > 1 file changed, 12 insertions(+), 4 deletions(-) Here is the summary with links: - net: ftgmac100: support getting MAC address from NVMEM https://git.kernel.org/netdev/net-next/c/2cee73cef253 You are awesome, thank you!
diff --git a/drivers/net/ethernet/faraday/ftgmac100.c b/drivers/net/ethernet/faraday/ftgmac100.c index a03879a27b04..9135b918dd49 100644 --- a/drivers/net/ethernet/faraday/ftgmac100.c +++ b/drivers/net/ethernet/faraday/ftgmac100.c @@ -177,16 +177,20 @@ static void ftgmac100_write_mac_addr(struct ftgmac100 *priv, const u8 *mac) iowrite32(laddr, priv->base + FTGMAC100_OFFSET_MAC_LADR); } -static void ftgmac100_initial_mac(struct ftgmac100 *priv) +static int ftgmac100_initial_mac(struct ftgmac100 *priv) { u8 mac[ETH_ALEN]; unsigned int m; unsigned int l; + int err; - if (!device_get_ethdev_address(priv->dev, priv->netdev)) { + err = of_get_ethdev_address(priv->dev->of_node, priv->netdev); + if (err == -EPROBE_DEFER) + return err; + if (!err) { dev_info(priv->dev, "Read MAC address %pM from device tree\n", priv->netdev->dev_addr); - return; + return 0; } m = ioread32(priv->base + FTGMAC100_OFFSET_MAC_MADR); @@ -207,6 +211,8 @@ static void ftgmac100_initial_mac(struct ftgmac100 *priv) dev_info(priv->dev, "Generated random MAC address %pM\n", priv->netdev->dev_addr); } + + return 0; } static int ftgmac100_set_mac_addr(struct net_device *dev, void *p) @@ -1843,7 +1849,9 @@ static int ftgmac100_probe(struct platform_device *pdev) priv->aneg_pause = true; /* MAC address from chip or random one */ - ftgmac100_initial_mac(priv); + err = ftgmac100_initial_mac(priv); + if (err) + goto err_phy_connect; np = pdev->dev.of_node; if (np && (of_device_is_compatible(np, "aspeed,ast2400-mac") ||
Make use of of_get_ethdev_address() to support reading MAC address not only from the usual DT nodes but also from an NVMEM provider (e.g. using a dedicated area in an FRU EEPROM). Signed-off-by: Paul Fertser <fercerpav@gmail.com> --- drivers/net/ethernet/faraday/ftgmac100.c | 16 ++++++++++++---- 1 file changed, 12 insertions(+), 4 deletions(-)