Message ID | 20240121173006.273629-3-marek.vasut+renesas@mailbox.org |
---|---|
State | Superseded |
Delegated to: | Marek Vasut |
Headers | show |
Series | [1/3] net: phy: Make eth_phy_get_addr() generic | expand |
On 21/01/2024 17:29, Marek Vasut wrote: > In case the PHY is fully described in DT, use PHY MDIO address > from DT directly instead of always using auto-detection. This > also fixes the behavior of 'mdio list' in such DT setup, which > now prints the PHY connected to the MAC correctly. > > Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org> > --- > Cc: Joe Hershberger <joe.hershberger@ni.com> > Cc: Paul Barker <paul.barker.ct@bp.renesas.com> > Cc: Ramon Fried <rfried.dev@gmail.com> > --- > drivers/net/sh_eth.c | 10 ++++++++-- > 1 file changed, 8 insertions(+), 2 deletions(-) I think the commit summaries are mixed up here - this patch says 'ravb' but changes 'sh_eth.c' and vice-versa with the other patch. > > diff --git a/drivers/net/sh_eth.c b/drivers/net/sh_eth.c > index 7b1f59dc498..5e2e3054ecf 100644 > --- a/drivers/net/sh_eth.c > +++ b/drivers/net/sh_eth.c > @@ -12,6 +12,7 @@ > #include <common.h> > #include <cpu_func.h> > #include <env.h> > +#include <eth_phy.h> > #include <log.h> > #include <malloc.h> > #include <net.h> > @@ -597,11 +598,16 @@ static int sh_eth_phy_config(struct udevice *dev) > struct sh_ether_priv *priv = dev_get_priv(dev); > struct eth_pdata *pdata = dev_get_plat(dev); > struct sh_eth_dev *eth = &priv->shdev; > - int ret = 0; > struct sh_eth_info *port_info = ð->port_info[eth->port]; > struct phy_device *phydev; > + int phy_addr; > + int ret = 0; > + > + phy_addr = eth_phy_get_addr(dev); > + if (phy_addr < 0) > + phy_addr = -1; > > - phydev = phy_connect(priv->bus, -1, dev, pdata->phy_interface); > + phydev = phy_connect(priv->bus, phy_addr, dev, pdata->phy_interface); > if (!phydev) > return -ENODEV; > I have a local patch to handle (phy_addr < 0) in phy_connect_phy_id() and lookup the address there. I think that's the better place to handle this as it's a universal fix for all users of CONFIG_PHY_ETHERNET_ID. Unfortunately I've been distracted by some higher priority work but I'm hoping to get back to this soon. I can do some quick testing with that patch and send it this week if that's helpful. Thanks, Paul
On 1/21/24 22:37, Paul Barker wrote: > On 21/01/2024 17:29, Marek Vasut wrote: >> In case the PHY is fully described in DT, use PHY MDIO address >> from DT directly instead of always using auto-detection. This >> also fixes the behavior of 'mdio list' in such DT setup, which >> now prints the PHY connected to the MAC correctly. >> >> Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org> >> --- >> Cc: Joe Hershberger <joe.hershberger@ni.com> >> Cc: Paul Barker <paul.barker.ct@bp.renesas.com> >> Cc: Ramon Fried <rfried.dev@gmail.com> >> --- >> drivers/net/sh_eth.c | 10 ++++++++-- >> 1 file changed, 8 insertions(+), 2 deletions(-) > > I think the commit summaries are mixed up here - this patch says 'ravb' > but changes 'sh_eth.c' and vice-versa with the other patch. Sigh ... too much git commit --fixup/rebase --autosquash . >> diff --git a/drivers/net/sh_eth.c b/drivers/net/sh_eth.c >> index 7b1f59dc498..5e2e3054ecf 100644 >> --- a/drivers/net/sh_eth.c >> +++ b/drivers/net/sh_eth.c >> @@ -12,6 +12,7 @@ >> #include <common.h> >> #include <cpu_func.h> >> #include <env.h> >> +#include <eth_phy.h> >> #include <log.h> >> #include <malloc.h> >> #include <net.h> >> @@ -597,11 +598,16 @@ static int sh_eth_phy_config(struct udevice *dev) >> struct sh_ether_priv *priv = dev_get_priv(dev); >> struct eth_pdata *pdata = dev_get_plat(dev); >> struct sh_eth_dev *eth = &priv->shdev; >> - int ret = 0; >> struct sh_eth_info *port_info = ð->port_info[eth->port]; >> struct phy_device *phydev; >> + int phy_addr; >> + int ret = 0; >> + >> + phy_addr = eth_phy_get_addr(dev); >> + if (phy_addr < 0) >> + phy_addr = -1; >> >> - phydev = phy_connect(priv->bus, -1, dev, pdata->phy_interface); >> + phydev = phy_connect(priv->bus, phy_addr, dev, pdata->phy_interface); >> if (!phydev) >> return -ENODEV; >> > > I have a local patch to handle (phy_addr < 0) in phy_connect_phy_id() > and lookup the address there. I think that's the better place to handle > this as it's a universal fix for all users of CONFIG_PHY_ETHERNET_ID. > Unfortunately I've been distracted by some higher priority work but I'm > hoping to get back to this soon. > > I can do some quick testing with that patch and send it this week if > that's helpful. Please do send it, yes.
diff --git a/drivers/net/sh_eth.c b/drivers/net/sh_eth.c index 7b1f59dc498..5e2e3054ecf 100644 --- a/drivers/net/sh_eth.c +++ b/drivers/net/sh_eth.c @@ -12,6 +12,7 @@ #include <common.h> #include <cpu_func.h> #include <env.h> +#include <eth_phy.h> #include <log.h> #include <malloc.h> #include <net.h> @@ -597,11 +598,16 @@ static int sh_eth_phy_config(struct udevice *dev) struct sh_ether_priv *priv = dev_get_priv(dev); struct eth_pdata *pdata = dev_get_plat(dev); struct sh_eth_dev *eth = &priv->shdev; - int ret = 0; struct sh_eth_info *port_info = ð->port_info[eth->port]; struct phy_device *phydev; + int phy_addr; + int ret = 0; + + phy_addr = eth_phy_get_addr(dev); + if (phy_addr < 0) + phy_addr = -1; - phydev = phy_connect(priv->bus, -1, dev, pdata->phy_interface); + phydev = phy_connect(priv->bus, phy_addr, dev, pdata->phy_interface); if (!phydev) return -ENODEV;
In case the PHY is fully described in DT, use PHY MDIO address from DT directly instead of always using auto-detection. This also fixes the behavior of 'mdio list' in such DT setup, which now prints the PHY connected to the MAC correctly. Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org> --- Cc: Joe Hershberger <joe.hershberger@ni.com> Cc: Paul Barker <paul.barker.ct@bp.renesas.com> Cc: Ramon Fried <rfried.dev@gmail.com> --- drivers/net/sh_eth.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-)