These functions shouldn't be called with a NULL pointer as argument in
the first place, but this happened at least twice so far[1][2], because
drivers weren't tested in fixed-link setups without PHYs.

To improve user experience, let's return -EIO instead of crashing
outright. The proper fix is still to adapt these drivers, so they
do not try to access non-existent PHYs.

[1]: commit 6269556397 ("net: eqos: let phy_resume handle fixed link phys")
[2]: 
https://lore.barebox.org/barebox/20241010-fix-designware-fixed-link-v1-1-de2df5485...@pengutronix.de/T/#u

Signed-off-by: Ahmad Fatoum <[email protected]>
---
 drivers/net/phy/mdio_bus.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/drivers/net/phy/mdio_bus.c b/drivers/net/phy/mdio_bus.c
index 30d5aeacff0d..68652e92a830 100644
--- a/drivers/net/phy/mdio_bus.c
+++ b/drivers/net/phy/mdio_bus.c
@@ -441,6 +441,9 @@ int mdiobus_read(struct mii_bus *bus, int addr, u32 regnum)
 {
        int ret;
 
+       if (!bus)
+               return -EIO;
+
        slice_acquire(&bus->slice);
 
        ret = bus->read(bus, addr, regnum);
@@ -461,6 +464,9 @@ int mdiobus_write(struct mii_bus *bus, int addr, u32 
regnum, u16 val)
 {
        int ret;
 
+       if (!bus)
+               return -EIO;
+
        slice_acquire(&bus->slice);
 
        ret = bus->write(bus, addr, regnum, val);
-- 
2.39.5


Reply via email to