On 2024/10/29 21:47, Johan Hovold wrote: > On Thu, Oct 24, 2024 at 10:39:29PM +0800, Zijun Hu wrote: >> From: Zijun Hu <quic_zijuhu@xxxxxxxxxxx> >> >> It will leak refcount of OF node @args.np for _of_phy_get() not to >> decrease refcount increased by previous of_parse_phandle_with_args() >> when returns due to of_device_is_compatible() error. >> >> Fix by adding of_node_put() before the error return. >> >> Fixes: b7563e2796f8 ("phy: work around 'phys' references to usb-nop-xceiv devices") >> Cc: stable@xxxxxxxxxxxxxxx >> Signed-off-by: Zijun Hu <quic_zijuhu@xxxxxxxxxxx> >> --- >> drivers/phy/phy-core.c | 5 ++++- >> 1 file changed, 4 insertions(+), 1 deletion(-) >> >> diff --git a/drivers/phy/phy-core.c b/drivers/phy/phy-core.c >> index 52ca590a58b9..967878b78797 100644 >> --- a/drivers/phy/phy-core.c >> +++ b/drivers/phy/phy-core.c >> @@ -629,8 +629,11 @@ static struct phy *_of_phy_get(struct device_node *np, int index) >> return ERR_PTR(-ENODEV); >> >> /* This phy type handled by the usb-phy subsystem for now */ >> - if (of_device_is_compatible(args.np, "usb-nop-xceiv")) >> + if (of_device_is_compatible(args.np, "usb-nop-xceiv")) { >> + /* Put refcount above of_parse_phandle_with_args() got */ > > No need for a comment as this is already handled in the later paths. > will remove it within v3 >> + of_node_put(args.np); > > For that reason you should probably initialise ret and add a new label > out_put_node that you jump to here instead. > will try to do it within v3. >> return ERR_PTR(-ENODEV); >> + } >> >> mutex_lock(&phy_provider_mutex); >> phy_provider = of_phy_provider_lookup(args.np); > > Johan