On 26.02.2019 09:43, Russell King - ARM Linux admin wrote:
> On Tue, Feb 26, 2019 at 08:25:41AM +0100, Heiner Kallweit wrote:
>> When debugging an issue I found implausible values in state->pause.
>> Reason in that state->pause isn't initialized and later only single
>> bits are changed. Also the struct itself isn't initialized in
>> phylink_resolve(). So better initialize state->pause.
>
> mac_link_state() is expected to always set this, but this is safer.
>
I've seen cases where mac_link_state() just initializes the relevant
bits. Therefore there's no functional problem, but it's not nice
(and disturbs in debug output) if all other bits have a random value.
> Maybe also set state->speed to SPEED_UNKNOWN, state->duplex to
> DUPLEX_UNKNOWN and state->an_complete to zero?
>
OK. Will include this in a v3.
>>
>> v2:
>> - use right function name in subject
>>
>> Fixes: 9525ae83959b ("phylink: add phylink infrastructure")
>> Signed-off-by: Heiner Kallweit <[email protected]>
>> ---
>> drivers/net/phy/phylink.c | 1 +
>> 1 file changed, 1 insertion(+)
>>
>> diff --git a/drivers/net/phy/phylink.c b/drivers/net/phy/phylink.c
>> index 59d175a5b..a9954c205 100644
>> --- a/drivers/net/phy/phylink.c
>> +++ b/drivers/net/phy/phylink.c
>> @@ -324,6 +324,7 @@ static int phylink_get_mac_state(struct phylink *pl,
>> struct phylink_link_state *
>> linkmode_zero(state->lp_advertising);
>> state->interface = pl->link_config.interface;
>> state->an_enabled = pl->link_config.an_enabled;
>> + state->pause = MLO_PAUSE_NONE;
>> state->link = 1;
>>
>> return pl->ops->mac_link_state(ndev, state);
>> --
>> 2.20.1
>>
>>
>
