On Mon, 14 Feb 2011, Paul Bolle wrote:
> On a laptop I see these errors on (most) resumes:
> hub 3-0:1.0: over-current change on port 1
> hub 3-0:1.0: over-current change on port 2
>
> The error message doesn't explain what action should be taken. Neither
> does the code surrounding this message. A quick look at the (large) USB
> specification didn't make it clear to me why this should be an error.
> All other over-current related messages in this file are at dev_dbg()
> level. There are no obvious other problems with USB after resume. I have
> no idea what an over-current change actually is, so it's just a guess
> that this message can be downgraded to dev_dbg() level.
"Over-current" means the device plugged into the hub's port is drawing
more electrical current than it is allowed to. "Over-current change"
means that the over-current status has changed, i.e., at some point
since the last time we looked at the hub, an over-current status
existed.
In general not much can be done about over-current problems, other than
unplugging the device that's drinking too much current. In my
experience these conditions are almost always very short-lived (a
fraction of a second), in which case no recovery action is needed.
> Signed-off-by: Paul Bolle <pebolle@xxxxxxxxxx>
> ---
> drivers/usb/core/hub.c | 2 +-
> 1 files changed, 1 insertions(+), 1 deletions(-)
>
> diff --git a/drivers/usb/core/hub.c b/drivers/usb/core/hub.c
> index d041c68..17ed38a 100644
> --- a/drivers/usb/core/hub.c
> +++ b/drivers/usb/core/hub.c
> @@ -3412,7 +3412,7 @@ static void hub_events(void)
> }
>
> if (portchange & USB_PORT_STAT_C_OVERCURRENT) {
> - dev_err (hub_dev,
> + dev_dbg (hub_dev,
> "over-current change on port %d\n",
> i);
> clear_port_feature(hdev, i,
I'm not sure. It's true that this is almost always harmless, and the
corresponding statement for hub over-current rather than port
over-current is indeed dev_dbg. On the other hand, it potentially
does indicate a genuine error condition.
Maybe we should delay for 500 ms and then check the port status again.
It would make sense to log an error message if the over-current status
persists, otherwise dev_dbg would be fine.
Alan Stern
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
