[PATCH] input: bcm5974 turn wellspring mode off if failed to start traffic

[PATCH] input: bcm5974 turn wellspring mode off if failed to start traffic

[Luo Jinghua]

The bcm5974 driver doesn't work on my MacbookPro6,2 for some reason. It
failed to start traffic and left the wellsprint mode on. As a result,
the keyboard is slow to response and become unusable.

Signed-off-by: Luo Jinghua <sunmoon1997@gmail.com>
---
 drivers/input/mouse/bcm5974.c |    1 +
 1 files changed, 1 insertions(+), 0 deletions(-)

diff --git a/drivers/input/mouse/bcm5974.c b/drivers/input/mouse/bcm5974.c
index 6dedded..5518089 100644
--- a/drivers/input/mouse/bcm5974.c
+++ b/drivers/input/mouse/bcm5974.c
@@ -595,6 +595,7 @@ static int bcm5974_start_traffic(struct bcm5974 *dev)
 
 err_kill_bt:
 	usb_kill_urb(dev->bt_urb);
+	bcm5974_wellspring_mode(dev, false);
 error:
 	return -EIO;
 }
-- 
1.7.1

Re: [PATCH] input: bcm5974 turn wellspring mode off if failed to start traffic

[Dmitry Torokhov]

Hi Luo,

On Wed, Jun 02, 2010 at 03:45:59AM +0800, Luo Jinghua wrote:
> The bcm5974 driver doesn't work on my MacbookPro6,2 for some reason. It
> failed to start traffic and left the wellsprint mode on. As a result,
> the keyboard is slow to response and become unusable.
> 
> Signed-off-by: Luo Jinghua <sunmoon1997@gmail.com>
> ---
>  drivers/input/mouse/bcm5974.c |    1 +
>  1 files changed, 1 insertions(+), 0 deletions(-)
> 
> diff --git a/drivers/input/mouse/bcm5974.c b/drivers/input/mouse/bcm5974.c
> index 6dedded..5518089 100644
> --- a/drivers/input/mouse/bcm5974.c
> +++ b/drivers/input/mouse/bcm5974.c
> @@ -595,6 +595,7 @@ static int bcm5974_start_traffic(struct bcm5974 *dev)
>  
>  err_kill_bt:
>  	usb_kill_urb(dev->bt_urb);
> +	bcm5974_wellspring_mode(dev, false);
>  error:
>  	return -EIO;
>  }

Could you see what URB it failed to submit and what error code was
returned? Maybe it will give us idea what went wrong.

-- 
Dmitry

Re: [PATCH] input: bcm5974 turn wellspring mode off if failed to start traffic

[Henrik Rydberg]

Luo Jinghua wrote:
[...]
> So the error code is the ENOSPC.

Interesting. A couple of years back, there were a few reports on strange usb
behavior on Macbooks that turned out to be a hardware problem, a bad internal
usb connector. Could that be the case here?

Since ENOSPC is pretty rare, narrowing it down code-wise seems possible. It
could for instance be this path happening:

usb_submit_urb
  usb_hcd_submit_urb
     hcd->driver->urb_enqueue
       ehci_urb_enqueue
         intr_submit
           qh_schedule
             fail

Henrik