mbox series

[0/2] USB: gadget core: Call ->disconnect() from usb_gadget_disconnect()

Message ID Pine.LNX.4.44L0.1808101525010.1425-100000@iolanthe.rowland.org (mailing list archive)
Headers show
Series USB: gadget core: Call ->disconnect() from usb_gadget_disconnect() | expand

Message

Alan Stern Aug. 10, 2018, 7:32 p.m. UTC 
Felipe:

The following patches concern ->disconnect() callbacks made when a 
gadget's D+ pullup is turned off.  Currently we don't have a fixed rule 
for such callbacks.

The first patch makes usb_gadget_disconnect() issue the callback 
automatically, so that it will always happen and UDC drivers don't need 
to worry about it.

The second patch removes a callback from net2280's pullup routine, 
because it is now unneeded.

Other UDC drivers may also need to have their callbacks removed; the
only one I have checked is dummy-hcd (which is okay).

Alan Stern

Comments

Felipe Balbi Oct. 2, 2018, 7:44 a.m. UTC | #1 
Hi,

Alan Stern <stern@rowland.harvard.edu> writes:
> Felipe:
>
> The following patches concern ->disconnect() callbacks made when a 
> gadget's D+ pullup is turned off.  Currently we don't have a fixed rule 
> for such callbacks.
>
> The first patch makes usb_gadget_disconnect() issue the callback 
> automatically, so that it will always happen and UDC drivers don't need 
> to worry about it.
>
> The second patch removes a callback from net2280's pullup routine, 
> because it is now unneeded.
>
> Other UDC drivers may also need to have their callbacks removed; the
> only one I have checked is dummy-hcd (which is okay).

I never got 2/2. Only 1/2 and this. Care to resend?
Alan Stern Oct. 2, 2018, 2:17 p.m. UTC | #2 
On Tue, 2 Oct 2018, Felipe Balbi wrote:

> Hi,
> 
> Alan Stern <stern@rowland.harvard.edu> writes:
> > Felipe:
> >
> > The following patches concern ->disconnect() callbacks made when a 
> > gadget's D+ pullup is turned off.  Currently we don't have a fixed rule 
> > for such callbacks.
> >
> > The first patch makes usb_gadget_disconnect() issue the callback 
> > automatically, so that it will always happen and UDC drivers don't need 
> > to worry about it.
> >
> > The second patch removes a callback from net2280's pullup routine, 
> > because it is now unneeded.
> >
> > Other UDC drivers may also need to have their callbacks removed; the
> > only one I have checked is dummy-hcd (which is okay).
> 
> I never got 2/2. Only 1/2 and this. Care to resend?

Patches 1/2 and 2/2 to follow...

Alan Stern