[linux-usb-devel] Re: Asynchronous API changes

2003-03-12 Thread Greg KH
On Thu, Mar 06, 2003 at 11:23:16AM -0500, Alan Stern wrote: > > Okay, so this means that the driver has to allocate and free the urb. > What used to be a simple function call has now turned into (without error > checking): > > urb = usb_allocate_urb(...); > usb_submit_bulk_msg(urb,

[linux-usb-devel] Re: Asynchronous API changes

2003-03-09 Thread Alan Stern
On Sat, 8 Mar 2003, David Brownell wrote: > Is there an extra "A" in the "Subject:" then? :) Yes, sorry about that. Slip of the fingers. > And notice that those timeouts are just reasons to abort/unlink > requests Indeed. I can't think offhand of any other reasons for having timeouts for US

[linux-usb-devel] Re: Asynchronous API changes

2003-03-08 Thread David Brownell
Alan Stern wrote: David and Greg: Stepping back and trying to look at the bigger picture led to the following thoughts regarding the synchronous API. Is there an extra "A" in the "Subject:" then? :) First of all, what's the reason for including usb_control_msg() and usb_bulk_msg() in the API