1USB_UNLINK_URB(9) USB Core APIs USB_UNLINK_URB(9)
2
6 usb_unlink_urb - abort/cancel a transfer request for an endpoint
7
9 int usb_unlink_urb(struct urb * urb);
10
12 urb
13 pointer to urb describing a previously submitted request, may be
14 NULL
15
17 This routine cancels an in-progress request. URBs complete only once
18 per submission, and may be canceled only once per submission.
19 Successful cancellation means termination of urb will be expedited and
20 the completion handler will be called with a status code indicating
21 that the request has been canceled (rather than any other code).
22 Drivers should not call this routine or related routines, such as
24 usb_kill_urb or usb_unlink_anchored_urbs, after their disconnect method
25 has returned. The disconnect function should synchronize with a
26 driver's I/O routines to insure that all URB-related activity has
27 completed before it returns.
28 This request is asynchronous, however the HCD might call the ->complete
30 callback during unlink. Therefore when drivers call usb_unlink_urb,
31 they must not hold any locks that may be taken by the completion
32 function. Success is indicated by returning -EINPROGRESS, at which time
33 the URB will probably not yet have been given back to the device
34 driver. When it is eventually called, the completion function will see
35 urb->status == -ECONNRESET. Failure is indicated by usb_unlink_urb
36 returning any other value. Unlinking will fail when urb is not
37 currently “linked” (i.e., it was never submitted, or it was unlinked
38 before, or the hardware is already finished with it), even if the
39 completion handler has not yet run.
40 The URB must not be deallocated while this routine is running. In
42 particular, when a driver calls this routine, it must insure that the
43 completion handler cannot deallocate the URB.
44
46 -EINPROGRESS on success. See description for other values on failure.
47
49 [The behaviors and guarantees described below do not apply to virtual
50 root hubs but only to endpoint queues for physical USB devices.]
51 Host Controller Drivers (HCDs) place all the URBs for a particular
53 endpoint in a queue. Normally the queue advances as the controller
54 hardware processes each request. But when an URB terminates with an
55 error its queue generally stops (see below), at least until that URB's
56 completion routine returns. It is guaranteed that a stopped queue will
57 not restart until all its unlinked URBs have been fully retired, with
58 their completion routines run, even if that's not until some time after
59 the original completion handler returns. The same behavior and
60 guarantee apply when an URB terminates because it was unlinked.
61 Bulk and interrupt endpoint queues are guaranteed to stop whenever an
63 URB terminates with any sort of error, including -ECONNRESET, -ENOENT,
64 and -EREMOTEIO. Control endpoint queues behave the same way except that
65 they are not guaranteed to stop for -EREMOTEIO errors. Queues for
66 isochronous endpoints are treated differently, because they must
67 advance at fixed rates. Such queues do not stop when an URB encounters
68 an error or is unlinked. An unlinked isochronous URB may leave a gap in
69 the stream of packets; it is undefined whether such gaps can be filled
70 in.
71 Note that early termination of an URB because a short packet was
73 received will generate a -EREMOTEIO error if and only if the
74 URB_SHORT_NOT_OK flag is set. By setting this flag, USB device drivers
75 can build deep queues for large or complex bulk transfers and clean
76 them up reliably after any sort of aborted transfer by unlinking all
77 pending URBs at the first fault.
78 When a control URB terminates with an error other than -EREMOTEIO, it
80 is quite likely that the status stage of the transfer will not take
81 place.
82
84Kernel Hackers Manual 3.10 June 2019 USB_UNLINK_URB(9)