1USB_SUBMIT_URB(9) USB Core APIs USB_SUBMIT_URB(9)
2
6 usb_submit_urb - issue an asynchronous transfer request for an endpoint
7
9 int usb_submit_urb(struct urb * urb, gfp_t mem_flags);
10
12 urb
13 pointer to the urb describing the request
14 mem_flags
16 the type of memory to allocate, see kmalloc for a list of valid
17 options for this.
18
20 This submits a transfer request, and transfers control of the URB
21 describing that request to the USB subsystem. Request completion will
22 be indicated later, asynchronously, by calling the completion handler.
23 The three types of completion are success, error, and unlink (a
24 software-induced fault, also called “request cancellation”).
25 URBs may be submitted in interrupt context.
27 The caller must have correctly initialized the URB before submitting
29 it. Functions such as usb_fill_bulk_urb and usb_fill_control_urb are
30 available to ensure that most fields are correctly initialized, for the
31 particular kind of transfer, although they will not initialize any
32 transfer flags.
33 Successful submissions return 0; otherwise this routine returns a
35 negative error number. If the submission is successful, the complete
36 callback from the URB will be called exactly once, when the USB core
37 and Host Controller Driver (HCD) are finished with the URB. When the
38 completion function is called, control of the URB is returned to the
39 device driver which issued the request. The completion handler may then
40 immediately free or reuse that URB.
41 With few exceptions, USB device drivers should never access URB fields
43 provided by usbcore or the HCD until its complete is called. The
44 exceptions relate to periodic transfer scheduling. For both interrupt
45 and isochronous urbs, as part of successful URB submission
46 urb->interval is modified to reflect the actual transfer period used
47 (normally some power of two units). And for isochronous urbs,
48 urb->start_frame is modified to reflect when the URB´s transfers were
49 scheduled to start. Not all isochronous transfer scheduling policies
50 will work, but most host controller drivers should easily handle ISO
51 queues going from now until 10-200 msec into the future.
52 For control endpoints, the synchronous usb_control_msg call is often
54 used (in non-interrupt context) instead of this call. That is often
55 used through convenience wrappers, for the requests that are
56 standardized in the USB 2.0 specification. For bulk endpoints, a
57 synchronous usb_bulk_msg call is available.
58
60 URBs may be submitted to endpoints before previous ones complete, to
61 minimize the impact of interrupt latencies and system overhead on data
62 throughput. With that queuing policy, an endpoint´s queue would never
63 be empty. This is required for continuous isochronous data streams, and
64 may also be required for some kinds of interrupt transfers. Such
65 queuing also maximizes bandwidth utilization by letting USB controllers
66 start work on later requests before driver software has finished the
67 completion processing for earlier (successful) requests.
68 As of Linux 2.6, all USB endpoint transfer queues support depths
70 greater than one. This was previously a HCD-specific behavior, except
71 for ISO transfers. Non-isochronous endpoint queues are inactive during
72 cleanup after faults (transfer errors or cancellation).
73
75 Periodic transfers (interrupt or isochronous) are performed repeatedly,
76 using the interval specified in the urb. Submitting the first urb to
77 the endpoint reserves the bandwidth necessary to make those transfers.
78 If the USB subsystem can´t allocate sufficient bandwidth to perform the
79 periodic request, submitting such a periodic request should fail.
80 For devices under xHCI, the bandwidth is reserved at configuration
82 time, or when the alt setting is selected. If there is not enough bus
83 bandwidth, the configuration/alt setting request will fail. Therefore,
84 submissions to periodic endpoints on devices under xHCI should never
85 fail due to bandwidth constraints.
86 Device drivers must explicitly request that repetition, by ensuring
88 that some URB is always on the endpoint´s queue (except possibly for
89 short periods during completion callacks). When there is no longer an
90 urb queued, the endpoint´s bandwidth reservation is canceled. This
91 means drivers can use their completion handlers to ensure they keep
92 bandwidth they need, by reinitializing and resubmitting the
93 just-completed urb until the driver longer needs that periodic
94 bandwidth.
95
97 The general rules for how to decide which mem_flags to use are the same
98 as for kmalloc. There are four different possible values; GFP_KERNEL,
99 GFP_NOFS, GFP_NOIO and GFP_ATOMIC.
100 GFP_NOFS is not ever used, as it has not been implemented yet.
102 GFP_ATOMIC is used when (a) you are inside a completion handler, an
104 interrupt, bottom half, tasklet or timer, or (b) you are holding a
105 spinlock or rwlock (does not apply to semaphores), or (c)
106 current->state != TASK_RUNNING, this is the case only after you´ve
107 changed it.
108 GFP_NOIO is used in the block io path and error handling of storage
110 devices.
111 All other situations use GFP_KERNEL.
113 Some more specific rules for mem_flags can be inferred, such as (1)
115 start_xmit, timeout, and receive methods of network drivers must use
116 GFP_ATOMIC (they are called with a spinlock held); (2) queuecommand
117 methods of scsi drivers must use GFP_ATOMIC (also called with a
118 spinlock held); (3) If you use a kernel thread with a network driver
119 you must use GFP_NOIO, unless (b) or (c) apply; (4) after you have done
120 a down you can use GFP_KERNEL, unless (b) or (c) apply or your are in a
121 storage driver´s block io path; (5) USB probe and disconnect can use
122 GFP_KERNEL unless (b) or (c) apply; and (6) changing firmware on a
123 running storage or net device uses GFP_NOIO, unless b) or c) apply
124
126Kernel Hackers Manual 2.6. June 2019 USB_SUBMIT_URB(9)