On Wed, Oct 14, 2020 at 06:41:23PM +0200, Sebastian Andrzej Siewior wrote: > On 2020-10-14 12:27:21 [-0400], Alan Stern wrote: > > > --- a/drivers/usb/core/hcd.c > > > +++ b/drivers/usb/core/hcd.c > > > @@ -746,9 +746,6 @@ static int rh_call_control (struct usb_h > > > * Root Hub interrupt transfers are polled using a timer if the > > > * driver requests it; otherwise the driver is responsible for > > > * calling usb_hcd_poll_rh_status() when an event occurs. > > > - * > > > - * Completions are called in_interrupt(), but they may or may not > > > - * be in_irq(). > > > > This comment should not be removed; instead it should be changed to say > > that completion handlers are called with interrupts disabled. > > The timer callback: > rh_timer_func() -> usb_hcd_poll_rh_status() > > invokes the function with enabled interrupts. Well, it doesn't change the interrupt settings. It might call usb_hcd_poll_rh_status() with interrupts enabled or disabled, depending on how it was called originally. But that wasn't what I meant. usb_hcd_poll_rh_status() calls usb_hcd_giveback_urb() with interrupts disabled always, and that routine may call __usb_hcd_giveback_urb(), which calls urb->complete(urb); In this case the completion handler would be invoked with interrupts disabled. Alternatively, __usb_hcd_giveback_urb() may be invoked from a BH handler, in which case the completion handler will run in softirq context with interrupts enabled. So I guess it would be best to say that completion handlers may be called with interrupts enabled or disabled. Or you might want to put such a comment in __usb_hcd_giveback_urb(). > > > @@ -1691,7 +1690,6 @@ static void usb_giveback_urb_bh(unsigned > > > * @hcd: host controller returning the URB > > > * @urb: urb being returned to the USB device driver. > > > * @status: completion status code for the URB. > > > - * Context: in_interrupt() > > > > The comment should be changed to say that the routine runs in a BH > > handler (or however you want to express it). > > Do you mean usb_hcd_giveback_urb() runs in BH context or that the > completion callback of the URB runs in BH context? Actually I meant that usb_hcd_giveback_urb_bh() runs in BH context. Sorry, I got confused about the location of this hunk. To be explicit: The comment for usb_hcd_giveback_urb() should say that the function expects to be called with interrupts disabled (whether the context is task, atomic, BH, interrupt, etc. doesn't matter). > The completion callback of the URB may run in BH or IRQ context > depending on HCD. > > > > --- a/drivers/usb/core/message.c > > > +++ b/drivers/usb/core/message.c > > > > > @@ -934,7 +939,7 @@ int usb_get_device_descriptor(struct usb > > > /* > > > * usb_set_isoch_delay - informs the device of the packet transmit delay > > > * @dev: the device whose delay is to be informed > > > - * Context: !in_interrupt() > > > + * Context: can sleep > > > > Why is this comment different from all the others? > > It says !in_interrupt() which is also true for preempt-disabled regions. > But the caller must not have preemption disabled. "can sleep" is more > obvious as what it needs. But all the other comments in this patch say: * Context: task context, might sleep. Why doesn't this comment say the same thing? Alan Stern