On 30.08.21 18:09, Alan Stern wrote: > On Mon, Aug 30, 2021 at 05:11:53PM +0200, Oliver Neukum wrote: >> On 30.08.21 16:46, Alan Stern wrote: >>> Do you have alternative suggestions? I don't think we want special >>> interruptible versions of usb_control_msg() and usb_bulk_msg() just for >>> use by usbfs. >> Hi, >> >> why not just pass a flag? > We could. But you're ignoring the question I asked earlier in that > email: Is a 5-second uninterruptible delay (the default USB control > timeout) acceptable in general? We cannot change the five seconds, so this boils down to whether we can always handle signals when we need to send control messages. The answer to that is negative. Suspend/resume, block IO, probe, ioctl(). This list will be rather long. Regards Oliver
