Just over a year ago (!) I had this brief exchange with Alan Stern:
>> It seems that the signal that can be used with USBDEVFS_DISCSIGNAL and
>> in usbdevfs_urb.signr is limited to the real-time signals SIGRTMIN to
>> SIGRTMAX. What's the rationale for this restriction? I believe that a
>> process can kill() itself with any signal number, can't it? I was
>> planning to use SIGIO for usbdevfs_urb.signr and SIGTERM (uncaught) for
>> USBDEVFS_DISCSIGNAL. I don't think I'll have a problem with using
>> SIGRTMIN+n instead, but I'm curious to know if there's some subtle
>> problem with the non-real-time signals that I should be aware of.
>
> I don't know of any reason for this restriction.
Since no-one else could think of a reason either, I offer the following
patch which allows any signal to be used with USBDEVFS_DISCSIGNAL and
usbdevfs_urb.signr.
Signed-off-by: Phil Endecott <usbpatch@chezphil.org>
Cc: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
return -EINVAL;
if (!uurb->buffer)
return -EINVAL;
- if (uurb->signr != 0 && (uurb->signr < SIGRTMIN ||
- uurb->signr > SIGRTMAX))
- return -EINVAL;
if (!(uurb->type == USBDEVFS_URB_TYPE_CONTROL &&
(uurb->endpoint & ~USB_ENDPOINT_DIR_MASK) == 0)) {
ifnum = findintfep(ps->dev, uurb->endpoint);
if (copy_from_user(&ds, arg, sizeof(ds)))
return -EFAULT;
- if (ds.signr != 0 && (ds.signr < SIGRTMIN || ds.signr > SIGRTMAX))
- return -EINVAL;
ps->discsignr = ds.signr;
ps->disccontext = ds.context;
return 0;