Hi!
> In usb.c, we could patch usb_bulk_msg() to check for
> a NULL <dev> and return -EINVAL if NULL.
>
> [usb_submit_urb() already hanles dev == NULL.]
>
> In usb.h, we could change __create_pipe() to return 0
> if <dev> is NULL. That would take care of all uses of
> usb_snd{type}pipe(). Or should we limit it to
> usb_sndbulkpipe() ?
>
> This latter change won't cause an error (just avoid an
> oops).
>
> Is anyone concerned about these suggested changes,
> either from a performance standpoint or the behavior
> of __create_pipe()?
>
> Matt, Pavel -- have either of you tried to force the
> problem with these kinds of changes? The patches
> are trivial, of course.
I can tell you that I'll not be able to reproduce the problem with
this kind of changes, (I'm far away from my machine.) just because it
was pretty hard to create the problem even without that tests. (I had
to put mdelays to code to make failures more common :-).
Pavel
--
The best software in life is free (not shareware)! Pavel
GCM d? s-: !g p?:+ au- a--@ w+ v- C++@ UL+++ L++ N++ E++ W--- M- Y- R+
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
