On 24 June 2011 14:40, Dave Anderson <anderson(a)redhat.com> wrote:
 And for that matter, since tty->read_cnt is 0 above,
 your original question as to how that code path was taken to
 begin with is also valid.
 
Surely that implies tty->read_cnt has been modified since it was tested, and
hence you're looking at concurrency issues?