Is it possible to make tty discipline shutdowns only in
normal kernel mode? (i.e. in close, in explicit TIOCSETD etc.)
No bottom halves, and, apparently, no masked interrupts.
F.e. I see no reasons why tty discipline is shutdown in tty_do_hangup,
when it is caused not by user level request but by hardware hangup.
It is my major fault, I missed this vital point.
(mainly, because in_interrupt() in earlier 2.1.* kernels
did not return correct state, but it is bad justification :( )
It's very unpleasant fact, it requires either more complicated
interaction between network driver and tty module or premature
overlocking of networking code. In any case, a lot of work.
I suspect, tty module can be changed with less pain. Is it correct?
Alexey Kuznetsov
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu