Messages in this thread |  | | From | Arnd Bergmann <> | Subject | Re: tty_lock held during transmit wait in close: still unresolved | Date | Fri, 27 May 2011 15:53:17 +0200 |
| |
On Friday 27 May 2011, Jiri Slaby wrote: > Actually the best would be to get rid of tty_lock completely. Here I > have a patch removing tty_lock from wait_until_sent path. I've been > using a kernel with the patch applied for about month without any > problems. But I have only 8250-based serial.
I agree that this would be the best option. The only reason I had for holding the tty_lock in wait_until_sent is that it the BKL has been held there tradiditonally. I could not find anything that would break without holding it, but I also didn't look very hard.
Arnd
|  |