From: Arne R. <ag...@po...> - 2009-01-31 07:48:29
|
[Unintentionally dropped iet-dev the first time. Resending] Am Freitag, den 30.01.2009, 20:21 +0100 schrieb Andreas Florath: > Hello Arne, > > think I found the problem: > > in kernel/target.c you added > > if (!err) > nthread_stop(target); > > The ! is wrong here. It should be: > > if (err) > nthread_stop(target); > > Attached you can find version 4 of the patch ;-) > This works for me with and without the worker pool. Great job, thank you. > BUT: The BUG (posted 2009-01-28 20:09) still remains: when there is a > problem creating the threads for the worker pool, this BUG occurs. My > first analyze: there might be a raise condition between > nthread_stop() and some nthread usage. Ok, just to see if I got your point: the patch fixes the oops during "normal operation", but in case sth. goes wrong you assume it will reoccur? I'll look into that as well. Unless there turns out to be an issue I will merge the patch. Thanks again, Arne |