Subject: Re: 1.5ALPHA_2 i386 panic in suspendsched()
To: enami tsugutomo <email@example.com>
From: Manuel Bouyer <firstname.lastname@example.org>
Date: 09/18/2000 09:50:13
On Mon, Sep 18, 2000 at 12:59:20PM +0900, enami tsugutomo wrote:
> Manuel Bouyer <email@example.com> writes:
> > Going this way, I guess the run queue locking could happen later too.
> BTW, why suspendsched() is necessary to traverse run queues? I think
> it is enough to remove from run queue when traversing allprocs if it
> is on a queue. At least, it is bad that processes runnable but
> swapped out are left runnable after suspendsched().
If I understood it properly, you don't have a way when traversing allprocs
to know whenever a proc is on a run queue or not: a process may be SRUN but
swapped out, in which case it's not on a run queue.
Manuel Bouyer, LIP6, Universite Paris VI. Manuel.Bouyer@lip6.fr