Re: I/O permission control in OSKit-Mach

[Marcus Brinkmann]

On Thu, Oct 18, 2001 at 03:32:05PM -0400, Roland McGrath wrote:
> > Also, what about i386_io_perm_modify vs another cpu switching to another
> > thread in the same task?  The bad kinds of interactions I am worried
> > about is not so much the io_perm interface with itself, but with
> > task_terminate on the one hand and the scheduler on the other hand.  Do I
> > use a simple lock in the machine_task and make them all use it?
> 
> Yeah, that is probably what you need to do.  Needs some more though.
> Stick in an XXX comment or a #warning #if NCPUS>1 and we can think about it
> again when working on making SMP really work.

One thing that comes to mind is that it probably should diddle all bitmaps
in all TSSs for processors which run threads in the current task.  Otherwise
the task would have I/O ports enabled, and the thread calling
i386_io_port_modify would have them enabled, but other currently running
threads on other processors would not be enabled until they are rescheduled.

Thanks,
Marcus

-- 
`Rhubarb is no Egyptian god.' Debian http://www.debian.org brinkmd@debian.org
Marcus Brinkmann              GNU    http://www.gnu.org    marcus@gnu.org
Marcus.Brinkmann@ruhr-uni-bochum.de
http://www.marcus-brinkmann.de