Re: SMP and threads..

Adam K Kirchhoff <[email protected]> writes:
> Does anyone know if development is ongoing to resolve the conflict between
> SMP and kernel threads?

What you mean is userland pthreads operating with our new snazzy lwp
infrastructure that Nathan built, not "kernel threads". I think Nathan
is indeed working on fixing the problem, but I do not know his
schedule on that. Perhaps he will answer.

For reference, "kernel threads" are processes executing entirely
inside the kernel with no userland component. For example, you will
note in an average ps ax:

  PID TT STAT     TIME COMMAND
    0 ?? DKs   0:00.08 [swapper]
    1 ?? Ss    0:00.03 init 
    2 ?? DK    0:00.00 [pms0]
    3 ?? DK    0:00.00 [atapibus0]
    4 ?? DK    0:00.01 [usb0]
    5 ?? DK    0:00.00 [usbtask]
    6 ?? DK    0:00.00 [cardslot0]
    7 ?? DK    0:00.00 [cardslot1]
    8 ?? DK    0:25.91 [apm0]
    9 ?? DK    0:00.54 [pagedaemon]
   10 ?? DK    0:03.44 [reaper]
   11 ?? DK    1:55.16 [ioflush]
   12 ?? DK    0:00.72 [aiodoned]

Everything listed in []s is a kernel thread.


> Ever since the kernel threading was merged in,
> all threaded apps solidly lock up my box.
> 
> Adam
> 
> 

-- 
Perry E. Metzger		[email protected]