scheduler (sched_4bsd) questions

[Stephan Uphoff]

On Tue, 2004-10-05 at 09:03, Peter Holm wrote: 
> On Mon, Oct 04, 2004 at 12:42:53PM -0700, Julian Elischer wrote:
> 
> OK, I got a crash dump now, after a few modifications to kern_shutdown.c
> 
> There are however a few strange things worth noticing:
> 
> 1) The are no panic string:
> 
> Mounted root from ufs:/dev/ad0s1a.
> pid 1146: corrected slot count (2->1)
> [thread 100796]
> Stopped at      sched_add+0x13: movl    0x14c(%esi),%ebx
> 
> 2) The gdb stack trace gets a bit weird at:
> 
> #8  0xc07812da in calltrap () at ../../../i386/i386/exception.s:140
> #9  0xc05f0018 in flock (td=0x0, uap=0x0) at ../../../kern/kern_descrip.c:2138
> #10 0xc0619fd1 in setrunqueue (td=0xc2319180, flags=0x0) at kern_switch.c:521
> #11 0xc061921f in sched_wakeup (td=0xc2319180) at ../../../kern/sched_4bsd.c:859
> 
> Where did flock() come from?
> 
> The full console output is at http://www.holm.cc/stress/log/cons82.html
> 
> - Peter

I am still puzzled.
My newest pet theory is that the sorting of the kg_runq is corrupted
before setrunqueue is called.
Directly changing td_priority while the thread is on the run queue would
be an explanation.
However the only instance that I found is what I think is a rare
condition where sleepq_resume_thread may be called while the thread is
on a runqueue. (John - what did I miss this time ...)

Peter could you try this patch?

Index: subr_sleepqueue.c
===================================================================
RCS file: /cvsroot/src/sys/kern/subr_sleepqueue.c,v
retrieving revision 1.11
diff -u -r1.11 subr_sleepqueue.c
--- subr_sleepqueue.c   19 Aug 2004 11:31:41 -0000      1.11
+++ subr_sleepqueue.c   10 Oct 2004 18:18:55 -0000
@@ -642,7 +642,7 @@
        /* Adjust priority if requested. */
        MPASS(pri == -1 || (pri >= PRI_MIN && pri <= PRI_MAX));
        if (pri != -1 && td->td_priority > pri)
-               td->td_priority = pri;
+               sched_prio(td, pri);
        setrunnable(td);
        mtx_unlock_spin(&sched_lock);
}

Should it crash again could you walk the kg_runq to verify the sorting?

Thanks
	Stephan