From owner-p4-projects@FreeBSD.ORG  Wed Nov  2 17:29:52 2005
Return-Path: <owner-p4-projects@FreeBSD.ORG>
X-Original-To: p4-projects@freebsd.org
Delivered-To: p4-projects@freebsd.org
Received: by hub.freebsd.org (Postfix, from userid 32767)
	id 8A45716A421; Wed,  2 Nov 2005 17:29:51 +0000 (GMT)
X-Original-To: perforce@freebsd.org
Delivered-To: perforce@freebsd.org
Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125])
	by hub.freebsd.org (Postfix) with ESMTP id 4960116A420
	for <perforce@freebsd.org>; Wed,  2 Nov 2005 17:29:51 +0000 (GMT)
	(envelope-from bb+lists.freebsd.perforce@cyrus.watson.org)
Received: from repoman.freebsd.org (repoman.freebsd.org [216.136.204.115])
	by mx1.FreeBSD.org (Postfix) with ESMTP id 33F1143D62
	for <perforce@freebsd.org>; Wed,  2 Nov 2005 17:29:48 +0000 (GMT)
	(envelope-from bb+lists.freebsd.perforce@cyrus.watson.org)
Received: from repoman.freebsd.org (localhost [127.0.0.1])
	by repoman.freebsd.org (8.13.1/8.13.1) with ESMTP id jA2HTlOB095810
	for <perforce@freebsd.org>; Wed, 2 Nov 2005 17:29:48 GMT
	(envelope-from bb+lists.freebsd.perforce@cyrus.watson.org)
Received: (from perforce@localhost)
	by repoman.freebsd.org (8.13.1/8.13.1/Submit) id jA2HTleT095807
	for perforce@freebsd.org; Wed, 2 Nov 2005 17:29:47 GMT
	(envelope-from bb+lists.freebsd.perforce@cyrus.watson.org)
Date: Wed, 2 Nov 2005 17:29:47 GMT
Message-Id: <200511021729.jA2HTleT095807@repoman.freebsd.org>
X-Authentication-Warning: repoman.freebsd.org: perforce set sender to
	bb+lists.freebsd.perforce@cyrus.watson.org using -f
From: Robert Watson <rwatson@FreeBSD.org>
To: Perforce Change Reviews <perforce@freebsd.org>
Cc: 
Subject: PERFORCE change 86239 for review
X-BeenThere: p4-projects@freebsd.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: p4 projects tree changes <p4-projects.freebsd.org>
List-Unsubscribe: <http://lists.freebsd.org/mailman/listinfo/p4-projects>,
	<mailto:p4-projects-request@freebsd.org?subject=unsubscribe>
List-Archive: <http://lists.freebsd.org/pipermail/p4-projects>
List-Post: <mailto:p4-projects@freebsd.org>
List-Help: <mailto:p4-projects-request@freebsd.org?subject=help>
List-Subscribe: <http://lists.freebsd.org/mailman/listinfo/p4-projects>,
	<mailto:p4-projects-request@freebsd.org?subject=subscribe>
X-List-Received-Date: Wed, 02 Nov 2005 17:29:52 -0000

http://perforce.freebsd.org/chv.cgi?CH=86239

Change 86239 by rwatson@rwatson_zoo on 2005/11/02 17:29:20

	When hinted not to preempt, also don't IPI.

Affected files ...

.. //depot/projects/netsmp/src/sys/kern/sched_4bsd.c#3 edit

Differences ...

==== //depot/projects/netsmp/src/sys/kern/sched_4bsd.c#3 (text+ko) ====

@@ -1159,29 +1159,40 @@
 		cpu = NOCPU;
 		ke->ke_runq = &runq;
 	}
-	
-	if (single_cpu && (cpu != PCPU_GET(cpuid))) {
-	        kick_other_cpu(td->td_priority,cpu);
-	} else {
-		
-		if (!single_cpu) {
-			cpumask_t me = PCPU_GET(cpumask);
-			int idle = idle_cpus_mask & me;	
+
+	/*
+	 * Once we've established where to schedule the thread, we have to
+	 * decide if we want to preempt the current thread, or IPI another
+	 * CPU to start immediate execution there.
+	 *
+	 * Currently, suppressing preemption on the scheduling also
+	 * suppresses IPIs to shedule on another CPU, as it is assumed that
+	 * starting the thread running at once will result in poor timing of
+	 * events, lock contention, etc.
+	 */
+	if (!(flags & SRQ_NOPREEMPT)) {
+		if (single_cpu && (cpu != PCPU_GET(cpuid))) {
+		        kick_other_cpu(td->td_priority,cpu);
+		} else {
+			if (!single_cpu) {
+				cpumask_t me = PCPU_GET(cpumask);
+				int idle = idle_cpus_mask & me;	
 
-			if (!idle && ((flags & SRQ_INTR) == 0) &&
-			    (idle_cpus_mask & ~(hlt_cpus_mask | me)))
-				forwarded = forward_wakeup(cpu);
-		}
+				if (!idle && ((flags & SRQ_INTR) == 0) &&
+				    (idle_cpus_mask & ~(hlt_cpus_mask | me)))
+					forwarded = forward_wakeup(cpu);
+			}
 
-		if (!forwarded) {
-			if ((flags & SRQ_YIELDING) == 0 &&
-			    (flags & SRQ_NOPREEMPT) == 0 &&
-			    maybe_preempt(td))
-				return;
-			else
-				maybe_resched(td);
+			if (!forwarded) {
+				if ((flags & SRQ_YIELDING) == 0 &&
+				    maybe_preempt(td))
+					return;
+				else
+					maybe_resched(td);
+			}
 		}
-	}
+	} else
+		maybe_resched(td);
 	
 	if ((td->td_proc->p_flag & P_NOLOAD) == 0)
 		sched_load_add();