Aren
/
NetBSD
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

sa_switch: remove cpu hand-off in "Case 2". 

it causes severe unfairness in some cases.
eg. when a process has many lwps doing poll.

no objections on tech-kern@.
This commit is contained in:
yamt 2005-10-08 06:37:12 +00:00
parent c610dbbdc2
commit 2db36492c6
1 changed files with 4 additions and 14 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

18
sys/kern/kern_sa.c
View File

@ -1,4 +1,4 @@
/* $NetBSD: kern_sa.c,v 1.66 2005/10/02 17:51:27 chs Exp $ */
/* $NetBSD: kern_sa.c,v 1.67 2005/10/08 06:37:12 yamt Exp $ */
/*-
* Copyright (c) 2001, 2004, 2005 The NetBSD Foundation, Inc.
@ -37,7 +37,7 @@
*/
#include <sys/cdefs.h>
__KERNEL_RCSID(0, "$NetBSD: kern_sa.c,v 1.66 2005/10/02 17:51:27 chs Exp $");
__KERNEL_RCSID(0, "$NetBSD: kern_sa.c,v 1.67 2005/10/08 06:37:12 yamt Exp $");
#include <sys/param.h>
#include <sys/systm.h>
@ -977,20 +977,10 @@ sa_switch(struct lwp *l, struct sadata_upcall *sau, int type)
* on the VP, but we're going back to sleep without
* having returned to userland and delivering the
* SA_UNBLOCKED upcall (select and poll cause this
* kind of behavior a lot). We just switch back to the
* LWP that had been running and let it have another
* go. If the LWP on the VP was idling, don't make it
* run again, though.
* kind of behavior a lot).
*/
freesau = sau;
if (vp->savp_lwp->l_flag & L_SA_YIELD)
l2 = NULL;
else {
/* XXXUPSXXX Unfair advantage for l2 ? */
l2 = vp->savp_lwp;
if (l2->l_stat != LSRUN || (l2->l_flag & L_INMEM) == 0)
l2 = NULL;
}
l2 = NULL;
} else {
/* NOTREACHED */
panic("sa_vp empty");