Source-Changes-HG archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

[src/netbsd-3]: src/sys/uvm Pullup rev 1.90 (requested by chs in ticket #914)



details:   https://anonhg.NetBSD.org/src/rev/13c71bff5ec5
branches:  netbsd-3
changeset: 577408:13c71bff5ec5
user:      jmc <jmc%NetBSD.org@localhost>
date:      Fri Oct 28 19:49:21 2005 +0000

description:
Pullup rev 1.90 (requested by chs in ticket #914)
 Remove the assertion in uvm_swapout_threads() about LSONPROC lwps
 not running on the same CPU as the swapper. l_stat is protected by
 sched_lock, which isn't held here, so we can race with that lwp
 starting to run and see its l_cpu not updated yet, as in PR 31870.
 we check l_stat again in uvm_swapout() while holding sched_lock,
 so the race itself is harmless.

diffstat:

 sys/uvm/uvm_glue.c |  5 ++---
 1 files changed, 2 insertions(+), 3 deletions(-)

diffs (26 lines):

diff -r 1bbbba5040f0 -r 13c71bff5ec5 sys/uvm/uvm_glue.c
--- a/sys/uvm/uvm_glue.c        Fri Oct 28 19:46:55 2005 +0000
+++ b/sys/uvm/uvm_glue.c        Fri Oct 28 19:49:21 2005 +0000
@@ -1,4 +1,4 @@
-/*     $NetBSD: uvm_glue.c,v 1.83.4.1 2005/05/22 19:39:28 snj Exp $    */
+/*     $NetBSD: uvm_glue.c,v 1.83.4.2 2005/10/28 19:49:21 jmc Exp $    */
 
 /*
  * Copyright (c) 1997 Charles D. Cranor and Washington University.
@@ -67,7 +67,7 @@
  */
 
 #include <sys/cdefs.h>
-__KERNEL_RCSID(0, "$NetBSD: uvm_glue.c,v 1.83.4.1 2005/05/22 19:39:28 snj Exp $");
+__KERNEL_RCSID(0, "$NetBSD: uvm_glue.c,v 1.83.4.2 2005/10/28 19:49:21 jmc Exp $");
 
 #include "opt_kgdb.h"
 #include "opt_kstack.h"
@@ -603,7 +603,6 @@
                        continue;
                switch (l->l_stat) {
                case LSONPROC:
-                       KDASSERT(l->l_cpu != curcpu());
                        continue;
 
                case LSRUN:



Home | Main Index | Thread Index | Old Index