kern/45577: system unstable: pid 0 is killed: exceeded RLIMIT_CPU

[taca%back-street.net@localhost]

>Number:         45577
>Category:       kern
>Synopsis:       system unstable: pid 0 is killed: exceeded RLIMIT_CPU
>Confidential:   no
>Severity:       critical
>Priority:       high
>Responsible:    kern-bug-people
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Sun Nov 06 11:30:00 +0000 2011
>Originator:     Takahiro Kambe
>Release:        NetBSD 5.99.56
>Organization:
Takahiro Kambe
>Environment:
        
        
System: NetBSD edge.back-street.net 5.99.56 NetBSD 5.99.56 (VMWARE-F3) #109: 
Sun Nov 6 18:35:36 JST 2011 
taca%edge.back-street.net@localhost:/data/amd64/obj/sys/arch/amd64/compile/VMWARE-F3
 amd64
Architecture: x86_64
Machine: amd64
>Description:
        Latest kernel is unstable with these messages:

pid 0 is killed: exceeded RLIMIT_CPU
pid 1021 is killed: exceeded RLIMIT_CPU
...

>How-To-Repeat:
        Use kernel 6th Nov. 2011
>Fix:
        It seems that there are comparing 64bit time_t with long.

Index: sys/kern/kern_synch.c
===================================================================
RCS file: /cvsroot/src/sys/kern/kern_synch.c,v
retrieving revision 1.295
diff -u -u -r1.295 kern_synch.c
--- sys/kern/kern_synch.c       5 Oct 2011 20:37:40 -0000       1.295
+++ sys/kern/kern_synch.c       6 Nov 2011 11:21:02 -0000
@@ -1198,7 +1198,7 @@
        PROCLIST_FOREACH(p, &allproc) {
                struct lwp *l;
                struct rlimit *rlim;
-               long runtm;
+               time_t runtm;
                int sig;
 
                /* Increment sleep time (if sleeping), ignore overflow. */

>Unformatted: