Re: BUG on booting 2.6.27-rc1

[Jeremy Fitzhardinge <jeremy@goop.org>]

Steve Wise wrote:
> Anybody seen this?
>
> Its on x86_64, FC7 distro quad core cpu.  I have 2 similar systems, 
> and one boots up ok but the other hits this on -rc1 right off the bat.
>
>
>
>
> Starting udev: ------------[ cut here ]------------
> kernel BUG at kernel/stop_machine.c:151!
> invalid opcode: 0000 [1] SMP
> CPU 0
> Modules linked in: dm_snapshot dm_zero dm_mirror dm_log dm_mod 
> pata_jmicron ata_generic ata_piix libata sd_mod scsi_mod ext3 jbd 
> mbcache [last unloaded: scsi_wait_scan]
> Pid: 960, comm: modprobe Not tainted 2.6.27-rc1 #7
> RIP: 0010:[<ffffffff8105be33>]  [<ffffffff8105be33>] 
> __stop_machine+0x134/0x1ea
> RSP: 0018:ffff88012e5dbd68  EFLAGS: 00010286
> RAX: 00000000fffffff3 RBX: ffff88012b9b1d40 RCX: 0000000000000206

RAX = -EACCES

Presumably that's being returned from security_task_setscheduler(), 
since it doesn't otherwise appear in __sched_setscheduler().  We 
probably shouldn't call that if !user.

diff -r 4f507097262d kernel/sched.c
--- a/kernel/sched.c	Fri Aug 01 13:14:44 2008 -0700
+++ b/kernel/sched.c	Fri Aug 01 14:13:59 2008 -0700
@@ -4998,19 +4998,21 @@
 			return -EPERM;
 	}
 
-#ifdef CONFIG_RT_GROUP_SCHED
-	/*
-	 * Do not allow realtime tasks into groups that have no runtime
-	 * assigned.
-	 */
-	if (user
-	    && rt_policy(policy) && task_group(p)->rt_bandwidth.rt_runtime == 0)
-		return -EPERM;
-#endif
-
-	retval = security_task_setscheduler(p, policy, param);
-	if (retval)
-		return retval;
+	if (user) {
+#ifdef CONFIG_RT_GROUP_SCHED
+		/*
+		 * Do not allow realtime tasks into groups that have no runtime
+		 * assigned.
+		 */
+		if (rt_policy(policy) && task_group(p)->rt_bandwidth.rt_runtime == 0)
+			return -EPERM;
+#endif
+
+		retval = security_task_setscheduler(p, policy, param);
+		if (retval)
+			return retval;
+	}
+
 	/*
 	 * make sure no PI-waiters arrive (or leave) while we are
 	 * changing the priority of the task:



    J