Re: 2.6.14-rc3-rt2

[Daniel Walker]

On Tue, 2005-10-04 at 23:41 +0530, Dinakar Guniguntala wrote:
> On Tue, Oct 04, 2005 at 10:54:30AM -0700, Daniel Walker wrote:
> > On Tue, 2005-10-04 at 23:28 +0530, Dinakar Guniguntala wrote:
> > 
> > > Nope doesnt help. I booted with this code change and I get the
> > > same message. 
> > > 
> > > I saw that the code change is in #ifdef CONFIG_HIGH_RES_TIMERS.
> > > I have already disabled CONFIG_HIGH_RES_TIMERS as Thomas Gleixner 
> > > suggested
> > 
> > Which code is #ifdef'd ?
> 
> Your code was in function smp_apic_timer_ipi_interrupt (right?) that
> is under CONFIG_HIGH_RES_TIMERS which was disabled.

Yeah, I see what your saying now ..

> > 
> > Is there any diversity in these messages , or is it always the same? Is
> > the CPU# ever different?
> > 
> 
> Sorry I should have put this up before.
> 
> 
> BUG: auditd:3587, possible softlockup detected on CPU#2!
>  [<c0144448>] softlockup_detected+0x39/0x46 (8)
>  [<c0144526>] softlockup_tick+0xd1/0xd3 (20)
>  [<c01112c7>] smp_apic_timer_interrupt+0xc0/0xcc (24)
>  [<c010396c>] apic_timer_interrupt+0x1c/0x24 (24)
>  [<c0102e8a>] sysenter_past_esp+0x2f/0x75 (44)

This patch should handle both cases . I would think if this doesn't
silence it, then it's something else..

Index: linux-2.6.13/arch/i386/kernel/apic.c
===================================================================
--- linux-2.6.13.orig/arch/i386/kernel/apic.c
+++ linux-2.6.13/arch/i386/kernel/apic.c
@@ -1153,6 +1153,14 @@ fastcall notrace void smp_apic_timer_ipi
 #if 0
 	profile_tick(CPU_PROFILING, regs);
 #endif
+	/*
+	 * If the task is currently running in user mode, don't
+	 * detect soft lockups.  If CONFIG_DETECT_SOFTLOCKUP is not
+	 * configured, this should be optimized out.
+	 */
+	if (user_mode(regs))
+		touch_light_softlockup_watchdog();
+
 	update_process_times(user_mode_vm(regs));
 	irq_exit();
 
@@ -1247,6 +1255,14 @@ inline void smp_local_timer_interrupt(st
 						per_cpu(prof_counter, cpu);
 		}
 #ifdef CONFIG_SMP
+		/*
+		 * If the task is currently running in user mode, don't
+		 * detect soft lockups.  If CONFIG_DETECT_SOFTLOCKUP is not
+		 * configured, this should be optimized out.
+		 */
+		if (user_mode(regs))
+			touch_light_softlockup_watchdog();
+
 		update_process_times(user_mode_vm(regs));
 #endif
 	}


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/