Re: 2.6.32-rc5-mmotm1101 - lockdep whinge during early boot

[Martin Schwidefsky <schwidefsky@googlemail.com>]

That is imho caused by this patch snippet from the linux-next.patch from mmotd:

diff -puN arch/x86/kernel/tsc.c~linux-next arch/x86/kernel/tsc.c
--- a/arch/x86/kernel/tsc.c~linux-next
+++ a/arch/x86/kernel/tsc.c
@@ -763,14 +763,12 @@ void mark_tsc_unstable(char *reason)
 {
        if (!tsc_unstable) {
                tsc_unstable = 1;
-               printk(KERN_INFO "Marking TSC unstable due to %s\n", reason);
+               printk("Marking TSC unstable due to %s\n", reason);
                /* Change only the rating, when not registered */
                if (clocksource_tsc.mult)
-                       clocksource_mark_unstable(&clocksource_tsc);
-               else {
-                       clocksource_tsc.flags |= CLOCK_SOURCE_UNSTABLE;
+                       clocksource_change_rating(&clocksource_tsc, 0);
+               else
                        clocksource_tsc.rating = 0;
-               }
        }
 }


The upstream variant of mark_tsc_unstable uses clocksource_mark_unstable
which does not take the mutex. The reason is the clocksource mutex vs. the
cpu_add_remove_lock spinlock.

-- 
blue skies,
  Martin