From: Alexander Beregalov <a.beregalov@xxxxxxxxx> Date: Thu, 18 Jun 2009 14:47:57 +0400 > 2009/6/18 David Miller <davem@xxxxxxxxxxxxx>: >> From: Alexander Beregalov <a.beregalov@xxxxxxxxx> >> Date: Wed, 17 Jun 2009 19:00:18 +0400 >> >>> Config is attached. >> >> Aha, the key is non-SMP build. >> >>> b696fdc259f0d94348a9327bed352fac44d4883d is first bad commit >>> sparc64: Defer cpu_data() setup until end of per-cpu data initialization. >>> >>> Does it make sense to you? >> >> I see the problem, please try this patch: >> >> sparc64: Fix UP bootup regression. > > Hm, no, still the same problem. The only thing that explains your bug is that cpu_data()->clock_tick is zero. And that would be zero if you have SMP disabled because of the bug I am fixing. of_fill_in_cpu_data() should find the right value and fill it in after my patch from paging_init(), which is long before time_init() which is where your problem triggers. See what value sparc64_init_timers() ends up using for "freq". Thanks! -- To unsubscribe from this list: send the line "unsubscribe sparclinux" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
