Hello Bo, On Mon, Sep 14, 2009 at 12:50:32PM +0200, Bo Hansen wrote: > I tried to apply your patch and enabled ftrace but have not been able to > reproduce > the panic. > It should also be noted that the panic is much more rare if I apply the > first patch you > sent. Revert this patch, only use the second. And optimally add a BUG() in hrtimer_interrupt_hanging (in kernel/hrtimer.c) after the printk. > The only panic I got was using your first patch and enabling ftrace. I > have not yet had > the time to try changing pr_info/pr_emerg. In the meantime I understood why the 2nd printk doesn't appear on the serial console. That's because "non-atomic" consoles don't print anything in atomic context (with PREEMPT_RT=y) > It seems to run stable when something like: "hrtimer: interrupt too > slow, forcing clock > min delta to 480771 ns" is found in the dmesg output, but I cannot be > sure as I don't > know if this is also written to the log just before the panic. Actually this is a bug because then min_delta_ns is decreased. With the BUG as suggested above this message would be enough to debug this further. > Unfortunately I don't have much time for testing this at the moment but > I'll get back > as soon as possible. :-| Best regards Uwe -- Pengutronix e.K. | Uwe Kleine-König | Industrial Linux Solutions | http://www.pengutronix.de/ | -- To unsubscribe from this list: send the line "unsubscribe linux-rt-users" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
