Re: [patch 1/2] Ignore stolen time in the softlockup watchdog
From: Prarit Bhargava <hidden>
Date: 2007-03-27 14:40:37
Also in:
lkml
Jeremy Fitzhardinge wrote:
quoted hunk ↗ jump to hunk
--- kernel/softlockup.c | 28 +++++++++++++++++++--------- 1 file changed, 19 insertions(+), 9 deletions(-) ===================================================================--- a/kernel/softlockup.c +++ b/kernel/softlockup.c@@ -17,8 +17,8 @@ static DEFINE_SPINLOCK(print_lock); -static DEFINE_PER_CPU(unsigned long, touch_timestamp); -static DEFINE_PER_CPU(unsigned long, print_timestamp); +static DEFINE_PER_CPU(unsigned long long, touch_timestamp); +static DEFINE_PER_CPU(unsigned long long, print_timestamp); static DEFINE_PER_CPU(struct task_struct *, watchdog_task); static int did_panic = 0;@@ -37,7 +37,7 @@ static struct notifier_block panic_block void touch_softlockup_watchdog(void) { - __raw_get_cpu_var(touch_timestamp) = jiffies; + __raw_get_cpu_var(touch_timestamp) = sched_clock(); }
I'd like to see this patch implement/fix touch_cpu_softlockup_watchdog and touch_softlockup_watchdog to mimic touch_nmi_watchdog's behaviour. See this now obsolete patch: http://lkml.org/lkml/2007/3/15/131 P.