RE: [BISECTION RESULT] sched: revert cpu_clock to pre-27ec4407790d075c325e1f4da0a19c56953cce23 state

Previous message: [thread] [date] [author]
Next message: [thread] [date] [author]
From: Luck, Tony
Date: Tuesday, August 5, 2008 - 7:59 am

+u64 sched_clock_cpu(int cpu)
+{
+       if (unlikely(!sched_clock_running))
+               return 0;
+
+       return sched_clock();
+}

This looks like it will solve the issue on the boot cpu.  It
leaves a potential hole as we bring up other cpus if[1] there
are any printk() calls before they get to cpu_init() to
initialize ar.kr3.  This code path is all under arch/ia64
though ... so if we shoot ourselves in the foot, it will be
our own fault.

-Tony

[1] Currently there is just one Dprintk() call in this path
at the start of start_secondary() ... I could put it inside
#ifndef CONFIG_PRINTK_TIME so as not to surprise anyone that
enabled it.
--
Previous message: [thread] [date] [author]
Next message: [thread] [date] [author]

Messages in current thread:
[BISECTION RESULT] sched: revert cpu_clock to pre-27ec4407 ..., Nishanth Aravamudan, (Mon Aug 4, 12:46 pm)
Re: [BISECTION RESULT] sched: revert cpu_clock to pre-27ec ..., Nishanth Aravamudan, (Mon Aug 4, 3:45 pm)
Re: [BISECTION RESULT] sched: revert cpu_clock to pre-27ec ..., Nishanth Aravamudan, (Mon Aug 4, 4:30 pm)
RE: [BISECTION RESULT] sched: revert cpu_clock to pre-27ec ..., Luck, Tony, (Tue Aug 5, 7:59 am)
Re: [BISECTION RESULT] sched: revert cpu_clock to pre-27ec ..., Nishanth Aravamudan, (Tue Aug 5, 10:34 am)
Re: [BISECTION RESULT] sched: revert cpu_clock to pre-27ec ..., Nishanth Aravamudan, (Wed Aug 13, 12:29 pm)
Re: [BISECTION RESULT] sched: revert cpu_clock to pre-27ec ..., Nishanth Aravamudan, (Tue Aug 19, 3:19 pm)