[Crash-utility] [PATCH] Add -r option to timer

Dave Anderson anderson at redhat.com
Mon Mar 11 20:36:20 UTC 2013



----- Original Message -----
> Hello Dave,
> 
> Please ignore the last mail. I change the style of expires, for "xx-xx"
> is a litter more difficult to be parsed. Please check.
> 
> At 2013-3-11 10:28, qiaonuohan wrote:
> > At 2013-3-9 2:26, Dave Anderson wrote:
> >> Shouldn't the NOW value for the ktime_get timers be adjusted appropriate
> >> to the EXPIRES values?
> >
> > Done. I made a mistaken here, thanks for pointing it out!
> >

Hi Qiao,

I've been tinkering around with this latest version, and it works pretty well.
I've changed a few things:

 (1) use option_not_supported() instead of command_not_supported() when used
     with pre-hrtimer kernels.
 (2) show the get_timer function in [brackets] on the same line as eacho of
     the "CLOCK: ... lines".
 (3) reversed the output of SOFTEXPIRES and EXPIRES, as it makes more
     sense to show the lower value of the range (SOFTEXPIRES) first.
 (4) displayed the NOW display (and changed it to "CURRENT") such that the
     value is properly aligned above with the SOFTEXPIRES/EXPIRES value,
     whichever is appropriate.
 (4) in the case of unused timers, don't bother displaying the current
     value or the headers -- it's just unnecessary clutter; it will just
     display the "CLOCK: 1  HRTIMER_CLOCK_BASE ..." line followed by
     "(empty)".

So for example, it would look like:

CPU: 63  HRTIMER_CPU_BASE: ffff88088e690f40
  CLOCK: 0  HRTIMER_CLOCK_BASE: ffff88088e690f48  [ktime_get_real]
  (empty)

  CLOCK: 1  HRTIMER_CLOCK_BASE: ffff88088e690f88  [ktime_get]
     CURRENT   
  3731204000000
   SOFTEXPIRES      EXPIRES         HRTIMER           FUNCTION    
  3731515000000  3731515000000  ffff88088e691040  ffffffff8109dff0  <tick_sched_timer>
  3736464999849  3736464999849  ffff88088e691260  ffffffff810d5fa0  <watchdog_timer_fn>

And I'll rewrite the help.c stuff, using i386 examples so that the output
examples can fit within 80 columns.

Thanks,
  Dave




More information about the Crash-utility mailing list