[PATCH] 2.5.43-tick

[PATCH] 2.5.43-tick

[Russell King]

This patch appears not to be in 2.5.43, but applies cleanly.

On some ARM platforms, CLOCK_TICK_RATE is not a constant (it is specified
at boot time), which means that TICK_USEC/TICK_NSEC is not constant.
We therefore can not initialise static variables with these definitions;
they must be done at run time.

 kernel/timer.c |    7 +++++--
 1 files changed, 5 insertions, 2 deletions

diff -ur orig/kernel/timer.c linux/kernel/timer.c
--- orig/kernel/timer.c	Wed Oct 16 09:17:13 2002
+++ linux/kernel/timer.c	Wed Oct 16 09:15:00 2002
@@ -376,8 +376,8 @@
 /*
  * Timekeeping variables
  */
-unsigned long tick_usec = TICK_USEC; 		/* ACTHZ   period (usec) */
-unsigned long tick_nsec = TICK_NSEC(TICK_USEC);	/* USER_HZ period (nsec) */
+unsigned long tick_usec;	/* ACTHZ   period (usec) */
+unsigned long tick_nsec;	/* USER_HZ period (nsec) */
 
 /* The current time */
 struct timespec xtime __attribute__ ((aligned (16)));
@@ -1069,6 +1069,9 @@
 void __init init_timers(void)
 {
 	int i, j;
+
+	tick_usec = TICK_USEC;
+	tick_nsec = TICK_NSEC(TICK_USEC);
 
 	for (i = 0; i < NR_CPUS; i++) {
 		tvec_base_t *base;

Re: [PATCH] 2.5.43-tick

[george anzinger]

Russell King wrote:
> 
> This patch appears not to be in 2.5.43, but applies cleanly.
> 
> On some ARM platforms, CLOCK_TICK_RATE is not a constant (it is specified
> at boot time), which means that TICK_USEC/TICK_NSEC is not constant.
> We therefore can not initialise static variables with these definitions;
> they must be done at run time.

I must be missing something here.  Why can't both be done? 
That is, leave timer.c alone and add code to the arch time
init to correct the values?

-g
> 
>  kernel/timer.c |    7 +++++--
>  1 files changed, 5 insertions, 2 deletions
> 
> diff -ur orig/kernel/timer.c linux/kernel/timer.c
> --- orig/kernel/timer.c Wed Oct 16 09:17:13 2002
> +++ linux/kernel/timer.c        Wed Oct 16 09:15:00 2002
> @@ -376,8 +376,8 @@
>  /*
>   * Timekeeping variables
>   */
> -unsigned long tick_usec = TICK_USEC;           /* ACTHZ   period (usec) */
> -unsigned long tick_nsec = TICK_NSEC(TICK_USEC);        /* USER_HZ period (nsec) */
> +unsigned long tick_usec;       /* ACTHZ   period (usec) */
> +unsigned long tick_nsec;       /* USER_HZ period (nsec) */
> 
>  /* The current time */
>  struct timespec xtime __attribute__ ((aligned (16)));
> @@ -1069,6 +1069,9 @@
>  void __init init_timers(void)
>  {
>         int i, j;
> +
> +       tick_usec = TICK_USEC;
> +       tick_nsec = TICK_NSEC(TICK_USEC);
> 
>         for (i = 0; i < NR_CPUS; i++) {
>                 tvec_base_t *base;
> 
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/

-- 
George Anzinger   george@mvista.com
High-res-timers: 
http://sourceforge.net/projects/high-res-timers/
Preemption patch:
http://www.kernel.org/pub/linux/kernel/people/rml