On Sat, 2009-11-21 at 22:49 +0800, Wu Zhangjin wrote:
> From: Wu Zhangjin <wuzhangjin@xxxxxxxxx>
>
> This patch adds a cnt32_to_63() and MIPS c0 count based sched_clock(),
> which can provide high resolution. and also, two new kernel options are
> added. the HR_SCHED_CLOCK is used to enable/disable this sched_clock(),
> and the HT_SCHED_CLOCK_UPDATE is used to allow whether update the
> sched_clock() automatically or not.
>
> Without it, the Ftrace for MIPS will give useless timestamp information.
>
> (Because cnt32_to_63() needs to be called at least once per half period
> to work properly, Differ from the old version, this v1 revision set up a
> kernel timer to ensure the requirement of some MIPSs which have short c0
> count period.)
[...]
>
> diff --git a/arch/mips/kernel/csrc-r4k.c b/arch/mips/kernel/csrc-r4k.c
> index e95a3cd..5048989 100644
> --- a/arch/mips/kernel/csrc-r4k.c
> +++ b/arch/mips/kernel/csrc-r4k.c
> @@ -6,10 +6,62 @@
> * Copyright (C) 2007 by Ralf Baechle
> */
> #include <linux/clocksource.h>
> +#include <linux/cnt32_to_63.h>
> +#include <linux/timer.h>
> #include <linux/init.h>
>
> #include <asm/time.h>
>
> +/*
> + * MIPS' sched_clock implementation.
> + *
> + * because cnt32_to_63() needs to be called at least once per half period to
> + * work properly, and some of the MIPS' frequency is very low, perhaps a kernel
> + * timer is needed to be set up to ensure this requirement is always met.
> + * please refer to arch/arm/plat-orion/time.c and include/linux/cnt32_to_63.h
> + */
> +static unsigned long __maybe_unused tclk2ns_scale;
> +static unsigned long __maybe_unused tclk2ns_scale_factor;
> +
need to be:
#ifdef CONFIG_HR_SCHED_CLOCK
> +unsigned long long notrace sched_clock(void)
> +{
> + unsigned long long v = cnt32_to_63(read_c0_count());
> + return (v * tclk2ns_scale) >> tclk2ns_scale_factor;
> +}
> +
#endif
Regards,
Wu Zhangjin
