linux-kernel - Re: [PATCH 2/2] tracing: add high precision version of trace_clock

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-Id: <1255767839.10178.12.camel@falcon>
Date:	Sat, 17 Oct 2009 16:23:59 +0800
From:	Wu Zhangjin <wuzhangjin@...il.com>
To:	linux-kernel@...r.kernel.org
Cc:	linux-mips <linux-mips@...ux-mips.org>,
	Ralf Baechle <ralf@...ux-mips.org>,
	Thomas Gleixner <tglx@...utronix.de>,
	Steven Rostedt <rostedt@...dmis.org>,
	Ingo Molnar <mingo@...e.hu>,
	Manuel Lauss <manuel.lauss@...glemail.com>,
	Atsushi Nemoto <anemo@....ocn.ne.jp>
Subject: Re: [PATCH 2/2] tracing: add high precision version of
 trace_clock_local() for MIPS

On Fri, 2009-10-16 at 19:38 +0800, Wu Zhangjin wrote:
> By default, trace_clock_local() call the sched_clock() to get timestamp,
> in x86, there is a tsc(64bit) based sched_clock(), but in MIPS, the
> sched_clock() is jiffies based, which can not give enough
> precision(about 10ms with 1000 HZ). so, we need to implement a higher
> precision version for MIPS.
> 
> in MIPS, there is a tsc like clock counter. based on it, we can
> implement a new trace_clock_local(), but because the clock counter is
> only 32bit long, we need to handle rollover for it.  fortunately, there
> a high-level abstraction(cyclecounter) of this action implemented in
> include/linux/clocksource.h, which helps to handle the details behind,
> we just use it!
> 
> Signed-off-by: Wu Zhangjin <wuzhangjin@...il.com>
> ---
>  arch/mips/kernel/Makefile      |    6 +++
>  arch/mips/kernel/trace_clock.c |   67 ++++++++++++++++++++++++++++++++++++++++
>  2 files changed, 73 insertions(+), 0 deletions(-)
>  create mode 100644 arch/mips/kernel/trace_clock.c
> 
> diff --git a/arch/mips/kernel/Makefile b/arch/mips/kernel/Makefile
> index 8e26e9c..f228868 100644
> --- a/arch/mips/kernel/Makefile
> +++ b/arch/mips/kernel/Makefile
> @@ -8,6 +8,10 @@ obj-y		+= cpu-probe.o branch.o entry.o genex.o irq.o process.o \
>  		   ptrace.o reset.o setup.o signal.o syscall.o \
>  		   time.o topology.o traps.o unaligned.o watch.o
>  
> +ifdef CONFIG_FUNCTION_TRACER
> +CFLAGS_REMOVE_trace_clock.o = -pg
> +endif
> +
>  obj-$(CONFIG_CEVT_BCM1480)	+= cevt-bcm1480.o
>  obj-$(CONFIG_CEVT_R4K_LIB)	+= cevt-r4k.o
>  obj-$(CONFIG_MIPS_MT_SMTC)	+= cevt-smtc.o
> @@ -24,6 +28,8 @@ obj-$(CONFIG_SYNC_R4K)		+= sync-r4k.o
>  obj-$(CONFIG_STACKTRACE)	+= stacktrace.o
>  obj-$(CONFIG_MODULES)		+= mips_ksyms.o module.o
>  
> +obj-$(CONFIG_FTRACE)		+= trace_clock.o
> +
>  obj-$(CONFIG_CPU_LOONGSON2)	+= r4k_fpu.o r4k_switch.o
>  obj-$(CONFIG_CPU_MIPS32)	+= r4k_fpu.o r4k_switch.o
>  obj-$(CONFIG_CPU_MIPS64)	+= r4k_fpu.o r4k_switch.o
> diff --git a/arch/mips/kernel/trace_clock.c b/arch/mips/kernel/trace_clock.c
> new file mode 100644
> index 0000000..435cd56
> --- /dev/null
> +++ b/arch/mips/kernel/trace_clock.c
> @@ -0,0 +1,67 @@
> +/*
> + * This file is subject to the terms and conditions of the GNU General Public
> + * License.  See the file "COPYING" in the main directory of this archive for
> + * more details.
> + *
> + * Copyright (C) 2009 Lemote Inc. & DSLab, Lanzhou University, China
> + * Author: Wu Zhangjin <wuzj@...ote.com>
> + */
> +
> +#include <linux/clocksource.h>
> +#include <linux/sched.h>
> +
> +#include <asm/time.h>
> +
> +/*
> + * MIPS clock counter based timecounter.
> + */
> +
> +struct timecounter mips_tc;
> +
> +/*
> + * trace_clock_local(): the simplest and least coherent tracing clock.
> + *
> + * Useful for tracing that does not cross to other CPUs nor
> + * does it go through idle events.
> + */
> +u64 trace_clock_local(void)
> +{
> +	unsigned long flags;
> +	u64 clock;
> +
> +	raw_local_irq_save(flags);
> +
> +	preempt_disable_notrace();
> +
> +	clock = timecounter_read(&mips_tc);
> +
> +	preempt_enable_no_resched_notrace();
> +
> +	raw_local_irq_restore(flags);
> +
> +	return clock;
> +}
> +
> +static cycle_t mips_cc_read(const struct cyclecounter *cc)
> +{
> +	return read_c0_count();
> +}
> +
> +static struct cyclecounter mips_cc = {
> +	.read = mips_cc_read,
> +	.mask = CLOCKSOURCE_MASK(32),
> +	.shift = 32,
> +};
> +
> +static int __init mips_timecounter_init(void)
> +{
> +	uint64_t mips_freq = mips_hpt_frequency;
> +

we must ensure mips_hpt_frequency is initialized before using it.

> +	mips_cc.mult = div_sc((unsigned long)mips_freq, NSEC_PER_SEC, 32);
> +
> +	timecounter_init(&mips_tc, &mips_cc, sched_clock());
> +
> +	return 0;
> +}
> +
> +arch_initcall(mips_timecounter_init);

mips_timecounter_init() must called before ftrace_init(), so can not put
it into *_initcall sections(which will be called in rest_init()).

Regards,
	Wu Zhangjin

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/