From: Christophe Leroy <christophe.leroy@c-s.fr>
To: Thomas Gleixner <tglx@linutronix.de>,
LKML <linux-kernel@vger.kernel.org>
Cc: x86@kernel.org, John Stultz <john.stultz@linaro.org>,
Vincenzo Frascino <vincenzo.frascino@arm.com>,
Andy Lutomirski <luto@kernel.org>,
Paolo Bonzini <pbonzini@redhat.com>,
Boris Ostrovsky <boris.ostrovsky@oracle.com>,
Juergen Gross <jgross@suse.com>,
Haiyang Zhang <haiyangz@microsoft.com>,
Sasha Levin <sashal@kernel.org>,
Ralf Baechle <ralf@linux-mips.org>,
Paul Burton <paulburton@kernel.org>,
James Hogan <jhogan@kernel.org>,
Russell King <linux@armlinux.org.uk>,
Catalin Marinas <catalin.marinas@arm.com>,
Will Deacon <will@kernel.org>,
Mark Rutland <mark.rutland@arm.com>,
Marc Zyngier <maz@kernel.org>
Subject: Re: [patch 09/15] clocksource: Add common vdso clock mode storage
Date: Wed, 15 Jan 2020 06:50:42 +0100 [thread overview]
Message-ID: <f6c5ce65-be49-add3-5959-85fa9cdc75a2@c-s.fr> (raw)
In-Reply-To: <20200114185947.500141436@linutronix.de>
Le 14/01/2020 à 19:52, Thomas Gleixner a écrit :
> All architectures which use the generic VDSO code have their own storage
> for the VDSO clock mode. That's pointless and just requires duplicate code.
>
> Provide generic storage for it. The new Kconfig symbol is intermediate and
> will be removed once all architectures are converted over.
>
> Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
> ---
> include/linux/clocksource.h | 12 +++++++++++-
> kernel/time/clocksource.c | 9 +++++++++
> kernel/time/vsyscall.c | 7 +++++++
> lib/vdso/Kconfig | 3 +++
> lib/vdso/gettimeofday.c | 13 +++++++++++--
> 5 files changed, 41 insertions(+), 3 deletions(-)
>
> --- a/include/linux/clocksource.h
> +++ b/include/linux/clocksource.h
> @@ -23,10 +23,19 @@
> struct clocksource;
> struct module;
>
> -#ifdef CONFIG_ARCH_CLOCKSOURCE_DATA
> +#if defined(CONFIG_ARCH_CLOCKSOURCE_DATA) || \
> + defined(CONFIG_GENERIC_VDSO_CLOCK_MODE)
> #include <asm/clocksource.h>
> #endif
>
> +enum vdso_clock_mode {
> + VDSO_CLOCKMODE_NONE,
> +#ifdef CONFIG_GENERIC_VDSO_CLOCK_MODE
> + VDSO_ARCH_CLOCKMODES,
> +#endif
> + VDSO_CLOCKMODE_MAX,
> +};
> +
> /**
> * struct clocksource - hardware abstraction for a free running counter
> * Provides mostly state-free accessors to the underlying hardware.
> @@ -97,6 +106,7 @@ struct clocksource {
> const char *name;
> struct list_head list;
> int rating;
> + enum vdso_clock_mode vdso_clock_mode;
> unsigned long flags;
>
> int (*enable)(struct clocksource *cs);
> --- a/kernel/time/clocksource.c
> +++ b/kernel/time/clocksource.c
> @@ -921,6 +921,15 @@ int __clocksource_register_scale(struct
>
> clocksource_arch_init(cs);
>
> +#ifdef CONFIG_GENERIC_VDSO_CLOCK_MODE
> + if (cs->vdso_clock_mode < 0 ||
> + cs->vdso_clock_mode >= VDSO_CLOCKMODE_MAX) {
> + pr_warn("clocksource %s registered with invalid VDSO mode %d. Disabling VDSO support.\n",
> + cs->name, cs->vdso_clock_mode);
> + cs->vdso_clock_mode = VDSO_CLOCKMODE_NONE;
> + }
> +#endif
> +
> /* Initialize mult/shift and max_idle_ns */
> __clocksource_update_freq_scale(cs, scale, freq);
>
> --- a/kernel/time/vsyscall.c
> +++ b/kernel/time/vsyscall.c
> @@ -72,12 +72,19 @@ void update_vsyscall(struct timekeeper *
> struct vdso_data *vdata = __arch_get_k_vdso_data();
> struct vdso_timestamp *vdso_ts;
> u64 nsec;
> + s32 mode;
gcc will claim 'mode' is unused when CONFIG_GENERIC_VDSO_CLOCK_MODE is
not defined.
>
> /* copy vsyscall data */
> vdso_write_begin(vdata);
>
> +#ifdef CONFIG_GENERIC_VDSO_CLOCK_MODE
> + mode = tk->tkr_mono.clock->vdso_clock_mode;
> + vdata[CS_HRES_COARSE].clock_mode = mode;
> + vdata[CS_RAW].clock_mode = mode;
> +#else
> vdata[CS_HRES_COARSE].clock_mode = __arch_get_clock_mode(tk);
> vdata[CS_RAW].clock_mode = __arch_get_clock_mode(tk);
> +#endif
Can we do :
#ifdef CONFIG_GENERIC_VDSO_CLOCK_MODE
mode = tk->tkr_mono.clock->vdso_clock_mode;
#else
mode = __arch_get_clock_mode(tk);
#endif
vdata[CS_HRES_COARSE].clock_mode = mode;
vdata[CS_RAW].clock_mode = mode;
Christophe
>
> /* CLOCK_REALTIME also required for time() */
> vdso_ts = &vdata[CS_HRES_COARSE].basetime[CLOCK_REALTIME];
> --- a/lib/vdso/Kconfig
> +++ b/lib/vdso/Kconfig
> @@ -30,4 +30,7 @@ config GENERIC_VDSO_TIME_NS
> Selected by architectures which support time namespaces in the
> VDSO
>
> +config GENERIC_VDSO_CLOCK_MODE
> + bool
> +
> endif
> --- a/lib/vdso/gettimeofday.c
> +++ b/lib/vdso/gettimeofday.c
> @@ -7,6 +7,7 @@
> #include <linux/time.h>
> #include <linux/kernel.h>
> #include <linux/hrtimer_defs.h>
> +#include <linux/clocksource.h>
> #include <vdso/datapage.h>
> #include <vdso/helpers.h>
>
> @@ -64,10 +65,14 @@ static int do_hres_timens(const struct v
>
> do {
> seq = vdso_read_begin(vd);
> + if (IS_ENABLED(CONFIG_GENERIC_VDSO_CLOCK_MODE) &&
> + vd->clock_mode == VDSO_CLOCKMODE_NONE)
> + return -1;
> cycles = __arch_get_hw_counter(vd->clock_mode);
> ns = vdso_ts->nsec;
> last = vd->cycle_last;
> - if (unlikely((s64)cycles < 0))
> + if (!IS_ENABLED(CONFIG_GENERIC_VDSO_CLOCK_MODE) &&
> + unlikely((s64)cycles < 0))
> return -1;
>
> ns += vdso_calc_delta(cycles, last, vd->mask, vd->mult);
> @@ -132,10 +137,14 @@ static __always_inline int do_hres(const
> }
> smp_rmb();
>
> + if (IS_ENABLED(CONFIG_GENERIC_VDSO_CLOCK_MODE) &&
> + vd->clock_mode == VDSO_CLOCKMODE_NONE)
> + return -1;
> cycles = __arch_get_hw_counter(vd->clock_mode);
> ns = vdso_ts->nsec;
> last = vd->cycle_last;
> - if (unlikely((s64)cycles < 0))
> + if (!IS_ENABLED(CONFIG_GENERIC_VDSO_CLOCK_MODE) &&
> + unlikely((s64)cycles < 0))
> return -1;
>
> ns += vdso_calc_delta(cycles, last, vd->mask, vd->mult);
>
next prev parent reply other threads:[~2020-01-15 5:50 UTC|newest]
Thread overview: 25+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-01-14 18:52 [patch 00/15] lib/vdso: Bugfix and consolidation Thomas Gleixner
2020-01-14 18:52 ` [patch 01/15] lib/vdso: Make __arch_update_vdso_data() logic understandable Thomas Gleixner
2020-01-17 15:07 ` [tip: timers/urgent] " tip-bot2 for Thomas Gleixner
2020-01-14 18:52 ` [patch 02/15] lib/vdso: Update coarse timekeeper unconditionally Thomas Gleixner
2020-01-17 15:07 ` [tip: timers/urgent] " tip-bot2 for Thomas Gleixner
2020-01-14 18:52 ` [patch 03/15] ARM: vdso: Remove unused function Thomas Gleixner
2020-01-14 18:52 ` [patch 04/15] lib/vdso: Allow the high resolution parts to be compiled out Thomas Gleixner
2020-01-14 18:52 ` [patch 05/15] ARM: vdso: Compile high resolution parts conditionally Thomas Gleixner
2020-01-14 18:52 ` [patch 06/15] MIPS: " Thomas Gleixner
2020-01-14 18:52 ` [patch 07/15] clocksource: Cleanup struct clocksource and documentation Thomas Gleixner
2020-01-14 18:52 ` [patch 08/15] x86/vdso: Move VDSO clocksource state tracking to callback Thomas Gleixner
2020-01-15 5:56 ` Jürgen Groß
2020-01-20 1:42 ` Michael Kelley
2020-01-14 18:52 ` [patch 09/15] clocksource: Add common vdso clock mode storage Thomas Gleixner
2020-01-15 5:50 ` Christophe Leroy [this message]
2020-01-15 7:59 ` Thomas Gleixner
2020-01-14 18:52 ` [patch 10/15] x86/vdso: Use generic VDSO " Thomas Gleixner
2020-01-15 5:57 ` Jürgen Groß
2020-01-14 18:52 ` [patch 11/15] mips: vdso: " Thomas Gleixner
2020-01-14 18:52 ` [patch 12/15] ARM/arm64: vdso: Use common vdso " Thomas Gleixner
2020-01-14 18:52 ` [patch 13/15] lib/vdso: Cleanup clock mode storage leftovers Thomas Gleixner
2020-01-14 18:52 ` [patch 14/15] lib/vdso: Avoid highres update if clocksource is not VDSO capable Thomas Gleixner
2020-01-15 5:54 ` Christophe Leroy
2020-01-14 18:52 ` [patch 15/15] lib/vdso: Move VCLOCK_TIMENS to vdso_clock_modes Thomas Gleixner
2020-01-15 5:58 ` Christophe Leroy
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=f6c5ce65-be49-add3-5959-85fa9cdc75a2@c-s.fr \
--to=christophe.leroy@c-s.fr \
--cc=boris.ostrovsky@oracle.com \
--cc=catalin.marinas@arm.com \
--cc=haiyangz@microsoft.com \
--cc=jgross@suse.com \
--cc=jhogan@kernel.org \
--cc=john.stultz@linaro.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux@armlinux.org.uk \
--cc=luto@kernel.org \
--cc=mark.rutland@arm.com \
--cc=maz@kernel.org \
--cc=paulburton@kernel.org \
--cc=pbonzini@redhat.com \
--cc=ralf@linux-mips.org \
--cc=sashal@kernel.org \
--cc=tglx@linutronix.de \
--cc=vincenzo.frascino@arm.com \
--cc=will@kernel.org \
--cc=x86@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).