public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
* [PATCH] arm64: kernel: Copy register_persistent_clock() to arm64 source subtree
@ 2014-11-07  0:21 Anatol Pomozov
  2014-11-07  0:25 ` Anatol Pomozov
  0 siblings, 1 reply; 6+ messages in thread
From: Anatol Pomozov @ 2014-11-07  0:21 UTC (permalink / raw)
  To: marc.zyngier, mark.rutland
  Cc: lorenzo.pieralisi, linux-kernel, Anatol Pomozov

This allows to port code that needs register_persistent_clock() to arm64

Signed-off-by: Anatol Pomozov <anatol.pomozov@gmail.com>
---
 arch/arm64/include/asm/time.h | 17 +++++++++++++++++
 arch/arm64/kernel/time.c      | 37 +++++++++++++++++++++++++++++++++++++
 2 files changed, 54 insertions(+)
 create mode 100644 arch/arm64/include/asm/time.h

diff --git a/arch/arm64/include/asm/time.h b/arch/arm64/include/asm/time.h
new file mode 100644
index 0000000..28015b8
--- /dev/null
+++ b/arch/arm64/include/asm/time.h
@@ -0,0 +1,17 @@
+/*
+ * Copyright (C) 2004 MontaVista Software, Inc.
+ * Copyright (C) 2014 The Chromium OS Authors.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+#ifndef __ASM_TIME_H
+#define __ASM_TIME_H
+
+struct timespec;
+typedef void (*clock_access_fn)(struct timespec *);
+extern int register_persistent_clock(clock_access_fn read_boot,
+		clock_access_fn read_persistent);
+
+#endif
diff --git a/arch/arm64/kernel/time.c b/arch/arm64/kernel/time.c
index 1a7125c..71a907e 100644
--- a/arch/arm64/kernel/time.c
+++ b/arch/arm64/kernel/time.c
@@ -40,6 +40,7 @@
 
 #include <asm/thread_info.h>
 #include <asm/stacktrace.h>
+#include <asm/time.h>
 
 #ifdef CONFIG_SMP
 unsigned long profile_pc(struct pt_regs *regs)
@@ -79,3 +80,39 @@ void __init time_init(void)
 	/* Calibrate the delay loop directly */
 	lpj_fine = arch_timer_rate / HZ;
 }
+
+static void dummy_clock_access(struct timespec *ts)
+{
+	ts->tv_sec = 0;
+	ts->tv_nsec = 0;
+}
+
+static clock_access_fn __read_persistent_clock = dummy_clock_access;
+static clock_access_fn __read_boot_clock = dummy_clock_access;
+
+void read_persistent_clock(struct timespec *ts)
+{
+	__read_persistent_clock(ts);
+}
+
+void read_boot_clock(struct timespec *ts)
+{
+	__read_boot_clock(ts);
+}
+
+int __init register_persistent_clock(clock_access_fn read_boot,
+				     clock_access_fn read_persistent)
+{
+	/* Only allow the clockaccess functions to be registered once */
+	if (__read_persistent_clock == dummy_clock_access &&
+	    __read_boot_clock == dummy_clock_access) {
+		if (read_boot)
+			__read_boot_clock = read_boot;
+		if (read_persistent)
+			__read_persistent_clock = read_persistent;
+
+		return 0;
+	}
+
+	return -EINVAL;
+}
-- 
2.1.0.rc2.206.gedb03e5


^ permalink raw reply related	[flat|nested] 6+ messages in thread

* Re: [PATCH] arm64: kernel: Copy register_persistent_clock() to arm64 source subtree
  2014-11-07  0:21 [PATCH] arm64: kernel: Copy register_persistent_clock() to arm64 source subtree Anatol Pomozov
@ 2014-11-07  0:25 ` Anatol Pomozov
  2014-11-07  4:16   ` Stephen Warren
  0 siblings, 1 reply; 6+ messages in thread
From: Anatol Pomozov @ 2014-11-07  0:25 UTC (permalink / raw)
  To: Marc Zyngier, mark.rutland, Stephen Warren
  Cc: Lorenzo Pieralisi, LKML, Anatol Pomozov

+Stephen

This patch is for tegra20_timer that uses register_persistent_clock().
I did not find any way to share the same arch code for arm/arm64.

Actually this register_persistent_clock() does not look arm specific
at all. Would it be better to move it somewhere outside of arch/?

On Thu, Nov 6, 2014 at 4:21 PM, Anatol Pomozov <anatol.pomozov@gmail.com> wrote:
> This allows to port code that needs register_persistent_clock() to arm64
>
> Signed-off-by: Anatol Pomozov <anatol.pomozov@gmail.com>
> ---
>  arch/arm64/include/asm/time.h | 17 +++++++++++++++++
>  arch/arm64/kernel/time.c      | 37 +++++++++++++++++++++++++++++++++++++
>  2 files changed, 54 insertions(+)
>  create mode 100644 arch/arm64/include/asm/time.h
>
> diff --git a/arch/arm64/include/asm/time.h b/arch/arm64/include/asm/time.h
> new file mode 100644
> index 0000000..28015b8
> --- /dev/null
> +++ b/arch/arm64/include/asm/time.h
> @@ -0,0 +1,17 @@
> +/*
> + * Copyright (C) 2004 MontaVista Software, Inc.
> + * Copyright (C) 2014 The Chromium OS Authors.
> + *
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License version 2 as
> + * published by the Free Software Foundation.
> + */
> +#ifndef __ASM_TIME_H
> +#define __ASM_TIME_H
> +
> +struct timespec;
> +typedef void (*clock_access_fn)(struct timespec *);
> +extern int register_persistent_clock(clock_access_fn read_boot,
> +               clock_access_fn read_persistent);
> +
> +#endif
> diff --git a/arch/arm64/kernel/time.c b/arch/arm64/kernel/time.c
> index 1a7125c..71a907e 100644
> --- a/arch/arm64/kernel/time.c
> +++ b/arch/arm64/kernel/time.c
> @@ -40,6 +40,7 @@
>
>  #include <asm/thread_info.h>
>  #include <asm/stacktrace.h>
> +#include <asm/time.h>
>
>  #ifdef CONFIG_SMP
>  unsigned long profile_pc(struct pt_regs *regs)
> @@ -79,3 +80,39 @@ void __init time_init(void)
>         /* Calibrate the delay loop directly */
>         lpj_fine = arch_timer_rate / HZ;
>  }
> +
> +static void dummy_clock_access(struct timespec *ts)
> +{
> +       ts->tv_sec = 0;
> +       ts->tv_nsec = 0;
> +}
> +
> +static clock_access_fn __read_persistent_clock = dummy_clock_access;
> +static clock_access_fn __read_boot_clock = dummy_clock_access;
> +
> +void read_persistent_clock(struct timespec *ts)
> +{
> +       __read_persistent_clock(ts);
> +}
> +
> +void read_boot_clock(struct timespec *ts)
> +{
> +       __read_boot_clock(ts);
> +}
> +
> +int __init register_persistent_clock(clock_access_fn read_boot,
> +                                    clock_access_fn read_persistent)
> +{
> +       /* Only allow the clockaccess functions to be registered once */
> +       if (__read_persistent_clock == dummy_clock_access &&
> +           __read_boot_clock == dummy_clock_access) {
> +               if (read_boot)
> +                       __read_boot_clock = read_boot;
> +               if (read_persistent)
> +                       __read_persistent_clock = read_persistent;
> +
> +               return 0;
> +       }
> +
> +       return -EINVAL;
> +}
> --
> 2.1.0.rc2.206.gedb03e5
>

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [PATCH] arm64: kernel: Copy register_persistent_clock() to arm64 source subtree
  2014-11-07  0:25 ` Anatol Pomozov
@ 2014-11-07  4:16   ` Stephen Warren
  2014-11-07 10:40     ` Mark Rutland
  0 siblings, 1 reply; 6+ messages in thread
From: Stephen Warren @ 2014-11-07  4:16 UTC (permalink / raw)
  To: Anatol Pomozov
  Cc: Marc Zyngier, mark.rutland, Lorenzo Pieralisi, LKML,
	linux-tegra@vger.kernel.org

On 11/06/2014 05:25 PM, Anatol Pomozov wrote:
> +Stephen
> 
> This patch is for tegra20_timer that uses register_persistent_clock().
> I did not find any way to share the same arch code for arm/arm64.
> 
> Actually this register_persistent_clock() does not look arm specific
> at all. Would it be better to move it somewhere outside of arch/?

No CC to linux-tegra@ or the other Tegra maintainers?

Yes, I think it'd be best not to have arch-specific APIs, or cut/paste
the same code into multiple places.

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [PATCH] arm64: kernel: Copy register_persistent_clock() to arm64 source subtree
  2014-11-07  4:16   ` Stephen Warren
@ 2014-11-07 10:40     ` Mark Rutland
  2014-11-07 16:24       ` Anatol Pomozov
  0 siblings, 1 reply; 6+ messages in thread
From: Mark Rutland @ 2014-11-07 10:40 UTC (permalink / raw)
  To: Stephen Warren
  Cc: Anatol Pomozov, Marc Zyngier, Lorenzo Pieralisi, LKML,
	linux-tegra@vger.kernel.org

On Fri, Nov 07, 2014 at 04:16:03AM +0000, Stephen Warren wrote:
> On 11/06/2014 05:25 PM, Anatol Pomozov wrote:
> > +Stephen
> > 
> > This patch is for tegra20_timer that uses register_persistent_clock().
> > I did not find any way to share the same arch code for arm/arm64.
> > 
> > Actually this register_persistent_clock() does not look arm specific
> > at all. Would it be better to move it somewhere outside of arch/?
> 
> No CC to linux-tegra@ or the other Tegra maintainers?
> 
> Yes, I think it'd be best not to have arch-specific APIs, or cut/paste
> the same code into multiple places.

Agreed. This looks in no way architecture specific, and having this in
common code would be preferable to copying.

Mark.

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [PATCH] arm64: kernel: Copy register_persistent_clock() to arm64 source subtree
  2014-11-07 10:40     ` Mark Rutland
@ 2014-11-07 16:24       ` Anatol Pomozov
  2014-11-07 16:38         ` Mark Rutland
  0 siblings, 1 reply; 6+ messages in thread
From: Anatol Pomozov @ 2014-11-07 16:24 UTC (permalink / raw)
  To: Mark Rutland
  Cc: Stephen Warren, Marc Zyngier, Lorenzo Pieralisi, LKML,
	linux-tegra@vger.kernel.org

Hi

On Fri, Nov 7, 2014 at 2:40 AM, Mark Rutland <mark.rutland@arm.com> wrote:
> On Fri, Nov 07, 2014 at 04:16:03AM +0000, Stephen Warren wrote:
>> On 11/06/2014 05:25 PM, Anatol Pomozov wrote:
>> > +Stephen
>> >
>> > This patch is for tegra20_timer that uses register_persistent_clock().
>> > I did not find any way to share the same arch code for arm/arm64.
>> >
>> > Actually this register_persistent_clock() does not look arm specific
>> > at all. Would it be better to move it somewhere outside of arch/?
>>
>> No CC to linux-tegra@ or the other Tegra maintainers?
>>
>> Yes, I think it'd be best not to have arch-specific APIs, or cut/paste
>> the same code into multiple places.
>
> Agreed. This looks in no way architecture specific, and having this in
> common code would be preferable to copying.

Where the code common for arm and arm64 should go?
drivers/???/arm_timekeeping.c ?

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [PATCH] arm64: kernel: Copy register_persistent_clock() to arm64 source subtree
  2014-11-07 16:24       ` Anatol Pomozov
@ 2014-11-07 16:38         ` Mark Rutland
  0 siblings, 0 replies; 6+ messages in thread
From: Mark Rutland @ 2014-11-07 16:38 UTC (permalink / raw)
  To: Anatol Pomozov
  Cc: Stephen Warren, Marc Zyngier, Lorenzo Pieralisi, LKML,
	linux-tegra@vger.kernel.org

On Fri, Nov 07, 2014 at 04:24:19PM +0000, Anatol Pomozov wrote:
> Hi
> 
> On Fri, Nov 7, 2014 at 2:40 AM, Mark Rutland <mark.rutland@arm.com> wrote:
> > On Fri, Nov 07, 2014 at 04:16:03AM +0000, Stephen Warren wrote:
> >> On 11/06/2014 05:25 PM, Anatol Pomozov wrote:
> >> > +Stephen
> >> >
> >> > This patch is for tegra20_timer that uses register_persistent_clock().
> >> > I did not find any way to share the same arch code for arm/arm64.
> >> >
> >> > Actually this register_persistent_clock() does not look arm specific
> >> > at all. Would it be better to move it somewhere outside of arch/?
> >>
> >> No CC to linux-tegra@ or the other Tegra maintainers?
> >>
> >> Yes, I think it'd be best not to have arch-specific APIs, or cut/paste
> >> the same code into multiple places.
> >
> > Agreed. This looks in no way architecture specific, and having this in
> > common code would be preferable to copying.
> 
> Where the code common for arm and arm64 should go?
> drivers/???/arm_timekeeping.c ?

I'd argue that this is in no way specific to ARM, and the current weak
functions live in kernel/time/timekeeping.c, so this should too.

Thanks,
Mark.

^ permalink raw reply	[flat|nested] 6+ messages in thread

end of thread, other threads:[~2014-11-07 16:38 UTC | newest]

Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2014-11-07  0:21 [PATCH] arm64: kernel: Copy register_persistent_clock() to arm64 source subtree Anatol Pomozov
2014-11-07  0:25 ` Anatol Pomozov
2014-11-07  4:16   ` Stephen Warren
2014-11-07 10:40     ` Mark Rutland
2014-11-07 16:24       ` Anatol Pomozov
2014-11-07 16:38         ` Mark Rutland

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox