Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1332366 > unrolled thread

[PATCH v42 5/6] clk: clk_put WARNs if user has not disabled clk

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.

Contents

  [PATCH v42 5/6] clk: clk_put WARNs if user has not disabled clk Michael Turquette <mturquette@baylibre.com> - 2016-02-11 22:20 +0100
    Re: [PATCH v42 5/6] clk: clk_put WARNs if user has not disabled clk Stephen Boyd <sboyd@codeaurora.org> - 2016-02-13 02:20 +0100
      Re: [PATCH v42 5/6] clk: clk_put WARNs if user has not disabled clk Geert Uytterhoeven <geert@linux-m68k.org> - 2016-02-14 16:00 +0100

#1332366 — [PATCH v42 5/6] clk: clk_put WARNs if user has not disabled clk

>From the clk_put kerneldoc in include/linux/clk.h:

"""
Note: drivers must ensure that all clk_enable calls made on this clock
source are balanced by clk_disable calls prior to calling this function.
"""

The common clock framework implementation of the clk.h api has per-user
reference counts for calls to clk_prepare and clk_disable. As such it
can enforce the requirement to properly call clk_disable and
clk_unprepare before calling clk_put.

Because this requirement is probably violated in many places, this patch
starts with a simple warning. Once offending code has been fixed this
check could additionally release the reference counts automatically.

Signed-off-by: Michael Turquette <mturquette@baylibre.com>
---
Changed in v2ish:
  * s/WARN_ON/WARN_ON_ONCE/ to reduce noise for pm-clocks

 drivers/clk/clk.c | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/drivers/clk/clk.c b/drivers/clk/clk.c
index 01183e3..d81f970 100644
--- a/drivers/clk/clk.c
+++ b/drivers/clk/clk.c
@@ -2824,6 +2824,14 @@ void __clk_put(struct clk *clk)
 	    clk->max_rate < clk->core->req_rate)
 		clk_core_set_rate_nolock(clk->core, clk->core->req_rate);
 
+	/*
+	 * before calling clk_put, all calls to clk_prepare and clk_enable from
+	 * a given user must be balanced with calls to clk_disable and
+	 * clk_unprepare by that same user
+	 */
+	WARN_ON_ONCE(clk->prepare_count);
+	WARN_ON_ONCE(clk->enable_count);
+
 	owner = clk->core->owner;
 	kref_put(&clk->core->ref, __clk_release);
 
-- 
2.1.4

[toc] | [next] | [standalone]

#1333242

On 02/11, Michael Turquette wrote:
> >From the clk_put kerneldoc in include/linux/clk.h:
> 
> """
> Note: drivers must ensure that all clk_enable calls made on this clock
> source are balanced by clk_disable calls prior to calling this function.
> """
> 
> The common clock framework implementation of the clk.h api has per-user
> reference counts for calls to clk_prepare and clk_disable. As such it
> can enforce the requirement to properly call clk_disable and
> clk_unprepare before calling clk_put.
> 
> Because this requirement is probably violated in many places, this patch
> starts with a simple warning. Once offending code has been fixed this
> check could additionally release the reference counts automatically.

Do we have any fixes for pm code in the works? I'm worried we're
going to be giving a warning and nobody will fix them or has a
plan to fix them.

> 
> Signed-off-by: Michael Turquette <mturquette@baylibre.com>
> ---

Reviewed-by: Stephen Boyd <sboyd@codeaurora.org>

-- 
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
a Linux Foundation Collaborative Project

[toc] | [prev] | [next] | [standalone]

#1333506

Hi Stephen, Mike,

On Sat, Feb 13, 2016 at 2:18 AM, Stephen Boyd <sboyd@codeaurora.org> wrote:
> On 02/11, Michael Turquette wrote:
>> >From the clk_put kerneldoc in include/linux/clk.h:
>>
>> """
>> Note: drivers must ensure that all clk_enable calls made on this clock
>> source are balanced by clk_disable calls prior to calling this function.
>> """
>>
>> The common clock framework implementation of the clk.h api has per-user
>> reference counts for calls to clk_prepare and clk_disable. As such it
>> can enforce the requirement to properly call clk_disable and
>> clk_unprepare before calling clk_put.
>>
>> Because this requirement is probably violated in many places, this patch
>> starts with a simple warning. Once offending code has been fixed this
>> check could additionally release the reference counts automatically.
>
> Do we have any fixes for pm code in the works? I'm worried we're
> going to be giving a warning and nobody will fix them or has a
> plan to fix them.
drivers/base/power/clock_ops.c
AFAIK not.

I've been running with the above patch for several months, and I had
to remove the two clk_put() calls in {en,dis}able_clock() in
drivers/base/power/clock_ops.c to get rid of the warnings when using the
legacy clock domain.

Fixing drivers/base/power/clock_ops.c is non-trivial though, as you need a
place to store the clk's reference obtained in enable_clock(), for later use in
disable_clock().

However, the plan is to make CONFIG_PM=y mandatory for Renesas ARM
SoCs with clock domains, which makes us no longer users of the legacy clock
domain.

Legacy SH and Davinci/Keystone/OMAP1 users may care, though...

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web