[PATCH] i2c: rcar: make sure clocks are on when doing hw init

[PATCH] i2c: rcar: make sure clocks are on when doing hw init

[Wolfram Sang]

From: Wolfram Sang <wsa+renesas@sang-engineering.com>

Reported-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
---
 drivers/i2c/busses/i2c-rcar.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/i2c/busses/i2c-rcar.c b/drivers/i2c/busses/i2c-rcar.c
index 616433d387cdb2..ef26e933b1e9cb 100644
--- a/drivers/i2c/busses/i2c-rcar.c
+++ b/drivers/i2c/busses/i2c-rcar.c
@@ -612,7 +612,10 @@ static int rcar_i2c_probe(struct platform_device *pdev)
 	if (IS_ERR(priv->io))
 		return PTR_ERR(priv->io);
 
+	pm_runtime_enable(dev);
+	pm_runtime_get_sync(dev);
 	rcar_i2c_init(priv);
+	pm_runtime_put(dev);
 
 	irq = platform_get_irq(pdev, 0);
 	init_waitqueue_head(&priv->wait);
@@ -634,7 +637,6 @@ static int rcar_i2c_probe(struct platform_device *pdev)
 		return ret;
 	}
 
-	pm_runtime_enable(dev);
 	platform_set_drvdata(pdev, priv);
 
 	ret = i2c_add_numbered_adapter(adap);
-- 
2.1.4

Re: [PATCH] i2c: rcar: make sure clocks are on when doing hw init

[Geert Uytterhoeven]

Hi Wolfram,

On Thu, Oct 29, 2015 at 8:19 PM, Wolfram Sang <wsa@the-dreams.de> wrote:
> From: Wolfram Sang <wsa+renesas@sang-engineering.com>
>
> Reported-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
> Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>

Thanks for your patch!

I suspect this bug was also the cause of the spurious failures of the R-Car
Gen2 regulator quirk code I experienced sometimes during the last few
days.

> ---
>  drivers/i2c/busses/i2c-rcar.c | 4 +++-
>  1 file changed, 3 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/i2c/busses/i2c-rcar.c b/drivers/i2c/busses/i2c-rcar.c
> index 616433d387cdb2..ef26e933b1e9cb 100644
> --- a/drivers/i2c/busses/i2c-rcar.c
> +++ b/drivers/i2c/busses/i2c-rcar.c
> @@ -612,7 +612,10 @@ static int rcar_i2c_probe(struct platform_device *pdev)
>         if (IS_ERR(priv->io))
>                 return PTR_ERR(priv->io);
>
> +       pm_runtime_enable(dev);
> +       pm_runtime_get_sync(dev);
>         rcar_i2c_init(priv);
> +       pm_runtime_put(dev);
>
>         irq = platform_get_irq(pdev, 0);
>         init_waitqueue_head(&priv->wait);
> @@ -634,7 +637,6 @@ static int rcar_i2c_probe(struct platform_device *pdev)
>                 return ret;

pm_runtime_disable() in error path?

>         }

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

Re: [PATCH] i2c: rcar: make sure clocks are on when doing hw init

[Kuninori Morimoto]

Hi, Wolfram

> From: Wolfram Sang <wsa+renesas@sang-engineering.com>
> 
> Reported-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
> Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
> ---
>  drivers/i2c/busses/i2c-rcar.c | 4 +++-
>  1 file changed, 3 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/i2c/busses/i2c-rcar.c b/drivers/i2c/busses/i2c-rcar.c
> index 616433d387cdb2..ef26e933b1e9cb 100644
> --- a/drivers/i2c/busses/i2c-rcar.c
> +++ b/drivers/i2c/busses/i2c-rcar.c
> @@ -612,7 +612,10 @@ static int rcar_i2c_probe(struct platform_device *pdev)
>  	if (IS_ERR(priv->io))
>  		return PTR_ERR(priv->io);
>  
> +	pm_runtime_enable(dev);
> +	pm_runtime_get_sync(dev);
>  	rcar_i2c_init(priv);
> +	pm_runtime_put(dev);

Please correct me if I'm misunderstanding, but in my experience,
above rcar_i2c_init() result can be removed if SoC has power-off feature,
and if IP doesn't keep register value power-off case ?
This case register doesn't keep myfunc_A's setting,
because it calls pm_runtime_put()

     pm_runtime_get_sync()
     myfunc_A()
     pm_runtime_put()
     ...
     pm_runtime_get_sync()
     myfunc_B()
     pm_runtime_put()

It should be this ?

     pm_runtime_get_sync()
     myfunc_A()
     myfunc_B()
     pm_runtime_put()

Re: [PATCH] i2c: rcar: make sure clocks are on when doing hw init

[Wolfram Sang]

[-- Attachment #1: Type: text/plain, Size: 1145 bytes --]

Morimoto-san,

> > +	pm_runtime_enable(dev);
> > +	pm_runtime_get_sync(dev);
> >  	rcar_i2c_init(priv);
> > +	pm_runtime_put(dev);
> 
> Please correct me if I'm misunderstanding, but in my experience,
> above rcar_i2c_init() result can be removed if SoC has power-off feature,
> and if IP doesn't keep register value power-off case ?
> This case register doesn't keep myfunc_A's setting,
> because it calls pm_runtime_put()
> 
>      pm_runtime_get_sync()
>      myfunc_A()
>      pm_runtime_put()
>      ...
>      pm_runtime_get_sync()
>      myfunc_B()
>      pm_runtime_put()
> 
> It should be this ?
> 
>      pm_runtime_get_sync()
>      myfunc_A()
>      myfunc_B()
>      pm_runtime_put()

This is for power-off case, right? Wouldn't it be clearer then to add a
resume function to pm_ops of the I2C driver which reinit the registers?
Doing it before every transfer might be simpler but is also a bit
implicit or subtle, I'd think.

And for completeness: The above is only for the power-off case. In
module-standby (MSTP bits), register values are always retained.
Correct?

Regards,

   Wolfram


[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]

Re: [PATCH] i2c: rcar: make sure clocks are on when doing hw init

[Kuninori Morimoto]

Hi Wolfram

> >      pm_runtime_get_sync()
> >      myfunc_A()
> >      pm_runtime_put()
> >      ...
> >      pm_runtime_get_sync()
> >      myfunc_B()
> >      pm_runtime_put()
> > 
> > It should be this ?
> > 
> >      pm_runtime_get_sync()
> >      myfunc_A()
> >      myfunc_B()
> >      pm_runtime_put()
> 
> This is for power-off case, right? Wouldn't it be clearer then to add a
> resume function to pm_ops of the I2C driver which reinit the registers?
> Doing it before every transfer might be simpler but is also a bit
> implicit or subtle, I'd think.
> 
> And for completeness: The above is only for the power-off case. In
> module-standby (MSTP bits), register values are always retained.
> Correct?

I'm not sure detail of pm_runtime, but it depends on
power-domain support ?
On SH-Mobile case, it automatically power-off:ed if all driver
user was gone. This feature is not yet supported on R-Car case I think.
So, both case are not problem at this point, but 1st case will be
problem if power-domain was suppored.
(If my memory was correct...)

Re: [PATCH] i2c: rcar: make sure clocks are on when doing hw init

[Wolfram Sang]

[-- Attachment #1: Type: text/plain, Size: 129 bytes --]


> On SH-Mobile case, it automatically power-off:ed if all driver
> user was gone.

It does that without calling some callback?


[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]

Re: [PATCH] i2c: rcar: make sure clocks are on when doing hw init

[Kuninori Morimoto]

Hi Wolfram

> > On SH-Mobile case, it automatically power-off:ed if all driver
> > user was gone.
> 
> It does that without calling some callback?

Do you mean you can call init() in callback ?
I don't care detail of design. But my concern was that
register value might be disappear during many get/put if
power-domain was implemented