Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Message-ID: <9bf87ee0e1c2a97b0a1b98734e34e456f84ea3d7.camel@gmail.com>
Subject: Re: [PATCH 3/7] Input: ep93xx_keypad: Prepare clock before using it
From:   Alexander Sverdlin <alexander.sverdlin@gmail.com>
To:     Dmitry Torokhov <dmitry.torokhov@gmail.com>
Cc:     Geert Uytterhoeven <geert@linux-m68k.org>,
        Nikita Shubin <nikita.shubin@maquefel.me>,
        linux-input@vger.kernel.org, linux-kernel@vger.kernel.org
Date:   Tue, 15 Jun 2021 09:46:51 +0200
In-Reply-To: <YMfQR8iM9be5Qg8+@google.com>
References: <20210613233041.128961-1-alexander.sverdlin@gmail.com>
         <20210613233041.128961-4-alexander.sverdlin@gmail.com>
         <YMfQR8iM9be5Qg8+@google.com>
Content-Type: text/plain; charset="UTF-8"
User-Agent: Evolution 3.40.2 
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Precedence: bulk

Hello Dmitry!

On Mon, 2021-06-14 at 14:55 -0700, Dmitry Torokhov wrote:
> > Use clk_prepare_enable()/clk_disable_unprepare() in preparation for switch
> > to Common Clock Framework.
> 
> Can this be merged standalone?

In principle, yes, but I thought it would be easier if the patches
would go via the same path as CCF conversion.

> > Signed-off-by: Alexander Sverdlin <alexander.sverdlin@gmail.com>
> > ---
> >   drivers/input/keyboard/ep93xx_keypad.c | 4 ++--
> >   1 file changed, 2 insertions(+), 2 deletions(-)
> > 
> > diff --git a/drivers/input/keyboard/ep93xx_keypad.c b/drivers/input/keyboard/ep93xx_keypad.c
> > index c8194333d612..e0e931e796fa 100644
> > --- a/drivers/input/keyboard/ep93xx_keypad.c
> > +++ b/drivers/input/keyboard/ep93xx_keypad.c
> > @@ -157,7 +157,7 @@ static int ep93xx_keypad_open(struct input_dev *pdev)
> >   
> >         if (!keypad->enabled) {
> >                 ep93xx_keypad_config(keypad);
> > -               clk_enable(keypad->clk);
> > +               clk_prepare_enable(keypad->clk);
> >                 keypad->enabled = true;
> >         }
> >   
> > @@ -169,7 +169,7 @@ static void ep93xx_keypad_close(struct input_dev *pdev)
> >         struct ep93xx_keypad *keypad = input_get_drvdata(pdev);
> >   
> >         if (keypad->enabled) {
> > -               clk_disable(keypad->clk);
> > +               clk_disable_unprepare(keypad->clk);
> >                 keypad->enabled = false;
> 
> While we are at it, I wonder about handling suspend/resume. I see that
> we disable the clock even if keyboard is configured as a wakeup source.
> Is it really capable of waking up the system when clock is off?

That what "EP93xx User’s Guide" says:

26.2.4 Low Power Mode
The key scanning block also supports a low power wake-up mode. In this mode, a key press
generates a wake up interrupt. The key scan interrupt should be masked. Because the wake
up interrupt is asynchronous, and depends on external keypad lines which may have a large
capacitance value, glitches may occur on the interrupt when transitioning to low power mode.
After transitioning, all clocks to the key scanning circuitry can be shut down.

-- 
Alexander Sverdlin.