mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: hui liu <quic_huliu@quicinc.com>
To: Heikki Krogerus <heikki.krogerus@linux.intel.com>
Cc: Bryan O'Donoghue <bryan.odonoghue@linaro.org>,
	Guenter Roeck <linux@roeck-us.net>,
	Andy Gross <agross@kernel.org>,
	Bjorn Andersson <andersson@kernel.org>,
	Konrad Dybcio <konrad.dybcio@linaro.org>,
	"Greg Kroah-Hartman" <gregkh@linuxfoundation.org>,
	<linux-arm-msm@vger.kernel.org>, <linux-usb@vger.kernel.org>,
	<linux-kernel@vger.kernel.org>, <quic_fenglinw@quicinc.com>,
	<subbaram@quicinc.com>
Subject: Re: [PATCH v2] usb: typec: qcom: check regulator enable status before disabling it
Date: Fri, 25 Aug 2023 18:03:02 +0800	[thread overview]
Message-ID: <0b6b864c-3e41-4b41-ac35-9ab8edb156a0@quicinc.com> (raw)
In-Reply-To: <ZOdt+dz7XXljFJcK@kuha.fi.intel.com>

Hi Heikki,

I will let Bryan to comment, I am using the driver to support the pdphy 
in SMB2352 and there is no external regulator required, so I am just 
using a dummy regulator device and I saw this unbalanced regulator 
disabling warnings, so my intention for this change is just fixing the 
warning message. However, I am fine with whatever suggestion you have, 
since the logic is straightforward, and I can make the changes once you 
have the agreement.

Thanks,
Hui

On 8/24/2023 10:49 PM, Heikki Krogerus wrote:
> On Thu, Aug 24, 2023 at 05:12:14PM +0300, Heikki Krogerus wrote:
>> On Thu, Aug 24, 2023 at 10:32:03AM +0800, Hui Liu via B4 Relay wrote:
>>> From: Hui Liu <quic_huliu@quicinc.com>
>>>
>>> Check regulator enable status before disabling it to avoid
>>> unbalanced regulator disable warnings.
>>>
>>> Reviewed-by: Guenter Roeck <linux@roeck-us.net>
>>> Fixes: a4422ff22142 ("usb: typec: qcom: Add Qualcomm PMIC Type-C driver")
>>> Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
>>> Signed-off-by: Hui Liu <quic_huliu@quicinc.com>
>>> ---
>>> Changes in v2:
>>> - Add Fixes tag
>>> - Link to v1: https://lore.kernel.org/r/20230823-qcom-tcpc-v1-1-fa81a09ca056@quicinc.com
>>> ---
>>>   drivers/usb/typec/tcpm/qcom/qcom_pmic_typec_pdphy.c | 3 ++-
>>>   1 file changed, 2 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/drivers/usb/typec/tcpm/qcom/qcom_pmic_typec_pdphy.c b/drivers/usb/typec/tcpm/qcom/qcom_pmic_typec_pdphy.c
>>> index bb0b8479d80f..ca616b17b5b6 100644
>>> --- a/drivers/usb/typec/tcpm/qcom/qcom_pmic_typec_pdphy.c
>>> +++ b/drivers/usb/typec/tcpm/qcom/qcom_pmic_typec_pdphy.c
>>> @@ -422,7 +422,8 @@ static int qcom_pmic_typec_pdphy_disable(struct pmic_typec_pdphy *pmic_typec_pdp
>>>   	ret = regmap_write(pmic_typec_pdphy->regmap,
>>>   			   pmic_typec_pdphy->base + USB_PDPHY_EN_CONTROL_REG, 0);
>>>   
>>> -	regulator_disable(pmic_typec_pdphy->vdd_pdphy);
>>> +	if (regulator_is_enabled(pmic_typec_pdphy->vdd_pdphy))
>>> +		regulator_disable(pmic_typec_pdphy->vdd_pdphy);
>>
>> Would it be an option to just enable the regulator in
>> qcom_pmic_typec_pdphy_start() and disable it in
>> qcom_pmic_typec_pdphy_stop()?
>>
>> Now the whole thing looks weird. That regulator is in practice
>> only disabled and then enabled in one and the same place -
>> pmic_typec_pdphy_reset(). It's not touched anywhere else. That makes
>> the above condition confusing to me. I may be missing something.
>>
>> At least more explanation is needed.
> 
> I took a closer look at these drivers, and I think I understand the
> code path now. This driver is made with an assumption that the
> regulator is "on" when the driver is probed, but in your case it's
> actually "off".
> 
> So there is something wrong here, but I don't know where the root
> cause is. If the regulator is really "on" when this driver is probed,
> then there should be another user for it somewhere (no?). In that case
> the driver can't just switch off the regulator like it does now - this
> part I think really has to be fixed (or explained).
> 
> The problem with your fix is that it will leave the regulator always
> on when the driver is removed, which it really can't do, not at least
> if the regulator was off by default.
> 
> I would propose this:
> 
> diff --git a/drivers/usb/typec/tcpm/qcom/qcom_pmic_typec_pdphy.c b/drivers/usb/typec/tcpm/qcom/qcom_pmic_typec_pdphy.c
> index bb0b8479d80f..bbe40634e821 100644
> --- a/drivers/usb/typec/tcpm/qcom/qcom_pmic_typec_pdphy.c
> +++ b/drivers/usb/typec/tcpm/qcom/qcom_pmic_typec_pdphy.c
> @@ -449,6 +449,10 @@ int qcom_pmic_typec_pdphy_start(struct pmic_typec_pdphy *pmic_typec_pdphy,
>   
>          pmic_typec_pdphy->tcpm_port = tcpm_port;
>   
> +       ret = regulator_enable(pmic_typec_pdphy->vdd_pdphy);
> +       if (ret)
> +               return ret;
> +
>          ret = pmic_typec_pdphy_reset(pmic_typec_pdphy);
>          if (ret)
>                  return ret;
> @@ -467,6 +471,7 @@ void qcom_pmic_typec_pdphy_stop(struct pmic_typec_pdphy *pmic_typec_pdphy)
>                  disable_irq(pmic_typec_pdphy->irq_data[i].irq);
>   
>          qcom_pmic_typec_pdphy_reset_on(pmic_typec_pdphy);
> +       regulator_disable(pmic_typec_pdphy->vdd_pdphy);
>   }
>   
>   struct pmic_typec_pdphy *qcom_pmic_typec_pdphy_alloc(struct device *dev)
> 
> 
> The problem with it is that the regulator is not going to be disabled
> if there really is another user for it when the component is expected
> to be reset. But as said above, if there really is an other user, then
> this driver simply can't just turn off the regulator.
> 
> thanks,
> 

  reply	other threads:[~2023-08-25 10:04 UTC|newest]

Thread overview: 9+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-08-24  2:32 Hui Liu via B4 Relay
2023-08-24  8:36 ` Bryan O'Donoghue
2023-08-24 14:12 ` Heikki Krogerus
2023-08-24 14:37   ` Bryan O'Donoghue
2023-08-24 14:49   ` Heikki Krogerus
2023-08-25 10:03     ` hui liu [this message]
2023-08-25 10:11       ` Bryan O'Donoghue
2023-08-28  5:51         ` hui liu
2023-08-28  9:43           ` Bryan O'Donoghue

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=0b6b864c-3e41-4b41-ac35-9ab8edb156a0@quicinc.com \
    --to=quic_huliu@quicinc.com \
    --cc=agross@kernel.org \
    --cc=andersson@kernel.org \
    --cc=bryan.odonoghue@linaro.org \
    --cc=gregkh@linuxfoundation.org \
    --cc=heikki.krogerus@linux.intel.com \
    --cc=konrad.dybcio@linaro.org \
    --cc=linux-arm-msm@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-usb@vger.kernel.org \
    --cc=linux@roeck-us.net \
    --cc=quic_fenglinw@quicinc.com \
    --cc=subbaram@quicinc.com \
    /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

all inboxes | Powered by JetHome®