Re: [PATCH v9] mmc: sdhci-of-dwcmshc: Add runtime PM operations

[Adrian Hunter <adrian.hunter@intel.com>]

On 18/08/23 12:00, Ulf Hansson wrote:
> On Thu, 17 Aug 2023 at 18:22, Liming Sun <limings@nvidia.com> wrote:
>>
>> This commit implements the runtime PM operations to disable eMMC
>> card clock when idle.
>>
>> Reviewed-by: David Thompson <davthompson@nvidia.com>
>> Signed-off-by: Liming Sun <limings@nvidia.com>
>> ---
>> v8->v9:
>>     - Address Adrian's comment to do the pm_runtime_put() in
>>       dwcmshc_resume() instead; Error path changes not included yet.
>> v7->v8:
>>     - Address Ulf's comment (option-1);
>>     - Updates for Adrian's comment to remove the force_suspend/resume
>>       in dwcmshc_resume()/dwcmshc_suspend(); Add comments for
>>       dwcmshc_resume()/dwcmshc_suspend();
>> v6->v7:
>>     - Address Ulf's comment;
>> v5->v6:
>>     - Address Adrian's more comments and add coordination between
>>       runtime PM and system PM;
>> v4->v5:
>>     - Address Adrian's comment to move the pm_enable to the end to
>>       avoid race;
>> v3->v4:
>>     - Fix compiling reported by 'kernel test robot';
>> v2->v3:
>>     - Revise the commit message;
>> v1->v2:
>>     Updates for comments from Ulf:
>>     - Make the runtime PM logic generic for sdhci-of-dwcmshc;
>> v1: Initial version.
>> ---
>>  drivers/mmc/host/sdhci-of-dwcmshc.c | 76 +++++++++++++++++++++++++++--
>>  1 file changed, 72 insertions(+), 4 deletions(-)
>>
>> diff --git a/drivers/mmc/host/sdhci-of-dwcmshc.c b/drivers/mmc/host/sdhci-of-dwcmshc.c
>> index e68cd87998c8..3b40f55ce2a4 100644
>> --- a/drivers/mmc/host/sdhci-of-dwcmshc.c
>> +++ b/drivers/mmc/host/sdhci-of-dwcmshc.c
>> @@ -15,6 +15,7 @@
>>  #include <linux/module.h>
>>  #include <linux/of.h>
>>  #include <linux/of_device.h>
>> +#include <linux/pm_runtime.h>
>>  #include <linux/reset.h>
>>  #include <linux/sizes.h>
>>
>> @@ -548,9 +549,13 @@ static int dwcmshc_probe(struct platform_device *pdev)
>>
>>         host->mmc->caps |= MMC_CAP_WAIT_WHILE_BUSY;
>>
>> +       pm_runtime_get_noresume(dev);
>> +       pm_runtime_set_active(dev);
>> +       pm_runtime_enable(dev);
>> +
>>         err = sdhci_setup_host(host);
>>         if (err)
>> -               goto err_clk;
>> +               goto err_rpm;
>>
>>         if (rk_priv)
>>                 dwcmshc_rk35xx_postinit(host, priv);
>> @@ -559,10 +564,15 @@ static int dwcmshc_probe(struct platform_device *pdev)
>>         if (err)
>>                 goto err_setup_host;
>>
>> +       pm_runtime_put(dev);
>> +
>>         return 0;
>>
>>  err_setup_host:
>>         sdhci_cleanup_host(host);
>> +err_rpm:
>> +       pm_runtime_disable(dev);
>> +       pm_runtime_put_noidle(dev);
>>  err_clk:
>>         clk_disable_unprepare(pltfm_host->clk);
>>         clk_disable_unprepare(priv->bus_clk);
>> @@ -602,9 +612,13 @@ static int dwcmshc_suspend(struct device *dev)
>>         struct rk35xx_priv *rk_priv = priv->priv;
>>         int ret;
>>
>> +       pm_runtime_get_sync(dev);
>> +
>>         ret = sdhci_suspend_host(host);
>> -       if (ret)
>> +       if (ret) {
>> +               pm_runtime_put(dev);
>>                 return ret;
>> +       }
>>
>>         clk_disable_unprepare(pltfm_host->clk);
>>         if (!IS_ERR(priv->bus_clk))
>> @@ -642,11 +656,65 @@ static int dwcmshc_resume(struct device *dev)
>>                         return ret;
>>         }
>>
>> -       return sdhci_resume_host(host);
>> +       ret = sdhci_resume_host(host);
>> +       if (ret)
>> +               return ret;
>> +
>> +       pm_runtime_put(dev);
> 
> To simplify the error path, I would suggest that you move the call to
> pm_runtime_put() to dwcmshc_suspend(). In fact what you need is just a
> call to pm_runtime_put_noidle(), somewhere after the call to
> pm_runtime_get_sync().
> 
> This is because runtime suspend is prevented by the PM core as it
> bumps the usage count with a pm_runtime_get_noresume() in the
> device_prepare() phase.

I thought you didn't want to assume that, because in that case
it can just be pm_runtime_resume() instead of pm_runtime_get_sync(),
and then no 'put' is needed at all.

> 
>> +
>> +       return 0;
>>  }
>>  #endif
> 
> [...]
> 
> Kind regards
> Uffe