linux-kernel - Re: [PATCH V1] mmc: sdhci-msm: Add system suspend/resume callbacks

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Date:   Thu, 06 Feb 2020 21:14:46 +0530
From:   sbhanu@...eaurora.org
To:     Stephen Boyd <swboyd@...omium.org>
Cc:     adrian.hunter@...el.com, ulf.hansson@...aro.org,
        asutoshd@...eaurora.org, stummala@...eaurora.org,
        vbadigan@...eaurora.org, sayalil@...eaurora.org,
        cang@...eaurora.org, rampraka@...eaurora.org,
        linux-mmc@...r.kernel.org, linux-kernel@...r.kernel.org,
        linux-arm-msm@...r.kernel.org, agross@...nel.org,
        bjorn.andersson@...aro.org
Subject: Re: [PATCH V1] mmc: sdhci-msm: Add system suspend/resume callbacks

On 2020-01-22 22:27, Stephen Boyd wrote:
> Quoting Shaik Sajida Bhanu (2020-01-21 06:30:22)
>> Add system suspend/resume callbacks to sdhci-msm platform driver.
> 
> Yes, but why? There are already suspend/resume callbacks so this is
> replacing them too.
> 

Sure. Will update the commit text.

The present suspend/resume callbacks are only disabling the clocks.

Since eMMC/SDcard would be powered off during system suspend, we can do 
little more like

resetting controller, disabling interrupts. So updating the existing 
callbacks.

>> 
>> Signed-off-by: Shaik Sajida Bhanu <sbhanu@...eaurora.org>
>> ---
>>  drivers/mmc/host/sdhci-msm.c | 47 
>> ++++++++++++++++++++++++++++++++++++++++++--
>>  1 file changed, 45 insertions(+), 2 deletions(-)
>> 
>> diff --git a/drivers/mmc/host/sdhci-msm.c 
>> b/drivers/mmc/host/sdhci-msm.c
>> index 71f29ba..4984857 100644
>> --- a/drivers/mmc/host/sdhci-msm.c
>> +++ b/drivers/mmc/host/sdhci-msm.c
>> @@ -2028,9 +2028,52 @@ static __maybe_unused int 
>> sdhci_msm_runtime_resume(struct device *dev)
>>         return 0;
>>  }
>> 
>> +static int sdhci_msm_suspend(struct device *dev)
>> +{
>> +       struct sdhci_host *host = dev_get_drvdata(dev);
>> +       struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
>> +       struct sdhci_msm_host *msm_host = 
>> sdhci_pltfm_priv(pltfm_host);
>> +       int ret = 0;
>> +
>> +       if (host->mmc->caps2 & MMC_CAP2_CQE) {
> 
> It would be nice if this if check was rolled into cqhci_suspend so that
> all the callers wouldn't have to check it.
> 
>> +               ret = cqhci_suspend(host->mmc);
>> +               if (ret)
>> +                       return ret;
>> +       }
>> +
>> +       disable_irq(msm_host->pwr_irq);
> 
> Why is the irq disabled? Please add a comment.
> 

Will add a comment.

During system suspend all SDHC interrupts can be disabled since device 
would be

in power down state.

>> +       ret = sdhci_suspend_host(host);
>> +       if (ret)
>> +               return ret;
>> +
>> +       return sdhci_msm_runtime_suspend(dev);
> 
> pm_runtime_force_suspend() does different things than just call the
> runtime suspend function for the driver. For example, it disables
> runtime PM on the device. Can you explain in the commit text how this 
> is
> a correct conversion?
> 

Will invoke pm_runtime_force_suspend() here instead of 
sdhci_msm_runtime_suspend()

>> +}
>> +
>> +static int sdhci_msm_resume(struct device *dev)
>> +{
>> +       struct sdhci_host *host = dev_get_drvdata(dev);
>> +       struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
>> +       struct sdhci_msm_host *msm_host = 
>> sdhci_pltfm_priv(pltfm_host);
>> +       int ret = 0;
>> +
>> +       ret = sdhci_msm_runtime_resume(dev);
>> +       if (ret)
>> +               return ret;
>> +
>> +       ret = sdhci_resume_host(host);
>> +       if (ret < 0)
>> +               return ret;
>> +       enable_irq(msm_host->pwr_irq);
> 
> Same question here about irq. Deserves a comment.
> 
>> +
>> +       if (host->mmc->caps2 & MMC_CAP2_CQE)
>> +               ret = cqhci_resume(host->mmc);
>> +
>> +       return ret;
>> +}
>> +
>>  static const struct dev_pm_ops sdhci_msm_pm_ops = {
>> -       SET_SYSTEM_SLEEP_PM_OPS(pm_runtime_force_suspend,
>> -                               pm_runtime_force_resume)
>> +       SET_SYSTEM_SLEEP_PM_OPS(sdhci_msm_suspend,
>> +                               sdhci_msm_resume)
>>         SET_RUNTIME_PM_OPS(sdhci_msm_runtime_suspend,
>>                            sdhci_msm_runtime_resume,
>>                            NULL)