Re: [PATCH V1] mmc: tegra: add sdhci tegra suspend and resume

[Thierry Reding]

On Tue, Apr 02, 2019 at 04:41:18PM +0000, Sowjanya Komatineni wrote:
> > On Mon, Apr 01, 2019 at 10:27:33PM -0700, Sowjanya Komatineni wrote:
> > > This patch adds suspend and resume PM ops for tegra SDHCI.
> > > 
> > > Signed-off-by: Sowjanya Komatineni <skomatineni@xxxxxxxxxx>
> > > ---
> > >  drivers/mmc/host/sdhci-tegra.c | 45 
> > > +++++++++++++++++++++++++++++++++++++++++-
> > >  1 file changed, 44 insertions(+), 1 deletion(-)
> > > 
> > > diff --git a/drivers/mmc/host/sdhci-tegra.c 
> > > b/drivers/mmc/host/sdhci-tegra.c index eafaaefab4a6..68263ea4a93e 
> > > 100644
> > > --- a/drivers/mmc/host/sdhci-tegra.c
> > > +++ b/drivers/mmc/host/sdhci-tegra.c
> > > @@ -1611,11 +1611,54 @@ static int sdhci_tegra_remove(struct platform_device *pdev)
> > >  	return 0;
> > >  }
> > >  
> > > +#ifdef CONFIG_PM_SLEEP
> > > +static int sdhci_tegra_suspend(struct device *dev) {
> > > +	struct sdhci_host *host = dev_get_drvdata(dev);
> > > +	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
> > > +	int ret;
> > > +
> > > +	if (host->mmc->caps2 & MMC_CAP2_CQE) {
> > > +		ret = cqhci_suspend(host->mmc);
> > > +		if (ret)
> > > +			return ret;
> > > +	}
> > > +
> > > +	ret = sdhci_suspend_host(host);
> > > +	if (ret)
> > > +		return ret;
> > > +
> > > +	clk_disable_unprepare(pltfm_host->clk);
> > > +	return 0;
> > > +}
> > > +
> > > +static int sdhci_tegra_resume(struct device *dev) {
> > > +	struct sdhci_host *host = dev_get_drvdata(dev);
> > > +	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
> > > +	int ret;
> > > +
> > > +	clk_prepare_enable(pltfm_host->clk);
> >
> > You need to check this for errors here.
> >
> > > +
> > > +	ret = sdhci_resume_host(host);
> > > +	if (ret)
> > > +		return ret;
> >
> > And clk_disable_unprepare() in case of failure.
> >
> > > +
> > > +	if (host->mmc->caps2 & MMC_CAP2_CQE)
> > > +		return cqhci_resume(host->mmc);
> >
> > Same here. Also make sure to sdhci_suspend_host() if cqhci_resume() fails. It's probably best to have a standard error cleanup path for this.
> >
> > Thierry
> 
> Thanks Thierry. Will add error check and cleanup for clock and sdhci_resume_host.
> Regarding, cqhci_resume always returns 0. Infact, it does nothing as resume happens on first cqhci_request.
> Since helper functions exists in cqhci driver, was using it for explicit resume call.

Yeah, I think it's good to use it. And I think it's also the safest
option to handle errors from it properly because somebody may eventually
add code to it that can fail. If we properly handle errors we'll be
prepared for when that happens.

Thierry

Attachment: signature.asc
Description: PGP signature