Re: [PATCH v3 2/7] mmc: sdhci-esdhc-imx: restore DLL override for DDR modes on resume

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] Runtime PM usage counter leak on the error path of `sdhci_esdhc_resume()`. While other commits in this series fixed pre-existing leaks in `sdhci_esdhc_suspend()` and `sdhci_esdhc_resume()`, the final patch in the series (commit "mmc: sdhci-esdhc-imx: fix resume error handling") newly introduced an early return if `pm_runtime_force_resume(dev)` fails. This bypasses the balancing `pm_runtime_put_autosuspend(dev)` call, permanently leaking the usage counter acquired during suspend.
--

commit 685c5877490d2d6169aac7e7a9adb332302c0a39
Author: Luke Wang <ziniu.wang_1@nxp.com>

mmc: sdhci-esdhc-imx: restore DLL override for DDR modes on resume

This commit extracts the DLL override setup into a helper function and
calls it during the resume path. This ensures that SDIO devices operating
in DDR mode retain their DLL override configuration if they keep power
during system suspend.

> diff --git a/drivers/mmc/host/sdhci-esdhc-imx.c b/drivers/mmc/host/sdhci-esdhc-imx.c
> index 6526d65538de5..a944351dbcdff 100644
> --- a/drivers/mmc/host/sdhci-esdhc-imx.c
> +++ b/drivers/mmc/host/sdhci-esdhc-imx.c

[ ... ]

> @@ -2123,9 +2132,18 @@ static int sdhci_esdhc_resume(struct device *dev)
>  	 * restore the saved tuning delay value for the device which keep
>  	 * power during system PM.
>  	 */
> -	if (mmc_card_keep_power(host->mmc) && esdhc_is_usdhc(imx_data))
> +	if (mmc_card_keep_power(host->mmc) && esdhc_is_usdhc(imx_data)) {
>  		sdhc_esdhc_tuning_restore(host);
>  
> +		/*
> +		 * Restore DLL override for DDR modes. hwinit unconditionally
> +		 * clears ESDHC_DLL_CTRL, but the card is still in DDR mode.
> +		 */
> +		if (host->timing == MMC_TIMING_UHS_DDR50 ||
> +		    host->timing == MMC_TIMING_MMC_DDR52)
> +			esdhc_set_dll_override(host);
> +	}
> +
>  	pm_runtime_put_autosuspend(dev);

[Severity: Medium]
Does this error path leak the runtime PM usage counter?

Looking at the beginning of sdhci_esdhc_resume():

	ret = pm_runtime_force_resume(dev);
	if (ret)
		return ret;

If pm_runtime_force_resume() fails, the function returns early and bypasses
this pm_runtime_put_autosuspend() call. Could this leave the usage counter
unbalanced and prevent the eSDHC controller from entering runtime suspend?

>  
>  	return ret;

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260629091954.3227151-1-ziniu.wang_1@oss.nxp.com?part=2