RE: [PATCH RESEND] scsi: ufs: core: Let delay value after LPM can be modified by vendor

From: Avri Altman
Date: Thu Oct 20 2022 - 03:18:14 EST


> From: Zhe Wang <zhe.wang1@xxxxxxxxxx>
>
> Some UFS devices require that the VCC should drop below 0.1V after
> turning off, otherwise device may not resume successfully. And
> because the power-off rate is different on different SOC platforms.
> Therefore, we hope that the delay can be modified by vendor to
> adjust the most appropriate delay value.
>
> Signed-off-by: Zhe Wang <zhe.wang1@xxxxxxxxxx>
> ---
> drivers/ufs/core/ufshcd.c | 11 +++++++++--
> include/ufs/ufshcd.h | 2 ++
> 2 files changed, 11 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/ufs/core/ufshcd.c b/drivers/ufs/core/ufshcd.c
> index 7256e6c43ca6..f6350231da0e 100644
> --- a/drivers/ufs/core/ufshcd.c
> +++ b/drivers/ufs/core/ufshcd.c
> @@ -89,6 +89,9 @@
> /* Polling time to wait for fDeviceInit */
> #define FDEVICEINIT_COMPL_TIMEOUT 1500 /* millisecs */
>
> +/* Default value of turn off VCC rail: 5000us */
> +#define UFS_VCC_TURNOFF_DELAY_US 5000
> +
> #define ufshcd_toggle_vreg(_dev, _vreg, _on) \
> ({ \
> int _ret; \
> @@ -7784,6 +7787,9 @@ static int ufs_get_device_desc(struct ufs_hba
> *hba)
>
> ufs_fixup_device_setup(hba);
>
> + if (hba->dev_quirks & UFS_DEVICE_QUIRK_DELAY_AFTER_LPM &&
> !hba->vcc_turnoff_delay)
> + hba->vcc_turnoff_delay = UFS_VCC_TURNOFF_DELAY_US;
Couldn't find where otherwise you are setting this value?
Also, UFS_DEVICE_QUIRK_DELAY_AFTER_LPM is only set for MTK.
Are you planning at some point adding your own host driver?

> +
> ufshcd_wb_probe(hba, desc_buf);
>
> ufshcd_temp_notif_probe(hba, desc_buf);
> @@ -8917,8 +8923,9 @@ static void ufshcd_vreg_set_lpm(struct ufs_hba
> *hba)
> * Some UFS devices require delay after VCC power rail is turned-off.
> */
> if (vcc_off && hba->vreg_info.vcc &&
> - hba->dev_quirks & UFS_DEVICE_QUIRK_DELAY_AFTER_LPM)
> - usleep_range(5000, 5100);
> + hba->dev_quirks & UFS_DEVICE_QUIRK_DELAY_AFTER_LPM &&
> + hba->vcc_turnoff_delay)
> + usleep_range(hba->vcc_turnoff_delay, hba->vcc_turnoff_delay +
> 100);
> }
>
> #ifdef CONFIG_PM
> diff --git a/include/ufs/ufshcd.h b/include/ufs/ufshcd.h
> index 9f28349ebcff..bfde3cb962fb 100644
> --- a/include/ufs/ufshcd.h
> +++ b/include/ufs/ufshcd.h
> @@ -828,6 +828,7 @@ struct ufs_hba_monitor {
> * device
> * @complete_put: whether or not to call ufshcd_rpm_put() from inside
> * ufshcd_resume_complete()
> + * @vcc_turnoff_delay: VCC turnoff delay value.
> */
> struct ufs_hba {
> void __iomem *mmio_base;
> @@ -975,6 +976,7 @@ struct ufs_hba {
> #endif
> u32 luns_avail;
> bool complete_put;
> + u32 vcc_turnoff_delay;
Can this be part of struct ufs_vreg instead?
Also maybe vcc_turnoff_delay_us?

Thanks,
Avri
> };
>
> /* Returns true if clocks can be gated. Otherwise false */
> --
> 2.17.1