Re: [PATCH 2/2] mmc: mtk-sd: add support for AN7581 MMC Host

[Christian Marangi <ansuelsmth@gmail.com>]

On Tue, Jan 21, 2025 at 06:25:48AM +0000, Andy-ld Lu (卢东) wrote:
> On Mon, 2025-01-20 at 18:29 +0100, Christian Marangi wrote:
> > External email : Please do not click links or open attachments until
> > you have verified the sender or the content.
> > 
> > 
> > On Thu, Jan 16, 2025 at 07:01:13AM +0000, Andy-ld Lu (卢东) wrote:
> > > On Wed, 2025-01-15 at 08:29 +0100, Christian Marangi wrote:
> > > > Add support for AN7581 MMC Host. The MMC Host controller is based
> > > > on
> > > > mt7622 with the difference of not having regulator supply and
> > > > state_uhs
> > > > pins and hclk clock.
> > > > 
> > > > Some minor fixes are applied to check if the state_uhs pins are
> > > > defined
> > > > and make hclk optional for the new airoha compatible.
> > > > 
> > > > Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
> > > > ---
> > > >  drivers/mmc/host/mtk-sd.c | 55 ++++++++++++++++++++++++++++++++-
> > > > ----
> > > > --
> > > >  1 file changed, 46 insertions(+), 9 deletions(-)
> > > > 
> > > > diff --git a/drivers/mmc/host/mtk-sd.c b/drivers/mmc/host/mtk-
> > > > sd.c
> > > > index efb0d2d5716b..9d6868883c91 100644
> > > > --- a/drivers/mmc/host/mtk-sd.c
> > > > +++ b/drivers/mmc/host/mtk-sd.c
> > > > @@ -666,6 +666,20 @@ static const struct mtk_mmc_compatible
> > > > mt8196_compat = {
> > > >     .support_new_rx = true,
> > > >  };
> > > > 
> > > > +static const struct mtk_mmc_compatible an7581_compat = {
> > > > +   .clk_div_bits = 12,
> > > > +   .recheck_sdio_irq = true,
> > > > +   .hs400_tune = false,
> > > > +   .pad_tune_reg = MSDC_PAD_TUNE0,
> > > > +   .async_fifo = true,
> > > > +   .data_tune = true,
> > > > +   .busy_check = true,
> > > > +   .stop_clk_fix = true,
> > > > +   .stop_dly_sel = 3,
> > > > +   .enhance_rx = true,
> > > > +   .support_64g = false,
> > > > +};
> > > > +
> > > >  static const struct of_device_id msdc_of_ids[] = {
> > > >     { .compatible = "mediatek,mt2701-mmc", .data =
> > > > &mt2701_compat},
> > > >     { .compatible = "mediatek,mt2712-mmc", .data =
> > > > &mt2712_compat},
> > > > @@ -680,7 +694,7 @@ static const struct of_device_id
> > > > msdc_of_ids[] =
> > > > {
> > > >     { .compatible = "mediatek,mt8183-mmc", .data =
> > > > &mt8183_compat},
> > > >     { .compatible = "mediatek,mt8196-mmc", .data =
> > > > &mt8196_compat},
> > > >     { .compatible = "mediatek,mt8516-mmc", .data =
> > > > &mt8516_compat},
> > > > -
> > > > +   { .compatible = "airoha,an7581-mmc", .data = &an7581_compat},
> > > >     {}
> > > >  };
> > > >  MODULE_DEVICE_TABLE(of, msdc_of_ids);
> > > > @@ -1600,6 +1614,10 @@ static int msdc_ops_switch_volt(struct
> > > > mmc_host *mmc, struct mmc_ios *ios)
> > > >     struct msdc_host *host = mmc_priv(mmc);
> > > >     int ret;
> > > > 
> > > > +   /* Skip setting supply if not supported */
> > > > +   if (!mmc->supply.vqmmc)
> > > > +           return 0;
> > > > +
> > > 
> > > Hi Christian,
> > > 
> > > I think here is no need. If you have not 'vqmmc' in the
> > > dts, IS_ERR(mmc->supply.vqmmc) would be -ENODEV and the
> > > corresponding
> > > flow would not be executed.
> > > 
> > > And another question, host->pins_default is just selected here,
> > > that
> > > would be lost.
> > > 
> > > >     if (!IS_ERR(mmc->supply.vqmmc)) {
> > > >             if (ios->signal_voltage != MMC_SIGNAL_VOLTAGE_330 &&
> > > >                 ios->signal_voltage != MMC_SIGNAL_VOLTAGE_180) {
> > > > @@ -1699,7 +1717,9 @@ static void msdc_enable_sdio_irq(struct
> > > > mmc_host *mmc, int enb)
> > > >                             dev_dbg(host->dev, "SDIO eint irq:
> > > > %d!\n", host->eint_irq);
> > > >                     }
> > > > 
> > > > -                   pinctrl_select_state(host->pinctrl, host-
> > > > > pins_uhs);
> > > > 
> > > > +                   /* Skip setting uhs pins if not supported */
> > > > +                   if (host->pins_uhs)
> > > > +                           pinctrl_select_state(host->pinctrl,
> > > > host->pins_uhs);
> > > >             } else {
> > > >                     dev_pm_clear_wake_irq(host->dev);
> > > >             }
> > > > @@ -2036,6 +2056,10 @@ static void msdc_ops_set_ios(struct
> > > > mmc_host
> > > > *mmc, struct mmc_ios *ios)
> > > > 
> > > >     msdc_set_buswidth(host, ios->bus_width);
> > > > 
> > > > +   /* Skip regulator if not supported */
> > > > +   if (!mmc->supply.vmmc)
> > > > +           goto skip_regulator;
> > > > +
> > > 
> > > No need too.
> > > 
> > > >     /* Suspend/Resume will do power off/on */
> > > >     switch (ios->power_mode) {
> > > >     case MMC_POWER_UP:
> > > > @@ -2071,6 +2095,7 @@ static void msdc_ops_set_ios(struct
> > > > mmc_host
> > > > *mmc, struct mmc_ios *ios)
> > > >             break;
> > > >     }
> > > > 
> > > > +skip_regulator:
> > > >     if (host->mclk != ios->clock || host->timing != ios->timing)
> > > >             msdc_set_mclk(host, ios->timing, ios->clock);
> > > >  }
> > > > @@ -2816,9 +2841,12 @@ static int msdc_of_clock_parse(struct
> > > > platform_device *pdev,
> > > >     if (IS_ERR(host->src_clk))
> > > >             return PTR_ERR(host->src_clk);
> > > > 
> > > > -   host->h_clk = devm_clk_get(&pdev->dev, "hclk");
> > > > -   if (IS_ERR(host->h_clk))
> > > > -           return PTR_ERR(host->h_clk);
> > > > +   /* AN7581 SoC doesn't have hclk */
> > > > +   if (!device_is_compatible(&pdev->dev, "airoha,an7581-mmc")) {
> > > > +           host->h_clk = devm_clk_get(&pdev->dev, "hclk");
> > > > +           if (IS_ERR(host->h_clk))
> > > > +                   return PTR_ERR(host->h_clk);
> > > > +   }
> > > 
> > > devm_clk_get_optional could be used to instead here, no need to use
> > > compatible to distinguish.
> > > 
> > 
> > I can make the hclk optional but I think this would affect also every
> > other compatible by hiding broken clock configuration.
> > 
> > > >     host->bus_clk = devm_clk_get_optional(&pdev->dev, "bus_clk");
> > > >     if (IS_ERR(host->bus_clk))
> > > > @@ -2926,10 +2954,13 @@ static int msdc_drv_probe(struct
> > > > platform_device *pdev)
> > > >             return PTR_ERR(host->pins_default);
> > > >     }
> > > > 
> > > > -   host->pins_uhs = pinctrl_lookup_state(host->pinctrl,
> > > > "state_uhs");
> > > > -   if (IS_ERR(host->pins_uhs)) {
> > > > -           dev_err(&pdev->dev, "Cannot find pinctrl uhs!\n");
> > > > -           return PTR_ERR(host->pins_uhs);
> > > > +   /* AN7581 doesn't have state_uhs pins */
> > > > +   if (!device_is_compatible(&pdev->dev, "airoha,an7581-mmc")) {
> > > > +           host->pins_uhs = pinctrl_lookup_state(host->pinctrl,
> > > > "state_uhs");
> > > > +           if (IS_ERR(host->pins_uhs)) {
> > > > +                   dev_err(&pdev->dev, "Cannot find pinctrl
> > > > uhs!\n");
> > > > +                   return PTR_ERR(host->pins_uhs);
> > > > +           }
> > > >     }
> > > 
> > > Could you consider to set a dummy 'state_uhs' same as
> > > 'state_default'
> > > in the dts, that you could not use compatible to distinguish here.
> > > 
> > 
> > This is problematic, correct me if I'm wrong, you are suggesting to
> > assign the emmc pins to both default and uhs? This is problematic as
> > the
> > pinctrl driver would complain that such pins are already assigned to
> > something. Also I don't think it's possible to assign these pins to a
> > dummy pin.
> > 
> Maybe I have not expressed clearly...What I mean is that you could set
> as below, and the content of &mmc_pins_uhs is just copied from
> &mmc_pins_default.
> 
> mmc@1fa0e000 {
> 	...
> 	pinctrl-names = "default", "state_uhs";
> 	pinctrl-0 = <&mmc_pins_default>;
> 	pinctrl-1 = <&mmc_pins_uhs>;
> }

Ok my bad. I did declared the second pin to pinctrl-0 instead of adding
pinctrl-1. With that it does work correctly.

> > > > 
> > > >     /* Support for SDIO eint irq ? */
> > > > @@ -3010,6 +3041,12 @@ static int msdc_drv_probe(struct
> > > > platform_device *pdev)
> > > >             dev_err(&pdev->dev, "Cannot ungate clocks!\n");
> > > >             goto release_clk;
> > > >     }
> > > > +
> > > > +   /* AN7581 without regulator require tune to OCR values */
> > > > +   if (device_is_compatible(&pdev->dev, "airoha,an7581-mmc") &&
> > > > +       !mmc->ocr_avail)
> > > > +           mmc->ocr_avail = MMC_VDD_32_33 | MMC_VDD_33_34;
> > > > +
> > > 
> > > Maybe you could use regulator-fixed in the dts and configure
> > > min/max
> > > voltage to get ocr_avail, no need to set hard code here.
> > > 
> > 
> > Also suggested by Wenbin Mei (梅文彬) and I just tried this.
> > 
> > This can't be done, fixed-regulator needs to have the same min and
> > max
> > voltage or they fail to probe sooo fixed-regulator saddly can't be
> > used
> > :(
> > 
> > I will send a new version of this with the other point corrected but
> > I
> > think a compatible and these additional if is a must :(
> If use the fixed regulator such as below, you will get the same
> ocr_avail as 'MMC_VDD_32_33 | MMC_VDD_33_34' through
> mmc_regulator_get_ocrmask().
> 
> vmmc_3v3: regulator-vmmc-3v3 {
> 	compatible = "regulator-fixed";
> 	regulator-name = "vmmc";
> 	regulator-min-microvolt = <3300000>;
> 	regulator-max-microvolt = <3300000>;
> 	regulator-always-on;
> }

Ok the code was a bit confusing but yes I can confirm that a 3.3 fixed
regulator define those 2 flags so also this is OK.

There is still the discussion about clock. You are totally against a new
compatible for the hclk? 

> > 
> > > >     msdc_init_hw(host);
> > > > 
> > > >     if (mmc->caps2 & MMC_CAP2_CQE) {
> > 
> > --
> >         Ansuel

-- 
	Ansuel