[PATCH 0/3] Adding UHS support for dw_mmc driver

[PATCH 0/3] Adding UHS support for dw_mmc driver

[Yuvaraj Kumar C D]

This series adds UHS support for dw_mmc driver.
Patch[1] reworks the handling of vmmc and vqmmc regulators.
Patch[2] handles the case where in some boards uses built-in CD line for
card detection and connected to a same voltage domain as of the IO rails.
Patch[3] was taken from chrome tree originally developed by Doug Anderson.

These patches were tested on exynos based chromebook with 3.16-rc1.

Doug Anderson (1):
  mmc: dw_mmc: Support voltage changes

Yuvaraj Kumar C D (2):
  mmc: dw_mmc: use mmc_regulator_get_supply to handle regulators
  mmc: dw_mmc: Dont cut off vqmmc and vmmc

 drivers/mmc/host/dw_mmc.c  |  239 +++++++++++++++++++++++++++++++++++---------
 drivers/mmc/host/dw_mmc.h  |    7 +-
 include/linux/mmc/dw_mmc.h |    2 +
 3 files changed, 201 insertions(+), 47 deletions(-)

-- 
1.7.10.4

[PATCH 1/3] mmc: dw_mmc: use mmc_regulator_get_supply to handle regulators

[Yuvaraj Kumar C D]

This patch makes use of mmc_regulator_get_supply() to handle
the vmmc and vqmmc regulators.Also it moves the code handling
the these regulators to dw_mci_set_ios().It turned on the vmmc
and vqmmc during MMC_POWER_UP and MMC_POWER_ON,and turned off
during MMC_POWER_OFF.

Signed-off-by: Yuvaraj Kumar C D <yuvaraj.cd@samsung.com>
---
 drivers/mmc/host/dw_mmc.c |   71 ++++++++++++++++++++++-----------------------
 drivers/mmc/host/dw_mmc.h |    2 ++
 2 files changed, 36 insertions(+), 37 deletions(-)

diff --git a/drivers/mmc/host/dw_mmc.c b/drivers/mmc/host/dw_mmc.c
index 1ac227c..f5cabce 100644
--- a/drivers/mmc/host/dw_mmc.c
+++ b/drivers/mmc/host/dw_mmc.c
@@ -937,6 +937,7 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
 	struct dw_mci_slot *slot = mmc_priv(mmc);
 	const struct dw_mci_drv_data *drv_data = slot->host->drv_data;
 	u32 regs;
+	int ret;
 
 	switch (ios->bus_width) {
 	case MMC_BUS_WIDTH_4:
@@ -975,16 +976,41 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
 
 	switch (ios->power_mode) {
 	case MMC_POWER_UP:
+		if ((!IS_ERR(mmc->supply.vmmc)) &&
+				!test_bit(DW_MMC_CARD_POWERED, &slot->flags)) {
+			ret = regulator_enable(mmc->supply.vmmc);
+			if (!ret)
+				set_bit(DW_MMC_CARD_POWERED, &slot->flags);
+		}
 		set_bit(DW_MMC_CARD_NEED_INIT, &slot->flags);
 		regs = mci_readl(slot->host, PWREN);
 		regs |= (1 << slot->id);
 		mci_writel(slot->host, PWREN, regs);
 		break;
 	case MMC_POWER_OFF:
+		if (!IS_ERR(mmc->supply.vqmmc) &&
+				test_bit(DW_MMC_IO_POWERED, &slot->flags)) {
+			ret = regulator_disable(mmc->supply.vqmmc);
+			if (!ret)
+				clear_bit(DW_MMC_IO_POWERED, &slot->flags);
+		}
+		if (!IS_ERR(mmc->supply.vmmc) &&
+				test_bit(DW_MMC_CARD_POWERED, &slot->flags)) {
+			ret = regulator_disable(mmc->supply.vmmc);
+			if (!ret)
+				clear_bit(DW_MMC_CARD_POWERED, &slot->flags);
+		}
 		regs = mci_readl(slot->host, PWREN);
 		regs &= ~(1 << slot->id);
 		mci_writel(slot->host, PWREN, regs);
 		break;
+	case MMC_POWER_ON:
+		if (!IS_ERR(mmc->supply.vqmmc) &&
+				!test_bit(DW_MMC_IO_POWERED, &slot->flags)) {
+			ret = regulator_enable(mmc->supply.vqmmc);
+			if (!ret)
+				set_bit(DW_MMC_IO_POWERED, &slot->flags);
+		}
 	default:
 		break;
 	}
@@ -2067,7 +2093,13 @@ static int dw_mci_init_slot(struct dw_mci *host, unsigned int id)
 		mmc->f_max = freq[1];
 	}
 
-	mmc->ocr_avail = MMC_VDD_32_33 | MMC_VDD_33_34;
+	/*if there are external regulators, get them*/
+	ret = mmc_regulator_get_supply(mmc);
+	if (ret == -EPROBE_DEFER)
+		goto err_setup_bus;
+
+	if (!mmc->ocr_avail)
+		mmc->ocr_avail = MMC_VDD_32_33 | MMC_VDD_33_34;
 
 	if (host->pdata->caps)
 		mmc->caps = host->pdata->caps;
@@ -2133,7 +2165,7 @@ static int dw_mci_init_slot(struct dw_mci *host, unsigned int id)
 
 err_setup_bus:
 	mmc_free_host(mmc);
-	return -EINVAL;
+	return ret;
 }
 
 static void dw_mci_cleanup_slot(struct dw_mci_slot *slot, unsigned int id)
@@ -2375,24 +2407,6 @@ int dw_mci_probe(struct dw_mci *host)
 		}
 	}
 
-	host->vmmc = devm_regulator_get_optional(host->dev, "vmmc");
-	if (IS_ERR(host->vmmc)) {
-		ret = PTR_ERR(host->vmmc);
-		if (ret == -EPROBE_DEFER)
-			goto err_clk_ciu;
-
-		dev_info(host->dev, "no vmmc regulator found: %d\n", ret);
-		host->vmmc = NULL;
-	} else {
-		ret = regulator_enable(host->vmmc);
-		if (ret) {
-			if (ret != -EPROBE_DEFER)
-				dev_err(host->dev,
-					"regulator_enable fail: %d\n", ret);
-			goto err_clk_ciu;
-		}
-	}
-
 	host->quirks = host->pdata->quirks;
 
 	spin_lock_init(&host->lock);
@@ -2536,8 +2550,6 @@ err_workqueue:
 err_dmaunmap:
 	if (host->use_dma && host->dma_ops->exit)
 		host->dma_ops->exit(host);
-	if (host->vmmc)
-		regulator_disable(host->vmmc);
 
 err_clk_ciu:
 	if (!IS_ERR(host->ciu_clk))
@@ -2573,9 +2585,6 @@ void dw_mci_remove(struct dw_mci *host)
 	if (host->use_dma && host->dma_ops->exit)
 		host->dma_ops->exit(host);
 
-	if (host->vmmc)
-		regulator_disable(host->vmmc);
-
 	if (!IS_ERR(host->ciu_clk))
 		clk_disable_unprepare(host->ciu_clk);
 
@@ -2592,9 +2601,6 @@ EXPORT_SYMBOL(dw_mci_remove);
  */
 int dw_mci_suspend(struct dw_mci *host)
 {
-	if (host->vmmc)
-		regulator_disable(host->vmmc);
-
 	return 0;
 }
 EXPORT_SYMBOL(dw_mci_suspend);
@@ -2603,15 +2609,6 @@ int dw_mci_resume(struct dw_mci *host)
 {
 	int i, ret;
 
-	if (host->vmmc) {
-		ret = regulator_enable(host->vmmc);
-		if (ret) {
-			dev_err(host->dev,
-				"failed to enable regulator: %d\n", ret);
-			return ret;
-		}
-	}
-
 	if (!dw_mci_ctrl_all_reset(host)) {
 		ret = -ENODEV;
 		return ret;
diff --git a/drivers/mmc/host/dw_mmc.h b/drivers/mmc/host/dw_mmc.h
index 738fa24..5c95d00 100644
--- a/drivers/mmc/host/dw_mmc.h
+++ b/drivers/mmc/host/dw_mmc.h
@@ -225,6 +225,8 @@ struct dw_mci_slot {
 	unsigned long		flags;
 #define DW_MMC_CARD_PRESENT	0
 #define DW_MMC_CARD_NEED_INIT	1
+#define DW_MMC_CARD_POWERED    2
+#define DW_MMC_IO_POWERED      3
 	int			id;
 	int			last_detect_state;
 };
-- 
1.7.10.4

[PATCH 1/3] mmc: dw_mmc: use mmc_regulator_get_supply to handle regulators

[Doug Anderson]

Yuvaraj,

On Mon, Jun 23, 2014 at 3:45 AM, Yuvaraj Kumar C D <yuvaraj.cd@gmail.com> wrote:
> This patch makes use of mmc_regulator_get_supply() to handle
> the vmmc and vqmmc regulators.Also it moves the code handling
> the these regulators to dw_mci_set_ios().It turned on the vmmc
> and vqmmc during MMC_POWER_UP and MMC_POWER_ON,and turned off
> during MMC_POWER_OFF.
>
> Signed-off-by: Yuvaraj Kumar C D <yuvaraj.cd@samsung.com>
> ---
>  drivers/mmc/host/dw_mmc.c |   71 ++++++++++++++++++++++-----------------------
>  drivers/mmc/host/dw_mmc.h |    2 ++
>  2 files changed, 36 insertions(+), 37 deletions(-)

Perhaps you could CC me on the whole series for the next version since
I was involved in privately reviewing previous versions?

Overall caveat for my review is that I'm nowhere near an SD/MMC expert.


> diff --git a/drivers/mmc/host/dw_mmc.c b/drivers/mmc/host/dw_mmc.c
> index 1ac227c..f5cabce 100644
> --- a/drivers/mmc/host/dw_mmc.c
> +++ b/drivers/mmc/host/dw_mmc.c
> @@ -937,6 +937,7 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
>         struct dw_mci_slot *slot = mmc_priv(mmc);
>         const struct dw_mci_drv_data *drv_data = slot->host->drv_data;
>         u32 regs;
> +       int ret;
>
>         switch (ios->bus_width) {
>         case MMC_BUS_WIDTH_4:
> @@ -975,16 +976,41 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
>
>         switch (ios->power_mode) {
>         case MMC_POWER_UP:
> +               if ((!IS_ERR(mmc->supply.vmmc)) &&
> +                               !test_bit(DW_MMC_CARD_POWERED, &slot->flags)) {
> +                       ret = regulator_enable(mmc->supply.vmmc);
> +                       if (!ret)
> +                               set_bit(DW_MMC_CARD_POWERED, &slot->flags);
> +               }

As per below, I'm not sure why you'd want to turn on vqmmc and vmmc at
different times.

Also: if you fail to turn on either of the regulators it feels like
you should print a pretty nasty error message since your device will
almost certainly not work.


>                 set_bit(DW_MMC_CARD_NEED_INIT, &slot->flags);
>                 regs = mci_readl(slot->host, PWREN);
>                 regs |= (1 << slot->id);
>                 mci_writel(slot->host, PWREN, regs);
>                 break;
>         case MMC_POWER_OFF:
> +               if (!IS_ERR(mmc->supply.vqmmc) &&
> +                               test_bit(DW_MMC_IO_POWERED, &slot->flags)) {
> +                       ret = regulator_disable(mmc->supply.vqmmc);
> +                       if (!ret)
> +                               clear_bit(DW_MMC_IO_POWERED, &slot->flags);
> +               }
> +               if (!IS_ERR(mmc->supply.vmmc) &&
> +                               test_bit(DW_MMC_CARD_POWERED, &slot->flags)) {
> +                       ret = regulator_disable(mmc->supply.vmmc);
> +                       if (!ret)
> +                               clear_bit(DW_MMC_CARD_POWERED, &slot->flags);
> +               }
>                 regs = mci_readl(slot->host, PWREN);
>                 regs &= ~(1 << slot->id);
>                 mci_writel(slot->host, PWREN, regs);
>                 break;
> +       case MMC_POWER_ON:
> +               if (!IS_ERR(mmc->supply.vqmmc) &&
> +                               !test_bit(DW_MMC_IO_POWERED, &slot->flags)) {
> +                       ret = regulator_enable(mmc->supply.vqmmc);
> +                       if (!ret)
> +                               set_bit(DW_MMC_IO_POWERED, &slot->flags);
> +               }
>         default:
>                 break;
>         }
> @@ -2067,7 +2093,13 @@ static int dw_mci_init_slot(struct dw_mci *host, unsigned int id)
>                 mmc->f_max = freq[1];
>         }
>
> -       mmc->ocr_avail = MMC_VDD_32_33 | MMC_VDD_33_34;
> +       /*if there are external regulators, get them*/
> +       ret = mmc_regulator_get_supply(mmc);
> +       if (ret == -EPROBE_DEFER)
> +               goto err_setup_bus;
> +
> +       if (!mmc->ocr_avail)
> +               mmc->ocr_avail = MMC_VDD_32_33 | MMC_VDD_33_34;
>
>         if (host->pdata->caps)
>                 mmc->caps = host->pdata->caps;
> @@ -2133,7 +2165,7 @@ static int dw_mci_init_slot(struct dw_mci *host, unsigned int id)
>
>  err_setup_bus:
>         mmc_free_host(mmc);
> -       return -EINVAL;
> +       return ret;
>  }
>
>  static void dw_mci_cleanup_slot(struct dw_mci_slot *slot, unsigned int id)
> @@ -2375,24 +2407,6 @@ int dw_mci_probe(struct dw_mci *host)
>                 }
>         }
>
> -       host->vmmc = devm_regulator_get_optional(host->dev, "vmmc");
> -       if (IS_ERR(host->vmmc)) {
> -               ret = PTR_ERR(host->vmmc);
> -               if (ret == -EPROBE_DEFER)
> -                       goto err_clk_ciu;
> -
> -               dev_info(host->dev, "no vmmc regulator found: %d\n", ret);
> -               host->vmmc = NULL;
> -       } else {
> -               ret = regulator_enable(host->vmmc);
> -               if (ret) {
> -                       if (ret != -EPROBE_DEFER)
> -                               dev_err(host->dev,
> -                                       "regulator_enable fail: %d\n", ret);
> -                       goto err_clk_ciu;
> -               }
> -       }
> -
>         host->quirks = host->pdata->quirks;
>
>         spin_lock_init(&host->lock);
> @@ -2536,8 +2550,6 @@ err_workqueue:
>  err_dmaunmap:
>         if (host->use_dma && host->dma_ops->exit)
>                 host->dma_ops->exit(host);
> -       if (host->vmmc)
> -               regulator_disable(host->vmmc);
>
>  err_clk_ciu:
>         if (!IS_ERR(host->ciu_clk))
> @@ -2573,9 +2585,6 @@ void dw_mci_remove(struct dw_mci *host)
>         if (host->use_dma && host->dma_ops->exit)
>                 host->dma_ops->exit(host);
>
> -       if (host->vmmc)
> -               regulator_disable(host->vmmc);
> -
>         if (!IS_ERR(host->ciu_clk))
>                 clk_disable_unprepare(host->ciu_clk);
>
> @@ -2592,9 +2601,6 @@ EXPORT_SYMBOL(dw_mci_remove);
>   */
>  int dw_mci_suspend(struct dw_mci *host)
>  {
> -       if (host->vmmc)
> -               regulator_disable(host->vmmc);
> -

Just to check: have you tested suspend/resume with the various
combinations of "keep-power-in-suspend" and "non-removable" with your
patch.  I remember some of the logic being a bit complicated...


>         return 0;
>  }
>  EXPORT_SYMBOL(dw_mci_suspend);
> @@ -2603,15 +2609,6 @@ int dw_mci_resume(struct dw_mci *host)
>  {
>         int i, ret;
>
> -       if (host->vmmc) {
> -               ret = regulator_enable(host->vmmc);
> -               if (ret) {
> -                       dev_err(host->dev,
> -                               "failed to enable regulator: %d\n", ret);
> -                       return ret;
> -               }
> -       }
> -
>         if (!dw_mci_ctrl_all_reset(host)) {
>                 ret = -ENODEV;
>                 return ret;
> diff --git a/drivers/mmc/host/dw_mmc.h b/drivers/mmc/host/dw_mmc.h
> index 738fa24..5c95d00 100644
> --- a/drivers/mmc/host/dw_mmc.h
> +++ b/drivers/mmc/host/dw_mmc.h

As I mentioned in my previous review, you should be removing "struct
regulator *vmmc; /* Power regulator */" from "struct dw_mci" since
you're no longer populating it.


> @@ -225,6 +225,8 @@ struct dw_mci_slot {
>         unsigned long           flags;
>  #define DW_MMC_CARD_PRESENT    0
>  #define DW_MMC_CARD_NEED_INIT  1
> +#define DW_MMC_CARD_POWERED    2
> +#define DW_MMC_IO_POWERED      3

I don't really think you should have two bits here.  From my
understanding of SD cards there should be very little reason to have
vqmmc and vmmc not powered at the same time.

If vqmmc is powered but vmmc is not powered then you'll get leakage
and the card can pull power through the IO lines which is bad for the
card.

I don't think that powering vmmc without vqmmc for a short time is
terribly bad but I can't quite see a good use case.  Essentially
you're powering the card but not able to talk to it, right?  I'm not
sure what the state of the IO lines would be (either driven low or
floating) since presumably the pullups on the lines are powered by
vqmmc too.


>         int                     id;
>         int                     last_detect_state;
>  };

[PATCH 1/3] mmc: dw_mmc: use mmc_regulator_get_supply to handle regulators

[Jaehoon Chung]

On 06/25/2014 03:00 AM, Doug Anderson wrote:
> Yuvaraj,
> 
> On Mon, Jun 23, 2014 at 3:45 AM, Yuvaraj Kumar C D <yuvaraj.cd@gmail.com> wrote:
>> This patch makes use of mmc_regulator_get_supply() to handle
>> the vmmc and vqmmc regulators.Also it moves the code handling
>> the these regulators to dw_mci_set_ios().It turned on the vmmc
>> and vqmmc during MMC_POWER_UP and MMC_POWER_ON,and turned off
>> during MMC_POWER_OFF.
>>
>> Signed-off-by: Yuvaraj Kumar C D <yuvaraj.cd@samsung.com>
>> ---
>>  drivers/mmc/host/dw_mmc.c |   71 ++++++++++++++++++++++-----------------------
>>  drivers/mmc/host/dw_mmc.h |    2 ++
>>  2 files changed, 36 insertions(+), 37 deletions(-)
> 
> Perhaps you could CC me on the whole series for the next version since
> I was involved in privately reviewing previous versions?
> 
> Overall caveat for my review is that I'm nowhere near an SD/MMC expert.
> 
> 
>> diff --git a/drivers/mmc/host/dw_mmc.c b/drivers/mmc/host/dw_mmc.c
>> index 1ac227c..f5cabce 100644
>> --- a/drivers/mmc/host/dw_mmc.c
>> +++ b/drivers/mmc/host/dw_mmc.c
>> @@ -937,6 +937,7 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
>>         struct dw_mci_slot *slot = mmc_priv(mmc);
>>         const struct dw_mci_drv_data *drv_data = slot->host->drv_data;
>>         u32 regs;
>> +       int ret;
>>
>>         switch (ios->bus_width) {
>>         case MMC_BUS_WIDTH_4:
>> @@ -975,16 +976,41 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
>>
>>         switch (ios->power_mode) {
>>         case MMC_POWER_UP:
>> +               if ((!IS_ERR(mmc->supply.vmmc)) &&
>> +                               !test_bit(DW_MMC_CARD_POWERED, &slot->flags)) {
>> +                       ret = regulator_enable(mmc->supply.vmmc);
>> +                       if (!ret)
>> +                               set_bit(DW_MMC_CARD_POWERED, &slot->flags);

MMC_CARD_POWERED? I didn't know why it needs.

>> +               }
> 
> As per below, I'm not sure why you'd want to turn on vqmmc and vmmc at
> different times.
In my case, in order to prevent the H/W mis-operation, it turns on vqmmc and vmmc at different times.

> 
> Also: if you fail to turn on either of the regulators it feels like
> you should print a pretty nasty error message since your device will
> almost certainly not work.
> 
> 
>>                 set_bit(DW_MMC_CARD_NEED_INIT, &slot->flags);
>>                 regs = mci_readl(slot->host, PWREN);
>>                 regs |= (1 << slot->id);
>>                 mci_writel(slot->host, PWREN, regs);
>>                 break;
>>         case MMC_POWER_OFF:
>> +               if (!IS_ERR(mmc->supply.vqmmc) &&
>> +                               test_bit(DW_MMC_IO_POWERED, &slot->flags)) {
>> +                       ret = regulator_disable(mmc->supply.vqmmc);
>> +                       if (!ret)
>> +                               clear_bit(DW_MMC_IO_POWERED, &slot->flags);
>> +               }
>> +               if (!IS_ERR(mmc->supply.vmmc) &&
>> +                               test_bit(DW_MMC_CARD_POWERED, &slot->flags)) {
>> +                       ret = regulator_disable(mmc->supply.vmmc);
>> +                       if (!ret)
>> +                               clear_bit(DW_MMC_CARD_POWERED, &slot->flags);
>> +               }
>>                 regs = mci_readl(slot->host, PWREN);
>>                 regs &= ~(1 << slot->id);
>>                 mci_writel(slot->host, PWREN, regs);
>>                 break;
>> +       case MMC_POWER_ON:
>> +               if (!IS_ERR(mmc->supply.vqmmc) &&
>> +                               !test_bit(DW_MMC_IO_POWERED, &slot->flags)) {
>> +                       ret = regulator_enable(mmc->supply.vqmmc);
>> +                       if (!ret)
>> +                               set_bit(DW_MMC_IO_POWERED, &slot->flags);
>> +               }
>>         default:
>>                 break;
>>         }
>> @@ -2067,7 +2093,13 @@ static int dw_mci_init_slot(struct dw_mci *host, unsigned int id)
>>                 mmc->f_max = freq[1];
>>         }
>>
>> -       mmc->ocr_avail = MMC_VDD_32_33 | MMC_VDD_33_34;
>> +       /*if there are external regulators, get them*/
>> +       ret = mmc_regulator_get_supply(mmc);
>> +       if (ret == -EPROBE_DEFER)
>> +               goto err_setup_bus;
>> +
>> +       if (!mmc->ocr_avail)
>> +               mmc->ocr_avail = MMC_VDD_32_33 | MMC_VDD_33_34;
>>
>>         if (host->pdata->caps)
>>                 mmc->caps = host->pdata->caps;
>> @@ -2133,7 +2165,7 @@ static int dw_mci_init_slot(struct dw_mci *host, unsigned int id)
>>
>>  err_setup_bus:
>>         mmc_free_host(mmc);
>> -       return -EINVAL;
>> +       return ret;
>>  }
>>
>>  static void dw_mci_cleanup_slot(struct dw_mci_slot *slot, unsigned int id)
>> @@ -2375,24 +2407,6 @@ int dw_mci_probe(struct dw_mci *host)
>>                 }
>>         }
>>
>> -       host->vmmc = devm_regulator_get_optional(host->dev, "vmmc");
>> -       if (IS_ERR(host->vmmc)) {
>> -               ret = PTR_ERR(host->vmmc);
>> -               if (ret == -EPROBE_DEFER)
>> -                       goto err_clk_ciu;
>> -
>> -               dev_info(host->dev, "no vmmc regulator found: %d\n", ret);
>> -               host->vmmc = NULL;
>> -       } else {
>> -               ret = regulator_enable(host->vmmc);
>> -               if (ret) {
>> -                       if (ret != -EPROBE_DEFER)
>> -                               dev_err(host->dev,
>> -                                       "regulator_enable fail: %d\n", ret);
>> -                       goto err_clk_ciu;
>> -               }
>> -       }
>> -
>>         host->quirks = host->pdata->quirks;
>>
>>         spin_lock_init(&host->lock);
>> @@ -2536,8 +2550,6 @@ err_workqueue:
>>  err_dmaunmap:
>>         if (host->use_dma && host->dma_ops->exit)
>>                 host->dma_ops->exit(host);
>> -       if (host->vmmc)
>> -               regulator_disable(host->vmmc);
>>
>>  err_clk_ciu:
>>         if (!IS_ERR(host->ciu_clk))
>> @@ -2573,9 +2585,6 @@ void dw_mci_remove(struct dw_mci *host)
>>         if (host->use_dma && host->dma_ops->exit)
>>                 host->dma_ops->exit(host);
>>
>> -       if (host->vmmc)
>> -               regulator_disable(host->vmmc);
>> -
>>         if (!IS_ERR(host->ciu_clk))
>>                 clk_disable_unprepare(host->ciu_clk);
>>
>> @@ -2592,9 +2601,6 @@ EXPORT_SYMBOL(dw_mci_remove);
>>   */
>>  int dw_mci_suspend(struct dw_mci *host)
>>  {
>> -       if (host->vmmc)
>> -               regulator_disable(host->vmmc);
>> -
> 
> Just to check: have you tested suspend/resume with the various
> combinations of "keep-power-in-suspend" and "non-removable" with your
> patch.  I remember some of the logic being a bit complicated...
> 
> 
>>         return 0;
>>  }
>>  EXPORT_SYMBOL(dw_mci_suspend);
>> @@ -2603,15 +2609,6 @@ int dw_mci_resume(struct dw_mci *host)
>>  {
>>         int i, ret;
>>
>> -       if (host->vmmc) {
>> -               ret = regulator_enable(host->vmmc);
>> -               if (ret) {
>> -                       dev_err(host->dev,
>> -                               "failed to enable regulator: %d\n", ret);
>> -                       return ret;
>> -               }
>> -       }
>> -
>>         if (!dw_mci_ctrl_all_reset(host)) {
>>                 ret = -ENODEV;
>>                 return ret;
>> diff --git a/drivers/mmc/host/dw_mmc.h b/drivers/mmc/host/dw_mmc.h
>> index 738fa24..5c95d00 100644
>> --- a/drivers/mmc/host/dw_mmc.h
>> +++ b/drivers/mmc/host/dw_mmc.h
> 
> As I mentioned in my previous review, you should be removing "struct
> regulator *vmmc; /* Power regulator */" from "struct dw_mci" since
> you're no longer populating it.
> 
> 
>> @@ -225,6 +225,8 @@ struct dw_mci_slot {
>>         unsigned long           flags;
>>  #define DW_MMC_CARD_PRESENT    0
>>  #define DW_MMC_CARD_NEED_INIT  1
>> +#define DW_MMC_CARD_POWERED    2
>> +#define DW_MMC_IO_POWERED      3
> 
> I don't really think you should have two bits here.  From my
> understanding of SD cards there should be very little reason to have
> vqmmc and vmmc not powered at the same time.
> 
> If vqmmc is powered but vmmc is not powered then you'll get leakage
> and the card can pull power through the IO lines which is bad for the
> card.
> 
> I don't think that powering vmmc without vqmmc for a short time is
> terribly bad but I can't quite see a good use case.  Essentially
> you're powering the card but not able to talk to it, right?  I'm not
> sure what the state of the IO lines would be (either driven low or
> floating) since presumably the pullups on the lines are powered by
> vqmmc too.
> 
> 
>>         int                     id;
>>         int                     last_detect_state;
>>  };
> --
> To unsubscribe from this list: send the line "unsubscribe linux-mmc" in
> the body of a message to majordomo at vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 

[PATCH 1/3] mmc: dw_mmc: use mmc_regulator_get_supply to handle regulators

[Doug Anderson]

Jaehoon,

On Tue, Jun 24, 2014 at 8:18 PM, Jaehoon Chung <jh80.chung@samsung.com> wrote:
> On 06/25/2014 03:00 AM, Doug Anderson wrote:
>> Yuvaraj,
>>
>> On Mon, Jun 23, 2014 at 3:45 AM, Yuvaraj Kumar C D <yuvaraj.cd@gmail.com> wrote:
>>> This patch makes use of mmc_regulator_get_supply() to handle
>>> the vmmc and vqmmc regulators.Also it moves the code handling
>>> the these regulators to dw_mci_set_ios().It turned on the vmmc
>>> and vqmmc during MMC_POWER_UP and MMC_POWER_ON,and turned off
>>> during MMC_POWER_OFF.
>>>
>>> Signed-off-by: Yuvaraj Kumar C D <yuvaraj.cd@samsung.com>
>>> ---
>>>  drivers/mmc/host/dw_mmc.c |   71 ++++++++++++++++++++++-----------------------
>>>  drivers/mmc/host/dw_mmc.h |    2 ++
>>>  2 files changed, 36 insertions(+), 37 deletions(-)
>>
>> Perhaps you could CC me on the whole series for the next version since
>> I was involved in privately reviewing previous versions?
>>
>> Overall caveat for my review is that I'm nowhere near an SD/MMC expert.
>>
>>
>>> diff --git a/drivers/mmc/host/dw_mmc.c b/drivers/mmc/host/dw_mmc.c
>>> index 1ac227c..f5cabce 100644
>>> --- a/drivers/mmc/host/dw_mmc.c
>>> +++ b/drivers/mmc/host/dw_mmc.c
>>> @@ -937,6 +937,7 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
>>>         struct dw_mci_slot *slot = mmc_priv(mmc);
>>>         const struct dw_mci_drv_data *drv_data = slot->host->drv_data;
>>>         u32 regs;
>>> +       int ret;
>>>
>>>         switch (ios->bus_width) {
>>>         case MMC_BUS_WIDTH_4:
>>> @@ -975,16 +976,41 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
>>>
>>>         switch (ios->power_mode) {
>>>         case MMC_POWER_UP:
>>> +               if ((!IS_ERR(mmc->supply.vmmc)) &&
>>> +                               !test_bit(DW_MMC_CARD_POWERED, &slot->flags)) {
>>> +                       ret = regulator_enable(mmc->supply.vmmc);
>>> +                       if (!ret)
>>> +                               set_bit(DW_MMC_CARD_POWERED, &slot->flags);
>
> MMC_CARD_POWERED? I didn't know why it needs.

I think the idea was to make sure that regulator enables/disables were
balanced.  ...but if the mmc core does this for us then we probably
don't need to worry.


>>> +               }
>>
>> As per below, I'm not sure why you'd want to turn on vqmmc and vmmc at
>> different times.
> In my case, in order to prevent the H/W mis-operation, it turns on vqmmc and vmmc at different times.

OK!  Can you explain more?  What instance is one powered but not the other?

I know that we talked to an SD Card manufacturer who told us that
powering their IO lines without their main power rail was a bad thing
(and we in fact saw this causing problems on SD cards).  I would guess
that means that you're either powering vmmc before vqmmc or that vqmmc
somehow doesn't directly enable pullups for IO lines.

If you need vmmc before vqmmc then I'm curious about why...

-Doug

[PATCH 1/3] mmc: dw_mmc: use mmc_regulator_get_supply to handle regulators

[Seungwon Jeon]

Hi Yuvaraj.

This patch looks like similar Jaehoon's.
Is it resending?
Anyway, I added some comments below.


On Wed, June 25, 2014, Jaehoon Chung wrote:
> On 06/25/2014 03:00 AM, Doug Anderson wrote:
> > Yuvaraj,
> >
> > On Mon, Jun 23, 2014 at 3:45 AM, Yuvaraj Kumar C D <yuvaraj.cd@gmail.com> wrote:
> >> This patch makes use of mmc_regulator_get_supply() to handle
> >> the vmmc and vqmmc regulators.Also it moves the code handling
> >> the these regulators to dw_mci_set_ios().It turned on the vmmc
> >> and vqmmc during MMC_POWER_UP and MMC_POWER_ON,and turned off
> >> during MMC_POWER_OFF.
> >>
> >> Signed-off-by: Yuvaraj Kumar C D <yuvaraj.cd@samsung.com>
> >> ---
> >>  drivers/mmc/host/dw_mmc.c |   71 ++++++++++++++++++++++-----------------------
> >>  drivers/mmc/host/dw_mmc.h |    2 ++
> >>  2 files changed, 36 insertions(+), 37 deletions(-)
> >
> > Perhaps you could CC me on the whole series for the next version since
> > I was involved in privately reviewing previous versions?
> >
> > Overall caveat for my review is that I'm nowhere near an SD/MMC expert.
> >
> >
> >> diff --git a/drivers/mmc/host/dw_mmc.c b/drivers/mmc/host/dw_mmc.c
> >> index 1ac227c..f5cabce 100644
> >> --- a/drivers/mmc/host/dw_mmc.c
> >> +++ b/drivers/mmc/host/dw_mmc.c
> >> @@ -937,6 +937,7 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
> >>         struct dw_mci_slot *slot = mmc_priv(mmc);
> >>         const struct dw_mci_drv_data *drv_data = slot->host->drv_data;
> >>         u32 regs;
> >> +       int ret;
> >>
> >>         switch (ios->bus_width) {
> >>         case MMC_BUS_WIDTH_4:
> >> @@ -975,16 +976,41 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
> >>
> >>         switch (ios->power_mode) {
> >>         case MMC_POWER_UP:
> >> +               if ((!IS_ERR(mmc->supply.vmmc)) &&
> >> +                               !test_bit(DW_MMC_CARD_POWERED, &slot->flags)) {
> >> +                       ret = regulator_enable(mmc->supply.vmmc);
> >> +                       if (!ret)
> >> +                               set_bit(DW_MMC_CARD_POWERED, &slot->flags);
> 
> MMC_CARD_POWERED? I didn't know why it needs.
Also, ios->vdd is missed. Use mmc_regulator_set_ocr() instead of regulator_enable() for vmmc.
And with mmc_regulator_set_ocr(), we don't need to check additional flag for enable/disable.
It does that already.

> 
> >> +               }
> >
> > As per below, I'm not sure why you'd want to turn on vqmmc and vmmc at
> > different times.
> In my case, in order to prevent the H/W mis-operation, it turns on vqmmc and vmmc at different times.
> 
> >
> > Also: if you fail to turn on either of the regulators it feels like
> > you should print a pretty nasty error message since your device will
> > almost certainly not work.
> >
> >
> >>                 set_bit(DW_MMC_CARD_NEED_INIT, &slot->flags);
> >>                 regs = mci_readl(slot->host, PWREN);
> >>                 regs |= (1 << slot->id);
> >>                 mci_writel(slot->host, PWREN, regs);
> >>                 break;
> >>         case MMC_POWER_OFF:
> >> +               if (!IS_ERR(mmc->supply.vqmmc) &&
> >> +                               test_bit(DW_MMC_IO_POWERED, &slot->flags)) {
> >> +                       ret = regulator_disable(mmc->supply.vqmmc);
> >> +                       if (!ret)
> >> +                               clear_bit(DW_MMC_IO_POWERED, &slot->flags);
> >> +               }
> >> +               if (!IS_ERR(mmc->supply.vmmc) &&
> >> +                               test_bit(DW_MMC_CARD_POWERED, &slot->flags)) {
> >> +                       ret = regulator_disable(mmc->supply.vmmc);
> >> +                       if (!ret)
> >> +                               clear_bit(DW_MMC_CARD_POWERED, &slot->flags);
> >> +               }
> >>                 regs = mci_readl(slot->host, PWREN);
> >>                 regs &= ~(1 << slot->id);
> >>                 mci_writel(slot->host, PWREN, regs);
> >>                 break;
> >> +       case MMC_POWER_ON:
> >> +               if (!IS_ERR(mmc->supply.vqmmc) &&
> >> +                               !test_bit(DW_MMC_IO_POWERED, &slot->flags)) {
You can use regulator_is_enabled() instead of flag bit, DW_MMC_IO_POWERED.
Important thing is that if powering vmmc failed at MMC_POWER_UP, vqmmc should not be powered.
Please consider Doug's mentions below.

> >> +                       ret = regulator_enable(mmc->supply.vqmmc);
> >> +                       if (!ret)
> >> +                               set_bit(DW_MMC_IO_POWERED, &slot->flags);
> >> +               }
> >>         default:
> >>                 break;
> >>         }
> >> @@ -2067,7 +2093,13 @@ static int dw_mci_init_slot(struct dw_mci *host, unsigned int id)
> >>                 mmc->f_max = freq[1];
> >>         }
> >>
> >> -       mmc->ocr_avail = MMC_VDD_32_33 | MMC_VDD_33_34;
> >> +       /*if there are external regulators, get them*/
> >> +       ret = mmc_regulator_get_supply(mmc);
> >> +       if (ret == -EPROBE_DEFER)
> >> +               goto err_setup_bus;
> >> +
> >> +       if (!mmc->ocr_avail)
> >> +               mmc->ocr_avail = MMC_VDD_32_33 | MMC_VDD_33_34;
> >>
> >>         if (host->pdata->caps)
> >>                 mmc->caps = host->pdata->caps;
> >> @@ -2133,7 +2165,7 @@ static int dw_mci_init_slot(struct dw_mci *host, unsigned int id)
> >>
> >>  err_setup_bus:
> >>         mmc_free_host(mmc);
> >> -       return -EINVAL;
> >> +       return ret;
> >>  }
> >>
> >>  static void dw_mci_cleanup_slot(struct dw_mci_slot *slot, unsigned int id)
> >> @@ -2375,24 +2407,6 @@ int dw_mci_probe(struct dw_mci *host)
> >>                 }
> >>         }
> >>
> >> -       host->vmmc = devm_regulator_get_optional(host->dev, "vmmc");
> >> -       if (IS_ERR(host->vmmc)) {
> >> -               ret = PTR_ERR(host->vmmc);
> >> -               if (ret == -EPROBE_DEFER)
> >> -                       goto err_clk_ciu;
> >> -
> >> -               dev_info(host->dev, "no vmmc regulator found: %d\n", ret);
> >> -               host->vmmc = NULL;
> >> -       } else {
> >> -               ret = regulator_enable(host->vmmc);
> >> -               if (ret) {
> >> -                       if (ret != -EPROBE_DEFER)
> >> -                               dev_err(host->dev,
> >> -                                       "regulator_enable fail: %d\n", ret);
> >> -                       goto err_clk_ciu;
> >> -               }
> >> -       }
> >> -
> >>         host->quirks = host->pdata->quirks;
> >>
> >>         spin_lock_init(&host->lock);
> >> @@ -2536,8 +2550,6 @@ err_workqueue:
> >>  err_dmaunmap:
> >>         if (host->use_dma && host->dma_ops->exit)
> >>                 host->dma_ops->exit(host);
> >> -       if (host->vmmc)
> >> -               regulator_disable(host->vmmc);
> >>
> >>  err_clk_ciu:
> >>         if (!IS_ERR(host->ciu_clk))
> >> @@ -2573,9 +2585,6 @@ void dw_mci_remove(struct dw_mci *host)
> >>         if (host->use_dma && host->dma_ops->exit)
> >>                 host->dma_ops->exit(host);
> >>
> >> -       if (host->vmmc)
> >> -               regulator_disable(host->vmmc);
> >> -
> >>         if (!IS_ERR(host->ciu_clk))
> >>                 clk_disable_unprepare(host->ciu_clk);
> >>
> >> @@ -2592,9 +2601,6 @@ EXPORT_SYMBOL(dw_mci_remove);
> >>   */
> >>  int dw_mci_suspend(struct dw_mci *host)
> >>  {
> >> -       if (host->vmmc)
> >> -               regulator_disable(host->vmmc);
> >> -
> >
> > Just to check: have you tested suspend/resume with the various
> > combinations of "keep-power-in-suspend" and "non-removable" with your
> > patch.  I remember some of the logic being a bit complicated...
Please check it and give some updates/result in next version.

Thanks,
Seungwon Jeon

> >
> >
> >>         return 0;
> >>  }
> >>  EXPORT_SYMBOL(dw_mci_suspend);
> >> @@ -2603,15 +2609,6 @@ int dw_mci_resume(struct dw_mci *host)
> >>  {
> >>         int i, ret;
> >>
> >> -       if (host->vmmc) {
> >> -               ret = regulator_enable(host->vmmc);
> >> -               if (ret) {
> >> -                       dev_err(host->dev,
> >> -                               "failed to enable regulator: %d\n", ret);
> >> -                       return ret;
> >> -               }
> >> -       }
> >> -
> >>         if (!dw_mci_ctrl_all_reset(host)) {
> >>                 ret = -ENODEV;
> >>                 return ret;
> >> diff --git a/drivers/mmc/host/dw_mmc.h b/drivers/mmc/host/dw_mmc.h
> >> index 738fa24..5c95d00 100644
> >> --- a/drivers/mmc/host/dw_mmc.h
> >> +++ b/drivers/mmc/host/dw_mmc.h
> >
> > As I mentioned in my previous review, you should be removing "struct
> > regulator *vmmc; /* Power regulator */" from "struct dw_mci" since
> > you're no longer populating it.
> >
> >
> >> @@ -225,6 +225,8 @@ struct dw_mci_slot {
> >>         unsigned long           flags;
> >>  #define DW_MMC_CARD_PRESENT    0
> >>  #define DW_MMC_CARD_NEED_INIT  1
> >> +#define DW_MMC_CARD_POWERED    2
> >> +#define DW_MMC_IO_POWERED      3
> >
> > I don't really think you should have two bits here.  From my
> > understanding of SD cards there should be very little reason to have
> > vqmmc and vmmc not powered at the same time.
> >
> > If vqmmc is powered but vmmc is not powered then you'll get leakage
> > and the card can pull power through the IO lines which is bad for the
> > card.
> >
> > I don't think that powering vmmc without vqmmc for a short time is
> > terribly bad but I can't quite see a good use case.  Essentially
> > you're powering the card but not able to talk to it, right?  I'm not
> > sure what the state of the IO lines would be (either driven low or
> > floating) since presumably the pullups on the lines are powered by
> > vqmmc too.
> >
> >
> >>         int                     id;
> >>         int                     last_detect_state;
> >>  };
> > --
> > To unsubscribe from this list: send the line "unsubscribe linux-mmc" in
> > the body of a message to majordomo at vger.kernel.org
> > More majordomo info at  http://vger.kernel.org/majordomo-info.html
> >
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-mmc" in
> the body of a message to majordomo at vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 1/3] mmc: dw_mmc: use mmc_regulator_get_supply to handle regulators

[Doug Anderson]

Seungwon,

On Wed, Jun 25, 2014 at 4:18 AM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
>> >> +       case MMC_POWER_ON:
>> >> +               if (!IS_ERR(mmc->supply.vqmmc) &&
>> >> +                               !test_bit(DW_MMC_IO_POWERED, &slot->flags)) {
> You can use regulator_is_enabled() instead of flag bit, DW_MMC_IO_POWERED.

I'd be a little worried about regulator_is_enabled() since regulators
are reference counted.  What if someone else is sharing this
regulator?  The regulator might happen to be enabled when you check it
but unless you add your own dw_mmc reference count they might turn it
off.

> Important thing is that if powering vmmc failed at MMC_POWER_UP, vqmmc should not be powered.

Good point.

[PATCH 1/3] mmc: dw_mmc: use mmc_regulator_get_supply to handle regulators

[Seungwon Jeon]

Hi Doug,

On Thu, June 26, 2014, Doug Anderson wrote:
> Seungwon,
> 
> On Wed, Jun 25, 2014 at 4:18 AM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
> >> >> +       case MMC_POWER_ON:
> >> >> +               if (!IS_ERR(mmc->supply.vqmmc) &&
> >> >> +                               !test_bit(DW_MMC_IO_POWERED, &slot->flags)) {
> > You can use regulator_is_enabled() instead of flag bit, DW_MMC_IO_POWERED.
> 
> I'd be a little worried about regulator_is_enabled() since regulators
> are reference counted.  What if someone else is sharing this
> regulator?  The regulator might happen to be enabled when you check it
> but unless you add your own dw_mmc reference count they might turn it
> off.
Cool, that's a possibility. Some assumption may need.
If mmc's core can guarantee its balance, I think we don't need to consider some flag.

Thanks,
Seungwon Jeon
> 
> > Important thing is that if powering vmmc failed at MMC_POWER_UP, vqmmc should not be powered.
> 
> Good point.
> --
> To unsubscribe from this list: send the line "unsubscribe linux-mmc" in
> the body of a message to majordomo at vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 1/3] mmc: dw_mmc: use mmc_regulator_get_supply to handle regulators

[Doug Anderson]

Seungwon,

On Thu, Jun 26, 2014 at 3:30 AM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
> Hi Doug,
>
> On Thu, June 26, 2014, Doug Anderson wrote:
>> Seungwon,
>>
>> On Wed, Jun 25, 2014 at 4:18 AM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
>> >> >> +       case MMC_POWER_ON:
>> >> >> +               if (!IS_ERR(mmc->supply.vqmmc) &&
>> >> >> +                               !test_bit(DW_MMC_IO_POWERED, &slot->flags)) {
>> > You can use regulator_is_enabled() instead of flag bit, DW_MMC_IO_POWERED.
>>
>> I'd be a little worried about regulator_is_enabled() since regulators
>> are reference counted.  What if someone else is sharing this
>> regulator?  The regulator might happen to be enabled when you check it
>> but unless you add your own dw_mmc reference count they might turn it
>> off.
> Cool, that's a possibility. Some assumption may need.
> If mmc's core can guarantee its balance, I think we don't need to consider some flag.

I notice that the mmc core seems to keep a flag itself for vdd (the
mmc->regulator_enabled flag).  That would imply that the core thought
it was important to have the extra flag and that we should keep our
own flag for vqmmc.

-Doug

[PATCH 1/3] mmc: dw_mmc: use mmc_regulator_get_supply to handle regulators

[Seungwon Jeon]

On Fri, June 27, 2014, Doug Anderson wrote:
> Seungwon,
> 
> On Thu, Jun 26, 2014 at 3:30 AM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
> > Hi Doug,
> >
> > On Thu, June 26, 2014, Doug Anderson wrote:
> >> Seungwon,
> >>
> >> On Wed, Jun 25, 2014 at 4:18 AM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
> >> >> >> +       case MMC_POWER_ON:
> >> >> >> +               if (!IS_ERR(mmc->supply.vqmmc) &&
> >> >> >> +                               !test_bit(DW_MMC_IO_POWERED, &slot->flags)) {
> >> > You can use regulator_is_enabled() instead of flag bit, DW_MMC_IO_POWERED.
> >>
> >> I'd be a little worried about regulator_is_enabled() since regulators
> >> are reference counted.  What if someone else is sharing this
> >> regulator?  The regulator might happen to be enabled when you check it
> >> but unless you add your own dw_mmc reference count they might turn it
> >> off.
> > Cool, that's a possibility. Some assumption may need.
> > If mmc's core can guarantee its balance, I think we don't need to consider some flag.
> 
> I notice that the mmc core seems to keep a flag itself for vdd (the
> mmc->regulator_enabled flag).  That would imply that the core thought
> it was important to have the extra flag and that we should keep our
> own flag for vqmmc.
Ok, Good.

Thanks,
Seungwon Jeon
> 
> -Doug
> --
> To unsubscribe from this list: send the line "unsubscribe linux-mmc" in
> the body of a message to majordomo at vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 1/3] mmc: dw_mmc: use mmc_regulator_get_supply to handle regulators

[Yuvaraj Kumar]

Doug

On Tue, Jun 24, 2014 at 11:30 PM, Doug Anderson <dianders@chromium.org> wrote:
> Yuvaraj,
>
> On Mon, Jun 23, 2014 at 3:45 AM, Yuvaraj Kumar C D <yuvaraj.cd@gmail.com> wrote:
>> This patch makes use of mmc_regulator_get_supply() to handle
>> the vmmc and vqmmc regulators.Also it moves the code handling
>> the these regulators to dw_mci_set_ios().It turned on the vmmc
>> and vqmmc during MMC_POWER_UP and MMC_POWER_ON,and turned off
>> during MMC_POWER_OFF.
>>
>> Signed-off-by: Yuvaraj Kumar C D <yuvaraj.cd@samsung.com>
>> ---
>>  drivers/mmc/host/dw_mmc.c |   71 ++++++++++++++++++++++-----------------------
>>  drivers/mmc/host/dw_mmc.h |    2 ++
>>  2 files changed, 36 insertions(+), 37 deletions(-)
>
> Perhaps you could CC me on the whole series for the next version since
> I was involved in privately reviewing previous versions?
It was just accidental missing you in the CC .Surely i will add you in
CC for next versions.
>
> Overall caveat for my review is that I'm nowhere near an SD/MMC expert.
>
>
>> diff --git a/drivers/mmc/host/dw_mmc.c b/drivers/mmc/host/dw_mmc.c
>> index 1ac227c..f5cabce 100644
>> --- a/drivers/mmc/host/dw_mmc.c
>> +++ b/drivers/mmc/host/dw_mmc.c
>> @@ -937,6 +937,7 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
>>         struct dw_mci_slot *slot = mmc_priv(mmc);
>>         const struct dw_mci_drv_data *drv_data = slot->host->drv_data;
>>         u32 regs;
>> +       int ret;
>>
>>         switch (ios->bus_width) {
>>         case MMC_BUS_WIDTH_4:
>> @@ -975,16 +976,41 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
>>
>>         switch (ios->power_mode) {
>>         case MMC_POWER_UP:
>> +               if ((!IS_ERR(mmc->supply.vmmc)) &&
>> +                               !test_bit(DW_MMC_CARD_POWERED, &slot->flags)) {
>> +                       ret = regulator_enable(mmc->supply.vmmc);
>> +                       if (!ret)
>> +                               set_bit(DW_MMC_CARD_POWERED, &slot->flags);
>> +               }
>
> As per below, I'm not sure why you'd want to turn on vqmmc and vmmc at
> different times.
As you can see people's have different opinion on this.When i had a
look at the other drivers in the subsystem which does in the same flow
as above.However i will change in the next version.
>
> Also: if you fail to turn on either of the regulators it feels like
> you should print a pretty nasty error message since your device will
> almost certainly not work.
Yes. I will add a error message.
>
>
>>                 set_bit(DW_MMC_CARD_NEED_INIT, &slot->flags);
>>                 regs = mci_readl(slot->host, PWREN);
>>                 regs |= (1 << slot->id);
>>                 mci_writel(slot->host, PWREN, regs);
>>                 break;
>>         case MMC_POWER_OFF:
>> +               if (!IS_ERR(mmc->supply.vqmmc) &&
>> +                               test_bit(DW_MMC_IO_POWERED, &slot->flags)) {
>> +                       ret = regulator_disable(mmc->supply.vqmmc);
>> +                       if (!ret)
>> +                               clear_bit(DW_MMC_IO_POWERED, &slot->flags);
>> +               }
>> +               if (!IS_ERR(mmc->supply.vmmc) &&
>> +                               test_bit(DW_MMC_CARD_POWERED, &slot->flags)) {
>> +                       ret = regulator_disable(mmc->supply.vmmc);
>> +                       if (!ret)
>> +                               clear_bit(DW_MMC_CARD_POWERED, &slot->flags);
>> +               }
>>                 regs = mci_readl(slot->host, PWREN);
>>                 regs &= ~(1 << slot->id);
>>                 mci_writel(slot->host, PWREN, regs);
>>                 break;
>> +       case MMC_POWER_ON:
>> +               if (!IS_ERR(mmc->supply.vqmmc) &&
>> +                               !test_bit(DW_MMC_IO_POWERED, &slot->flags)) {
>> +                       ret = regulator_enable(mmc->supply.vqmmc);
>> +                       if (!ret)
>> +                               set_bit(DW_MMC_IO_POWERED, &slot->flags);
>> +               }
>>         default:
>>                 break;
>>         }
>> @@ -2067,7 +2093,13 @@ static int dw_mci_init_slot(struct dw_mci *host, unsigned int id)
>>                 mmc->f_max = freq[1];
>>         }
>>
>> -       mmc->ocr_avail = MMC_VDD_32_33 | MMC_VDD_33_34;
>> +       /*if there are external regulators, get them*/
>> +       ret = mmc_regulator_get_supply(mmc);
>> +       if (ret == -EPROBE_DEFER)
>> +               goto err_setup_bus;
>> +
>> +       if (!mmc->ocr_avail)
>> +               mmc->ocr_avail = MMC_VDD_32_33 | MMC_VDD_33_34;
>>
>>         if (host->pdata->caps)
>>                 mmc->caps = host->pdata->caps;
>> @@ -2133,7 +2165,7 @@ static int dw_mci_init_slot(struct dw_mci *host, unsigned int id)
>>
>>  err_setup_bus:
>>         mmc_free_host(mmc);
>> -       return -EINVAL;
>> +       return ret;
>>  }
>>
>>  static void dw_mci_cleanup_slot(struct dw_mci_slot *slot, unsigned int id)
>> @@ -2375,24 +2407,6 @@ int dw_mci_probe(struct dw_mci *host)
>>                 }
>>         }
>>
>> -       host->vmmc = devm_regulator_get_optional(host->dev, "vmmc");
>> -       if (IS_ERR(host->vmmc)) {
>> -               ret = PTR_ERR(host->vmmc);
>> -               if (ret == -EPROBE_DEFER)
>> -                       goto err_clk_ciu;
>> -
>> -               dev_info(host->dev, "no vmmc regulator found: %d\n", ret);
>> -               host->vmmc = NULL;
>> -       } else {
>> -               ret = regulator_enable(host->vmmc);
>> -               if (ret) {
>> -                       if (ret != -EPROBE_DEFER)
>> -                               dev_err(host->dev,
>> -                                       "regulator_enable fail: %d\n", ret);
>> -                       goto err_clk_ciu;
>> -               }
>> -       }
>> -
>>         host->quirks = host->pdata->quirks;
>>
>>         spin_lock_init(&host->lock);
>> @@ -2536,8 +2550,6 @@ err_workqueue:
>>  err_dmaunmap:
>>         if (host->use_dma && host->dma_ops->exit)
>>                 host->dma_ops->exit(host);
>> -       if (host->vmmc)
>> -               regulator_disable(host->vmmc);
>>
>>  err_clk_ciu:
>>         if (!IS_ERR(host->ciu_clk))
>> @@ -2573,9 +2585,6 @@ void dw_mci_remove(struct dw_mci *host)
>>         if (host->use_dma && host->dma_ops->exit)
>>                 host->dma_ops->exit(host);
>>
>> -       if (host->vmmc)
>> -               regulator_disable(host->vmmc);
>> -
>>         if (!IS_ERR(host->ciu_clk))
>>                 clk_disable_unprepare(host->ciu_clk);
>>
>> @@ -2592,9 +2601,6 @@ EXPORT_SYMBOL(dw_mci_remove);
>>   */
>>  int dw_mci_suspend(struct dw_mci *host)
>>  {
>> -       if (host->vmmc)
>> -               regulator_disable(host->vmmc);
>> -
>
> Just to check: have you tested suspend/resume with the various
> combinations of "keep-power-in-suspend" and "non-removable" with your
> patch.  I remember some of the logic being a bit complicated...
Tested suspend/resume with the non-removable.Will do more testing with
combination of "keep-power-in-suspend" and "non-removable".
>
>
>>         return 0;
>>  }
>>  EXPORT_SYMBOL(dw_mci_suspend);
>> @@ -2603,15 +2609,6 @@ int dw_mci_resume(struct dw_mci *host)
>>  {
>>         int i, ret;
>>
>> -       if (host->vmmc) {
>> -               ret = regulator_enable(host->vmmc);
>> -               if (ret) {
>> -                       dev_err(host->dev,
>> -                               "failed to enable regulator: %d\n", ret);
>> -                       return ret;
>> -               }
>> -       }
>> -
>>         if (!dw_mci_ctrl_all_reset(host)) {
>>                 ret = -ENODEV;
>>                 return ret;
>> diff --git a/drivers/mmc/host/dw_mmc.h b/drivers/mmc/host/dw_mmc.h
>> index 738fa24..5c95d00 100644
>> --- a/drivers/mmc/host/dw_mmc.h
>> +++ b/drivers/mmc/host/dw_mmc.h
>
> As I mentioned in my previous review, you should be removing "struct
> regulator *vmmc; /* Power regulator */" from "struct dw_mci" since
> you're no longer populating it.
ok
>
>
>> @@ -225,6 +225,8 @@ struct dw_mci_slot {
>>         unsigned long           flags;
>>  #define DW_MMC_CARD_PRESENT    0
>>  #define DW_MMC_CARD_NEED_INIT  1
>> +#define DW_MMC_CARD_POWERED    2
>> +#define DW_MMC_IO_POWERED      3
>
> I don't really think you should have two bits here.  From my
> understanding of SD cards there should be very little reason to have
> vqmmc and vmmc not powered at the same time.
I think if i can use mmc_regulator_set_ocr(), we don't need additional
flag.But for tps65090 mmc_regulator_get_ocr() and
mmc_regulator_set_ocr() is failing as its a fixed-regulator.
>
> If vqmmc is powered but vmmc is not powered then you'll get leakage
> and the card can pull power through the IO lines which is bad for the
> card.
>
> I don't think that powering vmmc without vqmmc for a short time is
> terribly bad but I can't quite see a good use case.  Essentially
> you're powering the card but not able to talk to it, right?  I'm not
> sure what the state of the IO lines would be (either driven low or
> floating) since presumably the pullups on the lines are powered by
> vqmmc too.
>
>
>>         int                     id;
>>         int                     last_detect_state;
>>  };

[PATCH 1/3] mmc: dw_mmc: use mmc_regulator_get_supply to handle regulators

[Doug Anderson]

Yuvaraj,

On Thu, Jun 26, 2014 at 4:21 AM, Yuvaraj Kumar <yuvaraj.cd@gmail.com> wrote:
> Doug
>
> On Tue, Jun 24, 2014 at 11:30 PM, Doug Anderson <dianders@chromium.org> wrote:
>> Yuvaraj,
>>
>> On Mon, Jun 23, 2014 at 3:45 AM, Yuvaraj Kumar C D <yuvaraj.cd@gmail.com> wrote:
>>> This patch makes use of mmc_regulator_get_supply() to handle
>>> the vmmc and vqmmc regulators.Also it moves the code handling
>>> the these regulators to dw_mci_set_ios().It turned on the vmmc
>>> and vqmmc during MMC_POWER_UP and MMC_POWER_ON,and turned off
>>> during MMC_POWER_OFF.
>>>
>>> Signed-off-by: Yuvaraj Kumar C D <yuvaraj.cd@samsung.com>
>>> ---
>>>  drivers/mmc/host/dw_mmc.c |   71 ++++++++++++++++++++++-----------------------
>>>  drivers/mmc/host/dw_mmc.h |    2 ++
>>>  2 files changed, 36 insertions(+), 37 deletions(-)
>>
>> Perhaps you could CC me on the whole series for the next version since
>> I was involved in privately reviewing previous versions?
> It was just accidental missing you in the CC .Surely i will add you in
> CC for next versions.
>>
>> Overall caveat for my review is that I'm nowhere near an SD/MMC expert.
>>
>>
>>> diff --git a/drivers/mmc/host/dw_mmc.c b/drivers/mmc/host/dw_mmc.c
>>> index 1ac227c..f5cabce 100644
>>> --- a/drivers/mmc/host/dw_mmc.c
>>> +++ b/drivers/mmc/host/dw_mmc.c
>>> @@ -937,6 +937,7 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
>>>         struct dw_mci_slot *slot = mmc_priv(mmc);
>>>         const struct dw_mci_drv_data *drv_data = slot->host->drv_data;
>>>         u32 regs;
>>> +       int ret;
>>>
>>>         switch (ios->bus_width) {
>>>         case MMC_BUS_WIDTH_4:
>>> @@ -975,16 +976,41 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
>>>
>>>         switch (ios->power_mode) {
>>>         case MMC_POWER_UP:
>>> +               if ((!IS_ERR(mmc->supply.vmmc)) &&
>>> +                               !test_bit(DW_MMC_CARD_POWERED, &slot->flags)) {
>>> +                       ret = regulator_enable(mmc->supply.vmmc);
>>> +                       if (!ret)
>>> +                               set_bit(DW_MMC_CARD_POWERED, &slot->flags);
>>> +               }
>>
>> As per below, I'm not sure why you'd want to turn on vqmmc and vmmc at
>> different times.
> As you can see people's have different opinion on this.When i had a
> look at the other drivers in the subsystem which does in the same flow
> as above.However i will change in the next version.

Given my self proclaimed lack of SD/MMC knowledge, if others have a
good reason for doing them separate then you should do it that way.
So far I haven't heard that reason but I certainly could be wrong.


>>> @@ -225,6 +225,8 @@ struct dw_mci_slot {
>>>         unsigned long           flags;
>>>  #define DW_MMC_CARD_PRESENT    0
>>>  #define DW_MMC_CARD_NEED_INIT  1
>>> +#define DW_MMC_CARD_POWERED    2
>>> +#define DW_MMC_IO_POWERED      3
>>
>> I don't really think you should have two bits here.  From my
>> understanding of SD cards there should be very little reason to have
>> vqmmc and vmmc not powered at the same time.
> I think if i can use mmc_regulator_set_ocr(), we don't need additional
> flag.But for tps65090 mmc_regulator_get_ocr() and
> mmc_regulator_set_ocr() is failing as its a fixed-regulator.

Can you explain more about what's failing?  It sure looks like mmc
core is supposed to handle this given comments below

/*
* If we're using a fixed/static regulator, don't call
* regulator_set_voltage; it would fail.
*/
voltage = regulator_get_voltage(supply);

if (!regulator_can_change_voltage(supply))
  min_uV = max_uV = voltage;


-Doug

[PATCH 1/3] mmc: dw_mmc: use mmc_regulator_get_supply to handle regulators

[Yuvaraj Kumar]

On Thu, Jun 26, 2014 at 9:48 PM, Doug Anderson <dianders@chromium.org> wrote:
> Yuvaraj,
>
> On Thu, Jun 26, 2014 at 4:21 AM, Yuvaraj Kumar <yuvaraj.cd@gmail.com> wrote:
>> Doug
>>
>> On Tue, Jun 24, 2014 at 11:30 PM, Doug Anderson <dianders@chromium.org> wrote:
>>> Yuvaraj,
>>>
>>> On Mon, Jun 23, 2014 at 3:45 AM, Yuvaraj Kumar C D <yuvaraj.cd@gmail.com> wrote:
>>>> This patch makes use of mmc_regulator_get_supply() to handle
>>>> the vmmc and vqmmc regulators.Also it moves the code handling
>>>> the these regulators to dw_mci_set_ios().It turned on the vmmc
>>>> and vqmmc during MMC_POWER_UP and MMC_POWER_ON,and turned off
>>>> during MMC_POWER_OFF.
>>>>
>>>> Signed-off-by: Yuvaraj Kumar C D <yuvaraj.cd@samsung.com>
>>>> ---
>>>>  drivers/mmc/host/dw_mmc.c |   71 ++++++++++++++++++++++-----------------------
>>>>  drivers/mmc/host/dw_mmc.h |    2 ++
>>>>  2 files changed, 36 insertions(+), 37 deletions(-)
>>>
>>> Perhaps you could CC me on the whole series for the next version since
>>> I was involved in privately reviewing previous versions?
>> It was just accidental missing you in the CC .Surely i will add you in
>> CC for next versions.
>>>
>>> Overall caveat for my review is that I'm nowhere near an SD/MMC expert.
>>>
>>>
>>>> diff --git a/drivers/mmc/host/dw_mmc.c b/drivers/mmc/host/dw_mmc.c
>>>> index 1ac227c..f5cabce 100644
>>>> --- a/drivers/mmc/host/dw_mmc.c
>>>> +++ b/drivers/mmc/host/dw_mmc.c
>>>> @@ -937,6 +937,7 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
>>>>         struct dw_mci_slot *slot = mmc_priv(mmc);
>>>>         const struct dw_mci_drv_data *drv_data = slot->host->drv_data;
>>>>         u32 regs;
>>>> +       int ret;
>>>>
>>>>         switch (ios->bus_width) {
>>>>         case MMC_BUS_WIDTH_4:
>>>> @@ -975,16 +976,41 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
>>>>
>>>>         switch (ios->power_mode) {
>>>>         case MMC_POWER_UP:
>>>> +               if ((!IS_ERR(mmc->supply.vmmc)) &&
>>>> +                               !test_bit(DW_MMC_CARD_POWERED, &slot->flags)) {
>>>> +                       ret = regulator_enable(mmc->supply.vmmc);
>>>> +                       if (!ret)
>>>> +                               set_bit(DW_MMC_CARD_POWERED, &slot->flags);
>>>> +               }
>>>
>>> As per below, I'm not sure why you'd want to turn on vqmmc and vmmc at
>>> different times.
>> As you can see people's have different opinion on this.When i had a
>> look at the other drivers in the subsystem which does in the same flow
>> as above.However i will change in the next version.
>
> Given my self proclaimed lack of SD/MMC knowledge, if others have a
> good reason for doing them separate then you should do it that way.
> So far I haven't heard that reason but I certainly could be wrong.
>
>
>>>> @@ -225,6 +225,8 @@ struct dw_mci_slot {
>>>>         unsigned long           flags;
>>>>  #define DW_MMC_CARD_PRESENT    0
>>>>  #define DW_MMC_CARD_NEED_INIT  1
>>>> +#define DW_MMC_CARD_POWERED    2
>>>> +#define DW_MMC_IO_POWERED      3
>>>
>>> I don't really think you should have two bits here.  From my
>>> understanding of SD cards there should be very little reason to have
>>> vqmmc and vmmc not powered at the same time.
>> I think if i can use mmc_regulator_set_ocr(), we don't need additional
>> flag.But for tps65090 mmc_regulator_get_ocr() and
>> mmc_regulator_set_ocr() is failing as its a fixed-regulator.
>
> Can you explain more about what's failing?  It sure looks like mmc
> core is supposed to handle this given comments below
>
> /*
> * If we're using a fixed/static regulator, don't call
> * regulator_set_voltage; it would fail.
> */
tps65090 driver does not register through fixed regulator framework.It
uses normal regulator framework and supports only
enable/disable/is_enabled callbacks.Also it lacks certain callbacks
for list_voltage, get_voltage ,set_voltage etc.
[    2.306476] dwmmc_exynos 12220000.mmc: Failed getting OCR mask: -22
[    2.393403] dwmmc_exynos 12220000.mmc: could not set regulator OCR (-22)
For the above reason,regulator framework treats fet4 as unused
regulator and disables the vmmc regulator.
> voltage = regulator_get_voltage(supply);
>
> if (!regulator_can_change_voltage(supply))
>   min_uV = max_uV = voltage;
>
>
> -Doug

[PATCH 1/3] mmc: dw_mmc: use mmc_regulator_get_supply to handle regulators

[Doug Anderson]

Yuvaraj,

On Fri, Jun 27, 2014 at 3:59 AM, Yuvaraj Kumar <yuvaraj.cd@gmail.com> wrote:
>>> mmc_regulator_set_ocr() is failing as its a fixed-regulator.
>>
>> Can you explain more about what's failing?  It sure looks like mmc
>> core is supposed to handle this given comments below
>>
>> /*
>> * If we're using a fixed/static regulator, don't call
>> * regulator_set_voltage; it would fail.
>> */
> tps65090 driver does not register through fixed regulator framework.It
> uses normal regulator framework and supports only
> enable/disable/is_enabled callbacks.Also it lacks certain callbacks
> for list_voltage, get_voltage ,set_voltage etc.
> [    2.306476] dwmmc_exynos 12220000.mmc: Failed getting OCR mask: -22
> [    2.393403] dwmmc_exynos 12220000.mmc: could not set regulator OCR (-22)
> For the above reason,regulator framework treats fet4 as unused
> regulator and disables the vmmc regulator.

Ah.  Perhaps tps65090 needs to be fixed then...  I'm not sure any
other drivers cared before so maybe that's why it was never caught?

-Doug

[PATCH 1/3] mmc: dw_mmc: use mmc_regulator_get_supply to handle regulators

[Javier Martinez Canillas]

Hello Yuvaraj,

On Sat, Jun 28, 2014 at 12:47 AM, Doug Anderson <dianders@chromium.org> wrote:
> On Fri, Jun 27, 2014 at 3:59 AM, Yuvaraj Kumar <yuvaraj.cd@gmail.com> wrote:
>>>> mmc_regulator_set_ocr() is failing as its a fixed-regulator.
>>>
>>> Can you explain more about what's failing?  It sure looks like mmc
>>> core is supposed to handle this given comments below
>>>
>>> /*
>>> * If we're using a fixed/static regulator, don't call
>>> * regulator_set_voltage; it would fail.
>>> */
>> tps65090 driver does not register through fixed regulator framework.It
>> uses normal regulator framework and supports only
>> enable/disable/is_enabled callbacks.Also it lacks certain callbacks
>> for list_voltage, get_voltage ,set_voltage etc.
>> [    2.306476] dwmmc_exynos 12220000.mmc: Failed getting OCR mask: -22
>> [    2.393403] dwmmc_exynos 12220000.mmc: could not set regulator OCR (-22)
>> For the above reason,regulator framework treats fet4 as unused
>> regulator and disables the vmmc regulator.
>
> Ah.  Perhaps tps65090 needs to be fixed then...  I'm not sure any
> other drivers cared before so maybe that's why it was never caught?
>

I've been looking at this and as Doug and you said it seems that
nobody cared and since the tps65090 fets are actually just load
switches the driver only implementes the .enable and .disable
functions handlers. Also since their output voltage is just equal to
their input supply, that information was not present neither in the
driver nor in the Device Tree.

But when fet4 is used as the vmmc-supply phandle,
mmc_regulator_get_supply() calls mmc_regulator_get_ocrmask() [0] which
expects to fetch the regulator voltage count and current voltage as
you had explained so the function was failing.

Now I don't think that the driver needs to implement the
{get,set,list}_voltage callbacks. If a tps65090 regulator has both
regulator-min-microvolt and regulator-max-microvolt properties from
the generic regulator DT binding, and the value is the same, it can be
assumed that the regulator is fixed and the fixed_uV field can be set
to the output voltage and n_voltages to 1. That's everything that the
regulator core needs from a fixed regulator in order to make
regulator_get_voltage() to succeed:

static int _regulator_get_voltage(struct regulator_dev *rdev)
{
...
    if (rdev->desc->ops->get_voltage_sel) {
        sel = rdev->desc->ops->get_voltage_sel(rdev);
        if (sel < 0)
            return sel;
        ret = rdev->desc->ops->list_voltage(rdev, sel);
    } else if (rdev->desc->ops->get_voltage) {
        ret = rdev->desc->ops->get_voltage(rdev);
    } else if (rdev->desc->ops->list_voltage) {
        ret = rdev->desc->ops->list_voltage(rdev, 0);
    } else if (rdev->desc->fixed_uV && (rdev->desc->n_voltages == 1)) {
        ret = rdev->desc->fixed_uV;
    } else {
        return -EINVAL;
    }
...
}

What do you think about patch [1]?

So, with that patch mmc_regulator_get_supply() does not fail anymore
and also does not break DT backward compatibility since it only has
effect on the regulators that define their min and max voltage and not
in the ones that don't like is the case in old DTB.

Now, it is an RFC since I don't know if this is the correct solution.
Even though I've seen that min == max seems to imply that the
regulator is fixed, I wonder if is better to have an explicit generic
DT property (regulator-fixed-microvolt?) that is used to set fixed_uV
in of_get_regulation_constraints(). Since this seems to be a general
problem and not only related to the tps65090 device or the Peach
Pit/Pi use case. If you agree that it's a good solution then I can
post it as a proper patch.

I've also pushed this patch and the ones adding the needed DTS bits
for Peach Pit and Pi DTS to my personal repository so you can test it.
The branch is:

http://cgit.collabora.com/git/user/javier/linux.git/log/?h=tps65090-fixes

You also need to replace regulator_enable(mmc->supply.vmmc) and
regulator_disable(mmc->supply.vmmc) by mmc_regulator_set_ocr(mmc,
mmc->supply.vmmc, ios->vdd) and mmc_regulator_set_ocr(mmc,
mmc->supply.vmmc, 0) respectively in your original patch.

Best regards,
Javier

[0]: http://lxr.free-electrons.com/source/drivers/mmc/core/core.c#L1215
[1]:
>From 3d2e6079cc8c372da946d430d43d36af99e7a582 Mon Sep 17 00:00:00 2001
From: Javier Martinez Canillas <javier.martinez@collabora.co.uk>
Date: Tue, 22 Jul 2014 19:16:47 +0200
Subject: [RFC] regulator: tps65090: Allow regulators voltage to be
 queried

The tps65090 device has some regulators with a fixed output
voltage and others with an adjustable output voltage. But the
regulators with adjustable output just provide the voltage of
their input supply so they really are fixed regulators within
a supported voltage range that depends on their input voltage.

That is why the driver only provides an .enable and .disable
function handlers and not a .{get,set,list}_voltage handlers.

But there is code in the kernel that expects to query the
output voltage for all regulators. For example the function
mmc_regulator_get_ocrmask() is executed if a regulator is
used as a vmmc supply and this functions tries to fetch the
number of voltages supported by the regulator and its current
voltage but fails since the driver does not provide this data.

The .{list,get}_voltage function handler are actually not even
needed for fixed regulators since in this case the regulator
core just need a fixed voltage and a single voltage selector.

So, let's allow output voltage to be defined using the generic
regulator-min-microvolt and regulator-max-microvolt  properties
and use the same semantic for fixed regulator as explained in
Documentation/devicetree/bindings/regulator/fixed-regulator.txt.
Where having an equal value for min and max microvolt means that
the regulator is fixed and has a single voltage selector.

Signed-off-by: Javier Martinez Canillas <javier.martinez@collabora.co.uk>
---
 drivers/regulator/tps65090-regulator.c | 14 ++++++++++++++
 1 file changed, 14 insertions(+)

diff --git a/drivers/regulator/tps65090-regulator.c
b/drivers/regulator/tps65090-regulator.c
index 2064b3f..673948d 100644
--- a/drivers/regulator/tps65090-regulator.c
+++ b/drivers/regulator/tps65090-regulator.c
@@ -408,6 +408,7 @@ static int tps65090_regulator_probe(struct
platform_device *pdev)
     struct of_regulator_match *tps65090_reg_matches = NULL;
     int num;
     int ret;
+    int min_uV, max_uV;

     dev_dbg(&pdev->dev, "Probing regulator\n");

@@ -435,6 +436,19 @@ static int tps65090_regulator_probe(struct
platform_device *pdev)
             ri->overcurrent_wait_valid =
                 tps_pdata->overcurrent_wait_valid;
             ri->overcurrent_wait = tps_pdata->overcurrent_wait;
+
+            min_uV = tps_pdata->reg_init_data->constraints.min_uV;
+            max_uV = tps_pdata->reg_init_data->constraints.max_uV;
+
+            /*
+             * A regulator whose regulator-min-microvolt and
+             * regulator-max-microvolt properties are the same
+             * is a fixed regulator and has a single voltage rail.
+             */
+            if (min_uV && max_uV && min_uV == max_uV) {
+                ri->desc->fixed_uV = min_uV;
+                ri->desc->n_voltages = 1;
+            }
         }

         /*
-- 
2.0.0.rc2

[PATCH 1/3] mmc: dw_mmc: use mmc_regulator_get_supply to handle regulators

[Jaehoon Chung]

On 06/27/2014 01:18 AM, Doug Anderson wrote:
> Yuvaraj,
> 
> On Thu, Jun 26, 2014 at 4:21 AM, Yuvaraj Kumar <yuvaraj.cd@gmail.com> wrote:
>> Doug
>>
>> On Tue, Jun 24, 2014 at 11:30 PM, Doug Anderson <dianders@chromium.org> wrote:
>>> Yuvaraj,
>>>
>>> On Mon, Jun 23, 2014 at 3:45 AM, Yuvaraj Kumar C D <yuvaraj.cd@gmail.com> wrote:
>>>> This patch makes use of mmc_regulator_get_supply() to handle
>>>> the vmmc and vqmmc regulators.Also it moves the code handling
>>>> the these regulators to dw_mci_set_ios().It turned on the vmmc
>>>> and vqmmc during MMC_POWER_UP and MMC_POWER_ON,and turned off
>>>> during MMC_POWER_OFF.
>>>>
>>>> Signed-off-by: Yuvaraj Kumar C D <yuvaraj.cd@samsung.com>
>>>> ---
>>>>  drivers/mmc/host/dw_mmc.c |   71 ++++++++++++++++++++++-----------------------
>>>>  drivers/mmc/host/dw_mmc.h |    2 ++
>>>>  2 files changed, 36 insertions(+), 37 deletions(-)
>>>
>>> Perhaps you could CC me on the whole series for the next version since
>>> I was involved in privately reviewing previous versions?
>> It was just accidental missing you in the CC .Surely i will add you in
>> CC for next versions.
>>>
>>> Overall caveat for my review is that I'm nowhere near an SD/MMC expert.
>>>
>>>
>>>> diff --git a/drivers/mmc/host/dw_mmc.c b/drivers/mmc/host/dw_mmc.c
>>>> index 1ac227c..f5cabce 100644
>>>> --- a/drivers/mmc/host/dw_mmc.c
>>>> +++ b/drivers/mmc/host/dw_mmc.c
>>>> @@ -937,6 +937,7 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
>>>>         struct dw_mci_slot *slot = mmc_priv(mmc);
>>>>         const struct dw_mci_drv_data *drv_data = slot->host->drv_data;
>>>>         u32 regs;
>>>> +       int ret;
>>>>
>>>>         switch (ios->bus_width) {
>>>>         case MMC_BUS_WIDTH_4:
>>>> @@ -975,16 +976,41 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
>>>>
>>>>         switch (ios->power_mode) {
>>>>         case MMC_POWER_UP:
>>>> +               if ((!IS_ERR(mmc->supply.vmmc)) &&
>>>> +                               !test_bit(DW_MMC_CARD_POWERED, &slot->flags)) {
>>>> +                       ret = regulator_enable(mmc->supply.vmmc);
>>>> +                       if (!ret)
>>>> +                               set_bit(DW_MMC_CARD_POWERED, &slot->flags);
>>>> +               }
>>>
>>> As per below, I'm not sure why you'd want to turn on vqmmc and vmmc at
>>> different times.
>> As you can see people's have different opinion on this.When i had a
>> look at the other drivers in the subsystem which does in the same flow
>> as above.However i will change in the next version.
> 
> Given my self proclaimed lack of SD/MMC knowledge, if others have a
> good reason for doing them separate then you should do it that way.
> So far I haven't heard that reason but I certainly could be wrong.

At first time, i had believed nothing problem that it turns on vmmc and vqmmc at different time.
It could have the potential problem.

Best Regards,
Jaehoon Chung
> 
> 
>>>> @@ -225,6 +225,8 @@ struct dw_mci_slot {
>>>>         unsigned long           flags;
>>>>  #define DW_MMC_CARD_PRESENT    0
>>>>  #define DW_MMC_CARD_NEED_INIT  1
>>>> +#define DW_MMC_CARD_POWERED    2
>>>> +#define DW_MMC_IO_POWERED      3
>>>
>>> I don't really think you should have two bits here.  From my
>>> understanding of SD cards there should be very little reason to have
>>> vqmmc and vmmc not powered at the same time.
>> I think if i can use mmc_regulator_set_ocr(), we don't need additional
>> flag.But for tps65090 mmc_regulator_get_ocr() and
>> mmc_regulator_set_ocr() is failing as its a fixed-regulator.
> 
> Can you explain more about what's failing?  It sure looks like mmc
> core is supposed to handle this given comments below
> 
> /*
> * If we're using a fixed/static regulator, don't call
> * regulator_set_voltage; it would fail.
> */
> voltage = regulator_get_voltage(supply);
> 
> if (!regulator_can_change_voltage(supply))
>   min_uV = max_uV = voltage;
> 
> 
> -Doug
> 

[PATCH 1/3] mmc: dw_mmc: use mmc_regulator_get_supply to handle regulators

[Doug Anderson]

Jaehoon,

On Mon, Jun 30, 2014 at 5:13 AM, Jaehoon Chung <jh80.chung@samsung.com> wrote:
> On 06/27/2014 01:18 AM, Doug Anderson wrote:
>> Yuvaraj,
>>
>> On Thu, Jun 26, 2014 at 4:21 AM, Yuvaraj Kumar <yuvaraj.cd@gmail.com> wrote:
>>> Doug
>>>
>>> On Tue, Jun 24, 2014 at 11:30 PM, Doug Anderson <dianders@chromium.org> wrote:
>>>> Yuvaraj,
>>>>
>>>> On Mon, Jun 23, 2014 at 3:45 AM, Yuvaraj Kumar C D <yuvaraj.cd@gmail.com> wrote:
>>>>> This patch makes use of mmc_regulator_get_supply() to handle
>>>>> the vmmc and vqmmc regulators.Also it moves the code handling
>>>>> the these regulators to dw_mci_set_ios().It turned on the vmmc
>>>>> and vqmmc during MMC_POWER_UP and MMC_POWER_ON,and turned off
>>>>> during MMC_POWER_OFF.
>>>>>
>>>>> Signed-off-by: Yuvaraj Kumar C D <yuvaraj.cd@samsung.com>
>>>>> ---
>>>>>  drivers/mmc/host/dw_mmc.c |   71 ++++++++++++++++++++++-----------------------
>>>>>  drivers/mmc/host/dw_mmc.h |    2 ++
>>>>>  2 files changed, 36 insertions(+), 37 deletions(-)
>>>>
>>>> Perhaps you could CC me on the whole series for the next version since
>>>> I was involved in privately reviewing previous versions?
>>> It was just accidental missing you in the CC .Surely i will add you in
>>> CC for next versions.
>>>>
>>>> Overall caveat for my review is that I'm nowhere near an SD/MMC expert.
>>>>
>>>>
>>>>> diff --git a/drivers/mmc/host/dw_mmc.c b/drivers/mmc/host/dw_mmc.c
>>>>> index 1ac227c..f5cabce 100644
>>>>> --- a/drivers/mmc/host/dw_mmc.c
>>>>> +++ b/drivers/mmc/host/dw_mmc.c
>>>>> @@ -937,6 +937,7 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
>>>>>         struct dw_mci_slot *slot = mmc_priv(mmc);
>>>>>         const struct dw_mci_drv_data *drv_data = slot->host->drv_data;
>>>>>         u32 regs;
>>>>> +       int ret;
>>>>>
>>>>>         switch (ios->bus_width) {
>>>>>         case MMC_BUS_WIDTH_4:
>>>>> @@ -975,16 +976,41 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
>>>>>
>>>>>         switch (ios->power_mode) {
>>>>>         case MMC_POWER_UP:
>>>>> +               if ((!IS_ERR(mmc->supply.vmmc)) &&
>>>>> +                               !test_bit(DW_MMC_CARD_POWERED, &slot->flags)) {
>>>>> +                       ret = regulator_enable(mmc->supply.vmmc);
>>>>> +                       if (!ret)
>>>>> +                               set_bit(DW_MMC_CARD_POWERED, &slot->flags);
>>>>> +               }
>>>>
>>>> As per below, I'm not sure why you'd want to turn on vqmmc and vmmc at
>>>> different times.
>>> As you can see people's have different opinion on this.When i had a
>>> look at the other drivers in the subsystem which does in the same flow
>>> as above.However i will change in the next version.
>>
>> Given my self proclaimed lack of SD/MMC knowledge, if others have a
>> good reason for doing them separate then you should do it that way.
>> So far I haven't heard that reason but I certainly could be wrong.
>
> At first time, i had believed nothing problem that it turns on vmmc and vqmmc at different time.
> It could have the potential problem.

OK.  As I said I'm nowhere near an expert on SD/MMC, so if there's
something out there that says that turning vmmc on before vqmmc is the
right thing to do then I guess that's the answer.  I would still be
very curious to get more details on what the potential problem is.
Could you provide a reference to documentation that says vmmc should
be on before vqmmc or describe a situation where it's important?

Thanks!

-Doug

[PATCH 1/3] mmc: dw_mmc: use mmc_regulator_get_supply to handle regulators

[Jaehoon Chung]

Hi Doug.

On 07/01/2014 12:06 AM, Doug Anderson wrote:
> Jaehoon,
> 
> On Mon, Jun 30, 2014 at 5:13 AM, Jaehoon Chung <jh80.chung@samsung.com> wrote:
>> On 06/27/2014 01:18 AM, Doug Anderson wrote:
>>> Yuvaraj,
>>>
>>> On Thu, Jun 26, 2014 at 4:21 AM, Yuvaraj Kumar <yuvaraj.cd@gmail.com> wrote:
>>>> Doug
>>>>
>>>> On Tue, Jun 24, 2014 at 11:30 PM, Doug Anderson <dianders@chromium.org> wrote:
>>>>> Yuvaraj,
>>>>>
>>>>> On Mon, Jun 23, 2014 at 3:45 AM, Yuvaraj Kumar C D <yuvaraj.cd@gmail.com> wrote:
>>>>>> This patch makes use of mmc_regulator_get_supply() to handle
>>>>>> the vmmc and vqmmc regulators.Also it moves the code handling
>>>>>> the these regulators to dw_mci_set_ios().It turned on the vmmc
>>>>>> and vqmmc during MMC_POWER_UP and MMC_POWER_ON,and turned off
>>>>>> during MMC_POWER_OFF.
>>>>>>
>>>>>> Signed-off-by: Yuvaraj Kumar C D <yuvaraj.cd@samsung.com>
>>>>>> ---
>>>>>>  drivers/mmc/host/dw_mmc.c |   71 ++++++++++++++++++++++-----------------------
>>>>>>  drivers/mmc/host/dw_mmc.h |    2 ++
>>>>>>  2 files changed, 36 insertions(+), 37 deletions(-)
>>>>>
>>>>> Perhaps you could CC me on the whole series for the next version since
>>>>> I was involved in privately reviewing previous versions?
>>>> It was just accidental missing you in the CC .Surely i will add you in
>>>> CC for next versions.
>>>>>
>>>>> Overall caveat for my review is that I'm nowhere near an SD/MMC expert.
>>>>>
>>>>>
>>>>>> diff --git a/drivers/mmc/host/dw_mmc.c b/drivers/mmc/host/dw_mmc.c
>>>>>> index 1ac227c..f5cabce 100644
>>>>>> --- a/drivers/mmc/host/dw_mmc.c
>>>>>> +++ b/drivers/mmc/host/dw_mmc.c
>>>>>> @@ -937,6 +937,7 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
>>>>>>         struct dw_mci_slot *slot = mmc_priv(mmc);
>>>>>>         const struct dw_mci_drv_data *drv_data = slot->host->drv_data;
>>>>>>         u32 regs;
>>>>>> +       int ret;
>>>>>>
>>>>>>         switch (ios->bus_width) {
>>>>>>         case MMC_BUS_WIDTH_4:
>>>>>> @@ -975,16 +976,41 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
>>>>>>
>>>>>>         switch (ios->power_mode) {
>>>>>>         case MMC_POWER_UP:
>>>>>> +               if ((!IS_ERR(mmc->supply.vmmc)) &&
>>>>>> +                               !test_bit(DW_MMC_CARD_POWERED, &slot->flags)) {
>>>>>> +                       ret = regulator_enable(mmc->supply.vmmc);
>>>>>> +                       if (!ret)
>>>>>> +                               set_bit(DW_MMC_CARD_POWERED, &slot->flags);
>>>>>> +               }
>>>>>
>>>>> As per below, I'm not sure why you'd want to turn on vqmmc and vmmc at
>>>>> different times.
>>>> As you can see people's have different opinion on this.When i had a
>>>> look at the other drivers in the subsystem which does in the same flow
>>>> as above.However i will change in the next version.
>>>
>>> Given my self proclaimed lack of SD/MMC knowledge, if others have a
>>> good reason for doing them separate then you should do it that way.
>>> So far I haven't heard that reason but I certainly could be wrong.
>>
>> At first time, i had believed nothing problem that it turns on vmmc and vqmmc at different time.
>> It could have the potential problem.
> 
> OK.  As I said I'm nowhere near an expert on SD/MMC, so if there's
> something out there that says that turning vmmc on before vqmmc is the
> right thing to do then I guess that's the answer.  I would still be
> very curious to get more details on what the potential problem is.
> Could you provide a reference to documentation that says vmmc should
> be on before vqmmc or describe a situation where it's important?

Actually i didn't have any documentation.
According to eMMC spec, vmmc and vqmmc should be used with same power supply.
In this case, it will turn on vmmc and vqmmc at same time.
If vqmmc is used with the I/O power supply, the sequence isn't important, 
but vmmc and vqmmc need to wait until stabling the power.

I didn't know Potential problem is what problem.
(Potential problem is mentioned from our H/W team member. H/W mis-operating?)
I didn't have the expert H/W knowledge.

Thanks!

Bset Regards,
Jaehoon Chung

> 
> Thanks!
> 
> -Doug
> --
> To unsubscribe from this list: send the line "unsubscribe linux-mmc" in
> the body of a message to majordomo at vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 

[PATCH 2/3] mmc: dw_mmc: Dont cut off vqmmc and vmmc

[Yuvaraj Kumar C D]

On exynos 5250 and 5420 based boards which uses built-in CD# line
for card detection.But unfortunately CD# line is on the same voltage
rails as of I/O voltage rails.When we cut off vqmmc,the consequent
card detection will break in these boards.

Also if we let alone the vqmmc turned on when vmmc turned off, the
card could have half way powered and this can damage the card.So
this patch adds a check so that, if the board used the built-in
card detection mechanism i.e through CDETECT, it will not turned
down vqmmc and vmmc both.

Signed-off-by: Yuvaraj Kumar C D <yuvaraj.cd@samsung.com>
---
 drivers/mmc/host/dw_mmc.c |   23 +++++++++++++++++++++++
 1 file changed, 23 insertions(+)

diff --git a/drivers/mmc/host/dw_mmc.c b/drivers/mmc/host/dw_mmc.c
index f5cabce..e034bce 100644
--- a/drivers/mmc/host/dw_mmc.c
+++ b/drivers/mmc/host/dw_mmc.c
@@ -932,6 +932,25 @@ static void dw_mci_request(struct mmc_host *mmc, struct mmc_request *mrq)
 	spin_unlock_bh(&host->lock);
 }
 
+/*
+ * some of the boards use controller CD line for card detection.Unfortunately
+ * CD line is bind to the same volatge domain as of the IO lines.If we turn off
+ * IO voltage domain, CD line wont work.
+ * Return true when controller CD line is used for card detection or return
+ * false.
+ */
+static bool dw_mci_builtin_cd(struct dw_mci_slot *slot)
+{
+	struct dw_mci_board *brd = slot->host->pdata;
+	struct mmc_host *mmc = slot->mmc;
+
+	if ((brd->quirks & DW_MCI_QUIRK_BROKEN_CARD_DETECTION) &&
+			!IS_ERR_VALUE(mmc_gpio_get_cd(mmc)))
+		return true;
+	else
+		return false;
+}
+
 static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
 {
 	struct dw_mci_slot *slot = mmc_priv(mmc);
@@ -988,6 +1007,10 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
 		mci_writel(slot->host, PWREN, regs);
 		break;
 	case MMC_POWER_OFF:
+		if (dw_mci_builtin_cd(slot) &&
+				!test_bit(DW_MMC_CARD_PRESENT, &slot->flags))
+			return;
+
 		if (!IS_ERR(mmc->supply.vqmmc) &&
 				test_bit(DW_MMC_IO_POWERED, &slot->flags)) {
 			ret = regulator_disable(mmc->supply.vqmmc);
-- 
1.7.10.4

[PATCH 2/3] mmc: dw_mmc: Dont cut off vqmmc and vmmc

[Doug Anderson]

Yuvaraj,

On Mon, Jun 23, 2014 at 3:45 AM, Yuvaraj Kumar C D <yuvaraj.cd@gmail.com> wrote:
> On exynos 5250 and 5420 based boards which uses built-in CD# line
> for card detection.But unfortunately CD# line is on the same voltage
> rails as of I/O voltage rails.When we cut off vqmmc,the consequent
> card detection will break in these boards.
>
> Also if we let alone the vqmmc turned on when vmmc turned off, the
> card could have half way powered and this can damage the card.So
> this patch adds a check so that, if the board used the built-in
> card detection mechanism i.e through CDETECT, it will not turned
> down vqmmc and vmmc both.
>
> Signed-off-by: Yuvaraj Kumar C D <yuvaraj.cd@samsung.com>
> ---
>  drivers/mmc/host/dw_mmc.c |   23 +++++++++++++++++++++++
>  1 file changed, 23 insertions(+)

Yes, but...

As I mentioned in our separate email thread about this you're now
preventing mmc_power_cycle() from working properly.

IMHO you need the patch I sent you back on April 24th (was it that
long ago?).  Due to the brokenness of exynos (and anyone else that
powers CD off of vqmmc) you need to extend the MMC core to
differentiate several different types of "power off":

* power off because no card is plugged in: you should keep your card on.
* power off because you're power cycling: you should power off your card.
* power off because the system is suspending: you should power off your card.

...the third bullet point is something I hadn't though of until just
now and probably isn't addressed in my old patch...


Also: as we've discussed privately: you could imagine someone
designing an exynos5-based board where they've put the CD on a
separate GPIO.  On a system like this then all these hacks won't be
necessary.

---

Verbatim from my email about this topic on May 28th:

There are two important cases to handle:

1. Properly power cycle both vmmc and vqmmc (at the same time!) in
mmc_power_cycle().

2. DON'T power off for mmc_power_off() unless it's part of
mmc_power_cycle().  Specifically note that mmc_power_off() is called
in a whole bunch of places other than mmc_power_cycle().  For
instance, if we fail to probe a card we'll call mmc_power_off().  All
of these _must_ not turn off power to card detect or else we won't be
able to see future card insertions.  Note that mmc_power_off() might
be called on its own even when a card is inserted.  Look at the end of
mmc_rescan_try_freq().  It will be called if there are errors
attaching a card.


Rules to always remember:

* On all boards you should turn on vmmc and vqmmc at the same time.
It's illegal to have vqmmc on without vmmc and not good to have vmmc
on without vqmmc.

* On exynos you must ensure that vqmmc is on at all times, except
during power cycling of a card.  If vqmmc is not on you can't detect
card insertions or removals since the card detect line won't be
powered.

* To handle errors, you must ensure that mmc_power_cycle() actually
cycles power to the card.

[PATCH 3/3] mmc: dw_mmc: Support voltage changes

[Yuvaraj Kumar C D]

From: Doug Anderson <dianders@chromium.org>

For UHS cards we need the ability to switch voltages from 3.3V to
1.8V.  Add support to the dw_mmc driver to handle this.  Note that
dw_mmc needs a little bit of extra code since the interface needs a
special bit programmed to the CMD register while CMD11 is progressing.
This means adding a few extra states to the state machine to track.

Signed-off-by: Doug Anderson <dianders@chromium.org>
Signed-off-by: Yuvaraj Kumar C D <yuvaraj.cd@samsung.com>

---
 drivers/mmc/host/dw_mmc.c  |  145 +++++++++++++++++++++++++++++++++++++++++---
 drivers/mmc/host/dw_mmc.h  |    5 +-
 include/linux/mmc/dw_mmc.h |    2 +
 3 files changed, 142 insertions(+), 10 deletions(-)

diff --git a/drivers/mmc/host/dw_mmc.c b/drivers/mmc/host/dw_mmc.c
index e034bce..38eb548 100644
--- a/drivers/mmc/host/dw_mmc.c
+++ b/drivers/mmc/host/dw_mmc.c
@@ -29,6 +29,7 @@
 #include <linux/irq.h>
 #include <linux/mmc/host.h>
 #include <linux/mmc/mmc.h>
+#include <linux/mmc/sd.h>
 #include <linux/mmc/sdio.h>
 #include <linux/mmc/dw_mmc.h>
 #include <linux/bitops.h>
@@ -235,10 +236,13 @@ err:
 }
 #endif /* defined(CONFIG_DEBUG_FS) */
 
+static void mci_send_cmd(struct dw_mci_slot *slot, u32 cmd, u32 arg);
+
 static u32 dw_mci_prepare_command(struct mmc_host *mmc, struct mmc_command *cmd)
 {
 	struct mmc_data	*data;
 	struct dw_mci_slot *slot = mmc_priv(mmc);
+	struct dw_mci *host = slot->host;
 	const struct dw_mci_drv_data *drv_data = slot->host->drv_data;
 	u32 cmdr;
 	cmd->error = -EINPROGRESS;
@@ -254,6 +258,32 @@ static u32 dw_mci_prepare_command(struct mmc_host *mmc, struct mmc_command *cmd)
 	else if (cmd->opcode != MMC_SEND_STATUS && cmd->data)
 		cmdr |= SDMMC_CMD_PRV_DAT_WAIT;
 
+	if (cmd->opcode == SD_SWITCH_VOLTAGE) {
+		u32 clk_en_a;
+
+		/* Special bit makes CMD11 not die */
+		cmdr |= SDMMC_CMD_VOLT_SWITCH;
+
+		/* Change state to continue to handle CMD11 weirdness */
+		WARN_ON(slot->host->state != STATE_SENDING_CMD);
+		slot->host->state = STATE_SENDING_CMD11;
+
+		/*
+		 * We need to disable clock stop while doing voltage switch
+		 * according to 7.4.1.2 Voltage Switch Normal Scenario.
+		 *
+		 * It's assumed that by the next time the CLKENA is updated
+		 * (when we set the clock next) that the voltage change will
+		 * be over, so we don't bother setting any bits to synchronize
+		 * with dw_mci_setup_bus().
+		 */
+		clk_en_a = mci_readl(host, CLKENA);
+		clk_en_a &= ~(SDMMC_CLKEN_LOW_PWR << slot->id);
+		mci_writel(host, CLKENA, clk_en_a);
+		mci_send_cmd(slot, SDMMC_CMD_UPD_CLK |
+			     SDMMC_CMD_PRV_DAT_WAIT, 0);
+	}
+
 	if (cmd->flags & MMC_RSP_PRESENT) {
 		/* We expect a response, so set this bit */
 		cmdr |= SDMMC_CMD_RESP_EXP;
@@ -776,11 +806,15 @@ static void dw_mci_setup_bus(struct dw_mci_slot *slot, bool force_clkinit)
 	unsigned int clock = slot->clock;
 	u32 div;
 	u32 clk_en_a;
+	u32 sdmmc_cmd_bits = SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT;
+
+	/* We must continue to set bit 28 in CMD until the change is complete */
+	if (host->state == STATE_WAITING_CMD11_DONE)
+		sdmmc_cmd_bits |= SDMMC_CMD_VOLT_SWITCH;
 
 	if (!clock) {
 		mci_writel(host, CLKENA, 0);
-		mci_send_cmd(slot,
-			     SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT, 0);
+		mci_send_cmd(slot, sdmmc_cmd_bits, 0);
 	} else if (clock != host->current_speed || force_clkinit) {
 		div = host->bus_hz / clock;
 		if (host->bus_hz % clock && host->bus_hz > clock)
@@ -804,15 +838,13 @@ static void dw_mci_setup_bus(struct dw_mci_slot *slot, bool force_clkinit)
 		mci_writel(host, CLKSRC, 0);
 
 		/* inform CIU */
-		mci_send_cmd(slot,
-			     SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT, 0);
+		mci_send_cmd(slot, sdmmc_cmd_bits, 0);
 
 		/* set clock to desired speed */
 		mci_writel(host, CLKDIV, div);
 
 		/* inform CIU */
-		mci_send_cmd(slot,
-			     SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT, 0);
+		mci_send_cmd(slot, sdmmc_cmd_bits, 0);
 
 		/* enable clock; only low power if no SDIO */
 		clk_en_a = SDMMC_CLKEN_ENABLE << slot->id;
@@ -821,8 +853,7 @@ static void dw_mci_setup_bus(struct dw_mci_slot *slot, bool force_clkinit)
 		mci_writel(host, CLKENA, clk_en_a);
 
 		/* inform CIU */
-		mci_send_cmd(slot,
-			     SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT, 0);
+		mci_send_cmd(slot, sdmmc_cmd_bits, 0);
 
 		/* keep the clock with reflecting clock dividor */
 		slot->__clk_old = clock << div;
@@ -898,6 +929,17 @@ static void dw_mci_queue_request(struct dw_mci *host, struct dw_mci_slot *slot,
 
 	slot->mrq = mrq;
 
+	if (host->state == STATE_WAITING_CMD11_DONE) {
+		dev_warn(&slot->mmc->class_dev,
+			 "Voltage change didn't complete\n");
+		/*
+		 * this case isn't expected to happen, so we can
+		 * either crash here or just try to continue on
+		 * in the closest possible state
+		 */
+		host->state = STATE_IDLE;
+	}
+
 	if (host->state == STATE_IDLE) {
 		host->state = STATE_SENDING_CMD;
 		dw_mci_start_request(host, slot);
@@ -993,6 +1035,9 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
 	/* Slot specific timing and width adjustment */
 	dw_mci_setup_bus(slot, false);
 
+	if (slot->host->state == STATE_WAITING_CMD11_DONE && ios->clock != 0)
+		slot->host->state = STATE_IDLE;
+
 	switch (ios->power_mode) {
 	case MMC_POWER_UP:
 		if ((!IS_ERR(mmc->supply.vmmc)) &&
@@ -1039,6 +1084,68 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
 	}
 }
 
+static int dw_mci_card_busy(struct mmc_host *mmc)
+{
+	struct dw_mci_slot *slot = mmc_priv(mmc);
+	u32 status;
+
+	/*
+	 * Check the busy bit which is low when DAT[3:0]
+	 * (the data lines) are 0000
+	 */
+	status = mci_readl(slot->host, STATUS);
+
+	return !!(status & SDMMC_STATUS_BUSY);
+}
+
+static int dw_mci_switch_voltage(struct mmc_host *mmc, struct mmc_ios *ios)
+{
+	struct dw_mci_slot *slot = mmc_priv(mmc);
+	struct dw_mci *host = slot->host;
+	u32 uhs;
+	u32 v18 = SDMMC_UHS_18V << slot->id;
+	int min_uv, max_uv;
+	int ret;
+
+	/*
+	 * Program the voltage.  Note that some instances of dw_mmc may use
+	 * the UHS_REG for this.  For other instances (like exynos) the UHS_REG
+	 * does no harm but you need to set the regulator directly.  Try both.
+	 */
+	uhs = mci_readl(host, UHS_REG);
+	if (ios->signal_voltage == MMC_SIGNAL_VOLTAGE_330) {
+		min_uv = 2700000;
+		max_uv = 3600000;
+		uhs &= ~v18;
+	} else {
+		min_uv = 1700000;
+		max_uv = 1950000;
+		uhs |= v18;
+	}
+	if (!IS_ERR(mmc->supply.vqmmc)) {
+		ret = regulator_set_voltage(mmc->supply.vqmmc, min_uv, max_uv);
+
+		/*
+		 * Only complain if regulator claims that it's not in the 1.8V
+		 * range.  This avoids a bunch of errors in the case that
+		 * we've got a fixed 1.8V regulator but the core SD code still
+		 * thinks it ought to try to switch to 3.3 and then back to 1.8
+		 */
+		if (ret) {
+			int cur_voltage = 0;
+
+			cur_voltage = regulator_get_voltage(mmc->supply.vqmmc);
+			if (cur_voltage < 1700000 || cur_voltage > 1950000)
+				dev_warn(&mmc->class_dev,
+					 "Regulator set error %d: %d - %d\n",
+					 ret, min_uv, max_uv);
+		}
+	}
+	mci_writel(host, UHS_REG, uhs);
+
+	return 0;
+}
+
 static int dw_mci_get_ro(struct mmc_host *mmc)
 {
 	int read_only;
@@ -1180,6 +1287,9 @@ static const struct mmc_host_ops dw_mci_ops = {
 	.get_cd			= dw_mci_get_cd,
 	.enable_sdio_irq	= dw_mci_enable_sdio_irq,
 	.execute_tuning		= dw_mci_execute_tuning,
+	.card_busy		= dw_mci_card_busy,
+	.start_signal_voltage_switch = dw_mci_switch_voltage,
+
 };
 
 static void dw_mci_request_end(struct dw_mci *host, struct mmc_request *mrq)
@@ -1203,7 +1313,11 @@ static void dw_mci_request_end(struct dw_mci *host, struct mmc_request *mrq)
 		dw_mci_start_request(host, slot);
 	} else {
 		dev_vdbg(host->dev, "list empty\n");
-		host->state = STATE_IDLE;
+
+		if (host->state == STATE_SENDING_CMD11)
+			host->state = STATE_WAITING_CMD11_DONE;
+		else
+			host->state = STATE_IDLE;
 	}
 
 	spin_unlock(&host->lock);
@@ -1314,8 +1428,10 @@ static void dw_mci_tasklet_func(unsigned long priv)
 
 		switch (state) {
 		case STATE_IDLE:
+		case STATE_WAITING_CMD11_DONE:
 			break;
 
+		case STATE_SENDING_CMD11:
 		case STATE_SENDING_CMD:
 			if (!test_and_clear_bit(EVENT_CMD_COMPLETE,
 						&host->pending_events))
@@ -1870,6 +1986,15 @@ static irqreturn_t dw_mci_interrupt(int irq, void *dev_id)
 	}
 
 	if (pending) {
+		/* Check volt switch first, since it can look like an error */
+		if ((host->state == STATE_SENDING_CMD11) &&
+		    (pending & SDMMC_INT_VOLT_SWITCH)) {
+			mci_writel(host, RINTSTS, SDMMC_INT_VOLT_SWITCH);
+			pending &= ~SDMMC_INT_VOLT_SWITCH;
+			dw_mci_cmd_interrupt(host,
+					     pending | SDMMC_INT_CMD_DONE);
+		}
+
 		if (pending & DW_MCI_CMD_ERROR_FLAGS) {
 			mci_writel(host, RINTSTS, DW_MCI_CMD_ERROR_FLAGS);
 			host->cmd_status = pending;
@@ -1975,7 +2100,9 @@ static void dw_mci_work_routine_card(struct work_struct *work)
 
 					switch (host->state) {
 					case STATE_IDLE:
+					case STATE_WAITING_CMD11_DONE:
 						break;
+					case STATE_SENDING_CMD11:
 					case STATE_SENDING_CMD:
 						mrq->cmd->error = -ENOMEDIUM;
 						if (!mrq->data)
diff --git a/drivers/mmc/host/dw_mmc.h b/drivers/mmc/host/dw_mmc.h
index 5c95d00..af1d35f 100644
--- a/drivers/mmc/host/dw_mmc.h
+++ b/drivers/mmc/host/dw_mmc.h
@@ -99,6 +99,7 @@
 #define SDMMC_INT_HLE			BIT(12)
 #define SDMMC_INT_FRUN			BIT(11)
 #define SDMMC_INT_HTO			BIT(10)
+#define SDMMC_INT_VOLT_SWITCH		BIT(10) /* overloads bit 10! */
 #define SDMMC_INT_DRTO			BIT(9)
 #define SDMMC_INT_RTO			BIT(8)
 #define SDMMC_INT_DCRC			BIT(7)
@@ -113,6 +114,7 @@
 /* Command register defines */
 #define SDMMC_CMD_START			BIT(31)
 #define SDMMC_CMD_USE_HOLD_REG	BIT(29)
+#define SDMMC_CMD_VOLT_SWITCH		BIT(28)
 #define SDMMC_CMD_CCS_EXP		BIT(23)
 #define SDMMC_CMD_CEATA_RD		BIT(22)
 #define SDMMC_CMD_UPD_CLK		BIT(21)
@@ -129,6 +131,7 @@
 #define SDMMC_CMD_INDX(n)		((n) & 0x1F)
 /* Status register defines */
 #define SDMMC_GET_FCNT(x)		(((x)>>17) & 0x1FFF)
+#define SDMMC_STATUS_BUSY		BIT(9)
 /* FIFOTH register defines */
 #define SDMMC_SET_FIFOTH(m, r, t)	(((m) & 0x7) << 28 | \
 					 ((r) & 0xFFF) << 16 | \
@@ -149,7 +152,7 @@
 #define SDMMC_GET_VERID(x)		((x) & 0xFFFF)
 /* Card read threshold */
 #define SDMMC_SET_RD_THLD(v, x)		(((v) & 0x1FFF) << 16 | (x))
-
+#define SDMMC_UHS_18V			BIT(0)
 /* Register access macros */
 #define mci_readl(dev, reg)			\
 	__raw_readl((dev)->regs + SDMMC_##reg)
diff --git a/include/linux/mmc/dw_mmc.h b/include/linux/mmc/dw_mmc.h
index babaea9..9e31564 100644
--- a/include/linux/mmc/dw_mmc.h
+++ b/include/linux/mmc/dw_mmc.h
@@ -26,6 +26,8 @@ enum dw_mci_state {
 	STATE_DATA_BUSY,
 	STATE_SENDING_STOP,
 	STATE_DATA_ERROR,
+	STATE_SENDING_CMD11,
+	STATE_WAITING_CMD11_DONE,
 };
 
 enum {
-- 
1.7.10.4

[PATCH 3/3] mmc: dw_mmc: Support voltage changes

[Doug Anderson]

Yuvaraj,

On Mon, Jun 23, 2014 at 3:45 AM, Yuvaraj Kumar C D <yuvaraj.cd@gmail.com> wrote:
> From: Doug Anderson <dianders@chromium.org>
>
> For UHS cards we need the ability to switch voltages from 3.3V to
> 1.8V.  Add support to the dw_mmc driver to handle this.  Note that
> dw_mmc needs a little bit of extra code since the interface needs a
> special bit programmed to the CMD register while CMD11 is progressing.
> This means adding a few extra states to the state machine to track.
>
> Signed-off-by: Doug Anderson <dianders@chromium.org>
> Signed-off-by: Yuvaraj Kumar C D <yuvaraj.cd@samsung.com>
>
> ---
>  drivers/mmc/host/dw_mmc.c  |  145 +++++++++++++++++++++++++++++++++++++++++---
>  drivers/mmc/host/dw_mmc.h  |    5 +-
>  include/linux/mmc/dw_mmc.h |    2 +
>  3 files changed, 142 insertions(+), 10 deletions(-)

I will note to other interested parties that in the ChromeOS tree we
never actually cleaned up the MMC voltage change stuff enough to make
use of this functionality to actually get UHS cards working (though
the code is exercised as part of WiFi).


> @@ -254,6 +258,32 @@ static u32 dw_mci_prepare_command(struct mmc_host *mmc, struct mmc_command *cmd)
>         else if (cmd->opcode != MMC_SEND_STATUS && cmd->data)
>                 cmdr |= SDMMC_CMD_PRV_DAT_WAIT;
>
> +       if (cmd->opcode == SD_SWITCH_VOLTAGE) {
> +               u32 clk_en_a;
> +
> +               /* Special bit makes CMD11 not die */
> +               cmdr |= SDMMC_CMD_VOLT_SWITCH;
> +
> +               /* Change state to continue to handle CMD11 weirdness */
> +               WARN_ON(slot->host->state != STATE_SENDING_CMD);
> +               slot->host->state = STATE_SENDING_CMD11;
> +
> +               /*
> +                * We need to disable clock stop while doing voltage switch
> +                * according to 7.4.1.2 Voltage Switch Normal Scenario.

I think upstream hasn't tended to like the chapter number references,
though Grant really liked them locally.

> +static int dw_mci_switch_voltage(struct mmc_host *mmc, struct mmc_ios *ios)
> +{
> +       struct dw_mci_slot *slot = mmc_priv(mmc);
> +       struct dw_mci *host = slot->host;
> +       u32 uhs;
> +       u32 v18 = SDMMC_UHS_18V << slot->id;
> +       int min_uv, max_uv;
> +       int ret;
> +
> +       /*
> +        * Program the voltage.  Note that some instances of dw_mmc may use
> +        * the UHS_REG for this.  For other instances (like exynos) the UHS_REG
> +        * does no harm but you need to set the regulator directly.  Try both.

I've only worked with exynos which doesn't use UHS_REG for anything,
so the setting of this bit was a best-guess.  If anyone actually had a
chip that uses it then please comment / test.

[PATCH 3/3] mmc: dw_mmc: Support voltage changes

[Seungwon Jeon]

On Mon, June 23, 2014, Yuvaraj Kumar C D wrote:
> Subject: [PATCH 3/3] mmc: dw_mmc: Support voltage changes
> 
> From: Doug Anderson <dianders@chromium.org>
> 
> For UHS cards we need the ability to switch voltages from 3.3V to
> 1.8V.  Add support to the dw_mmc driver to handle this.  Note that
> dw_mmc needs a little bit of extra code since the interface needs a
> special bit programmed to the CMD register while CMD11 is progressing.
> This means adding a few extra states to the state machine to track.

Overall new additional states makes it complicated.
Can we do that in other way?

> 
> Signed-off-by: Doug Anderson <dianders@chromium.org>
> Signed-off-by: Yuvaraj Kumar C D <yuvaraj.cd@samsung.com>
> 
> ---
>  drivers/mmc/host/dw_mmc.c  |  145 +++++++++++++++++++++++++++++++++++++++++---
>  drivers/mmc/host/dw_mmc.h  |    5 +-
>  include/linux/mmc/dw_mmc.h |    2 +
>  3 files changed, 142 insertions(+), 10 deletions(-)
> 
> diff --git a/drivers/mmc/host/dw_mmc.c b/drivers/mmc/host/dw_mmc.c
> index e034bce..38eb548 100644
> --- a/drivers/mmc/host/dw_mmc.c
> +++ b/drivers/mmc/host/dw_mmc.c
> @@ -29,6 +29,7 @@
>  #include <linux/irq.h>
>  #include <linux/mmc/host.h>
>  #include <linux/mmc/mmc.h>
> +#include <linux/mmc/sd.h>
>  #include <linux/mmc/sdio.h>
>  #include <linux/mmc/dw_mmc.h>
>  #include <linux/bitops.h>
> @@ -235,10 +236,13 @@ err:
>  }
>  #endif /* defined(CONFIG_DEBUG_FS) */
> 
> +static void mci_send_cmd(struct dw_mci_slot *slot, u32 cmd, u32 arg);
> +
>  static u32 dw_mci_prepare_command(struct mmc_host *mmc, struct mmc_command *cmd)
>  {
>  	struct mmc_data	*data;
>  	struct dw_mci_slot *slot = mmc_priv(mmc);
> +	struct dw_mci *host = slot->host;
>  	const struct dw_mci_drv_data *drv_data = slot->host->drv_data;
>  	u32 cmdr;
>  	cmd->error = -EINPROGRESS;
> @@ -254,6 +258,32 @@ static u32 dw_mci_prepare_command(struct mmc_host *mmc, struct mmc_command *cmd)
>  	else if (cmd->opcode != MMC_SEND_STATUS && cmd->data)
>  		cmdr |= SDMMC_CMD_PRV_DAT_WAIT;
> 
> +	if (cmd->opcode == SD_SWITCH_VOLTAGE) {
> +		u32 clk_en_a;
> +
> +		/* Special bit makes CMD11 not die */
> +		cmdr |= SDMMC_CMD_VOLT_SWITCH;
> +
> +		/* Change state to continue to handle CMD11 weirdness */
> +		WARN_ON(slot->host->state != STATE_SENDING_CMD);
> +		slot->host->state = STATE_SENDING_CMD11;
> +
> +		/*
> +		 * We need to disable clock stop while doing voltage switch
> +		 * according to 7.4.1.2 Voltage Switch Normal Scenario.
> +		 *
> +		 * It's assumed that by the next time the CLKENA is updated
> +		 * (when we set the clock next) that the voltage change will
> +		 * be over, so we don't bother setting any bits to synchronize
> +		 * with dw_mci_setup_bus().
> +		 */
> +		clk_en_a = mci_readl(host, CLKENA);
> +		clk_en_a &= ~(SDMMC_CLKEN_LOW_PWR << slot->id);
> +		mci_writel(host, CLKENA, clk_en_a);
> +		mci_send_cmd(slot, SDMMC_CMD_UPD_CLK |
> +			     SDMMC_CMD_PRV_DAT_WAIT, 0);
dw_mci_disable_low_power() can be used here.

> +	}
> +
>  	if (cmd->flags & MMC_RSP_PRESENT) {
>  		/* We expect a response, so set this bit */
>  		cmdr |= SDMMC_CMD_RESP_EXP;
> @@ -776,11 +806,15 @@ static void dw_mci_setup_bus(struct dw_mci_slot *slot, bool force_clkinit)
>  	unsigned int clock = slot->clock;
>  	u32 div;
>  	u32 clk_en_a;
> +	u32 sdmmc_cmd_bits = SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT;
> +
> +	/* We must continue to set bit 28 in CMD until the change is complete */
> +	if (host->state == STATE_WAITING_CMD11_DONE)
> +		sdmmc_cmd_bits |= SDMMC_CMD_VOLT_SWITCH;
I didn't get the reason SDMMC_CMD_VOLT_SWITCH is needed during clock update(enable/disable)
Can you explain it in details?

> 
>  	if (!clock) {
>  		mci_writel(host, CLKENA, 0);
> -		mci_send_cmd(slot,
> -			     SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT, 0);
> +		mci_send_cmd(slot, sdmmc_cmd_bits, 0);
>  	} else if (clock != host->current_speed || force_clkinit) {
>  		div = host->bus_hz / clock;
>  		if (host->bus_hz % clock && host->bus_hz > clock)
> @@ -804,15 +838,13 @@ static void dw_mci_setup_bus(struct dw_mci_slot *slot, bool force_clkinit)
>  		mci_writel(host, CLKSRC, 0);
> 
>  		/* inform CIU */
> -		mci_send_cmd(slot,
> -			     SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT, 0);
> +		mci_send_cmd(slot, sdmmc_cmd_bits, 0);
> 
>  		/* set clock to desired speed */
>  		mci_writel(host, CLKDIV, div);
> 
>  		/* inform CIU */
> -		mci_send_cmd(slot,
> -			     SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT, 0);
> +		mci_send_cmd(slot, sdmmc_cmd_bits, 0);
> 
>  		/* enable clock; only low power if no SDIO */
>  		clk_en_a = SDMMC_CLKEN_ENABLE << slot->id;
> @@ -821,8 +853,7 @@ static void dw_mci_setup_bus(struct dw_mci_slot *slot, bool force_clkinit)
>  		mci_writel(host, CLKENA, clk_en_a);
> 
>  		/* inform CIU */
> -		mci_send_cmd(slot,
> -			     SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT, 0);
> +		mci_send_cmd(slot, sdmmc_cmd_bits, 0);
> 
>  		/* keep the clock with reflecting clock dividor */
>  		slot->__clk_old = clock << div;
> @@ -898,6 +929,17 @@ static void dw_mci_queue_request(struct dw_mci *host, struct dw_mci_slot *slot,
> 
>  	slot->mrq = mrq;
> 
> +	if (host->state == STATE_WAITING_CMD11_DONE) {
> +		dev_warn(&slot->mmc->class_dev,
> +			 "Voltage change didn't complete\n");
> +		/*
> +		 * this case isn't expected to happen, so we can
> +		 * either crash here or just try to continue on
> +		 * in the closest possible state
> +		 */
> +		host->state = STATE_IDLE;
> +	}
> +
>  	if (host->state == STATE_IDLE) {
>  		host->state = STATE_SENDING_CMD;
>  		dw_mci_start_request(host, slot);
> @@ -993,6 +1035,9 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
>  	/* Slot specific timing and width adjustment */
>  	dw_mci_setup_bus(slot, false);
> 
> +	if (slot->host->state == STATE_WAITING_CMD11_DONE && ios->clock != 0)
> +		slot->host->state = STATE_IDLE;
> +
>  	switch (ios->power_mode) {
>  	case MMC_POWER_UP:
>  		if ((!IS_ERR(mmc->supply.vmmc)) &&
> @@ -1039,6 +1084,68 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
>  	}
>  }
> 
> +static int dw_mci_card_busy(struct mmc_host *mmc)
> +{
> +	struct dw_mci_slot *slot = mmc_priv(mmc);
> +	u32 status;
> +
> +	/*
> +	 * Check the busy bit which is low when DAT[3:0]
> +	 * (the data lines) are 0000
> +	 */
> +	status = mci_readl(slot->host, STATUS);
> +
> +	return !!(status & SDMMC_STATUS_BUSY);
> +}
> +
> +static int dw_mci_switch_voltage(struct mmc_host *mmc, struct mmc_ios *ios)
> +{
> +	struct dw_mci_slot *slot = mmc_priv(mmc);
> +	struct dw_mci *host = slot->host;
> +	u32 uhs;
> +	u32 v18 = SDMMC_UHS_18V << slot->id;
> +	int min_uv, max_uv;
> +	int ret;
> +
> +	/*
> +	 * Program the voltage.  Note that some instances of dw_mmc may use
> +	 * the UHS_REG for this.  For other instances (like exynos) the UHS_REG
> +	 * does no harm but you need to set the regulator directly.  Try both.
> +	 */
> +	uhs = mci_readl(host, UHS_REG);
> +	if (ios->signal_voltage == MMC_SIGNAL_VOLTAGE_330) {
> +		min_uv = 2700000;
> +		max_uv = 3600000;
> +		uhs &= ~v18;
> +	} else {
> +		min_uv = 1700000;
> +		max_uv = 1950000;
> +		uhs |= v18;
> +	}
> +	if (!IS_ERR(mmc->supply.vqmmc)) {
> +		ret = regulator_set_voltage(mmc->supply.vqmmc, min_uv, max_uv);
> +
> +		/*
> +		 * Only complain if regulator claims that it's not in the 1.8V
> +		 * range.  This avoids a bunch of errors in the case that
> +		 * we've got a fixed 1.8V regulator but the core SD code still
> +		 * thinks it ought to try to switch to 3.3 and then back to 1.8
> +		 */
> +		if (ret) {
I think if ret is error, printing message and returning error is good.
Currently, just returning '0' though it fails.

> +			int cur_voltage = 0;
> +
> +			cur_voltage = regulator_get_voltage(mmc->supply.vqmmc);
> +			if (cur_voltage < 1700000 || cur_voltage > 1950000)
> +				dev_warn(&mmc->class_dev,
> +					 "Regulator set error %d: %d - %d\n",
> +					 ret, min_uv, max_uv);
> +		}
> +	}
> +	mci_writel(host, UHS_REG, uhs);
> +
> +	return 0;
> +}
> +
>  static int dw_mci_get_ro(struct mmc_host *mmc)
>  {
>  	int read_only;
> @@ -1180,6 +1287,9 @@ static const struct mmc_host_ops dw_mci_ops = {
>  	.get_cd			= dw_mci_get_cd,
>  	.enable_sdio_irq	= dw_mci_enable_sdio_irq,
>  	.execute_tuning		= dw_mci_execute_tuning,
> +	.card_busy		= dw_mci_card_busy,
> +	.start_signal_voltage_switch = dw_mci_switch_voltage,
> +
>  };
> 
>  static void dw_mci_request_end(struct dw_mci *host, struct mmc_request *mrq)
> @@ -1203,7 +1313,11 @@ static void dw_mci_request_end(struct dw_mci *host, struct mmc_request *mrq)
>  		dw_mci_start_request(host, slot);
>  	} else {
>  		dev_vdbg(host->dev, "list empty\n");
> -		host->state = STATE_IDLE;
> +
> +		if (host->state == STATE_SENDING_CMD11)
> +			host->state = STATE_WAITING_CMD11_DONE;
> +		else
> +			host->state = STATE_IDLE;
>  	}
> 
>  	spin_unlock(&host->lock);
> @@ -1314,8 +1428,10 @@ static void dw_mci_tasklet_func(unsigned long priv)
> 
>  		switch (state) {
>  		case STATE_IDLE:
> +		case STATE_WAITING_CMD11_DONE:
>  			break;
> 
> +		case STATE_SENDING_CMD11:
>  		case STATE_SENDING_CMD:
>  			if (!test_and_clear_bit(EVENT_CMD_COMPLETE,
>  						&host->pending_events))
> @@ -1870,6 +1986,15 @@ static irqreturn_t dw_mci_interrupt(int irq, void *dev_id)
>  	}
> 
>  	if (pending) {
> +		/* Check volt switch first, since it can look like an error */
> +		if ((host->state == STATE_SENDING_CMD11) &&
> +		    (pending & SDMMC_INT_VOLT_SWITCH)) {
> +			mci_writel(host, RINTSTS, SDMMC_INT_VOLT_SWITCH);
> +			pending &= ~SDMMC_INT_VOLT_SWITCH;
> +			dw_mci_cmd_interrupt(host,
> +					     pending | SDMMC_INT_CMD_DONE);
Artificially adding SDMMC_INT_CMD_DONE to pending is needed to complete cmd handling?
Is there any reason?

Thanks,
Seungwon Jeon

> +		}
> +
>  		if (pending & DW_MCI_CMD_ERROR_FLAGS) {
>  			mci_writel(host, RINTSTS, DW_MCI_CMD_ERROR_FLAGS);
>  			host->cmd_status = pending;
> @@ -1975,7 +2100,9 @@ static void dw_mci_work_routine_card(struct work_struct *work)
> 
>  					switch (host->state) {
>  					case STATE_IDLE:
> +					case STATE_WAITING_CMD11_DONE:
>  						break;
> +					case STATE_SENDING_CMD11:
>  					case STATE_SENDING_CMD:
>  						mrq->cmd->error = -ENOMEDIUM;
>  						if (!mrq->data)
> diff --git a/drivers/mmc/host/dw_mmc.h b/drivers/mmc/host/dw_mmc.h
> index 5c95d00..af1d35f 100644
> --- a/drivers/mmc/host/dw_mmc.h
> +++ b/drivers/mmc/host/dw_mmc.h
> @@ -99,6 +99,7 @@
>  #define SDMMC_INT_HLE			BIT(12)
>  #define SDMMC_INT_FRUN			BIT(11)
>  #define SDMMC_INT_HTO			BIT(10)
> +#define SDMMC_INT_VOLT_SWITCH		BIT(10) /* overloads bit 10! */
>  #define SDMMC_INT_DRTO			BIT(9)
>  #define SDMMC_INT_RTO			BIT(8)
>  #define SDMMC_INT_DCRC			BIT(7)
> @@ -113,6 +114,7 @@
>  /* Command register defines */
>  #define SDMMC_CMD_START			BIT(31)
>  #define SDMMC_CMD_USE_HOLD_REG	BIT(29)
> +#define SDMMC_CMD_VOLT_SWITCH		BIT(28)
>  #define SDMMC_CMD_CCS_EXP		BIT(23)
>  #define SDMMC_CMD_CEATA_RD		BIT(22)
>  #define SDMMC_CMD_UPD_CLK		BIT(21)
> @@ -129,6 +131,7 @@
>  #define SDMMC_CMD_INDX(n)		((n) & 0x1F)
>  /* Status register defines */
>  #define SDMMC_GET_FCNT(x)		(((x)>>17) & 0x1FFF)
> +#define SDMMC_STATUS_BUSY		BIT(9)
>  /* FIFOTH register defines */
>  #define SDMMC_SET_FIFOTH(m, r, t)	(((m) & 0x7) << 28 | \
>  					 ((r) & 0xFFF) << 16 | \
> @@ -149,7 +152,7 @@
>  #define SDMMC_GET_VERID(x)		((x) & 0xFFFF)
>  /* Card read threshold */
>  #define SDMMC_SET_RD_THLD(v, x)		(((v) & 0x1FFF) << 16 | (x))
> -
> +#define SDMMC_UHS_18V			BIT(0)
>  /* Register access macros */
>  #define mci_readl(dev, reg)			\
>  	__raw_readl((dev)->regs + SDMMC_##reg)
> diff --git a/include/linux/mmc/dw_mmc.h b/include/linux/mmc/dw_mmc.h
> index babaea9..9e31564 100644
> --- a/include/linux/mmc/dw_mmc.h
> +++ b/include/linux/mmc/dw_mmc.h
> @@ -26,6 +26,8 @@ enum dw_mci_state {
>  	STATE_DATA_BUSY,
>  	STATE_SENDING_STOP,
>  	STATE_DATA_ERROR,
> +	STATE_SENDING_CMD11,
> +	STATE_WAITING_CMD11_DONE,
>  };
> 
>  enum {
> --
> 1.7.10.4
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-mmc" in
> the body of a message to majordomo at vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 3/3] mmc: dw_mmc: Support voltage changes

[Doug Anderson]

Seungwon,

On Wed, Jun 25, 2014 at 6:08 AM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
> On Mon, June 23, 2014, Yuvaraj Kumar C D wrote:
>> Subject: [PATCH 3/3] mmc: dw_mmc: Support voltage changes
>>
>> From: Doug Anderson <dianders@chromium.org>
>>
>> For UHS cards we need the ability to switch voltages from 3.3V to
>> 1.8V.  Add support to the dw_mmc driver to handle this.  Note that
>> dw_mmc needs a little bit of extra code since the interface needs a
>> special bit programmed to the CMD register while CMD11 is progressing.
>> This means adding a few extra states to the state machine to track.
>
> Overall new additional states makes it complicated.
> Can we do that in other way?

That was the best I was able to figure out when I thought this
through.  If you have ideas for doing it another way I'd imagine that
Yuvaraj would be happy to take your feedback.


>> Signed-off-by: Doug Anderson <dianders@chromium.org>
>> Signed-off-by: Yuvaraj Kumar C D <yuvaraj.cd@samsung.com>
>>
>> ---
>>  drivers/mmc/host/dw_mmc.c  |  145 +++++++++++++++++++++++++++++++++++++++++---
>>  drivers/mmc/host/dw_mmc.h  |    5 +-
>>  include/linux/mmc/dw_mmc.h |    2 +
>>  3 files changed, 142 insertions(+), 10 deletions(-)
>>
>> diff --git a/drivers/mmc/host/dw_mmc.c b/drivers/mmc/host/dw_mmc.c
>> index e034bce..38eb548 100644
>> --- a/drivers/mmc/host/dw_mmc.c
>> +++ b/drivers/mmc/host/dw_mmc.c
>> @@ -29,6 +29,7 @@
>>  #include <linux/irq.h>
>>  #include <linux/mmc/host.h>
>>  #include <linux/mmc/mmc.h>
>> +#include <linux/mmc/sd.h>
>>  #include <linux/mmc/sdio.h>
>>  #include <linux/mmc/dw_mmc.h>
>>  #include <linux/bitops.h>
>> @@ -235,10 +236,13 @@ err:
>>  }
>>  #endif /* defined(CONFIG_DEBUG_FS) */
>>
>> +static void mci_send_cmd(struct dw_mci_slot *slot, u32 cmd, u32 arg);
>> +
>>  static u32 dw_mci_prepare_command(struct mmc_host *mmc, struct mmc_command *cmd)
>>  {
>>       struct mmc_data *data;
>>       struct dw_mci_slot *slot = mmc_priv(mmc);
>> +     struct dw_mci *host = slot->host;
>>       const struct dw_mci_drv_data *drv_data = slot->host->drv_data;
>>       u32 cmdr;
>>       cmd->error = -EINPROGRESS;
>> @@ -254,6 +258,32 @@ static u32 dw_mci_prepare_command(struct mmc_host *mmc, struct mmc_command *cmd)
>>       else if (cmd->opcode != MMC_SEND_STATUS && cmd->data)
>>               cmdr |= SDMMC_CMD_PRV_DAT_WAIT;
>>
>> +     if (cmd->opcode == SD_SWITCH_VOLTAGE) {
>> +             u32 clk_en_a;
>> +
>> +             /* Special bit makes CMD11 not die */
>> +             cmdr |= SDMMC_CMD_VOLT_SWITCH;
>> +
>> +             /* Change state to continue to handle CMD11 weirdness */
>> +             WARN_ON(slot->host->state != STATE_SENDING_CMD);
>> +             slot->host->state = STATE_SENDING_CMD11;
>> +
>> +             /*
>> +              * We need to disable clock stop while doing voltage switch
>> +              * according to 7.4.1.2 Voltage Switch Normal Scenario.
>> +              *
>> +              * It's assumed that by the next time the CLKENA is updated
>> +              * (when we set the clock next) that the voltage change will
>> +              * be over, so we don't bother setting any bits to synchronize
>> +              * with dw_mci_setup_bus().
>> +              */
>> +             clk_en_a = mci_readl(host, CLKENA);
>> +             clk_en_a &= ~(SDMMC_CLKEN_LOW_PWR << slot->id);
>> +             mci_writel(host, CLKENA, clk_en_a);
>> +             mci_send_cmd(slot, SDMMC_CMD_UPD_CLK |
>> +                          SDMMC_CMD_PRV_DAT_WAIT, 0);
> dw_mci_disable_low_power() can be used here.

Ah.  I guess we don't have that locally anymore.  Locally we have variants on:
* https://patchwork.kernel.org/patch/3070311/
* https://patchwork.kernel.org/patch/3070251/
* https://patchwork.kernel.org/patch/3070221/

...which removed that function.  ...but I guess upstream never picked
up those patches, huh?  Looking back it looks like you had some
feedback and it needed another spin but somehow fell off my plate.  :(

Maybe this is something Yuvaraj would like to pick up?

>
>> +     }
>> +
>>       if (cmd->flags & MMC_RSP_PRESENT) {
>>               /* We expect a response, so set this bit */
>>               cmdr |= SDMMC_CMD_RESP_EXP;
>> @@ -776,11 +806,15 @@ static void dw_mci_setup_bus(struct dw_mci_slot *slot, bool force_clkinit)
>>       unsigned int clock = slot->clock;
>>       u32 div;
>>       u32 clk_en_a;
>> +     u32 sdmmc_cmd_bits = SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT;
>> +
>> +     /* We must continue to set bit 28 in CMD until the change is complete */
>> +     if (host->state == STATE_WAITING_CMD11_DONE)
>> +             sdmmc_cmd_bits |= SDMMC_CMD_VOLT_SWITCH;
> I didn't get the reason SDMMC_CMD_VOLT_SWITCH is needed during clock update(enable/disable)
> Can you explain it in details?

Simply put: if I didn't do this then the system hung during voltage
switch.  It was not documented in the version of the IP manual that I
had access to and it took me a bunch of digging / trial and error to
figure this out.  Whatever this bit does internally it's important to
set it while the voltage change is happening.  Note that this need was
the whole reason for adding the extra state to the state machine.

Perhaps Yuvaraj can try without it and I'd assume he'll report the
same freeze.


>>       if (!clock) {
>>               mci_writel(host, CLKENA, 0);
>> -             mci_send_cmd(slot,
>> -                          SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT, 0);
>> +             mci_send_cmd(slot, sdmmc_cmd_bits, 0);
>>       } else if (clock != host->current_speed || force_clkinit) {
>>               div = host->bus_hz / clock;
>>               if (host->bus_hz % clock && host->bus_hz > clock)
>> @@ -804,15 +838,13 @@ static void dw_mci_setup_bus(struct dw_mci_slot *slot, bool force_clkinit)
>>               mci_writel(host, CLKSRC, 0);
>>
>>               /* inform CIU */
>> -             mci_send_cmd(slot,
>> -                          SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT, 0);
>> +             mci_send_cmd(slot, sdmmc_cmd_bits, 0);
>>
>>               /* set clock to desired speed */
>>               mci_writel(host, CLKDIV, div);
>>
>>               /* inform CIU */
>> -             mci_send_cmd(slot,
>> -                          SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT, 0);
>> +             mci_send_cmd(slot, sdmmc_cmd_bits, 0);
>>
>>               /* enable clock; only low power if no SDIO */
>>               clk_en_a = SDMMC_CLKEN_ENABLE << slot->id;
>> @@ -821,8 +853,7 @@ static void dw_mci_setup_bus(struct dw_mci_slot *slot, bool force_clkinit)
>>               mci_writel(host, CLKENA, clk_en_a);
>>
>>               /* inform CIU */
>> -             mci_send_cmd(slot,
>> -                          SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT, 0);
>> +             mci_send_cmd(slot, sdmmc_cmd_bits, 0);
>>
>>               /* keep the clock with reflecting clock dividor */
>>               slot->__clk_old = clock << div;
>> @@ -898,6 +929,17 @@ static void dw_mci_queue_request(struct dw_mci *host, struct dw_mci_slot *slot,
>>
>>       slot->mrq = mrq;
>>
>> +     if (host->state == STATE_WAITING_CMD11_DONE) {
>> +             dev_warn(&slot->mmc->class_dev,
>> +                      "Voltage change didn't complete\n");
>> +             /*
>> +              * this case isn't expected to happen, so we can
>> +              * either crash here or just try to continue on
>> +              * in the closest possible state
>> +              */
>> +             host->state = STATE_IDLE;
>> +     }
>> +
>>       if (host->state == STATE_IDLE) {
>>               host->state = STATE_SENDING_CMD;
>>               dw_mci_start_request(host, slot);
>> @@ -993,6 +1035,9 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
>>       /* Slot specific timing and width adjustment */
>>       dw_mci_setup_bus(slot, false);
>>
>> +     if (slot->host->state == STATE_WAITING_CMD11_DONE && ios->clock != 0)
>> +             slot->host->state = STATE_IDLE;
>> +
>>       switch (ios->power_mode) {
>>       case MMC_POWER_UP:
>>               if ((!IS_ERR(mmc->supply.vmmc)) &&
>> @@ -1039,6 +1084,68 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
>>       }
>>  }
>>
>> +static int dw_mci_card_busy(struct mmc_host *mmc)
>> +{
>> +     struct dw_mci_slot *slot = mmc_priv(mmc);
>> +     u32 status;
>> +
>> +     /*
>> +      * Check the busy bit which is low when DAT[3:0]
>> +      * (the data lines) are 0000
>> +      */
>> +     status = mci_readl(slot->host, STATUS);
>> +
>> +     return !!(status & SDMMC_STATUS_BUSY);
>> +}
>> +
>> +static int dw_mci_switch_voltage(struct mmc_host *mmc, struct mmc_ios *ios)
>> +{
>> +     struct dw_mci_slot *slot = mmc_priv(mmc);
>> +     struct dw_mci *host = slot->host;
>> +     u32 uhs;
>> +     u32 v18 = SDMMC_UHS_18V << slot->id;
>> +     int min_uv, max_uv;
>> +     int ret;
>> +
>> +     /*
>> +      * Program the voltage.  Note that some instances of dw_mmc may use
>> +      * the UHS_REG for this.  For other instances (like exynos) the UHS_REG
>> +      * does no harm but you need to set the regulator directly.  Try both.
>> +      */
>> +     uhs = mci_readl(host, UHS_REG);
>> +     if (ios->signal_voltage == MMC_SIGNAL_VOLTAGE_330) {
>> +             min_uv = 2700000;
>> +             max_uv = 3600000;
>> +             uhs &= ~v18;
>> +     } else {
>> +             min_uv = 1700000;
>> +             max_uv = 1950000;
>> +             uhs |= v18;
>> +     }
>> +     if (!IS_ERR(mmc->supply.vqmmc)) {
>> +             ret = regulator_set_voltage(mmc->supply.vqmmc, min_uv, max_uv);
>> +
>> +             /*
>> +              * Only complain if regulator claims that it's not in the 1.8V
>> +              * range.  This avoids a bunch of errors in the case that
>> +              * we've got a fixed 1.8V regulator but the core SD code still
>> +              * thinks it ought to try to switch to 3.3 and then back to 1.8
>> +              */
>> +             if (ret) {
> I think if ret is error, printing message and returning error is good.
> Currently, just returning '0' though it fails.
>
>> +                     int cur_voltage = 0;
>> +
>> +                     cur_voltage = regulator_get_voltage(mmc->supply.vqmmc);
>> +                     if (cur_voltage < 1700000 || cur_voltage > 1950000)
>> +                             dev_warn(&mmc->class_dev,
>> +                                      "Regulator set error %d: %d - %d\n",
>> +                                      ret, min_uv, max_uv);
>> +             }
>> +     }
>> +     mci_writel(host, UHS_REG, uhs);
>> +
>> +     return 0;
>> +}
>> +
>>  static int dw_mci_get_ro(struct mmc_host *mmc)
>>  {
>>       int read_only;
>> @@ -1180,6 +1287,9 @@ static const struct mmc_host_ops dw_mci_ops = {
>>       .get_cd                 = dw_mci_get_cd,
>>       .enable_sdio_irq        = dw_mci_enable_sdio_irq,
>>       .execute_tuning         = dw_mci_execute_tuning,
>> +     .card_busy              = dw_mci_card_busy,
>> +     .start_signal_voltage_switch = dw_mci_switch_voltage,
>> +
>>  };
>>
>>  static void dw_mci_request_end(struct dw_mci *host, struct mmc_request *mrq)
>> @@ -1203,7 +1313,11 @@ static void dw_mci_request_end(struct dw_mci *host, struct mmc_request *mrq)
>>               dw_mci_start_request(host, slot);
>>       } else {
>>               dev_vdbg(host->dev, "list empty\n");
>> -             host->state = STATE_IDLE;
>> +
>> +             if (host->state == STATE_SENDING_CMD11)
>> +                     host->state = STATE_WAITING_CMD11_DONE;
>> +             else
>> +                     host->state = STATE_IDLE;
>>       }
>>
>>       spin_unlock(&host->lock);
>> @@ -1314,8 +1428,10 @@ static void dw_mci_tasklet_func(unsigned long priv)
>>
>>               switch (state) {
>>               case STATE_IDLE:
>> +             case STATE_WAITING_CMD11_DONE:
>>                       break;
>>
>> +             case STATE_SENDING_CMD11:
>>               case STATE_SENDING_CMD:
>>                       if (!test_and_clear_bit(EVENT_CMD_COMPLETE,
>>                                               &host->pending_events))
>> @@ -1870,6 +1986,15 @@ static irqreturn_t dw_mci_interrupt(int irq, void *dev_id)
>>       }
>>
>>       if (pending) {
>> +             /* Check volt switch first, since it can look like an error */
>> +             if ((host->state == STATE_SENDING_CMD11) &&
>> +                 (pending & SDMMC_INT_VOLT_SWITCH)) {
>> +                     mci_writel(host, RINTSTS, SDMMC_INT_VOLT_SWITCH);
>> +                     pending &= ~SDMMC_INT_VOLT_SWITCH;
>> +                     dw_mci_cmd_interrupt(host,
>> +                                          pending | SDMMC_INT_CMD_DONE);
> Artificially adding SDMMC_INT_CMD_DONE to pending is needed to complete cmd handling?
> Is there any reason?

I don't remember this for sure since I wrote this a long time ago.
Maybe it's not needed?  I may have just been modeling on other
interrupts.

-Doug

[PATCH 3/3] mmc: dw_mmc: Support voltage changes

[Seungwon Jeon]

On Thu, June 26, 2014, Doug Anderson wrote:
> Seungwon,
> 
> On Wed, Jun 25, 2014 at 6:08 AM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
> > On Mon, June 23, 2014, Yuvaraj Kumar C D wrote:
> >> Subject: [PATCH 3/3] mmc: dw_mmc: Support voltage changes
> >>
> >> From: Doug Anderson <dianders@chromium.org>
> >>
> >> For UHS cards we need the ability to switch voltages from 3.3V to
> >> 1.8V.  Add support to the dw_mmc driver to handle this.  Note that
> >> dw_mmc needs a little bit of extra code since the interface needs a
> >> special bit programmed to the CMD register while CMD11 is progressing.
> >> This means adding a few extra states to the state machine to track.
> >
> > Overall new additional states makes it complicated.
> > Can we do that in other way?
> 
> That was the best I was able to figure out when I thought this
> through.  If you have ideas for doing it another way I'd imagine that
> Yuvaraj would be happy to take your feedback.
Let's clean up SDMMC_CMD_VOLT_SWITCH.
In turn, we may remove state-handling simply.

> 
> 
> >> Signed-off-by: Doug Anderson <dianders@chromium.org>
> >> Signed-off-by: Yuvaraj Kumar C D <yuvaraj.cd@samsung.com>
> >>
> >> ---
> >>  drivers/mmc/host/dw_mmc.c  |  145 +++++++++++++++++++++++++++++++++++++++++---
> >>  drivers/mmc/host/dw_mmc.h  |    5 +-
> >>  include/linux/mmc/dw_mmc.h |    2 +
> >>  3 files changed, 142 insertions(+), 10 deletions(-)
> >>
> >> diff --git a/drivers/mmc/host/dw_mmc.c b/drivers/mmc/host/dw_mmc.c
> >> index e034bce..38eb548 100644
> >> --- a/drivers/mmc/host/dw_mmc.c
> >> +++ b/drivers/mmc/host/dw_mmc.c
> >> @@ -29,6 +29,7 @@
> >>  #include <linux/irq.h>
> >>  #include <linux/mmc/host.h>
> >>  #include <linux/mmc/mmc.h>
> >> +#include <linux/mmc/sd.h>
> >>  #include <linux/mmc/sdio.h>
> >>  #include <linux/mmc/dw_mmc.h>
> >>  #include <linux/bitops.h>
> >> @@ -235,10 +236,13 @@ err:
> >>  }
> >>  #endif /* defined(CONFIG_DEBUG_FS) */
> >>
> >> +static void mci_send_cmd(struct dw_mci_slot *slot, u32 cmd, u32 arg);
> >> +
> >>  static u32 dw_mci_prepare_command(struct mmc_host *mmc, struct mmc_command *cmd)
> >>  {
> >>       struct mmc_data *data;
> >>       struct dw_mci_slot *slot = mmc_priv(mmc);
> >> +     struct dw_mci *host = slot->host;
> >>       const struct dw_mci_drv_data *drv_data = slot->host->drv_data;
> >>       u32 cmdr;
> >>       cmd->error = -EINPROGRESS;
> >> @@ -254,6 +258,32 @@ static u32 dw_mci_prepare_command(struct mmc_host *mmc, struct mmc_command
> *cmd)
> >>       else if (cmd->opcode != MMC_SEND_STATUS && cmd->data)
> >>               cmdr |= SDMMC_CMD_PRV_DAT_WAIT;
> >>
> >> +     if (cmd->opcode == SD_SWITCH_VOLTAGE) {
> >> +             u32 clk_en_a;
> >> +
> >> +             /* Special bit makes CMD11 not die */
> >> +             cmdr |= SDMMC_CMD_VOLT_SWITCH;
> >> +
> >> +             /* Change state to continue to handle CMD11 weirdness */
> >> +             WARN_ON(slot->host->state != STATE_SENDING_CMD);
> >> +             slot->host->state = STATE_SENDING_CMD11;
> >> +
> >> +             /*
> >> +              * We need to disable clock stop while doing voltage switch
> >> +              * according to 7.4.1.2 Voltage Switch Normal Scenario.
> >> +              *
> >> +              * It's assumed that by the next time the CLKENA is updated
> >> +              * (when we set the clock next) that the voltage change will
> >> +              * be over, so we don't bother setting any bits to synchronize
> >> +              * with dw_mci_setup_bus().
> >> +              */
> >> +             clk_en_a = mci_readl(host, CLKENA);
> >> +             clk_en_a &= ~(SDMMC_CLKEN_LOW_PWR << slot->id);
> >> +             mci_writel(host, CLKENA, clk_en_a);
> >> +             mci_send_cmd(slot, SDMMC_CMD_UPD_CLK |
> >> +                          SDMMC_CMD_PRV_DAT_WAIT, 0);
> > dw_mci_disable_low_power() can be used here.
> 
> Ah.  I guess we don't have that locally anymore.  Locally we have variants on:
I'm checking on cjb/mmc branch.

> * https://patchwork.kernel.org/patch/3070311/
> * https://patchwork.kernel.org/patch/3070251/
> * https://patchwork.kernel.org/patch/3070221/
> 
> ...which removed that function.  ...but I guess upstream never picked
> up those patches, huh?  Looking back it looks like you had some
> feedback and it needed another spin but somehow fell off my plate.  :(
> 
> Maybe this is something Yuvaraj would like to pick up?
It's long ago. I remember that there is no progress since my last comment.
In case of patch "3070221", I want to pick up for next Kernel.

> 
> >
> >> +     }
> >> +
> >>       if (cmd->flags & MMC_RSP_PRESENT) {
> >>               /* We expect a response, so set this bit */
> >>               cmdr |= SDMMC_CMD_RESP_EXP;
> >> @@ -776,11 +806,15 @@ static void dw_mci_setup_bus(struct dw_mci_slot *slot, bool force_clkinit)
> >>       unsigned int clock = slot->clock;
> >>       u32 div;
> >>       u32 clk_en_a;
> >> +     u32 sdmmc_cmd_bits = SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT;
> >> +
> >> +     /* We must continue to set bit 28 in CMD until the change is complete */
> >> +     if (host->state == STATE_WAITING_CMD11_DONE)
> >> +             sdmmc_cmd_bits |= SDMMC_CMD_VOLT_SWITCH;
> > I didn't get the reason SDMMC_CMD_VOLT_SWITCH is needed during clock update(enable/disable)
> > Can you explain it in details?
> 
> Simply put: if I didn't do this then the system hung during voltage
> switch.  It was not documented in the version of the IP manual that I
> had access to and it took me a bunch of digging / trial and error to
> figure this out.  Whatever this bit does internally it's important to
> set it while the voltage change is happening.  Note that this need was
> the whole reason for adding the extra state to the state machine.
> 
> Perhaps Yuvaraj can try without it and I'd assume he'll report the
> same freeze.
Clarify the necessity of SDMMC_CMD_VOLT_SWITCH.
As far as I experience, we didn't apply this bit for several projects.

> 
> 
> >>       if (!clock) {
> >>               mci_writel(host, CLKENA, 0);
> >> -             mci_send_cmd(slot,
> >> -                          SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT, 0);
> >> +             mci_send_cmd(slot, sdmmc_cmd_bits, 0);
> >>       } else if (clock != host->current_speed || force_clkinit) {
> >>               div = host->bus_hz / clock;
> >>               if (host->bus_hz % clock && host->bus_hz > clock)
> >> @@ -804,15 +838,13 @@ static void dw_mci_setup_bus(struct dw_mci_slot *slot, bool force_clkinit)
> >>               mci_writel(host, CLKSRC, 0);
> >>
> >>               /* inform CIU */
> >> -             mci_send_cmd(slot,
> >> -                          SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT, 0);
> >> +             mci_send_cmd(slot, sdmmc_cmd_bits, 0);
> >>
> >>               /* set clock to desired speed */
> >>               mci_writel(host, CLKDIV, div);
> >>
> >>               /* inform CIU */
> >> -             mci_send_cmd(slot,
> >> -                          SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT, 0);
> >> +             mci_send_cmd(slot, sdmmc_cmd_bits, 0);
> >>
> >>               /* enable clock; only low power if no SDIO */
> >>               clk_en_a = SDMMC_CLKEN_ENABLE << slot->id;
> >> @@ -821,8 +853,7 @@ static void dw_mci_setup_bus(struct dw_mci_slot *slot, bool force_clkinit)
> >>               mci_writel(host, CLKENA, clk_en_a);
> >>
> >>               /* inform CIU */
> >> -             mci_send_cmd(slot,
> >> -                          SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT, 0);
> >> +             mci_send_cmd(slot, sdmmc_cmd_bits, 0);
> >>
> >>               /* keep the clock with reflecting clock dividor */
> >>               slot->__clk_old = clock << div;
> >> @@ -898,6 +929,17 @@ static void dw_mci_queue_request(struct dw_mci *host, struct dw_mci_slot *slot,
> >>
> >>       slot->mrq = mrq;
> >>
> >> +     if (host->state == STATE_WAITING_CMD11_DONE) {
> >> +             dev_warn(&slot->mmc->class_dev,
> >> +                      "Voltage change didn't complete\n");
> >> +             /*
> >> +              * this case isn't expected to happen, so we can
> >> +              * either crash here or just try to continue on
> >> +              * in the closest possible state
> >> +              */
> >> +             host->state = STATE_IDLE;
> >> +     }
> >> +
> >>       if (host->state == STATE_IDLE) {
> >>               host->state = STATE_SENDING_CMD;
> >>               dw_mci_start_request(host, slot);
> >> @@ -993,6 +1035,9 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
> >>       /* Slot specific timing and width adjustment */
> >>       dw_mci_setup_bus(slot, false);
> >>
> >> +     if (slot->host->state == STATE_WAITING_CMD11_DONE && ios->clock != 0)
> >> +             slot->host->state = STATE_IDLE;
> >> +
> >>       switch (ios->power_mode) {
> >>       case MMC_POWER_UP:
> >>               if ((!IS_ERR(mmc->supply.vmmc)) &&
> >> @@ -1039,6 +1084,68 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
> >>       }
> >>  }
> >>
> >> +static int dw_mci_card_busy(struct mmc_host *mmc)
> >> +{
> >> +     struct dw_mci_slot *slot = mmc_priv(mmc);
> >> +     u32 status;
> >> +
> >> +     /*
> >> +      * Check the busy bit which is low when DAT[3:0]
> >> +      * (the data lines) are 0000
> >> +      */
> >> +     status = mci_readl(slot->host, STATUS);
> >> +
> >> +     return !!(status & SDMMC_STATUS_BUSY);
> >> +}
> >> +
> >> +static int dw_mci_switch_voltage(struct mmc_host *mmc, struct mmc_ios *ios)
> >> +{
> >> +     struct dw_mci_slot *slot = mmc_priv(mmc);
> >> +     struct dw_mci *host = slot->host;
> >> +     u32 uhs;
> >> +     u32 v18 = SDMMC_UHS_18V << slot->id;
> >> +     int min_uv, max_uv;
> >> +     int ret;
> >> +
> >> +     /*
> >> +      * Program the voltage.  Note that some instances of dw_mmc may use
> >> +      * the UHS_REG for this.  For other instances (like exynos) the UHS_REG
> >> +      * does no harm but you need to set the regulator directly.  Try both.
> >> +      */
> >> +     uhs = mci_readl(host, UHS_REG);
> >> +     if (ios->signal_voltage == MMC_SIGNAL_VOLTAGE_330) {
> >> +             min_uv = 2700000;
> >> +             max_uv = 3600000;
> >> +             uhs &= ~v18;
> >> +     } else {
> >> +             min_uv = 1700000;
> >> +             max_uv = 1950000;
> >> +             uhs |= v18;
> >> +     }
> >> +     if (!IS_ERR(mmc->supply.vqmmc)) {
> >> +             ret = regulator_set_voltage(mmc->supply.vqmmc, min_uv, max_uv);
> >> +
> >> +             /*
> >> +              * Only complain if regulator claims that it's not in the 1.8V
> >> +              * range.  This avoids a bunch of errors in the case that
> >> +              * we've got a fixed 1.8V regulator but the core SD code still
> >> +              * thinks it ought to try to switch to 3.3 and then back to 1.8
> >> +              */
> >> +             if (ret) {
> > I think if ret is error, printing message and returning error is good.
> > Currently, just returning '0' though it fails.
Any feedback?

> >
> >> +                     int cur_voltage = 0;
> >> +
> >> +                     cur_voltage = regulator_get_voltage(mmc->supply.vqmmc);
> >> +                     if (cur_voltage < 1700000 || cur_voltage > 1950000)
> >> +                             dev_warn(&mmc->class_dev,
> >> +                                      "Regulator set error %d: %d - %d\n",
> >> +                                      ret, min_uv, max_uv);
> >> +             }
> >> +     }
> >> +     mci_writel(host, UHS_REG, uhs);
> >> +
> >> +     return 0;
> >> +}
> >> +
> >>  static int dw_mci_get_ro(struct mmc_host *mmc)
> >>  {
> >>       int read_only;
> >> @@ -1180,6 +1287,9 @@ static const struct mmc_host_ops dw_mci_ops = {
> >>       .get_cd                 = dw_mci_get_cd,
> >>       .enable_sdio_irq        = dw_mci_enable_sdio_irq,
> >>       .execute_tuning         = dw_mci_execute_tuning,
> >> +     .card_busy              = dw_mci_card_busy,
> >> +     .start_signal_voltage_switch = dw_mci_switch_voltage,
> >> +
> >>  };
> >>
> >>  static void dw_mci_request_end(struct dw_mci *host, struct mmc_request *mrq)
> >> @@ -1203,7 +1313,11 @@ static void dw_mci_request_end(struct dw_mci *host, struct mmc_request *mrq)
> >>               dw_mci_start_request(host, slot);
> >>       } else {
> >>               dev_vdbg(host->dev, "list empty\n");
> >> -             host->state = STATE_IDLE;
> >> +
> >> +             if (host->state == STATE_SENDING_CMD11)
> >> +                     host->state = STATE_WAITING_CMD11_DONE;
> >> +             else
> >> +                     host->state = STATE_IDLE;
> >>       }
> >>
> >>       spin_unlock(&host->lock);
> >> @@ -1314,8 +1428,10 @@ static void dw_mci_tasklet_func(unsigned long priv)
> >>
> >>               switch (state) {
> >>               case STATE_IDLE:
> >> +             case STATE_WAITING_CMD11_DONE:
> >>                       break;
> >>
> >> +             case STATE_SENDING_CMD11:
> >>               case STATE_SENDING_CMD:
> >>                       if (!test_and_clear_bit(EVENT_CMD_COMPLETE,
> >>                                               &host->pending_events))
> >> @@ -1870,6 +1986,15 @@ static irqreturn_t dw_mci_interrupt(int irq, void *dev_id)
> >>       }
> >>
> >>       if (pending) {
> >> +             /* Check volt switch first, since it can look like an error */
> >> +             if ((host->state == STATE_SENDING_CMD11) &&
> >> +                 (pending & SDMMC_INT_VOLT_SWITCH)) {
> >> +                     mci_writel(host, RINTSTS, SDMMC_INT_VOLT_SWITCH);
> >> +                     pending &= ~SDMMC_INT_VOLT_SWITCH;
> >> +                     dw_mci_cmd_interrupt(host,
> >> +                                          pending | SDMMC_INT_CMD_DONE);
> > Artificially adding SDMMC_INT_CMD_DONE to pending is needed to complete cmd handling?
> > Is there any reason?
> 
> I don't remember this for sure since I wrote this a long time ago.
> Maybe it's not needed?  I may have just been modeling on other
> interrupts.
If there is no specific reason, please remove it.

Thanks,
Seungwon Jeon

> 
> -Doug
> --
> To unsubscribe from this list: send the line "unsubscribe linux-mmc" in
> the body of a message to majordomo at vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 3/3] mmc: dw_mmc: Support voltage changes

[Doug Anderson]

Seungwon,

On Thu, Jun 26, 2014 at 3:41 AM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
> On Thu, June 26, 2014, Doug Anderson wrote:
>> Seungwon,
>>
>> On Wed, Jun 25, 2014 at 6:08 AM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
>> > On Mon, June 23, 2014, Yuvaraj Kumar C D wrote:
>> >> Subject: [PATCH 3/3] mmc: dw_mmc: Support voltage changes
>> >>
>> >> From: Doug Anderson <dianders@chromium.org>
>> >>
>> >> For UHS cards we need the ability to switch voltages from 3.3V to
>> >> 1.8V.  Add support to the dw_mmc driver to handle this.  Note that
>> >> dw_mmc needs a little bit of extra code since the interface needs a
>> >> special bit programmed to the CMD register while CMD11 is progressing.
>> >> This means adding a few extra states to the state machine to track.
>> >
>> > Overall new additional states makes it complicated.
>> > Can we do that in other way?
>>
>> That was the best I was able to figure out when I thought this
>> through.  If you have ideas for doing it another way I'd imagine that
>> Yuvaraj would be happy to take your feedback.
> Let's clean up SDMMC_CMD_VOLT_SWITCH.
> In turn, we may remove state-handling simply.
>
>>
>>
>> >> Signed-off-by: Doug Anderson <dianders@chromium.org>
>> >> Signed-off-by: Yuvaraj Kumar C D <yuvaraj.cd@samsung.com>
>> >>
>> >> ---
>> >>  drivers/mmc/host/dw_mmc.c  |  145 +++++++++++++++++++++++++++++++++++++++++---
>> >>  drivers/mmc/host/dw_mmc.h  |    5 +-
>> >>  include/linux/mmc/dw_mmc.h |    2 +
>> >>  3 files changed, 142 insertions(+), 10 deletions(-)
>> >>
>> >> diff --git a/drivers/mmc/host/dw_mmc.c b/drivers/mmc/host/dw_mmc.c
>> >> index e034bce..38eb548 100644
>> >> --- a/drivers/mmc/host/dw_mmc.c
>> >> +++ b/drivers/mmc/host/dw_mmc.c
>> >> @@ -29,6 +29,7 @@
>> >>  #include <linux/irq.h>
>> >>  #include <linux/mmc/host.h>
>> >>  #include <linux/mmc/mmc.h>
>> >> +#include <linux/mmc/sd.h>
>> >>  #include <linux/mmc/sdio.h>
>> >>  #include <linux/mmc/dw_mmc.h>
>> >>  #include <linux/bitops.h>
>> >> @@ -235,10 +236,13 @@ err:
>> >>  }
>> >>  #endif /* defined(CONFIG_DEBUG_FS) */
>> >>
>> >> +static void mci_send_cmd(struct dw_mci_slot *slot, u32 cmd, u32 arg);
>> >> +
>> >>  static u32 dw_mci_prepare_command(struct mmc_host *mmc, struct mmc_command *cmd)
>> >>  {
>> >>       struct mmc_data *data;
>> >>       struct dw_mci_slot *slot = mmc_priv(mmc);
>> >> +     struct dw_mci *host = slot->host;
>> >>       const struct dw_mci_drv_data *drv_data = slot->host->drv_data;
>> >>       u32 cmdr;
>> >>       cmd->error = -EINPROGRESS;
>> >> @@ -254,6 +258,32 @@ static u32 dw_mci_prepare_command(struct mmc_host *mmc, struct mmc_command
>> *cmd)
>> >>       else if (cmd->opcode != MMC_SEND_STATUS && cmd->data)
>> >>               cmdr |= SDMMC_CMD_PRV_DAT_WAIT;
>> >>
>> >> +     if (cmd->opcode == SD_SWITCH_VOLTAGE) {
>> >> +             u32 clk_en_a;
>> >> +
>> >> +             /* Special bit makes CMD11 not die */
>> >> +             cmdr |= SDMMC_CMD_VOLT_SWITCH;
>> >> +
>> >> +             /* Change state to continue to handle CMD11 weirdness */
>> >> +             WARN_ON(slot->host->state != STATE_SENDING_CMD);
>> >> +             slot->host->state = STATE_SENDING_CMD11;
>> >> +
>> >> +             /*
>> >> +              * We need to disable clock stop while doing voltage switch
>> >> +              * according to 7.4.1.2 Voltage Switch Normal Scenario.
>> >> +              *
>> >> +              * It's assumed that by the next time the CLKENA is updated
>> >> +              * (when we set the clock next) that the voltage change will
>> >> +              * be over, so we don't bother setting any bits to synchronize
>> >> +              * with dw_mci_setup_bus().
>> >> +              */
>> >> +             clk_en_a = mci_readl(host, CLKENA);
>> >> +             clk_en_a &= ~(SDMMC_CLKEN_LOW_PWR << slot->id);
>> >> +             mci_writel(host, CLKENA, clk_en_a);
>> >> +             mci_send_cmd(slot, SDMMC_CMD_UPD_CLK |
>> >> +                          SDMMC_CMD_PRV_DAT_WAIT, 0);
>> > dw_mci_disable_low_power() can be used here.
>>
>> Ah.  I guess we don't have that locally anymore.  Locally we have variants on:
> I'm checking on cjb/mmc branch.
>
>> * https://patchwork.kernel.org/patch/3070311/
>> * https://patchwork.kernel.org/patch/3070251/
>> * https://patchwork.kernel.org/patch/3070221/
>>
>> ...which removed that function.  ...but I guess upstream never picked
>> up those patches, huh?  Looking back it looks like you had some
>> feedback and it needed another spin but somehow fell off my plate.  :(
>>
>> Maybe this is something Yuvaraj would like to pick up?
> It's long ago. I remember that there is no progress since my last comment.
> In case of patch "3070221", I want to pick up for next Kernel.

Sounds like Yuvaraj has agreed to look at addressing your comments.  Thanks!


>> >> +     }
>> >> +
>> >>       if (cmd->flags & MMC_RSP_PRESENT) {
>> >>               /* We expect a response, so set this bit */
>> >>               cmdr |= SDMMC_CMD_RESP_EXP;
>> >> @@ -776,11 +806,15 @@ static void dw_mci_setup_bus(struct dw_mci_slot *slot, bool force_clkinit)
>> >>       unsigned int clock = slot->clock;
>> >>       u32 div;
>> >>       u32 clk_en_a;
>> >> +     u32 sdmmc_cmd_bits = SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT;
>> >> +
>> >> +     /* We must continue to set bit 28 in CMD until the change is complete */
>> >> +     if (host->state == STATE_WAITING_CMD11_DONE)
>> >> +             sdmmc_cmd_bits |= SDMMC_CMD_VOLT_SWITCH;
>> > I didn't get the reason SDMMC_CMD_VOLT_SWITCH is needed during clock update(enable/disable)
>> > Can you explain it in details?
>>
>> Simply put: if I didn't do this then the system hung during voltage
>> switch.  It was not documented in the version of the IP manual that I
>> had access to and it took me a bunch of digging / trial and error to
>> figure this out.  Whatever this bit does internally it's important to
>> set it while the voltage change is happening.  Note that this need was
>> the whole reason for adding the extra state to the state machine.
>>
>> Perhaps Yuvaraj can try without it and I'd assume he'll report the
>> same freeze.
> Clarify the necessity of SDMMC_CMD_VOLT_SWITCH.
> As far as I experience, we didn't apply this bit for several projects.

I just tried this locally on our ChromeOS 3.8 kernel (which has quite
a few backports and matches dw_mmc upstream pretty closely).  When I
take out "sdmmc_cmd_bits |= SDMMC_CMD_VOLT_SWITCH" and bootup it fails
to bringup WiFi chip.  It prints messages like:

[    4.400842] mmc_host mmc1: Timeout sending command (cmd 0x202000
arg 0x0 status 0x80202000)

I will let Yuvaraj comment about his testing too.


>> >> +static int dw_mci_switch_voltage(struct mmc_host *mmc, struct mmc_ios *ios)
>> >> +{
>> >> +     struct dw_mci_slot *slot = mmc_priv(mmc);
>> >> +     struct dw_mci *host = slot->host;
>> >> +     u32 uhs;
>> >> +     u32 v18 = SDMMC_UHS_18V << slot->id;
>> >> +     int min_uv, max_uv;
>> >> +     int ret;
>> >> +
>> >> +     /*
>> >> +      * Program the voltage.  Note that some instances of dw_mmc may use
>> >> +      * the UHS_REG for this.  For other instances (like exynos) the UHS_REG
>> >> +      * does no harm but you need to set the regulator directly.  Try both.
>> >> +      */
>> >> +     uhs = mci_readl(host, UHS_REG);
>> >> +     if (ios->signal_voltage == MMC_SIGNAL_VOLTAGE_330) {
>> >> +             min_uv = 2700000;
>> >> +             max_uv = 3600000;
>> >> +             uhs &= ~v18;
>> >> +     } else {
>> >> +             min_uv = 1700000;
>> >> +             max_uv = 1950000;
>> >> +             uhs |= v18;
>> >> +     }
>> >> +     if (!IS_ERR(mmc->supply.vqmmc)) {
>> >> +             ret = regulator_set_voltage(mmc->supply.vqmmc, min_uv, max_uv);
>> >> +
>> >> +             /*
>> >> +              * Only complain if regulator claims that it's not in the 1.8V
>> >> +              * range.  This avoids a bunch of errors in the case that
>> >> +              * we've got a fixed 1.8V regulator but the core SD code still
>> >> +              * thinks it ought to try to switch to 3.3 and then back to 1.8
>> >> +              */
>> >> +             if (ret) {
>> > I think if ret is error, printing message and returning error is good.
>> > Currently, just returning '0' though it fails.
> Any feedback?

Whoops, right.  I think you're right that in the case it warns it
should also return the error code.  In the case it doesn't warn it
shouldn't.  Also: possibly we should use a trick like the mmc core
does and use "regulator_count_voltages" to figure out if we're going
to be able to switch voltages...

>> > Artificially adding SDMMC_INT_CMD_DONE to pending is needed to complete cmd handling?
>> > Is there any reason?
>>
>> I don't remember this for sure since I wrote this a long time ago.
>> Maybe it's not needed?  I may have just been modeling on other
>> interrupts.
> If there is no specific reason, please remove it.

OK, we'll see how Yuvaraj's testing goes without it.  My incredibly
brief testing in our local Chrome OS 3.8 tree shows that the WiFi is
not detected properly if I comment out the line:
  dw_mci_cmd_interrupt(host,
    pending | SDMMC_INT_CMD_DONE);

It may be sufficient to simply schedule the tasklet or to do some
other subset of dw_mci_cmd_interrupt().  Yuvaraj can confirm on the
latest kernel and also investigate further.


-Doug

[PATCH 3/3] mmc: dw_mmc: Support voltage changes

[Yuvaraj Kumar]

On Thu, Jun 26, 2014 at 10:20 PM, Doug Anderson <dianders@chromium.org> wrote:
> Seungwon,
>
> On Thu, Jun 26, 2014 at 3:41 AM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
>> On Thu, June 26, 2014, Doug Anderson wrote:
>>> Seungwon,
>>>
>>> On Wed, Jun 25, 2014 at 6:08 AM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
>>> > On Mon, June 23, 2014, Yuvaraj Kumar C D wrote:
>>> >> Subject: [PATCH 3/3] mmc: dw_mmc: Support voltage changes
>>> >>
>>> >> From: Doug Anderson <dianders@chromium.org>
>>> >>
>>> >> For UHS cards we need the ability to switch voltages from 3.3V to
>>> >> 1.8V.  Add support to the dw_mmc driver to handle this.  Note that
>>> >> dw_mmc needs a little bit of extra code since the interface needs a
>>> >> special bit programmed to the CMD register while CMD11 is progressing.
>>> >> This means adding a few extra states to the state machine to track.
>>> >
>>> > Overall new additional states makes it complicated.
>>> > Can we do that in other way?
>>>
>>> That was the best I was able to figure out when I thought this
>>> through.  If you have ideas for doing it another way I'd imagine that
>>> Yuvaraj would be happy to take your feedback.
>> Let's clean up SDMMC_CMD_VOLT_SWITCH.
>> In turn, we may remove state-handling simply.
>>
>>>
>>>
>>> >> Signed-off-by: Doug Anderson <dianders@chromium.org>
>>> >> Signed-off-by: Yuvaraj Kumar C D <yuvaraj.cd@samsung.com>
>>> >>
>>> >> ---
>>> >>  drivers/mmc/host/dw_mmc.c  |  145 +++++++++++++++++++++++++++++++++++++++++---
>>> >>  drivers/mmc/host/dw_mmc.h  |    5 +-
>>> >>  include/linux/mmc/dw_mmc.h |    2 +
>>> >>  3 files changed, 142 insertions(+), 10 deletions(-)
>>> >>
>>> >> diff --git a/drivers/mmc/host/dw_mmc.c b/drivers/mmc/host/dw_mmc.c
>>> >> index e034bce..38eb548 100644
>>> >> --- a/drivers/mmc/host/dw_mmc.c
>>> >> +++ b/drivers/mmc/host/dw_mmc.c
>>> >> @@ -29,6 +29,7 @@
>>> >>  #include <linux/irq.h>
>>> >>  #include <linux/mmc/host.h>
>>> >>  #include <linux/mmc/mmc.h>
>>> >> +#include <linux/mmc/sd.h>
>>> >>  #include <linux/mmc/sdio.h>
>>> >>  #include <linux/mmc/dw_mmc.h>
>>> >>  #include <linux/bitops.h>
>>> >> @@ -235,10 +236,13 @@ err:
>>> >>  }
>>> >>  #endif /* defined(CONFIG_DEBUG_FS) */
>>> >>
>>> >> +static void mci_send_cmd(struct dw_mci_slot *slot, u32 cmd, u32 arg);
>>> >> +
>>> >>  static u32 dw_mci_prepare_command(struct mmc_host *mmc, struct mmc_command *cmd)
>>> >>  {
>>> >>       struct mmc_data *data;
>>> >>       struct dw_mci_slot *slot = mmc_priv(mmc);
>>> >> +     struct dw_mci *host = slot->host;
>>> >>       const struct dw_mci_drv_data *drv_data = slot->host->drv_data;
>>> >>       u32 cmdr;
>>> >>       cmd->error = -EINPROGRESS;
>>> >> @@ -254,6 +258,32 @@ static u32 dw_mci_prepare_command(struct mmc_host *mmc, struct mmc_command
>>> *cmd)
>>> >>       else if (cmd->opcode != MMC_SEND_STATUS && cmd->data)
>>> >>               cmdr |= SDMMC_CMD_PRV_DAT_WAIT;
>>> >>
>>> >> +     if (cmd->opcode == SD_SWITCH_VOLTAGE) {
>>> >> +             u32 clk_en_a;
>>> >> +
>>> >> +             /* Special bit makes CMD11 not die */
>>> >> +             cmdr |= SDMMC_CMD_VOLT_SWITCH;
>>> >> +
>>> >> +             /* Change state to continue to handle CMD11 weirdness */
>>> >> +             WARN_ON(slot->host->state != STATE_SENDING_CMD);
>>> >> +             slot->host->state = STATE_SENDING_CMD11;
>>> >> +
>>> >> +             /*
>>> >> +              * We need to disable clock stop while doing voltage switch
>>> >> +              * according to 7.4.1.2 Voltage Switch Normal Scenario.
>>> >> +              *
>>> >> +              * It's assumed that by the next time the CLKENA is updated
>>> >> +              * (when we set the clock next) that the voltage change will
>>> >> +              * be over, so we don't bother setting any bits to synchronize
>>> >> +              * with dw_mci_setup_bus().
>>> >> +              */
>>> >> +             clk_en_a = mci_readl(host, CLKENA);
>>> >> +             clk_en_a &= ~(SDMMC_CLKEN_LOW_PWR << slot->id);
>>> >> +             mci_writel(host, CLKENA, clk_en_a);
>>> >> +             mci_send_cmd(slot, SDMMC_CMD_UPD_CLK |
>>> >> +                          SDMMC_CMD_PRV_DAT_WAIT, 0);
>>> > dw_mci_disable_low_power() can be used here.
>>>
>>> Ah.  I guess we don't have that locally anymore.  Locally we have variants on:
>> I'm checking on cjb/mmc branch.
>>
>>> * https://patchwork.kernel.org/patch/3070311/
>>> * https://patchwork.kernel.org/patch/3070251/
>>> * https://patchwork.kernel.org/patch/3070221/
>>>
>>> ...which removed that function.  ...but I guess upstream never picked
>>> up those patches, huh?  Looking back it looks like you had some
>>> feedback and it needed another spin but somehow fell off my plate.  :(
>>>
>>> Maybe this is something Yuvaraj would like to pick up?
>> It's long ago. I remember that there is no progress since my last comment.
>> In case of patch "3070221", I want to pick up for next Kernel.
>
> Sounds like Yuvaraj has agreed to look at addressing your comments.  Thanks!
>
>
>>> >> +     }
>>> >> +
>>> >>       if (cmd->flags & MMC_RSP_PRESENT) {
>>> >>               /* We expect a response, so set this bit */
>>> >>               cmdr |= SDMMC_CMD_RESP_EXP;
>>> >> @@ -776,11 +806,15 @@ static void dw_mci_setup_bus(struct dw_mci_slot *slot, bool force_clkinit)
>>> >>       unsigned int clock = slot->clock;
>>> >>       u32 div;
>>> >>       u32 clk_en_a;
>>> >> +     u32 sdmmc_cmd_bits = SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT;
>>> >> +
>>> >> +     /* We must continue to set bit 28 in CMD until the change is complete */
>>> >> +     if (host->state == STATE_WAITING_CMD11_DONE)
>>> >> +             sdmmc_cmd_bits |= SDMMC_CMD_VOLT_SWITCH;
>>> > I didn't get the reason SDMMC_CMD_VOLT_SWITCH is needed during clock update(enable/disable)
>>> > Can you explain it in details?
>>>
>>> Simply put: if I didn't do this then the system hung during voltage
>>> switch.  It was not documented in the version of the IP manual that I
>>> had access to and it took me a bunch of digging / trial and error to
>>> figure this out.  Whatever this bit does internally it's important to
>>> set it while the voltage change is happening.  Note that this need was
>>> the whole reason for adding the extra state to the state machine.
>>>
>>> Perhaps Yuvaraj can try without it and I'd assume he'll report the
>>> same freeze.
>> Clarify the necessity of SDMMC_CMD_VOLT_SWITCH.
>> As far as I experience, we didn't apply this bit for several projects.
>
> I just tried this locally on our ChromeOS 3.8 kernel (which has quite
> a few backports and matches dw_mmc upstream pretty closely).  When I
> take out "sdmmc_cmd_bits |= SDMMC_CMD_VOLT_SWITCH" and bootup it fails
> to bringup WiFi chip.  It prints messages like:
>
> [    4.400842] mmc_host mmc1: Timeout sending command (cmd 0x202000
> arg 0x0 status 0x80202000)
>
> I will let Yuvaraj comment about his testing too.
To make it simple and verify, i have just enabled eMMC and SD channels
on 3.16-rc1 and tested.
Without "sdmmc_cmd_bits |= SDMMC_CMD_VOLT_SWITCH" bootup fails.
[    3.015971] mmc_host mmc1: Timeout sending command (cmd 0x202000
arg 0x0 status 0x80202000)
[    3.530974] mmc_host mmc1: Timeout sending command (cmd 0x202000
arg 0x0 status 0x80202000)
>
>
>>> >> +static int dw_mci_switch_voltage(struct mmc_host *mmc, struct mmc_ios *ios)
>>> >> +{
>>> >> +     struct dw_mci_slot *slot = mmc_priv(mmc);
>>> >> +     struct dw_mci *host = slot->host;
>>> >> +     u32 uhs;
>>> >> +     u32 v18 = SDMMC_UHS_18V << slot->id;
>>> >> +     int min_uv, max_uv;
>>> >> +     int ret;
>>> >> +
>>> >> +     /*
>>> >> +      * Program the voltage.  Note that some instances of dw_mmc may use
>>> >> +      * the UHS_REG for this.  For other instances (like exynos) the UHS_REG
>>> >> +      * does no harm but you need to set the regulator directly.  Try both.
>>> >> +      */
>>> >> +     uhs = mci_readl(host, UHS_REG);
>>> >> +     if (ios->signal_voltage == MMC_SIGNAL_VOLTAGE_330) {
>>> >> +             min_uv = 2700000;
>>> >> +             max_uv = 3600000;
>>> >> +             uhs &= ~v18;
>>> >> +     } else {
>>> >> +             min_uv = 1700000;
>>> >> +             max_uv = 1950000;
>>> >> +             uhs |= v18;
>>> >> +     }
>>> >> +     if (!IS_ERR(mmc->supply.vqmmc)) {
>>> >> +             ret = regulator_set_voltage(mmc->supply.vqmmc, min_uv, max_uv);
>>> >> +
>>> >> +             /*
>>> >> +              * Only complain if regulator claims that it's not in the 1.8V
>>> >> +              * range.  This avoids a bunch of errors in the case that
>>> >> +              * we've got a fixed 1.8V regulator but the core SD code still
>>> >> +              * thinks it ought to try to switch to 3.3 and then back to 1.8
>>> >> +              */
>>> >> +             if (ret) {
>>> > I think if ret is error, printing message and returning error is good.
>>> > Currently, just returning '0' though it fails.
>> Any feedback?
>
> Whoops, right.  I think you're right that in the case it warns it
> should also return the error code.  In the case it doesn't warn it
> shouldn't.  Also: possibly we should use a trick like the mmc core
> does and use "regulator_count_voltages" to figure out if we're going
> to be able to switch voltages...
>
>>> > Artificially adding SDMMC_INT_CMD_DONE to pending is needed to complete cmd handling?
>>> > Is there any reason?
>>>
>>> I don't remember this for sure since I wrote this a long time ago.
>>> Maybe it's not needed?  I may have just been modeling on other
>>> interrupts.
>> If there is no specific reason, please remove it.
>
> OK, we'll see how Yuvaraj's testing goes without it.  My incredibly
> brief testing in our local Chrome OS 3.8 tree shows that the WiFi is
> not detected properly if I comment out the line:
>   dw_mci_cmd_interrupt(host,
>     pending | SDMMC_INT_CMD_DONE);
>
> It may be sufficient to simply schedule the tasklet or to do some
> other subset of dw_mci_cmd_interrupt().  Yuvaraj can confirm on the
> latest kernel and also investigate further.
>
>
> -Doug

[PATCH 3/3] mmc: dw_mmc: Support voltage changes

[Seungwon Jeon]

Hi Yuvaraj,

On Fri, June 27, 2014, Yuvaraj Kumar wrote:
> On Thu, Jun 26, 2014 at 10:20 PM, Doug Anderson <dianders@chromium.org> wrote:
> > Seungwon,
> >
> > On Thu, Jun 26, 2014 at 3:41 AM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
> >> On Thu, June 26, 2014, Doug Anderson wrote:
> >>> Seungwon,
> >>>
> >>> On Wed, Jun 25, 2014 at 6:08 AM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
> >>> > On Mon, June 23, 2014, Yuvaraj Kumar C D wrote:
> >>> >> Subject: [PATCH 3/3] mmc: dw_mmc: Support voltage changes
> >>> >>
> >>> >> From: Doug Anderson <dianders@chromium.org>
> >>> >>
> >>> >> For UHS cards we need the ability to switch voltages from 3.3V to
> >>> >> 1.8V.  Add support to the dw_mmc driver to handle this.  Note that
> >>> >> dw_mmc needs a little bit of extra code since the interface needs a
> >>> >> special bit programmed to the CMD register while CMD11 is progressing.
> >>> >> This means adding a few extra states to the state machine to track.
> >>> >
> >>> > Overall new additional states makes it complicated.
> >>> > Can we do that in other way?
> >>>
> >>> That was the best I was able to figure out when I thought this
> >>> through.  If you have ideas for doing it another way I'd imagine that
> >>> Yuvaraj would be happy to take your feedback.
> >> Let's clean up SDMMC_CMD_VOLT_SWITCH.
> >> In turn, we may remove state-handling simply.
> >>
> >>>
> >>>
> >>> >> Signed-off-by: Doug Anderson <dianders@chromium.org>
> >>> >> Signed-off-by: Yuvaraj Kumar C D <yuvaraj.cd@samsung.com>
> >>> >>
> >>> >> ---
> >>> >>  drivers/mmc/host/dw_mmc.c  |  145 +++++++++++++++++++++++++++++++++++++++++---
> >>> >>  drivers/mmc/host/dw_mmc.h  |    5 +-
> >>> >>  include/linux/mmc/dw_mmc.h |    2 +
> >>> >>  3 files changed, 142 insertions(+), 10 deletions(-)
> >>> >>
> >>> >> diff --git a/drivers/mmc/host/dw_mmc.c b/drivers/mmc/host/dw_mmc.c
> >>> >> index e034bce..38eb548 100644
> >>> >> --- a/drivers/mmc/host/dw_mmc.c
> >>> >> +++ b/drivers/mmc/host/dw_mmc.c
> >>> >> @@ -29,6 +29,7 @@
> >>> >>  #include <linux/irq.h>
> >>> >>  #include <linux/mmc/host.h>
> >>> >>  #include <linux/mmc/mmc.h>
> >>> >> +#include <linux/mmc/sd.h>
> >>> >>  #include <linux/mmc/sdio.h>
> >>> >>  #include <linux/mmc/dw_mmc.h>
> >>> >>  #include <linux/bitops.h>
> >>> >> @@ -235,10 +236,13 @@ err:
> >>> >>  }
> >>> >>  #endif /* defined(CONFIG_DEBUG_FS) */
> >>> >>
> >>> >> +static void mci_send_cmd(struct dw_mci_slot *slot, u32 cmd, u32 arg);
> >>> >> +
> >>> >>  static u32 dw_mci_prepare_command(struct mmc_host *mmc, struct mmc_command *cmd)
> >>> >>  {
> >>> >>       struct mmc_data *data;
> >>> >>       struct dw_mci_slot *slot = mmc_priv(mmc);
> >>> >> +     struct dw_mci *host = slot->host;
> >>> >>       const struct dw_mci_drv_data *drv_data = slot->host->drv_data;
> >>> >>       u32 cmdr;
> >>> >>       cmd->error = -EINPROGRESS;
> >>> >> @@ -254,6 +258,32 @@ static u32 dw_mci_prepare_command(struct mmc_host *mmc, struct mmc_command
> >>> *cmd)
> >>> >>       else if (cmd->opcode != MMC_SEND_STATUS && cmd->data)
> >>> >>               cmdr |= SDMMC_CMD_PRV_DAT_WAIT;
> >>> >>
> >>> >> +     if (cmd->opcode == SD_SWITCH_VOLTAGE) {
> >>> >> +             u32 clk_en_a;
> >>> >> +
> >>> >> +             /* Special bit makes CMD11 not die */
> >>> >> +             cmdr |= SDMMC_CMD_VOLT_SWITCH;
> >>> >> +
> >>> >> +             /* Change state to continue to handle CMD11 weirdness */
> >>> >> +             WARN_ON(slot->host->state != STATE_SENDING_CMD);
> >>> >> +             slot->host->state = STATE_SENDING_CMD11;
> >>> >> +
> >>> >> +             /*
> >>> >> +              * We need to disable clock stop while doing voltage switch
> >>> >> +              * according to 7.4.1.2 Voltage Switch Normal Scenario.
> >>> >> +              *
> >>> >> +              * It's assumed that by the next time the CLKENA is updated
> >>> >> +              * (when we set the clock next) that the voltage change will
> >>> >> +              * be over, so we don't bother setting any bits to synchronize
> >>> >> +              * with dw_mci_setup_bus().
> >>> >> +              */
> >>> >> +             clk_en_a = mci_readl(host, CLKENA);
> >>> >> +             clk_en_a &= ~(SDMMC_CLKEN_LOW_PWR << slot->id);
> >>> >> +             mci_writel(host, CLKENA, clk_en_a);
> >>> >> +             mci_send_cmd(slot, SDMMC_CMD_UPD_CLK |
> >>> >> +                          SDMMC_CMD_PRV_DAT_WAIT, 0);
> >>> > dw_mci_disable_low_power() can be used here.
> >>>
> >>> Ah.  I guess we don't have that locally anymore.  Locally we have variants on:
> >> I'm checking on cjb/mmc branch.
> >>
> >>> * https://patchwork.kernel.org/patch/3070311/
> >>> * https://patchwork.kernel.org/patch/3070251/
> >>> * https://patchwork.kernel.org/patch/3070221/
> >>>
> >>> ...which removed that function.  ...but I guess upstream never picked
> >>> up those patches, huh?  Looking back it looks like you had some
> >>> feedback and it needed another spin but somehow fell off my plate.  :(
> >>>
> >>> Maybe this is something Yuvaraj would like to pick up?
> >> It's long ago. I remember that there is no progress since my last comment.
> >> In case of patch "3070221", I want to pick up for next Kernel.
> >
> > Sounds like Yuvaraj has agreed to look at addressing your comments.  Thanks!
> >
> >
> >>> >> +     }
> >>> >> +
> >>> >>       if (cmd->flags & MMC_RSP_PRESENT) {
> >>> >>               /* We expect a response, so set this bit */
> >>> >>               cmdr |= SDMMC_CMD_RESP_EXP;
> >>> >> @@ -776,11 +806,15 @@ static void dw_mci_setup_bus(struct dw_mci_slot *slot, bool force_clkinit)
> >>> >>       unsigned int clock = slot->clock;
> >>> >>       u32 div;
> >>> >>       u32 clk_en_a;
> >>> >> +     u32 sdmmc_cmd_bits = SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT;
> >>> >> +
> >>> >> +     /* We must continue to set bit 28 in CMD until the change is complete */
> >>> >> +     if (host->state == STATE_WAITING_CMD11_DONE)
> >>> >> +             sdmmc_cmd_bits |= SDMMC_CMD_VOLT_SWITCH;
> >>> > I didn't get the reason SDMMC_CMD_VOLT_SWITCH is needed during clock update(enable/disable)
> >>> > Can you explain it in details?
> >>>
> >>> Simply put: if I didn't do this then the system hung during voltage
> >>> switch.  It was not documented in the version of the IP manual that I
> >>> had access to and it took me a bunch of digging / trial and error to
> >>> figure this out.  Whatever this bit does internally it's important to
> >>> set it while the voltage change is happening.  Note that this need was
> >>> the whole reason for adding the extra state to the state machine.
> >>>
> >>> Perhaps Yuvaraj can try without it and I'd assume he'll report the
> >>> same freeze.
> >> Clarify the necessity of SDMMC_CMD_VOLT_SWITCH.
> >> As far as I experience, we didn't apply this bit for several projects.
> >
> > I just tried this locally on our ChromeOS 3.8 kernel (which has quite
> > a few backports and matches dw_mmc upstream pretty closely).  When I
> > take out "sdmmc_cmd_bits |= SDMMC_CMD_VOLT_SWITCH" and bootup it fails
> > to bringup WiFi chip.  It prints messages like:
> >
> > [    4.400842] mmc_host mmc1: Timeout sending command (cmd 0x202000
> > arg 0x0 status 0x80202000)
> >
> > I will let Yuvaraj comment about his testing too.
> To make it simple and verify, i have just enabled eMMC and SD channels
> on 3.16-rc1 and tested.
> Without "sdmmc_cmd_bits |= SDMMC_CMD_VOLT_SWITCH" bootup fails.
> [    3.015971] mmc_host mmc1: Timeout sending command (cmd 0x202000
> arg 0x0 status 0x80202000)
> [    3.530974] mmc_host mmc1: Timeout sending command (cmd 0x202000
> arg 0x0 status 0x80202000)
Thank you for test.
Hmm, it failed during clock update, right?.
Can you check HLE interrupt at this time?
I'll investigate for this.

> >
> >
> >>> >> +static int dw_mci_switch_voltage(struct mmc_host *mmc, struct mmc_ios *ios)
> >>> >> +{
> >>> >> +     struct dw_mci_slot *slot = mmc_priv(mmc);
> >>> >> +     struct dw_mci *host = slot->host;
> >>> >> +     u32 uhs;
> >>> >> +     u32 v18 = SDMMC_UHS_18V << slot->id;
> >>> >> +     int min_uv, max_uv;
> >>> >> +     int ret;
> >>> >> +
> >>> >> +     /*
> >>> >> +      * Program the voltage.  Note that some instances of dw_mmc may use
> >>> >> +      * the UHS_REG for this.  For other instances (like exynos) the UHS_REG
> >>> >> +      * does no harm but you need to set the regulator directly.  Try both.
> >>> >> +      */
> >>> >> +     uhs = mci_readl(host, UHS_REG);
> >>> >> +     if (ios->signal_voltage == MMC_SIGNAL_VOLTAGE_330) {
> >>> >> +             min_uv = 2700000;
> >>> >> +             max_uv = 3600000;
> >>> >> +             uhs &= ~v18;
> >>> >> +     } else {
> >>> >> +             min_uv = 1700000;
> >>> >> +             max_uv = 1950000;
> >>> >> +             uhs |= v18;
> >>> >> +     }
> >>> >> +     if (!IS_ERR(mmc->supply.vqmmc)) {
> >>> >> +             ret = regulator_set_voltage(mmc->supply.vqmmc, min_uv, max_uv);
> >>> >> +
> >>> >> +             /*
> >>> >> +              * Only complain if regulator claims that it's not in the 1.8V
> >>> >> +              * range.  This avoids a bunch of errors in the case that
> >>> >> +              * we've got a fixed 1.8V regulator but the core SD code still
> >>> >> +              * thinks it ought to try to switch to 3.3 and then back to 1.8
> >>> >> +              */
> >>> >> +             if (ret) {
> >>> > I think if ret is error, printing message and returning error is good.
> >>> > Currently, just returning '0' though it fails.
> >> Any feedback?
> >
> > Whoops, right.  I think you're right that in the case it warns it
> > should also return the error code.  In the case it doesn't warn it
> > shouldn't.  Also: possibly we should use a trick like the mmc core
> > does and use "regulator_count_voltages" to figure out if we're going
> > to be able to switch voltages...
> >
> >>> > Artificially adding SDMMC_INT_CMD_DONE to pending is needed to complete cmd handling?
> >>> > Is there any reason?
> >>>
> >>> I don't remember this for sure since I wrote this a long time ago.
> >>> Maybe it's not needed?  I may have just been modeling on other
> >>> interrupts.
> >> If there is no specific reason, please remove it.
> >
> > OK, we'll see how Yuvaraj's testing goes without it.  My incredibly
> > brief testing in our local Chrome OS 3.8 tree shows that the WiFi is
> > not detected properly if I comment out the line:
> >   dw_mci_cmd_interrupt(host,
> >     pending | SDMMC_INT_CMD_DONE);
> >
> > It may be sufficient to simply schedule the tasklet or to do some
> > other subset of dw_mci_cmd_interrupt().  Yuvaraj can confirm on the
> > latest kernel and also investigate further.
How about completing CMD without SDMMC_INT_CMD_DONE?

Thanks,
Seungwon Jeon

[PATCH 3/3] mmc: dw_mmc: Support voltage changes

[Jaehoon Chung]

On 06/27/2014 08:18 PM, Seungwon Jeon wrote:
> Hi Yuvaraj,
> 
> On Fri, June 27, 2014, Yuvaraj Kumar wrote:
>> On Thu, Jun 26, 2014 at 10:20 PM, Doug Anderson <dianders@chromium.org> wrote:
>>> Seungwon,
>>>
>>> On Thu, Jun 26, 2014 at 3:41 AM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
>>>> On Thu, June 26, 2014, Doug Anderson wrote:
>>>>> Seungwon,
>>>>>
>>>>> On Wed, Jun 25, 2014 at 6:08 AM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
>>>>>> On Mon, June 23, 2014, Yuvaraj Kumar C D wrote:
>>>>>>> Subject: [PATCH 3/3] mmc: dw_mmc: Support voltage changes
>>>>>>>
>>>>>>> From: Doug Anderson <dianders@chromium.org>
>>>>>>>
>>>>>>> For UHS cards we need the ability to switch voltages from 3.3V to
>>>>>>> 1.8V.  Add support to the dw_mmc driver to handle this.  Note that
>>>>>>> dw_mmc needs a little bit of extra code since the interface needs a
>>>>>>> special bit programmed to the CMD register while CMD11 is progressing.
>>>>>>> This means adding a few extra states to the state machine to track.
>>>>>>
>>>>>> Overall new additional states makes it complicated.
>>>>>> Can we do that in other way?
>>>>>
>>>>> That was the best I was able to figure out when I thought this
>>>>> through.  If you have ideas for doing it another way I'd imagine that
>>>>> Yuvaraj would be happy to take your feedback.
>>>> Let's clean up SDMMC_CMD_VOLT_SWITCH.
>>>> In turn, we may remove state-handling simply.
>>>>
>>>>>
>>>>>
>>>>>>> Signed-off-by: Doug Anderson <dianders@chromium.org>
>>>>>>> Signed-off-by: Yuvaraj Kumar C D <yuvaraj.cd@samsung.com>
>>>>>>>
>>>>>>> ---
>>>>>>>  drivers/mmc/host/dw_mmc.c  |  145 +++++++++++++++++++++++++++++++++++++++++---
>>>>>>>  drivers/mmc/host/dw_mmc.h  |    5 +-
>>>>>>>  include/linux/mmc/dw_mmc.h |    2 +
>>>>>>>  3 files changed, 142 insertions(+), 10 deletions(-)
>>>>>>>
>>>>>>> diff --git a/drivers/mmc/host/dw_mmc.c b/drivers/mmc/host/dw_mmc.c
>>>>>>> index e034bce..38eb548 100644
>>>>>>> --- a/drivers/mmc/host/dw_mmc.c
>>>>>>> +++ b/drivers/mmc/host/dw_mmc.c
>>>>>>> @@ -29,6 +29,7 @@
>>>>>>>  #include <linux/irq.h>
>>>>>>>  #include <linux/mmc/host.h>
>>>>>>>  #include <linux/mmc/mmc.h>
>>>>>>> +#include <linux/mmc/sd.h>
>>>>>>>  #include <linux/mmc/sdio.h>
>>>>>>>  #include <linux/mmc/dw_mmc.h>
>>>>>>>  #include <linux/bitops.h>
>>>>>>> @@ -235,10 +236,13 @@ err:
>>>>>>>  }
>>>>>>>  #endif /* defined(CONFIG_DEBUG_FS) */
>>>>>>>
>>>>>>> +static void mci_send_cmd(struct dw_mci_slot *slot, u32 cmd, u32 arg);
>>>>>>> +
>>>>>>>  static u32 dw_mci_prepare_command(struct mmc_host *mmc, struct mmc_command *cmd)
>>>>>>>  {
>>>>>>>       struct mmc_data *data;
>>>>>>>       struct dw_mci_slot *slot = mmc_priv(mmc);
>>>>>>> +     struct dw_mci *host = slot->host;
>>>>>>>       const struct dw_mci_drv_data *drv_data = slot->host->drv_data;
>>>>>>>       u32 cmdr;
>>>>>>>       cmd->error = -EINPROGRESS;
>>>>>>> @@ -254,6 +258,32 @@ static u32 dw_mci_prepare_command(struct mmc_host *mmc, struct mmc_command
>>>>> *cmd)
>>>>>>>       else if (cmd->opcode != MMC_SEND_STATUS && cmd->data)
>>>>>>>               cmdr |= SDMMC_CMD_PRV_DAT_WAIT;
>>>>>>>
>>>>>>> +     if (cmd->opcode == SD_SWITCH_VOLTAGE) {
>>>>>>> +             u32 clk_en_a;
>>>>>>> +
>>>>>>> +             /* Special bit makes CMD11 not die */
>>>>>>> +             cmdr |= SDMMC_CMD_VOLT_SWITCH;
>>>>>>> +
>>>>>>> +             /* Change state to continue to handle CMD11 weirdness */
>>>>>>> +             WARN_ON(slot->host->state != STATE_SENDING_CMD);
>>>>>>> +             slot->host->state = STATE_SENDING_CMD11;
>>>>>>> +
>>>>>>> +             /*
>>>>>>> +              * We need to disable clock stop while doing voltage switch
>>>>>>> +              * according to 7.4.1.2 Voltage Switch Normal Scenario.
>>>>>>> +              *
>>>>>>> +              * It's assumed that by the next time the CLKENA is updated
>>>>>>> +              * (when we set the clock next) that the voltage change will
>>>>>>> +              * be over, so we don't bother setting any bits to synchronize
>>>>>>> +              * with dw_mci_setup_bus().
>>>>>>> +              */
>>>>>>> +             clk_en_a = mci_readl(host, CLKENA);
>>>>>>> +             clk_en_a &= ~(SDMMC_CLKEN_LOW_PWR << slot->id);
>>>>>>> +             mci_writel(host, CLKENA, clk_en_a);
>>>>>>> +             mci_send_cmd(slot, SDMMC_CMD_UPD_CLK |
>>>>>>> +                          SDMMC_CMD_PRV_DAT_WAIT, 0);
>>>>>> dw_mci_disable_low_power() can be used here.
>>>>>
>>>>> Ah.  I guess we don't have that locally anymore.  Locally we have variants on:
>>>> I'm checking on cjb/mmc branch.
>>>>
>>>>> * https://patchwork.kernel.org/patch/3070311/
>>>>> * https://patchwork.kernel.org/patch/3070251/
>>>>> * https://patchwork.kernel.org/patch/3070221/
>>>>>
>>>>> ...which removed that function.  ...but I guess upstream never picked
>>>>> up those patches, huh?  Looking back it looks like you had some
>>>>> feedback and it needed another spin but somehow fell off my plate.  :(
>>>>>
>>>>> Maybe this is something Yuvaraj would like to pick up?
>>>> It's long ago. I remember that there is no progress since my last comment.
>>>> In case of patch "3070221", I want to pick up for next Kernel.
>>>
>>> Sounds like Yuvaraj has agreed to look at addressing your comments.  Thanks!
>>>
>>>
>>>>>>> +     }
>>>>>>> +
>>>>>>>       if (cmd->flags & MMC_RSP_PRESENT) {
>>>>>>>               /* We expect a response, so set this bit */
>>>>>>>               cmdr |= SDMMC_CMD_RESP_EXP;
>>>>>>> @@ -776,11 +806,15 @@ static void dw_mci_setup_bus(struct dw_mci_slot *slot, bool force_clkinit)
>>>>>>>       unsigned int clock = slot->clock;
>>>>>>>       u32 div;
>>>>>>>       u32 clk_en_a;
>>>>>>> +     u32 sdmmc_cmd_bits = SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT;
>>>>>>> +
>>>>>>> +     /* We must continue to set bit 28 in CMD until the change is complete */
>>>>>>> +     if (host->state == STATE_WAITING_CMD11_DONE)
>>>>>>> +             sdmmc_cmd_bits |= SDMMC_CMD_VOLT_SWITCH;
>>>>>> I didn't get the reason SDMMC_CMD_VOLT_SWITCH is needed during clock update(enable/disable)
>>>>>> Can you explain it in details?
>>>>>
>>>>> Simply put: if I didn't do this then the system hung during voltage
>>>>> switch.  It was not documented in the version of the IP manual that I
>>>>> had access to and it took me a bunch of digging / trial and error to
>>>>> figure this out.  Whatever this bit does internally it's important to
>>>>> set it while the voltage change is happening.  Note that this need was
>>>>> the whole reason for adding the extra state to the state machine.
>>>>>
>>>>> Perhaps Yuvaraj can try without it and I'd assume he'll report the
>>>>> same freeze.
>>>> Clarify the necessity of SDMMC_CMD_VOLT_SWITCH.
>>>> As far as I experience, we didn't apply this bit for several projects.
>>>
>>> I just tried this locally on our ChromeOS 3.8 kernel (which has quite
>>> a few backports and matches dw_mmc upstream pretty closely).  When I
>>> take out "sdmmc_cmd_bits |= SDMMC_CMD_VOLT_SWITCH" and bootup it fails
>>> to bringup WiFi chip.  It prints messages like:
>>>
>>> [    4.400842] mmc_host mmc1: Timeout sending command (cmd 0x202000
>>> arg 0x0 status 0x80202000)
>>>
>>> I will let Yuvaraj comment about his testing too.
>> To make it simple and verify, i have just enabled eMMC and SD channels
>> on 3.16-rc1 and tested.
>> Without "sdmmc_cmd_bits |= SDMMC_CMD_VOLT_SWITCH" bootup fails.
>> [    3.015971] mmc_host mmc1: Timeout sending command (cmd 0x202000
>> arg 0x0 status 0x80202000)
>> [    3.530974] mmc_host mmc1: Timeout sending command (cmd 0x202000
>> arg 0x0 status 0x80202000)
> Thank you for test.
> Hmm, it failed during clock update, right?.

I think that clock update or clock disable is failed, isn't?

> Can you check HLE interrupt at this time?
> I'll investigate for this.

If HLE interrupt is occured, we need to control this. 
I knew that Mr.Jeon had discussed about controlling HLE. 
Also i didn't reproduce the HLE error on my board.
so if this problem is related with HLE, I think we can find the one of HLE error case.


Best Regards,
Jaehoon Chung

> 
>>>
>>>
>>>>>>> +static int dw_mci_switch_voltage(struct mmc_host *mmc, struct mmc_ios *ios)
>>>>>>> +{
>>>>>>> +     struct dw_mci_slot *slot = mmc_priv(mmc);
>>>>>>> +     struct dw_mci *host = slot->host;
>>>>>>> +     u32 uhs;
>>>>>>> +     u32 v18 = SDMMC_UHS_18V << slot->id;
>>>>>>> +     int min_uv, max_uv;
>>>>>>> +     int ret;
>>>>>>> +
>>>>>>> +     /*
>>>>>>> +      * Program the voltage.  Note that some instances of dw_mmc may use
>>>>>>> +      * the UHS_REG for this.  For other instances (like exynos) the UHS_REG
>>>>>>> +      * does no harm but you need to set the regulator directly.  Try both.
>>>>>>> +      */
>>>>>>> +     uhs = mci_readl(host, UHS_REG);
>>>>>>> +     if (ios->signal_voltage == MMC_SIGNAL_VOLTAGE_330) {
>>>>>>> +             min_uv = 2700000;
>>>>>>> +             max_uv = 3600000;
>>>>>>> +             uhs &= ~v18;
>>>>>>> +     } else {
>>>>>>> +             min_uv = 1700000;
>>>>>>> +             max_uv = 1950000;
>>>>>>> +             uhs |= v18;
>>>>>>> +     }
>>>>>>> +     if (!IS_ERR(mmc->supply.vqmmc)) {
>>>>>>> +             ret = regulator_set_voltage(mmc->supply.vqmmc, min_uv, max_uv);
>>>>>>> +
>>>>>>> +             /*
>>>>>>> +              * Only complain if regulator claims that it's not in the 1.8V
>>>>>>> +              * range.  This avoids a bunch of errors in the case that
>>>>>>> +              * we've got a fixed 1.8V regulator but the core SD code still
>>>>>>> +              * thinks it ought to try to switch to 3.3 and then back to 1.8
>>>>>>> +              */
>>>>>>> +             if (ret) {
>>>>>> I think if ret is error, printing message and returning error is good.
>>>>>> Currently, just returning '0' though it fails.
>>>> Any feedback?
>>>
>>> Whoops, right.  I think you're right that in the case it warns it
>>> should also return the error code.  In the case it doesn't warn it
>>> shouldn't.  Also: possibly we should use a trick like the mmc core
>>> does and use "regulator_count_voltages" to figure out if we're going
>>> to be able to switch voltages...
>>>
>>>>>> Artificially adding SDMMC_INT_CMD_DONE to pending is needed to complete cmd handling?
>>>>>> Is there any reason?
>>>>>
>>>>> I don't remember this for sure since I wrote this a long time ago.
>>>>> Maybe it's not needed?  I may have just been modeling on other
>>>>> interrupts.
>>>> If there is no specific reason, please remove it.
>>>
>>> OK, we'll see how Yuvaraj's testing goes without it.  My incredibly
>>> brief testing in our local Chrome OS 3.8 tree shows that the WiFi is
>>> not detected properly if I comment out the line:
>>>   dw_mci_cmd_interrupt(host,
>>>     pending | SDMMC_INT_CMD_DONE);
>>>
>>> It may be sufficient to simply schedule the tasklet or to do some
>>> other subset of dw_mci_cmd_interrupt().  Yuvaraj can confirm on the
>>> latest kernel and also investigate further.
> How about completing CMD without SDMMC_INT_CMD_DONE?
> 
> Thanks,
> Seungwon Jeon
> 
> 

[PATCH 3/3] mmc: dw_mmc: Support voltage changes

[Yuvaraj Kumar]

On Fri, Jun 27, 2014 at 4:48 PM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
> Hi Yuvaraj,
>
> On Fri, June 27, 2014, Yuvaraj Kumar wrote:
>> On Thu, Jun 26, 2014 at 10:20 PM, Doug Anderson <dianders@chromium.org> wrote:
>> > Seungwon,
>> >
>> > On Thu, Jun 26, 2014 at 3:41 AM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
>> >> On Thu, June 26, 2014, Doug Anderson wrote:
>> >>> Seungwon,
>> >>>
>> >>> On Wed, Jun 25, 2014 at 6:08 AM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
>> >>> > On Mon, June 23, 2014, Yuvaraj Kumar C D wrote:
>> >>> >> Subject: [PATCH 3/3] mmc: dw_mmc: Support voltage changes
>> >>> >>
>> >>> >> From: Doug Anderson <dianders@chromium.org>
>> >>> >>
>> >>> >> For UHS cards we need the ability to switch voltages from 3.3V to
>> >>> >> 1.8V.  Add support to the dw_mmc driver to handle this.  Note that
>> >>> >> dw_mmc needs a little bit of extra code since the interface needs a
>> >>> >> special bit programmed to the CMD register while CMD11 is progressing.
>> >>> >> This means adding a few extra states to the state machine to track.
>> >>> >
>> >>> > Overall new additional states makes it complicated.
>> >>> > Can we do that in other way?
>> >>>
>> >>> That was the best I was able to figure out when I thought this
>> >>> through.  If you have ideas for doing it another way I'd imagine that
>> >>> Yuvaraj would be happy to take your feedback.
>> >> Let's clean up SDMMC_CMD_VOLT_SWITCH.
>> >> In turn, we may remove state-handling simply.
>> >>
>> >>>
>> >>>
>> >>> >> Signed-off-by: Doug Anderson <dianders@chromium.org>
>> >>> >> Signed-off-by: Yuvaraj Kumar C D <yuvaraj.cd@samsung.com>
>> >>> >>
>> >>> >> ---
>> >>> >>  drivers/mmc/host/dw_mmc.c  |  145 +++++++++++++++++++++++++++++++++++++++++---
>> >>> >>  drivers/mmc/host/dw_mmc.h  |    5 +-
>> >>> >>  include/linux/mmc/dw_mmc.h |    2 +
>> >>> >>  3 files changed, 142 insertions(+), 10 deletions(-)
>> >>> >>
>> >>> >> diff --git a/drivers/mmc/host/dw_mmc.c b/drivers/mmc/host/dw_mmc.c
>> >>> >> index e034bce..38eb548 100644
>> >>> >> --- a/drivers/mmc/host/dw_mmc.c
>> >>> >> +++ b/drivers/mmc/host/dw_mmc.c
>> >>> >> @@ -29,6 +29,7 @@
>> >>> >>  #include <linux/irq.h>
>> >>> >>  #include <linux/mmc/host.h>
>> >>> >>  #include <linux/mmc/mmc.h>
>> >>> >> +#include <linux/mmc/sd.h>
>> >>> >>  #include <linux/mmc/sdio.h>
>> >>> >>  #include <linux/mmc/dw_mmc.h>
>> >>> >>  #include <linux/bitops.h>
>> >>> >> @@ -235,10 +236,13 @@ err:
>> >>> >>  }
>> >>> >>  #endif /* defined(CONFIG_DEBUG_FS) */
>> >>> >>
>> >>> >> +static void mci_send_cmd(struct dw_mci_slot *slot, u32 cmd, u32 arg);
>> >>> >> +
>> >>> >>  static u32 dw_mci_prepare_command(struct mmc_host *mmc, struct mmc_command *cmd)
>> >>> >>  {
>> >>> >>       struct mmc_data *data;
>> >>> >>       struct dw_mci_slot *slot = mmc_priv(mmc);
>> >>> >> +     struct dw_mci *host = slot->host;
>> >>> >>       const struct dw_mci_drv_data *drv_data = slot->host->drv_data;
>> >>> >>       u32 cmdr;
>> >>> >>       cmd->error = -EINPROGRESS;
>> >>> >> @@ -254,6 +258,32 @@ static u32 dw_mci_prepare_command(struct mmc_host *mmc, struct mmc_command
>> >>> *cmd)
>> >>> >>       else if (cmd->opcode != MMC_SEND_STATUS && cmd->data)
>> >>> >>               cmdr |= SDMMC_CMD_PRV_DAT_WAIT;
>> >>> >>
>> >>> >> +     if (cmd->opcode == SD_SWITCH_VOLTAGE) {
>> >>> >> +             u32 clk_en_a;
>> >>> >> +
>> >>> >> +             /* Special bit makes CMD11 not die */
>> >>> >> +             cmdr |= SDMMC_CMD_VOLT_SWITCH;
>> >>> >> +
>> >>> >> +             /* Change state to continue to handle CMD11 weirdness */
>> >>> >> +             WARN_ON(slot->host->state != STATE_SENDING_CMD);
>> >>> >> +             slot->host->state = STATE_SENDING_CMD11;
>> >>> >> +
>> >>> >> +             /*
>> >>> >> +              * We need to disable clock stop while doing voltage switch
>> >>> >> +              * according to 7.4.1.2 Voltage Switch Normal Scenario.
>> >>> >> +              *
>> >>> >> +              * It's assumed that by the next time the CLKENA is updated
>> >>> >> +              * (when we set the clock next) that the voltage change will
>> >>> >> +              * be over, so we don't bother setting any bits to synchronize
>> >>> >> +              * with dw_mci_setup_bus().
>> >>> >> +              */
>> >>> >> +             clk_en_a = mci_readl(host, CLKENA);
>> >>> >> +             clk_en_a &= ~(SDMMC_CLKEN_LOW_PWR << slot->id);
>> >>> >> +             mci_writel(host, CLKENA, clk_en_a);
>> >>> >> +             mci_send_cmd(slot, SDMMC_CMD_UPD_CLK |
>> >>> >> +                          SDMMC_CMD_PRV_DAT_WAIT, 0);
>> >>> > dw_mci_disable_low_power() can be used here.
>> >>>
>> >>> Ah.  I guess we don't have that locally anymore.  Locally we have variants on:
>> >> I'm checking on cjb/mmc branch.
>> >>
>> >>> * https://patchwork.kernel.org/patch/3070311/
>> >>> * https://patchwork.kernel.org/patch/3070251/
>> >>> * https://patchwork.kernel.org/patch/3070221/
>> >>>
>> >>> ...which removed that function.  ...but I guess upstream never picked
>> >>> up those patches, huh?  Looking back it looks like you had some
>> >>> feedback and it needed another spin but somehow fell off my plate.  :(
>> >>>
>> >>> Maybe this is something Yuvaraj would like to pick up?
>> >> It's long ago. I remember that there is no progress since my last comment.
>> >> In case of patch "3070221", I want to pick up for next Kernel.
>> >
>> > Sounds like Yuvaraj has agreed to look at addressing your comments.  Thanks!
>> >
>> >
>> >>> >> +     }
>> >>> >> +
>> >>> >>       if (cmd->flags & MMC_RSP_PRESENT) {
>> >>> >>               /* We expect a response, so set this bit */
>> >>> >>               cmdr |= SDMMC_CMD_RESP_EXP;
>> >>> >> @@ -776,11 +806,15 @@ static void dw_mci_setup_bus(struct dw_mci_slot *slot, bool force_clkinit)
>> >>> >>       unsigned int clock = slot->clock;
>> >>> >>       u32 div;
>> >>> >>       u32 clk_en_a;
>> >>> >> +     u32 sdmmc_cmd_bits = SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT;
>> >>> >> +
>> >>> >> +     /* We must continue to set bit 28 in CMD until the change is complete */
>> >>> >> +     if (host->state == STATE_WAITING_CMD11_DONE)
>> >>> >> +             sdmmc_cmd_bits |= SDMMC_CMD_VOLT_SWITCH;
>> >>> > I didn't get the reason SDMMC_CMD_VOLT_SWITCH is needed during clock update(enable/disable)
>> >>> > Can you explain it in details?
>> >>>
>> >>> Simply put: if I didn't do this then the system hung during voltage
>> >>> switch.  It was not documented in the version of the IP manual that I
>> >>> had access to and it took me a bunch of digging / trial and error to
>> >>> figure this out.  Whatever this bit does internally it's important to
>> >>> set it while the voltage change is happening.  Note that this need was
>> >>> the whole reason for adding the extra state to the state machine.
>> >>>
>> >>> Perhaps Yuvaraj can try without it and I'd assume he'll report the
>> >>> same freeze.
>> >> Clarify the necessity of SDMMC_CMD_VOLT_SWITCH.
>> >> As far as I experience, we didn't apply this bit for several projects.
>> >
>> > I just tried this locally on our ChromeOS 3.8 kernel (which has quite
>> > a few backports and matches dw_mmc upstream pretty closely).  When I
>> > take out "sdmmc_cmd_bits |= SDMMC_CMD_VOLT_SWITCH" and bootup it fails
>> > to bringup WiFi chip.  It prints messages like:
>> >
>> > [    4.400842] mmc_host mmc1: Timeout sending command (cmd 0x202000
>> > arg 0x0 status 0x80202000)
>> >
>> > I will let Yuvaraj comment about his testing too.
>> To make it simple and verify, i have just enabled eMMC and SD channels
>> on 3.16-rc1 and tested.
>> Without "sdmmc_cmd_bits |= SDMMC_CMD_VOLT_SWITCH" bootup fails.
>> [    3.015971] mmc_host mmc1: Timeout sending command (cmd 0x202000
>> arg 0x0 status 0x80202000)
>> [    3.530974] mmc_host mmc1: Timeout sending command (cmd 0x202000
>> arg 0x0 status 0x80202000)
> Thank you for test.
> Hmm, it failed during clock update, right?.
> Can you check HLE interrupt at this time?
> I'll investigate for this.
Yes, it failed during clock update.I am dumping the register contents
just before the dev_err statement.I can observe that 1st timeout does
not have HLE interrupt.But once we set bit 0 of UHS_REG subsequent
timeout's have HLE bit set.I am attaching the complete log.
>
>> >
>> >
>> >>> >> +static int dw_mci_switch_voltage(struct mmc_host *mmc, struct mmc_ios *ios)
>> >>> >> +{
>> >>> >> +     struct dw_mci_slot *slot = mmc_priv(mmc);
>> >>> >> +     struct dw_mci *host = slot->host;
>> >>> >> +     u32 uhs;
>> >>> >> +     u32 v18 = SDMMC_UHS_18V << slot->id;
>> >>> >> +     int min_uv, max_uv;
>> >>> >> +     int ret;
>> >>> >> +
>> >>> >> +     /*
>> >>> >> +      * Program the voltage.  Note that some instances of dw_mmc may use
>> >>> >> +      * the UHS_REG for this.  For other instances (like exynos) the UHS_REG
>> >>> >> +      * does no harm but you need to set the regulator directly.  Try both.
>> >>> >> +      */
>> >>> >> +     uhs = mci_readl(host, UHS_REG);
>> >>> >> +     if (ios->signal_voltage == MMC_SIGNAL_VOLTAGE_330) {
>> >>> >> +             min_uv = 2700000;
>> >>> >> +             max_uv = 3600000;
>> >>> >> +             uhs &= ~v18;
>> >>> >> +     } else {
>> >>> >> +             min_uv = 1700000;
>> >>> >> +             max_uv = 1950000;
>> >>> >> +             uhs |= v18;
>> >>> >> +     }
>> >>> >> +     if (!IS_ERR(mmc->supply.vqmmc)) {
>> >>> >> +             ret = regulator_set_voltage(mmc->supply.vqmmc, min_uv, max_uv);
>> >>> >> +
>> >>> >> +             /*
>> >>> >> +              * Only complain if regulator claims that it's not in the 1.8V
>> >>> >> +              * range.  This avoids a bunch of errors in the case that
>> >>> >> +              * we've got a fixed 1.8V regulator but the core SD code still
>> >>> >> +              * thinks it ought to try to switch to 3.3 and then back to 1.8
>> >>> >> +              */
>> >>> >> +             if (ret) {
>> >>> > I think if ret is error, printing message and returning error is good.
>> >>> > Currently, just returning '0' though it fails.
>> >> Any feedback?
>> >
>> > Whoops, right.  I think you're right that in the case it warns it
>> > should also return the error code.  In the case it doesn't warn it
>> > shouldn't.  Also: possibly we should use a trick like the mmc core
>> > does and use "regulator_count_voltages" to figure out if we're going
>> > to be able to switch voltages...
>> >
>> >>> > Artificially adding SDMMC_INT_CMD_DONE to pending is needed to complete cmd handling?
>> >>> > Is there any reason?
>> >>>
>> >>> I don't remember this for sure since I wrote this a long time ago.
>> >>> Maybe it's not needed?  I may have just been modeling on other
>> >>> interrupts.
>> >> If there is no specific reason, please remove it.
>> >
>> > OK, we'll see how Yuvaraj's testing goes without it.  My incredibly
>> > brief testing in our local Chrome OS 3.8 tree shows that the WiFi is
>> > not detected properly if I comment out the line:
>> >   dw_mci_cmd_interrupt(host,
>> >     pending | SDMMC_INT_CMD_DONE);
>> >
>> > It may be sufficient to simply schedule the tasklet or to do some
>> > other subset of dw_mci_cmd_interrupt().  Yuvaraj can confirm on the
>> > latest kernel and also investigate further.
> How about completing CMD without SDMMC_INT_CMD_DONE?
Without this tuning fails.
[    2.525284] mmc_host mmc0: Bus speed (slot 0) = 200000000Hz (slot
req 200000000Hz, actual 200000000HZ div = 0)
[  198.556573] random: nonblocking pool is initialized
[  240.226044] INFO: task kworker/u8:0:6 blocked for more than 120 seconds.
[  240.231276]       Not tainted 3.16.0-rc1-00010-g2f7a756-dirty #55
[  240.237359] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs"
disables this message.
[  240.245150] kworker/u8:0    D c03a58a0     0     6      2 0x00000000
[  240.251487] Workqueue: kmmcd mmc_rescan
[  240.255297] [<c03a58a0>] (__schedule) from [<c03a5330>]
(schedule_timeout+0x130/0x170)
[  240.263201] [<c03a5330>] (schedule_timeout) from [<c03a6540>]
(wait_for_common+0xbc/0x14c)
[  240.271441] [<c03a6540>] (wait_for_common) from [<c02ca6e4>]
(mmc_wait_for_req_done+0x6c/0xf0)
[  240.280031] [<c02ca6e4>] (mmc_wait_for_req_done) from [<c02e3f10>]
(dw_mci_exynos_execute_tuning+0x1d4/0x2c4)
[  240.289919] [<c02e3f10>] (dw_mci_exynos_execute_tuning) from
[<c02e0198>] (dw_mci_execute_tuning+0x54/0xb4)
[  240.299635] [<c02e0198>] (dw_mci_execute_tuning) from [<c02cf47c>]
(mmc_init_card+0xeec/0x14ac)
[  240.308309] [<c02cf47c>] (mmc_init_card) from [<c02cfc34>]
(mmc_attach_mmc+0x8c/0x160)
[  240.316202] [<c02cfc34>] (mmc_attach_mmc) from [<c02cca0c>]
(mmc_rescan+0x2b0/0x308)
[  240.323924] [<c02cca0c>] (mmc_rescan) from [<c0032f60>]
(process_one_work+0xf8/0x364)
[  240.331732] [<c0032f60>] (process_one_work) from [<c0033800>]
(worker_thread+0x50/0x5a0)
[  240.339799] [<c0033800>] (worker_thread) from [<c0038e7c>]
(kthread+0xd8/0xf0)
[  240.346999] [<c0038e7c>] (kthread) from [<c000e438>]
(ret_from_fork+0x14/0x3c)

>
> Thanks,
> Seungwon Jeon
>
-------------- next part --------------
[  360.231140] INFO: task swapper/0:1 blocked for more than 120 seconds.
[  360.236113]       Not tainted 3.16.0-rc1-00010-g2f7a756-dirty #52
[  360.242183] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[  360.249991] swapper/0       D c03a5858     0     1      0 0x00000000
[  360.256329] [<c03a5858>] (__schedule) from [<c03a52c8>] (schedule_timeout+0x110/0x170)
[  360.264222] [<c03a52c8>] (schedule_timeout) from [<c0028540>] (msleep+0x2c/0x38)
[  360.271596] [<c0028540>] (msleep) from [<c0503394>] (prepare_namespace+0x13c/0x1d4)
[  360.279229] [<c0503394>] (prepare_namespace) from [<c0502d54>] (kernel_init_freeable+0x1cc/0x1dc)
[  360.288077] [<c0502d54>] (kernel_init_freeable) from [<c03a0bcc>] (kernel_init+0xc/0xe8)
[  360.296145] [<c03a0bcc>] (kernel_init) from [<c000e438>] (ret_from_fork+0x14/0x3c)
[  480.231101] INFO: task swapper/0:1 blocked for more than 120 seconds.
[  480.236069]       Not tainted 3.16.0-rc1-00010-g2f7a756-dirty #52
[  480.242138] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[  480.249949] swapper/0       D c03a5858     0     1      0 0x00000000
[  480.256284] [<c03a5858>] (__schedule) from [<c03a52c8>] (schedule_timeout+0x110/0x170)
[  480.264178] [<c03a52c8>] (schedule_timeout) from [<c0028540>] (msleep+0x2c/0x38)
[  480.271551] [<c0028540>] (msleep) from [<c0503394>] (prepare_namespace+0x13c/0x1d4)
[  480.279185] [<c0503394>] (prepare_namespace) from [<c0502d54>] (kernel_init_freeable+0x1cc/0x1dc)
[  480.288033] [<c0502d54>] (kernel_init_freeable) from [<c03a0bcc>] (kernel_init+0xc/0xe8)
[  480.296101] [<c03a0bcc>] (kernel_init) from [<c000e438>] (ret_from_fork+0x14/0x3c)
[  600.231135] INFO: task swapper/0:1 blocked for more than 120 seconds.
[  600.236103]       Not tainted 3.16.0-rc1-00010-g2f7a756-dirty #52
[  600.242173] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[  600.249981] swapper/0       D c03a5858     0     1      0 0x00000000
[  600.256319] [<c03a5858>] (__schedule) from [<c03a52c8>] (schedule_timeout+0x110/0x170)
[  600.264212] [<c03a52c8>] (schedule_timeout) from [<c0028540>] (msleep+0x2c/0x38)
[  600.271585] [<c0028540>] (msleep) from [<c0503394>] (prepare_namespace+0x13c/0x1d4)
[  600.279219] [<c0503394>] (prepare_namespace) from [<c0502d54>] (kernel_init_freeable+0x1cc/0x1dc)
[  600.288068] [<c0502d54>] (kernel_init_freeable) from [<c03a0bcc>] (kernel_init+0xc/0xe8)
[  600.296136] [<c03a0bcc>] (kernel_init) from [<c000e438>] (ret_from_fork+0x14/0x3c)


U-Boot 2013.04 (Feb 21 2014 - 15:29:10) for Peach

CPU:	Exynos5420 at 900MHz

Board: Google Peach Pit, rev 6.0
I2C:   ready
DRAM:  2 GiB
PMIC max77802-pmic initialized
CPU:	Exynos5420 at 1800MHz
TPS65090 PMIC EC init
MMC:   EXYNOS DWMMC: 0, EXYNOS DWMMC: 1
SF: Detected W25Q32DW with page size 4 KiB, total 32 MiB
In:    cros-ec-keyb
Out:   lcd
Err:   lcd
SF: Detected W25Q32DW with page size 4 KiB, total 32 MiB
ELOG: Event(17) added with size 13
Net:   No ethernet found.
Hit any key to stop autoboot:  3 \b\b\b 2 \b\b\b 1 \b\b\b 0 
mmc1 is current device
2935183 bytes read in 158 ms (17.7 MiB/s)
## Loading kernel from FIT Image at 20008000 ...
   Using 'conf at 1' configuration
   Trying 'kernel at 1' kernel subimage
     Description:  unavailable
     Type:         Kernel Image (no loading done)
     Compression:  uncompressed
     Data Start:   0x200080a4
     Data Size:    2893976 Bytes = 2.8 MiB
   Verifying Hash Integrity ... OK
## Loading fdt from FIT Image at 20008000 ...
   Using 'conf at 1' configuration
   Trying 'fdt at 1' fdt subimage
     Description:  exynos5420-peach-pit.dtb
     Type:         Flat Device Tree
     Compression:  uncompressed
     Data Start:   0x202ca9f4
     Data Size:    40098 Bytes = 39.2 KiB
     Architecture: ARM
   Verifying Hash Integrity ... OK
   Booting using the fdt blob at 0x202ca9f4
   XIP Kernel Image (no loading done) ... OK
   Loading Device Tree to 3ffe6000, end 3ffff1d1 ... OK
boot_kernel.c: ft_board_setup: warning: g_crossystem_data is NULL

Starting kernel ...

Timer summary in microseconds:
       Mark    Elapsed  Stage
          0          0  reset
    114,935    114,935  board_init_f
    158,743     43,808  board_init_r
    245,247     86,504  id=64
    247,201      1,954  main_loop
  3,720,769  3,473,568  bootm_start
  3,720,770          1  id=1
  3,725,417      4,647  id=100
  3,725,420          3  id=101
  3,725,420          0  id=102
  3,728,731      3,311  id=110
  3,753,507     24,776  id=105
  3,753,508          1  id=106
  3,753,512          4  id=107
  3,753,513          1  id=109
  3,753,513          0  id=108
  3,757,905      4,392  id=90
  3,757,907          2  id=92
  3,757,907          0  id=91
  3,787,512     29,605  id=95
  3,787,515          3  id=96
  3,787,516          1  id=97
  3,787,518          2  id=98
  3,787,519          1  id=99
  3,796,747      9,228  id=7
  3,808,661     11,914  id=15
  3,811,322      2,661  start_kernel

Accumulated time:
                 6,921  SPI read
[    0.000000] Booting Linux on physical CPU 0x0
[    0.000000] Linux version 3.16.0-rc1-00010-g2f7a756-dirty (chromeos at yuvaraj) (gcc version 4.8.3 20140106 (prerelease) (crosstool-NG linaro-1.13.1-4.8-2014.01 - Linaro GCC 2013.11) ) #53 SMP PREEMPT Tue Jul 1 14:55:34 IST 2014
[    0.000000] CPU: ARMv7 Processor [412fc0f3] revision 3 (ARMv7), cr=10c5387d
[    0.000000] CPU: PIPT / VIPT nonaliasing data cache, PIPT instruction cache
[    0.000000] Machine model: Google Peach Pit Rev 6+
[    0.000000] Memory policy: Data cache writealloc
[    0.000000] On node 0 totalpages: 524288
[    0.000000] free_area_init_node: node 0, pgdat c0583300, node_mem_map ee7f4000
[    0.000000]   Normal zone: 1520 pages used for memmap
[    0.000000]   Normal zone: 0 pages reserved
[    0.000000]   Normal zone: 194560 pages, LIFO batch:31
[    0.000000]   HighMem zone: 2576 pages used for memmap
[    0.000000]   HighMem zone: 329728 pages, LIFO batch:31
[    0.000000] PERCPU: Embedded 7 pages/cpu @ee79f000 s7104 r8192 d13376 u32768
[    0.000000] pcpu-alloc: s7104 r8192 d13376 u32768 alloc=8*4096
[    0.000000] pcpu-alloc: [0] 0 [0] 1 [0] 2 [0] 3 
[    0.000000] Built 1 zonelists in Zone order, mobility grouping on.  Total pages: 522768
[    0.000000] Kernel command line: console=ttySAC3,115200 debug no_console_suspend root=/dev/mmcblk1p3 rootwait rw no_console_suspend clk_ignore_unused
[    0.000000] PID hash table entries: 4096 (order: 2, 16384 bytes)
[    0.000000] Dentry cache hash table entries: 131072 (order: 7, 524288 bytes)
[    0.000000] Inode-cache hash table entries: 65536 (order: 6, 262144 bytes)
[    0.000000] Memory: 2073612K/2097152K available (3849K kernel code, 255K rwdata, 1244K rodata, 258K init, 271K bss, 23540K reserved, 1318912K highmem)
[    0.000000] Virtual kernel memory layout:
[    0.000000]     vector  : 0xffff0000 - 0xffff1000   (   4 kB)
[    0.000000]     fixmap  : 0xffc00000 - 0xffe00000   (2048 kB)
[    0.000000]     vmalloc : 0xf0000000 - 0xff000000   ( 240 MB)
[    0.000000]     lowmem  : 0xc0000000 - 0xef800000   ( 760 MB)
[    0.000000]     pkmap   : 0xbfe00000 - 0xc0000000   (   2 MB)
[    0.000000]     modules : 0xbf000000 - 0xbfe00000   (  14 MB)
[    0.000000]       .text : 0xc0008000 - 0xc05017bc   (5094 kB)
[    0.000000]       .init : 0xc0502000 - 0xc0542bc0   ( 259 kB)
[    0.000000]       .data : 0xc0544000 - 0xc0583ea0   ( 256 kB)
[    0.000000]        .bss : 0xc0583eac - 0xc05c7d40   ( 272 kB)
[    0.000000] SLUB: HWalign=64, Order=0-3, MinObjects=0, CPUs=4, Nodes=1
[    0.000000] Preemptible hierarchical RCU implementation.
[    0.000000] 	RCU restricting CPUs from NR_CPUS=8 to nr_cpu_ids=4.
[    0.000000] RCU: Adjusting geometry for rcu_fanout_leaf=16, nr_cpu_ids=4
[    0.000000] NR_IRQS:16 nr_irqs:16 16
[    0.000000] L2C: failed to init: -19
[    0.000004] sched_clock: 64 bits at 24MHz, resolution 41ns, wraps every 2863311519744ns
[    0.000196] Console: colour dummy device 80x30
[    0.000211] Calibrating delay loop... 3594.64 BogoMIPS (lpj=8986624)
[    0.030003] pid_max: default: 32768 minimum: 301
[    0.030103] Mount-cache hash table entries: 2048 (order: 1, 8192 bytes)
[    0.030111] Mountpoint-cache hash table entries: 2048 (order: 1, 8192 bytes)
[    0.030434] CPU: Testing write buffer coherency: ok
[    0.030577] CPU0: update cpu_capacity 1024
[    0.030584] CPU0: thread -1, cpu 0, socket 0, mpidr 80000000
[    0.030681] Setting up static identity map for 0x203aa130 - 0x203aa188
[    0.030779] ARM CCI driver probed
[    0.030814] Exynos MCPM support installed
[    0.060235] CPU1: Booted secondary processor
[    0.090006] CPU1: update cpu_capacity 1024
[    0.090010] CPU1: thread -1, cpu 1, socket 0, mpidr 80000001
[    0.100228] CPU2: Booted secondary processor
[    0.130007] CPU2: update cpu_capacity 1024
[    0.130011] CPU2: thread -1, cpu 2, socket 0, mpidr 80000002
[    0.140237] CPU3: Booted secondary processor
[    0.170008] CPU3: update cpu_capacity 1024
[    0.170011] CPU3: thread -1, cpu 3, socket 0, mpidr 80000003
[    0.170070] Brought up 4 CPUs
[    0.170088] SMP: Total of 4 processors activated.
[    0.170093] CPU: All CPU(s) started in SVC mode.
[    0.170441] devtmpfs: initialized
[    0.175525] VFP support v0.3: implementor 41 architecture 4 part 30 variant f rev 0
[    0.175844] pinctrl core: initialized pinctrl subsystem
[    0.176196] regulator-dummy: no parameters
[    0.188219] NET: Registered protocol family 16
[    0.188367] DMA: preallocated 256 KiB pool for atomic coherent allocations
[    0.191769] gpiochip_add: registered GPIOs 0 to 7 on device: gpy7
[    0.191778] gpiochip_add: registered GPIOs 8 to 15 on device: gpx0
[    0.191787] gpiochip_add: registered GPIOs 16 to 23 on device: gpx1
[    0.191794] gpiochip_add: registered GPIOs 24 to 31 on device: gpx2
[    0.191802] gpiochip_add: registered GPIOs 32 to 39 on device: gpx3
[    0.192475] gpiochip_add: registered GPIOs 40 to 47 on device: gpc0
[    0.192484] gpiochip_add: registered GPIOs 48 to 55 on device: gpc1
[    0.192491] gpiochip_add: registered GPIOs 56 to 62 on device: gpc2
[    0.192498] gpiochip_add: registered GPIOs 63 to 66 on device: gpc3
[    0.192505] gpiochip_add: registered GPIOs 67 to 68 on device: gpc4
[    0.192513] gpiochip_add: registered GPIOs 69 to 76 on device: gpd1
[    0.192520] gpiochip_add: registered GPIOs 77 to 82 on device: gpy0
[    0.192526] gpiochip_add: registered GPIOs 83 to 86 on device: gpy1
[    0.192534] gpiochip_add: registered GPIOs 87 to 92 on device: gpy2
[    0.192541] gpiochip_add: registered GPIOs 93 to 100 on device: gpy3
[    0.192548] gpiochip_add: registered GPIOs 101 to 108 on device: gpy4
[    0.192555] gpiochip_add: registered GPIOs 109 to 116 on device: gpy5
[    0.192562] gpiochip_add: registered GPIOs 117 to 124 on device: gpy6
[    0.193044] gpiochip_add: registered GPIOs 125 to 132 on device: gpe0
[    0.193052] gpiochip_add: registered GPIOs 133 to 134 on device: gpe1
[    0.193060] gpiochip_add: registered GPIOs 135 to 140 on device: gpf0
[    0.193067] gpiochip_add: registered GPIOs 141 to 148 on device: gpf1
[    0.193073] gpiochip_add: registered GPIOs 149 to 156 on device: gpg0
[    0.193080] gpiochip_add: registered GPIOs 157 to 164 on device: gpg1
[    0.193087] gpiochip_add: registered GPIOs 165 to 166 on device: gpg2
[    0.193095] gpiochip_add: registered GPIOs 167 to 170 on device: gpj4
[    0.193533] gpiochip_add: registered GPIOs 171 to 178 on device: gpa0
[    0.193542] gpiochip_add: registered GPIOs 179 to 184 on device: gpa1
[    0.193549] gpiochip_add: registered GPIOs 185 to 192 on device: gpa2
[    0.193556] gpiochip_add: registered GPIOs 193 to 197 on device: gpb0
[    0.193563] gpiochip_add: registered GPIOs 198 to 202 on device: gpb1
[    0.193570] gpiochip_add: registered GPIOs 203 to 206 on device: gpb2
[    0.193578] gpiochip_add: registered GPIOs 207 to 214 on device: gpb3
[    0.193584] gpiochip_add: registered GPIOs 215 to 216 on device: gpb4
[    0.193592] gpiochip_add: registered GPIOs 217 to 224 on device: gph0
[    0.194166] gpiochip_add: registered GPIOs 225 to 231 on device: gpz
[    0.195182] exynos-audss-clk 3810000.audss-clock-controller: setup completed
[    0.202280] EXYNOS5420 PMU Initialized
[    0.225885] of_get_named_gpiod_flags exited with status 0
[    0.225989] P5.0V_USB3CON0: 5000 mV 
[    0.226120] of_get_named_gpiod_flags exited with status 0
[    0.226218] P5.0V_USB3CON1: 5000 mV 
[    0.226338] of_get_named_gpiod_flags exited with status 0
[    0.226435] mask-tpm-reset : no parameters
[    0.226534] of_get_named_gpiod_flags: can't parse gpios property of node '/fixed-regulator[0]'
[    0.226619] vbat-supply: no parameters
[    0.227865] SCSI subsystem initialized
[    0.228201] usbcore: registered new interface driver usbfs
[    0.228278] usbcore: registered new interface driver hub
[    0.228365] usbcore: registered new device driver usb
[    0.228679] s3c-i2c 12c80000.i2c: slave address 0x50
[    0.228691] s3c-i2c 12c80000.i2c: bus frequency set to 65 KHz
[    0.228819] s3c-i2c 12c80000.i2c: i2c-2: S3C I2C adapter
[    0.229783] Switched to clocksource mct-frc
[    0.239452] NET: Registered protocol family 2
[    0.239861] TCP established hash table entries: 8192 (order: 3, 32768 bytes)
[    0.239909] TCP bind hash table entries: 8192 (order: 5, 163840 bytes)
[    0.240000] TCP: Hash tables configured (established 8192 bind 8192)
[    0.240042] TCP: reno registered
[    0.240051] UDP hash table entries: 512 (order: 2, 24576 bytes)
[    0.240074] UDP-Lite hash table entries: 512 (order: 2, 24576 bytes)
[    0.240203] NET: Registered protocol family 1
[    0.241230] futex hash table entries: 1024 (order: 4, 65536 bytes)
[    0.251484] ROMFS MTD (C) 2007 Red Hat, Inc.
[    0.251664] msgmni has been set to 1474
[    0.252030] bounce: pool size: 64 pages
[    0.252039] io scheduler noop registered
[    0.252048] io scheduler deadline registered
[    0.252394] io scheduler cfq registered (default)
[    0.254527] dma-pl330 3880000.adma: Loaded driver for PL330 DMAC-2364208
[    0.254536] dma-pl330 3880000.adma: 	DBUFF-4x8bytes Num_Chans-6 Num_Peri-16 Num_Events-6
[    0.257230] dma-pl330 121a0000.pdma: Loaded driver for PL330 DMAC-2364208
[    0.257239] dma-pl330 121a0000.pdma: 	DBUFF-32x4bytes Num_Chans-8 Num_Peri-32 Num_Events-32
[    0.259761] dma-pl330 121b0000.pdma: Loaded driver for PL330 DMAC-2364208
[    0.259770] dma-pl330 121b0000.pdma: 	DBUFF-32x4bytes Num_Chans-8 Num_Peri-32 Num_Events-32
[    0.260576] dma-pl330 10800000.mdma: Loaded driver for PL330 DMAC-2364208
[    0.260585] dma-pl330 10800000.mdma: 	DBUFF-64x8bytes Num_Chans-8 Num_Peri-1 Num_Events-32
[    0.323551] Serial: 8250/16550 driver, 4 ports, IRQ sharing disabled
[    0.324867] 12c00000.serial: ttySAC0 at MMIO 0x12c00000 (irq = 83, base_baud = 0) is a S3C6400/10
[    0.325182] 12c10000.serial: ttySAC1 at MMIO 0x12c10000 (irq = 84, base_baud = 0) is a S3C6400/10
[    0.325489] 12c20000.serial: ttySAC2 at MMIO 0x12c20000 (irq = 85, base_baud = 0) is a S3C6400/10
[    0.325802] 12c30000.serial: ttySAC3 at MMIO 0x12c30000 (irq = 86, base_baud = 0) is a S3C6400/10
[    1.218428] console [ttySAC3] enabled
[    1.229272] brd: module loaded
[    1.233958] loop: module loaded
[    1.236558] of_get_named_gpiod_flags exited with status 0
[    1.241505] cros-ec-spi spi2.0: Chrome EC (SPI)
[    1.245893] usbcore: registered new interface driver asix
[    1.250963] usbcore: registered new interface driver ax88179_178a
[    1.257022] usbcore: registered new interface driver cdc_ether
[    1.262858] usbcore: registered new interface driver smsc75xx
[    1.268577] usbcore: registered new interface driver smsc95xx
[    1.274292] usbcore: registered new interface driver net1080
[    1.279929] usbcore: registered new interface driver cdc_subset
[    1.285828] usbcore: registered new interface driver zaurus
[    1.291402] usbcore: registered new interface driver cdc_ncm
[    1.297421] usb at 12000000 supply vdd33 not found, using dummy regulator
[    1.303517] usb at 12000000 supply vdd10 not found, using dummy regulator
[    1.310228] platform 12000000.dwc3: Driver dwc3 requests probe deferral
[    1.316791] usb at 12400000 supply vdd33 not found, using dummy regulator
[    1.323129] usb at 12400000 supply vdd10 not found, using dummy regulator
[    1.329836] platform 12400000.dwc3: Driver dwc3 requests probe deferral
[    1.336239] ehci_hcd: USB 2.0 'Enhanced' Host Controller (EHCI) Driver
[    1.342692] ehci-exynos: EHCI EXYNOS driver
[    1.346967] of_get_named_gpiod_flags: can't parse gpios property of node '/usb at 12110000[0]'
[    1.355197] exynos-ehci 12110000.usb: no usb2 phy configured
[    1.360825] platform 12110000.usb: Driver exynos-ehci requests probe deferral
[    1.368067] usbcore: registered new interface driver usb-storage
[    1.374492] mousedev: PS/2 mouse device common for all mice
[    1.379894] input: cros-ec-spi as /devices/12d40000.spi/spi_master/spi2/spi2.0/cros-ec-keyb.1/input/input0
[    1.389624] s3c-rtc 101e0000.rtc: rtc disabled, re-enabling
[    1.394891] s3c-rtc 101e0000.rtc: rtc core: registered s3c as rtc0
[    1.400821] s3c-rtc 101e0000.rtc: warning: invalid RTC value so initializing it
[    1.408986] of_get_named_gpiod_flags exited with status 0
[    1.413480] of_get_named_gpiod_flags exited with status 0
[    1.418861] of_get_named_gpiod_flags exited with status 0
[    1.424234] of_get_named_gpiod_flags exited with status 0
[    1.429617] of_get_named_gpiod_flags exited with status 0
[    1.434993] of_get_named_gpiod_flags exited with status 0
[    1.440377] of_get_named_gpiod_flags exited with status 0
[    1.445751] of_get_named_gpiod_flags exited with status 0
[    1.451343] max77xxx 4-0009: device found
[    1.455624] max77xxx 4-0009: no regulator-op-mode property property at LDO2
[    1.462165] max77xxx 4-0009: no regulator-op-mode property property at LDO4
[    1.469161] max77xxx 4-0009: no regulator-op-mode property property at LDO7
[    1.476474] max77xxx 4-0009: no regulator-op-mode property property at LDO18
[    1.483023] max77xxx 4-0009: no regulator-op-mode property property at LDO19
[    1.490094] max77xxx 4-0009: no regulator-op-mode property property at LDO21
[    1.497073] max77xxx 4-0009: no regulator-op-mode property property at LDO23
[    1.504093] max77xxx 4-0009: no regulator-op-mode property property at LDO24
[    1.511130] max77xxx 4-0009: no regulator-op-mode property property at LDO25
[    1.518149] max77xxx 4-0009: no regulator-op-mode property property at LDO26
[    1.525194] max77xxx 4-0009: no regulator-op-mode property property at LDO27
[    1.532201] max77xxx 4-0009: no regulator-op-mode property property at LDO28
[    1.539228] max77xxx 4-0009: no regulator-op-mode property property at LDO29
[    1.546309] max77xxx 4-0009: no regulator-op-mode property property at LDO32
[    1.553281] max77xxx 4-0009: no regulator-op-mode property property at LDO33
[    1.560309] max77xxx 4-0009: no regulator-op-mode property property at LDO34
[    1.567335] max77xxx 4-0009: no regulator-op-mode property property at LDO35
[    1.574924] max77xxx 4-0009: no regulator-op-mode property property at EN32KHZ_AP
[    1.581822] max77xxx 4-0009: no regulator-op-mode property property at EN32KHZ_CP
[    1.589543] vdd_1v0: 1000 mV 
[    1.592477] vdd_1v2_2: 1200 mV 
[    1.595686] vdd_1v8_3: 1800 mV 
[    1.598693] vdd_sd: 1800 <--> 2800 mV at 2800 mV 
[    1.603481] vdd_1v8_5: 1800 mV 
[    1.606606] vdd_1v8_6: 1800 mV 
[    1.609631] vdd_1v8_7: 1800 mV 
[    1.612863] vdd_ldo8: 1000 mV 
[    1.615895] vdd_ldo9: 1800 mV 
[    1.618909] vdd_ldo10: 1800 mV 
[    1.622054] vdd_ldo11: 1800 mV 
[    1.625181] vdd_ldo12: 3000 mV 
[    1.628276] vdd_ldo13: 1800 mV 
[    1.631426] vdd_ldo14: 1800 mV 
[    1.634524] vdd_ldo15: 1000 mV 
[    1.637674] vdd_g3ds: 900 <--> 1400 mV at 1000 mV 
[    1.642354] ldo_18: 1800 mV 
[    1.645212] ldo_19: 1800 mV 
[    1.648067] ldo_20: 1800 mV 
[    1.650936] ldo_21: 2800 mV 
[    1.653786] ld0_23: 3300 mV 
[    1.656673] ldo_24: 2800 mV 
[    1.659510] ldo_25: 3300 mV 
[    1.662477] ldo_26: 1200 mV 
[    1.665259] ldo_27: 1200 mV 
[    1.668104] ldo_28: 1800 mV 
[    1.670984] ldo_29: 1800 mV 
[    1.673910] vdd_mifs: 900 <--> 1300 mV at 1000 mV 
[    1.678615] ldo_32: 3000 mV 
[    1.681475] ldo_33: 2800 mV 
[    1.684321] ldo_34: 3000 mV 
[    1.687209] ldo_35: 1200 mV 
[    1.690485] vdd_mif: 800 <--> 1300 mV at 1000 mV 
[    1.695157] vdd_arm: 800 <--> 1500 mV at 1262 mV 
[    1.699826] vdd_int: 800 <--> 1400 mV at 1000 mV 
[    1.704487] vdd_g3d: 700 <--> 1400 mV at 1000 mV 
[    1.709013] vdd_1v2: 1200 mV 
[    1.712150] vdd_kfc: 800 <--> 1500 mV at 1000 mV 
[    1.716556] vdd_1v35: 1350 mV 
[    1.719673] vdd_emmc: 2850 mV 
[    1.722634] vdd_2v: 2000 mV 
[    1.725510] vdd_1v8: 1800 mV 
[    1.728283] en32khz_ap: no parameters
[    1.731792] en32khz_cp: no parameters
[    1.737923] tpm_i2c_infineon 9-0020: 1.2 TPM (device-id 0x1A)
[    1.764175] tpm_i2c_infineon 9-0020: Issuing TPM_STARTUP
[    1.930850] tps65090 0-0048: No cache defaults, reading back from HW
[    1.942070] of_get_named_gpiod_flags: can't parse gpios property of node '/spi at 12d40000/cros-ec at 0/i2c-tunnel/power-regulator at 48[0]'
[    1.952419] of_get_named_gpiod_flags: can't parse gpios property of node '/spi at 12d40000/cros-ec at 0/i2c-tunnel/power-regulator at 48[0]'
[    1.964221] of_get_named_gpiod_flags: can't parse gpios property of node '/spi at 12d40000/cros-ec at 0/i2c-tunnel/power-regulator at 48[0]'
[    1.976108] TPS65090_RAILSDCDC1: no parameters
[    1.980471] TPS65090_RAILSDCDC1: supplied by vbat-supply
[    1.989278] TPS65090_RAILSDCDC2: no parameters
[    1.992276] TPS65090_RAILSDCDC2: supplied by vbat-supply
[    2.001153] TPS65090_RAILSDCDC3: no parameters
[    2.004134] TPS65090_RAILSDCDC3: supplied by vbat-supply
[    2.012944] vcd_led: no parameters
[    2.014904] vcd_led: supplied by vbat-supply
[    2.033160] video_mid: no parameters
[    2.035298] video_mid: supplied by vbat-supply
[    2.053727] wwan_r: no parameters
[    2.055597] wwan_r: supplied by vbat-supply
[    2.063332] sdcard: no parameters
[    2.065211] sdcard: supplied by vbat-supply
[    2.072853] camout: no parameters
[    2.074709] camout: supplied by vbat-supply
[    2.082378] lcd_vdd: no parameters
[    2.084321] lcd_vdd: supplied by vbat-supply
[    2.102596] video_mid_1a: no parameters
[    2.104993] video_mid_1a: supplied by vbat-supply
[    2.113145] TPS65090_RAILSLDO1: no parameters
[    2.116061] TPS65090_RAILSLDO1: supplied by vbat-supply
[    2.121384] TPS65090_RAILSLDO2: no parameters
[    2.125599] TPS65090_RAILSLDO2: supplied by vbat-supply
[    2.131707] tps65090-charger tps65090-charger: Unable to get charger irq = -22
[    2.149999] tps65090-charger: probe of tps65090-charger failed with error -22
[    2.156234] device-mapper: ioctl: 4.27.0-ioctl (2013-10-30) initialised: dm-devel at redhat.com
[    2.164121] sdhci: Secure Digital Host Controller Interface driver
[    2.170240] sdhci: Copyright(c) Pierre Ossman
[    2.174712] Synopsys Designware Multimedia Card Interface Driver
[    2.180967] dwmmc_exynos 12200000.mmc: Using internal DMA controller.
[    2.186989] dwmmc_exynos 12200000.mmc: Version ID is 250a
[    2.192480] dwmmc_exynos 12200000.mmc: DW MMC controller at irq 107, 64 bit host data width, 64 deep fifo
[    2.201922] dwmmc_exynos 12200000.mmc: No vmmc regulator found
[    2.207718] dwmmc_exynos 12200000.mmc: No vqmmc regulator found
[    2.213616] of_get_named_gpiod_flags: can't parse gpios property of node '/mmc at 12200000[0]'
[    2.249863] dwmmc_exynos 12200000.mmc: 1 slots initialized
[    2.254040] dwmmc_exynos 12220000.mmc: Using internal DMA controller.
[    2.260311] dwmmc_exynos 12220000.mmc: Version ID is 250a
[    2.265803] dwmmc_exynos 12220000.mmc: DW MMC controller at irq 109, 64 bit host data width, 64 deep fifo
[    2.275314] dwmmc_exynos 12220000.mmc: Failed getting OCR mask: -22
[    2.281472] of_get_named_gpiod_flags: can't parse gpios property of node '/mmc at 12220000[0]'
[    2.289797] of_get_named_gpiod_flags: can't parse gpios property of node '/mmc at 12220000[0]'
[    2.292724] mmc0: BKOPS_EN bit is not set
[    2.294818] mmc_host mmc0: Bus speed (slot 0) = 50000000Hz (slot req 52000000Hz, actual 50000000HZ div = 0)
[    2.295201] mmc_host mmc0: Bus speed (slot 0) = 100000000Hz (slot req 52000000Hz, actual 50000000HZ div = 1)
[    2.295442] mmc0: new DDR MMC card at address 0001
[    2.296456] mmcblk0: mmc0:0001 MAG2GC 14.5 GiB 
[    2.296729] mmcblk0boot0: mmc0:0001 MAG2GC partition 1 4.00 MiB
[    2.296968] mmcblk0boot1: mmc0:0001 MAG2GC partition 2 4.00 MiB
[    2.297201] mmcblk0rpmb: mmc0:0001 MAG2GC partition 3 4.00 MiB
[    2.301775]  mmcblk0: p1 p2 p3 p4 p5 p6 p7
[    2.309324]  mmcblk0boot1: unknown partition table
[    2.311649]  mmcblk0boot0: unknown partition table
[    2.394847] dwmmc_exynos 12220000.mmc: 1 slots initialized
[    2.399552] usbcore: registered new interface driver usbhid
[    2.404439] usbhid: USB HID core driver
[    2.408354] TCP: cubic registered
[    2.412561] NET: Registered protocol family 17
[    2.418000] NET: Registered protocol family 15
[    2.421158] Registering SWP/SWPB emulation handler
[    2.426354] platform 12000000.dwc3: Driver dwc3 requests probe deferral
[    2.432445] platform 12400000.dwc3: Driver dwc3 requests probe deferral
[    2.439030] of_get_named_gpiod_flags: can't parse gpios property of node '/usb at 12110000[0]'
[    2.447354] exynos-ehci 12110000.usb: no usb2 phy configured
[    2.453092] platform 12110000.usb: Driver exynos-ehci requests probe deferral
[    2.460375] of_get_named_gpiod_flags exited with status 0
[    2.465468] gpio-18 (Power): gpiod_set_debounce: missing set() or set_debounce() operations
[    2.475279] input: gpio-keys as /devices/gpio-keys/input/input1
[    2.480647] s3c-rtc 101e0000.rtc: setting system clock to 2000-01-01 00:00:01 UTC (946684801)
[    2.480650] platform 12000000.dwc3: Driver dwc3 requests probe deferral
[    2.480979] platform 12400000.dwc3: Driver dwc3 requests probe deferral
[    2.481285] of_get_named_gpiod_flags: can't parse gpios property of node '/usb at 12110000[0]'
[    2.481321] exynos-ehci 12110000.usb: no usb2 phy configured
[    2.481361] platform 12110000.usb: Driver exynos-ehci requests probe deferral
[    2.685427] TPS65090_RAILSLDO2: disabling
[    2.687991] TPS65090_RAILSLDO1: disabling
[    2.705531] TPS65090_RAILSDCDC3: disabling
[    2.708187] TPS65090_RAILSDCDC2: disabling
[    2.712305] TPS65090_RAILSDCDC1: disabling
[    2.716674] vdd_1v2_2: disabling
[    2.719861] P5.0V_USB3CON1: disabling
[    2.723210] P5.0V_USB3CON0: disabling
[    2.726898] clk: Not disabling unused clocks
[    2.735415] Waiting for root device /dev/mmcblk1p3...
[    2.944793] dwmmc_exynos 12220000.mmc: : ============== REGISTER DUMP ==============
[    2.951066] dwmmc_exynos 12220000.mmc: : CTRL:	0x00000010
[    2.956445] dwmmc_exynos 12220000.mmc: : PWREN:	0x00000001
[    2.961910] dwmmc_exynos 12220000.mmc: : CLKDIV:	0x0000003f
[    2.967461] dwmmc_exynos 12220000.mmc: : CLKSRC:	0x00000000
[    2.973013] dwmmc_exynos 12220000.mmc: : CLKENA:	0x00000000
[    2.978566] dwmmc_exynos 12220000.mmc: : TMOUT:	0xffffffff
[    2.984031] dwmmc_exynos 12220000.mmc: : CTYPE:	0x00000000
[    2.989496] dwmmc_exynos 12220000.mmc: : BLKSIZ:	0x00000200
[    2.995048] dwmmc_exynos 12220000.mmc: : BYTCNT:	0x00000200
[    3.000600] dwmmc_exynos 12220000.mmc: : INTMSK:	0x0000b7ff
[    3.006152] dwmmc_exynos 12220000.mmc: : CMDARG:	0x00000000
[    3.011704] dwmmc_exynos 12220000.mmc: : CMD:	0x80202000
[    3.016995] dwmmc_exynos 12220000.mmc: : RESP0:	0x00000320
[    3.022461] dwmmc_exynos 12220000.mmc: : RESP1:	0x3ac57f80
[    3.027926] dwmmc_exynos 12220000.mmc: : RESP2:	0x5b590000
[    3.033391] dwmmc_exynos 12220000.mmc: : RESP3:	0x400e0032
[    3.038857] dwmmc_exynos 12220000.mmc: : MINTSTS:	0x00000000
[    3.044496] dwmmc_exynos 12220000.mmc: : RINTSTS:	0x00010000
[    3.050135] dwmmc_exynos 12220000.mmc: : STATUS:	0x00005a06
[    3.055686] dwmmc_exynos 12220000.mmc: : FIFOTH:	0x201f0020
[    3.061238] dwmmc_exynos 12220000.mmc: : CDETECT:	0x00000000
[    3.066877] dwmmc_exynos 12220000.mmc: : WRTPRT:	0x00000000
[    3.072429] dwmmc_exynos 12220000.mmc: : GPIO:	0x00000000
[    3.077808] dwmmc_exynos 12220000.mmc: : TCBCNT:	0x00000200
[    3.083359] dwmmc_exynos 12220000.mmc: : TBBCNT:	0x00000000
[    3.088911] dwmmc_exynos 12220000.mmc: : DEBNCE:	0x00ffffff
[    3.094463] dwmmc_exynos 12220000.mmc: : USRID:	0xffffffff
[    3.099930] dwmmc_exynos 12220000.mmc: : VERID:	0x5342250a
[    3.105394] dwmmc_exynos 12220000.mmc: : HCON:	0x00e42541
[    3.110772] dwmmc_exynos 12220000.mmc: : UHS_REG:	0x00000000
[    3.116413] dwmmc_exynos 12220000.mmc: : BMOD:	0x00000200
[    3.121790] dwmmc_exynos 12220000.mmc: : PLDMND:	0x00000000
[    3.127341] dwmmc_exynos 12220000.mmc: : DBADDR:	0x4dbbb000
[    3.132894] dwmmc_exynos 12220000.mmc: : IDSTS:	0x00000000
[    3.138359] dwmmc_exynos 12220000.mmc: : IDINTEN:	0x00000103
[    3.143998] dwmmc_exynos 12220000.mmc: : DSCADDR:	0x00000000
[    3.149636] dwmmc_exynos 12220000.mmc: : BUFADDR:	0x00000000
[    3.155273] dwmmc_exynos 12220000.mmc: : CLKSEL:	0x03030002
[    3.160827] dwmmc_exynos 12220000.mmc: : ================= CMD REG =================
[    3.168547] dwmmc_exynos 12220000.mmc: : read/write        : read
[    3.174620] dwmmc_exynos 12220000.mmc: : data expected     : 0
[    3.180432] dwmmc_exynos 12220000.mmc: : cmd index         : 0
[    3.186245] dwmmc_exynos 12220000.mmc: : ================ STATUS REG ===============
[    3.193965] dwmmc_exynos 12220000.mmc: : fifocount         : 0
[    3.199785] dwmmc_exynos 12220000.mmc: : response index    : 11
[    3.205677] dwmmc_exynos 12220000.mmc: : data state mc busy: 0
[    3.211488] dwmmc_exynos 12220000.mmc: : data busy         : 1
[    3.217301] dwmmc_exynos 12220000.mmc: : data 3 state      : 0
[    3.223113] dwmmc_exynos 12220000.mmc: : command fsm state : 0
[    3.228926] dwmmc_exynos 12220000.mmc: : fifo full         : 0
[    3.234738] dwmmc_exynos 12220000.mmc: : fifo empty        : 1
[    3.240550] dwmmc_exynos 12220000.mmc: : fifo tx watermark : 1
[    3.246362] dwmmc_exynos 12220000.mmc: : fifo rx watermark : 0
[    3.252174] dwmmc_exynos 12220000.mmc: : ============== STATUS DUMP ================
[    3.259894] dwmmc_exynos 12220000.mmc: : cmd_status:      0x00000000
[    3.266228] dwmmc_exynos 12220000.mmc: : data_status:     0x00000000
[    3.272561] dwmmc_exynos 12220000.mmc: : pending_events:  0x00000000
[    3.278893] dwmmc_exynos 12220000.mmc: : completed_events:0x00000001
[    3.285226] dwmmc_exynos 12220000.mmc: : state:           7
[    3.290778] dwmmc_exynos 12220000.mmc: : ===========================================
[    3.298500] mmc_host mmc1: Timeout sending command (cmd 0x202000 arg 0x0 status 0x80202000)
[    3.347603] random: nonblocking pool is initialized
[    3.814791] dwmmc_exynos 12220000.mmc: : ============== REGISTER DUMP ==============
[    3.821068] dwmmc_exynos 12220000.mmc: : CTRL:	0x00000010
[    3.826447] dwmmc_exynos 12220000.mmc: : PWREN:	0x00000001
[    3.831913] dwmmc_exynos 12220000.mmc: : CLKDIV:	0x0000003f
[    3.837465] dwmmc_exynos 12220000.mmc: : CLKSRC:	0x00000000
[    3.843015] dwmmc_exynos 12220000.mmc: : CLKENA:	0x00000000
[    3.848569] dwmmc_exynos 12220000.mmc: : TMOUT:	0xffffffff
[    3.854034] dwmmc_exynos 12220000.mmc: : CTYPE:	0x00000000
[    3.859499] dwmmc_exynos 12220000.mmc: : BLKSIZ:	0x00000200
[    3.865050] dwmmc_exynos 12220000.mmc: : BYTCNT:	0x00000200
[    3.870601] dwmmc_exynos 12220000.mmc: : INTMSK:	0x0000b7ff
[    3.876153] dwmmc_exynos 12220000.mmc: : CMDARG:	0x00000000
[    3.881707] dwmmc_exynos 12220000.mmc: : CMD:	0x80202000
[    3.886998] dwmmc_exynos 12220000.mmc: : RESP0:	0x00000320
[    3.892462] dwmmc_exynos 12220000.mmc: : RESP1:	0x3ac57f80
[    3.897929] dwmmc_exynos 12220000.mmc: : RESP2:	0x5b590000
[    3.903395] dwmmc_exynos 12220000.mmc: : RESP3:	0x400e0032
[    3.908858] dwmmc_exynos 12220000.mmc: : MINTSTS:	0x00001000
[    3.914498] dwmmc_exynos 12220000.mmc: : RINTSTS:	0x00011000
[    3.920136] dwmmc_exynos 12220000.mmc: : STATUS:	0x00005a06
[    3.925689] dwmmc_exynos 12220000.mmc: : FIFOTH:	0x201f0020
[    3.931240] dwmmc_exynos 12220000.mmc: : CDETECT:	0x00000000
[    3.936879] dwmmc_exynos 12220000.mmc: : WRTPRT:	0x00000000
[    3.942432] dwmmc_exynos 12220000.mmc: : GPIO:	0x00000000
[    3.947811] dwmmc_exynos 12220000.mmc: : TCBCNT:	0x00000200
[    3.953362] dwmmc_exynos 12220000.mmc: : TBBCNT:	0x00000000
[    3.958913] dwmmc_exynos 12220000.mmc: : DEBNCE:	0x00ffffff
[    3.964466] dwmmc_exynos 12220000.mmc: : USRID:	0xffffffff
[    3.969931] dwmmc_exynos 12220000.mmc: : VERID:	0x5342250a
[    3.975395] dwmmc_exynos 12220000.mmc: : HCON:	0x00e42541
[    3.980775] dwmmc_exynos 12220000.mmc: : UHS_REG:	0x00000001
[    3.986414] dwmmc_exynos 12220000.mmc: : BMOD:	0x00000200
[    3.991792] dwmmc_exynos 12220000.mmc: : PLDMND:	0x00000000
[    3.997343] dwmmc_exynos 12220000.mmc: : DBADDR:	0x4dbbb000
[    4.002895] dwmmc_exynos 12220000.mmc: : IDSTS:	0x00000000
[    4.008362] dwmmc_exynos 12220000.mmc: : IDINTEN:	0x00000103
[    4.014000] dwmmc_exynos 12220000.mmc: : DSCADDR:	0x00000000
[    4.019639] dwmmc_exynos 12220000.mmc: : BUFADDR:	0x00000000
[    4.025278] dwmmc_exynos 12220000.mmc: : CLKSEL:	0x03030002
[    4.030830] dwmmc_exynos 12220000.mmc: : ================= CMD REG =================
[    4.038549] dwmmc_exynos 12220000.mmc: : read/write        : read
[    4.044622] dwmmc_exynos 12220000.mmc: : data expected     : 0
[    4.050435] dwmmc_exynos 12220000.mmc: : cmd index         : 0
[    4.056247] dwmmc_exynos 12220000.mmc: : ================ STATUS REG ===============
[    4.063968] dwmmc_exynos 12220000.mmc: : fifocount         : 0
[    4.069789] dwmmc_exynos 12220000.mmc: : response index    : 11
[    4.075679] dwmmc_exynos 12220000.mmc: : data state mc busy: 0
[    4.081491] dwmmc_exynos 12220000.mmc: : data busy         : 1
[    4.087303] dwmmc_exynos 12220000.mmc: : data 3 state      : 0
[    4.093116] dwmmc_exynos 12220000.mmc: : command fsm state : 0
[    4.098929] dwmmc_exynos 12220000.mmc: : fifo full         : 0
[    4.104740] dwmmc_exynos 12220000.mmc: : fifo empty        : 1
[    4.110553] dwmmc_exynos 12220000.mmc: : fifo tx watermark : 1
[    4.116365] dwmmc_exynos 12220000.mmc: : fifo rx watermark : 0
[    4.122176] dwmmc_exynos 12220000.mmc: : ============== STATUS DUMP ================
[    4.129898] dwmmc_exynos 12220000.mmc: : cmd_status:      0x00000000
[    4.136230] dwmmc_exynos 12220000.mmc: : data_status:     0x00000000
[    4.142563] dwmmc_exynos 12220000.mmc: : pending_events:  0x00000000
[    4.148895] dwmmc_exynos 12220000.mmc: : completed_events:0x00000001
[    4.155229] dwmmc_exynos 12220000.mmc: : state:           7
[    4.160779] dwmmc_exynos 12220000.mmc: : ===========================================
[    4.168502] mmc_host mmc1: Timeout sending command (cmd 0x202000 arg 0x0 status 0x80202000)
[    4.674791] dwmmc_exynos 12220000.mmc: : ============== REGISTER DUMP ==============
[    4.681065] dwmmc_exynos 12220000.mmc: : CTRL:	0x00000010
[    4.686442] dwmmc_exynos 12220000.mmc: : PWREN:	0x00000001
[    4.691907] dwmmc_exynos 12220000.mmc: : CLKDIV:	0x0000003f
[    4.697460] dwmmc_exynos 12220000.mmc: : CLKSRC:	0x00000000
[    4.703012] dwmmc_exynos 12220000.mmc: : CLKENA:	0x00000000
[    4.708563] dwmmc_exynos 12220000.mmc: : TMOUT:	0xffffffff
[    4.714029] dwmmc_exynos 12220000.mmc: : CTYPE:	0x00000000
[    4.719493] dwmmc_exynos 12220000.mmc: : BLKSIZ:	0x00000200
[    4.725047] dwmmc_exynos 12220000.mmc: : BYTCNT:	0x00000200
[    4.730598] dwmmc_exynos 12220000.mmc: : INTMSK:	0x0000b7ff
[    4.736150] dwmmc_exynos 12220000.mmc: : CMDARG:	0x00000000
[    4.741701] dwmmc_exynos 12220000.mmc: : CMD:	0x80202000
[    4.746994] dwmmc_exynos 12220000.mmc: : RESP0:	0x00000320
[    4.752459] dwmmc_exynos 12220000.mmc: : RESP1:	0x3ac57f80
[    4.757925] dwmmc_exynos 12220000.mmc: : RESP2:	0x5b590000
[    4.763389] dwmmc_exynos 12220000.mmc: : RESP3:	0x400e0032
[    4.768856] dwmmc_exynos 12220000.mmc: : MINTSTS:	0x00001000
[    4.774494] dwmmc_exynos 12220000.mmc: : RINTSTS:	0x00011000
[    4.780133] dwmmc_exynos 12220000.mmc: : STATUS:	0x00005a06
[    4.785685] dwmmc_exynos 12220000.mmc: : FIFOTH:	0x201f0020
[    4.791236] dwmmc_exynos 12220000.mmc: : CDETECT:	0x00000000
[    4.796876] dwmmc_exynos 12220000.mmc: : WRTPRT:	0x00000000
[    4.802428] dwmmc_exynos 12220000.mmc: : GPIO:	0x00000000
[    4.807806] dwmmc_exynos 12220000.mmc: : TCBCNT:	0x00000200
[    4.813358] dwmmc_exynos 12220000.mmc: : TBBCNT:	0x00000000
[    4.818909] dwmmc_exynos 12220000.mmc: : DEBNCE:	0x00ffffff
[    4.824462] dwmmc_exynos 12220000.mmc: : USRID:	0xffffffff
[    4.829927] dwmmc_exynos 12220000.mmc: : VERID:	0x5342250a
[    4.835392] dwmmc_exynos 12220000.mmc: : HCON:	0x00e42541
[    4.840770] dwmmc_exynos 12220000.mmc: : UHS_REG:	0x00000001
[    4.846410] dwmmc_exynos 12220000.mmc: : BMOD:	0x00000200
[    4.851788] dwmmc_exynos 12220000.mmc: : PLDMND:	0x00000000
[    4.857340] dwmmc_exynos 12220000.mmc: : DBADDR:	0x4dbbb000
[    4.862893] dwmmc_exynos 12220000.mmc: : IDSTS:	0x00000000
[    4.868357] dwmmc_exynos 12220000.mmc: : IDINTEN:	0x00000103
[    4.873996] dwmmc_exynos 12220000.mmc: : DSCADDR:	0x00000000
[    4.879636] dwmmc_exynos 12220000.mmc: : BUFADDR:	0x00000000
[    4.885274] dwmmc_exynos 12220000.mmc: : CLKSEL:	0x03030002
[    4.890826] dwmmc_exynos 12220000.mmc: : ================= CMD REG =================
[    4.898545] dwmmc_exynos 12220000.mmc: : read/write        : read
[    4.904619] dwmmc_exynos 12220000.mmc: : data expected     : 0
[    4.910430] dwmmc_exynos 12220000.mmc: : cmd index         : 0
[    4.916243] dwmmc_exynos 12220000.mmc: : ================ STATUS REG ===============
[    4.923963] dwmmc_exynos 12220000.mmc: : fifocount         : 0
[    4.929793] dwmmc_exynos 12220000.mmc: : response index    : 11
[    4.935675] dwmmc_exynos 12220000.mmc: : data state mc busy: 0
[    4.941487] dwmmc_exynos 12220000.mmc: : data busy         : 1
[    4.947298] dwmmc_exynos 12220000.mmc: : data 3 state      : 0
[    4.953112] dwmmc_exynos 12220000.mmc: : command fsm state : 0
[    4.958923] dwmmc_exynos 12220000.mmc: : fifo full         : 0
[    4.964736] dwmmc_exynos 12220000.mmc: : fifo empty        : 1
[    4.970548] dwmmc_exynos 12220000.mmc: : fifo tx watermark : 1
[    4.976361] dwmmc_exynos 12220000.mmc: : fifo rx watermark : 0
[    4.982173] dwmmc_exynos 12220000.mmc: : ============== STATUS DUMP ================
[    4.989893] dwmmc_exynos 12220000.mmc: : cmd_status:      0x00000000
[    4.996226] dwmmc_exynos 12220000.mmc: : data_status:     0x00000000
[    5.002559] dwmmc_exynos 12220000.mmc: : pending_events:  0x00000000
[    5.008891] dwmmc_exynos 12220000.mmc: : completed_events:0x00000001
[    5.015225] dwmmc_exynos 12220000.mmc: : state:           7
[    5.020777] dwmmc_exynos 12220000.mmc: : ===========================================
[    5.028498] mmc_host mmc1: Timeout sending command (cmd 0x202000 arg 0x0 status 0x80202000)
[    5.534791] dwmmc_exynos 12220000.mmc: : ============== REGISTER DUMP ==============
[    5.541069] dwmmc_exynos 12220000.mmc: : CTRL:	0x00000010
[    5.546447] dwmmc_exynos 12220000.mmc: : PWREN:	0x00000001
[    5.551913] dwmmc_exynos 12220000.mmc: : CLKDIV:	0x0000003f
[    5.557465] dwmmc_exynos 12220000.mmc: : CLKSRC:	0x00000000
[    5.563017] dwmmc_exynos 12220000.mmc: : CLKENA:	0x00000000
[    5.568569] dwmmc_exynos 12220000.mmc: : TMOUT:	0xffffffff
[    5.574034] dwmmc_exynos 12220000.mmc: : CTYPE:	0x00000000
[    5.579499] dwmmc_exynos 12220000.mmc: : BLKSIZ:	0x00000200
[    5.585051] dwmmc_exynos 12220000.mmc: : BYTCNT:	0x00000200
[    5.590603] dwmmc_exynos 12220000.mmc: : INTMSK:	0x0000b7ff
[    5.596155] dwmmc_exynos 12220000.mmc: : CMDARG:	0x00000000
[    5.601707] dwmmc_exynos 12220000.mmc: : CMD:	0x80202000
[    5.607000] dwmmc_exynos 12220000.mmc: : RESP0:	0x00000320
[    5.612464] dwmmc_exynos 12220000.mmc: : RESP1:	0x3ac57f80
[    5.617929] dwmmc_exynos 12220000.mmc: : RESP2:	0x5b590000
[    5.623395] dwmmc_exynos 12220000.mmc: : RESP3:	0x400e0032
[    5.628861] dwmmc_exynos 12220000.mmc: : MINTSTS:	0x00001000
[    5.634498] dwmmc_exynos 12220000.mmc: : RINTSTS:	0x00011000
[    5.640138] dwmmc_exynos 12220000.mmc: : STATUS:	0x00005a06
[    5.645688] dwmmc_exynos 12220000.mmc: : FIFOTH:	0x201f0020
[    5.651242] dwmmc_exynos 12220000.mmc: : CDETECT:	0x00000000
[    5.656881] dwmmc_exynos 12220000.mmc: : WRTPRT:	0x00000000
[    5.662432] dwmmc_exynos 12220000.mmc: : GPIO:	0x00000000
[    5.667810] dwmmc_exynos 12220000.mmc: : TCBCNT:	0x00000200
[    5.673362] dwmmc_exynos 12220000.mmc: : TBBCNT:	0x00000000
[    5.678914] dwmmc_exynos 12220000.mmc: : DEBNCE:	0x00ffffff
[    5.684467] dwmmc_exynos 12220000.mmc: : USRID:	0xffffffff
[    5.689932] dwmmc_exynos 12220000.mmc: : VERID:	0x5342250a
[    5.695396] dwmmc_exynos 12220000.mmc: : HCON:	0x00e42541
[    5.700775] dwmmc_exynos 12220000.mmc: : UHS_REG:	0x00000001
[    5.706415] dwmmc_exynos 12220000.mmc: : BMOD:	0x00000200
[    5.711793] dwmmc_exynos 12220000.mmc: : PLDMND:	0x00000000
[    5.717345] dwmmc_exynos 12220000.mmc: : DBADDR:	0x4dbbb000
[    5.722897] dwmmc_exynos 12220000.mmc: : IDSTS:	0x00000000
[    5.728363] dwmmc_exynos 12220000.mmc: : IDINTEN:	0x00000103
[    5.734001] dwmmc_exynos 12220000.mmc: : DSCADDR:	0x00000000
[    5.739640] dwmmc_exynos 12220000.mmc: : BUFADDR:	0x00000000
[    5.745279] dwmmc_exynos 12220000.mmc: : CLKSEL:	0x03030002
[    5.750830] dwmmc_exynos 12220000.mmc: : ================= CMD REG =================
[    5.758550] dwmmc_exynos 12220000.mmc: : read/write        : read
[    5.764623] dwmmc_exynos 12220000.mmc: : data expected     : 0
[    5.770436] dwmmc_exynos 12220000.mmc: : cmd index         : 0
[    5.776248] dwmmc_exynos 12220000.mmc: : ================ STATUS REG ===============
[    5.783969] dwmmc_exynos 12220000.mmc: : fifocount         : 0
[    5.789791] dwmmc_exynos 12220000.mmc: : response index    : 11
[    5.795680] dwmmc_exynos 12220000.mmc: : data state mc busy: 0
[    5.801492] dwmmc_exynos 12220000.mmc: : data busy         : 1
[    5.807303] dwmmc_exynos 12220000.mmc: : data 3 state      : 0
[    5.813116] dwmmc_exynos 12220000.mmc: : command fsm state : 0
[    5.818929] dwmmc_exynos 12220000.mmc: : fifo full         : 0
[    5.824740] dwmmc_exynos 12220000.mmc: : fifo empty        : 1
[    5.830553] dwmmc_exynos 12220000.mmc: : fifo tx watermark : 1
[    5.836365] dwmmc_exynos 12220000.mmc: : fifo rx watermark : 0
[    5.842182] dwmmc_exynos 12220000.mmc: : ============== STATUS DUMP ================
[    5.842187] dwmmc_exynos 12220000.mmc: : cmd_status:      0x00000000
[    5.842193] dwmmc_exynos 12220000.mmc: : data_status:     0x00000000
[    5.842197] dwmmc_exynos 12220000.mmc: : pending_events:  0x00000000
[    5.842202] dwmmc_exynos 12220000.mmc: : completed_events:0x00000001
[    5.842211] dwmmc_exynos 12220000.mmc: : state:           7
[    5.842220] dwmmc_exynos 12220000.mmc: : ===========================================
[    5.842230] mmc_host mmc1: Timeout sending command (cmd 0x202000 arg 0x0 status 0x80202000)


U-Boot 2013.04 (Feb 21 2014 - 15:29:10) for Peach

CPU:	Exynos5420 at 900MHz

Board: Google Peach Pit, rev 6.0
I2C:   ready
DRAM:  2 GiB
PMIC max77802-pmic initialized
CPU:	Exynos5420 at 1800MHz
TPS65090 PMIC EC init
MMC:   EXYNOS DWMMC: 0, EXYNOS DWMMC: 1
SF: Detected W25Q32DW with page size 4 KiB, total 32 MiB
In:    cros-ec-keyb
Out:   lcd
Err:   lcd
SF: Detected W25Q32DW with page size 4 KiB, total 32 MiB
ELOG: Event(17) added with size 13
Net:   No ethernet found.
Hit any key to stop autoboot:  3 \b\b\b 0 
Exynos DP init done
Peach # 
Peach # 

[PATCH 3/3] mmc: dw_mmc: Support voltage changes

[Seungwon Jeon]

On Tue, July 01, 2014. Yuvaraj Kumar wrote:
> On Fri, Jun 27, 2014 at 4:48 PM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
> > Hi Yuvaraj,
> >
> > On Fri, June 27, 2014, Yuvaraj Kumar wrote:
> >> On Thu, Jun 26, 2014 at 10:20 PM, Doug Anderson <dianders@chromium.org> wrote:
> >> > Seungwon,
> >> >
> >> > On Thu, Jun 26, 2014 at 3:41 AM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
> >> >> On Thu, June 26, 2014, Doug Anderson wrote:
> >> >>> Seungwon,
> >> >>>
> >> >>> On Wed, Jun 25, 2014 at 6:08 AM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
> >> >>> > On Mon, June 23, 2014, Yuvaraj Kumar C D wrote:
> >> >>> >> Subject: [PATCH 3/3] mmc: dw_mmc: Support voltage changes
> >> >>> >>
> >> >>> >> From: Doug Anderson <dianders@chromium.org>
> >> >>> >>
> >> >>> >> For UHS cards we need the ability to switch voltages from 3.3V to
> >> >>> >> 1.8V.  Add support to the dw_mmc driver to handle this.  Note that
> >> >>> >> dw_mmc needs a little bit of extra code since the interface needs a
> >> >>> >> special bit programmed to the CMD register while CMD11 is progressing.
> >> >>> >> This means adding a few extra states to the state machine to track.
> >> >>> >
> >> >>> > Overall new additional states makes it complicated.
> >> >>> > Can we do that in other way?
> >> >>>
> >> >>> That was the best I was able to figure out when I thought this
> >> >>> through.  If you have ideas for doing it another way I'd imagine that
> >> >>> Yuvaraj would be happy to take your feedback.
> >> >> Let's clean up SDMMC_CMD_VOLT_SWITCH.
> >> >> In turn, we may remove state-handling simply.
> >> >>
> >> >>>
> >> >>>
> >> >>> >> Signed-off-by: Doug Anderson <dianders@chromium.org>
> >> >>> >> Signed-off-by: Yuvaraj Kumar C D <yuvaraj.cd@samsung.com>
> >> >>> >>
> >> >>> >> ---
> >> >>> >>  drivers/mmc/host/dw_mmc.c  |  145 +++++++++++++++++++++++++++++++++++++++++---
> >> >>> >>  drivers/mmc/host/dw_mmc.h  |    5 +-
> >> >>> >>  include/linux/mmc/dw_mmc.h |    2 +
> >> >>> >>  3 files changed, 142 insertions(+), 10 deletions(-)
> >> >>> >>
> >> >>> >> diff --git a/drivers/mmc/host/dw_mmc.c b/drivers/mmc/host/dw_mmc.c
> >> >>> >> index e034bce..38eb548 100644
> >> >>> >> --- a/drivers/mmc/host/dw_mmc.c
> >> >>> >> +++ b/drivers/mmc/host/dw_mmc.c
> >> >>> >> @@ -29,6 +29,7 @@
> >> >>> >>  #include <linux/irq.h>
> >> >>> >>  #include <linux/mmc/host.h>
> >> >>> >>  #include <linux/mmc/mmc.h>
> >> >>> >> +#include <linux/mmc/sd.h>
> >> >>> >>  #include <linux/mmc/sdio.h>
> >> >>> >>  #include <linux/mmc/dw_mmc.h>
> >> >>> >>  #include <linux/bitops.h>
> >> >>> >> @@ -235,10 +236,13 @@ err:
> >> >>> >>  }
> >> >>> >>  #endif /* defined(CONFIG_DEBUG_FS) */
> >> >>> >>
> >> >>> >> +static void mci_send_cmd(struct dw_mci_slot *slot, u32 cmd, u32 arg);
> >> >>> >> +
> >> >>> >>  static u32 dw_mci_prepare_command(struct mmc_host *mmc, struct mmc_command *cmd)
> >> >>> >>  {
> >> >>> >>       struct mmc_data *data;
> >> >>> >>       struct dw_mci_slot *slot = mmc_priv(mmc);
> >> >>> >> +     struct dw_mci *host = slot->host;
> >> >>> >>       const struct dw_mci_drv_data *drv_data = slot->host->drv_data;
> >> >>> >>       u32 cmdr;
> >> >>> >>       cmd->error = -EINPROGRESS;
> >> >>> >> @@ -254,6 +258,32 @@ static u32 dw_mci_prepare_command(struct mmc_host *mmc, struct
> mmc_command
> >> >>> *cmd)
> >> >>> >>       else if (cmd->opcode != MMC_SEND_STATUS && cmd->data)
> >> >>> >>               cmdr |= SDMMC_CMD_PRV_DAT_WAIT;
> >> >>> >>
> >> >>> >> +     if (cmd->opcode == SD_SWITCH_VOLTAGE) {
> >> >>> >> +             u32 clk_en_a;
> >> >>> >> +
> >> >>> >> +             /* Special bit makes CMD11 not die */
> >> >>> >> +             cmdr |= SDMMC_CMD_VOLT_SWITCH;
> >> >>> >> +
> >> >>> >> +             /* Change state to continue to handle CMD11 weirdness */
> >> >>> >> +             WARN_ON(slot->host->state != STATE_SENDING_CMD);
> >> >>> >> +             slot->host->state = STATE_SENDING_CMD11;
> >> >>> >> +
> >> >>> >> +             /*
> >> >>> >> +              * We need to disable clock stop while doing voltage switch
> >> >>> >> +              * according to 7.4.1.2 Voltage Switch Normal Scenario.
> >> >>> >> +              *
> >> >>> >> +              * It's assumed that by the next time the CLKENA is updated
> >> >>> >> +              * (when we set the clock next) that the voltage change will
> >> >>> >> +              * be over, so we don't bother setting any bits to synchronize
> >> >>> >> +              * with dw_mci_setup_bus().
> >> >>> >> +              */
> >> >>> >> +             clk_en_a = mci_readl(host, CLKENA);
> >> >>> >> +             clk_en_a &= ~(SDMMC_CLKEN_LOW_PWR << slot->id);
> >> >>> >> +             mci_writel(host, CLKENA, clk_en_a);
> >> >>> >> +             mci_send_cmd(slot, SDMMC_CMD_UPD_CLK |
> >> >>> >> +                          SDMMC_CMD_PRV_DAT_WAIT, 0);
> >> >>> > dw_mci_disable_low_power() can be used here.
> >> >>>
> >> >>> Ah.  I guess we don't have that locally anymore.  Locally we have variants on:
> >> >> I'm checking on cjb/mmc branch.
> >> >>
> >> >>> * https://patchwork.kernel.org/patch/3070311/
> >> >>> * https://patchwork.kernel.org/patch/3070251/
> >> >>> * https://patchwork.kernel.org/patch/3070221/
> >> >>>
> >> >>> ...which removed that function.  ...but I guess upstream never picked
> >> >>> up those patches, huh?  Looking back it looks like you had some
> >> >>> feedback and it needed another spin but somehow fell off my plate.  :(
> >> >>>
> >> >>> Maybe this is something Yuvaraj would like to pick up?
> >> >> It's long ago. I remember that there is no progress since my last comment.
> >> >> In case of patch "3070221", I want to pick up for next Kernel.
> >> >
> >> > Sounds like Yuvaraj has agreed to look at addressing your comments.  Thanks!
> >> >
> >> >
> >> >>> >> +     }
> >> >>> >> +
> >> >>> >>       if (cmd->flags & MMC_RSP_PRESENT) {
> >> >>> >>               /* We expect a response, so set this bit */
> >> >>> >>               cmdr |= SDMMC_CMD_RESP_EXP;
> >> >>> >> @@ -776,11 +806,15 @@ static void dw_mci_setup_bus(struct dw_mci_slot *slot, bool
> force_clkinit)
> >> >>> >>       unsigned int clock = slot->clock;
> >> >>> >>       u32 div;
> >> >>> >>       u32 clk_en_a;
> >> >>> >> +     u32 sdmmc_cmd_bits = SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT;
> >> >>> >> +
> >> >>> >> +     /* We must continue to set bit 28 in CMD until the change is complete */
> >> >>> >> +     if (host->state == STATE_WAITING_CMD11_DONE)
> >> >>> >> +             sdmmc_cmd_bits |= SDMMC_CMD_VOLT_SWITCH;
> >> >>> > I didn't get the reason SDMMC_CMD_VOLT_SWITCH is needed during clock update(enable/disable)
> >> >>> > Can you explain it in details?
> >> >>>
> >> >>> Simply put: if I didn't do this then the system hung during voltage
> >> >>> switch.  It was not documented in the version of the IP manual that I
> >> >>> had access to and it took me a bunch of digging / trial and error to
> >> >>> figure this out.  Whatever this bit does internally it's important to
> >> >>> set it while the voltage change is happening.  Note that this need was
> >> >>> the whole reason for adding the extra state to the state machine.
> >> >>>
> >> >>> Perhaps Yuvaraj can try without it and I'd assume he'll report the
> >> >>> same freeze.
> >> >> Clarify the necessity of SDMMC_CMD_VOLT_SWITCH.
> >> >> As far as I experience, we didn't apply this bit for several projects.
> >> >
> >> > I just tried this locally on our ChromeOS 3.8 kernel (which has quite
> >> > a few backports and matches dw_mmc upstream pretty closely).  When I
> >> > take out "sdmmc_cmd_bits |= SDMMC_CMD_VOLT_SWITCH" and bootup it fails
> >> > to bringup WiFi chip.  It prints messages like:
> >> >
> >> > [    4.400842] mmc_host mmc1: Timeout sending command (cmd 0x202000
> >> > arg 0x0 status 0x80202000)
> >> >
> >> > I will let Yuvaraj comment about his testing too.
> >> To make it simple and verify, i have just enabled eMMC and SD channels
> >> on 3.16-rc1 and tested.
> >> Without "sdmmc_cmd_bits |= SDMMC_CMD_VOLT_SWITCH" bootup fails.
> >> [    3.015971] mmc_host mmc1: Timeout sending command (cmd 0x202000
> >> arg 0x0 status 0x80202000)
> >> [    3.530974] mmc_host mmc1: Timeout sending command (cmd 0x202000
> >> arg 0x0 status 0x80202000)
> > Thank you for test.
> > Hmm, it failed during clock update, right?.
> > Can you check HLE interrupt at this time?
> > I'll investigate for this.
> Yes, it failed during clock update.I am dumping the register contents
> just before the dev_err statement.I can observe that 1st timeout does
> not have HLE interrupt.But once we set bit 0 of UHS_REG subsequent
> timeout's have HLE bit set.I am attaching the complete log.

Before disabling the clocks, we ensure that the card is not busy.(It's from Synopsys's databook)
I guess SDMMC_CMD_VOLT_SWITCH flag seems to affect to avoid this limitation if you have no problem.
But, I met the following log with HEL when I apply and test this patch .

"mmc_host mmc1: Timeout sending command (cmd 0x202000 arg 0x0 status 0x80202000)"
I need to look into it more deeply.

> >
> >> >
> >> >
> >> >>> >> +static int dw_mci_switch_voltage(struct mmc_host *mmc, struct mmc_ios *ios)
> >> >>> >> +{
> >> >>> >> +     struct dw_mci_slot *slot = mmc_priv(mmc);
> >> >>> >> +     struct dw_mci *host = slot->host;
> >> >>> >> +     u32 uhs;
> >> >>> >> +     u32 v18 = SDMMC_UHS_18V << slot->id;
> >> >>> >> +     int min_uv, max_uv;
> >> >>> >> +     int ret;
> >> >>> >> +
> >> >>> >> +     /*
> >> >>> >> +      * Program the voltage.  Note that some instances of dw_mmc may use
> >> >>> >> +      * the UHS_REG for this.  For other instances (like exynos) the UHS_REG
> >> >>> >> +      * does no harm but you need to set the regulator directly.  Try both.
> >> >>> >> +      */
> >> >>> >> +     uhs = mci_readl(host, UHS_REG);
> >> >>> >> +     if (ios->signal_voltage == MMC_SIGNAL_VOLTAGE_330) {
> >> >>> >> +             min_uv = 2700000;
> >> >>> >> +             max_uv = 3600000;
> >> >>> >> +             uhs &= ~v18;
> >> >>> >> +     } else {
> >> >>> >> +             min_uv = 1700000;
> >> >>> >> +             max_uv = 1950000;
> >> >>> >> +             uhs |= v18;
> >> >>> >> +     }
> >> >>> >> +     if (!IS_ERR(mmc->supply.vqmmc)) {
> >> >>> >> +             ret = regulator_set_voltage(mmc->supply.vqmmc, min_uv, max_uv);
> >> >>> >> +
> >> >>> >> +             /*
> >> >>> >> +              * Only complain if regulator claims that it's not in the 1.8V
> >> >>> >> +              * range.  This avoids a bunch of errors in the case that
> >> >>> >> +              * we've got a fixed 1.8V regulator but the core SD code still
> >> >>> >> +              * thinks it ought to try to switch to 3.3 and then back to 1.8
> >> >>> >> +              */
> >> >>> >> +             if (ret) {
> >> >>> > I think if ret is error, printing message and returning error is good.
> >> >>> > Currently, just returning '0' though it fails.
> >> >> Any feedback?
> >> >
> >> > Whoops, right.  I think you're right that in the case it warns it
> >> > should also return the error code.  In the case it doesn't warn it
> >> > shouldn't.  Also: possibly we should use a trick like the mmc core
> >> > does and use "regulator_count_voltages" to figure out if we're going
> >> > to be able to switch voltages...
> >> >
> >> >>> > Artificially adding SDMMC_INT_CMD_DONE to pending is needed to complete cmd handling?
> >> >>> > Is there any reason?
> >> >>>
> >> >>> I don't remember this for sure since I wrote this a long time ago.
> >> >>> Maybe it's not needed?  I may have just been modeling on other
> >> >>> interrupts.
> >> >> If there is no specific reason, please remove it.
> >> >
> >> > OK, we'll see how Yuvaraj's testing goes without it.  My incredibly
> >> > brief testing in our local Chrome OS 3.8 tree shows that the WiFi is
> >> > not detected properly if I comment out the line:
> >> >   dw_mci_cmd_interrupt(host,
> >> >     pending | SDMMC_INT_CMD_DONE);
> >> >
> >> > It may be sufficient to simply schedule the tasklet or to do some
> >> > other subset of dw_mci_cmd_interrupt().  Yuvaraj can confirm on the
> >> > latest kernel and also investigate further.
> > How about completing CMD without SDMMC_INT_CMD_DONE?
> Without this tuning fails.
> [    2.525284] mmc_host mmc0: Bus speed (slot 0) = 200000000Hz (slot
> req 200000000Hz, actual 200000000HZ div = 0)
> [  198.556573] random: nonblocking pool is initialized
> [  240.226044] INFO: task kworker/u8:0:6 blocked for more than 120 seconds.
> [  240.231276]       Not tainted 3.16.0-rc1-00010-g2f7a756-dirty #55
> [  240.237359] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs"
> disables this message.
> [  240.245150] kworker/u8:0    D c03a58a0     0     6      2 0x00000000
> [  240.251487] Workqueue: kmmcd mmc_rescan
> [  240.255297] [<c03a58a0>] (__schedule) from [<c03a5330>]
> (schedule_timeout+0x130/0x170)
> [  240.263201] [<c03a5330>] (schedule_timeout) from [<c03a6540>]
> (wait_for_common+0xbc/0x14c)
> [  240.271441] [<c03a6540>] (wait_for_common) from [<c02ca6e4>]
> (mmc_wait_for_req_done+0x6c/0xf0)
> [  240.280031] [<c02ca6e4>] (mmc_wait_for_req_done) from [<c02e3f10>]
> (dw_mci_exynos_execute_tuning+0x1d4/0x2c4)
> [  240.289919] [<c02e3f10>] (dw_mci_exynos_execute_tuning) from
> [<c02e0198>] (dw_mci_execute_tuning+0x54/0xb4)
> [  240.299635] [<c02e0198>] (dw_mci_execute_tuning) from [<c02cf47c>]
> (mmc_init_card+0xeec/0x14ac)
> [  240.308309] [<c02cf47c>] (mmc_init_card) from [<c02cfc34>]
> (mmc_attach_mmc+0x8c/0x160)
> [  240.316202] [<c02cfc34>] (mmc_attach_mmc) from [<c02cca0c>]
> (mmc_rescan+0x2b0/0x308)
> [  240.323924] [<c02cca0c>] (mmc_rescan) from [<c0032f60>]
> (process_one_work+0xf8/0x364)
> [  240.331732] [<c0032f60>] (process_one_work) from [<c0033800>]
> (worker_thread+0x50/0x5a0)
> [  240.339799] [<c0033800>] (worker_thread) from [<c0038e7c>]
> (kthread+0xd8/0xf0)
> [  240.346999] [<c0038e7c>] (kthread) from [<c000e438>]
> (ret_from_fork+0x14/0x3c)

Thank you for test.
I want to clear it why CMD_DONE is related to tuning.

Thanks,
Seungwon Jeon

[PATCH 3/3] mmc: dw_mmc: Support voltage changes

[Seungwon Jeon]

On Fri, July 04, 2014, Seungwon Jeon wrote:
> On Tue, July 01, 2014. Yuvaraj Kumar wrote:
> > On Fri, Jun 27, 2014 at 4:48 PM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
> > > Hi Yuvaraj,
> > >
> > > On Fri, June 27, 2014, Yuvaraj Kumar wrote:
> > >> On Thu, Jun 26, 2014 at 10:20 PM, Doug Anderson <dianders@chromium.org> wrote:
> > >> > Seungwon,
> > >> >
> > >> > On Thu, Jun 26, 2014 at 3:41 AM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
> > >> >> On Thu, June 26, 2014, Doug Anderson wrote:
> > >> >>> Seungwon,
> > >> >>>
> > >> >>> On Wed, Jun 25, 2014 at 6:08 AM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
> > >> >>> > On Mon, June 23, 2014, Yuvaraj Kumar C D wrote:
> > >> >>> >> Subject: [PATCH 3/3] mmc: dw_mmc: Support voltage changes
> > >> >>> >>
> > >> >>> >> From: Doug Anderson <dianders@chromium.org>
> > >> >>> >>
> > >> >>> >> For UHS cards we need the ability to switch voltages from 3.3V to
> > >> >>> >> 1.8V.  Add support to the dw_mmc driver to handle this.  Note that
> > >> >>> >> dw_mmc needs a little bit of extra code since the interface needs a
> > >> >>> >> special bit programmed to the CMD register while CMD11 is progressing.
> > >> >>> >> This means adding a few extra states to the state machine to track.
> > >> >>> >
> > >> >>> > Overall new additional states makes it complicated.
> > >> >>> > Can we do that in other way?
> > >> >>>
> > >> >>> That was the best I was able to figure out when I thought this
> > >> >>> through.  If you have ideas for doing it another way I'd imagine that
> > >> >>> Yuvaraj would be happy to take your feedback.
> > >> >> Let's clean up SDMMC_CMD_VOLT_SWITCH.
> > >> >> In turn, we may remove state-handling simply.
> > >> >>
> > >> >>>
> > >> >>>
> > >> >>> >> Signed-off-by: Doug Anderson <dianders@chromium.org>
> > >> >>> >> Signed-off-by: Yuvaraj Kumar C D <yuvaraj.cd@samsung.com>
> > >> >>> >>
> > >> >>> >> ---
> > >> >>> >>  drivers/mmc/host/dw_mmc.c  |  145 +++++++++++++++++++++++++++++++++++++++++---
> > >> >>> >>  drivers/mmc/host/dw_mmc.h  |    5 +-
> > >> >>> >>  include/linux/mmc/dw_mmc.h |    2 +
> > >> >>> >>  3 files changed, 142 insertions(+), 10 deletions(-)
> > >> >>> >>
> > >> >>> >> diff --git a/drivers/mmc/host/dw_mmc.c b/drivers/mmc/host/dw_mmc.c
> > >> >>> >> index e034bce..38eb548 100644
> > >> >>> >> --- a/drivers/mmc/host/dw_mmc.c
> > >> >>> >> +++ b/drivers/mmc/host/dw_mmc.c
> > >> >>> >> @@ -29,6 +29,7 @@
> > >> >>> >>  #include <linux/irq.h>
> > >> >>> >>  #include <linux/mmc/host.h>
> > >> >>> >>  #include <linux/mmc/mmc.h>
> > >> >>> >> +#include <linux/mmc/sd.h>
> > >> >>> >>  #include <linux/mmc/sdio.h>
> > >> >>> >>  #include <linux/mmc/dw_mmc.h>
> > >> >>> >>  #include <linux/bitops.h>
> > >> >>> >> @@ -235,10 +236,13 @@ err:
> > >> >>> >>  }
> > >> >>> >>  #endif /* defined(CONFIG_DEBUG_FS) */
> > >> >>> >>
> > >> >>> >> +static void mci_send_cmd(struct dw_mci_slot *slot, u32 cmd, u32 arg);
> > >> >>> >> +
> > >> >>> >>  static u32 dw_mci_prepare_command(struct mmc_host *mmc, struct mmc_command *cmd)
> > >> >>> >>  {
> > >> >>> >>       struct mmc_data *data;
> > >> >>> >>       struct dw_mci_slot *slot = mmc_priv(mmc);
> > >> >>> >> +     struct dw_mci *host = slot->host;
> > >> >>> >>       const struct dw_mci_drv_data *drv_data = slot->host->drv_data;
> > >> >>> >>       u32 cmdr;
> > >> >>> >>       cmd->error = -EINPROGRESS;
> > >> >>> >> @@ -254,6 +258,32 @@ static u32 dw_mci_prepare_command(struct mmc_host *mmc, struct
> > mmc_command
> > >> >>> *cmd)
> > >> >>> >>       else if (cmd->opcode != MMC_SEND_STATUS && cmd->data)
> > >> >>> >>               cmdr |= SDMMC_CMD_PRV_DAT_WAIT;
> > >> >>> >>
> > >> >>> >> +     if (cmd->opcode == SD_SWITCH_VOLTAGE) {
> > >> >>> >> +             u32 clk_en_a;
> > >> >>> >> +
> > >> >>> >> +             /* Special bit makes CMD11 not die */
> > >> >>> >> +             cmdr |= SDMMC_CMD_VOLT_SWITCH;
> > >> >>> >> +
> > >> >>> >> +             /* Change state to continue to handle CMD11 weirdness */
> > >> >>> >> +             WARN_ON(slot->host->state != STATE_SENDING_CMD);
> > >> >>> >> +             slot->host->state = STATE_SENDING_CMD11;
> > >> >>> >> +
> > >> >>> >> +             /*
> > >> >>> >> +              * We need to disable clock stop while doing voltage switch
> > >> >>> >> +              * according to 7.4.1.2 Voltage Switch Normal Scenario.
> > >> >>> >> +              *
> > >> >>> >> +              * It's assumed that by the next time the CLKENA is updated
> > >> >>> >> +              * (when we set the clock next) that the voltage change will
> > >> >>> >> +              * be over, so we don't bother setting any bits to synchronize
> > >> >>> >> +              * with dw_mci_setup_bus().
> > >> >>> >> +              */
> > >> >>> >> +             clk_en_a = mci_readl(host, CLKENA);
> > >> >>> >> +             clk_en_a &= ~(SDMMC_CLKEN_LOW_PWR << slot->id);
> > >> >>> >> +             mci_writel(host, CLKENA, clk_en_a);
> > >> >>> >> +             mci_send_cmd(slot, SDMMC_CMD_UPD_CLK |
> > >> >>> >> +                          SDMMC_CMD_PRV_DAT_WAIT, 0);
> > >> >>> > dw_mci_disable_low_power() can be used here.
> > >> >>>
> > >> >>> Ah.  I guess we don't have that locally anymore.  Locally we have variants on:
> > >> >> I'm checking on cjb/mmc branch.
> > >> >>
> > >> >>> * https://patchwork.kernel.org/patch/3070311/
> > >> >>> * https://patchwork.kernel.org/patch/3070251/
> > >> >>> * https://patchwork.kernel.org/patch/3070221/
> > >> >>>
> > >> >>> ...which removed that function.  ...but I guess upstream never picked
> > >> >>> up those patches, huh?  Looking back it looks like you had some
> > >> >>> feedback and it needed another spin but somehow fell off my plate.  :(
> > >> >>>
> > >> >>> Maybe this is something Yuvaraj would like to pick up?
> > >> >> It's long ago. I remember that there is no progress since my last comment.
> > >> >> In case of patch "3070221", I want to pick up for next Kernel.
> > >> >
> > >> > Sounds like Yuvaraj has agreed to look at addressing your comments.  Thanks!
> > >> >
> > >> >
> > >> >>> >> +     }
> > >> >>> >> +
> > >> >>> >>       if (cmd->flags & MMC_RSP_PRESENT) {
> > >> >>> >>               /* We expect a response, so set this bit */
> > >> >>> >>               cmdr |= SDMMC_CMD_RESP_EXP;
> > >> >>> >> @@ -776,11 +806,15 @@ static void dw_mci_setup_bus(struct dw_mci_slot *slot, bool
> > force_clkinit)
> > >> >>> >>       unsigned int clock = slot->clock;
> > >> >>> >>       u32 div;
> > >> >>> >>       u32 clk_en_a;
> > >> >>> >> +     u32 sdmmc_cmd_bits = SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT;
> > >> >>> >> +
> > >> >>> >> +     /* We must continue to set bit 28 in CMD until the change is complete */
> > >> >>> >> +     if (host->state == STATE_WAITING_CMD11_DONE)
> > >> >>> >> +             sdmmc_cmd_bits |= SDMMC_CMD_VOLT_SWITCH;
> > >> >>> > I didn't get the reason SDMMC_CMD_VOLT_SWITCH is needed during clock update(enable/disable)
> > >> >>> > Can you explain it in details?
> > >> >>>
> > >> >>> Simply put: if I didn't do this then the system hung during voltage
> > >> >>> switch.  It was not documented in the version of the IP manual that I
> > >> >>> had access to and it took me a bunch of digging / trial and error to
> > >> >>> figure this out.  Whatever this bit does internally it's important to
> > >> >>> set it while the voltage change is happening.  Note that this need was
> > >> >>> the whole reason for adding the extra state to the state machine.
> > >> >>>
> > >> >>> Perhaps Yuvaraj can try without it and I'd assume he'll report the
> > >> >>> same freeze.
> > >> >> Clarify the necessity of SDMMC_CMD_VOLT_SWITCH.
> > >> >> As far as I experience, we didn't apply this bit for several projects.
> > >> >
> > >> > I just tried this locally on our ChromeOS 3.8 kernel (which has quite
> > >> > a few backports and matches dw_mmc upstream pretty closely).  When I
> > >> > take out "sdmmc_cmd_bits |= SDMMC_CMD_VOLT_SWITCH" and bootup it fails
> > >> > to bringup WiFi chip.  It prints messages like:
> > >> >
> > >> > [    4.400842] mmc_host mmc1: Timeout sending command (cmd 0x202000
> > >> > arg 0x0 status 0x80202000)
> > >> >
> > >> > I will let Yuvaraj comment about his testing too.
> > >> To make it simple and verify, i have just enabled eMMC and SD channels
> > >> on 3.16-rc1 and tested.
> > >> Without "sdmmc_cmd_bits |= SDMMC_CMD_VOLT_SWITCH" bootup fails.
> > >> [    3.015971] mmc_host mmc1: Timeout sending command (cmd 0x202000
> > >> arg 0x0 status 0x80202000)
> > >> [    3.530974] mmc_host mmc1: Timeout sending command (cmd 0x202000
> > >> arg 0x0 status 0x80202000)
> > > Thank you for test.
> > > Hmm, it failed during clock update, right?.
> > > Can you check HLE interrupt at this time?
> > > I'll investigate for this.
> > Yes, it failed during clock update.I am dumping the register contents
> > just before the dev_err statement.I can observe that 1st timeout does
> > not have HLE interrupt.But once we set bit 0 of UHS_REG subsequent
> > timeout's have HLE bit set.I am attaching the complete log.
> 
> Before disabling the clocks, we ensure that the card is not busy.(It's from Synopsys's databook)
> I guess SDMMC_CMD_VOLT_SWITCH flag seems to affect to avoid this limitation if you have no problem.
> But, I met the following log with HEL when I apply and test this patch .
> 
> "mmc_host mmc1: Timeout sending command (cmd 0x202000 arg 0x0 status 0x80202000)"
> I need to look into it more deeply.
I confirmed that SDMMC_CMD_VOLT_SWITCH is needed during clock-off/on in voltage switching sequence
even though there is no description for this in Synopsys's databook.
And the above-mentioned message("mmc_host mmc1: Timeout sending command (cmd 0x202000 arg 0x0 status 0x80202000)")
is happened on mmc_power_cycle(), which should be done because of busy status of DAT[3:0].
It indicates clock-off/on sequence fails when executing mmc_power_cycle().

> 
> > >
> > >> >
> > >> >
> > >> >>> >> +static int dw_mci_switch_voltage(struct mmc_host *mmc, struct mmc_ios *ios)
> > >> >>> >> +{
> > >> >>> >> +     struct dw_mci_slot *slot = mmc_priv(mmc);
> > >> >>> >> +     struct dw_mci *host = slot->host;
> > >> >>> >> +     u32 uhs;
> > >> >>> >> +     u32 v18 = SDMMC_UHS_18V << slot->id;
> > >> >>> >> +     int min_uv, max_uv;
> > >> >>> >> +     int ret;
> > >> >>> >> +
> > >> >>> >> +     /*
> > >> >>> >> +      * Program the voltage.  Note that some instances of dw_mmc may use
> > >> >>> >> +      * the UHS_REG for this.  For other instances (like exynos) the UHS_REG
> > >> >>> >> +      * does no harm but you need to set the regulator directly.  Try both.
> > >> >>> >> +      */
> > >> >>> >> +     uhs = mci_readl(host, UHS_REG);
> > >> >>> >> +     if (ios->signal_voltage == MMC_SIGNAL_VOLTAGE_330) {
> > >> >>> >> +             min_uv = 2700000;
> > >> >>> >> +             max_uv = 3600000;
> > >> >>> >> +             uhs &= ~v18;
> > >> >>> >> +     } else {
> > >> >>> >> +             min_uv = 1700000;
> > >> >>> >> +             max_uv = 1950000;
> > >> >>> >> +             uhs |= v18;
> > >> >>> >> +     }
> > >> >>> >> +     if (!IS_ERR(mmc->supply.vqmmc)) {
> > >> >>> >> +             ret = regulator_set_voltage(mmc->supply.vqmmc, min_uv, max_uv);
> > >> >>> >> +
> > >> >>> >> +             /*
> > >> >>> >> +              * Only complain if regulator claims that it's not in the 1.8V
> > >> >>> >> +              * range.  This avoids a bunch of errors in the case that
> > >> >>> >> +              * we've got a fixed 1.8V regulator but the core SD code still
> > >> >>> >> +              * thinks it ought to try to switch to 3.3 and then back to 1.8
> > >> >>> >> +              */
> > >> >>> >> +             if (ret) {
> > >> >>> > I think if ret is error, printing message and returning error is good.
> > >> >>> > Currently, just returning '0' though it fails.
> > >> >> Any feedback?
> > >> >
> > >> > Whoops, right.  I think you're right that in the case it warns it
> > >> > should also return the error code.  In the case it doesn't warn it
> > >> > shouldn't.  Also: possibly we should use a trick like the mmc core
> > >> > does and use "regulator_count_voltages" to figure out if we're going
> > >> > to be able to switch voltages...
> > >> >
> > >> >>> > Artificially adding SDMMC_INT_CMD_DONE to pending is needed to complete cmd handling?
> > >> >>> > Is there any reason?
> > >> >>>
> > >> >>> I don't remember this for sure since I wrote this a long time ago.
> > >> >>> Maybe it's not needed?  I may have just been modeling on other
> > >> >>> interrupts.
> > >> >> If there is no specific reason, please remove it.
> > >> >
> > >> > OK, we'll see how Yuvaraj's testing goes without it.  My incredibly
> > >> > brief testing in our local Chrome OS 3.8 tree shows that the WiFi is
> > >> > not detected properly if I comment out the line:
> > >> >   dw_mci_cmd_interrupt(host,
> > >> >     pending | SDMMC_INT_CMD_DONE);
> > >> >
> > >> > It may be sufficient to simply schedule the tasklet or to do some
> > >> > other subset of dw_mci_cmd_interrupt().  Yuvaraj can confirm on the
> > >> > latest kernel and also investigate further.
> > > How about completing CMD without SDMMC_INT_CMD_DONE?
> > Without this tuning fails.
> > [    2.525284] mmc_host mmc0: Bus speed (slot 0) = 200000000Hz (slot
> > req 200000000Hz, actual 200000000HZ div = 0)
> > [  198.556573] random: nonblocking pool is initialized
> > [  240.226044] INFO: task kworker/u8:0:6 blocked for more than 120 seconds.
> > [  240.231276]       Not tainted 3.16.0-rc1-00010-g2f7a756-dirty #55
> > [  240.237359] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs"
> > disables this message.
> > [  240.245150] kworker/u8:0    D c03a58a0     0     6      2 0x00000000
> > [  240.251487] Workqueue: kmmcd mmc_rescan
> > [  240.255297] [<c03a58a0>] (__schedule) from [<c03a5330>]
> > (schedule_timeout+0x130/0x170)
> > [  240.263201] [<c03a5330>] (schedule_timeout) from [<c03a6540>]
> > (wait_for_common+0xbc/0x14c)
> > [  240.271441] [<c03a6540>] (wait_for_common) from [<c02ca6e4>]
> > (mmc_wait_for_req_done+0x6c/0xf0)
> > [  240.280031] [<c02ca6e4>] (mmc_wait_for_req_done) from [<c02e3f10>]
> > (dw_mci_exynos_execute_tuning+0x1d4/0x2c4)
> > [  240.289919] [<c02e3f10>] (dw_mci_exynos_execute_tuning) from
> > [<c02e0198>] (dw_mci_execute_tuning+0x54/0xb4)
> > [  240.299635] [<c02e0198>] (dw_mci_execute_tuning) from [<c02cf47c>]
> > (mmc_init_card+0xeec/0x14ac)
> > [  240.308309] [<c02cf47c>] (mmc_init_card) from [<c02cfc34>]
> > (mmc_attach_mmc+0x8c/0x160)
> > [  240.316202] [<c02cfc34>] (mmc_attach_mmc) from [<c02cca0c>]
> > (mmc_rescan+0x2b0/0x308)
> > [  240.323924] [<c02cca0c>] (mmc_rescan) from [<c0032f60>]
> > (process_one_work+0xf8/0x364)
> > [  240.331732] [<c0032f60>] (process_one_work) from [<c0033800>]
> > (worker_thread+0x50/0x5a0)
> > [  240.339799] [<c0033800>] (worker_thread) from [<c0038e7c>]
> > (kthread+0xd8/0xf0)
> > [  240.346999] [<c0038e7c>] (kthread) from [<c000e438>]
> > (ret_from_fork+0x14/0x3c)
> 
> Thank you for test.
> I want to clear it why CMD_DONE is related to tuning.
In my case, SDMMC_INT_CMD_DONE doesn't affect.
And do you check VOLT_SWITCH interrupt occurs twice during voltage-switching.
For last one, if-statement in dw_mci_interrupt() disturbs clearing interrupt.
if ((host->state == STATE_SENDING_CMD11) &&

In addition, I suggest adding another flag instead of using host's state.
Can you consider it?

Thanks,
Seungwon Jeon

[PATCH 3/3] mmc: dw_mmc: Support voltage changes

[Yuvaraj Kumar]

On Mon, Jul 7, 2014 at 10:53 AM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
> On Fri, July 04, 2014, Seungwon Jeon wrote:
>> On Tue, July 01, 2014. Yuvaraj Kumar wrote:
>> > On Fri, Jun 27, 2014 at 4:48 PM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
>> > > Hi Yuvaraj,
>> > >
>> > > On Fri, June 27, 2014, Yuvaraj Kumar wrote:
>> > >> On Thu, Jun 26, 2014 at 10:20 PM, Doug Anderson <dianders@chromium.org> wrote:
>> > >> > Seungwon,
>> > >> >
>> > >> > On Thu, Jun 26, 2014 at 3:41 AM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
>> > >> >> On Thu, June 26, 2014, Doug Anderson wrote:
>> > >> >>> Seungwon,
>> > >> >>>
>> > >> >>> On Wed, Jun 25, 2014 at 6:08 AM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
>> > >> >>> > On Mon, June 23, 2014, Yuvaraj Kumar C D wrote:
>> > >> >>> >> Subject: [PATCH 3/3] mmc: dw_mmc: Support voltage changes
>> > >> >>> >>
>> > >> >>> >> From: Doug Anderson <dianders@chromium.org>
>> > >> >>> >>
>> > >> >>> >> For UHS cards we need the ability to switch voltages from 3.3V to
>> > >> >>> >> 1.8V.  Add support to the dw_mmc driver to handle this.  Note that
>> > >> >>> >> dw_mmc needs a little bit of extra code since the interface needs a
>> > >> >>> >> special bit programmed to the CMD register while CMD11 is progressing.
>> > >> >>> >> This means adding a few extra states to the state machine to track.
>> > >> >>> >
>> > >> >>> > Overall new additional states makes it complicated.
>> > >> >>> > Can we do that in other way?
>> > >> >>>
>> > >> >>> That was the best I was able to figure out when I thought this
>> > >> >>> through.  If you have ideas for doing it another way I'd imagine that
>> > >> >>> Yuvaraj would be happy to take your feedback.
>> > >> >> Let's clean up SDMMC_CMD_VOLT_SWITCH.
>> > >> >> In turn, we may remove state-handling simply.
>> > >> >>
>> > >> >>>
>> > >> >>>
>> > >> >>> >> Signed-off-by: Doug Anderson <dianders@chromium.org>
>> > >> >>> >> Signed-off-by: Yuvaraj Kumar C D <yuvaraj.cd@samsung.com>
>> > >> >>> >>
>> > >> >>> >> ---
>> > >> >>> >>  drivers/mmc/host/dw_mmc.c  |  145 +++++++++++++++++++++++++++++++++++++++++---
>> > >> >>> >>  drivers/mmc/host/dw_mmc.h  |    5 +-
>> > >> >>> >>  include/linux/mmc/dw_mmc.h |    2 +
>> > >> >>> >>  3 files changed, 142 insertions(+), 10 deletions(-)
>> > >> >>> >>
>> > >> >>> >> diff --git a/drivers/mmc/host/dw_mmc.c b/drivers/mmc/host/dw_mmc.c
>> > >> >>> >> index e034bce..38eb548 100644
>> > >> >>> >> --- a/drivers/mmc/host/dw_mmc.c
>> > >> >>> >> +++ b/drivers/mmc/host/dw_mmc.c
>> > >> >>> >> @@ -29,6 +29,7 @@
>> > >> >>> >>  #include <linux/irq.h>
>> > >> >>> >>  #include <linux/mmc/host.h>
>> > >> >>> >>  #include <linux/mmc/mmc.h>
>> > >> >>> >> +#include <linux/mmc/sd.h>
>> > >> >>> >>  #include <linux/mmc/sdio.h>
>> > >> >>> >>  #include <linux/mmc/dw_mmc.h>
>> > >> >>> >>  #include <linux/bitops.h>
>> > >> >>> >> @@ -235,10 +236,13 @@ err:
>> > >> >>> >>  }
>> > >> >>> >>  #endif /* defined(CONFIG_DEBUG_FS) */
>> > >> >>> >>
>> > >> >>> >> +static void mci_send_cmd(struct dw_mci_slot *slot, u32 cmd, u32 arg);
>> > >> >>> >> +
>> > >> >>> >>  static u32 dw_mci_prepare_command(struct mmc_host *mmc, struct mmc_command *cmd)
>> > >> >>> >>  {
>> > >> >>> >>       struct mmc_data *data;
>> > >> >>> >>       struct dw_mci_slot *slot = mmc_priv(mmc);
>> > >> >>> >> +     struct dw_mci *host = slot->host;
>> > >> >>> >>       const struct dw_mci_drv_data *drv_data = slot->host->drv_data;
>> > >> >>> >>       u32 cmdr;
>> > >> >>> >>       cmd->error = -EINPROGRESS;
>> > >> >>> >> @@ -254,6 +258,32 @@ static u32 dw_mci_prepare_command(struct mmc_host *mmc, struct
>> > mmc_command
>> > >> >>> *cmd)
>> > >> >>> >>       else if (cmd->opcode != MMC_SEND_STATUS && cmd->data)
>> > >> >>> >>               cmdr |= SDMMC_CMD_PRV_DAT_WAIT;
>> > >> >>> >>
>> > >> >>> >> +     if (cmd->opcode == SD_SWITCH_VOLTAGE) {
>> > >> >>> >> +             u32 clk_en_a;
>> > >> >>> >> +
>> > >> >>> >> +             /* Special bit makes CMD11 not die */
>> > >> >>> >> +             cmdr |= SDMMC_CMD_VOLT_SWITCH;
>> > >> >>> >> +
>> > >> >>> >> +             /* Change state to continue to handle CMD11 weirdness */
>> > >> >>> >> +             WARN_ON(slot->host->state != STATE_SENDING_CMD);
>> > >> >>> >> +             slot->host->state = STATE_SENDING_CMD11;
>> > >> >>> >> +
>> > >> >>> >> +             /*
>> > >> >>> >> +              * We need to disable clock stop while doing voltage switch
>> > >> >>> >> +              * according to 7.4.1.2 Voltage Switch Normal Scenario.
>> > >> >>> >> +              *
>> > >> >>> >> +              * It's assumed that by the next time the CLKENA is updated
>> > >> >>> >> +              * (when we set the clock next) that the voltage change will
>> > >> >>> >> +              * be over, so we don't bother setting any bits to synchronize
>> > >> >>> >> +              * with dw_mci_setup_bus().
>> > >> >>> >> +              */
>> > >> >>> >> +             clk_en_a = mci_readl(host, CLKENA);
>> > >> >>> >> +             clk_en_a &= ~(SDMMC_CLKEN_LOW_PWR << slot->id);
>> > >> >>> >> +             mci_writel(host, CLKENA, clk_en_a);
>> > >> >>> >> +             mci_send_cmd(slot, SDMMC_CMD_UPD_CLK |
>> > >> >>> >> +                          SDMMC_CMD_PRV_DAT_WAIT, 0);
>> > >> >>> > dw_mci_disable_low_power() can be used here.
>> > >> >>>
>> > >> >>> Ah.  I guess we don't have that locally anymore.  Locally we have variants on:
>> > >> >> I'm checking on cjb/mmc branch.
>> > >> >>
>> > >> >>> * https://patchwork.kernel.org/patch/3070311/
>> > >> >>> * https://patchwork.kernel.org/patch/3070251/
>> > >> >>> * https://patchwork.kernel.org/patch/3070221/
>> > >> >>>
>> > >> >>> ...which removed that function.  ...but I guess upstream never picked
>> > >> >>> up those patches, huh?  Looking back it looks like you had some
>> > >> >>> feedback and it needed another spin but somehow fell off my plate.  :(
>> > >> >>>
>> > >> >>> Maybe this is something Yuvaraj would like to pick up?
>> > >> >> It's long ago. I remember that there is no progress since my last comment.
>> > >> >> In case of patch "3070221", I want to pick up for next Kernel.
>> > >> >
>> > >> > Sounds like Yuvaraj has agreed to look at addressing your comments.  Thanks!
>> > >> >
>> > >> >
>> > >> >>> >> +     }
>> > >> >>> >> +
>> > >> >>> >>       if (cmd->flags & MMC_RSP_PRESENT) {
>> > >> >>> >>               /* We expect a response, so set this bit */
>> > >> >>> >>               cmdr |= SDMMC_CMD_RESP_EXP;
>> > >> >>> >> @@ -776,11 +806,15 @@ static void dw_mci_setup_bus(struct dw_mci_slot *slot, bool
>> > force_clkinit)
>> > >> >>> >>       unsigned int clock = slot->clock;
>> > >> >>> >>       u32 div;
>> > >> >>> >>       u32 clk_en_a;
>> > >> >>> >> +     u32 sdmmc_cmd_bits = SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT;
>> > >> >>> >> +
>> > >> >>> >> +     /* We must continue to set bit 28 in CMD until the change is complete */
>> > >> >>> >> +     if (host->state == STATE_WAITING_CMD11_DONE)
>> > >> >>> >> +             sdmmc_cmd_bits |= SDMMC_CMD_VOLT_SWITCH;
>> > >> >>> > I didn't get the reason SDMMC_CMD_VOLT_SWITCH is needed during clock update(enable/disable)
>> > >> >>> > Can you explain it in details?
>> > >> >>>
>> > >> >>> Simply put: if I didn't do this then the system hung during voltage
>> > >> >>> switch.  It was not documented in the version of the IP manual that I
>> > >> >>> had access to and it took me a bunch of digging / trial and error to
>> > >> >>> figure this out.  Whatever this bit does internally it's important to
>> > >> >>> set it while the voltage change is happening.  Note that this need was
>> > >> >>> the whole reason for adding the extra state to the state machine.
>> > >> >>>
>> > >> >>> Perhaps Yuvaraj can try without it and I'd assume he'll report the
>> > >> >>> same freeze.
>> > >> >> Clarify the necessity of SDMMC_CMD_VOLT_SWITCH.
>> > >> >> As far as I experience, we didn't apply this bit for several projects.
>> > >> >
>> > >> > I just tried this locally on our ChromeOS 3.8 kernel (which has quite
>> > >> > a few backports and matches dw_mmc upstream pretty closely).  When I
>> > >> > take out "sdmmc_cmd_bits |= SDMMC_CMD_VOLT_SWITCH" and bootup it fails
>> > >> > to bringup WiFi chip.  It prints messages like:
>> > >> >
>> > >> > [    4.400842] mmc_host mmc1: Timeout sending command (cmd 0x202000
>> > >> > arg 0x0 status 0x80202000)
>> > >> >
>> > >> > I will let Yuvaraj comment about his testing too.
>> > >> To make it simple and verify, i have just enabled eMMC and SD channels
>> > >> on 3.16-rc1 and tested.
>> > >> Without "sdmmc_cmd_bits |= SDMMC_CMD_VOLT_SWITCH" bootup fails.
>> > >> [    3.015971] mmc_host mmc1: Timeout sending command (cmd 0x202000
>> > >> arg 0x0 status 0x80202000)
>> > >> [    3.530974] mmc_host mmc1: Timeout sending command (cmd 0x202000
>> > >> arg 0x0 status 0x80202000)
>> > > Thank you for test.
>> > > Hmm, it failed during clock update, right?.
>> > > Can you check HLE interrupt at this time?
>> > > I'll investigate for this.
>> > Yes, it failed during clock update.I am dumping the register contents
>> > just before the dev_err statement.I can observe that 1st timeout does
>> > not have HLE interrupt.But once we set bit 0 of UHS_REG subsequent
>> > timeout's have HLE bit set.I am attaching the complete log.
>>
>> Before disabling the clocks, we ensure that the card is not busy.(It's from Synopsys's databook)
>> I guess SDMMC_CMD_VOLT_SWITCH flag seems to affect to avoid this limitation if you have no problem.
>> But, I met the following log with HEL when I apply and test this patch .
>>
>> "mmc_host mmc1: Timeout sending command (cmd 0x202000 arg 0x0 status 0x80202000)"
>> I need to look into it more deeply.
> I confirmed that SDMMC_CMD_VOLT_SWITCH is needed during clock-off/on in voltage switching sequence
> even though there is no description for this in Synopsys's databook.
> And the above-mentioned message("mmc_host mmc1: Timeout sending command (cmd 0x202000 arg 0x0 status 0x80202000)")
> is happened on mmc_power_cycle(), which should be done because of busy status of DAT[3:0].
> It indicates clock-off/on sequence fails when executing mmc_power_cycle().
>
>>
>> > >
>> > >> >
>> > >> >
>> > >> >>> >> +static int dw_mci_switch_voltage(struct mmc_host *mmc, struct mmc_ios *ios)
>> > >> >>> >> +{
>> > >> >>> >> +     struct dw_mci_slot *slot = mmc_priv(mmc);
>> > >> >>> >> +     struct dw_mci *host = slot->host;
>> > >> >>> >> +     u32 uhs;
>> > >> >>> >> +     u32 v18 = SDMMC_UHS_18V << slot->id;
>> > >> >>> >> +     int min_uv, max_uv;
>> > >> >>> >> +     int ret;
>> > >> >>> >> +
>> > >> >>> >> +     /*
>> > >> >>> >> +      * Program the voltage.  Note that some instances of dw_mmc may use
>> > >> >>> >> +      * the UHS_REG for this.  For other instances (like exynos) the UHS_REG
>> > >> >>> >> +      * does no harm but you need to set the regulator directly.  Try both.
>> > >> >>> >> +      */
>> > >> >>> >> +     uhs = mci_readl(host, UHS_REG);
>> > >> >>> >> +     if (ios->signal_voltage == MMC_SIGNAL_VOLTAGE_330) {
>> > >> >>> >> +             min_uv = 2700000;
>> > >> >>> >> +             max_uv = 3600000;
>> > >> >>> >> +             uhs &= ~v18;
>> > >> >>> >> +     } else {
>> > >> >>> >> +             min_uv = 1700000;
>> > >> >>> >> +             max_uv = 1950000;
>> > >> >>> >> +             uhs |= v18;
>> > >> >>> >> +     }
>> > >> >>> >> +     if (!IS_ERR(mmc->supply.vqmmc)) {
>> > >> >>> >> +             ret = regulator_set_voltage(mmc->supply.vqmmc, min_uv, max_uv);
>> > >> >>> >> +
>> > >> >>> >> +             /*
>> > >> >>> >> +              * Only complain if regulator claims that it's not in the 1.8V
>> > >> >>> >> +              * range.  This avoids a bunch of errors in the case that
>> > >> >>> >> +              * we've got a fixed 1.8V regulator but the core SD code still
>> > >> >>> >> +              * thinks it ought to try to switch to 3.3 and then back to 1.8
>> > >> >>> >> +              */
>> > >> >>> >> +             if (ret) {
>> > >> >>> > I think if ret is error, printing message and returning error is good.
>> > >> >>> > Currently, just returning '0' though it fails.
>> > >> >> Any feedback?
>> > >> >
>> > >> > Whoops, right.  I think you're right that in the case it warns it
>> > >> > should also return the error code.  In the case it doesn't warn it
>> > >> > shouldn't.  Also: possibly we should use a trick like the mmc core
>> > >> > does and use "regulator_count_voltages" to figure out if we're going
>> > >> > to be able to switch voltages...
>> > >> >
>> > >> >>> > Artificially adding SDMMC_INT_CMD_DONE to pending is needed to complete cmd handling?
>> > >> >>> > Is there any reason?
>> > >> >>>
>> > >> >>> I don't remember this for sure since I wrote this a long time ago.
>> > >> >>> Maybe it's not needed?  I may have just been modeling on other
>> > >> >>> interrupts.
>> > >> >> If there is no specific reason, please remove it.
>> > >> >
>> > >> > OK, we'll see how Yuvaraj's testing goes without it.  My incredibly
>> > >> > brief testing in our local Chrome OS 3.8 tree shows that the WiFi is
>> > >> > not detected properly if I comment out the line:
>> > >> >   dw_mci_cmd_interrupt(host,
>> > >> >     pending | SDMMC_INT_CMD_DONE);
>> > >> >
>> > >> > It may be sufficient to simply schedule the tasklet or to do some
>> > >> > other subset of dw_mci_cmd_interrupt().  Yuvaraj can confirm on the
>> > >> > latest kernel and also investigate further.
>> > > How about completing CMD without SDMMC_INT_CMD_DONE?
>> > Without this tuning fails.
>> > [    2.525284] mmc_host mmc0: Bus speed (slot 0) = 200000000Hz (slot
>> > req 200000000Hz, actual 200000000HZ div = 0)
>> > [  198.556573] random: nonblocking pool is initialized
>> > [  240.226044] INFO: task kworker/u8:0:6 blocked for more than 120 seconds.
>> > [  240.231276]       Not tainted 3.16.0-rc1-00010-g2f7a756-dirty #55
>> > [  240.237359] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs"
>> > disables this message.
>> > [  240.245150] kworker/u8:0    D c03a58a0     0     6      2 0x00000000
>> > [  240.251487] Workqueue: kmmcd mmc_rescan
>> > [  240.255297] [<c03a58a0>] (__schedule) from [<c03a5330>]
>> > (schedule_timeout+0x130/0x170)
>> > [  240.263201] [<c03a5330>] (schedule_timeout) from [<c03a6540>]
>> > (wait_for_common+0xbc/0x14c)
>> > [  240.271441] [<c03a6540>] (wait_for_common) from [<c02ca6e4>]
>> > (mmc_wait_for_req_done+0x6c/0xf0)
>> > [  240.280031] [<c02ca6e4>] (mmc_wait_for_req_done) from [<c02e3f10>]
>> > (dw_mci_exynos_execute_tuning+0x1d4/0x2c4)
>> > [  240.289919] [<c02e3f10>] (dw_mci_exynos_execute_tuning) from
>> > [<c02e0198>] (dw_mci_execute_tuning+0x54/0xb4)
>> > [  240.299635] [<c02e0198>] (dw_mci_execute_tuning) from [<c02cf47c>]
>> > (mmc_init_card+0xeec/0x14ac)
>> > [  240.308309] [<c02cf47c>] (mmc_init_card) from [<c02cfc34>]
>> > (mmc_attach_mmc+0x8c/0x160)
>> > [  240.316202] [<c02cfc34>] (mmc_attach_mmc) from [<c02cca0c>]
>> > (mmc_rescan+0x2b0/0x308)
>> > [  240.323924] [<c02cca0c>] (mmc_rescan) from [<c0032f60>]
>> > (process_one_work+0xf8/0x364)
>> > [  240.331732] [<c0032f60>] (process_one_work) from [<c0033800>]
>> > (worker_thread+0x50/0x5a0)
>> > [  240.339799] [<c0033800>] (worker_thread) from [<c0038e7c>]
>> > (kthread+0xd8/0xf0)
>> > [  240.346999] [<c0038e7c>] (kthread) from [<c000e438>]
>> > (ret_from_fork+0x14/0x3c)
>>
>> Thank you for test.
>> I want to clear it why CMD_DONE is related to tuning.
> In my case, SDMMC_INT_CMD_DONE doesn't affect.
I rechecked this.SDMMC_INT_CMD_DONE flag doesn't affect.Simply
sheduling the tasklet is enough.Sorry,I misunderstood your comments
earlier.
> And do you check VOLT_SWITCH interrupt occurs twice during voltage-switching.
No.Only once.
> For last one, if-statement in dw_mci_interrupt() disturbs clearing interrupt.
> if ((host->state == STATE_SENDING_CMD11) &&
>
> In addition, I suggest adding another flag instead of using host's state.
> Can you consider it?
Sure.I will check this.
>
> Thanks,
> Seungwon Jeon
>

[PATCH 3/3] mmc: dw_mmc: Support voltage changes

[Seungwon Jeon]

On Tue, July 08, 2014,Yuvaraj Kumar wrote:
> On Mon, Jul 7, 2014 at 10:53 AM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
> > On Fri, July 04, 2014, Seungwon Jeon wrote:
> >> On Tue, July 01, 2014. Yuvaraj Kumar wrote:
> >> > On Fri, Jun 27, 2014 at 4:48 PM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
> >> > > Hi Yuvaraj,
> >> > >
> >> > > On Fri, June 27, 2014, Yuvaraj Kumar wrote:
> >> > >> On Thu, Jun 26, 2014 at 10:20 PM, Doug Anderson <dianders@chromium.org> wrote:
> >> > >> > Seungwon,
> >> > >> >
> >> > >> > On Thu, Jun 26, 2014 at 3:41 AM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
> >> > >> >> On Thu, June 26, 2014, Doug Anderson wrote:
> >> > >> >>> Seungwon,
> >> > >> >>>
> >> > >> >>> On Wed, Jun 25, 2014 at 6:08 AM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
> >> > >> >>> > On Mon, June 23, 2014, Yuvaraj Kumar C D wrote:
> >> > >> >>> >> Subject: [PATCH 3/3] mmc: dw_mmc: Support voltage changes
> >> > >> >>> >>
> >> > >> >>> >> From: Doug Anderson <dianders@chromium.org>
> >> > >> >>> >>
> >> > >> >>> >> For UHS cards we need the ability to switch voltages from 3.3V to
> >> > >> >>> >> 1.8V.  Add support to the dw_mmc driver to handle this.  Note that
> >> > >> >>> >> dw_mmc needs a little bit of extra code since the interface needs a
> >> > >> >>> >> special bit programmed to the CMD register while CMD11 is progressing.
> >> > >> >>> >> This means adding a few extra states to the state machine to track.
> >> > >> >>> >
> >> > >> >>> > Overall new additional states makes it complicated.
> >> > >> >>> > Can we do that in other way?
> >> > >> >>>
> >> > >> >>> That was the best I was able to figure out when I thought this
> >> > >> >>> through.  If you have ideas for doing it another way I'd imagine that
> >> > >> >>> Yuvaraj would be happy to take your feedback.
> >> > >> >> Let's clean up SDMMC_CMD_VOLT_SWITCH.
> >> > >> >> In turn, we may remove state-handling simply.
> >> > >> >>
> >> > >> >>>
> >> > >> >>>
> >> > >> >>> >> Signed-off-by: Doug Anderson <dianders@chromium.org>
> >> > >> >>> >> Signed-off-by: Yuvaraj Kumar C D <yuvaraj.cd@samsung.com>
> >> > >> >>> >>
> >> > >> >>> >> ---
> >> > >> >>> >>  drivers/mmc/host/dw_mmc.c  |  145 +++++++++++++++++++++++++++++++++++++++++---
> >> > >> >>> >>  drivers/mmc/host/dw_mmc.h  |    5 +-
> >> > >> >>> >>  include/linux/mmc/dw_mmc.h |    2 +
> >> > >> >>> >>  3 files changed, 142 insertions(+), 10 deletions(-)
> >> > >> >>> >>
> >> > >> >>> >> diff --git a/drivers/mmc/host/dw_mmc.c b/drivers/mmc/host/dw_mmc.c
> >> > >> >>> >> index e034bce..38eb548 100644
> >> > >> >>> >> --- a/drivers/mmc/host/dw_mmc.c
> >> > >> >>> >> +++ b/drivers/mmc/host/dw_mmc.c
> >> > >> >>> >> @@ -29,6 +29,7 @@
> >> > >> >>> >>  #include <linux/irq.h>
> >> > >> >>> >>  #include <linux/mmc/host.h>
> >> > >> >>> >>  #include <linux/mmc/mmc.h>
> >> > >> >>> >> +#include <linux/mmc/sd.h>
> >> > >> >>> >>  #include <linux/mmc/sdio.h>
> >> > >> >>> >>  #include <linux/mmc/dw_mmc.h>
> >> > >> >>> >>  #include <linux/bitops.h>
> >> > >> >>> >> @@ -235,10 +236,13 @@ err:
> >> > >> >>> >>  }
> >> > >> >>> >>  #endif /* defined(CONFIG_DEBUG_FS) */
> >> > >> >>> >>
> >> > >> >>> >> +static void mci_send_cmd(struct dw_mci_slot *slot, u32 cmd, u32 arg);
> >> > >> >>> >> +
> >> > >> >>> >>  static u32 dw_mci_prepare_command(struct mmc_host *mmc, struct mmc_command *cmd)
> >> > >> >>> >>  {
> >> > >> >>> >>       struct mmc_data *data;
> >> > >> >>> >>       struct dw_mci_slot *slot = mmc_priv(mmc);
> >> > >> >>> >> +     struct dw_mci *host = slot->host;
> >> > >> >>> >>       const struct dw_mci_drv_data *drv_data = slot->host->drv_data;
> >> > >> >>> >>       u32 cmdr;
> >> > >> >>> >>       cmd->error = -EINPROGRESS;
> >> > >> >>> >> @@ -254,6 +258,32 @@ static u32 dw_mci_prepare_command(struct mmc_host *mmc, struct
> >> > mmc_command
> >> > >> >>> *cmd)
> >> > >> >>> >>       else if (cmd->opcode != MMC_SEND_STATUS && cmd->data)
> >> > >> >>> >>               cmdr |= SDMMC_CMD_PRV_DAT_WAIT;
> >> > >> >>> >>
> >> > >> >>> >> +     if (cmd->opcode == SD_SWITCH_VOLTAGE) {
> >> > >> >>> >> +             u32 clk_en_a;
> >> > >> >>> >> +
> >> > >> >>> >> +             /* Special bit makes CMD11 not die */
> >> > >> >>> >> +             cmdr |= SDMMC_CMD_VOLT_SWITCH;
> >> > >> >>> >> +
> >> > >> >>> >> +             /* Change state to continue to handle CMD11 weirdness */
> >> > >> >>> >> +             WARN_ON(slot->host->state != STATE_SENDING_CMD);
> >> > >> >>> >> +             slot->host->state = STATE_SENDING_CMD11;
> >> > >> >>> >> +
> >> > >> >>> >> +             /*
> >> > >> >>> >> +              * We need to disable clock stop while doing voltage switch
> >> > >> >>> >> +              * according to 7.4.1.2 Voltage Switch Normal Scenario.
> >> > >> >>> >> +              *
> >> > >> >>> >> +              * It's assumed that by the next time the CLKENA is updated
> >> > >> >>> >> +              * (when we set the clock next) that the voltage change will
> >> > >> >>> >> +              * be over, so we don't bother setting any bits to synchronize
> >> > >> >>> >> +              * with dw_mci_setup_bus().
> >> > >> >>> >> +              */
> >> > >> >>> >> +             clk_en_a = mci_readl(host, CLKENA);
> >> > >> >>> >> +             clk_en_a &= ~(SDMMC_CLKEN_LOW_PWR << slot->id);
> >> > >> >>> >> +             mci_writel(host, CLKENA, clk_en_a);
> >> > >> >>> >> +             mci_send_cmd(slot, SDMMC_CMD_UPD_CLK |
> >> > >> >>> >> +                          SDMMC_CMD_PRV_DAT_WAIT, 0);
> >> > >> >>> > dw_mci_disable_low_power() can be used here.
> >> > >> >>>
> >> > >> >>> Ah.  I guess we don't have that locally anymore.  Locally we have variants on:
> >> > >> >> I'm checking on cjb/mmc branch.
> >> > >> >>
> >> > >> >>> * https://patchwork.kernel.org/patch/3070311/
> >> > >> >>> * https://patchwork.kernel.org/patch/3070251/
> >> > >> >>> * https://patchwork.kernel.org/patch/3070221/
> >> > >> >>>
> >> > >> >>> ...which removed that function.  ...but I guess upstream never picked
> >> > >> >>> up those patches, huh?  Looking back it looks like you had some
> >> > >> >>> feedback and it needed another spin but somehow fell off my plate.  :(
> >> > >> >>>
> >> > >> >>> Maybe this is something Yuvaraj would like to pick up?
> >> > >> >> It's long ago. I remember that there is no progress since my last comment.
> >> > >> >> In case of patch "3070221", I want to pick up for next Kernel.
> >> > >> >
> >> > >> > Sounds like Yuvaraj has agreed to look at addressing your comments.  Thanks!
> >> > >> >
> >> > >> >
> >> > >> >>> >> +     }
> >> > >> >>> >> +
> >> > >> >>> >>       if (cmd->flags & MMC_RSP_PRESENT) {
> >> > >> >>> >>               /* We expect a response, so set this bit */
> >> > >> >>> >>               cmdr |= SDMMC_CMD_RESP_EXP;
> >> > >> >>> >> @@ -776,11 +806,15 @@ static void dw_mci_setup_bus(struct dw_mci_slot *slot, bool
> >> > force_clkinit)
> >> > >> >>> >>       unsigned int clock = slot->clock;
> >> > >> >>> >>       u32 div;
> >> > >> >>> >>       u32 clk_en_a;
> >> > >> >>> >> +     u32 sdmmc_cmd_bits = SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT;
> >> > >> >>> >> +
> >> > >> >>> >> +     /* We must continue to set bit 28 in CMD until the change is complete */
> >> > >> >>> >> +     if (host->state == STATE_WAITING_CMD11_DONE)
> >> > >> >>> >> +             sdmmc_cmd_bits |= SDMMC_CMD_VOLT_SWITCH;
> >> > >> >>> > I didn't get the reason SDMMC_CMD_VOLT_SWITCH is needed during clock
> update(enable/disable)
> >> > >> >>> > Can you explain it in details?
> >> > >> >>>
> >> > >> >>> Simply put: if I didn't do this then the system hung during voltage
> >> > >> >>> switch.  It was not documented in the version of the IP manual that I
> >> > >> >>> had access to and it took me a bunch of digging / trial and error to
> >> > >> >>> figure this out.  Whatever this bit does internally it's important to
> >> > >> >>> set it while the voltage change is happening.  Note that this need was
> >> > >> >>> the whole reason for adding the extra state to the state machine.
> >> > >> >>>
> >> > >> >>> Perhaps Yuvaraj can try without it and I'd assume he'll report the
> >> > >> >>> same freeze.
> >> > >> >> Clarify the necessity of SDMMC_CMD_VOLT_SWITCH.
> >> > >> >> As far as I experience, we didn't apply this bit for several projects.
> >> > >> >
> >> > >> > I just tried this locally on our ChromeOS 3.8 kernel (which has quite
> >> > >> > a few backports and matches dw_mmc upstream pretty closely).  When I
> >> > >> > take out "sdmmc_cmd_bits |= SDMMC_CMD_VOLT_SWITCH" and bootup it fails
> >> > >> > to bringup WiFi chip.  It prints messages like:
> >> > >> >
> >> > >> > [    4.400842] mmc_host mmc1: Timeout sending command (cmd 0x202000
> >> > >> > arg 0x0 status 0x80202000)
> >> > >> >
> >> > >> > I will let Yuvaraj comment about his testing too.
> >> > >> To make it simple and verify, i have just enabled eMMC and SD channels
> >> > >> on 3.16-rc1 and tested.
> >> > >> Without "sdmmc_cmd_bits |= SDMMC_CMD_VOLT_SWITCH" bootup fails.
> >> > >> [    3.015971] mmc_host mmc1: Timeout sending command (cmd 0x202000
> >> > >> arg 0x0 status 0x80202000)
> >> > >> [    3.530974] mmc_host mmc1: Timeout sending command (cmd 0x202000
> >> > >> arg 0x0 status 0x80202000)
> >> > > Thank you for test.
> >> > > Hmm, it failed during clock update, right?.
> >> > > Can you check HLE interrupt at this time?
> >> > > I'll investigate for this.
> >> > Yes, it failed during clock update.I am dumping the register contents
> >> > just before the dev_err statement.I can observe that 1st timeout does
> >> > not have HLE interrupt.But once we set bit 0 of UHS_REG subsequent
> >> > timeout's have HLE bit set.I am attaching the complete log.
> >>
> >> Before disabling the clocks, we ensure that the card is not busy.(It's from Synopsys's databook)
> >> I guess SDMMC_CMD_VOLT_SWITCH flag seems to affect to avoid this limitation if you have no problem.
> >> But, I met the following log with HEL when I apply and test this patch .
> >>
> >> "mmc_host mmc1: Timeout sending command (cmd 0x202000 arg 0x0 status 0x80202000)"
> >> I need to look into it more deeply.
> > I confirmed that SDMMC_CMD_VOLT_SWITCH is needed during clock-off/on in voltage switching sequence
> > even though there is no description for this in Synopsys's databook.
> > And the above-mentioned message("mmc_host mmc1: Timeout sending command (cmd 0x202000 arg 0x0 status
> 0x80202000)")
> > is happened on mmc_power_cycle(), which should be done because of busy status of DAT[3:0].
> > It indicates clock-off/on sequence fails when executing mmc_power_cycle().
Do you have any idea for this error case?

> >
> >>
> >> > >
> >> > >> >
> >> > >> >
> >> > >> >>> >> +static int dw_mci_switch_voltage(struct mmc_host *mmc, struct mmc_ios *ios)
> >> > >> >>> >> +{
> >> > >> >>> >> +     struct dw_mci_slot *slot = mmc_priv(mmc);
> >> > >> >>> >> +     struct dw_mci *host = slot->host;
> >> > >> >>> >> +     u32 uhs;
> >> > >> >>> >> +     u32 v18 = SDMMC_UHS_18V << slot->id;
> >> > >> >>> >> +     int min_uv, max_uv;
> >> > >> >>> >> +     int ret;
> >> > >> >>> >> +
> >> > >> >>> >> +     /*
> >> > >> >>> >> +      * Program the voltage.  Note that some instances of dw_mmc may use
> >> > >> >>> >> +      * the UHS_REG for this.  For other instances (like exynos) the UHS_REG
> >> > >> >>> >> +      * does no harm but you need to set the regulator directly.  Try both.
> >> > >> >>> >> +      */
> >> > >> >>> >> +     uhs = mci_readl(host, UHS_REG);
> >> > >> >>> >> +     if (ios->signal_voltage == MMC_SIGNAL_VOLTAGE_330) {
> >> > >> >>> >> +             min_uv = 2700000;
> >> > >> >>> >> +             max_uv = 3600000;
> >> > >> >>> >> +             uhs &= ~v18;
> >> > >> >>> >> +     } else {
> >> > >> >>> >> +             min_uv = 1700000;
> >> > >> >>> >> +             max_uv = 1950000;
> >> > >> >>> >> +             uhs |= v18;
> >> > >> >>> >> +     }
> >> > >> >>> >> +     if (!IS_ERR(mmc->supply.vqmmc)) {
> >> > >> >>> >> +             ret = regulator_set_voltage(mmc->supply.vqmmc, min_uv, max_uv);
> >> > >> >>> >> +
> >> > >> >>> >> +             /*
> >> > >> >>> >> +              * Only complain if regulator claims that it's not in the 1.8V
> >> > >> >>> >> +              * range.  This avoids a bunch of errors in the case that
> >> > >> >>> >> +              * we've got a fixed 1.8V regulator but the core SD code still
> >> > >> >>> >> +              * thinks it ought to try to switch to 3.3 and then back to 1.8
> >> > >> >>> >> +              */
> >> > >> >>> >> +             if (ret) {
> >> > >> >>> > I think if ret is error, printing message and returning error is good.
> >> > >> >>> > Currently, just returning '0' though it fails.
> >> > >> >> Any feedback?
> >> > >> >
> >> > >> > Whoops, right.  I think you're right that in the case it warns it
> >> > >> > should also return the error code.  In the case it doesn't warn it
> >> > >> > shouldn't.  Also: possibly we should use a trick like the mmc core
> >> > >> > does and use "regulator_count_voltages" to figure out if we're going
> >> > >> > to be able to switch voltages...
> >> > >> >
> >> > >> >>> > Artificially adding SDMMC_INT_CMD_DONE to pending is needed to complete cmd handling?
> >> > >> >>> > Is there any reason?
> >> > >> >>>
> >> > >> >>> I don't remember this for sure since I wrote this a long time ago.
> >> > >> >>> Maybe it's not needed?  I may have just been modeling on other
> >> > >> >>> interrupts.
> >> > >> >> If there is no specific reason, please remove it.
> >> > >> >
> >> > >> > OK, we'll see how Yuvaraj's testing goes without it.  My incredibly
> >> > >> > brief testing in our local Chrome OS 3.8 tree shows that the WiFi is
> >> > >> > not detected properly if I comment out the line:
> >> > >> >   dw_mci_cmd_interrupt(host,
> >> > >> >     pending | SDMMC_INT_CMD_DONE);
> >> > >> >
> >> > >> > It may be sufficient to simply schedule the tasklet or to do some
> >> > >> > other subset of dw_mci_cmd_interrupt().  Yuvaraj can confirm on the
> >> > >> > latest kernel and also investigate further.
> >> > > How about completing CMD without SDMMC_INT_CMD_DONE?
> >> > Without this tuning fails.
> >> > [    2.525284] mmc_host mmc0: Bus speed (slot 0) = 200000000Hz (slot
> >> > req 200000000Hz, actual 200000000HZ div = 0)
> >> > [  198.556573] random: nonblocking pool is initialized
> >> > [  240.226044] INFO: task kworker/u8:0:6 blocked for more than 120 seconds.
> >> > [  240.231276]       Not tainted 3.16.0-rc1-00010-g2f7a756-dirty #55
> >> > [  240.237359] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs"
> >> > disables this message.
> >> > [  240.245150] kworker/u8:0    D c03a58a0     0     6      2 0x00000000
> >> > [  240.251487] Workqueue: kmmcd mmc_rescan
> >> > [  240.255297] [<c03a58a0>] (__schedule) from [<c03a5330>]
> >> > (schedule_timeout+0x130/0x170)
> >> > [  240.263201] [<c03a5330>] (schedule_timeout) from [<c03a6540>]
> >> > (wait_for_common+0xbc/0x14c)
> >> > [  240.271441] [<c03a6540>] (wait_for_common) from [<c02ca6e4>]
> >> > (mmc_wait_for_req_done+0x6c/0xf0)
> >> > [  240.280031] [<c02ca6e4>] (mmc_wait_for_req_done) from [<c02e3f10>]
> >> > (dw_mci_exynos_execute_tuning+0x1d4/0x2c4)
> >> > [  240.289919] [<c02e3f10>] (dw_mci_exynos_execute_tuning) from
> >> > [<c02e0198>] (dw_mci_execute_tuning+0x54/0xb4)
> >> > [  240.299635] [<c02e0198>] (dw_mci_execute_tuning) from [<c02cf47c>]
> >> > (mmc_init_card+0xeec/0x14ac)
> >> > [  240.308309] [<c02cf47c>] (mmc_init_card) from [<c02cfc34>]
> >> > (mmc_attach_mmc+0x8c/0x160)
> >> > [  240.316202] [<c02cfc34>] (mmc_attach_mmc) from [<c02cca0c>]
> >> > (mmc_rescan+0x2b0/0x308)
> >> > [  240.323924] [<c02cca0c>] (mmc_rescan) from [<c0032f60>]
> >> > (process_one_work+0xf8/0x364)
> >> > [  240.331732] [<c0032f60>] (process_one_work) from [<c0033800>]
> >> > (worker_thread+0x50/0x5a0)
> >> > [  240.339799] [<c0033800>] (worker_thread) from [<c0038e7c>]
> >> > (kthread+0xd8/0xf0)
> >> > [  240.346999] [<c0038e7c>] (kthread) from [<c000e438>]
> >> > (ret_from_fork+0x14/0x3c)
> >>
> >> Thank you for test.
> >> I want to clear it why CMD_DONE is related to tuning.
> > In my case, SDMMC_INT_CMD_DONE doesn't affect.
> I rechecked this.SDMMC_INT_CMD_DONE flag doesn't affect.Simply
> sheduling the tasklet is enough.Sorry,I misunderstood your comments
> earlier.
> > And do you check VOLT_SWITCH interrupt occurs twice during voltage-switching.
> No.Only once.

VOLT_SWITCH interrupt occurs totally twice. First is under CMD11. And second is after 1.8V voltage completed.
I actually found twice. Please check Voltage Switch Normal Scenario in the manual.


Thanks,
Seungwon Jeon

> > For last one, if-statement in dw_mci_interrupt() disturbs clearing interrupt.
> > if ((host->state == STATE_SENDING_CMD11) &&
> >
> > In addition, I suggest adding another flag instead of using host's state.
> > Can you consider it?
> Sure.I will check this.
> >
> > Thanks,
> > Seungwon Jeon
> >
> --
> To unsubscribe from this list: send the line "unsubscribe linux-mmc" in
> the body of a message to majordomo at vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 3/3] mmc: dw_mmc: Support voltage changes

[Yuvaraj Kumar]

On Wed, Jun 25, 2014 at 11:16 PM, Doug Anderson <dianders@chromium.org> wrote:
> Seungwon,
>
> On Wed, Jun 25, 2014 at 6:08 AM, Seungwon Jeon <tgih.jun@samsung.com> wrote:
>> On Mon, June 23, 2014, Yuvaraj Kumar C D wrote:
>>> Subject: [PATCH 3/3] mmc: dw_mmc: Support voltage changes
>>>
>>> From: Doug Anderson <dianders@chromium.org>
>>>
>>> For UHS cards we need the ability to switch voltages from 3.3V to
>>> 1.8V.  Add support to the dw_mmc driver to handle this.  Note that
>>> dw_mmc needs a little bit of extra code since the interface needs a
>>> special bit programmed to the CMD register while CMD11 is progressing.
>>> This means adding a few extra states to the state machine to track.
>>
>> Overall new additional states makes it complicated.
>> Can we do that in other way?
>
> That was the best I was able to figure out when I thought this
> through.  If you have ideas for doing it another way I'd imagine that
> Yuvaraj would be happy to take your feedback.
Yes.
>
>
>>> Signed-off-by: Doug Anderson <dianders@chromium.org>
>>> Signed-off-by: Yuvaraj Kumar C D <yuvaraj.cd@samsung.com>
>>>
>>> ---
>>>  drivers/mmc/host/dw_mmc.c  |  145 +++++++++++++++++++++++++++++++++++++++++---
>>>  drivers/mmc/host/dw_mmc.h  |    5 +-
>>>  include/linux/mmc/dw_mmc.h |    2 +
>>>  3 files changed, 142 insertions(+), 10 deletions(-)
>>>
>>> diff --git a/drivers/mmc/host/dw_mmc.c b/drivers/mmc/host/dw_mmc.c
>>> index e034bce..38eb548 100644
>>> --- a/drivers/mmc/host/dw_mmc.c
>>> +++ b/drivers/mmc/host/dw_mmc.c
>>> @@ -29,6 +29,7 @@
>>>  #include <linux/irq.h>
>>>  #include <linux/mmc/host.h>
>>>  #include <linux/mmc/mmc.h>
>>> +#include <linux/mmc/sd.h>
>>>  #include <linux/mmc/sdio.h>
>>>  #include <linux/mmc/dw_mmc.h>
>>>  #include <linux/bitops.h>
>>> @@ -235,10 +236,13 @@ err:
>>>  }
>>>  #endif /* defined(CONFIG_DEBUG_FS) */
>>>
>>> +static void mci_send_cmd(struct dw_mci_slot *slot, u32 cmd, u32 arg);
>>> +
>>>  static u32 dw_mci_prepare_command(struct mmc_host *mmc, struct mmc_command *cmd)
>>>  {
>>>       struct mmc_data *data;
>>>       struct dw_mci_slot *slot = mmc_priv(mmc);
>>> +     struct dw_mci *host = slot->host;
>>>       const struct dw_mci_drv_data *drv_data = slot->host->drv_data;
>>>       u32 cmdr;
>>>       cmd->error = -EINPROGRESS;
>>> @@ -254,6 +258,32 @@ static u32 dw_mci_prepare_command(struct mmc_host *mmc, struct mmc_command *cmd)
>>>       else if (cmd->opcode != MMC_SEND_STATUS && cmd->data)
>>>               cmdr |= SDMMC_CMD_PRV_DAT_WAIT;
>>>
>>> +     if (cmd->opcode == SD_SWITCH_VOLTAGE) {
>>> +             u32 clk_en_a;
>>> +
>>> +             /* Special bit makes CMD11 not die */
>>> +             cmdr |= SDMMC_CMD_VOLT_SWITCH;
>>> +
>>> +             /* Change state to continue to handle CMD11 weirdness */
>>> +             WARN_ON(slot->host->state != STATE_SENDING_CMD);
>>> +             slot->host->state = STATE_SENDING_CMD11;
>>> +
>>> +             /*
>>> +              * We need to disable clock stop while doing voltage switch
>>> +              * according to 7.4.1.2 Voltage Switch Normal Scenario.
>>> +              *
>>> +              * It's assumed that by the next time the CLKENA is updated
>>> +              * (when we set the clock next) that the voltage change will
>>> +              * be over, so we don't bother setting any bits to synchronize
>>> +              * with dw_mci_setup_bus().
>>> +              */
>>> +             clk_en_a = mci_readl(host, CLKENA);
>>> +             clk_en_a &= ~(SDMMC_CLKEN_LOW_PWR << slot->id);
>>> +             mci_writel(host, CLKENA, clk_en_a);
>>> +             mci_send_cmd(slot, SDMMC_CMD_UPD_CLK |
>>> +                          SDMMC_CMD_PRV_DAT_WAIT, 0);
>> dw_mci_disable_low_power() can be used here.
>
> Ah.  I guess we don't have that locally anymore.  Locally we have variants on:
> * https://patchwork.kernel.org/patch/3070311/
> * https://patchwork.kernel.org/patch/3070251/
> * https://patchwork.kernel.org/patch/3070221/
>
> ...which removed that function.  ...but I guess upstream never picked
> up those patches, huh?  Looking back it looks like you had some
> feedback and it needed another spin but somehow fell off my plate.  :(
>
> Maybe this is something Yuvaraj would like to pick up?
ok.
>
>>
>>> +     }
>>> +
>>>       if (cmd->flags & MMC_RSP_PRESENT) {
>>>               /* We expect a response, so set this bit */
>>>               cmdr |= SDMMC_CMD_RESP_EXP;
>>> @@ -776,11 +806,15 @@ static void dw_mci_setup_bus(struct dw_mci_slot *slot, bool force_clkinit)
>>>       unsigned int clock = slot->clock;
>>>       u32 div;
>>>       u32 clk_en_a;
>>> +     u32 sdmmc_cmd_bits = SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT;
>>> +
>>> +     /* We must continue to set bit 28 in CMD until the change is complete */
>>> +     if (host->state == STATE_WAITING_CMD11_DONE)
>>> +             sdmmc_cmd_bits |= SDMMC_CMD_VOLT_SWITCH;
>> I didn't get the reason SDMMC_CMD_VOLT_SWITCH is needed during clock update(enable/disable)
>> Can you explain it in details?
>
> Simply put: if I didn't do this then the system hung during voltage
> switch.  It was not documented in the version of the IP manual that I
> had access to and it took me a bunch of digging / trial and error to
> figure this out.  Whatever this bit does internally it's important to
> set it while the voltage change is happening.  Note that this need was
> the whole reason for adding the extra state to the state machine.
>
> Perhaps Yuvaraj can try without it and I'd assume he'll report the
> same freeze.
sure, i will try this.
>
>
>>>       if (!clock) {
>>>               mci_writel(host, CLKENA, 0);
>>> -             mci_send_cmd(slot,
>>> -                          SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT, 0);
>>> +             mci_send_cmd(slot, sdmmc_cmd_bits, 0);
>>>       } else if (clock != host->current_speed || force_clkinit) {
>>>               div = host->bus_hz / clock;
>>>               if (host->bus_hz % clock && host->bus_hz > clock)
>>> @@ -804,15 +838,13 @@ static void dw_mci_setup_bus(struct dw_mci_slot *slot, bool force_clkinit)
>>>               mci_writel(host, CLKSRC, 0);
>>>
>>>               /* inform CIU */
>>> -             mci_send_cmd(slot,
>>> -                          SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT, 0);
>>> +             mci_send_cmd(slot, sdmmc_cmd_bits, 0);
>>>
>>>               /* set clock to desired speed */
>>>               mci_writel(host, CLKDIV, div);
>>>
>>>               /* inform CIU */
>>> -             mci_send_cmd(slot,
>>> -                          SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT, 0);
>>> +             mci_send_cmd(slot, sdmmc_cmd_bits, 0);
>>>
>>>               /* enable clock; only low power if no SDIO */
>>>               clk_en_a = SDMMC_CLKEN_ENABLE << slot->id;
>>> @@ -821,8 +853,7 @@ static void dw_mci_setup_bus(struct dw_mci_slot *slot, bool force_clkinit)
>>>               mci_writel(host, CLKENA, clk_en_a);
>>>
>>>               /* inform CIU */
>>> -             mci_send_cmd(slot,
>>> -                          SDMMC_CMD_UPD_CLK | SDMMC_CMD_PRV_DAT_WAIT, 0);
>>> +             mci_send_cmd(slot, sdmmc_cmd_bits, 0);
>>>
>>>               /* keep the clock with reflecting clock dividor */
>>>               slot->__clk_old = clock << div;
>>> @@ -898,6 +929,17 @@ static void dw_mci_queue_request(struct dw_mci *host, struct dw_mci_slot *slot,
>>>
>>>       slot->mrq = mrq;
>>>
>>> +     if (host->state == STATE_WAITING_CMD11_DONE) {
>>> +             dev_warn(&slot->mmc->class_dev,
>>> +                      "Voltage change didn't complete\n");
>>> +             /*
>>> +              * this case isn't expected to happen, so we can
>>> +              * either crash here or just try to continue on
>>> +              * in the closest possible state
>>> +              */
>>> +             host->state = STATE_IDLE;
>>> +     }
>>> +
>>>       if (host->state == STATE_IDLE) {
>>>               host->state = STATE_SENDING_CMD;
>>>               dw_mci_start_request(host, slot);
>>> @@ -993,6 +1035,9 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
>>>       /* Slot specific timing and width adjustment */
>>>       dw_mci_setup_bus(slot, false);
>>>
>>> +     if (slot->host->state == STATE_WAITING_CMD11_DONE && ios->clock != 0)
>>> +             slot->host->state = STATE_IDLE;
>>> +
>>>       switch (ios->power_mode) {
>>>       case MMC_POWER_UP:
>>>               if ((!IS_ERR(mmc->supply.vmmc)) &&
>>> @@ -1039,6 +1084,68 @@ static void dw_mci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
>>>       }
>>>  }
>>>
>>> +static int dw_mci_card_busy(struct mmc_host *mmc)
>>> +{
>>> +     struct dw_mci_slot *slot = mmc_priv(mmc);
>>> +     u32 status;
>>> +
>>> +     /*
>>> +      * Check the busy bit which is low when DAT[3:0]
>>> +      * (the data lines) are 0000
>>> +      */
>>> +     status = mci_readl(slot->host, STATUS);
>>> +
>>> +     return !!(status & SDMMC_STATUS_BUSY);
>>> +}
>>> +
>>> +static int dw_mci_switch_voltage(struct mmc_host *mmc, struct mmc_ios *ios)
>>> +{
>>> +     struct dw_mci_slot *slot = mmc_priv(mmc);
>>> +     struct dw_mci *host = slot->host;
>>> +     u32 uhs;
>>> +     u32 v18 = SDMMC_UHS_18V << slot->id;
>>> +     int min_uv, max_uv;
>>> +     int ret;
>>> +
>>> +     /*
>>> +      * Program the voltage.  Note that some instances of dw_mmc may use
>>> +      * the UHS_REG for this.  For other instances (like exynos) the UHS_REG
>>> +      * does no harm but you need to set the regulator directly.  Try both.
>>> +      */
>>> +     uhs = mci_readl(host, UHS_REG);
>>> +     if (ios->signal_voltage == MMC_SIGNAL_VOLTAGE_330) {
>>> +             min_uv = 2700000;
>>> +             max_uv = 3600000;
>>> +             uhs &= ~v18;
>>> +     } else {
>>> +             min_uv = 1700000;
>>> +             max_uv = 1950000;
>>> +             uhs |= v18;
>>> +     }
>>> +     if (!IS_ERR(mmc->supply.vqmmc)) {
>>> +             ret = regulator_set_voltage(mmc->supply.vqmmc, min_uv, max_uv);
>>> +
>>> +             /*
>>> +              * Only complain if regulator claims that it's not in the 1.8V
>>> +              * range.  This avoids a bunch of errors in the case that
>>> +              * we've got a fixed 1.8V regulator but the core SD code still
>>> +              * thinks it ought to try to switch to 3.3 and then back to 1.8
>>> +              */
>>> +             if (ret) {
>> I think if ret is error, printing message and returning error is good.
>> Currently, just returning '0' though it fails.
ok
>>
>>> +                     int cur_voltage = 0;
>>> +
>>> +                     cur_voltage = regulator_get_voltage(mmc->supply.vqmmc);
>>> +                     if (cur_voltage < 1700000 || cur_voltage > 1950000)
>>> +                             dev_warn(&mmc->class_dev,
>>> +                                      "Regulator set error %d: %d - %d\n",
>>> +                                      ret, min_uv, max_uv);
>>> +             }
>>> +     }
>>> +     mci_writel(host, UHS_REG, uhs);
>>> +
>>> +     return 0;
>>> +}
>>> +
>>>  static int dw_mci_get_ro(struct mmc_host *mmc)
>>>  {
>>>       int read_only;
>>> @@ -1180,6 +1287,9 @@ static const struct mmc_host_ops dw_mci_ops = {
>>>       .get_cd                 = dw_mci_get_cd,
>>>       .enable_sdio_irq        = dw_mci_enable_sdio_irq,
>>>       .execute_tuning         = dw_mci_execute_tuning,
>>> +     .card_busy              = dw_mci_card_busy,
>>> +     .start_signal_voltage_switch = dw_mci_switch_voltage,
>>> +
>>>  };
>>>
>>>  static void dw_mci_request_end(struct dw_mci *host, struct mmc_request *mrq)
>>> @@ -1203,7 +1313,11 @@ static void dw_mci_request_end(struct dw_mci *host, struct mmc_request *mrq)
>>>               dw_mci_start_request(host, slot);
>>>       } else {
>>>               dev_vdbg(host->dev, "list empty\n");
>>> -             host->state = STATE_IDLE;
>>> +
>>> +             if (host->state == STATE_SENDING_CMD11)
>>> +                     host->state = STATE_WAITING_CMD11_DONE;
>>> +             else
>>> +                     host->state = STATE_IDLE;
>>>       }
>>>
>>>       spin_unlock(&host->lock);
>>> @@ -1314,8 +1428,10 @@ static void dw_mci_tasklet_func(unsigned long priv)
>>>
>>>               switch (state) {
>>>               case STATE_IDLE:
>>> +             case STATE_WAITING_CMD11_DONE:
>>>                       break;
>>>
>>> +             case STATE_SENDING_CMD11:
>>>               case STATE_SENDING_CMD:
>>>                       if (!test_and_clear_bit(EVENT_CMD_COMPLETE,
>>>                                               &host->pending_events))
>>> @@ -1870,6 +1986,15 @@ static irqreturn_t dw_mci_interrupt(int irq, void *dev_id)
>>>       }
>>>
>>>       if (pending) {
>>> +             /* Check volt switch first, since it can look like an error */
>>> +             if ((host->state == STATE_SENDING_CMD11) &&
>>> +                 (pending & SDMMC_INT_VOLT_SWITCH)) {
>>> +                     mci_writel(host, RINTSTS, SDMMC_INT_VOLT_SWITCH);
>>> +                     pending &= ~SDMMC_INT_VOLT_SWITCH;
>>> +                     dw_mci_cmd_interrupt(host,
>>> +                                          pending | SDMMC_INT_CMD_DONE);
>> Artificially adding SDMMC_INT_CMD_DONE to pending is needed to complete cmd handling?
>> Is there any reason?
>
> I don't remember this for sure since I wrote this a long time ago.
> Maybe it's not needed?  I may have just been modeling on other
> interrupts.
>
> -Doug