Re: [PATCH v4 1/7] mmc: dt-bindings: add Mediatek MMC bindings

Re: [PATCH v4 1/7] mmc: dt-bindings: add Mediatek MMC bindings

[Ulf Hansson]

On 19 May 2015 at 08:36, Chaotian Jing <chaotian.jing@mediatek.com> wrote:
> Document the device-tree binding of Mediatek MMC host
>
> Signed-off-by: Chaotian Jing <chaotian.jing@mediatek.com>
> ---
>  Documentation/devicetree/bindings/mmc/mtk-sd.txt | 33 ++++++++++++++++++++++++
>  1 file changed, 33 insertions(+)
>  create mode 100644 Documentation/devicetree/bindings/mmc/mtk-sd.txt
>
> diff --git a/Documentation/devicetree/bindings/mmc/mtk-sd.txt b/Documentation/devicetree/bindings/mmc/mtk-sd.txt
> new file mode 100644
> index 0000000..ba6d668
> --- /dev/null
> +++ b/Documentation/devicetree/bindings/mmc/mtk-sd.txt
> @@ -0,0 +1,33 @@
> +* MTK MMC controller
> +
> +The MTK  MSDC can act as a MMC controller
> +to support MMC, SD, and SDIO types of memory cards.
> +
> +This file documents differences between the core properties in mmc.txt
> +and the properties used by the msdc driver.
> +
> +Required properties:
> +- compatible: Should be "mediatek,mt8173-mmc","mediatek,mt8135-mmc".
> +- interrupts: Should contain MSDC interrupt number
> +- clocks: MSDC source clock, HCLK, HCLK is mandatory when it is not controlled by Infra

This seems strange. Either it's mandatory since it's under the
"Required properties" section or it should be moved to the "Optional
properties" section.

> +- clock-names: "source", "hclk"
> +- pinctrl-names: should be "default", "state_uhs"
> +- pinctrl-0: should contain default/high speed pin ctrl
> +- pinctrl-1: should contain uhs mode pin ctrl
> +- vmmc-supply: power to the Core
> +
> +Optional properties:
> +- vqmmc-supply: power to the IO(for SD/SDIO)
> +
> +Examples:
> +mmc0: mmc@11230000 {
> +       compatible = "mediatek,mt8173-mmc", "mediatek,mt8135-mmc";
> +       reg = <0 0x11230000 0 0x108>;
> +       interrupts = <GIC_SPI 39 IRQ_TYPE_LEVEL_LOW>;
> +       vmmc-supply = <&mt6397_vemc_3v3_reg>;
> +       clocks = <&pericfg CLK_PERI_MSDC30_0>, <&topckgen CLK_TOP_MSDC50_0_H_SEL>;
> +       clock-names = "source", "hclk";
> +       pinctrl-names = "default", "state_uhs";
> +       pinctrl-0 = <&mmc0_pins_default>;
> +       pinctrl-1 = <&mmc0_pins_uhs>;
> +};
> --
> 1.8.1.1.dirty
>

Kind regards
Uffe

Re: [PATCH v4 3/7] mmc: mediatek: Add PM support for MMC driver

[Ulf Hansson]

On 19 May 2015 at 08:36, Chaotian Jing <chaotian.jing@mediatek.com> wrote:
> Add PM support for Mediatek MMC driver
> Save/restore registers when PM
>
> Signed-off-by: Chaotian Jing <chaotian.jing@mediatek.com>
> ---
>  drivers/mmc/host/mtk-sd.c | 99 +++++++++++++++++++++++++++++++++++++++++++++--
>  1 file changed, 95 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/mmc/host/mtk-sd.c b/drivers/mmc/host/mtk-sd.c
> index 26e9590..c0b2e2d 100644
> --- a/drivers/mmc/host/mtk-sd.c
> +++ b/drivers/mmc/host/mtk-sd.c
> @@ -22,6 +22,7 @@
>  #include <linux/of_gpio.h>
>  #include <linux/pinctrl/consumer.h>
>  #include <linux/platform_device.h>
> +#include <linux/pm_runtime.h>
>  #include <linux/regulator/consumer.h>
>  #include <linux/spinlock.h>
>
> @@ -212,6 +213,7 @@
>  #define MSDC_ASYNC_FLAG (0x1 << 1)
>  #define MSDC_MMAP_FLAG (0x1 << 2)
>
> +#define MTK_MMC_AUTOSUSPEND_DELAY      50
>  #define CMD_TIMEOUT         (HZ/10 * 5)        /* 100ms x5 */
>  #define DAT_TIMEOUT         (HZ    * 5)        /* 1000ms x5 */
>
> @@ -254,6 +256,15 @@ struct msdc_dma {
>         dma_addr_t bd_addr;     /* the physical address of bd array */
>  };
>
> +struct msdc_save_para {
> +       u32 msdc_cfg;
> +       u32 iocon;
> +       u32 sdc_cfg;
> +       u32 pad_tune;
> +       u32 patch_bit0;
> +       u32 patch_bit1;
> +};
> +
>  struct msdc_host {
>         struct device *dev;
>         struct mmc_host *mmc;   /* mmc structure */
> @@ -287,6 +298,8 @@ struct msdc_host {
>         bool vqmmc_enabled;
>         bool sclk_enabled;      /* source clock enabled */
>         bool hclk_enabled;      /* Hclk enabled */
> +       bool in_suspend;        /* used for apply restore regs */

You shouldn't need to track this.

> +       struct msdc_save_para save_para; /* used when gate HCLK */
>  };
>
>  static void sdr_set_bits(void __iomem *reg, u32 bs)
> @@ -459,6 +472,30 @@ static void msdc_set_timeout(struct msdc_host *host, u32 ns, u32 clks)
>         sdr_set_field(host->base + SDC_CFG, SDC_CFG_DTOC, timeout);
>  }
>
> +static void msdc_save_reg(struct msdc_host *host)
> +{
> +       host->save_para.msdc_cfg = readl(host->base + MSDC_CFG);
> +       host->save_para.iocon = readl(host->base + MSDC_IOCON);
> +       host->save_para.sdc_cfg = readl(host->base + SDC_CFG);
> +       host->save_para.pad_tune = readl(host->base + MSDC_PAD_TUNE);
> +       host->save_para.patch_bit0 = readl(host->base + MSDC_PATCH_BIT);
> +       host->save_para.patch_bit1 = readl(host->base + MSDC_PATCH_BIT1);
> +       host->in_suspend = true;
> +}
> +
> +static void msdc_restore_reg(struct msdc_host *host)
> +{
> +       if (host->in_suspend == false)
> +               return;
> +       writel(host->save_para.msdc_cfg, host->base + MSDC_CFG);
> +       writel(host->save_para.iocon, host->base + MSDC_IOCON);
> +       writel(host->save_para.sdc_cfg, host->base + SDC_CFG);
> +       writel(host->save_para.pad_tune, host->base + MSDC_PAD_TUNE);
> +       writel(host->save_para.patch_bit0, host->base + MSDC_PATCH_BIT);
> +       writel(host->save_para.patch_bit1, host->base + MSDC_PATCH_BIT1);
> +       host->in_suspend = false;
> +}
> +
>  static void msdc_disable_src_clk(struct msdc_host *host)
>  {
>         if (host->sclk_enabled) {
> @@ -479,6 +516,10 @@ static void msdc_enable_src_clk(struct msdc_host *host)
>
>  static void msdc_gate_clock(struct msdc_host *host)
>  {
> +       /* Save register first, only when clk is on */
> +       if (host->sclk_enabled)
> +               msdc_save_reg(host);

Please move msdc_save_reg() outside the msdc_gate_clock() function,
since I think these things should be somewhat decoupled for each
other.

> +
>         msdc_disable_src_clk(host);
>         if (!IS_ERR(host->h_clk) && host->hclk_enabled) {
>                 clk_disable_unprepare(host->h_clk);
> @@ -493,6 +534,8 @@ static void msdc_ungate_clock(struct msdc_host *host)
>                 host->hclk_enabled = true;
>         }
>         msdc_enable_src_clk(host);
> +       /* Then restore register */
> +       msdc_restore_reg(host);

As above, please do msdc_restore_reg() outside of msdc_ungate_clock().

>  }
>
>  static void msdc_set_mclk(struct msdc_host *host, int ddr, u32 hz)
> @@ -710,6 +753,9 @@ static void msdc_request_done(struct msdc_host *host, struct mmc_request *mrq)
>         if (mrq->data)
>                 msdc_unprepare_data(host, mrq);
>         mmc_request_done(host->mmc, mrq);
> +
> +       pm_runtime_mark_last_busy(host->dev);
> +       pm_runtime_put_autosuspend(host->dev);
>  }
>
>  /* returns true if command is fully handled; returns false otherwise */
> @@ -866,6 +912,8 @@ static void msdc_ops_request(struct mmc_host *mmc, struct mmc_request *mrq)
>                 WARN_ON(host->mrq);
>         spin_unlock_irqrestore(&host->lock, flags);
>
> +       pm_runtime_get_sync(host->dev);
> +
>         if (mrq->data)
>                 msdc_prepare_data(host, mrq);
>
> @@ -1130,7 +1178,8 @@ static void msdc_init_hw(struct msdc_host *host)
>         sdr_set_field(host->base + MSDC_PATCH_BIT, MSDC_CKGEN_MSDC_DLY_SEL, 1);
>         writel(0xffff0089, host->base + MSDC_PATCH_BIT1);
>         /* Configure to enable SDIO mode.
> -          it's must otherwise sdio cmd5 failed */
> +        * it's must otherwise sdio cmd5 failed
> +        */

White space.

>         sdr_set_bits(host->base + SDC_CFG, SDC_CFG_SDIO);
>
>         /* disable detect SDIO device interrupt function */
> @@ -1185,6 +1234,8 @@ static void msdc_ops_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
>         int ret;
>         u32 ddr = 0;
>
> +       pm_runtime_get_sync(host->dev);
> +
>         if (ios->timing == MMC_TIMING_UHS_DDR50 ||
>                         ios->timing == MMC_TIMING_MMC_DDR52)
>                 ddr = 1;
> @@ -1200,7 +1251,7 @@ static void msdc_ops_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
>                                         ios->vdd);
>                         if (ret) {
>                                 dev_err(host->dev, "Failed to set vmmc power!\n");
> -                               return;
> +                               goto end;
>                         }
>                 }
>                 break;
> @@ -1234,6 +1285,10 @@ static void msdc_ops_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
>
>         if (host->mclk != ios->clock || host->ddr != ddr)
>                 msdc_set_mclk(host, ddr, ios->clock);
> +
> +end:
> +       pm_runtime_mark_last_busy(host->dev);
> +       pm_runtime_put_autosuspend(host->dev);
>  }
>
>  static struct mmc_host_ops mt_msdc_ops = {
> @@ -1355,12 +1410,18 @@ static int msdc_drv_probe(struct platform_device *pdev)
>         if (ret)
>                 goto release;
>
> +       pm_runtime_set_active(host->dev);
> +       pm_runtime_set_autosuspend_delay(host->dev, MTK_MMC_AUTOSUSPEND_DELAY);
> +       pm_runtime_use_autosuspend(host->dev);
> +       pm_runtime_enable(host->dev);
>         ret = mmc_add_host(mmc);
> +
>         if (ret)
> -               goto release;
> +               goto end;
>
>         return 0;
> -
> +end:
> +       pm_runtime_disable(host->dev);
>  release:
>         platform_set_drvdata(pdev, NULL);
>         msdc_deinit_hw(host);
> @@ -1388,10 +1449,15 @@ static int msdc_drv_remove(struct platform_device *pdev)
>         mmc = platform_get_drvdata(pdev);
>         host = mmc_priv(mmc);
>
> +       pm_runtime_get_sync(host->dev);
> +
>         platform_set_drvdata(pdev, NULL);
>         mmc_remove_host(host->mmc);
>         msdc_deinit_hw(host);
> +       msdc_gate_clock(host);

As per comment on patch 2. This clock gating should be a part of that
patch instead.

>
> +       pm_runtime_disable(host->dev);
> +       pm_runtime_put_noidle(host->dev);
>         dma_free_coherent(&pdev->dev,
>                         sizeof(struct mt_gpdma_desc),
>                         host->dma.gpd, host->dma.gpd_addr);
> @@ -1403,6 +1469,30 @@ static int msdc_drv_remove(struct platform_device *pdev)
>         return 0;
>  }
>
> +#ifdef CONFIG_PM
> +static int msdc_runtime_suspend(struct device *dev)
> +{
> +       struct mmc_host *mmc = dev_get_drvdata(dev);
> +       struct msdc_host *host = mmc_priv(mmc);
> +
> +       msdc_gate_clock(host);
> +       return 0;
> +}
> +
> +static int msdc_runtime_resume(struct device *dev)
> +{
> +       struct mmc_host *mmc = dev_get_drvdata(dev);
> +       struct msdc_host *host = mmc_priv(mmc);
> +
> +       msdc_ungate_clock(host);
> +       return 0;
> +}
> +#endif
> +
> +static const struct dev_pm_ops msdc_dev_pm_ops = {
> +       SET_RUNTIME_PM_OPS(msdc_runtime_suspend, msdc_runtime_resume, NULL)
> +};
> +
>  static const struct of_device_id msdc_of_ids[] = {
>         {   .compatible = "mediatek,mt8135-mmc", },
>         {}
> @@ -1414,6 +1504,7 @@ static struct platform_driver mt_msdc_driver = {
>         .driver = {
>                 .name = "mtk-msdc",
>                 .of_match_table = msdc_of_ids,
> +               .pm = &msdc_dev_pm_ops,
>         },
>  };
>
> --
> 1.8.1.1.dirty
>

Kind regards
Uffe

Re: [PATCH v4 3/7] mmc: mediatek: Add PM support for MMC driver

[Sascha Hauer]

Hi Chaotian,

On Tue, May 19, 2015 at 02:36:47PM +0800, Chaotian Jing wrote:
> Add PM support for Mediatek MMC driver
> Save/restore registers when PM

Has anyone asked you to add a new driver in two steps? I find it hard to
review if a driver is introduced and changed right afterwards.
Presumably you find it hard to maintain the patches like this also.
Please ignore if this is indeed requested by someone.

Sascha

-- 
Pengutronix e.K.                           |                             |
Industrial Linux Solutions                 | http://www.pengutronix.de/  |
Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0    |
Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |

Re: [PATCH v4 2/7] mmc: mediatek: Add Mediatek MMC driver

[Sascha Hauer]

On Tue, May 19, 2015 at 02:36:46PM +0800, Chaotian Jing wrote:
> Add Mediatek MMC driver code
> Support eMMC/SD/SDIO
> 
> Signed-off-by: Chaotian Jing <chaotian.jing@mediatek.com>
> ---
>  drivers/mmc/host/Kconfig  |    8 +
>  drivers/mmc/host/Makefile |    1 +
>  drivers/mmc/host/mtk-sd.c | 1422 +++++++++++++++++++++++++++++++++++++++++++++
>  3 files changed, 1431 insertions(+)
>  create mode 100644 drivers/mmc/host/mtk-sd.c
> 
> +
> +	host->src_clk = devm_clk_get(&pdev->dev, "source");
> +	if (IS_ERR(host->src_clk)) {
> +		ret = PTR_ERR(host->src_clk);
> +		goto host_free;
> +	}
> +
> +	host->h_clk = devm_clk_get(&pdev->dev, "hclk");
> +	if (IS_ERR(host->h_clk)) {
> +		/* host->h_clk is optional, Only for MSDC0/3 at MT8173 */
> +		dev_dbg(&pdev->dev,
> +				"Invalied hclk from the device tree!\n");
> +	}

s/Invalied/Invalid/

According to my reference manual the controller always needs a hclk. It
seems on some controllers it is just not software controllable. If
that's the case you should always provide a hclk to the driver and make
this clock mandatory.

Sascha

-- 
Pengutronix e.K.                           |                             |
Industrial Linux Solutions                 | http://www.pengutronix.de/  |
Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0    |
Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |

Re: [PATCH v4 2/7] mmc: mediatek: Add Mediatek MMC driver

[Russell King - ARM Linux]

On Tue, May 19, 2015 at 02:36:46PM +0800, Chaotian Jing wrote:
> +struct mt_bdma_desc {
> +	u32 first_u32;
> +#define BDMA_DESC_EOL		(0x1 << 0)
> +#define BDMA_DESC_CHECKSUM	(0xff << 8) /* bit8 ~ bit15 */
> +#define BDMA_DESC_BLKPAD	(0x1 << 17)
> +#define BDMA_DESC_DWPAD		(0x1 << 18)
> +	u32 next;
> +	u32 ptr;
> +	u32 second_u32;
> +#define BDMA_DESC_BUFLEN	(0xffff) /* bit0 ~ bit15 */
> +};

"first_u32" and "second_u32" aren't descriptive names for these, and we can
see from the above that they're actually wrong (the second_u32 would be the
"next" pointer.)  Please try to come up with better names for structure
members.

> +static void sdr_set_bits(void __iomem *reg, u32 bs)
> +{
> +	u32 val = readl(reg);
> +
> +	val |= bs;
> +	writel(val, reg);
> +}
> +
> +static void sdr_clr_bits(void __iomem *reg, u32 bs)
> +{
> +	u32 val = readl(reg);
> +
> +	val &= ~(u32)bs;
> +	writel(val, reg);
> +}

What makes these read-modify-write operations safe?

That (u32) cast is not required.

> +static void sdr_set_field(void __iomem *reg, u32 field, u32 val)
> +{
> +	unsigned int tv = readl(reg);
> +
> +	tv &= ~field;
> +	tv |= ((val) << (ffs((unsigned int)field) - 1));
> +	writel(tv, reg);
> +}

Or this safe?

If you have another thread of execution (eg, an interrupt, or another CPU)
running concurrently, writing to the same register that the above code is
trying to modify, your results will be indeterminant.

> +static inline void msdc_dma_setup(struct msdc_host *host, struct msdc_dma *dma,
> +		struct mmc_data *data)
> +{
> +	u32 sglen, j;

Please don't use a 32-bit integer for things which you don't require to be
32-bit.  data->sg_len is "unsigned int", so sglen and j should be too,
since you're using those to deal with data->sg_len.  In any case, you
may wish to review this given a comment below.

> +	u32 dma_address, dma_len;

So DMA addresses and DMA lengths are 32-bit integers here.  This is not
entirely correct. DMA addresses can be 32 or 64-bit depending on the
kernel configuration.  Hence, they're typed as dma_addr_t.  DMA length
is unsigned int.

> +	struct scatterlist *sg;
> +	struct mt_gpdma_desc *gpd;
> +	struct mt_bdma_desc *bd;
> +
> +	sglen = data->sg_len;
> +	sg = data->sg;
> +
> +	gpd = dma->gpd;
> +	bd = dma->bd;
> +
> +	/* modify gpd */
> +	gpd->first_u32 |= GPDMA_DESC_HWO;
> +	gpd->first_u32 |= GPDMA_DESC_BDP;
> +	/* need to clear first. use these bits to calc checksum */
> +	gpd->first_u32 &= ~GPDMA_DESC_CHECKSUM;
> +	gpd->first_u32 |= msdc_dma_calcs((u8 *) gpd, 16) << 8;
> +
> +	/* modify bd */
> +	for (j = 0; j < sglen; j++) {

This is not how you iterate over scatterlists.  There is a helper for this.

	for_each_sg(data->sg, sg, sglen, j) {

and omit the sg++ below.

> +		dma_address = sg_dma_address(sg);
> +		dma_len = sg_dma_len(sg);

It may make some sense to BUG_ON(dma_address > (u32)~0) here, to catch
a DMA address greater than 32-bit, which would be a data-corrupting event.

> +
> +		/* init bd */
> +		bd[j].first_u32 &= ~BDMA_DESC_BLKPAD;
> +		bd[j].first_u32 &= ~BDMA_DESC_DWPAD;
> +		bd[j].ptr = (u32)dma_address;
> +		bd[j].second_u32 &= ~BDMA_DESC_BUFLEN;
> +		bd[j].second_u32 |= (dma_len & BDMA_DESC_BUFLEN);

I hope that you've told the rest of the system what your maximum buffer
length for any scatterlist entry is.

> +
> +		if (j == sglen - 1) /* the last bd */
> +			bd[j].first_u32 |= BDMA_DESC_EOL;
> +		else
> +			bd[j].first_u32 &= ~BDMA_DESC_EOL;
> +
> +		/* checksume need to clear first */
> +		bd[j].first_u32 &= ~BDMA_DESC_CHECKSUM;
> +		bd[j].first_u32 |= msdc_dma_calcs((u8 *)(&bd[j]), 16) << 8;
> +		sg++;
> +	}
> +
> +	sdr_set_field(host->base + MSDC_DMA_CFG, MSDC_DMA_CFG_DECSEN, 1);
> +	sdr_set_field(host->base + MSDC_DMA_CTRL, MSDC_DMA_CTRL_BRUSTSZ,
> +			MSDC_BURST_64B);
> +	sdr_set_field(host->base + MSDC_DMA_CTRL, MSDC_DMA_CTRL_MODE, 1);

Here, you have the overhead of reading MSDC_DMA_CTRL twice, and writing it
back twice - and the read/write operations you are using are fully ordered,
which means that you're having to wait for your reads and writes to go all
the way to the hardware.  This is inefficient.  I'm sure it could be
improved.

As a general note, readl/writel are barriered operations, which also touch
the L2 cache.  You really ough to be using readl_relaxed()/writel_relaxed()
if you don't need their guarantees.

You _will_ need their guarantees on whichever access starts any data
transfer from memory (maybe the one below?) or reads any status which
is a result of writing data back to memory (eg, reading a DMA status
register.)

> +
> +	writel((u32)dma->gpd_addr, host->base + MSDC_DMA_SA);
> +}
> +
> +static void msdc_prepare_data(struct msdc_host *host, struct mmc_request *mrq)
> +{
> +	struct mmc_data *data = mrq->data;
> +
> +	if (!(data->host_cookie & MSDC_PREPARE_FLAG)) {
> +		bool read = (data->flags & MMC_DATA_READ) != 0;
> +
> +		data->host_cookie |= MSDC_PREPARE_FLAG;
> +		dma_map_sg(host->dev, data->sg, data->sg_len,
> +			   read ? DMA_FROM_DEVICE : DMA_TO_DEVICE);

The return value of dma_map_sg() really needs to be stored, and used when
you subsequently walk the scatterlist.  This is a bug.

The DMA API is permitted to coalesce scatterlist entries if it so chooses,
which results in fewer entries needed to be walked when programming the
DMA hardware.

However, the paired dma_unmap_sg() _must_ get the original data->sg_len,
so you may not overwrite data->sg_len with this.

This is probably a long standing MMC core bug, as struct mmc_data really
needs an additional member to get this right.  Ulf, please fix this.

Lastly, what if dma_map_sg() fails (it's allowed to.)

> +	}
> +}
> +
> +static void msdc_unprepare_data(struct msdc_host *host, struct mmc_request *mrq)
> +{
> +	struct mmc_data *data = mrq->data;
> +
> +	if (data->host_cookie & MSDC_ASYNC_FLAG)
> +		return;
> +
> +	if ((data->host_cookie & MSDC_PREPARE_FLAG)) {
> +		bool read = (data->flags & MMC_DATA_READ) != 0;
> +
> +		dma_unmap_sg(host->dev, data->sg, data->sg_len,
> +			     read ? DMA_FROM_DEVICE : DMA_TO_DEVICE);
> +		data->host_cookie &= ~MSDC_PREPARE_FLAG;
> +	}
> +}
> +
> +/* clock control primitives */
> +static void msdc_set_timeout(struct msdc_host *host, u32 ns, u32 clks)
> +{
> +	u32 timeout, clk_ns;
> +	u32 mode = 0;
> +
> +	host->timeout_ns = ns;
> +	host->timeout_clks = clks;
> +	if (host->sclk == 0) {
> +		timeout = 0;
> +	} else {
> +		clk_ns  = 1000000000UL / host->sclk;
> +		timeout = (ns + clk_ns - 1) / clk_ns + clks;
> +		/* in 1048576 sclk cycle unit */
> +		timeout = (timeout + (0x1 << 20) - 1) >> 20;
> +		sdr_get_field(host->base + MSDC_CFG, MSDC_CFG_CKMOD, &mode);
> +		/*DDR mode will double the clk cycles for data timeout */
> +		timeout = mode >= 2 ? timeout * 2 : timeout;
> +		timeout = timeout > 1 ? timeout - 1 : 0;
> +		timeout = timeout > 255 ? 255 : timeout;
> +	}
> +	sdr_set_field(host->base + SDC_CFG, SDC_CFG_DTOC, timeout);
> +}
> +
> +static void msdc_disable_src_clk(struct msdc_host *host)
> +{
> +	if (host->sclk_enabled) {
> +		clk_disable_unprepare(host->src_clk);
> +		host->sclk_enabled = false;
> +	}
> +}
> +
> +static void msdc_enable_src_clk(struct msdc_host *host)
> +{
> +	if (!host->sclk_enabled) {
> +		clk_prepare_enable(host->src_clk);
> +		host->sclk_enabled = true;
> +	}
> +	while (!(readl(host->base + MSDC_CFG) & MSDC_CFG_CKSTB))
> +		cpu_relax();
> +}
> +
> +static void msdc_gate_clock(struct msdc_host *host)
> +{
> +	msdc_disable_src_clk(host);
> +	if (!IS_ERR(host->h_clk) && host->hclk_enabled) {
> +		clk_disable_unprepare(host->h_clk);
> +		host->hclk_enabled = false;
> +	}
> +}
> +
> +static void msdc_ungate_clock(struct msdc_host *host)
> +{
> +	if (!IS_ERR(host->h_clk) && !host->hclk_enabled) {
> +		clk_prepare_enable(host->h_clk);
> +		host->hclk_enabled = true;
> +	}
> +	msdc_enable_src_clk(host);
> +}
> +
> +static void msdc_set_mclk(struct msdc_host *host, int ddr, u32 hz)
> +{
> +	u32 mode;
> +	u32 flags;
> +	u32 div;
> +	u32 sclk;
> +	u32 hclk = host->hclk;
> +
> +	if (!hz) {
> +		dev_dbg(host->dev, "set mclk to 0\n");
> +		host->mclk = 0;
> +		msdc_disable_src_clk(host);
> +		return;
> +	}
> +
> +	flags = readl(host->base + MSDC_INTEN);
> +	sdr_clr_bits(host->base + MSDC_INTEN, flags);
> +	if (ddr) { /* may need to modify later */
> +		mode = 0x2; /* ddr mode and use divisor */
> +		if (hz >= (hclk >> 2)) {
> +			div = 0; /* mean div = 1/4 */
> +			sclk = hclk >> 2; /* sclk = clk / 4 */
> +		} else {
> +			div = (hclk + ((hz << 2) - 1)) / (hz << 2);
> +			sclk = (hclk >> 2) / div;
> +			div = (div >> 1);
> +		}
> +	} else if (hz >= hclk) {
> +		mode = 0x1; /* no divisor */
> +		div = 0;
> +		sclk = hclk;
> +	} else {
> +		mode = 0x0; /* use divisor */
> +		if (hz >= (hclk >> 1)) {
> +			div = 0; /* mean div = 1/2 */
> +			sclk = hclk >> 1; /* sclk = clk / 2 */
> +		} else {
> +			div = (hclk + ((hz << 2) - 1)) / (hz << 2);
> +			sclk = (hclk >> 2) / div;
> +		}
> +	}
> +	sdr_set_field(host->base + MSDC_CFG, MSDC_CFG_CKMOD | MSDC_CFG_CKDIV,
> +			(mode << 8) | (div % 0xff));
> +	msdc_enable_src_clk(host);
> +	host->sclk = sclk;
> +	host->mclk = hz;
> +	host->ddr = ddr;
> +	/* need because clk changed. */
> +	msdc_set_timeout(host, host->timeout_ns, host->timeout_clks);
> +	sdr_set_bits(host->base + MSDC_INTEN, flags);
> +
> +	dev_dbg(host->dev, "sclk: %d, ddr: %d\n", host->sclk, ddr);
> +}
> +
> +static inline u32 msdc_cmd_find_resp(struct msdc_host *host,
> +		struct mmc_request *mrq, struct mmc_command *cmd)
> +{
> +	u32 resp;
> +
> +	switch (mmc_resp_type(cmd)) {
> +		/* Actually, R1, R5, R6, R7 are the same */
> +	case MMC_RSP_R1:
> +		resp = 0x1;
> +		break;
> +	case MMC_RSP_R1B:
> +		resp = 0x7;
> +		break;
> +	case MMC_RSP_R2:
> +		resp = 0x2;
> +		break;
> +	case MMC_RSP_R3:
> +		resp = 0x3;
> +		break;
> +	case MMC_RSP_NONE:
> +	default:
> +		resp = 0x0;
> +		break;
> +	}
> +
> +	return resp;
> +}
> +
> +static inline u32 msdc_cmd_prepare_raw_cmd(struct msdc_host *host,
> +		struct mmc_request *mrq, struct mmc_command *cmd)
> +{
> +	/* rawcmd :
> +	 * vol_swt << 30 | auto_cmd << 28 | blklen << 16 | go_irq << 15 |
> +	 * stop << 14 | rw << 13 | dtype << 11 | rsptyp << 7 | brk << 6 | opcode
> +	 */
> +	u32 opcode = cmd->opcode;
> +	u32 resp = msdc_cmd_find_resp(host, mrq, cmd);
> +	u32 rawcmd = (opcode & 0x3f) | ((resp & 0x7) << 7);
> +
> +	host->cmd_rsp = resp;
> +
> +	if ((opcode == SD_IO_RW_DIRECT && cmd->flags == (unsigned int) -1) ||
> +			opcode == MMC_STOP_TRANSMISSION)
> +		rawcmd |= (0x1 << 14);
> +	else if (opcode == SD_SWITCH_VOLTAGE)
> +		rawcmd |= (0x1 << 30);
> +	else if ((opcode == SD_APP_SEND_SCR) ||
> +			(opcode == SD_APP_SEND_NUM_WR_BLKS) ||
> +			(opcode == SD_SWITCH &&
> +			(mmc_cmd_type(cmd) == MMC_CMD_ADTC)) ||
> +			(opcode == SD_APP_SD_STATUS &&
> +			(mmc_cmd_type(cmd) == MMC_CMD_ADTC)) ||
> +			(opcode == MMC_SEND_EXT_CSD &&
> +			(mmc_cmd_type(cmd) == MMC_CMD_ADTC)))
> +		rawcmd |= (0x1 << 11);
> +
> +	if (cmd->data) {
> +		struct mmc_data *data = cmd->data;
> +
> +		if (mmc_op_multi(opcode)) {
> +			if (mmc_card_mmc(host->mmc->card) && mrq->sbc &&
> +					!(mrq->sbc->arg & 0xFFFF0000))
> +				rawcmd |= 0x2 << 28; /* AutoCMD23 */
> +		}
> +
> +		rawcmd |= ((data->blksz & 0xFFF) << 16);
> +		if (data->flags & MMC_DATA_WRITE)
> +			rawcmd |= (0x1 << 13);
> +		if (data->blocks > 1)
> +			rawcmd |= (0x2 << 11);
> +		else
> +			rawcmd |= (0x1 << 11);
> +		/* Always use dma mode */
> +		sdr_clr_bits(host->base + MSDC_CFG, MSDC_CFG_PIO);
> +
> +		if (((host->timeout_ns != data->timeout_ns) ||
> +				(host->timeout_clks != data->timeout_clks))) {

Please align the open parens where required:

+		if (((host->timeout_ns != data->timeout_ns) ||
+		     (host->timeout_clks != data->timeout_clks))) {

and please get rid of excessive parens.  The compiler knows how to evaluate
this without all those parens which just make this harder to read:

+		if (host->timeout_ns != data->timeout_ns ||
+		    host->timeout_clks != data->timeout_clks) {

will do.  There's no need for the braces around this either, as it is a
single statement:

> +			msdc_set_timeout(host, data->timeout_ns,
> +					data->timeout_clks);
> +		}
> +
> +		writel(data->blocks, host->base + SDC_BLK_NUM);
> +	}
> +	return rawcmd;
> +}
> +
> +static void msdc_start_data(struct msdc_host *host, struct mmc_request *mrq,
> +			    struct mmc_command *cmd, struct mmc_data *data)
> +{
> +	bool read;
> +	unsigned long flags;
> +
> +	spin_lock_irqsave(&host->lock, flags);
> +	if (!host->data)
> +		host->data = data;
> +	else
> +		WARN_ON(host->data);
> +	spin_unlock_irqrestore(&host->lock, flags);

Doesn't this code go horribly wrong if host->data was not NULL here?  What's
wrong with:

	WARN_ON(host->data);
	host->data = data;

without the spinlock?  What is the spinlock protecting anyway?  As I said
above, if host->data was not already NULL, then you have serious problems
here, so you must somehow guarantee that host->data is already known to
be NULL at this point.  If not, you have a race condition with or without
the spinlock.

> +	read = data->flags & MMC_DATA_READ;
> +
> +	mod_delayed_work(system_wq, &host->req_timeout, DAT_TIMEOUT);
> +	msdc_dma_setup(host, &host->dma, data);
> +	sdr_set_bits(host->base + MSDC_INTEN, data_ints_mask);
> +	mb(); /* wait for pending IO to finish */

What is this memory barrier trying to achieve that writel() hasn't
already done for you?

> +	sdr_set_field(host->base + MSDC_DMA_CTRL, MSDC_DMA_CTRL_START, 1);
> +	dev_dbg(host->dev, "DMA start\n");
> +	dev_dbg(host->dev, "%s: cmd=%d DMA data: %d blocks; read=%d\n",
> +			__func__, cmd->opcode, data->blocks, read);
> +}
> +
> +static int msdc_auto_cmd_done(struct msdc_host *host, int events,
> +		struct mmc_command *cmd)
> +{
> +	u32 *rsp = cmd->resp;
> +
> +	rsp[0] = readl(host->base + SDC_ACMD_RESP);
> +
> +	if (events & MSDC_INT_ACMDRDY) {
> +		cmd->error = 0;
> +	} else {
> +		msdc_reset_hw(host);
> +		if (events & MSDC_INT_ACMDCRCERR) {
> +			cmd->error = -EILSEQ;
> +			host->error |= REQ_STOP_EIO;
> +		} else if (events & MSDC_INT_ACMDTMO) {
> +			cmd->error = -ETIMEDOUT;
> +			host->error |= REQ_STOP_TMO;
> +		}
> +		dev_err(host->dev,
> +			"%s: AUTO_CMD%d arg=%08X; rsp %08X; cmd_error=%d\n",
> +			__func__, cmd->opcode, cmd->arg, rsp[0], cmd->error);
> +	}
> +	return cmd->error;
> +}
> +
> +static void msdc_track_cmd_data(struct msdc_host *host,
> +				struct mmc_command *cmd, struct mmc_data *data)
> +{
> +	if (host->error)
> +		dev_dbg(host->dev, "%s: cmd=%d arg=%08X; host->error=0x%08X\n",
> +			__func__, cmd->opcode, cmd->arg, host->error);
> +}
> +
> +static void msdc_request_done(struct msdc_host *host, struct mmc_request *mrq)
> +{
> +	unsigned long flags;
> +	bool ret;
> +
> +	ret = cancel_delayed_work(&host->req_timeout);
> +	if (!ret) {
> +		/* delay work already running */
> +		return;
> +	}
> +	spin_lock_irqsave(&host->lock, flags);
> +	host->mrq = NULL;
> +	spin_unlock_irqrestore(&host->lock, flags);
> +
> +	msdc_track_cmd_data(host, mrq->cmd, mrq->data);
> +	if (mrq->data)
> +		msdc_unprepare_data(host, mrq);
> +	mmc_request_done(host->mmc, mrq);
> +}
> +
> +/* returns true if command is fully handled; returns false otherwise */
> +static bool msdc_cmd_done(struct msdc_host *host, int events,
> +			  struct mmc_request *mrq, struct mmc_command *cmd)
> +{
> +	bool done = false;
> +	bool sbc_error;
> +	unsigned long flags;
> +	u32 *rsp = cmd->resp;
> +
> +	if (mrq->sbc && cmd == mrq->cmd &&
> +			(events & (MSDC_INT_ACMDRDY | MSDC_INT_ACMDCRCERR
> +				   | MSDC_INT_ACMDTMO)))
> +		msdc_auto_cmd_done(host, events, mrq->sbc);
> +
> +	sbc_error = mrq->sbc && mrq->sbc->error;
> +
> +	if (!sbc_error && !(events & (MSDC_INT_CMDRDY
> +					| MSDC_INT_RSPCRCERR
> +					| MSDC_INT_CMDTMO)))
> +		return done;
> +
> +	spin_lock_irqsave(&host->lock, flags);
> +	done = !host->cmd;
> +	host->cmd = NULL;
> +	spin_unlock_irqrestore(&host->lock, flags);
> +
> +	if (done)
> +		return true;
> +
> +	sdr_clr_bits(host->base + MSDC_INTEN, MSDC_INTEN_CMDRDY |
> +			MSDC_INTEN_RSPCRCERR | MSDC_INTEN_CMDTMO |
> +			MSDC_INTEN_ACMDRDY | MSDC_INTEN_ACMDCRCERR |
> +			MSDC_INTEN_ACMDTMO);
> +	writel(cmd->arg, host->base + SDC_ARG);
> +
> +	switch (host->cmd_rsp) {
> +	case 0:
> +		break;
> +	case 2:
> +		rsp[0] = readl(host->base + SDC_RESP3);
> +		rsp[1] = readl(host->base + SDC_RESP2);
> +		rsp[2] = readl(host->base + SDC_RESP1);
> +		rsp[3] = readl(host->base + SDC_RESP0);
> +		break;
> +	default: /* Response types 1, 3, 4, 5, 6, 7(1b) */
> +		rsp[0] = readl(host->base + SDC_RESP0);
> +		break;
> +	}

Is there something wrong with a simpler implementation of this:

	if (cmd->flags & MMC_RSP_PRESENT) {
		if (cmd->flags & MMC_RSP_136) {
			rsp[0] = readl(host->base + SDC_RESP3);
			rsp[1] = readl(host->base + SDC_RESP2);
			rsp[2] = readl(host->base + SDC_RESP1);
			rsp[3] = readl(host->base + SDC_RESP0);
		} else {
			rsp[0] = readl(host->base + SDC_RESP0);
		}
	}

I originally split the response types up by their properties to allow
such an approach in drivers, so they didn't have to decode lots of
different response types.

> +	if (!sbc_error && !(events & MSDC_INT_CMDRDY)) {
> +		msdc_reset_hw(host);
> +		if (events & MSDC_INT_RSPCRCERR) {
> +			cmd->error = -EILSEQ;
> +			host->error |= REQ_CMD_EIO;
> +		} else if (events & MSDC_INT_CMDTMO) {
> +			cmd->error = -ETIMEDOUT;
> +			host->error |= REQ_CMD_TMO;
> +		}
> +	}
> +	if (cmd->error)
> +		dev_dbg(host->dev,
> +				"%s: cmd=%d arg=%08X; rsp %08X; cmd_error=%d\n",
> +				__func__, cmd->opcode, cmd->arg, rsp[0],
> +				cmd->error);
> +
> +	msdc_cmd_next(host, mrq, cmd);
> +	done = true;
> +	return done;

No need to set a local variable only then to return it.

	return true;

will suffice.

> +}
> +
> +static inline bool msdc_cmd_is_ready(struct msdc_host *host,
> +		struct mmc_request *mrq, struct mmc_command *cmd)
> +{
> +	unsigned long tmo = jiffies + msecs_to_jiffies(20);

Where does this 20ms come from?  Please document it....

> +
> +	while ((readl(host->base + SDC_STS) & SDC_STS_CMDBUSY)
> +			&& time_before(jiffies, tmo))
> +		continue;

Instead of continue,
		cpu_relax();

> +
> +	if (readl(host->base + SDC_STS) & SDC_STS_CMDBUSY) {
> +		dev_err(host->dev, "CMD bus busy detected\n");
> +		host->error |= REQ_CMD_BUSY;
> +		msdc_cmd_done(host, MSDC_INT_CMDTMO, mrq, cmd);
> +		return false;
> +	}
> +
> +	if (mmc_resp_type(cmd) == MMC_RSP_R1B || cmd->data) {
> +		/* R1B or with data, should check SDCBUSY */
> +		while (readl(host->base + SDC_STS) & SDC_STS_SDCBUSY)
> +			cpu_relax();
> +	}
> +	return true;
> +}
> +
> +static void msdc_start_command(struct msdc_host *host,
> +		struct mmc_request *mrq, struct mmc_command *cmd)
> +{
> +	u32 rawcmd;
> +	unsigned long flags;
> +
> +	spin_lock_irqsave(&host->lock, flags);
> +	if (host->cmd)
> +		WARN_ON(host->cmd);
> +	else
> +		host->cmd = cmd;
> +	spin_unlock_irqrestore(&host->lock, flags);

Same comments as for host->data above.

> +
> +	if (!msdc_cmd_is_ready(host, mrq, cmd))
> +		return;
> +
> +	if ((readl(host->base + MSDC_FIFOCS) & MSDC_FIFOCS_TXCNT) >> 16
> +		|| (readl(host->base + MSDC_FIFOCS) & MSDC_FIFOCS_RXCNT)) {

"||" should be on the preceding line, and same comment about lining up the
opening parens.  Anything that makes if() statements easier to read.
In any case, the second half of that doesn't need the () around it.

> +		dev_err(host->dev, "TX/RX FIFO non-empty before start of IO. Reset\n");
> +		msdc_reset_hw(host);
> +	}
> +
> +	cmd->error = 0;
> +	rawcmd = msdc_cmd_prepare_raw_cmd(host, mrq, cmd);
> +	mod_delayed_work(system_wq, &host->req_timeout, DAT_TIMEOUT);
> +
> +	sdr_set_bits(host->base + MSDC_INTEN, MSDC_INTEN_CMDRDY |
> +			MSDC_INTEN_RSPCRCERR | MSDC_INTEN_CMDTMO |
> +			MSDC_INTEN_ACMDRDY | MSDC_INTEN_ACMDCRCERR |
> +			MSDC_INTEN_ACMDTMO);
> +	writel(cmd->arg, host->base + SDC_ARG);
> +	writel(rawcmd, host->base + SDC_CMD);
> +}
> +
> +static void msdc_cmd_next(struct msdc_host *host,
> +		struct mmc_request *mrq, struct mmc_command *cmd)
> +{
> +	if (cmd->error || (mrq->sbc && mrq->sbc->error))
> +		msdc_request_done(host, mrq);
> +	else if (cmd == mrq->sbc)
> +		msdc_start_command(host, mrq, mrq->cmd);
> +	else if (!cmd->data)
> +		msdc_request_done(host, mrq);
> +	else
> +		msdc_start_data(host, mrq, cmd, cmd->data);
> +}
> +
> +static void msdc_ops_request(struct mmc_host *mmc, struct mmc_request *mrq)
> +{
> +	unsigned long flags;
> +	struct msdc_host *host = mmc_priv(mmc);
> +
> +	host->error = 0;
> +
> +	spin_lock_irqsave(&host->lock, flags);
> +	if (!host->mrq)
> +		host->mrq = mrq;
> +	else
> +		WARN_ON(host->mrq);
> +	spin_unlock_irqrestore(&host->lock, flags);

Same comments as for host->data above.

> +
> +	if (mrq->data)
> +		msdc_prepare_data(host, mrq);
> +
> +	/* if SBC is required, we have HW option and SW option.
> +	 * if HW option is enabled, and SBC does not have "special" flags,
> +	 * use HW option,  otherwise use SW option
> +	 */
> +	if (mrq->sbc && (!mmc_card_mmc(mmc->card) ||
> +	    (mrq->sbc->arg & 0xFFFF0000)))
> +		msdc_start_command(host, mrq, mrq->sbc);
> +	else
> +		msdc_start_command(host, mrq, mrq->cmd);
> +}
> +
> +static void msdc_pre_req(struct mmc_host *mmc, struct mmc_request *mrq,
> +		bool is_first_req)
> +{
> +	struct msdc_host *host = mmc_priv(mmc);
> +	struct mmc_data *data = mrq->data;
> +
> +	if (!data)
> +		return;
> +
> +	msdc_prepare_data(host, mrq);
> +	data->host_cookie |= MSDC_ASYNC_FLAG;
> +}
> +
> +static void msdc_post_req(struct mmc_host *mmc, struct mmc_request *mrq,
> +		int err)
> +{
> +	struct msdc_host *host = mmc_priv(mmc);
> +	struct mmc_data *data;
> +
> +	data = mrq->data;
> +	if (!data)
> +		return;
> +	if (data->host_cookie) {
> +		data->host_cookie &= ~MSDC_ASYNC_FLAG;
> +		msdc_unprepare_data(host, mrq);
> +	}
> +}
> +
> +static void msdc_data_xfer_next(struct msdc_host *host,
> +				struct mmc_request *mrq, struct mmc_data *data)
> +{
> +	if (mmc_op_multi(mrq->cmd->opcode) && mrq->stop && !mrq->stop->error &&
> +			(!data->bytes_xfered || !mrq->sbc))

Please:
+	if (mmc_op_multi(mrq->cmd->opcode) && mrq->stop && !mrq->stop->error &&
+	    (!data->bytes_xfered || !mrq->sbc))

> +		msdc_start_command(host, mrq, mrq->stop);
> +	else
> +		msdc_request_done(host, mrq);
> +}
> +
> +static bool msdc_data_xfer_done(struct msdc_host *host, u32 events,
> +				struct mmc_request *mrq, struct mmc_data *data)
> +{
> +	struct mmc_command *stop = data->stop;
> +	unsigned long flags;
> +	bool done;
> +
> +	bool check_data = (events &
> +	    (MSDC_INT_XFER_COMPL | MSDC_INT_DATCRCERR | MSDC_INT_DATTMO
> +	     | MSDC_INT_DMA_BDCSERR | MSDC_INT_DMA_GPDCSERR
> +	     | MSDC_INT_DMA_PROTECT));

This is totally wrong.  You need to turn the bitwise value into a bool
before assigning to a bool variable.  Also, no blank line between this
and the preceding bool

> +
> +	spin_lock_irqsave(&host->lock, flags);
> +	done = !host->data;
> +	if (check_data)
> +		host->data = NULL;
> +	spin_unlock_irqrestore(&host->lock, flags);
> +
> +	if (done)
> +		return true;
> +
> +	if (check_data || (stop && stop->error)) {
> +		dev_dbg(host->dev, "DMA status: 0x%8X\n",
> +				readl(host->base + MSDC_DMA_CFG));
> +		sdr_set_field(host->base + MSDC_DMA_CTRL, MSDC_DMA_CTRL_STOP,
> +				1);
> +		while (readl(host->base + MSDC_DMA_CFG) & MSDC_DMA_CFG_STS)
> +			;

			cpu_relax();

> +		mb(); /* wait for pending IO to finish */

Why is this necessary (remember what I said about readl() above.)

> +		sdr_clr_bits(host->base + MSDC_INTEN, data_ints_mask);
> +		dev_dbg(host->dev, "DMA stop\n");
> +
> +		if ((events & MSDC_INT_XFER_COMPL) && (!stop || !stop->error)) {
> +			data->bytes_xfered = data->blocks * data->blksz;
> +		} else {
> +			dev_err(host->dev, "interrupt events: %x\n", events);
> +			msdc_reset_hw(host);
> +			host->error |= REQ_DAT_ERR;
> +			data->bytes_xfered = 0;
> +
> +			if (events & MSDC_INT_DATTMO)
> +				data->error = -ETIMEDOUT;
> +
> +			dev_err(host->dev, "%s: cmd=%d; blocks=%d",
> +				__func__, mrq->cmd->opcode, data->blocks);
> +			dev_err(host->dev, "data_error=%d xfer_size=%d\n",
> +					(int)data->error, data->bytes_xfered);
> +		}
> +
> +		msdc_data_xfer_next(host, mrq, data);
> +		done = true;
> +	}
> +	return done;
> +}
> +
> +static void msdc_set_buswidth(struct msdc_host *host, u32 width)
> +{
> +	u32 val = readl(host->base + SDC_CFG);
> +
> +	val &= ~SDC_CFG_BUSWIDTH;
> +
> +	switch (width) {
> +	default:
> +	case MMC_BUS_WIDTH_1:
> +		val |= (MSDC_BUS_1BITS << 16);
> +		break;
> +	case MMC_BUS_WIDTH_4:
> +		val |= (MSDC_BUS_4BITS << 16);
> +		break;
> +	case MMC_BUS_WIDTH_8:
> +		val |= (MSDC_BUS_8BITS << 16);
> +		break;
> +	}
> +
> +	writel(val, host->base + SDC_CFG);
> +	dev_dbg(host->dev, "Bus Width = %d", width);
> +}
> +
> +static int msdc_ops_switch_volt(struct mmc_host *mmc, struct mmc_ios *ios)
> +{
> +	struct msdc_host *host = mmc_priv(mmc);
> +	int min_uv, max_uv;
> +	int ret = 0;
> +
> +	if (!IS_ERR(mmc->supply.vqmmc)) {
> +		if (ios->signal_voltage == MMC_SIGNAL_VOLTAGE_330) {
> +			min_uv = 3300000;
> +			max_uv = 3300000;
> +		} else if (ios->signal_voltage == MMC_SIGNAL_VOLTAGE_180) {
> +			min_uv = 1800000;
> +			max_uv = 1800000;
> +		} else {
> +			dev_err(host->dev, "Unsupported signal voltage!\n");
> +			return -EINVAL;
> +		}
> +
> +		ret = regulator_set_voltage(mmc->supply.vqmmc, min_uv, max_uv);
> +		if (ret) {
> +			dev_err(host->dev,
> +					"Regulator set error %d: %d - %d\n",
> +					ret, min_uv, max_uv);
> +		}
> +	}
> +	return ret;
> +}
> +
> +static int msdc_card_busy(struct mmc_host *mmc)
> +{
> +	struct msdc_host *host = mmc_priv(mmc);
> +	u32 status = readl(host->base + MSDC_PS);
> +
> +	/* check if any pin between dat[0:3] is low */
> +	if (((status >> 16) & 0xf) != 0xf)
> +		return 1;
> +
> +	return 0;
> +}
> +
> +static void msdc_request_timeout(struct work_struct *work)
> +{
> +	struct msdc_host *host = container_of(work, struct msdc_host,
> +			req_timeout.work);
> +
> +	/* simulate HW timeout status */
> +	dev_err(host->dev, "%s: aborting cmd/data/mrq\n", __func__);
> +	if (host->mrq) {
> +		dev_err(host->dev, "%s: aborting mrq=%p cmd=%d\n", __func__,
> +				host->mrq, host->mrq->cmd->opcode);
> +		if (host->cmd) {
> +			dev_err(host->dev, "%s: aborting cmd=%d\n",
> +					__func__, host->cmd->opcode);
> +			msdc_cmd_done(host, MSDC_INT_CMDTMO, host->mrq,
> +					host->cmd);
> +		} else if (host->data) {
> +			dev_err(host->dev, "%s: abort data: cmd%d; %d blocks\n",
> +					__func__, host->mrq->cmd->opcode,
> +					host->data->blocks);
> +			msdc_data_xfer_done(host, MSDC_INT_DATTMO, host->mrq,
> +					host->data);
> +		}
> +	}
> +}
> +
> +static irqreturn_t msdc_irq(int irq, void *dev_id)
> +{
> +	struct msdc_host *host = (struct msdc_host *) dev_id;
> +
> +	while (true) {
> +		unsigned long flags;
> +		struct mmc_request *mrq;
> +		struct mmc_command *cmd;
> +		struct mmc_data *data;
> +		u32 events, event_mask;
> +
> +		spin_lock_irqsave(&host->lock, flags);
> +		events = readl(host->base + MSDC_INT);
> +		event_mask = readl(host->base + MSDC_INTEN);
> +		/* clear interrupts */
> +		writel(events & event_mask, host->base + MSDC_INT);
> +
> +		mrq = host->mrq;
> +		cmd = host->cmd;
> +		data = host->data;
> +		spin_unlock_irqrestore(&host->lock, flags);
> +
> +		if (!(events & event_mask))
> +			break;
> +
> +		if (!mrq) {
> +			dev_err(host->dev,
> +				"%s: MRQ=NULL; events=%08X; event_mask=%08X\n",
> +				__func__, events, event_mask);
> +			WARN_ON(1);
> +			break;
> +		}
> +
> +		dev_dbg(host->dev, "%s: events=%08X\n", __func__, events);
> +
> +		if (cmd)
> +			msdc_cmd_done(host, events, mrq, cmd);
> +		else if (data)
> +			msdc_data_xfer_done(host, events, mrq, data);
> +	}
> +
> +	return IRQ_HANDLED;
> +}
> +
> +static void msdc_init_hw(struct msdc_host *host)
> +{
> +	u32 val;
> +	/* Configure to MMC/SD mode, clock free running */
> +	sdr_set_bits(host->base + MSDC_CFG, MSDC_CFG_MODE);
> +	sdr_set_bits(host->base + MSDC_CFG, MSDC_CFG_CKPDN);
> +
> +	/* Reset */
> +	msdc_reset_hw(host);
> +
> +	/* Disable card detection */
> +	sdr_clr_bits(host->base + MSDC_PS, MSDC_PS_CDEN);
> +
> +	/* Disable and clear all interrupts */
> +	writel(0, host->base + MSDC_INTEN);
> +	val = readl(host->base + MSDC_INT);
> +	writel(val, host->base + MSDC_INT);
> +
> +	writel(0, host->base + MSDC_PAD_TUNE);
> +	writel(0, host->base + MSDC_IOCON);
> +	sdr_set_field(host->base + MSDC_IOCON, MSDC_IOCON_DDLSEL, 1);
> +	writel(0x403c004f, host->base + MSDC_PATCH_BIT);
> +	sdr_set_field(host->base + MSDC_PATCH_BIT, MSDC_CKGEN_MSDC_DLY_SEL, 1);
> +	writel(0xffff0089, host->base + MSDC_PATCH_BIT1);
> +	/* Configure to enable SDIO mode.
> +	   it's must otherwise sdio cmd5 failed */
> +	sdr_set_bits(host->base + SDC_CFG, SDC_CFG_SDIO);
> +
> +	/* disable detect SDIO device interrupt function */
> +	sdr_clr_bits(host->base + SDC_CFG, SDC_CFG_SDIOIDE);
> +
> +	/* Configure to default data timeout */
> +	sdr_set_field(host->base + SDC_CFG, SDC_CFG_DTOC, 3);
> +
> +	dev_dbg(host->dev, "init hardware done!");
> +}
> +
> +static void msdc_deinit_hw(struct msdc_host *host)
> +{
> +	u32 val;
> +	/* Disable and clear all interrupts */
> +	writel(0, host->base + MSDC_INTEN);
> +
> +	val = readl(host->base + MSDC_INT);
> +	writel(val, host->base + MSDC_INT);
> +}
> +
> +/* init gpd and bd list in msdc_drv_probe */
> +static void msdc_init_gpd_bd(struct msdc_host *host, struct msdc_dma *dma)
> +{
> +	struct mt_gpdma_desc *gpd = dma->gpd;
> +	struct mt_bdma_desc *bd = dma->bd;
> +	int i;
> +
> +	memset(gpd, 0, sizeof(struct mt_gpdma_desc));
> +	gpd->next = 0; /* only one gpd */

Isn't that initialised by the above memset?

> +
> +	gpd->first_u32 |= GPDMA_DESC_BDP; /* hwo, cs, bd pointer */

And the value of gpd->first_u32 is guaranteed to be zero, so...

> +	gpd->ptr = (u32)dma->bd_addr; /* physical address */
> +
> +	memset(bd, 0, sizeof(struct mt_bdma_desc) * MAX_BD_NUM);
> +
> +	for (i = 0; i < (MAX_BD_NUM - 1); i++)
> +		bd[i].next = (u32)dma->bd_addr + sizeof(*bd) * (i + 1);
> +}
> +
> +static int timing_is_uhs(struct mmc_ios *ios)
> +{
> +	if (ios->signal_voltage != MMC_SIGNAL_VOLTAGE_330)
> +		return 1;
> +
> +	return 0;
> +}
> +
> +static void msdc_ops_set_ios(struct mmc_host *mmc, struct mmc_ios *ios)
> +{
> +	struct msdc_host *host = mmc_priv(mmc);
> +	int ret;
> +	u32 ddr = 0;
> +
> +	if (ios->timing == MMC_TIMING_UHS_DDR50 ||
> +			ios->timing == MMC_TIMING_MMC_DDR52)

Please align thusly:

	if (ios->timing == MMC_TIMING_UHS_DDR50 ||
	    ios->timing == MMC_TIMING_MMC_DDR52)

> +		ddr = 1;
> +
> +	msdc_set_buswidth(host, ios->bus_width);
> +
> +	/* Suspend/Resume will do power off/on */
> +	switch (ios->power_mode) {
> +	case MMC_POWER_UP:
> +		msdc_init_hw(host);
> +		if (!IS_ERR(mmc->supply.vmmc)) {
> +			ret = mmc_regulator_set_ocr(mmc, mmc->supply.vmmc,
> +					ios->vdd);
> +			if (ret) {
> +				dev_err(host->dev, "Failed to set vmmc power!\n");
> +				return;
> +			}
> +		}
> +		break;
> +	case MMC_POWER_ON:
> +		if (!IS_ERR(mmc->supply.vqmmc) && !host->vqmmc_enabled) {
> +			ret = regulator_enable(mmc->supply.vqmmc);
> +			if (ret)
> +				dev_err(host->dev, "Failed to set vqmmc power!\n");
> +			else
> +				host->vqmmc_enabled = true;
> +		}
> +		break;
> +	case MMC_POWER_OFF:
> +		if (!IS_ERR(mmc->supply.vmmc))
> +			mmc_regulator_set_ocr(mmc, mmc->supply.vmmc, 0);
> +
> +		if (!IS_ERR(mmc->supply.vqmmc) && host->vqmmc_enabled) {
> +			regulator_disable(mmc->supply.vqmmc);
> +			host->vqmmc_enabled = false;
> +		}
> +		break;
> +	default:
> +		break;
> +	}
> +
> +	/* Apply different pinctrl settings for different timing */
> +	if (timing_is_uhs(ios))
> +		pinctrl_select_state(host->pinctrl, host->pins_uhs);
> +	else
> +		pinctrl_select_state(host->pinctrl, host->pins_default);
> +
> +	if (host->mclk != ios->clock || host->ddr != ddr)
> +		msdc_set_mclk(host, ddr, ios->clock);
> +}
> +
> +static struct mmc_host_ops mt_msdc_ops = {
> +	.post_req = msdc_post_req,
> +	.pre_req = msdc_pre_req,
> +	.request = msdc_ops_request,
> +	.set_ios = msdc_ops_set_ios,
> +	.start_signal_voltage_switch = msdc_ops_switch_volt,
> +	.card_busy = msdc_card_busy,
> +};
> +
> +static int msdc_drv_probe(struct platform_device *pdev)
> +{
> +	struct mmc_host *mmc;
> +	struct msdc_host *host;
> +	struct resource *res;
> +	int ret;
> +
> +	if (!pdev->dev.of_node) {
> +		dev_err(&pdev->dev, "No DT found\n");
> +		return -EINVAL;
> +	}
> +	/* Allocate MMC host for this device */
> +	mmc = mmc_alloc_host(sizeof(struct msdc_host), &pdev->dev);
> +	if (!mmc)
> +		return -ENOMEM;
> +
> +	host = mmc_priv(mmc);
> +	ret = mmc_of_parse(mmc);
> +	if (ret)
> +		goto host_free;
> +
> +	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
> +	host->base = devm_ioremap_resource(&pdev->dev, res);
> +	if (IS_ERR(host->base)) {
> +		ret = PTR_ERR(host->base);
> +		goto host_free;
> +	}
> +
> +	ret = mmc_regulator_get_supply(mmc);
> +	if (ret == -EPROBE_DEFER)
> +		goto host_free;
> +
> +	host->src_clk = devm_clk_get(&pdev->dev, "source");
> +	if (IS_ERR(host->src_clk)) {
> +		ret = PTR_ERR(host->src_clk);
> +		goto host_free;
> +	}
> +
> +	host->h_clk = devm_clk_get(&pdev->dev, "hclk");
> +	if (IS_ERR(host->h_clk)) {
> +		/* host->h_clk is optional, Only for MSDC0/3 at MT8173 */
> +		dev_dbg(&pdev->dev,
> +				"Invalied hclk from the device tree!\n");

"Invalid"

> +	}
> +
> +	host->irq = platform_get_irq(pdev, 0);
> +	if (host->irq < 0) {
> +		ret = -EINVAL;
> +		goto host_free;
> +	}
> +
> +	host->pinctrl = devm_pinctrl_get(&pdev->dev);
> +	if (IS_ERR(host->pinctrl)) {
> +		ret = PTR_ERR(host->pinctrl);
> +		dev_err(&pdev->dev, "Cannot find pinctrl!\n");
> +		goto host_free;
> +	}
> +
> +	host->pins_default = pinctrl_lookup_state(host->pinctrl, "default");
> +	if (IS_ERR(host->pins_default)) {
> +		ret = PTR_ERR(host->pins_default);
> +		dev_err(&pdev->dev, "Cannot find pinctrl default!\n");
> +		goto host_free;
> +	}
> +
> +	host->pins_uhs = pinctrl_lookup_state(host->pinctrl, "state_uhs");
> +	if (IS_ERR(host->pins_uhs)) {
> +		ret = PTR_ERR(host->pins_uhs);
> +		dev_err(&pdev->dev, "Cannot find pinctrl uhs!\n");
> +		goto host_free;
> +	}
> +
> +	host->dev = &pdev->dev;
> +	host->mmc = mmc;
> +	host->hclk = clk_get_rate(host->src_clk);
> +	/* Set host parameters to mmc */
> +	mmc->ops = &mt_msdc_ops;
> +	mmc->f_min = host->hclk / (4 * 255);
> +
> +	mmc->caps |= MMC_CAP_ERASE | MMC_CAP_CMD23;
> +	/* MMC core transfer sizes tunable parameters */
> +	mmc->max_segs = MAX_BD_NUM;
> +	mmc->max_seg_size = (64 * 1024 - 512);

Isn't this related to BDMA_DESC_BUFLEN ?

> +	mmc->max_blk_size = 2048;
> +	mmc->max_req_size = 512 * 1024;
> +	mmc->max_blk_count = mmc->max_req_size;

I've not checked these, I hope they're correct.

> +
> +	host->timeout_clks = 3 * 1048576;
> +	host->dma.gpd = dma_alloc_coherent(&pdev->dev,
> +				sizeof(struct mt_gpdma_desc),
> +				&host->dma.gpd_addr, GFP_KERNEL);
> +	host->dma.bd = dma_alloc_coherent(&pdev->dev,
> +				MAX_BD_NUM * sizeof(struct mt_bdma_desc),
> +				&host->dma.bd_addr, GFP_KERNEL);
> +	if ((!host->dma.gpd) || (!host->dma.bd)) {

No need for additional parens.

> +		ret = -ENOMEM;
> +		goto release_mem;
> +	}
> +	msdc_init_gpd_bd(host, &host->dma);
> +	INIT_DELAYED_WORK(&host->req_timeout, msdc_request_timeout);
> +	spin_lock_init(&host->lock);
> +
> +	platform_set_drvdata(pdev, mmc);
> +	msdc_ungate_clock(host);
> +
> +	ret = devm_request_irq(&pdev->dev, (unsigned int) host->irq, msdc_irq,

No need for cast to unsigned int.

-- 
FTTC broadband for 0.8mile line: currently at 10.5Mbps down 400kbps up
according to speedtest.net.

Re: [PATCH v4 3/7] mmc: mediatek: Add PM support for MMC driver

[Ulf Hansson]

On 19 May 2015 at 13:04, Sascha Hauer <s.hauer@pengutronix.de> wrote:
> Hi Chaotian,
>
> On Tue, May 19, 2015 at 02:36:47PM +0800, Chaotian Jing wrote:
>> Add PM support for Mediatek MMC driver
>> Save/restore registers when PM
>
> Has anyone asked you to add a new driver in two steps? I find it hard to
> review if a driver is introduced and changed right afterwards.
> Presumably you find it hard to maintain the patches like this also.
> Please ignore if this is indeed requested by someone.

I requested it.

Primarily because I want to make sure we deal with runtime/system PM
support properly. I guess I have seen too many cases where
functionality is broken, unless compiled with CONFIG_PM* set.

Kind regards
Uffe

Re: [PATCH v4 2/7] mmc: mediatek: Add Mediatek MMC driver

[Ulf Hansson]

[...]

>> You are invoking msdc_gate_clock() and msdc_ungate_clock() in a
>> balanced manner, thus hclk_enabled is redundant. Please remove it.
>
> on drv->probe(), already invoke the msdc_ungate_clock(), so, when the
> runtime pm resume invoke the msdc_ungate_clock(), the hclk already
> enabled.


That's why you invoke pm_runtime_set_active() during ->probe() when
deploying PM support in patch3. It's not an issue then.

[...]

>> I assume it's possible to gate the clock by updating a MSDC register
>> instead!? That would be prefereable since then you can leave clock
>> gating/ungating via the clk API, to be dealt from runtime PM. That
>> would also make "sclk_enabled" in the struct msdc_host redundant.
>>
>> Adopting to above, obviously requires MSDC to be able to ungate the
>> clock by also updating a MSDC register. I assume that's possible as
>> well!?
>>
> We can set the bit1 of MSDC_CFG, when this bit is 0, the bus clock was
> gated to 0 if no command or data is transmitted.
> And, from our designer, when we operate the MSDC register, we need make
> sure both HCLK and source are enabled, if source clock was disabled,
> write some MSDC registers will have no effect(eg. send CMD, without
> source clock, not only cannot send CMD, but also cannot get CMD timeout
> interrupt.)

Thanks, that answered my question. As I understand it you should be
able to adopt to my propsual.

[...]

>> > +{
>> > +       unsigned long tmo = jiffies + msecs_to_jiffies(20);
>> > +
>> > +       while ((readl(host->base + SDC_STS) & SDC_STS_CMDBUSY)
>> > +                       && time_before(jiffies, tmo))
>> > +               continue;
>> > +
>> > +       if (readl(host->base + SDC_STS) & SDC_STS_CMDBUSY) {
>> > +               dev_err(host->dev, "CMD bus busy detected\n");
>> > +               host->error |= REQ_CMD_BUSY;
>> > +               msdc_cmd_done(host, MSDC_INT_CMDTMO, mrq, cmd);
>> > +               return false;
>> > +       }
>> > +
>> > +       if (mmc_resp_type(cmd) == MMC_RSP_R1B || cmd->data) {
>> > +               /* R1B or with data, should check SDCBUSY */
>> > +               while (readl(host->base + SDC_STS) & SDC_STS_SDCBUSY)
>> > +                       cpu_relax();
>> > +       }
>>
>> MSDC seems to be handling card busy detection in HW, right?
>>
> Do not have this ability, HW only know if CMD/DAT is low, but do not
> have any interrupt for it,

I see, but doesn't the above polling mean that msdc will not propagate
the response until the card have stopped signal busy? That's what
MMC_CAP_WAIT_WHILE_BUSY shall be used for.

Perhaps you should remove the above polling, and rely on the MMC core
to poll with CMD13 instead?

>> If so, you should enable MMC_CAP_WAIT_WHILE_BUSY and set
>> "max_busy_timeout" to DAT_TIMEOUT to inform the mmc core about it.
>>

[...]

Kind regards
Uffe

Re: [PATCH v4 2/7] mmc: mediatek: Add Mediatek MMC driver

[Ulf Hansson]

[...]

>> >> > +{
>> >> > +       unsigned long tmo = jiffies + msecs_to_jiffies(20);
>> >> > +
>> >> > +       while ((readl(host->base + SDC_STS) & SDC_STS_CMDBUSY)
>> >> > +                       && time_before(jiffies, tmo))
>> >> > +               continue;
>> >> > +
>> >> > +       if (readl(host->base + SDC_STS) & SDC_STS_CMDBUSY) {
>> >> > +               dev_err(host->dev, "CMD bus busy detected\n");
>> >> > +               host->error |= REQ_CMD_BUSY;
>> >> > +               msdc_cmd_done(host, MSDC_INT_CMDTMO, mrq, cmd);
>> >> > +               return false;
>> >> > +       }
>> >> > +
>> >> > +       if (mmc_resp_type(cmd) == MMC_RSP_R1B || cmd->data) {
>> >> > +               /* R1B or with data, should check SDCBUSY */
>> >> > +               while (readl(host->base + SDC_STS) & SDC_STS_SDCBUSY)
>> >> > +                       cpu_relax();
>> >> > +       }
>> >>
>> >> MSDC seems to be handling card busy detection in HW, right?
>> >>
>> > Do not have this ability, HW only know if CMD/DAT is low, but do not
>> > have any interrupt for it,
>>
>> I see, but doesn't the above polling mean that msdc will not propagate
>> the response until the card have stopped signal busy? That's what
>> MMC_CAP_WAIT_WHILE_BUSY shall be used for.
>>
> As you see, we only check the "busy state" BEFORE issue a R1B command or
> with data command, but do not check if AFTER the request was done, that
> would do not match the "MMC_CAP_WAIT_WHILE_BUSY"(eg. CMD5 to sleep)

Okay, fair enough. Still, I don't understand why you want to do that?

> In addition, about CMD5, I find that the suspend/resume flow of EMMC is
> stranger in new kernel version, when suspend, it may issue CMD5 to enter
> sleep mode, then power off MMC, but when resume, it will
> re-initialization, So that why need do the redundant CMD5 in suspend ?

Both CMD0 and CMD5 are valid as wakeup commands.

To be able to use CMD5, the VCCQ regulator must be kept enabled in the
sleep state. That's when it becomes a bit tricky, due to the range of
different host drivers and SoCs for how VCCQ is managed.

To be safe we have chosen to use CMD0, since it works for *all* cases
no matter if VCCQ get gated or not.

Moreover, using CMD5 as the wakeup command would require added
complexity in the code dealing with suspend/resume. I don't think the
effort is worth it, at least until someone has proven that the resume
time is greatly decreased by using CMD5.

>
>> Perhaps you should remove the above polling, and rely on the MMC core
>> to poll with CMD13 instead?
> before any read/write command, core will issue CMD13 to confirm card
> status, here is just only do double confirm to avoid HW issue.

What HW issue and why do you need to double confirm? It seems strange.

Kind regards
Uffe

Re: [PATCH v4 2/7] mmc: mediatek: Add Mediatek MMC driver

[Ulf Hansson]

On 27 May 2015 at 13:34, Chaotian Jing <chaotian.jing@mediatek.com> wrote:
> On Tue, 2015-05-26 at 14:33 +0200, Ulf Hansson wrote:
>> [...]
>>
>> >> >> > +{
>> >> >> > +       unsigned long tmo = jiffies + msecs_to_jiffies(20);
>> >> >> > +
>> >> >> > +       while ((readl(host->base + SDC_STS) & SDC_STS_CMDBUSY)
>> >> >> > +                       && time_before(jiffies, tmo))
>> >> >> > +               continue;
>> >> >> > +
>> >> >> > +       if (readl(host->base + SDC_STS) & SDC_STS_CMDBUSY) {
>> >> >> > +               dev_err(host->dev, "CMD bus busy detected\n");
>> >> >> > +               host->error |= REQ_CMD_BUSY;
>> >> >> > +               msdc_cmd_done(host, MSDC_INT_CMDTMO, mrq, cmd);
>> >> >> > +               return false;
>> >> >> > +       }
>> >> >> > +
>> >> >> > +       if (mmc_resp_type(cmd) == MMC_RSP_R1B || cmd->data) {
>> >> >> > +               /* R1B or with data, should check SDCBUSY */
>> >> >> > +               while (readl(host->base + SDC_STS) & SDC_STS_SDCBUSY)
>> >> >> > +                       cpu_relax();
>> >> >> > +       }
>> >> >>
>> >> >> MSDC seems to be handling card busy detection in HW, right?
>> >> >>
>> >> > Do not have this ability, HW only know if CMD/DAT is low, but do not
>> >> > have any interrupt for it,
>> >>
>> >> I see, but doesn't the above polling mean that msdc will not propagate
>> >> the response until the card have stopped signal busy? That's what
>> >> MMC_CAP_WAIT_WHILE_BUSY shall be used for.
>> >>
>> > As you see, we only check the "busy state" BEFORE issue a R1B command or
>> > with data command, but do not check if AFTER the request was done, that
>> > would do not match the "MMC_CAP_WAIT_WHILE_BUSY"(eg. CMD5 to sleep)
>>
>> Okay, fair enough. Still, I don't understand why you want to do that?
>>
> Below is the description of these 2 bits:
> 1       CMDBSY  CMDBUSY SD Command line busy status
> S/W should always read this bit to make sure the command line is not
> busy before sending the next command.
> If the command is R1B or data read/write command, S/W should check
> SDCBUSY bit too.
> Note: When Auto command 12 is enabled, this bit will be asserted
> immediately after SDC_CMD is written and de-asserted after auto-command
> 12 finishes.
> 1'b0: No transmission is going on CMD line on SD bus
> 1'b1: There exists transmission going on CMD line on SD bus
> 0       SDCBSY  SDCBUSY SD controller busy status
> 1'b0: SD controller is idle
> 1'b1: SD controller is busy
>

Thanks, for elaborating. I get it know. :-)

>> > In addition, about CMD5, I find that the suspend/resume flow of EMMC is
>> > stranger in new kernel version, when suspend, it may issue CMD5 to enter
>> > sleep mode, then power off MMC, but when resume, it will
>> > re-initialization, So that why need do the redundant CMD5 in suspend ?
>>
>> Both CMD0 and CMD5 are valid as wakeup commands.
>>
>> To be able to use CMD5, the VCCQ regulator must be kept enabled in the
>> sleep state. That's when it becomes a bit tricky, due to the range of
>> different host drivers and SoCs for how VCCQ is managed.
>>
>> To be safe we have chosen to use CMD0, since it works for *all* cases
>> no matter if VCCQ get gated or not.
>>
>> Moreover, using CMD5 as the wakeup command would require added
>> complexity in the code dealing with suspend/resume. I don't think the
>> effort is worth it, at least until someone has proven that the resume
>> time is greatly decreased by using CMD5.
>>
>> >
>> >> Perhaps you should remove the above polling, and rely on the MMC core
>> >> to poll with CMD13 instead?
>> > before any read/write command, core will issue CMD13 to confirm card
>> > status, here is just only do double confirm to avoid HW issue.
>>
>> What HW issue and why do you need to double confirm? It seems strange.
> Some fake SD card, even the CMD13 get card status is in transfer state,
> but when we do read/write, the data0 was pull low.

Okay. So in this case your mmc host driver will take care of this. That's fine.

My concern is then how other mmc host drivers will deal with these
kind of cards that violates the spec. It would be nice if you could
add a card quirk for these cards, if you still recall what cards that
were causing these issues.

>
> By the way, do you think need do some spinlock protect in our code ?
> I cannot find a good reason to use spin lock, it seems no need
> it,although there will be many thread(mmcqd, mmcqd/0boot1,) and the irq
> handler.

>From mmc core layer point of view, you shouldn't be needing any spinlocks.

Whether your driver has a potential concurrent access of the same
registers, I don't know. If that's the case you probably want to
protect this from happening.

Kind regards
Uffe

Re: [PATCH v4 2/7] mmc: mediatek: Add Mediatek MMC driver

[Sascha Hauer]

On Thu, Jun 04, 2015 at 10:54:03AM +0800, Chaotian Jing wrote:
> On Tue, 2015-05-19 at 13:15 +0200, Sascha Hauer wrote:
> > On Tue, May 19, 2015 at 02:36:46PM +0800, Chaotian Jing wrote:
> > > Add Mediatek MMC driver code
> > > Support eMMC/SD/SDIO
> > > 
> > > Signed-off-by: Chaotian Jing <chaotian.jing@mediatek.com>
> > > ---
> > >  drivers/mmc/host/Kconfig  |    8 +
> > >  drivers/mmc/host/Makefile |    1 +
> > >  drivers/mmc/host/mtk-sd.c | 1422 +++++++++++++++++++++++++++++++++++++++++++++
> > >  3 files changed, 1431 insertions(+)
> > >  create mode 100644 drivers/mmc/host/mtk-sd.c
> > > 
> > > +
> > > +	host->src_clk = devm_clk_get(&pdev->dev, "source");
> > > +	if (IS_ERR(host->src_clk)) {
> > > +		ret = PTR_ERR(host->src_clk);
> > > +		goto host_free;
> > > +	}
> > > +
> > > +	host->h_clk = devm_clk_get(&pdev->dev, "hclk");
> > > +	if (IS_ERR(host->h_clk)) {
> > > +		/* host->h_clk is optional, Only for MSDC0/3 at MT8173 */
> > > +		dev_dbg(&pdev->dev,
> > > +				"Invalied hclk from the device tree!\n");
> > > +	}
> > 
> > s/Invalied/Invalid/
> > 
> > According to my reference manual the controller always needs a hclk. It
> > seems on some controllers it is just not software controllable. If
> > that's the case you should always provide a hclk to the driver and make
> > this clock mandatory.
> For MT8173, MSDC1, MSDC2's HCLK are in Infra, software cannot control
> it. So just set the HCLK to clk_null ?

Yes, please.

Sascha

-- 
Pengutronix e.K.                           |                             |
Industrial Linux Solutions                 | http://www.pengutronix.de/  |
Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0    |
Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |