Patchwork [3/3] mmc: esdhc i.MX: Support 8bit mode

login
register
mail settings
Submitter Sascha Hauer
Date Sept. 24, 2012, 7:22 a.m.
Message ID <1348471345-30785-4-git-send-email-s.hauer@pengutronix.de>
Download mbox | patch
Permalink /patch/186311/
State New
Headers show

Comments

Sascha Hauer - Sept. 24, 2012, 7:22 a.m.
The i.MX esdhc has a nonstandard bit layout for the SDHCI_HOST_CONTROL
register. To support 8bit bus width on i.MX populate the platform_bus_width
callback. This is tested on an i.MX25, but should according to the datasheets
work on the other i.MX using this hardware aswell. The i.MX6, while having
a SDHCI_SPEC_300 controller, still uses the same nonstandard register layout.

Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
---
 .../devicetree/bindings/mmc/fsl-imx-esdhc.txt      |    1 +
 arch/arm/plat-mxc/include/mach/esdhc.h             |    1 +
 drivers/mmc/host/sdhci-esdhc-imx.c                 |   56 ++++++++++++++++++--
 3 files changed, 55 insertions(+), 3 deletions(-)
Shawn Guo - Sept. 25, 2012, 7:15 a.m.
On Mon, Sep 24, 2012 at 09:22:25AM +0200, Sascha Hauer wrote:
> The i.MX esdhc has a nonstandard bit layout for the SDHCI_HOST_CONTROL
> register. To support 8bit bus width on i.MX populate the platform_bus_width
> callback. This is tested on an i.MX25, but should according to the datasheets
> work on the other i.MX using this hardware aswell. The i.MX6, while having
> a SDHCI_SPEC_300 controller, still uses the same nonstandard register layout.
> 
> Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
> ---
>  .../devicetree/bindings/mmc/fsl-imx-esdhc.txt      |    1 +
>  arch/arm/plat-mxc/include/mach/esdhc.h             |    1 +
>  drivers/mmc/host/sdhci-esdhc-imx.c                 |   56 ++++++++++++++++++--
>  3 files changed, 55 insertions(+), 3 deletions(-)
> 
> diff --git a/Documentation/devicetree/bindings/mmc/fsl-imx-esdhc.txt b/Documentation/devicetree/bindings/mmc/fsl-imx-esdhc.txt
> index 1dd6225..c989994 100644
> --- a/Documentation/devicetree/bindings/mmc/fsl-imx-esdhc.txt
> +++ b/Documentation/devicetree/bindings/mmc/fsl-imx-esdhc.txt
> @@ -12,6 +12,7 @@ Required properties:
>  Optional properties:
>  - fsl,cd-controller : Indicate to use controller internal card detection
>  - fsl,wp-controller : Indicate to use controller internal write protection
> +- bus-width : Maximum supported bus width. Defaults to 4 if omitted
>
This is a common mmc property documented in bindings/mmc/mmc.txt.
Instead of duplicating the documentation, we should try to make our
implementation conform to the common definition of the property.

>  Examples:
>  
> diff --git a/arch/arm/plat-mxc/include/mach/esdhc.h b/arch/arm/plat-mxc/include/mach/esdhc.h
> index aaf9748..b4a0521 100644
> --- a/arch/arm/plat-mxc/include/mach/esdhc.h
> +++ b/arch/arm/plat-mxc/include/mach/esdhc.h
> @@ -39,5 +39,6 @@ struct esdhc_platform_data {
>  	unsigned int cd_gpio;
>  	enum wp_types wp_type;
>  	enum cd_types cd_type;
> +	int max_bus_width;
>  };
>  #endif /* __ASM_ARCH_IMX_ESDHC_H */
> diff --git a/drivers/mmc/host/sdhci-esdhc-imx.c b/drivers/mmc/host/sdhci-esdhc-imx.c
> index 4fe7312..b205fe9 100644
> --- a/drivers/mmc/host/sdhci-esdhc-imx.c
> +++ b/drivers/mmc/host/sdhci-esdhc-imx.c
> @@ -57,6 +57,13 @@
>   */
>  #define ESDHC_FLAG_MULTIBLK_NO_INT	(1 << 1)
>  
> +/*
> + * Freescale interpretation of the SDHCI_HOST_CONTROL register
> + */
#define FSL_SDHCI_CTRL_1BITBUS			(0x0 << 1)
...

> +#define FSL_SDHCI_CTRL_4BITBUS			(0x1 << 1)
> +#define FSL_SDHCI_CTRL_8BITBUS			(0x2 << 1)
> +#define FSL_SDHCI_CTRL_BUSWIDTH_MASK		(0x3 << 1)
> +
>  enum imx_esdhc_type {
>  	IMX25_ESDHC,
>  	IMX35_ESDHC,
> @@ -307,6 +314,7 @@ static void esdhc_writeb_le(struct sdhci_host *host, u8 val, int reg)
>  	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
>  	struct pltfm_imx_data *imx_data = pltfm_host->priv;
>  	u32 new_val;
> +	u32 mask;
>  
>  	switch (reg) {
>  	case SDHCI_POWER_CONTROL:
> @@ -318,7 +326,6 @@ static void esdhc_writeb_le(struct sdhci_host *host, u8 val, int reg)
>  	case SDHCI_HOST_CONTROL:
>  		/* FSL messed up here, so we can just keep those three */
>  		new_val = val & (SDHCI_CTRL_LED | \
> -				SDHCI_CTRL_4BITBUS | \
>  				SDHCI_CTRL_D3CD);
>  		/* ensure the endianess */
>  		new_val |= ESDHC_HOST_CONTROL_LE;
> @@ -328,7 +335,13 @@ static void esdhc_writeb_le(struct sdhci_host *host, u8 val, int reg)
>  			new_val |= (val & SDHCI_CTRL_DMA_MASK) << 5;
>  		}
>  
> -		esdhc_clrset_le(host, 0xffff, new_val, reg);
> +		/*
> +		 * Do not touch buswidth bits here. This is done in
> +		 * esdhc_pltfm_buswidth.
> +		 */
> +		mask = 0xffff & ~FSL_SDHCI_CTRL_BUSWIDTH_MASK;
> +
> +		esdhc_clrset_le(host, mask, new_val, reg);
>  		return;
>  	}
>  	esdhc_clrset_le(host, 0xff, val, reg);
> @@ -379,6 +392,27 @@ static unsigned int esdhc_pltfm_get_ro(struct sdhci_host *host)
>  	return -ENOSYS;
>  }
>  
> +static int esdhc_pltfm_bus_width(struct sdhci_host *host, int width)
> +{
> +	u32 ctrl;
> +
> +	switch (width) {
> +	case MMC_BUS_WIDTH_8:
> +		ctrl = FSL_SDHCI_CTRL_8BITBUS;
> +		break;
> +	case MMC_BUS_WIDTH_4:
> +		ctrl = FSL_SDHCI_CTRL_4BITBUS;
> +		break;
> +	default:
> +		ctrl = 0;

... 
		ctrl = FSL_SDHCI_CTRL_1BITBUS;

Any better?

> +		break;
> +	}
> +
> +	esdhc_clrset_le(host, FSL_SDHCI_CTRL_BUSWIDTH_MASK, ctrl, SDHCI_HOST_CONTROL);
> +
> +	return 0;
> +}
> +
>  static struct sdhci_ops sdhci_esdhc_ops = {
>  	.read_l = esdhc_readl_le,
>  	.read_w = esdhc_readw_le,
> @@ -389,6 +423,7 @@ static struct sdhci_ops sdhci_esdhc_ops = {
>  	.get_max_clock = esdhc_pltfm_get_max_clock,
>  	.get_min_clock = esdhc_pltfm_get_min_clock,
>  	.get_ro = esdhc_pltfm_get_ro,
> +	.platform_bus_width = esdhc_pltfm_bus_width,
>  };
>  
>  static struct sdhci_pltfm_data sdhci_esdhc_imx_pdata = {
> @@ -434,6 +469,8 @@ sdhci_esdhc_imx_probe_dt(struct platform_device *pdev,
>  	if (gpio_is_valid(boarddata->wp_gpio))
>  		boarddata->wp_type = ESDHC_WP_GPIO;
>  
> +	of_property_read_u32(np, "bus-width", &boarddata->max_bus_width);
> +
>  	return 0;
>  }
>  #else
> @@ -507,7 +544,7 @@ static int __devinit sdhci_esdhc_imx_probe(struct platform_device *pdev)
>  	if (is_imx25_esdhc(imx_data) || is_imx35_esdhc(imx_data))
>  		/* Fix errata ENGcm07207 present on i.MX25 and i.MX35 */
>  		host->quirks |= SDHCI_QUIRK_NO_MULTIBLOCK
> -			| SDHCI_QUIRK_BROKEN_ADMA;
> +			 | SDHCI_QUIRK_BROKEN_ADMA;
>  

Why this change?

Shawn

>  	if (is_imx53_esdhc(imx_data))
>  		imx_data->flags |= ESDHC_FLAG_MULTIBLK_NO_INT;
> @@ -577,6 +614,19 @@ static int __devinit sdhci_esdhc_imx_probe(struct platform_device *pdev)
>  		break;
>  	}
>  
> +	switch (boarddata->max_bus_width) {
> +	case 8:
> +		host->mmc->caps |= MMC_CAP_8_BIT_DATA | MMC_CAP_4_BIT_DATA;
> +		break;
> +	case 4:
> +	default:
> +		host->mmc->caps |= MMC_CAP_4_BIT_DATA;
> +		break;
> +	case 1:
> +		host->quirks |= SDHCI_QUIRK_FORCE_1_BIT_DATA;
> +		break;
> +	}
> +
>  	err = sdhci_add_host(host);
>  	if (err)
>  		goto err_add_host;
> -- 
> 1.7.10.4
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-mmc" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
Sascha Hauer - Sept. 25, 2012, 7:27 a.m.
On Tue, Sep 25, 2012 at 03:15:08PM +0800, Shawn Guo wrote:
> On Mon, Sep 24, 2012 at 09:22:25AM +0200, Sascha Hauer wrote:
> > The i.MX esdhc has a nonstandard bit layout for the SDHCI_HOST_CONTROL
> > register. To support 8bit bus width on i.MX populate the platform_bus_width
> > callback. This is tested on an i.MX25, but should according to the datasheets
> > work on the other i.MX using this hardware aswell. The i.MX6, while having
> > a SDHCI_SPEC_300 controller, still uses the same nonstandard register layout.
> > 
> > Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
> > ---
> >  .../devicetree/bindings/mmc/fsl-imx-esdhc.txt      |    1 +
> >  arch/arm/plat-mxc/include/mach/esdhc.h             |    1 +
> >  drivers/mmc/host/sdhci-esdhc-imx.c                 |   56 ++++++++++++++++++--
> >  3 files changed, 55 insertions(+), 3 deletions(-)
> > 
> > diff --git a/Documentation/devicetree/bindings/mmc/fsl-imx-esdhc.txt b/Documentation/devicetree/bindings/mmc/fsl-imx-esdhc.txt
> > index 1dd6225..c989994 100644
> > --- a/Documentation/devicetree/bindings/mmc/fsl-imx-esdhc.txt
> > +++ b/Documentation/devicetree/bindings/mmc/fsl-imx-esdhc.txt
> > @@ -12,6 +12,7 @@ Required properties:
> >  Optional properties:
> >  - fsl,cd-controller : Indicate to use controller internal card detection
> >  - fsl,wp-controller : Indicate to use controller internal write protection
> > +- bus-width : Maximum supported bus width. Defaults to 4 if omitted
> >
> This is a common mmc property documented in bindings/mmc/mmc.txt.
> Instead of duplicating the documentation, we should try to make our
> implementation conform to the common definition of the property.

It is conform to the common definition, so all I have to do is drop the
line duplicating the docs. That's easy ;)

> > +		break;
> > +	case MMC_BUS_WIDTH_4:
> > +		ctrl = FSL_SDHCI_CTRL_4BITBUS;
> > +		break;
> > +	default:
> > +		ctrl = 0;
> 
> ... 
> 		ctrl = FSL_SDHCI_CTRL_1BITBUS;
> 
> Any better?

ok.

> >  	if (is_imx25_esdhc(imx_data) || is_imx35_esdhc(imx_data))
> >  		/* Fix errata ENGcm07207 present on i.MX25 and i.MX35 */
> >  		host->quirks |= SDHCI_QUIRK_NO_MULTIBLOCK
> > -			| SDHCI_QUIRK_BROKEN_ADMA;
> > +			 | SDHCI_QUIRK_BROKEN_ADMA;
> >  
> 
> Why this change?

Accident. Will drop.

Sascha
Shawn Guo - Sept. 25, 2012, 7:33 a.m.
On Tue, Sep 25, 2012 at 09:27:15AM +0200, Sascha Hauer wrote:
> > >  Optional properties:
> > >  - fsl,cd-controller : Indicate to use controller internal card detection
> > >  - fsl,wp-controller : Indicate to use controller internal write protection
> > > +- bus-width : Maximum supported bus width. Defaults to 4 if omitted
> > >
> > This is a common mmc property documented in bindings/mmc/mmc.txt.
> > Instead of duplicating the documentation, we should try to make our
> > implementation conform to the common definition of the property.
> 
> It is conform to the common definition, so all I have to do is drop the
> line duplicating the docs. That's easy ;)
> 
Maybe not.  The common binding defines it as a required property while
the patch implements it as an optional one.

Shawn
Sascha Hauer - Sept. 25, 2012, 7:38 a.m.
On Tue, Sep 25, 2012 at 03:33:19PM +0800, Shawn Guo wrote:
> On Tue, Sep 25, 2012 at 09:27:15AM +0200, Sascha Hauer wrote:
> > > >  Optional properties:
> > > >  - fsl,cd-controller : Indicate to use controller internal card detection
> > > >  - fsl,wp-controller : Indicate to use controller internal write protection
> > > > +- bus-width : Maximum supported bus width. Defaults to 4 if omitted
> > > >
> > > This is a common mmc property documented in bindings/mmc/mmc.txt.
> > > Instead of duplicating the documentation, we should try to make our
> > > implementation conform to the common definition of the property.
> > 
> > It is conform to the common definition, so all I have to do is drop the
> > line duplicating the docs. That's easy ;)
> > 
> Maybe not.  The common binding defines it as a required property while
> the patch implements it as an optional one.

So you want to make the code conform to the common spec which has
bus-width as a required property. Is this really worth it to add an
incompatible change to our devicetrees?

Sascha
Shawn Guo - Sept. 25, 2012, 7:45 a.m.
On Tue, Sep 25, 2012 at 09:38:08AM +0200, Sascha Hauer wrote:
> So you want to make the code conform to the common spec which has
> bus-width as a required property. Is this really worth it to add an
> incompatible change to our devicetrees?
> 
I do want to have our implementation conform to common spec.  At the
current stage devicetrees are not stable anyway, so we should make
them right as much as possible before they start getting maintained
separately from kernel tree, where we will be asked to maintain
devicetree compatibility a lot harder.

Shawn
Chris Ball - Sept. 25, 2012, 7:52 a.m.
Hi,

On Tue, Sep 25 2012, Sascha Hauer wrote:
>> > > > +- bus-width : Maximum supported bus width. Defaults to 4 if omitted
>> > > >
>> > > This is a common mmc property documented in bindings/mmc/mmc.txt.
>> > > Instead of duplicating the documentation, we should try to make our
>> > > implementation conform to the common definition of the property.
>> > 
>> > It is conform to the common definition, so all I have to do is drop the
>> > line duplicating the docs. That's easy ;)
>> > 
>> Maybe not.  The common binding defines it as a required property while
>> the patch implements it as an optional one.
>
> So you want to make the code conform to the common spec which has
> bus-width as a required property. Is this really worth it to add an
> incompatible change to our devicetrees?

We could also break the stalemate by just changing the common spec to
have bus-width become optional, default 4.  It's not going to break the
code of anyone who's been treating it as required.  I don't think there
there was a principled reason behind making it required.

Thanks,

- Chris.
Shawn Guo - Sept. 25, 2012, 7:53 a.m.
On Tue, Sep 25, 2012 at 03:52:09AM -0400, Chris Ball wrote:
> We could also break the stalemate by just changing the common spec to
> have bus-width become optional, default 4.  It's not going to break the
> code of anyone who's been treating it as required.  I don't think there
> there was a principled reason behind making it required.
> 
That would be the best :)

Shawn
Chris Ball - Sept. 25, 2012, 8 a.m.
Hi,

On Tue, Sep 25 2012, Shawn Guo wrote:
> On Tue, Sep 25, 2012 at 03:52:09AM -0400, Chris Ball wrote:
>> We could also break the stalemate by just changing the common spec to
>> have bus-width become optional, default 4.  It's not going to break the
>> code of anyone who's been treating it as required.  I don't think there
>> there was a principled reason behind making it required.
>> 
> That would be the best :)

Ah, but sdhci-s3c and dw_mmc have been assuming that bus-width defaults
1 when it's not supplied.  That's going to make it complicated.

- Chris.
Sascha Hauer - Sept. 25, 2012, 8:05 a.m.
On Tue, Sep 25, 2012 at 03:45:50PM +0800, Shawn Guo wrote:
> On Tue, Sep 25, 2012 at 09:38:08AM +0200, Sascha Hauer wrote:
> > So you want to make the code conform to the common spec which has
> > bus-width as a required property. Is this really worth it to add an
> > incompatible change to our devicetrees?
> > 
> I do want to have our implementation conform to common spec.  At the
> current stage devicetrees are not stable anyway, so we should make
> them right as much as possible before they start getting maintained
> separately from kernel tree, where we will be asked to maintain
> devicetree compatibility a lot harder.

Ok, how about this:

- I add a bus-width = <4> property to all i.MX dtsi files
- we merge the patch adding bus-width as an optional property
- Make the bus-width option mandatory in the next step.

This way we can merge the mmc related patches through the mmc tree and
the dts changes through arm-soc.

Sascha
Shawn Guo - Sept. 25, 2012, 8:50 a.m.
On Tue, Sep 25, 2012 at 10:05:25AM +0200, Sascha Hauer wrote:
> Ok, how about this:
> 
> - I add a bus-width = <4> property to all i.MX dtsi files
> - we merge the patch adding bus-width as an optional property
> - Make the bus-width option mandatory in the next step.
> 
> This way we can merge the mmc related patches through the mmc tree and
> the dts changes through arm-soc.
> 
Sounds good to me.  How the last step should be done will depend
on the common spec.  Per Chris, there are already two drivers
implementing the property as optional.  Not sure if Chris will change
spec to have the property defined as optional (probably default to 1?).

Shawn
Sascha Hauer - Sept. 25, 2012, 9:45 a.m.
On Tue, Sep 25, 2012 at 04:50:52PM +0800, Shawn Guo wrote:
> On Tue, Sep 25, 2012 at 10:05:25AM +0200, Sascha Hauer wrote:
> > Ok, how about this:
> > 
> > - I add a bus-width = <4> property to all i.MX dtsi files
> > - we merge the patch adding bus-width as an optional property
> > - Make the bus-width option mandatory in the next step.
> > 
> > This way we can merge the mmc related patches through the mmc tree and
> > the dts changes through arm-soc.
> > 
> Sounds good to me.  How the last step should be done will depend
> on the common spec.  Per Chris, there are already two drivers
> implementing the property as optional.  Not sure if Chris will change
> spec to have the property defined as optional (probably default to 1?).

It won't hurt if we add the property now for all i.MX. Then Chris is
free to do whatever he wants (at least from the i.MX perspective)

Sascha
Shawn Guo - Dec. 26, 2012, 2:56 a.m.
On Tue, Sep 25, 2012 at 04:00:18AM -0400, Chris Ball wrote:
> Hi,
> 
> On Tue, Sep 25 2012, Shawn Guo wrote:
> > On Tue, Sep 25, 2012 at 03:52:09AM -0400, Chris Ball wrote:
> >> We could also break the stalemate by just changing the common spec to
> >> have bus-width become optional, default 4.  It's not going to break the
> >> code of anyone who's been treating it as required.  I don't think there
> >> there was a principled reason behind making it required.
> >> 
> > That would be the best :)
> 
> Ah, but sdhci-s3c and dw_mmc have been assuming that bus-width defaults
> 1 when it's not supplied.  That's going to make it complicated.
> 
Okay, we will do the same for esdhc driver, and also submit a change to
the common binding to make it clear.

Shawn

Patch

diff --git a/Documentation/devicetree/bindings/mmc/fsl-imx-esdhc.txt b/Documentation/devicetree/bindings/mmc/fsl-imx-esdhc.txt
index 1dd6225..c989994 100644
--- a/Documentation/devicetree/bindings/mmc/fsl-imx-esdhc.txt
+++ b/Documentation/devicetree/bindings/mmc/fsl-imx-esdhc.txt
@@ -12,6 +12,7 @@  Required properties:
 Optional properties:
 - fsl,cd-controller : Indicate to use controller internal card detection
 - fsl,wp-controller : Indicate to use controller internal write protection
+- bus-width : Maximum supported bus width. Defaults to 4 if omitted
 
 Examples:
 
diff --git a/arch/arm/plat-mxc/include/mach/esdhc.h b/arch/arm/plat-mxc/include/mach/esdhc.h
index aaf9748..b4a0521 100644
--- a/arch/arm/plat-mxc/include/mach/esdhc.h
+++ b/arch/arm/plat-mxc/include/mach/esdhc.h
@@ -39,5 +39,6 @@  struct esdhc_platform_data {
 	unsigned int cd_gpio;
 	enum wp_types wp_type;
 	enum cd_types cd_type;
+	int max_bus_width;
 };
 #endif /* __ASM_ARCH_IMX_ESDHC_H */
diff --git a/drivers/mmc/host/sdhci-esdhc-imx.c b/drivers/mmc/host/sdhci-esdhc-imx.c
index 4fe7312..b205fe9 100644
--- a/drivers/mmc/host/sdhci-esdhc-imx.c
+++ b/drivers/mmc/host/sdhci-esdhc-imx.c
@@ -57,6 +57,13 @@ 
  */
 #define ESDHC_FLAG_MULTIBLK_NO_INT	(1 << 1)
 
+/*
+ * Freescale interpretation of the SDHCI_HOST_CONTROL register
+ */
+#define FSL_SDHCI_CTRL_4BITBUS			(0x1 << 1)
+#define FSL_SDHCI_CTRL_8BITBUS			(0x2 << 1)
+#define FSL_SDHCI_CTRL_BUSWIDTH_MASK		(0x3 << 1)
+
 enum imx_esdhc_type {
 	IMX25_ESDHC,
 	IMX35_ESDHC,
@@ -307,6 +314,7 @@  static void esdhc_writeb_le(struct sdhci_host *host, u8 val, int reg)
 	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
 	struct pltfm_imx_data *imx_data = pltfm_host->priv;
 	u32 new_val;
+	u32 mask;
 
 	switch (reg) {
 	case SDHCI_POWER_CONTROL:
@@ -318,7 +326,6 @@  static void esdhc_writeb_le(struct sdhci_host *host, u8 val, int reg)
 	case SDHCI_HOST_CONTROL:
 		/* FSL messed up here, so we can just keep those three */
 		new_val = val & (SDHCI_CTRL_LED | \
-				SDHCI_CTRL_4BITBUS | \
 				SDHCI_CTRL_D3CD);
 		/* ensure the endianess */
 		new_val |= ESDHC_HOST_CONTROL_LE;
@@ -328,7 +335,13 @@  static void esdhc_writeb_le(struct sdhci_host *host, u8 val, int reg)
 			new_val |= (val & SDHCI_CTRL_DMA_MASK) << 5;
 		}
 
-		esdhc_clrset_le(host, 0xffff, new_val, reg);
+		/*
+		 * Do not touch buswidth bits here. This is done in
+		 * esdhc_pltfm_buswidth.
+		 */
+		mask = 0xffff & ~FSL_SDHCI_CTRL_BUSWIDTH_MASK;
+
+		esdhc_clrset_le(host, mask, new_val, reg);
 		return;
 	}
 	esdhc_clrset_le(host, 0xff, val, reg);
@@ -379,6 +392,27 @@  static unsigned int esdhc_pltfm_get_ro(struct sdhci_host *host)
 	return -ENOSYS;
 }
 
+static int esdhc_pltfm_bus_width(struct sdhci_host *host, int width)
+{
+	u32 ctrl;
+
+	switch (width) {
+	case MMC_BUS_WIDTH_8:
+		ctrl = FSL_SDHCI_CTRL_8BITBUS;
+		break;
+	case MMC_BUS_WIDTH_4:
+		ctrl = FSL_SDHCI_CTRL_4BITBUS;
+		break;
+	default:
+		ctrl = 0;
+		break;
+	}
+
+	esdhc_clrset_le(host, FSL_SDHCI_CTRL_BUSWIDTH_MASK, ctrl, SDHCI_HOST_CONTROL);
+
+	return 0;
+}
+
 static struct sdhci_ops sdhci_esdhc_ops = {
 	.read_l = esdhc_readl_le,
 	.read_w = esdhc_readw_le,
@@ -389,6 +423,7 @@  static struct sdhci_ops sdhci_esdhc_ops = {
 	.get_max_clock = esdhc_pltfm_get_max_clock,
 	.get_min_clock = esdhc_pltfm_get_min_clock,
 	.get_ro = esdhc_pltfm_get_ro,
+	.platform_bus_width = esdhc_pltfm_bus_width,
 };
 
 static struct sdhci_pltfm_data sdhci_esdhc_imx_pdata = {
@@ -434,6 +469,8 @@  sdhci_esdhc_imx_probe_dt(struct platform_device *pdev,
 	if (gpio_is_valid(boarddata->wp_gpio))
 		boarddata->wp_type = ESDHC_WP_GPIO;
 
+	of_property_read_u32(np, "bus-width", &boarddata->max_bus_width);
+
 	return 0;
 }
 #else
@@ -507,7 +544,7 @@  static int __devinit sdhci_esdhc_imx_probe(struct platform_device *pdev)
 	if (is_imx25_esdhc(imx_data) || is_imx35_esdhc(imx_data))
 		/* Fix errata ENGcm07207 present on i.MX25 and i.MX35 */
 		host->quirks |= SDHCI_QUIRK_NO_MULTIBLOCK
-			| SDHCI_QUIRK_BROKEN_ADMA;
+			 | SDHCI_QUIRK_BROKEN_ADMA;
 
 	if (is_imx53_esdhc(imx_data))
 		imx_data->flags |= ESDHC_FLAG_MULTIBLK_NO_INT;
@@ -577,6 +614,19 @@  static int __devinit sdhci_esdhc_imx_probe(struct platform_device *pdev)
 		break;
 	}
 
+	switch (boarddata->max_bus_width) {
+	case 8:
+		host->mmc->caps |= MMC_CAP_8_BIT_DATA | MMC_CAP_4_BIT_DATA;
+		break;
+	case 4:
+	default:
+		host->mmc->caps |= MMC_CAP_4_BIT_DATA;
+		break;
+	case 1:
+		host->quirks |= SDHCI_QUIRK_FORCE_1_BIT_DATA;
+		break;
+	}
+
 	err = sdhci_add_host(host);
 	if (err)
 		goto err_add_host;