[PATCH 3/3 v2] musb: AM35x: Workaround for fifo read issue

[PATCH 3/3 v2] musb: AM35x: Workaround for fifo read issue

[Ajay Kumar Gupta]

AM35x supports only 32bit read operations so we need to have
workaround for 8bit and 16bit read operations.

Signed-off-by: Ajay Kumar Gupta <ajay.gupta-l0cyMroinI0@public.gmane.org>
---
Patch created against linus'tree + all musb patches in Greg's queue 
Changes from v1:
	- removed unnecessary parens.
	- Removed 'memcpy' for 32 bit read loops.

 drivers/usb/musb/am3517.c    |   30 ++++++++++++++++++++++++++++++
 drivers/usb/musb/musb_core.c |    2 ++
 2 files changed, 32 insertions(+), 0 deletions(-)

diff --git a/drivers/usb/musb/am3517.c b/drivers/usb/musb/am3517.c
index b74e664..3299c66 100644
--- a/drivers/usb/musb/am3517.c
+++ b/drivers/usb/musb/am3517.c
@@ -515,3 +515,33 @@ void musb_platform_restore_context(struct musb_context_registers
 	phy_on();
 }
 #endif
+
+/* AM35x supports only 32bit read operation */
+void musb_read_fifo(struct musb_hw_ep *hw_ep, u16 len, u8 *dst)
+{
+	void __iomem *fifo = hw_ep->fifo;
+	u32		val;
+	int		i;
+
+	/* Read for 32bit-aligned destination address */
+	if (likely((0x03 & (unsigned long) dst) == 0) && len >= 4) {
+		readsl(fifo, dst, len >> 2);
+		dst += len & ~0x03;
+		len &= 0x03;
+	}
+	/*
+	 * Now read the rest 1 to 3 bytes or complete length if
+	 * unaligned address.
+	 */
+	if (len > 4) {
+		for (i = 0; i < (len >> 2); i++) {
+			*(u32 *) dst = musb_readl(fifo, 0);
+			dst += 4;
+		}
+		len %= 4;
+	}
+	if (len > 0) {
+		val = musb_readl(fifo, 0);
+		memcpy(dst, &val, len);
+	}
+}
diff --git a/drivers/usb/musb/musb_core.c b/drivers/usb/musb/musb_core.c
index 4093f6d..9c59a8e 100644
--- a/drivers/usb/musb/musb_core.c
+++ b/drivers/usb/musb/musb_core.c
@@ -262,6 +262,7 @@ void musb_write_fifo(struct musb_hw_ep *hw_ep, u16 len, const u8 *src)
 	}
 }
 
+#if !defined(CONFIG_MACH_OMAP3517EVM)
 /*
  * Unload an endpoint's FIFO
  */
@@ -299,6 +300,7 @@ void musb_read_fifo(struct musb_hw_ep *hw_ep, u16 len, u8 *dst)
 		readsb(fifo, dst, len);
 	}
 }
+#endif
 
 #endif	/* normal PIO */
 
-- 
1.6.2.4

--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH 3/3 v2] musb: AM35x: Workaround for fifo read issue

[Sergei Shtylyov]

Hello.

Ajay Kumar Gupta wrote:

> AM35x supports only 32bit read operations so we need to have
> workaround for 8bit and 16bit read operations.
>   

   But don't we need to override musb_readb() and musb_readw() then?

> Signed-off-by: Ajay Kumar Gupta <ajay.gupta-l0cyMroinI0@public.gmane.org>

WBR, Sergei

--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

RE: [PATCH 3/3 v2] musb: AM35x: Workaround for fifo read issue

[Gupta, Ajay Kumar]

Hi
> AM35x supports only 32bit read operations so we need to have
> workaround for 8bit and 16bit read operations.

>   But don't we need to override musb_readb() and musb_readw() then?

Yes, Correct. I do have another patch on that which I will cleanup and submit later.
Anyways with currebt three patch set, basic Host and device operation has been tested.

-Ajay

> Signed-off-by: Ajay Kumar Gupta <ajay.gupta@ti.com>
>WBR, Sergei

RE: [PATCH 3/3 v2] musb: AM35x: Workaround for fifo read issue

[Gupta, Ajay Kumar]

Hi,
> AM35x supports only 32bit read operations so we need to have
> workaround for 8bit and 16bit read operations.
> 
> Signed-off-by: Ajay Kumar Gupta <ajay.gupta@ti.com>
> ---
> Patch created against linus'tree + all musb patches in Greg's queue
> Changes from v1:
> 	- removed unnecessary parens.
> 	- Removed 'memcpy' for 32 bit read loops.
> 
>  drivers/usb/musb/am3517.c    |   30 ++++++++++++++++++++++++++++++
>  drivers/usb/musb/musb_core.c |    2 ++
>  2 files changed, 32 insertions(+), 0 deletions(-)
> 
> diff --git a/drivers/usb/musb/am3517.c b/drivers/usb/musb/am3517.c
> index b74e664..3299c66 100644
> --- a/drivers/usb/musb/am3517.c
> +++ b/drivers/usb/musb/am3517.c
[...]
> +	}
> +}
> diff --git a/drivers/usb/musb/musb_core.c b/drivers/usb/musb/musb_core.c
> index 4093f6d..9c59a8e 100644
> --- a/drivers/usb/musb/musb_core.c
> +++ b/drivers/usb/musb/musb_core.c
> @@ -262,6 +262,7 @@ void musb_write_fifo(struct musb_hw_ep *hw_ep, u16
> len, const u8 *src)
>  	}
>  }
> 
> +#if !defined(CONFIG_MACH_OMAP3517EVM)

Fixes/workaround based on CONFIG_MACH_OMAP3517EVM will be good only for
OMAP3517EVM and would not scale well to other boards based on AM35x.
(As commented earlier by Sergei)

I just got to know of another board "LIZARD" based on AM35x so we really
need to find a solution for this. 

This problem is due to the fact that AM35x is based on OMAP35x but musb ip
Is updated to RTL1.8 using CPPI4.1 DMA engine thus we need to have a
config option to differentiate musb ips between actual OMAP3 and AM35x
platforms.

I am thinking of adding new config option OMAP_MUSB_RTL18 which should be
selected by all the boards based on AM35x in arch/arm/mach-omap2/Kconfig.

The same config option can be used for all the workaround/fixes specific
to AM35x musb platform.
 
------------------
diff --git a/arch/arm/mach-omap2/Kconfig b/arch/arm/mach-omap2/Kconfig
index b72ae06..3ab1156 100644
--- a/arch/arm/mach-omap2/Kconfig
+++ b/arch/arm/mach-omap2/Kconfig
@@ -95,6 +95,7 @@ config MACH_OMAP3517EVM
        bool "OMAP3517/ AM3517 EVM board"
        depends on ARCH_OMAP3 && ARCH_OMAP34XX
        select OMAP_PACKAGE_CBB
+       select OMAP_MUSB_RTL18

 config PMIC_TPS65023
        bool "TPS65023 Power Module"
------------------

Does anyone has a better option to fix this issue or any comment on this
approach?


Regards,
Ajay
>  /*
>   * Unload an endpoint's FIFO
>   */
> @@ -299,6 +300,7 @@ void musb_read_fifo(struct musb_hw_ep *hw_ep, u16 len,
> u8 *dst)
>  		readsb(fifo, dst, len);
>  	}
>  }
> +#endif
> 
>  #endif	/* normal PIO */
> 
> --
> 1.6.2.4

Re: [PATCH 3/3 v2] musb: AM35x: Workaround for fifo read issue

[Sergei Shtylyov]

Hello.

Gupta, Ajay Kumar wrote:

>> AM35x supports only 32bit read operations so we need to have
>> workaround for 8bit and 16bit read operations.

>> Signed-off-by: Ajay Kumar Gupta <ajay.gupta-l0cyMroinI0@public.gmane.org>
>> ---
>> Patch created against linus'tree + all musb patches in Greg's queue
>> Changes from v1:
>> 	- removed unnecessary parens.
>> 	- Removed 'memcpy' for 32 bit read loops.

>>  drivers/usb/musb/am3517.c    |   30 ++++++++++++++++++++++++++++++
>>  drivers/usb/musb/musb_core.c |    2 ++
>>  2 files changed, 32 insertions(+), 0 deletions(-)

>> diff --git a/drivers/usb/musb/am3517.c b/drivers/usb/musb/am3517.c
>> index b74e664..3299c66 100644
>> --- a/drivers/usb/musb/am3517.c
>> +++ b/drivers/usb/musb/am3517.c
> [...]
>> +	}
>> +}
>> diff --git a/drivers/usb/musb/musb_core.c b/drivers/usb/musb/musb_core.c
>> index 4093f6d..9c59a8e 100644
>> --- a/drivers/usb/musb/musb_core.c
>> +++ b/drivers/usb/musb/musb_core.c
>> @@ -262,6 +262,7 @@ void musb_write_fifo(struct musb_hw_ep *hw_ep, u16
>> len, const u8 *src)
>>  	}
>>  }
>>
>> +#if !defined(CONFIG_MACH_OMAP3517EVM)

> Fixes/workaround based on CONFIG_MACH_OMAP3517EVM will be good only for
> OMAP3517EVM and would not scale well to other boards based on AM35x.
> (As commented earlier by Sergei)

> I just got to know of another board "LIZARD" based on AM35x so we really
> need to find a solution for this. 

> This problem is due to the fact that AM35x is based on OMAP35x but musb ip
> Is updated to RTL1.8 using CPPI4.1 DMA engine

    Is that really the only difference?

> thus we need to have a
> config option to differentiate musb ips between actual OMAP3 and AM35x
> platforms.

> I am thinking of adding new config option OMAP_MUSB_RTL18 which should be
> selected by all the boards based on AM35x in arch/arm/mach-omap2/Kconfig.

> The same config option can be used for all the workaround/fixes specific
> to AM35x musb platform.

> ------------------
> diff --git a/arch/arm/mach-omap2/Kconfig b/arch/arm/mach-omap2/Kconfig
> index b72ae06..3ab1156 100644
> --- a/arch/arm/mach-omap2/Kconfig
> +++ b/arch/arm/mach-omap2/Kconfig
> @@ -95,6 +95,7 @@ config MACH_OMAP3517EVM
>         bool "OMAP3517/ AM3517 EVM board"
>         depends on ARCH_OMAP3 && ARCH_OMAP34XX
>         select OMAP_PACKAGE_CBB
> +       select OMAP_MUSB_RTL18
> 
>  config PMIC_TPS65023
>         bool "TPS65023 Power Module"
> ------------------
> 
> Does anyone has a better option to fix this issue or any comment on this
> approach?

    Why not introduce CONFIG_ARCH_AM35x instead?

WBR, Sergei

--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

RE: [PATCH 3/3 v2] musb: AM35x: Workaround for fifo read issue

[Gupta, Ajay Kumar]

Hi,
> Gupta, Ajay Kumar wrote:
> 
> >> AM35x supports only 32bit read operations so we need to have
> >> workaround for 8bit and 16bit read operations.
> 
> >> Signed-off-by: Ajay Kumar Gupta <ajay.gupta@ti.com>
[...]
> >>
> >> +#if !defined(CONFIG_MACH_OMAP3517EVM)
> 
> > Fixes/workaround based on CONFIG_MACH_OMAP3517EVM will be good only for
> > OMAP3517EVM and would not scale well to other boards based on AM35x.
> > (As commented earlier by Sergei)
> 
> > I just got to know of another board "LIZARD" based on AM35x so we really
> > need to find a solution for this.
> 
> > This problem is due to the fact that AM35x is based on OMAP35x but musb
> ip
> > Is updated to RTL1.8 using CPPI4.1 DMA engine
> 
>     Is that really the only difference?

Other difference: USB PHY is built inside the ip itself.

Additionally, there is an errata which restricts byte/word (8/16 bit)
data read which is due to the bus interface on AM35x.
 
> 
> > thus we need to have a
> > config option to differentiate musb ips between actual OMAP3 and AM35x
> > platforms.
> 
> > I am thinking of adding new config option OMAP_MUSB_RTL18 which should
> be
> > selected by all the boards based on AM35x in arch/arm/mach-
> omap2/Kconfig.
> 
> > The same config option can be used for all the workaround/fixes specific
> > to AM35x musb platform.
> 
> > ------------------
> > diff --git a/arch/arm/mach-omap2/Kconfig b/arch/arm/mach-omap2/Kconfig
> > index b72ae06..3ab1156 100644
> > --- a/arch/arm/mach-omap2/Kconfig
> > +++ b/arch/arm/mach-omap2/Kconfig
> > @@ -95,6 +95,7 @@ config MACH_OMAP3517EVM
> >         bool "OMAP3517/ AM3517 EVM board"
> >         depends on ARCH_OMAP3 && ARCH_OMAP34XX
> >         select OMAP_PACKAGE_CBB
> > +       select OMAP_MUSB_RTL18
> >
> >  config PMIC_TPS65023
> >         bool "TPS65023 Power Module"
> > ------------------
> >
> > Does anyone has a better option to fix this issue or any comment on this
> > approach?
> 
>     Why not introduce CONFIG_ARCH_AM35x instead?

This is fine with me but don't know if Tony would agree with it.

-Ajay
> 
> WBR, Sergei

RE: [PATCH 3/3 v2] musb: AM35x: Workaround for fifo read issue

[Gadiyar, Anand]

 
Gupta, Ajay Kumar wrote:
> Fixes/workaround based on CONFIG_MACH_OMAP3517EVM will be good only for
> OMAP3517EVM and would not scale well to other boards based on AM35x.
> (As commented earlier by Sergei)
> 
> I just got to know of another board "LIZARD" based on AM35x so we really
> need to find a solution for this. 
> 
> This problem is due to the fact that AM35x is based on OMAP35x but musb ip
> Is updated to RTL1.8 using CPPI4.1 DMA engine thus we need to have a
> config option to differentiate musb ips between actual OMAP3 and AM35x
> platforms.
> 
> I am thinking of adding new config option OMAP_MUSB_RTL18 which should be
> selected by all the boards based on AM35x in arch/arm/mach-omap2/Kconfig.
> 
> The same config option can be used for all the workaround/fixes specific
> to AM35x musb platform.
>  
> ------------------
> diff --git a/arch/arm/mach-omap2/Kconfig b/arch/arm/mach-omap2/Kconfig
> index b72ae06..3ab1156 100644
> --- a/arch/arm/mach-omap2/Kconfig
> +++ b/arch/arm/mach-omap2/Kconfig
> @@ -95,6 +95,7 @@ config MACH_OMAP3517EVM
>         bool "OMAP3517/ AM3517 EVM board"
>         depends on ARCH_OMAP3 && ARCH_OMAP34XX
>         select OMAP_PACKAGE_CBB
> +       select OMAP_MUSB_RTL18
> 
>  config PMIC_TPS65023
>         bool "TPS65023 Power Module"
> ------------------
> 
> Does anyone has a better option to fix this issue or any 
> comment on this approach?
> 

It would not scale very well - we already have multi-omap builds
that select support for multiple boards. If the AM35x boards are
part of such builds, then mutually exclusive config options won't
work - already n8x0 MUSB is exclusive with 243x/omap3.

If it is possible to detect the AM35x at runtime, then we could
handle this well. Also, a similar set of changes will be needed for
the DMA code as well (right now we can pick only one of the DMA
engines at a time).

- Anand

RE: [PATCH 3/3 v2] musb: AM35x: Workaround for fifo read issue

[Gupta, Ajay Kumar]

Hi,
> > Fixes/workaround based on CONFIG_MACH_OMAP3517EVM will be good only for
> > OMAP3517EVM and would not scale well to other boards based on AM35x.
> > (As commented earlier by Sergei)
> >
> > I just got to know of another board "LIZARD" based on AM35x so we really
> > need to find a solution for this.
> >
> > This problem is due to the fact that AM35x is based on OMAP35x but musb
> ip
> > Is updated to RTL1.8 using CPPI4.1 DMA engine thus we need to have a
> > config option to differentiate musb ips between actual OMAP3 and AM35x
> > platforms.
> >
> > I am thinking of adding new config option OMAP_MUSB_RTL18 which should
> be
> > selected by all the boards based on AM35x in arch/arm/mach-
> omap2/Kconfig.
> >
> > The same config option can be used for all the workaround/fixes specific
> > to AM35x musb platform.
> >
> > ------------------
> > diff --git a/arch/arm/mach-omap2/Kconfig b/arch/arm/mach-omap2/Kconfig
> > index b72ae06..3ab1156 100644
> > --- a/arch/arm/mach-omap2/Kconfig
> > +++ b/arch/arm/mach-omap2/Kconfig
> > @@ -95,6 +95,7 @@ config MACH_OMAP3517EVM
> >         bool "OMAP3517/ AM3517 EVM board"
> >         depends on ARCH_OMAP3 && ARCH_OMAP34XX
> >         select OMAP_PACKAGE_CBB
> > +       select OMAP_MUSB_RTL18
> >
> >  config PMIC_TPS65023
> >         bool "TPS65023 Power Module"
> > ------------------
> >
> > Does anyone has a better option to fix this issue or any
> > comment on this approach?
> >
> 
> It would not scale very well - we already have multi-omap builds
> that select support for multiple boards. If the AM35x boards are
> part of such builds, then mutually exclusive config options won't
> work - already n8x0 MUSB is exclusive with 243x/omap3.

AM35x musb needs different initialization sequence which involves
PHY configuration etc. being done in drivers/usb/musb/am3517.c.
am3517.c also need to handle different ISR routine and all CPPi4.1
DMA related programmings.

> 
> If it is possible to detect the AM35x at runtime, then we could
> handle this well.
> Also, a similar set of changes will be needed for
> the DMA code as well (right now we can pick only one of the DMA
> engines at a time).

We do have such mechanism cpu_is_omap3517() but how about compilation
flags for am3517.c? Do you intend to have different initialization
sequence, ISR function and CPPI4.1 DMA programmings within
drivers/usb/musb/omap2430.c ?

-Ajay
> 
> - Anand

Re: [PATCH 3/3 v2] musb: AM35x: Workaround for fifo read issue

[Felipe Balbi]

On Thu, May 20, 2010 at 11:41:14AM +0200, ext Gadiyar, Anand wrote:
>It would not scale very well - we already have multi-omap builds
>that select support for multiple boards. If the AM35x boards are
>part of such builds, then mutually exclusive config options won't
>work - already n8x0 MUSB is exclusive with 243x/omap3.
>
>If it is possible to detect the AM35x at runtime, then we could
>handle this well. Also, a similar set of changes will be needed for
>the DMA code as well (right now we can pick only one of the DMA
>engines at a time).

it's time to split out platform code from musb core. I was thinking of 
having omap2430.c blackfin.c tusb6010.c davinci.c be a platform_device 
that instantiates musb_hdrc platform_device. It would also ioremap() the 
area and pass the gotten/enabled clock to musb. Then we can have all of 
the platforms enabled since the board files would pass down the 
platform_device for the platform code. Something like:

arch/arm/mach-omap2/usb-musb.c:

static struct platform_device omap_musb = {
	.dev	= {
		.name	= "omap-hsusbotg",	/* using the IP block name */
	},
	[..]
};

[..]

	platform_device_register(&omap_musb);

drivers/usb/musb/omap2430.c:

static struct musb_platform_data omap_musb_data;

static int __init omap_musb_probe(struct platform_device *pdev)
{
	clk = clk_get(&pdev->dev, "ick");
	base = ioremap(res->start, resource_size(res));

	omap_musb_data.clk = clk;
	omap_musb_data.base = base;

	[... all other necessary fields, like mode, etc]

	musb_pdev = platform_device_alloc("musb_hdrc", -1);
	musb_pdev->dev.parent = &pdev->dev;
	platform_device_add_data(musb_pdev, &omap_musb_data);
	platform_device_add(musb_pdev);

	return 0;
}

static int __exit omap_musb_remove(struct platform_device *pdev)
{
	clk_put(clk);
	iounmap(base);

	platform_device_del(musb_pdev);

	return 0;
}

[...]

io functions could also be passed through this platform_device so that 
we remove the insane amounts of ifdefs. Also context save/restore could 
be done a per-platform basis. Since the platform-code would be a parent 
to musb_hdrc, platform-code's suspend would be called before musb_hdrc 
suspend, then we would save context at that point already.

Any other ifdefferry on anything would be easier to remove with this 
approach and it also means we can have one musb_hdrc.ko for all (well 
arm-based only) boards and be sure it would work. We would just need to 
put something similar down for dma engines.

-- 
balbi

DefectiveByDesign.org
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH 3/3 v2] musb: AM35x: Workaround for fifo read issue

[Felipe Balbi]

On Thu, May 20, 2010 at 12:06:30PM +0200, Balbi Felipe (Nokia-D/Helsinki) wrote:
>arm-based only) boards and be sure it would work. We would just need to
>put something similar down for dma engines.

and of course, the same platform-code could instantiate a dma engine 
platform_device but that's trickier, I guess, since the base address 
would have to be used by two drivers, then platform code would have to 
handle synchronization.

-- 
balbi

DefectiveByDesign.org
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

RE: [PATCH 3/3 v2] musb: AM35x: Workaround for fifo read issue

[Gupta, Ajay Kumar]

Hi,
> >It would not scale very well - we already have multi-omap builds
> >that select support for multiple boards. If the AM35x boards are
> >part of such builds, then mutually exclusive config options won't
> >work - already n8x0 MUSB is exclusive with 243x/omap3.
> >
> >If it is possible to detect the AM35x at runtime, then we could
> >handle this well. Also, a similar set of changes will be needed for
> >the DMA code as well (right now we can pick only one of the DMA
> >engines at a time).
> 
> it's time to split out platform code from musb core. I was thinking of
> having omap2430.c blackfin.c tusb6010.c davinci.c be a platform_device
> that instantiates musb_hdrc platform_device. It would also ioremap() the
> area and pass the gotten/enabled clock to musb. Then we can have all of
> the platforms enabled since the board files would pass down the
> platform_device for the platform code. Something like:

This approach would anyway not help the original issue discussed
in this thread. We need to have some config option to differentiate
musb ips between OMAP3 and AM35x.

Another issue: Currently almost 90% of the code is same between
drivers/usb/musb/da8xx.c and drivers/usb/musb/am3517.c. any idea
on how to avoid duplication?  

Can we merge these two files and have compilation flags based on musb
capability alone (like HAS_CPPI41) rather than CONFIG_ARCH_xx? This
approach would also help the original issue of discussed in this thread.

-Ajay
> 
> arch/arm/mach-omap2/usb-musb.c:
> 
[..]

[PATCH 1/3 v2] AM35x: Add musb support

[Ajay Kumar Gupta]

AM35x has musb interface (version 1.8) and uses CPPI41 DMA engine.
It supports upto 500mA of power in host mode.

Signed-off-by: Ajay Kumar Gupta <ajay.gupta@ti.com>
---
Changes from v1:
	- Added PHY clock and OTGMODE settings in board files
 arch/arm/mach-omap2/board-am3517evm.c |   42 +++++++++++++++++++++++++++++++++
 arch/arm/mach-omap2/usb-musb.c        |    4 +++
 arch/arm/plat-omap/include/plat/usb.h |   21 ++++++++++++++++
 3 files changed, 67 insertions(+), 0 deletions(-)

diff --git a/arch/arm/mach-omap2/board-am3517evm.c b/arch/arm/mach-omap2/board-am3517evm.c
index 837f294..c66ac72 100644
--- a/arch/arm/mach-omap2/board-am3517evm.c
+++ b/arch/arm/mach-omap2/board-am3517evm.c
@@ -32,6 +32,7 @@
 #include <plat/common.h>
 #include <plat/usb.h>
 #include <plat/display.h>
+#include <plat/control.h>
 
 #include "mux.h"
 
@@ -278,6 +279,12 @@ static void __init am3517_evm_init_irq(void)
 	omap_gpio_init();
 }
 
+static struct omap_musb_board_data musb_board_data = {
+	.interface_type         = MUSB_INTERFACE_ULPI,
+	.mode                   = MUSB_OTG,
+	.power                  = 500,
+};
+
 static const struct ehci_hcd_omap_platform_data ehci_pdata __initconst = {
 	.port_mode[0] = EHCI_HCD_OMAP_MODE_PHY,
 #if defined(CONFIG_PANEL_SHARP_LQ043T1DG01) || \
@@ -296,12 +303,45 @@ static const struct ehci_hcd_omap_platform_data ehci_pdata __initconst = {
 
 #ifdef CONFIG_OMAP_MUX
 static struct omap_board_mux board_mux[] __initdata = {
+	/* USB OTG DRVVBUS offset = 0x212 */
+	OMAP3_MUX(SAD2D_MCAD23, OMAP_MUX_MODE0 | OMAP_PIN_INPUT_PULLDOWN),
 	{ .reg_offset = OMAP_MUX_TERMINATOR },
 };
 #else
 #define board_mux	NULL
 #endif
 
+static __init void am3517_evm_musb_init(void)
+{
+	u32 devconf2;
+
+	/*
+	 * Set up USB clock/mode in the DEVCONF2 register.
+	 */
+	devconf2 = omap_ctrl_readl(AM35XX_CONTROL_DEVCONF2);
+
+	/* USB2.0 PHY reference clock is 13 MHz */
+	devconf2 &= ~CONF2_REFFREQ;
+	devconf2 |=  CONF2_REFFREQ_13MHZ;
+
+	/*
+	 * We have to override VBUS/ID signals when MUSB is configured into the
+	 * host-only mode -- ID pin will float if no cable is connected, so the
+	 * controller won't be able to drive VBUS thinking that it's a B-device.
+	 * Otherwise, we want to use the OTG mode and enable VBUS comparators.
+	 */
+	devconf2 &= ~CONF2_OTGMODE;
+#ifdef	CONFIG_USB_MUSB_HOST
+	devconf2 |=  CONF2_FORCE_HOST;
+#else
+	devconf2 |=  CONF2_SESENDEN | CONF2_VBDTCTEN;
+#endif
+
+	omap_ctrl_writel(devconf2, AM35XX_CONTROL_DEVCONF2);
+
+	usb_musb_init(&musb_board_data);
+}
+
 static void __init am3517_evm_init(void)
 {
 	omap3_mux_init(board_mux, OMAP_PACKAGE_CBB);
@@ -323,6 +363,8 @@ static void __init am3517_evm_init(void)
 
 	i2c_register_board_info(1, am3517evm_i2c_boardinfo,
 				ARRAY_SIZE(am3517evm_i2c_boardinfo));
+	/* MUSB */
+	am3517_evm_musb_init();
 }
 
 static void __init am3517_evm_map_io(void)
diff --git a/arch/arm/mach-omap2/usb-musb.c b/arch/arm/mach-omap2/usb-musb.c
index 6d41fa7..710020d 100644
--- a/arch/arm/mach-omap2/usb-musb.c
+++ b/arch/arm/mach-omap2/usb-musb.c
@@ -28,6 +28,7 @@
 
 #include <mach/hardware.h>
 #include <mach/irqs.h>
+#include <mach/am35xx.h>
 #include <plat/mux.h>
 #include <plat/usb.h>
 
@@ -90,6 +91,9 @@ void __init usb_musb_init(struct omap_musb_board_data *board_data)
 {
 	if (cpu_is_omap243x()) {
 		musb_resources[0].start = OMAP243X_HS_BASE;
+	} else if (cpu_is_omap3517()) {
+		musb_resources[0].start = AM35XX_IPSS_USBOTGSS_BASE;
+		musb_resources[1].start = INT_35XX_USBOTG_IRQ;
 	} else if (cpu_is_omap34xx()) {
 		musb_resources[0].start = OMAP34XX_HSUSB_OTG_BASE;
 	} else if (cpu_is_omap44xx()) {
diff --git a/arch/arm/plat-omap/include/plat/usb.h b/arch/arm/plat-omap/include/plat/usb.h
index 876ca8d..c0e416c 100644
--- a/arch/arm/plat-omap/include/plat/usb.h
+++ b/arch/arm/plat-omap/include/plat/usb.h
@@ -167,5 +167,26 @@ void omap_usb_init(struct omap_usb_config *pdata);
 #	define	USBT2TLL5PI		(1 << 17)
 #	define	USB0PUENACTLOI		(1 << 16)
 #	define	USBSTANDBYCTRL		(1 << 15)
+/* AM3517 */
+/* USB 2.0 PHY Control */
+#define CONF2_PHY_GPIOMODE	(1 << 23)
+#define CONF2_OTGMODE		(3 << 14)
+#define CONF2_NO_OVERRIDE	(0 << 14)
+#define CONF2_FORCE_HOST	(1 << 14)
+#define CONF2_FORCE_DEVICE	(2 << 14)
+#define CONF2_FORCE_HOST_VBUS_LOW (3 << 14)
+#define CONF2_SESENDEN		(1 << 13)
+#define CONF2_VBDTCTEN		(1 << 12)
+#define CONF2_REFFREQ_24MHZ	(2 << 8)
+#define CONF2_REFFREQ_26MHZ	(7 << 8)
+#define CONF2_REFFREQ_13MHZ	(6 << 8)
+#define CONF2_REFFREQ		(0xf << 8)
+#define CONF2_PHYCLKGD		(1 << 7)
+#define CONF2_VBUSSENSE		(1 << 6)
+#define CONF2_PHY_PLLON		(1 << 5)
+#define CONF2_RESET		(1 << 4)
+#define CONF2_PHYPWRDN		(1 << 3)
+#define CONF2_OTGPWRDN		(1 << 2)
+#define CONF2_DATPOL		(1 << 1)
 
 #endif	/* __ASM_ARCH_OMAP_USB_H */
-- 
1.6.2.4

[PATCH 2/3 v2] musb: add musb support for AM35x

[Ajay Kumar Gupta]

AM35x is based on OMAP35x but has an updated musb interface which
uses CPPI4.1 DMA engine.

Current patch supports only PIO mode transfers.

Signed-off-by: Ajay Kumar Gupta <ajay.gupta-l0cyMroinI0@public.gmane.org>
---
Changes from v1: (Based on Sergei's comment)
	- Moved PHY clock and OTGMODE settings to board files
	- Added clk_disable/put in exit path
	- Removed unwanted code related to vbus
	- Removed unhandled interrupt check
	- Added timeout in phy_on()
I had to use MACH_OMAP3517EVM as there is no ARCH_xxx config specific to
AM3517 platforms other than ARCH_OMAP3.

 drivers/usb/musb/Kconfig  |    4 +-
 drivers/usb/musb/Makefile |    4 +
 drivers/usb/musb/am3517.c |  517 +++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 523 insertions(+), 2 deletions(-)
 create mode 100644 drivers/usb/musb/am3517.c

diff --git a/drivers/usb/musb/Kconfig b/drivers/usb/musb/Kconfig
index ea3a882..5ea3d45 100644
--- a/drivers/usb/musb/Kconfig
+++ b/drivers/usb/musb/Kconfig
@@ -10,7 +10,7 @@ comment "Enable Host or Gadget support to see Inventra options"
 config USB_MUSB_HDRC
 	depends on (USB || USB_GADGET)
 	depends on (ARM || (BF54x && !BF544) || (BF52x && !BF522 && !BF523))
-	select NOP_USB_XCEIV if (ARCH_DAVINCI || MACH_OMAP3EVM || BLACKFIN)
+	select NOP_USB_XCEIV if (ARCH_DAVINCI || MACH_OMAP3EVM || BLACKFIN || MACH_OMAP3517EVM)
 	select TWL4030_USB if MACH_OMAP_3430SDP
 	select USB_OTG_UTILS
 	tristate 'Inventra Highspeed Dual Role Controller (TI, ADI, ...)'
@@ -140,7 +140,7 @@ config USB_MUSB_HDRC_HCD
 config MUSB_PIO_ONLY
 	bool 'Disable DMA (always use PIO)'
 	depends on USB_MUSB_HDRC
-	default y if USB_TUSB6010
+	default USB_TUSB6010 || MACH_OMAP3517EVM
 	help
 	  All data is copied between memory and FIFO by the CPU.
 	  DMA controllers are ignored.
diff --git a/drivers/usb/musb/Makefile b/drivers/usb/musb/Makefile
index 3a485da..52a8cf6 100644
--- a/drivers/usb/musb/Makefile
+++ b/drivers/usb/musb/Makefile
@@ -19,7 +19,11 @@ ifeq ($(CONFIG_ARCH_OMAP2430),y)
 endif
 
 ifeq ($(CONFIG_ARCH_OMAP3430),y)
+   ifeq ($(CONFIG_MACH_OMAP3517EVM),y)
+	musb_hdrc-objs  += am3517.o
+   else
 	musb_hdrc-objs	+= omap2430.o
+   endif
 endif
 
 ifeq ($(CONFIG_BF54x),y)
diff --git a/drivers/usb/musb/am3517.c b/drivers/usb/musb/am3517.c
new file mode 100644
index 0000000..dd9e883
--- /dev/null
+++ b/drivers/usb/musb/am3517.c
@@ -0,0 +1,517 @@
+/*
+ * Texas Instruments AM3517 "glue layer"
+ *
+ * Copyright (c) 2010, by Texas Instruments
+ *
+ * Based on the DA8xx "glue layer" code.
+ * Copyright (C) 2005-2006 by Texas Instruments
+ * Copyright (c) 2008, MontaVista Software, Inc. <source-Igf4POYTYCDQT0dZR+AlfA@public.gmane.org>
+ *
+ * This file is part of the Inventra Controller Driver for Linux.
+ *
+ * The Inventra Controller Driver for Linux is free software; you
+ * can redistribute it and/or modify it under the terms of the GNU
+ * General Public License version 2 as published by the Free Software
+ * Foundation.
+ *
+ * The Inventra Controller Driver for Linux is distributed in
+ * the hope that it will be useful, but WITHOUT ANY WARRANTY;
+ * without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public
+ * License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with The Inventra Controller Driver for Linux ; if not,
+ * write to the Free Software Foundation, Inc., 59 Temple Place,
+ * Suite 330, Boston, MA  02111-1307  USA
+ *
+ */
+
+#include <linux/init.h>
+#include <linux/clk.h>
+#include <linux/io.h>
+
+#include <plat/control.h>
+#include <plat/usb.h>
+
+#include "musb_core.h"
+
+/*
+ * AM3517 specific definitions
+ */
+/* USB 2.0 OTG module registers */
+#define USB_REVISION_REG	0x00
+#define USB_CTRL_REG		0x04
+#define USB_STAT_REG		0x08
+#define USB_EMULATION_REG	0x0c
+/* 0x10 Reserved */
+#define USB_AUTOREQ_REG		0x14
+#define USB_SRP_FIX_TIME_REG	0x18
+#define USB_TEARDOWN_REG	0x1c
+#define EP_INTR_SRC_REG		0x20
+#define EP_INTR_SRC_SET_REG	0x24
+#define EP_INTR_SRC_CLEAR_REG	0x28
+#define EP_INTR_MASK_REG	0x2c
+#define EP_INTR_MASK_SET_REG	0x30
+#define EP_INTR_MASK_CLEAR_REG	0x34
+#define EP_INTR_SRC_MASKED_REG	0x38
+#define CORE_INTR_SRC_REG	0x40
+#define CORE_INTR_SRC_SET_REG	0x44
+#define CORE_INTR_SRC_CLEAR_REG	0x48
+#define CORE_INTR_MASK_REG	0x4c
+#define CORE_INTR_MASK_SET_REG	0x50
+#define CORE_INTR_MASK_CLEAR_REG 0x54
+#define CORE_INTR_SRC_MASKED_REG 0x58
+/* 0x5c Reserved */
+#define USB_END_OF_INTR_REG	0x60
+
+/* Control register bits */
+#define USB_SOFT_RESET_MASK	1
+
+/* USB interrupt register bits */
+#define USB_INTR_USB_SHIFT	16
+#define USB_INTR_USB_MASK	(0x1ff << USB_INTR_USB_SHIFT)
+#define USB_INTR_DRVVBUS	0x100
+#define USB_INTR_RX_SHIFT	16
+#define USB_INTR_TX_SHIFT	0
+#define AM3517_TX_EP_MASK	0xffff		/* EP0 + 15 Tx EPs */
+#define AM3517_RX_EP_MASK	0xfffe		/* 15 Rx EPs */
+#define AM3517_TX_INTR_MASK	(AM3517_TX_EP_MASK << USB_INTR_TX_SHIFT)
+#define AM3517_RX_INTR_MASK	(AM3517_RX_EP_MASK << USB_INTR_RX_SHIFT)
+
+#define USB_MENTOR_CORE_OFFSET	0x400
+
+static inline void phy_on(void)
+{
+	unsigned long timeout = jiffies + msecs_to_jiffies(100);
+	u32 devconf2;
+
+	/*
+	 * Start the on-chip PHY and its PLL.
+	 */
+	devconf2 = omap_ctrl_readl(AM35XX_CONTROL_DEVCONF2);
+
+	devconf2 &= ~(CONF2_RESET | CONF2_PHYPWRDN | CONF2_OTGPWRDN |
+			CONF2_PHY_GPIOMODE);
+	devconf2 |= CONF2_PHY_PLLON | CONF2_DATPOL;
+
+	omap_ctrl_writel(devconf2, AM35XX_CONTROL_DEVCONF2);
+
+	DBG(1, "Waiting for PHY clock good...\n");
+	while (!(omap_ctrl_readl(AM35XX_CONTROL_DEVCONF2)
+			& CONF2_PHYCLKGD)) {
+		cpu_relax();
+
+		if (time_after(jiffies, timeout)) {
+			DBG(1, "musb PHY clock good timed out\n");
+			return;
+		}
+	}
+}
+
+static inline void phy_off(void)
+{
+	u32 devconf2;
+
+	/*
+	 * Power down the on-chip PHY.
+	 */
+	devconf2 = omap_ctrl_readl(AM35XX_CONTROL_DEVCONF2);
+
+	devconf2 &= ~CONF2_PHY_PLLON;
+	devconf2 |=  CONF2_PHYPWRDN | CONF2_OTGPWRDN;
+	omap_ctrl_writel(devconf2, AM35XX_CONTROL_DEVCONF2);
+}
+
+/**
+ * musb_platform_enable - enable interrupts
+ */
+void musb_platform_enable(struct musb *musb)
+{
+	void __iomem *reg_base = musb->ctrl_base;
+	u32 epmask, coremask;
+
+	/* Workaround: setup IRQs through both register sets. */
+	epmask = ((musb->epmask & AM3517_TX_EP_MASK) << USB_INTR_TX_SHIFT) |
+	       ((musb->epmask & AM3517_RX_EP_MASK) << USB_INTR_RX_SHIFT);
+	coremask = USB_INTR_USB_MASK;
+
+	musb_writel(reg_base, EP_INTR_MASK_SET_REG, epmask);
+	musb_writel(reg_base, CORE_INTR_MASK_SET_REG, coremask);
+
+	/* Force the DRVVBUS IRQ so we can start polling for ID change. */
+	if (is_otg_enabled(musb))
+		musb_writel(reg_base, CORE_INTR_SRC_SET_REG,
+			    USB_INTR_DRVVBUS << USB_INTR_USB_SHIFT);
+}
+
+/**
+ * musb_platform_disable - disable HDRC and flush interrupts
+ */
+void musb_platform_disable(struct musb *musb)
+{
+	void __iomem *reg_base = musb->ctrl_base;
+
+	musb_writel(reg_base, CORE_INTR_MASK_CLEAR_REG, USB_INTR_USB_MASK);
+	musb_writel(reg_base, EP_INTR_MASK_CLEAR_REG,
+			 AM3517_TX_INTR_MASK | AM3517_RX_INTR_MASK);
+	musb_writeb(musb->mregs, MUSB_DEVCTL, 0);
+	musb_writel(reg_base, USB_END_OF_INTR_REG, 0);
+}
+
+#ifdef CONFIG_USB_MUSB_HDRC_HCD
+#define portstate(stmt) 	stmt
+#else
+#define portstate(stmt)
+#endif
+
+static void am3517_set_vbus(struct musb *musb, int is_on)
+{
+	WARN_ON(is_on && is_peripheral_active(musb));
+}
+
+#define	POLL_SECONDS	2
+
+static struct timer_list otg_workaround;
+
+static void otg_timer(unsigned long _musb)
+{
+	struct musb		*musb = (void *)_musb;
+	void __iomem		*mregs = musb->mregs;
+	u8			devctl;
+	unsigned long		flags;
+
+	/* We poll because AM3517's won't expose several OTG-critical
+	 * status change events (from the transceiver) otherwise.
+	 */
+	devctl = musb_readb(mregs, MUSB_DEVCTL);
+	DBG(7, "Poll devctl %02x (%s)\n", devctl, otg_state_string(musb));
+
+	spin_lock_irqsave(&musb->lock, flags);
+	switch (musb->xceiv->state) {
+	case OTG_STATE_A_WAIT_BCON:
+		devctl &= ~MUSB_DEVCTL_SESSION;
+		musb_writeb(musb->mregs, MUSB_DEVCTL, devctl);
+
+		devctl = musb_readb(musb->mregs, MUSB_DEVCTL);
+		if (devctl & MUSB_DEVCTL_BDEVICE) {
+			musb->xceiv->state = OTG_STATE_B_IDLE;
+			MUSB_DEV_MODE(musb);
+		} else {
+			musb->xceiv->state = OTG_STATE_A_IDLE;
+			MUSB_HST_MODE(musb);
+		}
+		break;
+	case OTG_STATE_A_WAIT_VFALL:
+		musb->xceiv->state = OTG_STATE_A_WAIT_VRISE;
+		musb_writel(musb->ctrl_base, CORE_INTR_SRC_SET_REG,
+			    MUSB_INTR_VBUSERROR << USB_INTR_USB_SHIFT);
+		break;
+	case OTG_STATE_B_IDLE:
+		if (!is_peripheral_enabled(musb))
+			break;
+
+		devctl = musb_readb(mregs, MUSB_DEVCTL);
+		if (devctl & MUSB_DEVCTL_BDEVICE)
+			mod_timer(&otg_workaround, jiffies + POLL_SECONDS * HZ);
+		else
+			musb->xceiv->state = OTG_STATE_A_IDLE;
+		break;
+	default:
+		break;
+	}
+	spin_unlock_irqrestore(&musb->lock, flags);
+}
+
+void musb_platform_try_idle(struct musb *musb, unsigned long timeout)
+{
+	static unsigned long last_timer;
+
+	if (!is_otg_enabled(musb))
+		return;
+
+	if (timeout == 0)
+		timeout = jiffies + msecs_to_jiffies(3);
+
+	/* Never idle if active, or when VBUS timeout is not set as host */
+	if (musb->is_active || (musb->a_wait_bcon == 0 &&
+				musb->xceiv->state == OTG_STATE_A_WAIT_BCON)) {
+		DBG(4, "%s active, deleting timer\n", otg_state_string(musb));
+		del_timer(&otg_workaround);
+		last_timer = jiffies;
+		return;
+	}
+
+	if (time_after(last_timer, timeout) && timer_pending(&otg_workaround)) {
+		DBG(4, "Longer idle timer already pending, ignoring...\n");
+		return;
+	}
+	last_timer = timeout;
+
+	DBG(4, "%s inactive, starting idle timer for %u ms\n",
+	    otg_state_string(musb), jiffies_to_msecs(timeout - jiffies));
+	mod_timer(&otg_workaround, timeout);
+}
+
+static irqreturn_t am3517_interrupt(int irq, void *hci)
+{
+	struct musb  *musb = hci;
+	void __iomem *reg_base = musb->ctrl_base;
+	unsigned long flags;
+	irqreturn_t ret = IRQ_NONE;
+	u32 epintr, usbintr, lvl_intr;
+
+	spin_lock_irqsave(&musb->lock, flags);
+
+	/* Get endpoint interrupts */
+	epintr = musb_readl(reg_base, EP_INTR_SRC_MASKED_REG);
+
+	if (epintr) {
+		musb_writel(reg_base, EP_INTR_SRC_CLEAR_REG, epintr);
+
+		musb->int_rx =
+			(epintr & AM3517_RX_INTR_MASK) >> USB_INTR_RX_SHIFT;
+		musb->int_tx =
+			(epintr & AM3517_TX_INTR_MASK) >> USB_INTR_TX_SHIFT;
+	}
+
+	/* Get usb core interrupts */
+	usbintr = musb_readl(reg_base, CORE_INTR_SRC_MASKED_REG);
+	if (!usbintr && !epintr)
+		goto eoi;
+
+	if (usbintr) {
+		musb_writel(reg_base, CORE_INTR_SRC_CLEAR_REG, usbintr);
+
+		musb->int_usb =
+			(usbintr & USB_INTR_USB_MASK) >> USB_INTR_USB_SHIFT;
+	}
+	/*
+	 * DRVVBUS IRQs are the only proxy we have (a very poor one!) for
+	 * AM3517's missing ID change IRQ.  We need an ID change IRQ to
+	 * switch appropriately between halves of the OTG state machine.
+	 * Managing DEVCTL.SESSION per Mentor docs requires that we know its
+	 * value but DEVCTL.BDEVICE is invalid without DEVCTL.SESSION set.
+	 * Also, DRVVBUS pulses for SRP (but not at 5V) ...
+	 */
+	if (usbintr & (USB_INTR_DRVVBUS << USB_INTR_USB_SHIFT)) {
+		int drvvbus = musb_readl(reg_base, USB_STAT_REG);
+		void __iomem *mregs = musb->mregs;
+		u8 devctl = musb_readb(mregs, MUSB_DEVCTL);
+		int err;
+
+		err = is_host_enabled(musb) && (musb->int_usb &
+						MUSB_INTR_VBUSERROR);
+		if (err) {
+			/*
+			 * The Mentor core doesn't debounce VBUS as needed
+			 * to cope with device connect current spikes. This
+			 * means it's not uncommon for bus-powered devices
+			 * to get VBUS errors during enumeration.
+			 *
+			 * This is a workaround, but newer RTL from Mentor
+			 * seems to allow a better one: "re"-starting sessions
+			 * without waiting for VBUS to stop registering in
+			 * devctl.
+			 */
+			musb->int_usb &= ~MUSB_INTR_VBUSERROR;
+			musb->xceiv->state = OTG_STATE_A_WAIT_VFALL;
+			mod_timer(&otg_workaround, jiffies + POLL_SECONDS * HZ);
+			WARNING("VBUS error workaround (delay coming)\n");
+		} else if (is_host_enabled(musb) && drvvbus) {
+			musb->is_active = 1;
+			MUSB_HST_MODE(musb);
+			musb->xceiv->default_a = 1;
+			musb->xceiv->state = OTG_STATE_A_WAIT_VRISE;
+			portstate(musb->port1_status |= USB_PORT_STAT_POWER);
+			del_timer(&otg_workaround);
+		} else {
+			musb->is_active = 0;
+			MUSB_DEV_MODE(musb);
+			musb->xceiv->default_a = 0;
+			musb->xceiv->state = OTG_STATE_B_IDLE;
+			portstate(musb->port1_status &= ~USB_PORT_STAT_POWER);
+		}
+
+		/* NOTE: this must complete power-on within 100 ms. */
+		DBG(2, "VBUS %s (%s)%s, devctl %02x\n",
+				drvvbus ? "on" : "off",
+				otg_state_string(musb),
+				err ? " ERROR" : "",
+				devctl);
+		ret = IRQ_HANDLED;
+	}
+
+	if (musb->int_tx || musb->int_rx || musb->int_usb)
+		ret |= musb_interrupt(musb);
+
+ eoi:
+	/* EOI needs to be written for the IRQ to be re-asserted. */
+	if (ret == IRQ_HANDLED || epintr || usbintr) {
+		/* clear level interrupt */
+		lvl_intr = omap_ctrl_readl(AM35XX_CONTROL_LVL_INTR_CLEAR);
+		lvl_intr |= AM35XX_USBOTGSS_INT_CLR;
+		omap_ctrl_writel(lvl_intr, AM35XX_CONTROL_LVL_INTR_CLEAR);
+		/* write EOI */
+		musb_writel(reg_base, USB_END_OF_INTR_REG, 0);
+	}
+
+	/* Poll for ID change */
+	if (is_otg_enabled(musb) && musb->xceiv->state == OTG_STATE_B_IDLE)
+		mod_timer(&otg_workaround, jiffies + POLL_SECONDS * HZ);
+
+	spin_unlock_irqrestore(&musb->lock, flags);
+
+	return ret;
+}
+
+int musb_platform_set_mode(struct musb *musb, u8 musb_mode)
+{
+	u32 devconf2 = omap_ctrl_readl(AM35XX_CONTROL_DEVCONF2);
+
+	devconf2 &= ~CONF2_OTGMODE;
+	switch (musb_mode) {
+#ifdef	CONFIG_USB_MUSB_HDRC_HCD
+	case MUSB_HOST:		/* Force VBUS valid, ID = 0 */
+		devconf2 |= CONF2_FORCE_HOST;
+		break;
+#endif
+#ifdef	CONFIG_USB_GADGET_MUSB_HDRC
+	case MUSB_PERIPHERAL:	/* Force VBUS valid, ID = 1 */
+		devconf2 |= CONF2_FORCE_DEVICE;
+		break;
+#endif
+#ifdef	CONFIG_USB_MUSB_OTG
+	case MUSB_OTG:		/* Don't override the VBUS/ID comparators */
+		devconf2 |= CONF2_NO_OVERRIDE;
+		break;
+#endif
+	default:
+		DBG(2, "Trying to set unsupported mode %u\n", musb_mode);
+	}
+
+	omap_ctrl_writel(devconf2, AM35XX_CONTROL_DEVCONF2);
+	return 0;
+}
+
+int __init musb_platform_init(struct musb *musb)
+{
+	void __iomem *reg_base = musb->ctrl_base;
+	struct clk              *otg_fck;
+	u32 rev, lvl_intr, sw_reset;
+
+	musb->mregs += USB_MENTOR_CORE_OFFSET;
+
+	if (musb->set_clock)
+		musb->set_clock(musb->clock, 1);
+	else
+		clk_enable(musb->clock);
+	DBG(2, "usbotg_ck=%lud\n", clk_get_rate(musb->clock));
+
+	otg_fck = clk_get(musb->controller, "fck");
+	clk_enable(otg_fck);
+	DBG(2, "usbotg_phy_ck=%lud\n", clk_get_rate(otg_fck));
+
+	/* Returns zero if e.g. not clocked */
+	rev = musb_readl(reg_base, USB_REVISION_REG);
+	if (!rev)
+		return -ENODEV;
+
+	usb_nop_xceiv_register();
+	musb->xceiv = otg_get_transceiver();
+	if (!musb->xceiv)
+		return -ENODEV;
+
+	if (is_host_enabled(musb))
+		setup_timer(&otg_workaround, otg_timer, (unsigned long) musb);
+
+	musb->board_set_vbus = am3517_set_vbus;
+
+	/* Global reset */
+	sw_reset = omap_ctrl_readl(AM35XX_CONTROL_IP_SW_RESET);
+
+	sw_reset |= AM35XX_USBOTGSS_SW_RST;
+	omap_ctrl_writel(sw_reset, AM35XX_CONTROL_IP_SW_RESET);
+
+	sw_reset &= ~AM35XX_USBOTGSS_SW_RST;
+	omap_ctrl_writel(sw_reset, AM35XX_CONTROL_IP_SW_RESET);
+
+	/* Reset the controller */
+	musb_writel(reg_base, USB_CTRL_REG, USB_SOFT_RESET_MASK);
+
+	/* Start the on-chip PHY and its PLL. */
+	phy_on();
+
+	msleep(5);
+
+	musb->isr = am3517_interrupt;
+
+	/* clear level interrupt */
+	lvl_intr = omap_ctrl_readl(AM35XX_CONTROL_LVL_INTR_CLEAR);
+	lvl_intr |= AM35XX_USBOTGSS_INT_CLR;
+	omap_ctrl_writel(lvl_intr, AM35XX_CONTROL_LVL_INTR_CLEAR);
+	return 0;
+}
+
+int musb_platform_exit(struct musb *musb)
+{
+	struct clk *otg_fck = clk_get(musb->controller, "fck");
+
+	if (is_host_enabled(musb))
+		del_timer_sync(&otg_workaround);
+
+	/* Delay to avoid problems with module reload... */
+	if (is_host_enabled(musb) && musb->xceiv->default_a) {
+		u8 devctl, warn = 0;
+		int delay;
+
+		/*
+		 * If there's no peripheral connected, VBUS can take a
+		 * long time to fall...
+		 */
+		for (delay = 30; delay > 0; delay--) {
+			devctl = musb_readb(musb->mregs, MUSB_DEVCTL);
+			if (!(devctl & MUSB_DEVCTL_VBUS))
+				goto done;
+			if ((devctl & MUSB_DEVCTL_VBUS) != warn) {
+				warn = devctl & MUSB_DEVCTL_VBUS;
+				DBG(1, "VBUS %d\n",
+					warn >> MUSB_DEVCTL_VBUS_SHIFT);
+			}
+			msleep(1000);
+		}
+
+		/* In OTG mode, another host might be connected... */
+		DBG(1, "VBUS off timeout (devctl %02x)\n", devctl);
+	}
+done:
+	phy_off();
+
+	usb_nop_xceiv_unregister();
+
+	if (musb->set_clock)
+		musb->set_clock(musb->clock, 0);
+	else
+		clk_disable(musb->clock);
+
+	if (otg_fck) {
+		clk_put(otg_fck);
+		clk_disable(otg_fck);
+	}
+
+	return 0;
+}
+
+#ifdef CONFIG_PM
+void musb_platform_save_context(struct musb_context_registers
+		*musb_context)
+{
+	phy_off();
+}
+
+void musb_platform_restore_context(struct musb_context_registers
+		*musb_context)
+{
+	phy_on();
+}
+#endif
-- 
1.6.2.4

--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 3/3 v2] musb: AM35x: Workaround for fifo read issue

[Ajay Kumar Gupta]

AM35x supports only 32bit read operations so we need to have
workaround for 8bit and 16bit read operations.

Signed-off-by: Ajay Kumar Gupta <ajay.gupta@ti.com>
---
 drivers/usb/musb/am3517.c    |   30 ++++++++++++++++++++++++++++++
 drivers/usb/musb/musb_core.c |    2 ++
 2 files changed, 32 insertions(+), 0 deletions(-)

diff --git a/drivers/usb/musb/am3517.c b/drivers/usb/musb/am3517.c
index dd9e883..24bdc2e 100644
--- a/drivers/usb/musb/am3517.c
+++ b/drivers/usb/musb/am3517.c
@@ -515,3 +515,33 @@ void musb_platform_restore_context(struct musb_context_registers
 	phy_on();
 }
 #endif
+
+/* AM35x supports only 32bit read operation */
+void musb_read_fifo(struct musb_hw_ep *hw_ep, u16 len, u8 *dst)
+{
+	void __iomem *fifo = hw_ep->fifo;
+	u32		val;
+	int		i;
+
+	/* Read for 32bit-aligned destination address */
+	if ((likely((0x03 & (unsigned long) dst) == 0)) && len >= 4) {
+		readsl(fifo, dst, len >> 2);
+		dst += (len & ~0x03);
+		len &= 0x03;
+	}
+	/* Now read the rest 1 to 3 bytes or complete length if
+	 * unaligned address.
+	 */
+	if (len > 4) {
+		for (i = 0; i < (len >> 2); i++) {
+			val = musb_readl(fifo, 0);
+			memcpy(dst, &val, 4);
+			dst += 4;
+		}
+		len %= 4;
+	}
+	if (len > 0) {
+		val = musb_readl(fifo, 0);
+		memcpy(dst, &val, len);
+	}
+}
diff --git a/drivers/usb/musb/musb_core.c b/drivers/usb/musb/musb_core.c
index 705cc4a..bc2cf14 100644
--- a/drivers/usb/musb/musb_core.c
+++ b/drivers/usb/musb/musb_core.c
@@ -191,6 +191,7 @@ void musb_write_fifo(struct musb_hw_ep *hw_ep, u16 len, const u8 *src)
 	}
 }
 
+#if !defined(CONFIG_MACH_OMAP3517EVM)
 /*
  * Unload an endpoint's FIFO
  */
@@ -228,6 +229,7 @@ void musb_read_fifo(struct musb_hw_ep *hw_ep, u16 len, u8 *dst)
 		readsb(fifo, dst, len);
 	}
 }
+#endif
 
 #endif	/* normal PIO */
 
-- 
1.6.2.4

Re: [PATCH 3/3 v2] musb: AM35x: Workaround for fifo read issue

[Venkatraman S]

Ajay Kumar Gupta <ajay.gupta@ti.com> wrote:
> AM35x supports only 32bit read operations so we need to have
> workaround for 8bit and 16bit read operations.
>
> Signed-off-by: Ajay Kumar Gupta <ajay.gupta@ti.com>
> ---
>  drivers/usb/musb/am3517.c    |   30 ++++++++++++++++++++++++++++++
>  drivers/usb/musb/musb_core.c |    2 ++
>  2 files changed, 32 insertions(+), 0 deletions(-)
>
> diff --git a/drivers/usb/musb/am3517.c b/drivers/usb/musb/am3517.c
> index dd9e883..24bdc2e 100644
> --- a/drivers/usb/musb/am3517.c
> +++ b/drivers/usb/musb/am3517.c
> @@ -515,3 +515,33 @@ void musb_platform_restore_context(struct musb_context_registers
>        phy_on();
>  }
>  #endif
> +
> +/* AM35x supports only 32bit read operation */
> +void musb_read_fifo(struct musb_hw_ep *hw_ep, u16 len, u8 *dst)
> +{
> +       void __iomem *fifo = hw_ep->fifo;
> +       u32             val;
> +       int             i;
> +
> +       /* Read for 32bit-aligned destination address */
> +       if ((likely((0x03 & (unsigned long) dst) == 0)) && len >= 4) {
> +               readsl(fifo, dst, len >> 2);
> +               dst += (len & ~0x03);
> +               len &= 0x03;
> +       }
> +       /* Now read the rest 1 to 3 bytes or complete length if
check multiline comments style
> +        * unaligned address.
> +        */
> +       if (len > 4) {
> +               for (i = 0; i < (len >> 2); i++) {
> +                       val = musb_readl(fifo, 0);
> +                       memcpy(dst, &val, 4);
> +                       dst += 4;
> +               }
> +               len %= 4;
> +       }
> +       if (len > 0) {
> +               val = musb_readl(fifo, 0);
> +               memcpy(dst, &val, len);
> +       }
> +}
> diff --git a/drivers/usb/musb/musb_core.c b/drivers/usb/musb/musb_core.c
> index 705cc4a..bc2cf14 100644
> --- a/drivers/usb/musb/musb_core.c
> +++ b/drivers/usb/musb/musb_core.c
> @@ -191,6 +191,7 @@ void musb_write_fifo(struct musb_hw_ep *hw_ep, u16 len, const u8 *src)
>        }
>  }
>
> +#if !defined(CONFIG_MACH_OMAP3517EVM)
>  /*
>  * Unload an endpoint's FIFO
>  */
> @@ -228,6 +229,7 @@ void musb_read_fifo(struct musb_hw_ep *hw_ep, u16 len, u8 *dst)
>                readsb(fifo, dst, len);
>        }
>  }
> +#endif
>
>  #endif /* normal PIO */
>
> --
> 1.6.2.4
>
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html