[PATCH 2/4] ARM: mach-shmobile: r8a7779: add SATA support

[PATCH 2/4] ARM: mach-shmobile: r8a7779: add SATA support

[Sergei Shtylyov]

From: Vladimir Barinov <vladimir.barinov@cogentembedded.com>

Add SATA clock and platform device resources on r8a7779 SoC.
Add entry to r8a7779_auxdata_lookup[], so that devm_clk_get() in the driver
still works when we're using the device tree.
 
Signed-off-by: Vladimir Barinov <vladimir.barinov@cogentembedded.com>
Signed-off-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>

---
 arch/arm/mach-shmobile/clock-r8a7779.c |    3 +++
 arch/arm/mach-shmobile/setup-r8a7779.c |   22 ++++++++++++++++++++++
 2 files changed, 25 insertions(+)

Index: renesas/arch/arm/mach-shmobile/clock-r8a7779.c
===================================================================
--- renesas.orig/arch/arm/mach-shmobile/clock-r8a7779.c
+++ renesas/arch/arm/mach-shmobile/clock-r8a7779.c
@@ -87,6 +87,7 @@ static struct clk div4_clks[DIV4_NR] = {
 };
 
 enum { MSTP323, MSTP322, MSTP321, MSTP320,
+	MSTP115,
 	MSTP101, MSTP100,
 	MSTP030,
 	MSTP029, MSTP028, MSTP027, MSTP026, MSTP025, MSTP024, MSTP023, MSTP022, MSTP021,
@@ -99,6 +100,7 @@ static struct clk mstp_clks[MSTP_NR] = {
 	[MSTP322] = SH_CLK_MSTP32(&div4_clks[DIV4_P], MSTPCR3, 22, 0), /* SDHI1 */
 	[MSTP321] = SH_CLK_MSTP32(&div4_clks[DIV4_P], MSTPCR3, 21, 0), /* SDHI2 */
 	[MSTP320] = SH_CLK_MSTP32(&div4_clks[DIV4_P], MSTPCR3, 20, 0), /* SDHI3 */
+	[MSTP115] = SH_CLK_MSTP32(&div4_clks[DIV4_P], MSTPCR1, 15, 0), /* SATA */
 	[MSTP101] = SH_CLK_MSTP32(&div4_clks[DIV4_P], MSTPCR1,  1, 0), /* USB2 */
 	[MSTP100] = SH_CLK_MSTP32(&div4_clks[DIV4_P], MSTPCR1,  0, 0), /* USB0/1 */
 	[MSTP030] = SH_CLK_MSTP32(&div4_clks[DIV4_P], MSTPCR0, 30, 0), /* I2C0 */
@@ -156,6 +158,7 @@ static struct clk_lookup lookups[] = {
 	CLKDEV_CON_ID("peripheral_clk",	&div4_clks[DIV4_P]),
 
 	/* MSTP32 clocks */
+	CLKDEV_DEV_ID("sata_rcar", &mstp_clks[MSTP115]), /* SATA */
 	CLKDEV_DEV_ID("ehci-platform.1", &mstp_clks[MSTP101]), /* USB EHCI port2 */
 	CLKDEV_DEV_ID("ohci-platform.1", &mstp_clks[MSTP101]), /* USB OHCI port2 */
 	CLKDEV_DEV_ID("ehci-platform.0", &mstp_clks[MSTP100]), /* USB EHCI port0/1 */
Index: renesas/arch/arm/mach-shmobile/setup-r8a7779.c
===================================================================
--- renesas.orig/arch/arm/mach-shmobile/setup-r8a7779.c
+++ renesas/arch/arm/mach-shmobile/setup-r8a7779.c
@@ -322,6 +322,26 @@ static struct platform_device i2c3_devic
 	.num_resources	= ARRAY_SIZE(rcar_i2c3_res),
 };
 
+static struct resource sata_resources[] = {
+	[0] = {
+		.name	= "rcar-sata",
+		.start	= 0xfc600000,
+		.end	= 0xfc601fff,
+		.flags	= IORESOURCE_MEM,
+	},
+	[1] = {
+		.start	= gic_spi(100),
+		.flags	= IORESOURCE_IRQ,
+	},
+};
+
+static struct platform_device sata_device = {
+	.name		= "sata_rcar",
+	.id		= -1,
+	.resource	= sata_resources,
+	.num_resources	= ARRAY_SIZE(sata_resources),
+};
+
 static struct platform_device *r8a7779_early_devices_dt[] __initdata = {
 	&scif0_device,
 	&scif1_device,
@@ -338,6 +358,7 @@ static struct platform_device *r8a7779_e
 	&i2c1_device,
 	&i2c2_device,
 	&i2c3_device,
+	&sata_device,
 };
 
 void __init r8a7779_add_standard_devices(void)
@@ -404,6 +425,7 @@ void __init r8a7779_add_early_devices_dt
 }
 
 static const struct of_dev_auxdata r8a7779_auxdata_lookup[] __initconst = {
+	OF_DEV_AUXDATA("renesas,rcar-sata", 0xfc600000, "sata_rcar", NULL),
 	{},
 };
 

[PATCH 2/4] ARM: mach-shmobile: r8a7779: add SATA support

[Kuninori Morimoto]

Hi Sergei

> From: Vladimir Barinov <vladimir.barinov@cogentembedded.com>
> 
> Add SATA clock and platform device resources on r8a7779 SoC.
> Add entry to r8a7779_auxdata_lookup[], so that devm_clk_get() in the driver
> still works when we're using the device tree.
>  
> Signed-off-by: Vladimir Barinov <vladimir.barinov@cogentembedded.com>
> Signed-off-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>
(snip)

>  	/* MSTP32 clocks */
> +	CLKDEV_DEV_ID("sata_rcar", &mstp_clks[MSTP115]), /* SATA */
>  	CLKDEV_DEV_ID("ehci-platform.1", &mstp_clks[MSTP101]), /* USB EHCI port2 */
>  	CLKDEV_DEV_ID("ohci-platform.1", &mstp_clks[MSTP101]), /* USB OHCI port2 */
>  	CLKDEV_DEV_ID("ehci-platform.0", &mstp_clks[MSTP100]), /* USB EHCI port0/1 */
(snip)
>  static const struct of_dev_auxdata r8a7779_auxdata_lookup[] __initconst = {
> +	OF_DEV_AUXDATA("renesas,rcar-sata", 0xfc600000, "sata_rcar", NULL),

??
Is this settings really required for DT ??

I guess you can remove it, and add

 +	CLKDEV_DEV_ID("sata_rcar", &mstp_clks[MSTP115]), // for platform
 +	CLKDEV_DEV_ID("fc600000.sata_rcar", &mstp_clks[MSTP115]), // for DT

And...

> [1/4] ARM: mach-shmobile: r8a7779: SATA DT configuration
> [2/4] ARM: mach-shmobile: r8a7779: add SATA support
> [3/4] libata: add R-Car SATA driver
> [4/4] ARM: mach-shmobile: marzen: add SATA support

I believe [3/4] patch should be base patch ?

Best regards
---
Kuninori Morimoto

[PATCH 2/4] ARM: mach-shmobile: r8a7779: add SATA support

[Sergei Shtylyov]

Hello.

On 18-02-2013 5:23, Kuninori Morimoto wrote:

>> From: Vladimir Barinov <vladimir.barinov@cogentembedded.com>

>> Add SATA clock and platform device resources on r8a7779 SoC.
>> Add entry to r8a7779_auxdata_lookup[], so that devm_clk_get() in the driver
>> still works when we're using the device tree.

>> Signed-off-by: Vladimir Barinov <vladimir.barinov@cogentembedded.com>
>> Signed-off-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>
> (snip)

>>   	/* MSTP32 clocks */
>> +	CLKDEV_DEV_ID("sata_rcar", &mstp_clks[MSTP115]), /* SATA */
>>   	CLKDEV_DEV_ID("ehci-platform.1", &mstp_clks[MSTP101]), /* USB EHCI port2 */
>>   	CLKDEV_DEV_ID("ohci-platform.1", &mstp_clks[MSTP101]), /* USB OHCI port2 */
>>   	CLKDEV_DEV_ID("ehci-platform.0", &mstp_clks[MSTP100]), /* USB EHCI port0/1 */
> (snip)
>>   static const struct of_dev_auxdata r8a7779_auxdata_lookup[] __initconst = {
>> +	OF_DEV_AUXDATA("renesas,rcar-sata", 0xfc600000, "sata_rcar", NULL),

> ??
> Is this settings really required for DT ??

    Yes, TTBOMK, it's the last resort measure used in exctly this case.

> I guess you can remove it, and add

>   +	CLKDEV_DEV_ID("sata_rcar", &mstp_clks[MSTP115]), // for platform
>   +	CLKDEV_DEV_ID("fc600000.sata_rcar", &mstp_clks[MSTP115]), // for DT

    IMO, this neither looks nor scales well.

> And...

>> [1/4] ARM: mach-shmobile: r8a7779: SATA DT configuration
>> [2/4] ARM: mach-shmobile: r8a7779: add SATA support
>> [3/4] libata: add R-Car SATA driver
>> [4/4] ARM: mach-shmobile: marzen: add SATA support

> I believe [3/4] patch should be base patch ?

    You're probably right, I'll reorder the patches when posting V2.

> Best regards
> ---
> Kuninori Morimoto

WBR, Sergei

[PATCH 2/4] ARM: mach-shmobile: r8a7779: add SATA support

[Magnus Damm]

Hi Sergei,

Thanks for your efforts with this SATA driver.

On Mon, Feb 18, 2013 at 11:07 PM, Sergei Shtylyov
<sergei.shtylyov@cogentembedded.com> wrote:
> On 18-02-2013 5:23, Kuninori Morimoto wrote:
>>> From: Vladimir Barinov <vladimir.barinov@cogentembedded.com>
>>> Add SATA clock and platform device resources on r8a7779 SoC.
>>> Add entry to r8a7779_auxdata_lookup[], so that devm_clk_get() in the
>>> driver
>>> still works when we're using the device tree.
>
>
>>> Signed-off-by: Vladimir Barinov <vladimir.barinov@cogentembedded.com>
>>> Signed-off-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>
>>
>> (snip)
>
>
>>>         /* MSTP32 clocks */
>>> +       CLKDEV_DEV_ID("sata_rcar", &mstp_clks[MSTP115]), /* SATA */
>>>         CLKDEV_DEV_ID("ehci-platform.1", &mstp_clks[MSTP101]), /* USB
>>> EHCI port2 */
>>>         CLKDEV_DEV_ID("ohci-platform.1", &mstp_clks[MSTP101]), /* USB
>>> OHCI port2 */
>>>         CLKDEV_DEV_ID("ehci-platform.0", &mstp_clks[MSTP100]), /* USB
>>> EHCI port0/1 */
>>
>> (snip)
>>>
>>>   static const struct of_dev_auxdata r8a7779_auxdata_lookup[] __initconst
>>> = {
>>> +       OF_DEV_AUXDATA("renesas,rcar-sata", 0xfc600000, "sata_rcar",
>>> NULL),
>
>
>> ??
>> Is this settings really required for DT ??
>
>
>    Yes, TTBOMK, it's the last resort measure used in exctly this case.

Well, I have to agree with Morimoto-san here. Other vendors may chose
to use AUXDATA to map clocks, and I believe it makes sense in the case
of adding platform data as a workaround during transition to full DT
support. But for simply mapping clocks please follow the same style as
we have done so far, which is what Morimoto-san pointed out:

CLKDEV_DEV_ID("fc600000.sata_rcar", &mstp_clks[MSTP115]),

Also, I don't think SATA is needed as an early device so it should be
enough to register it late as a regular platform device. =)

Thanks,

/ magnus

[PATCH 2/4] ARM: mach-shmobile: r8a7779: add SATA support

[Kuninori Morimoto]

Hi Sergei, Magnus, Simon

> >>>   static const struct of_dev_auxdata r8a7779_auxdata_lookup[] __initconst
> >>> = {
> >>> +       OF_DEV_AUXDATA("renesas,rcar-sata", 0xfc600000, "sata_rcar",
> >>> NULL),
> >
> >
> >> ??
> >> Is this settings really required for DT ??
> >
> >
> >    Yes, TTBOMK, it's the last resort measure used in exctly this case.
> 
> Well, I have to agree with Morimoto-san here. Other vendors may chose
> to use AUXDATA to map clocks, and I believe it makes sense in the case
> of adding platform data as a workaround during transition to full DT
> support. But for simply mapping clocks please follow the same style as
> we have done so far, which is what Morimoto-san pointed out:
> 
> CLKDEV_DEV_ID("fc600000.sata_rcar", &mstp_clks[MSTP115]),
> 
> Also, I don't think SATA is needed as an early device so it should be
> enough to register it late as a regular platform device. =)

BTW, Sergei, which branch are you using for R-Car H1 DT booting ?
I'm not sure detail, but, I guess R-Car H1 DT detection needs
simon's latest branch.
If you are using old branch, DT detection will be failed ?

Simon, is my comment correct ?

Best regards
---
Kuninori Morimoto

[PATCH 2/4] ARM: mach-shmobile: r8a7779: add SATA support

[Simon Horman]

On Mon, Feb 18, 2013 at 11:21:19PM +0900, Magnus Damm wrote:
> Hi Sergei,
> 
> Thanks for your efforts with this SATA driver.
> 
> On Mon, Feb 18, 2013 at 11:07 PM, Sergei Shtylyov
> <sergei.shtylyov@cogentembedded.com> wrote:
> > On 18-02-2013 5:23, Kuninori Morimoto wrote:
> >>> From: Vladimir Barinov <vladimir.barinov@cogentembedded.com>
> >>> Add SATA clock and platform device resources on r8a7779 SoC.
> >>> Add entry to r8a7779_auxdata_lookup[], so that devm_clk_get() in the
> >>> driver
> >>> still works when we're using the device tree.
> >
> >
> >>> Signed-off-by: Vladimir Barinov <vladimir.barinov@cogentembedded.com>
> >>> Signed-off-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>
> >>
> >> (snip)
> >
> >
> >>>         /* MSTP32 clocks */
> >>> +       CLKDEV_DEV_ID("sata_rcar", &mstp_clks[MSTP115]), /* SATA */
> >>>         CLKDEV_DEV_ID("ehci-platform.1", &mstp_clks[MSTP101]), /* USB
> >>> EHCI port2 */
> >>>         CLKDEV_DEV_ID("ohci-platform.1", &mstp_clks[MSTP101]), /* USB
> >>> OHCI port2 */
> >>>         CLKDEV_DEV_ID("ehci-platform.0", &mstp_clks[MSTP100]), /* USB
> >>> EHCI port0/1 */
> >>
> >> (snip)
> >>>
> >>>   static const struct of_dev_auxdata r8a7779_auxdata_lookup[] __initconst
> >>> = {
> >>> +       OF_DEV_AUXDATA("renesas,rcar-sata", 0xfc600000, "sata_rcar",
> >>> NULL),
> >
> >
> >> ??
> >> Is this settings really required for DT ??
> >
> >
> >    Yes, TTBOMK, it's the last resort measure used in exctly this case.
> 
> Well, I have to agree with Morimoto-san here. Other vendors may chose
> to use AUXDATA to map clocks, and I believe it makes sense in the case
> of adding platform data as a workaround during transition to full DT
> support. But for simply mapping clocks please follow the same style as
> we have done so far, which is what Morimoto-san pointed out:
> 
> CLKDEV_DEV_ID("fc600000.sata_rcar", &mstp_clks[MSTP115]),

FWIW, this CLKDEV_DEV_ID() is consistent with how shmobile has
handled other cases so far. (Mostly because of the argument Magnus
makes above).

> Also, I don't think SATA is needed as an early device so it should be
> enough to register it late as a regular platform device. =)
> 
> Thanks,
> 
> / magnus
> --
> To unsubscribe from this list: send the line "unsubscribe linux-sh" in
> the body of a message to majordomo at vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 

[PATCH 2/4] ARM: mach-shmobile: r8a7779: add SATA support

[Simon Horman]

On Mon, Feb 18, 2013 at 04:40:16PM -0800, Kuninori Morimoto wrote:
> 
> Hi Sergei, Magnus, Simon
> 
> > >>>   static const struct of_dev_auxdata r8a7779_auxdata_lookup[] __initconst
> > >>> = {
> > >>> +       OF_DEV_AUXDATA("renesas,rcar-sata", 0xfc600000, "sata_rcar",
> > >>> NULL),
> > >
> > >
> > >> ??
> > >> Is this settings really required for DT ??
> > >
> > >
> > >    Yes, TTBOMK, it's the last resort measure used in exctly this case.
> > 
> > Well, I have to agree with Morimoto-san here. Other vendors may chose
> > to use AUXDATA to map clocks, and I believe it makes sense in the case
> > of adding platform data as a workaround during transition to full DT
> > support. But for simply mapping clocks please follow the same style as
> > we have done so far, which is what Morimoto-san pointed out:
> > 
> > CLKDEV_DEV_ID("fc600000.sata_rcar", &mstp_clks[MSTP115]),
> > 
> > Also, I don't think SATA is needed as an early device so it should be
> > enough to register it late as a regular platform device. =)
> 
> BTW, Sergei, which branch are you using for R-Car H1 DT booting ?
> I'm not sure detail, but, I guess R-Car H1 DT detection needs
> simon's latest branch.
> If you are using old branch, DT detection will be failed ?
> 
> Simon, is my comment correct ?

I guess so too.

[PATCH 2/4] ARM: mach-shmobile: r8a7779: add SATA support

[Sergei Shtylyov]

Hello.

On 19-02-2013 4:40, Kuninori Morimoto wrote:

>>>>>    static const struct of_dev_auxdata r8a7779_auxdata_lookup[] __initconst
>>>>> = {
>>>>> +       OF_DEV_AUXDATA("renesas,rcar-sata", 0xfc600000, "sata_rcar",
>>>>> NULL),


>>>> ??
>>>> Is this settings really required for DT ??


>>>     Yes, TTBOMK, it's the last resort measure used in exctly this case.

>> Well, I have to agree with Morimoto-san here. Other vendors may chose
>> to use AUXDATA to map clocks, and I believe it makes sense in the case
>> of adding platform data as a workaround during transition to full DT
>> support. But for simply mapping clocks please follow the same style as
>> we have done so far, which is what Morimoto-san pointed out:

>> CLKDEV_DEV_ID("fc600000.sata_rcar", &mstp_clks[MSTP115]),

>> Also, I don't think SATA is needed as an early device so it should be
>> enough to register it late as a regular platform device. =)

> BTW, Sergei, which branch are you using for R-Car H1 DT booting ?
> I'm not sure detail, but, I guess R-Car H1 DT detection needs
> simon's latest branch.

    I noted in the cover letter that the patches are against the 'next' branch.

> If you are using old branch, DT detection will be failed ?

    Those old/new branches sound too vague for me.

WBR, Sergei