[PATCH] fsl_spi_init: Support non-QE processors

[PATCH] fsl_spi_init: Support non-QE processors

[Peter Korsgaard]

On non-QE processors (mpc831x/mpc834x) the SPI clock is the SoC clock.

Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
---
 arch/powerpc/sysdev/fsl_soc.c |   27 +++++++++++++++++++--------
 1 files changed, 19 insertions(+), 8 deletions(-)

diff --git a/arch/powerpc/sysdev/fsl_soc.c b/arch/powerpc/sysdev/fsl_soc.c
index a57fe56..59e4188 100644
--- a/arch/powerpc/sysdev/fsl_soc.c
+++ b/arch/powerpc/sysdev/fsl_soc.c
@@ -1220,14 +1220,17 @@ int __init fsl_spi_init(struct spi_board_info *board_infos,
 {
 	struct device_node *np;
 	unsigned int i;
-	const u32 *sysclk;
+	const u32 *qe_sysclk = 0, *soc_sysclk = 0;
 
 	np = of_find_node_by_type(NULL, "qe");
-	if (!np)
-		return -ENODEV;
+	if (np)
+		qe_sysclk = of_get_property(np, "bus-frequency", NULL);
+
+	np = of_find_node_by_type(NULL, "soc");
+	if (np)
+		soc_sysclk = of_get_property(np, "bus-frequency", NULL);
 
-	sysclk = of_get_property(np, "bus-frequency", NULL);
-	if (!sysclk)
+	if (!(qe_sysclk || soc_sysclk))
 		return -ENODEV;
 
 	for (np = NULL, i = 1;
@@ -1245,16 +1248,24 @@ int __init fsl_spi_init(struct spi_board_info *board_infos,
 
 		memset(res, 0, sizeof(res));
 
-		pdata.sysclk = *sysclk;
-
 		prop = of_get_property(np, "reg", NULL);
 		if (!prop)
 			goto err;
 		pdata.bus_num = *(u32 *)prop;
 
 		prop = of_get_property(np, "mode", NULL);
-		if (prop && !strcmp(prop, "qe"))
+		if (prop && !strcmp(prop, "qe")) {
 			pdata.qe_mode = 1;
+			if (qe_sysclk)
+				pdata.sysclk = *qe_sysclk;
+			else
+				goto err;
+		} else {
+			if (soc_sysclk)
+				pdata.sysclk = *soc_sysclk;
+			else
+				goto err;
+		}
 
 		for (j = 0; j < num_board_infos; j++) {
 			if (board_infos[j].bus_num == pdata.bus_num)
-- 
1.5.3.2


-- 
Bye, Peter Korsgaard

Re: [PATCH] fsl_spi_init: Support non-QE processors

[Grant Likely]

On 10/3/07, Peter Korsgaard <jacmet@sunsite.dk> wrote:
> On non-QE processors (mpc831x/mpc834x) the SPI clock is the SoC clock.
>
> Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
> ---
>  arch/powerpc/sysdev/fsl_soc.c |   27 +++++++++++++++++++--------
>  1 files changed, 19 insertions(+), 8 deletions(-)
>
> diff --git a/arch/powerpc/sysdev/fsl_soc.c b/arch/powerpc/sysdev/fsl_soc.c
> index a57fe56..59e4188 100644
> --- a/arch/powerpc/sysdev/fsl_soc.c
> +++ b/arch/powerpc/sysdev/fsl_soc.c
> @@ -1220,14 +1220,17 @@ int __init fsl_spi_init(struct spi_board_info *board_infos,
>  {
>         struct device_node *np;
>         unsigned int i;
> -       const u32 *sysclk;
> +       const u32 *qe_sysclk = 0, *soc_sysclk = 0;
>
>         np = of_find_node_by_type(NULL, "qe");
> -       if (!np)
> -               return -ENODEV;
> +       if (np)
> +               qe_sysclk = of_get_property(np, "bus-frequency", NULL);
> +
> +       np = of_find_node_by_type(NULL, "soc");
> +       if (np)
> +               soc_sysclk = of_get_property(np, "bus-frequency", NULL);

Why not just:

        np = of_find_node_by_type(NULL, "qe");
+       if (!np)
+               np = of_find_node_by_type(NULL, "soc");
        if (!np)
                return -ENODEV;

The other changes aren't needed that way.

g.

-- 
Grant Likely, B.Sc., P.Eng.
Secret Lab Technologies Ltd.
grant.likely@secretlab.ca
(403) 399-0195

Re: [PATCH] fsl_spi_init: Support non-QE processors

[Peter Korsgaard]

>>>>> "Grant" == Grant Likely <grant.likely@secretlab.ca> writes:

Hi,

 Grant> Why not just:

 Grant>         np = of_find_node_by_type(NULL, "qe");
 Grant> +       if (!np)
 Grant> +               np = of_find_node_by_type(NULL, "soc");
 Grant>         if (!np)
 Grant>                 return -ENODEV;

My first iteration did it like that, but then you don't get a -ENODEV
if the node is missing (and you'll end up using the wrong clock) and
it doesn't support processors with SPI on and off QE (if that
exists/will ever exist).

-- 
Bye, Peter Korsgaard

Re: [PATCH] fsl_spi_init: Support non-QE processors

[Grant Likely]

On 10/3/07, Peter Korsgaard <jacmet@sunsite.dk> wrote:
> >>>>> "Grant" == Grant Likely <grant.likely@secretlab.ca> writes:
>
> Hi,
>
>  Grant> Why not just:
>
>  Grant>         np = of_find_node_by_type(NULL, "qe");
>  Grant> +       if (!np)
>  Grant> +               np = of_find_node_by_type(NULL, "soc");
>  Grant>         if (!np)
>  Grant>                 return -ENODEV;
>
> My first iteration did it like that, but then you don't get a -ENODEV
> if the node is missing (and you'll end up using the wrong clock) and
> it doesn't support processors with SPI on and off QE (if that
> exists/will ever exist).

Okay, but you should at least be able confine your determination of
which sysclk value to use to one part of the function.  Otherwise, it
looks good.

Cheers,
g.

-- 
Grant Likely, B.Sc., P.Eng.
Secret Lab Technologies Ltd.
grant.likely@secretlab.ca
(403) 399-0195

Re: [PATCH] fsl_spi_init: Support non-QE processors

[Kumar Gala]

On Oct 3, 2007, at 1:17 PM, Grant Likely wrote:

> On 10/3/07, Peter Korsgaard <jacmet@sunsite.dk> wrote:
>>>>>>> "Grant" == Grant Likely <grant.likely@secretlab.ca> writes:
>>
>> Hi,
>>
>>  Grant> Why not just:
>>
>>  Grant>         np = of_find_node_by_type(NULL, "qe");
>>  Grant> +       if (!np)
>>  Grant> +               np = of_find_node_by_type(NULL, "soc");
>>  Grant>         if (!np)
>>  Grant>                 return -ENODEV;
>>
>> My first iteration did it like that, but then you don't get a -ENODEV
>> if the node is missing (and you'll end up using the wrong clock) and
>> it doesn't support processors with SPI on and off QE (if that
>> exists/will ever exist).
>
> Okay, but you should at least be able confine your determination of
> which sysclk value to use to one part of the function.  Otherwise, it
> looks good.

Peter, can you respin this w/Grant's modification.  I've grabbed the  
other patches and applied them.  waiting on this one.

- k

Re: [PATCH] fsl_spi_init: Support non-QE processors

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 453 bytes --]

On Wed, 03 Oct 2007 17:43:50 +0200 Peter Korsgaard <jacmet@sunsite.dk> wrote:
>
> @@ -1220,14 +1220,17 @@ int __init fsl_spi_init(struct spi_board_info *board_infos,
>  {
>  	struct device_node *np;
>  	unsigned int i;
> -	const u32 *sysclk;
> +	const u32 *qe_sysclk = 0, *soc_sysclk = 0;

Please use NULL when referring to pointers.

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: [PATCH] fsl_spi_init: Support non-QE processors

[Kumar Gala]

On Oct 3, 2007, at 11:01 PM, Stephen Rothwell wrote:

> On Wed, 03 Oct 2007 17:43:50 +0200 Peter Korsgaard  
> <jacmet@sunsite.dk> wrote:
>>
>> @@ -1220,14 +1220,17 @@ int __init fsl_spi_init(struct  
>> spi_board_info *board_infos,
>>  {
>>  	struct device_node *np;
>>  	unsigned int i;
>> -	const u32 *sysclk;
>> +	const u32 *qe_sysclk = 0, *soc_sysclk = 0;
>
> Please use NULL when referring to pointers.

Peter, any chance of getting a respin.  I'd like this to go into 2.6.24.

- k

Re: [PATCH] fsl_spi_init: Support non-QE processors

[Peter Korsgaard]

>>>>> "Kumar" == Kumar Gala <galak@kernel.crashing.org> writes:

 Kumar> On Oct 3, 2007, at 11:01 PM, Stephen Rothwell wrote:

 >> On Wed, 03 Oct 2007 17:43:50 +0200 Peter Korsgaard
 >> <jacmet@sunsite.dk> wrote:
 >>> 
 >>> @@ -1220,14 +1220,17 @@ int __init fsl_spi_init(struct
 >>> spi_board_info *board_infos,
 >>> {
 >>> struct device_node *np;
 >>> unsigned int i;
 >>> -	const u32 *sysclk;
 >>> +	const u32 *qe_sysclk = 0, *soc_sysclk = 0;
 >> 
 >> Please use NULL when referring to pointers.

 Kumar> Peter, any chance of getting a respin.  I'd like this to go
 Kumar> into 2.6.24.

Certainly. Sorry for the delay, I have been offline for 2 days
building my house ..
---
fsl_spi_init: Support non-QE processors

On non-QE processors (mpc831x/mpc834x) the SPI clock is the SoC clock.

Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
---
 arch/powerpc/sysdev/fsl_soc.c |    4 ++++
 1 files changed, 4 insertions(+), 0 deletions(-)

diff --git a/arch/powerpc/sysdev/fsl_soc.c b/arch/powerpc/sysdev/fsl_soc.c
index be5e0bd..3ace747 100644
--- a/arch/powerpc/sysdev/fsl_soc.c
+++ b/arch/powerpc/sysdev/fsl_soc.c
@@ -1222,8 +1222,12 @@ int __init fsl_spi_init(struct spi_board_info *board_infos,
 	unsigned int i;
 	const u32 *sysclk;
 
+	/* SPI controller is either clocked from QE or SoC clock */
 	np = of_find_node_by_type(NULL, "qe");
 	if (!np)
+		np = of_find_node_by_type(NULL, "soc");
+
+	if (!np)
 		return -ENODEV;
 
 	sysclk = of_get_property(np, "bus-frequency", NULL);
-- 
1.5.3.2

-- 
Bye, Peter Korsgaard

Re: [PATCH] fsl_spi_init: Support non-QE processors

[Kumar Gala]

On Oct 6, 2007, at 3:06 PM, Peter Korsgaard wrote:

>>>>>> "Kumar" == Kumar Gala <galak@kernel.crashing.org> writes:
>
>  Kumar> On Oct 3, 2007, at 11:01 PM, Stephen Rothwell wrote:
>
>>> On Wed, 03 Oct 2007 17:43:50 +0200 Peter Korsgaard
>>> <jacmet@sunsite.dk> wrote:
>>>>
>>>> @@ -1220,14 +1220,17 @@ int __init fsl_spi_init(struct
>>>> spi_board_info *board_infos,
>>>> {
>>>> struct device_node *np;
>>>> unsigned int i;
>>>> -	const u32 *sysclk;
>>>> +	const u32 *qe_sysclk = 0, *soc_sysclk = 0;
>>>
>>> Please use NULL when referring to pointers.
>
>  Kumar> Peter, any chance of getting a respin.  I'd like this to go
>  Kumar> into 2.6.24.
>
> Certainly. Sorry for the delay, I have been offline for 2 days
> building my house ..

applied.

No problem, sounds like fun.

If you get a chance can you test my for-2.6.24 board to make SPI is  
functional as you expect.

thanks

- k