[PATCH 1/2] ARM: pxa: ssp: provide platform_device_id for PXA3xx

[Daniel Mack]

Hi Haojian,

On 08/14/2014 04:08 AM, Haojian Zhuang wrote:
> On Thu, Aug 14, 2014 at 3:59 AM, Daniel Mack <zonque at gmail.com> wrote:
>> Provide an explicit match string for PXA3xx SSP ports.
>>
>> Signed-off-by: Daniel Mack <zonque at gmail.com>
>> ---
>>  arch/arm/plat-pxa/ssp.c | 1 +
>>  1 file changed, 1 insertion(+)
>>
>> diff --git a/arch/arm/plat-pxa/ssp.c b/arch/arm/plat-pxa/ssp.c
>> index 3ea0290..1f5ee17 100644
>> --- a/arch/arm/plat-pxa/ssp.c
>> +++ b/arch/arm/plat-pxa/ssp.c
>> @@ -258,6 +258,7 @@ static const struct platform_device_id ssp_id_table[] = {
>>         { "pxa25x-ssp",         PXA25x_SSP },
>>         { "pxa25x-nssp",        PXA25x_NSSP },
>>         { "pxa27x-ssp",         PXA27x_SSP },
>> +       { "pxa3xx-ssp",         PXA3xx_SSP },
>>         { "pxa168-ssp",         PXA168_SSP },
>>         { "pxa910-ssp",         PXA910_SSP },
>>         { },
>> --

> I don't agree on this. SSP1,SSP2,SSP3 on pxa3xx are same with pxa27x.
> So we reuse the name.
> For the new SSP4, we create pxa3xx_device_ssp4.

Well, the problem is that the code in sound/soc/pxa/pxa-ssp.c is full of
checks like "ssp->type == PXA3xx_SSP", but unless the devices are
instantiated via DT, this value is never set appropriately, not even for
ssp4.

That's a regression caused by 972a55b62 ("ASoC: fix pxa-ssp compiling
issue under mach-mmp"), btw, that added the PXA3xx_SSP enum value but
forgot to make sure it is actually used.

Hence, this commit could even get a stable tag for 3.5+.


Thanks,
Daniel