Re: Nokia N900 sound driver and ECI GPIOs

["Pali Rohár" <pali.rohar@gmail.com>]

[-- Attachment #1: Type: Text/Plain, Size: 2445 bytes --]

On Monday 02 January 2017 15:36:38 Peter Ujfalusi wrote:
> On 01/02/2017 02:58 PM, Pali Rohár wrote:
> > On Monday 02 January 2017 14:53:44 Peter Ujfalusi wrote:
> >> On 01/02/2017 10:51 AM, Pali Rohár wrote:
> >>> On Monday 02 January 2017 10:34:34 Peter Ujfalusi wrote:
> >>>> On 01/02/2017 12:36 AM, Pavel Machek wrote:
> >>>>> Are there multi-button headsets compatible with N900?
> >>>> 
> >>>> Not sure if n900 is capable of handling ECI communication, n9
> >>>> was. In n9 the PMIC/Audio chip had an ECI core integrated afaik
> >>> 
> >>> In N9 it is handled by twl5030 chip? Is there documentation for
> >>> that ECI/ACI part? N900 has twl4030 so we need some chip
> >>> documentation for adding support for N900...
> >> 
> >> n9 has twl5031 and the documentation is not public. But the kernel
> >> source from Nokia have the driver under drivers/mfd/twl5031-aci.c
> >> It is using the ECI stack to report the changes.
> >> twl4030 and twl5030 does not have support for ECI, only twl5031.
> > 
> > Ok, so Nokia N900's audio chip (twl4030) does not have support for
> > ECI.
> 
> TLV320AIC34 (two of them) on n900 to be precise, n9 uses twl5031 and
> tlv320dac33.
> 
> > But still as Jarkko wrote there could be an option to supports ECI
> > via GPIO and ADC.
> 
> Yes, there is the option, but w/o documentation on the ECI protocol
> it is not going to be easy.

Is there documentation?

> And I'm sure there were a (good) reason the n900 does not support ECI.

What was that (good) reason? I would like to know it.

> If it could be done in a stable way the product would have support for it.

Hm... I would rather not to speak about decisions of product managers :-)
 
> > And there is still mysterious gpio 178 (RX51_ECI_SWITCH_1_GPIO)
> > which could help us?
> 
> Yeah, that is something I have no idea what it is for. It could be
> that the schema is using different pin mode for it? It might worth
> looking at the TRM on which pin the gpio_178 can go out and look for
> the possible modes? Most likely waste of time...

According to omap trm GPIO_178 is connected to mcspi2_clk. But
mcspi2_clk is missing in RX51 schematic too.

To verify that omap trm is correct: GPIO_177 (in omap trm) is connected
to mcspi1_cs3. And in RX51 schematic HEADPH_IND (gpio 177) is connected
to mcspi1_cs3. So looks like omap trm is correct there.

-- 
Pali Rohár
pali.rohar@gmail.com

[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 198 bytes --]