[RFC 3/5] board: starfive: support Milk-V Mars board

[Heinrich Schuchardt]

On 3/8/24 04:20, Heinrich Schuchardt wrote:
> On 3/8/24 03:06, E Shattow wrote:
>> P.S. Found the longer description at 
>> https://github.com/milkv-mars/mars-buildroot-sdk/commit/d381610c92827de01b25843786012351b3f35519 <https://github.com/milkv-mars/mars-buildroot-sdk/commit/d381610c92827de01b25843786012351b3f35519> as follows: "if configured as 24c04, its address will occupy 0x51, conflicting with the RTC chip pcf85063 on the IO-Board. Refer to: Documentation/misc-devices/eeprom.rst". What does that even mean, it is for Raspberry Pi Compute IO carrier only?
> 
> According to
> https://ww1.microchip.com/downloads/en/devicedoc/doc0180.pdf
> the chips differ in how their address lines are used. Either they select 
> a chip or a memory address.
> 
> The chip is at reg 0x50. So it could be that if you read the upper 
> memory of atmel,24c04 it would use reg 0x51.
> 
> https://datasheets.raspberrypi.com/cm4io/cm4io-datasheet.pdf says the 
> CM4IO board has an RTC at address 0x51.
> 
> Together with the commit message this really sounds like the smaller 
> chip size is faked due to incompatibility.

Running i2cdetect on my Mars Milk-V board shows that a device is 
responding to address 0x51:

$ sudo i2cdetect 5
I will probe file /dev/i2c-5.
I will probe address range 0x08-0x77.
Continue? [Y/n] y
      0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f
00:
10:
20:
30: -- -- -- -- -- -- UU --
40:
50: 50 51 -- -- -- -- -- -- -- -- -- -- -- -- -- --
60:
70:

> 
> As you mentioned to have access to a CM module I would like to know what 
> its serial number looks like to understand what if it can be used to 
> detect if we are running U-Boot on full board or a CM module.
> 
> For my Milk-V board I have:
> 
> MARS-V11-2340-D008E000-xxxxxxxx
> 
> Best regards
> 
> Heinrich
> 
> 
> 
> 
> 
> 
>