Re: New hardware support - ITE IT8689

Re: New hardware support - ITE IT8689

[Frank Crawford]

Pozdrawiam,

I've only just joined this list, so didn't see your request previously,
but contact me about monitoring an IT8689, as I have that in the set I'm
trying to get into the kernel, or you can test yourself if you grab it
from my git repo https://github.com/frankcrawford/it87

Note that as Guenter says, this is all from testing and guessing, and
there is the possibility that it may cause stability issues, although I
haven't seen any in my use.

Regards
Frank

Re: New hardware support - ITE IT8689

[Guenter Roeck]

Hi,

On 11/13/21 11:46 AM, Paweł Metelski wrote:
> Hi,
> 
> I've got this new hardware, Gigabyte Z690 UD DDR5.
> Apparently it employs a new SuperIO monitoring chip, an ITE IT8689:
> [    1.626587] gpio_it87: Unknown Chip found, Chip 8689 Revision 1
> 
> I'm running gentoo-sources-5.14.17 but it also seems missing on upstream git.
> 
> I could probably test your patches quite easily (I suppose I just need
> to build a module and load it, see dmesg output, try to spin up a fan,
> see if hwmon reports it, etc.), I could also try to patch it myself
> given a domain introduction (8y exp C dev, 0y exp kernel C dev).
> 

Gigabyte does not support Linux and does not make datasheets available,
and ITE usually does not even admit the existence of specific Super-IO
chips. On top of that, ITE Super-IO chips are somewhat notorious for issues
which can, if programmed the wrong way, result in random system resets.
This means the it87 driver in the Linux kernel is for all practical purposes
unmaintained, and there is pretty much zero chance that it will be possible
to add support for this chip to the driver.

Guenter

New hardware support - ITE IT8689

[Paweł Metelski]

Hi,

I've got this new hardware, Gigabyte Z690 UD DDR5.
Apparently it employs a new SuperIO monitoring chip, an ITE IT8689:
[    1.626587] gpio_it87: Unknown Chip found, Chip 8689 Revision 1

I'm running gentoo-sources-5.14.17 but it also seems missing on upstream git.

I could probably test your patches quite easily (I suppose I just need
to build a module and load it, see dmesg output, try to spin up a fan,
see if hwmon reports it, etc.), I could also try to patch it myself
given a domain introduction (8y exp C dev, 0y exp kernel C dev).

-- 
Pozdrawiam,
Paweł Metelski