so, i got a 6.1 kernel running, now wheres my keyboard?

[leandro friedrich]

Aight so first, hello

I recently found my old wm8650 netbook and had some fun with it.

i thought i might give getting mainline kernel to run a try, and it worked! currently running debian 10 on it and over ssh or with external keyboard its pretty decent actually

now heres my issue: i cant get the keyboard to work. iirc its connected over usb, so whats the issue? i read something here about it but couldnt figure out what was meant, so if anyone here has some ideas, id love to hear them

greetings from germany lol

[Alexey Charkov]

AFAIR the WM8650 netbooks had their keyboards connected to an internal
UHCI controller, separate from the EHCI one that all external ports
are driven from. Have you enabled the platform UHCI driver in kernel
config?

Best,
Alexey

[leandro friedrich]

So, from what i saw it should be enabled, but im getting 
"[    3.310000] platform-uhci: probe of d8007b00.uhci failed with error -2"

in dmesg, any idea what went wrong? EHCI seems to work given how external KB and usb etc. work

[Alexey Charkov]

On Thu, Nov 10, 2022 at 11:19 PM leandro friedrich
<leandro.fr...@gmail.com> wrote:
>
> So, from what i saw it should be enabled, but im getting
> "[ 3.310000] platform-uhci: probe of d8007b00.uhci failed with error -2"
> in dmesg, any idea what went wrong? EHCI seems to work given how external KB and usb etc. work

Well, -2 is -ENOENT. Looks like it's missing some resource required
for probing the device. If you can boot the device, maybe check what
device tree the kernel is seeing?

Like for example:

dtc -I fs -O dts /sys/firmware/devicetree/base

Best,
Alexey

[leandro friedrich]

should i send the output here in a file? or as text? i took some stuff i found on here for the config and the device tree by the way

[leandro friedrich]

i put it in a file, cant hurt

[Alexey Charkov]

On Fri, Nov 11, 2022 at 12:19 AM leandro friedrich
<leandro.fr...@gmail.com> wrote:
>
> i put it in a file, cant hurt
>
> leandro friedrich schrieb am Donnerstag, 10. November 2022 um 21:12:18 UTC+1:
>>
>> should i send the output here in a file? or as text? i took some stuff i found on here for the config and the device tree by the way

Normally inline plain text is easier to review and comment over email,
and also saves the trouble of downloading the file and opening it in
an editor, but nevermind :)

The device tree looks good, so it's weird that it cannot probe. Maybe
some module dependencies don't get sequenced quite right.

Could you please try compiling the UHCI driver as a module instead of
built-in, to make sure it loads later in the boot process? Let's see
what that gives.

Best,
Alexey

[leandro friedrich]

where do i even find the UHCI driver in the config? what would be its name?

[Alexey Charkov]

On Fri, Nov 11, 2022 at 5:39 PM leandro friedrich
<leandro.fr...@gmail.com> wrote:
>
> where do i even find the UHCI driver in the config? what would be its name?

It's CONFIG_USB_UHCI_HCD under USB Host Controller Drivers

[leandro friedrich]

first, sorry for responding this late, worked on another project of mine
so: i tried that, still no change

[Alexey Charkov]

On Tue, Nov 15, 2022 at 3:54 PM leandro friedrich
<leandro.fr...@gmail.com> wrote:
>
> first, sorry for responding this late, worked on another project of mine
> so: i tried that, still no change

Another hypothesis. Looks like the clk code that was added in 2018
might not allow for a graceful probe if no clock is provided (which is
the case for our WMT devices). What if you comment out the following
code block in drivers/usb/host/uhci-platform.c (around line 124):

/* Get and enable clock if any specified */
uhci->clk = devm_clk_get(&pdev->dev, NULL);
if (IS_ERR(uhci->clk)) {
ret = PTR_ERR(uhci->clk);
goto err_rmr;
}
ret = clk_prepare_enable(uhci->clk);
if (ret) {
dev_err(&pdev->dev, "Error couldn't enable clock (%d)\n", ret);
goto err_rmr;
}