package/linux-mainline: Bump to 6.2

[alistair23]

Bump the mainline based kernel to the 6.1 6.2 release.

Besides the mainline changes from 6.0 to 6.2, this includes a refactoring of the rohm,bd71815 code to use the mainline driver instead of the reMarkable version. It has also dropped a Koreader related hack from the kernel, see koreader/koreader#10012.

This is the 6.1 6.2 mainline kernel and 18 12 rM2 specific patches on top.

[Etn40ff]

I just ran this on 20221026104022. There are a couple of issues:

• a ⚠️ symbol appears next to the battery indicator in xochitl

• I get the following messages in dmesg

[   15.287260] UDC core: g_ether: couldn't find an available UDC or it's busy
[  231.122624] evbug: Event. Dev: input1, Type: 1, Code: 330, Value: 0
[  231.122635] evbug: Event. Dev: input1, Type: 0, Code: 0, Value: 0
[  234.253718] usbmisc_imx 30b20200.usbmisc: vbus is error
[  234.253764] usbmisc_imx 30b20200.usbmisc: Error occurs during detection: -22

actually dmesg output is spammed with evbugs messages.

Other than that all seems to be in order.

[alistair23]

Weird. I don't see those.

dmesg | grep g_ether    
[    3.610858] UDC core: g_ether: couldn't find an available UDC
[    4.899438] g_ether gadget.0: Ethernet Gadget, version: Memorial Day 2008
[    4.906347] g_ether gadget.0: g_ether ready

After a reboot do you still see the battery warning sign?

[Etn40ff]

I do but the message buffer is not spammed by them and ethernet over usb works:

reMarkable: ~/ dmesg | grep g_ether
[    2.620511] UDC core: g_ether: couldn't find an available UDC
[    3.493576] g_ether gadget.0: Ethernet Gadget, version: Memorial Day 2008
[    3.493599] g_ether gadget.0: g_ether ready
[   13.885486] UDC core: g_ether: couldn't find an available UDC or it's busy

Here are some pictures of the warning signs:

[alistair23]

Ok! Pushed an update to go all the way to the 6.2 kernel.

I double checked and in Xochitl 2.9.0.210 I don't see any warning signs for the battery. Let me know if you still see anything, although I think that's a Xochitl bug as we are now using the upstream driver for the power management

[Etn40ff]

I just gave this another spin. I still get the weird error symbol in 2.15.1.1189 but other than that it works as intended.

[alistair23]

I think this should be merged then. We can work on the warning symbol in the future

[alistair23]

Ping!

[alistair23]

Ping!

[Eeems]

I'd be happy to merge it if one of the maintainers with a rM2 is okay with that. @raisjn @matteodelabre @LinusCDE ?

[alistair23]

Anyone?

[alistair23]

Another ping

[alistair23]

Ping!

[Eeems]

I've asked the question a number of times on discord and had no reply about it, other than @LinusCDE saying he didn't think he was the right person to answer.

[Eeems]

@LinusCDE I think it would be fine for you to approve this at this point.

[LinusCDE]

Will rubber stamp, but have not done testing on custom kernels yet. If some unknown issue crops up, i can probably find the time to look into it.

[Eeems]

Will rubber stamp, but have not done testing on custom kernels yet. If some unknown issue crops up, i can probably find the time to look into it.

To clarify, I'm not looking for a rubber stamp. I'm looking for an explicit statement that you are comfortable or not with the battery warning when running this on 2.15.

[LinusCDE]

Was initially not sure, because it involves battery seemingly. I doubt that the kernel would need to do any crucial control for it, but on the rM 2 it's usually hard to say what they did in the name of making stuff sleeker and more responsive.

In this case I doubt it's an issue though. Looking at the dmesg log, the problem doesn't seem to be battery but actually the UDC, which according to this documenation should mean USB Device Controller. At first I thought this should not matter to the reMarkable anyway.

But in one way I would: The ethernet interface. Can someone confirm that the Ethernet functionality still works normally? If not I should have probably not approved it. Otherwise I think it should be fine.

[Eeems]

@Etn40ff since you are the one who can replicate, can you take a look?

[Etn40ff]

@Etn40ff since you are the one who can replicate, can you take a look?

I do not have my recovery gear with me at the moment; I will not be able to test this again before mid September.

[alistair23]

I use the Ethernet interface fairly often. I'm pretty sure it works fine with the 6.2 kernel

[LinusCDE]

Then there should be no issue merging this i guess. 👍

[Eeems]

Alright, good enough for me, I guess. Merging.