Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
  • Jul 18 15:33
    @kprinssu banned @whatnameisit
  • Sep 08 2021 12:53
    @kprinssu banned @sqlsec
  • Feb 01 2021 20:01
    @kprinssu banned @kdmitrienko
  • Nov 18 2020 13:40
    @kprinssu banned @PredatorAlpha777
  • Nov 15 2020 16:16
    @kprinssu banned @nikhil19rathore99
blankmac
@blankmac
Oh, is it an I2C trackpad?
Ben Raz
@ben9923
Yeah
blankmac
@blankmac
Yeah, that's a no go unfortunately. I worked with somebody on here for quite awhile trying to get I2C Alps to work without results. For whatever reason, it's not able to put the trackpad into precision mode from the HID interface. Probably needs to be done directly with an I2C bus command like the Elan.
Which would require writing a new kext.
Though I will say, he wasn't getting that particular error, just no input. The kext would show attached in the ioreg for him.
Ben Raz
@ben9923

@blankmac Ugh, that's a bummer.
Seems like Linux aren't doing anything special there, looks similar to your code.

Noticed you're not checking for BTNLESS here https://github.com/blankmac/AlpsT4USB/blob/410e2daac0aa56d0504f333ae710e9194b833e74/AlpsT4USB/AlpsT4USB.cpp#L137

BTW his IOReg shows he does have another I2C HID device :P
blankmac
@blankmac
Hmm, ok, when I looked at his IOReg, all I saw were nubs under his I2C devices. His trackpad is attaching under USB tho, so I'm not sure what the problem is.
Ben Raz
@ben9923
Oh the lack if BTNLESS will make it fail to start, probably not it...
But anyway this guy's device is BTNLESS, might be worth to see how it goes if it actually loads
Just for the sake of testing we can add kIOHIDVendorIDKey and see if the error message has changed
Can be done in Info.plist rather than code too
blankmac
@blankmac
His device id is 0x120d which is being checked for and should be supported. It's the same as the one on the Elite G1 that @kprinssu was using for a bit.
Ben Raz
@ben9923
image.png
Am I looking at the wrong IOReg? :/
blankmac
@blankmac
Either that or I am! lol
Ben Raz
@ben9923
lol
blankmac
@blankmac
Yeah, it was me, lol.
Ben Raz
@ben9923
Haha
blankmac
@blankmac
Ok, as written, it isn't trying to initialize his device at all it looks like.
I'll get his device ID added properly and it will probably attach but I doubt it will work.
Although the other guy had a U1 based device so who knows, maybe it will.
Ben Raz
@ben9923
Hopefully someone with this device in hand can make it work sometime
Though I didn't lose all my hope :P
blankmac
@blankmac
Yeah, I'm not sure what the disconnect is really. You're right that the Linux code isn't really any different so I'm not sure why it wouldn't get into precision mode.
Ben Raz
@ben9923
Let's hope this one does, I guess
Ben Raz
@ben9923
@blankmac Kext seems to panic on his machine. Do you have the debug symbols?
blankmac
@blankmac
I do but I'm not at home at the moment.
Ben Raz
@ben9923
I asked him for another log with keepsyms=1 in the meantime
blankmac
@blankmac
If he has Xcode installed he could just build a release version real quick.
Ben Raz
@ben9923
He's not available for now anyway :P
CoolStar
@coolstar
@ben9923 @blankmac I can confirm btw that VoodooGPIO 2.4.0 works perfectly here, whereas even 2.4.3 was having issues with sleep/wake breaking the touchpad after some time
Ben Raz
@ben9923
@coolstar I have a fix for the cannot be used as IRQ after sleep issue
Not sure if it's the one you're referring to tho
isi95010
@isi95010
Hi all. I mentioned in the main chat that I'd like to get the Cypress running on my C720. I feel honored to be added to this chat but I'll mention that as tech savvy of an end-user as I am, I have close to 0 coding skills. As Ben suggested I plan to use the Elan satellite as an example, and somehow port the device from 1.0.4 of voodooi2c. The plan makes sense in my mind, but I don't want to bother anyone to walk me through things - perhaps if you can suggest some online learnings I could check out to put me in a position to do this effectively, I would appreciate it. On the other hand, if anyone does have time to do it, I am available to test.
Ben Raz
@ben9923

Read the IOKit reference as a start, make sure you're familiar with C/C++
Then read the old Cyapa code, make sure you understand it
Then read the VoodooI2CELAN code and understand how a satellite works (Might want to read device nub code either?)
You might want to look at the old ELAN code for reference...

Finally, port the code. Looking at the ELAN diff should give you a better idea on how it should work.

isi95010
@isi95010
Thanks!
Ben Raz
@ben9923
Sure :)
blankmac
@blankmac
@coolstar no issues here with any version.
granted, my touch pad is USB tho.
@isi95010 Don't be discouraged or overwhelmed. I started at ground zero myself in the code department with the VoodooI2C project a few years ago.
krycho74
@krycho74
@blankmac Hi, are you developing your alps touchpad project? Would you have some time now and would like to experiment with my touchpad?
blankmac
@blankmac
@krycho74 All I did was update it to comply with the changes in VoodooInput. I'm really not too interested in trying to develop for hardware I don't have on hand, sorry. Hackintoshing, as a whole, is reaching the end of the road with Apple's move back to proprietary chips.
blankmac
@blankmac
Hey guys, seems the old methods of live logging to the console from the kernel don't work anymore (ie - IOLog). I'm trying to chase down this problem and would like to output some of the input data live -- blankmac/AlpsT4USB#7
Any advice?
Avery Black
@1Revenger1
I was having a similar issue with VoodooRMI - I ended up adding a prefix to all the logging data using a pre-processor macro so that all my debug statements began with "VRMI - " and then used sudo dmesg
you generally will want to use the boot-arg msgbuf=1548576 as well to increase the max size of the dmesg buffer. That's the best way I've found, though I would like iolog to work again ;-;
oops, msgbuf should be 1048576, not the number I gave before
Ben Raz
@ben9923
@blankmac I’ve been asking users to attach dmesg logs for a while now
An alternative to the boot arg should be injecting DebugEnhancer.kext. IIRC it allows an even bigger buffer for logs (it’s patching the kernel after all).
We should add it to our docs.
zhen-zen
@zhen-zen

Any advice?

Honestly, life could be much easier debugging with Catalina. Specifically the live monitor part. The dmesg don't support -w and the watch command is still unsatisfying.

blankmac
@blankmac
Ok, that was enough to get me the info I needed. Thanks guys.
Leonard Church
@343iChurch
Hiya. So I have a unique scenario in which my laptop has a PS/2 keyboard but an I2C touchpad. I cannot get them both working consistently due to conflicts between the kexts -- when the touchpad works, the keyboard is disabled until a wake/sleep cycle. Is there a way to properly use both VoodooI2C and VoodooPS2?