Voodooi2c: Problem pointer - ice lake notebook - i2c hid satellite

[Lorys89]

Describe the bug
A clear and concise description of what the bug is.
-Hello everyone and sorry for the inconvenience, but I have a problem with the touch pad of my laptop, it is of the ice lake generation, the gestures work, system preferences sees it well, but with the use of 1 finger, the pointer responds badly, in the video attachment you can see, moving vertically the pointer responds badly, and moving the pointer horizontally with the slightest movement goes very fast horizontally on the screen

Did you read the common errors documentation?
Please answer yes or no after reading https://voodooi2c.github.io/#Common%20Errors/Common%20Errors

• YES
  Did you read the troubleshooting documentation?
  Please answer yes or no after reading https://voodooi2c.github.io/#Troubleshooting/Troubleshooting.
  -YES

Have you searched the issue on Github, Gitter, or Google?
Please answer yes or no, this is to avoid creating duplicate Github issues.
-YES

System Environment
Please answer:

1. What is your operating system? - MACOS BIG SUR 11.2
2. What version of the drivers (including drivers) are you using? - VOODOOI2C AND VOODOOI2C HID 2.6.3
3. What is your hardware configuration? - DELL VOSTRO 5401 (ICE LAKE)
   -CPU : I7 1065G7
   -IGPU: INTEL IRIS PLUS G7
   -1° NVME FOR MACOS: SAMSUNG 970 EVO PLUS 500GB
   -2° NVME FOR WINDOWS 10: HYNIX BC511 512GB
   -RAM: KINGSTON HYPERX IMPACT DDR4 3200 32GB (2X16GB)
   -WIFI+BT: BROADCOM DW1830
   MY EFI ON GITHUB --> https://github.com/Lorys89/DELL_VOSTRO_5401-ICE-LAKE
   Troubleshooting Archive
   Please attach the troubleshooting archive (as per https://voodooi2c.github.io/#Troubleshooting/Troubleshooting) and any additional logs or videos demonstrating the problem. Please attach any extra files to Github and not anywhere else like Google Drive.
   LOG -->
   Log.log

IOREG -->
MacBook Pro 16,2 ice lake DELL.zip

IOREG WITH SSDT-I2CD -->
MacBook Pro 16,2 with I2CD.zip

VIDEO OF THE PROBLEM -->
problem i2c.mp4.zip

DSDT ORIGIN -->
DSDT.aml.zip

Additional context
Add any other context about the problem here.

[Lorys89]

@ben9923 @zhen-zen is this material okay? or do you want me to add more?

last log --->
Log.log

with ssdt i2c1 + log --->
SSDT-I2C1.zip
Log2.log

RunMeApp-OcDebug -->
debug_3834.zip

[Lorys89]

for me it's not a problem at acpi level but it will be kext/driver

info for my touch pad resolution --->

and the value of DELL XPS 9300 (ICE LAKE) touch pad (in this XPS icl work good i2c)--->

and the value of DELL INSPIRON 5593 (ICE LAKE) touch pad (in this INSPIRON icl work good i2c)--->

[Vorshim92]

Hello everyone! Here with the same identical problem on my HP Pavilion x360 Icelake! everything works, also the Touchscreen panel! Just the vertical issue on the trackpad! here my ioreg and rep: IOREG

[Lorys89]

Hello everyone! Here with the same identical problem on my HP Pavilion x360 Icelake! everything works, also the Touchscreen panel! Just the vertical issue on the trackpad! here my ioreg and rep: IOREG

we probably have the same i2c elan

[Lorys89]

@kprinssu seeing the info and files provided, what do you think?

[Lorys89]

@ben9923 WP i2c hid , see video. https://github.com/Lorys89/DELL_VOSTRO_5401-ICE-LAKE/raw/main/Screenshot/video_2021-01-18_23-48-22.mp4.zip

MacBook Pro di Lorys.zip

my friend @mbarbierato worked on kext i2c hid and the results are excellent, the speed must only be optimized, now I can already use it discreetly in my touch pad.

[Lorys89]

@ben9923 this pr https://github.com/VoodooI2C/VoodooI2CHID/pull/47 fix my problem :)
(my touch pad with dpi x and dpi y elevated, and logic and phisical differ a lot, it needs a new calculation, I have all the study material.)
i2cLog.log
ioreg MacBook Pro 16,2.zip
see video https://github.com/Lorys89/DELL_VOSTRO_5401-ICE-LAKE/raw/main/Screenshot/i2c%20hid%20work%20good.mp4.zip

[ben9923]

Problem lays in VoodooI2CHID's unit exponent factoring, not being able to handle anything lower than -2.
We probably also want to eliminate the scaling by x10 in the native engine and just let the satellites report the proper, already-scaled dimensions.

Reported physical X/Y in @Lorys89's device are 0x1bee0/0x10f18 (114.4cm/69.4cm), 10 times bigger than actual value...
As the Magic Trackpad 2 only reports 16 bit dimensions, actual reported values were 0xbee0/0x0f18 (48.9cm/3.9cm).
Such values obviously messed up with the input processing.

Fixed in https://github.com/VoodooI2C/VoodooI2CHID/commit/1674b569a863ba789982cc1d86f2c80c05d55c74 along with a change in VoodooI2C to not scale the value by x10.

Note: VoodooInput expects physical dimensions in 0.01 mm units.

Thanks @mbarbierato for the lead to mess with physical dimensions.
CC @kprinssu

[kprinssu]

@ben9923 the changes looks good. Feel free to do a new release!

[Lorys89]

thanks

[mbarbierato]

Thanks