Dr. Hurt

I’m having issues loading the kexts. If I have to guess, I’d say the injector kext is not loading the vanilla ar40 kext. The fixup kext seems to load. (Clover, SLE give same result).

 

AR9485, 10.13.5, lilu 1.2.7, boot flag set correctly.

 

It only worked once for me, died after reboot. Says no hardware installed, ioreg shows no kext loaded for airport.

 

Hope you could update the kexts to address this issue.

There are commits to Linux upstream that seem to address such issues. I’m not sure when/if I’ll have time to update the driver.

FWIW, I've submitted a fix to Clover (commit 4194).

 

Details:             #71            

@Dr. Hurt, can you update the source on github with your latest changes in v6? I noticed your Release-6-BETA branch is a little outdated and has some issues with sleep. I'd appreciate the source as I'm currently testing some changes for optimizing the touch input for clickpads (with the button under the trackpad)

 

Of course!! I'm sorry for not updating it in so long. I have a ton of studying

 

If you don't mind, I'll share my code here for now till I have time to sync/update on github.

 

This code has the fix for interrupt (As suggested by the Linux devs). This fixed most sleep issues.

sourceR6.zip

I still have problem with sleep. Work perfectly but touchpad stops responding after waking from sleep, keyboard works normally before and after sleep.

My hw is Dell ALPS touchpad v8, on Dell precision M7510.

 

That contradicts what you said in your last post. Which version are you using? 

 

Also, it would be helpful if you check log after sleep to see where its failing.

The sleep problem on some Latitudes is caused by a bad BIOS/EC. (According to Linux ALPS developers)

 

R6RC2:

- Ignore invalid packets (instead of buffering). FIXED SLEEP ON V7 LATITUDE CONFIRMED BY "bigahega".

 

(It should fix sleep on V8, V3 Rushmore as well.)

Known issues:

- V7 and V8 tracksticks disabled. (till someone fixes it or helps me fix it)

- V6 not supported.

- As usual, most settings in prefpane won't work. They're not made for ALPS.

 

R6RC2.zip

Well, I had a good look at the log and first thing I noticed is that x and y sometimes have unrealistic values. They cannot be >127 so I filtered those out.

 

3:

- Filter out any packet reporting x or y >127 (not sure if I should filter or just set to zero so let's experiment)

- remove V8 valid packet check from interrupt (yet another attempt at sleep).

- disable trackstick buttons for now.

=========

 

I wish I could see how x and y values change in response to trackstick direction.

You could use "Jake Lo 2" build and try to figure out a pattern. Things like x bigger or smaller than y, positive or negative, in relation to direction change.

 

Good Luck

Jake Lo 3.zip

Could you post a sample from the report in log? I wanna see if there are any wrong-looking packets that need filtering.

Also, did you have sleep problems with other systems too? V3 and V7? Does this version fix it?

@others:
Anyone else still having problems with sleep? (With jake lo 2 build)
Did it fix or break sleep for anyone?

@Jake Lo:
I made a mistake in last test build. Sry, try new attachment instead.
========

In fact, EVERYONE ELSE should use the "Jake Lo 2" build instead of R6RC1 because it actually fixes a very major mistake. (Affects stability and sleep). And PLEASE REPORT if you try it.

I am so sorry. I supposed you want to check the installation, the last ioreg is in normally using by r5 kext. This is in using R6c1. And I tried to check the s/l/e and l/e, then using the kext utility. But not found the hw, too. Check again, thanks.Kitkim’s MacBook Pro(R6c1.zip

 

According to this, everything is working correctly.

The driver is correctly announcing multitouch prefpane compatibility. (screenshot)

 

You may have an problem with your install (corruption or something) or maybe it's another issue which I don't know.

 

Edit: You may wanna try deleting settings as per this guide before considering further steps...

http://osxdaily.com/2014/08/01/mac-mouse-trackpad-weird-behavior-troubleshooting/

Tested R6RC1 on

E7470: V8  ID: E7=0x73 0x03 0x28, EC=0x73 0x01 0x13

 

-Keyboard / touchpad works; trackstick response but goes straight to lower right corner and stays there

-The right mouse buttons for both touchpad / trackstick don't work. 2 finger right click is hard to control / works but need a few tries to get the right speed and pressure

-After sleep keyboard continues working but touchpad/trackstick/ left/right mouse button stops working

-Had to use external USB mouse at this point

 

Dmesg shows "HID Activity Tickle ...."

 

If you want, here's another build which adds:

- Debug output for Trackstick.

- Minor changes to button handling.

- Reset buffer after sleep

 

When macOS starts, do NOT touch trackstick, then open console.

 

Move trackstick without touching any buttons and check console for ALPS: Trackstick report:

 

The point of this is to identify any incorrect packets that need to be filtered like with Rushmore.

Also check sleep again.

 

EDIT: Sry, I made a mistake with last version. Hope you haven't tested it yet. Use 2 instead.

Jake Lo.zip

Jake Lo 2.zip

Thanks. This is my iroeg.Kitkim’s MacBook Pro.zip

 

The ioreg shows that Release 5 is loaded!!!!

Something is wrong with your kext installation. Maybe duplicates, or you're confusing old and new kexts.

 

Delete any voodoops2 kext from:

/System/Library/Extensions

/Library/Extensions

In clover folder: /kexts/10.12 and /kexts/other

 

After making sure everything is deleted, empty trash and run kext utility 2.6.6 app (search google) to rebuild caches. 

WAIT till everything finishes. 

THEN: put R6RC1 kext in clover's "/kexts/10.12" folder.

Dr.hurt. I do not make trouble for you. But my preference can't detect the hw.

I'm not ignoring you.

I just don't have a fix because I don't know why that's happening to you. You're the only person with this issue. I think you have a problem with your mac install.

 

Upload an ioreg dump.

 

After lots of help from Tuurbo, I’m happy to announce that the Latitude sleep problem is probably fixed for good using Linux 4.4 code to maintain V7 and V8 compatibility.

At this point, it’s safe to assume that the driver is stable enough to be labelled as Release Candidate.

R6RC1:
- Patch Linux 4.4 Interrupt to fix sleep on Dell Latitude
- Reject bare PS2 packets (from external port)
- Fixed driver reset during startup (as you may have noticed in log, it was starting twice)
- Update V8 trackstick code from upstream.
- Enable V8 trackstick (to see if upstream changes make it any better)
- Separate dispatch code from alps to make things cleaner to work with
- More cleanups
- Disable all build optimizations

Known issues:
- Most preference pane settings do not work.
- V7 trackstick is disabled for now
- V8 trackstick is probably buggy
- V1 and V2 probably need some fixing

What to test:
As usual, test general performance and stability.
People with sleep issues should test again.
V8 trackstick users, please report if it works.

R6RC1.zip

R6B6 uses interrupt handling from Linux 3.13 which breaks support for V8 and causes bad performance for V7.

 

The sleep problem seems to be mostly with Dell Latitude laptops.

 

People with sleep problem, could you please try R6B6?!!

 

@Tuurbo:

Did you try both B10 and B10fixed? If not try the other and tell me if it makes a difference.

Also keep console open before sleep, and on wake check for ALPS messages. Perhaps it could give a clue about what's happening. Or is the driver not starting? (i.e. No messages after sleep)

Hi, Dr.hurt. Merry Christmas! I test this R6B10kext. It worked perfect after sleep and slide smoothly. And the multi-finger slide performed perfect. But the preference of the trackpad shows NO TRACKPAD FOUND, whichever I use the old pane or the new pane.

What hw version and OS X version?

 

I have reports from both sierra and el cap that pref pane works!!

 

Make sure you don't have conflicting kext or pref panes installed.

 

Merry Christmas everyone

Hello,

 

I installed Sierra, and my trackpad is not working (very slow, no multitouch...) with basic Voodoo kext by rehabman.

 

I have the ALPS V8, with this ID : E7=0x73 0x03 0x50, EC=0x73 0x02 0x02.

 

I used this kext (http://www.osx86.net/files/file/3998-voodoops2controller-acer-aspire-v3-771/) on Yosemite and it worked great. But on Sierra, I have a kernel panic.

 

Can I merge something ? How to upgrade the kext ?

 

Thanks

Have you tried R6 kext from here? It's supposed to work with V8.

With the help Piranha, we got Caps Lock working on older OS X and fixed sleep issue.

 

R6B10:

- Sample Rate to 40 packets/sec on V7 and V3 Rushmore

- TouchPad works AFTER sleep (as tested by Piranha)

- Caps Lock works on Sierra AND older OS X (as tested by Piranha)

- Fix accidental touch while typing

 

Known Issues: 

- As usual, most settings in preference panes will not work.

- Scroll will be very fast on OS X versions before Sierra. Edit plist to fix

- V1 and V2 are not working.

- V7 and V8 tracksticks are disabled for now

 

What to test?

General usability, performance, sleep...

 

EDIT:

R10B10fixed fixes a typo that caused pointer lag

R6B10.zip

R6B10fixed.zip

I made the change and although the immediate amount of scrolling on a two-finger tap is reduced, the probability of it registering as a two-finger tap is still the same about 1/20 attempts.

 

I think the issue here has more to do with the polling interval than the sensitivity of calculated scroll value.  I feel like there should be a brief time period (maybe 100ms) after a tap is made, that no scrolling is interpreted.

 

You just gave me an idea about a possible reason for why things are a bit different between my V3 and your V7

 

My V3 Pinnacle is set to a sample rate of 40 while V3 Rushmore and V7 are set to 100.

 

R6B9:

- Set Sample Rate to 40 packets/sec on V7 and V3 Rushmore (not sure if HW will actually accept that command).

- Hopefully fix mistake from last release.

- Fix accidental input while typing. 

 

As always, test sleep and general performance.

R6B9.zip

R6B8:

- Caps Lock fix only applies to Sierra (because it breaks older os x)  

- Very strict memory handling during touchpad initialization. (yet another shot at the sleep problem)

- Keyboard start before Touchpad again.

- Updated build settings

 

I honestly no longer know what to do about the sleep problem.

I'll just keep optimizing the driver and hope for the best.

 

Known Issues:

- Not all pref pane settings work. Never will...

- V7 and V8 Tracksticks are intentionally disabled for now.

- Scrolling is too fast on macOS versions before Sierra!!

R6B8.zip

Try with beta 6 too. To compare the effect of changing interrupt.

 

Another thing:

Before putting system to sleep, open console app (leave it open) and look for messages related to touchpad init as soon as system wakes.

Hopefully this'll give us a clue about where its failing. 

e5550 on 10.11 El Capitan

ALPS: Found a V7 TouchPad with ID: E7=0x73 0x03 0x0a, EC=0x88 0xba 0x2a

 

One issue I've been having with all of the R6 drivers is the two-finger scrolling sensitivity being too high.  I've set the scroll sensitivity in Mouse and Trackpad to the lowest setting, but almost every two finger tap results in a scroll.  This makes it almost impossible to perform a two finger tap to right-click since any detection of a scroll ignores the two finger tap as a right-click.  Is anyone else having this problem?

 

open plist in the trackpad kext (in plugin folder) and increase MultiFingerVericalDivisor and MultiFingerHorizontalDivisor to 13. I think I might have to create a separate profile for Sierra and older OS X since they're so different.

 

Question: Does touchpad work after sleep?

 

V7 on my E5450 sleep NOT work 

 

Did sleep work with beta 6?

The last version is working surprising well for me too

 

The Sierra Caps Lock fix breaks Caps Lock on older OS X. Will fix that in the final release.

 

As for mission control, go to "Preferences > keyboard > shortcuts > mission control" and reassign the shortcut (double click and perform gesture).

 

PS. I want as much feedback as I could get to move towards a final Release 6...

R6B7:

- Updated Interrupt to Linux again, added a hack to prevent desync. (Let's hope this fixes sleep craziness)

- Changed Keyboard type from Apple to Unknown. (for Bronxteck)

- Cleaned up base driver, moved essential settings to ALPS driver instead.

- V8 should work again (since it needs the new Interrupt).

 

This release should be quite stable as we approach final Release 6.

 

Test:

- Sleep

- V3 Rushmore users: sleep, lid closing and opening

- General usability and stability

R6B7.zip

I think I found out what was crashing the driver after sleep.

 

The ringBuffer Tail is not advanced after a bare 3-byte ps2 packet is sent which causes the protocol to become desynchronised and results in garbage data being sent to decode function!! At least thats what I think is happening!!

 

The strange thing is: where are the bare packets coming from? Does anyone here have external PS2 ports in their laptops? 

 

I'll try to fix this in a release later tonight.

 

Have a nice day everyone