Dr. Hurt

Try removing the entire ALPS profile from the plist (but make sure you don't corrupt the plist). Try only on Yosemite for now.

I build using xcode 6. Try booting with chameleon. Or disable *all* DSDT fixes in clover and revert battery related changes in the DSDT file.

Release 4: - Added a 0.1 second delay to prevent pointer movement while tapping (accidental drag) - Updated ALPS profile again - Reverts, cleanups This will be my last release for the foreseeable future. The driver is now working well enough for me and I have no time to maintain it further. Release 4.zip

R4RC3: - Added DecayingAverage+SimpleAverage together to further stabilise the pointer. - Re-added some code to possibly fix the KP - One more fix attempt before reverting 2-finger update Focus testing on pointer stability while tapping, 2-finger scroll and tap for right click. @wern: Did you try ACPIBattery v1.6? It fixed some issues I had (cycle count, current) @Jake Lo: I included the debug version in case this also KPs. I'll be gradually reverting some changes till we identify the cause. R4RC3.zip

Next Build: - Implemented decay before averaging. Pointer is MUCH smoother. - Reverted some of the 2-finger changes in my previous build. Results were just too inconsistent. @Jake Lo: I'll be releasing a debug version of the next build. Hopefully this'll tell us whats going on. Does the KP happen when you touch the touchpad? Or while the system is booting?

That device ID is probably Synaptics, not ALPS. Seeing this message means that the ALPS driver rejected the device and didn't load. Check the rest of the log or the ioreg and you'll probably find the Synaptics driver loaded. And yes, Synaptics touchpads are much better than ALPS (at least till V5). They have a much higher resolution and much more advanced firmware.

I guess we might need help from RehabMan on this one!! EDIT: @Jake Lo How's the KP issue with RC2? Anyway, here's some good news for you: I fixed 2-finger scroll for Rushmore. Now you can use any 2-fingers regardless of distance between them or contact area Will add it to the RC3. You'll absolutely LOVE it.

The driver you're using doesn't seem to be authentic. Please use the one compiled by RehabMan!!! https://bitbucket.org/RehabMan/os-x-acpi-battery-driver/downloads

R4RC2: Changes: - Re-added lock during probe (hopefully fixes the KP) - Added filter to bitmap to prevent multi- and single-touch from interfering. (2-finger scroll/tap is much better now). - Reorganise and cleanup (things are quite messy). R4RC2.zip

I added a filter to the bitmap processing to make finger number reporting more accurate. The result is more stable 2-finger tap and surprisingly, interia is much smoother!! I'll send the patch to Linux devs to review and possibly add to mainstream. @Jake Lo: Do you have any lag with inertia on Rushmore systems? Pointer jump on 2-finger tap? I wanna know if this patch is universally useful or just for regular V3. Will upload changes in R4RC2 later.

I'm using this to load the N5110 profile for the multimedia keys and breakless info: Method (_DSM, 4, NotSerialized) { If (LEqual (Arg2, Zero)) { Return (Buffer (One) { 0x03 }) } Return (Package (0x04) { "RM,oem-id", "DELL", "RM,oem-table-id", "WN09" }) }

He can only use that if he has the exact same hardware or else the system probably won't boot and the kernel could panic. PS. What do you think of R4?

I just cannot think of a reason why that is happening. Could you provide an ioreg from your system while the problem is happening? I wanna see if the Battery driver is properly loaded.

What are you using for battery status? ACPIBatteryManager.kext is the best. It may require a DSDT patch to work reliably though. Make sure you don't have any other kexts such as voodoobattery.kext I just can't think of a correlation between the two issues!!!

I just found out that I had an embarrassingly major mistake in the driver that caused it to clash with the base driver Fixed that and performance seems significantly and even tangibly improved!! This is a major change worthy of a fourth release. I'm still having trouble with momentum scroll though. Probably won't include it in R4 because its laggy and cannot be disabled via system preferences!! Update: Now it works and can be disabled from system preferences. R4RC1 coming tonight: (changelog compared to Release 3) - Fixed inertia scrolling (laggy, but can be disabled) - Reverted interrupt handler (not sure if this is what killed the touchpad, needs more testing). - Built using 10.7 SDK - Fixed interaction between Base Driver and ALPS driver!!! - Lowered resolution to 133, adjusted divisors & thresholds accordingly - Separate divisors for 2-finger scrolling and edge scrolling - Massive clean up, miscellaneous tweaks. Update: It's here. R4RC1.zip

Hmm, just did some testing and I think you're right. I was confusing smoothing with interia. Update: I got inertia working. But its quite sluggish!! Did it actually work with older drivers?

Here's a test version with the following changes: - Reverted to the old interrupt handler (the only thing I can think of that can cause the sleep issue) - Lowered SDK to 10.7 as per RehabMan's suggestion - Refactored scroll code to be exactly like synaptics. - Implemented command gate (fixed scroll and drag timers) @wern: whats wrong with intertia? Its working fine for me. Its quite subtle though. test1.zip

@Gia: Hmm, same hw version as me. Did you try on 10.10? Same issue? Does the same thing happen with bpedman's version? @Jake: Here's a test version built using the 10.9 SDK 10.9SDK.zip

@wern: Are you sure you don't have any other kexts installed? We have the same hardware if I remember correctly. Use ikext to install. Delete the mouse kext from the plugins folder before install. @Gia: It would be helpful if you provide the hw version of your touchpad. Search in console for "found ALPS ..." If I have to guess, I'll say its a V3 Rushmore. It had an update related to sleep in its interrupt handling. @Jake: I'm using the 10.10 SDK so maybe this is why the kext is not loading on older versions. Will change that later and post a test version.

First post updated with Release 3. The only change compared to the last RC is the trackstick speed. I'm really happy to see how this driver has matured

Thats good news How much faster should I make the trackstick? 2x faster or 4x faster? The target is same speed as touchpad. There's not much I can do to improve scrolling. It's just how the Linux driver works. As for older versions of OS X, I have the deployment target set to 10.6+ I haven't personally tested anything other that Yosemite. I guess we'll have to wait for more reports.

Linux 4.2 made a rather interesting addition to the V3 trackstick code. It seems like they're trying to prevent the trackstick from interfering with the touchpad. (Maybe thats what killed the touchpad earlier) R3RC4: - Corrected minor mistakes (including scroll direction). - Updated trackstick with Linux 4.2 - Removed the hardcoded divisor again. Divisor total reduced to 8 which should make the trackstick 2x as fast as RC3. If it breaks again, I'll just hardcode the divisor to 8 as a compromise. R3RC4.zip

R3RC3: I'm optimistic again!! - Fixed minor issue with Linux 4.2 bitmap - Updated V3 Rushmore decode to Linux 4.2 (new finger detection packets!!!) - Refactored V3 Pinnacle decode like Linux 4.2 - Hardcoded trackstick divisor to 16 (hopefully it won't kill touchpad again) - Re-enabled V1 & V2 2-finger scroll. Can't wait to hear what you guys think @Jake Lo: Remind me again why you have 2 different kexts for the Exx20 and Exx30? Are those two Rushmore? Does this version improve on that? PS. Ignore RC2, this fixes it. R3RC3.zip

R3RC2: - Updated bitmap processing to Linux 4.2 - Reverted trackstick changes again (should fix touchpad and pointer jump but will still be fast). UPDATE: - I made a little mistake while porting the Linux 4.2 bitmap. - I also realised that my trackstick speed code got lost somewhere while doing the reverts a while back. - Rushmore code has been updated in Linux 4.2 (I'll add that new stuff soon). RC3 coming tonight. R3RC2.zip

Touchpad dying probably has to do with the trackstick divisor again!! The pointer jumps when using the trackstick, right? I will revert the trackstick stuff again in the next RC. I'll also update the bitmap code from Linux 4.2!! Hopefully it'll solve the 2-finger problem once and for all.