Contribute
Register

HP EliteBook 8740w sleep and wake not working.

Status
Not open for further replies.
Joined
Dec 10, 2011
Messages
59
Motherboard
MSI Z170A SLI PLUS
CPU
Intel Core i5-6600K
Graphics
Zotac GeForce GTX 1070 8GB AMP! Edition
Mobile Phone
  1. Android
Hello everyone.

I have Mountain Lion installed on my HP EliteBook 8740w and have not yet managed to get sleep or wake to work anywhere close to decent.
When the default 10 minute inactivity timer expires the screen blanks but the computer seems to keep running.
The fans are still spinning and the activity LEDs don't show any change.
When I hit a key or move the mouse, the screen flickers once but remains blank.
Mac OS is still running fine but if I'm not lucky enough to manage to blindly select the reboot option I have to turn the machine off.

I'd really appreciate any help debugging this problem.
Thanks.

This is the hardware configuration:
CPU: Core i5 540M
RAM: 4GB DDR3
Motherboard: HP 1520, Intel QM57 Rev.06
BIOS: HP 68CAD Ver. F.21 12/05/2011
Graphics: ATI FirePro M7820 (about the same as ATI Mobility Radeon HD 5870)
Audio: IDT 92HD75B3X5 and ATI Radeon HDMI
Keyboard: PS/2 with HP QLB
TouchPad: PS/2 Synaptics
Ethernet: Intel 82577LM
WiFi: Intel Centrino Ultimate-N 6300 AGN
Modem: Agere Systems HDA
FireWire: Ricoh RL5C832
USB3: NEC uPD720200
WebCam: USB HP 2MP

I've attached the output of kextstat.
 

Attachments

  • kextstat.txt
    11 KB · Views: 205
Both the HP product documentation and system information tools I ran reported it as Synaptics PS/2 TouchPad with no model number specified.
I'm not sure how to get more precise information about it.

I tried the latest version of VoodooPS2Controller you created and it made the trackpad respond but it was rather buggy.
The movement was staggered, it filled the system log with messages about unrecognized input and made the system feel choppy.
Only the touchpad itself and the buttons below it worked. The nub in the middle of the keyboard and the mouse buttons directly below the keyboard didn't.
It also didn't discard input while the keyboard was being used, so I decided I was better off with the USB mouse.

If you're interested, I can install a debug version to help collect some logs.

Did you follow all instructions for installing?
 
Did you follow all instructions for installing?
I had actually missed the instructions because they were only at the GitHub page for the project source tree and I had gone straight for the download link.
It wouldn't be a bad idea to link to the download page from the installation instructions only. ;-)

Anyway, I just finished installing RehabMan-Voodoo-2013-0701/Release and RehabMan-FakeSMC-2013-0706/Release and the behaviour is about the same. The choppiness doesn't seem to be as bad, but functionally it's the same, including the very verbose logging.
I've attached the tail of system.log starting at the latest boot.
 

Attachments

  • syslog.txt.zip
    22.3 KB · Views: 99
I had actually missed the instructions because they were only at the GitHub page for the project source tree and I had gone straight for the download link.
It wouldn't be a bad idea to link to the download page from the installation instructions only. ;-)

Anyway, I just finished installing RehabMan-Voodoo-2013-0701/Release and RehabMan-FakeSMC-2013-0706/Release and the behaviour is about the same. The choppiness doesn't seem to be as bad, but functionally it's the same, including the very verbose logging.
I've attached the tail of system.log starting at the latest boot.

Your's is one of the weird Synaptics touchpads (with the trackstyk). We added support for one of the Probooks like yours, but I don't think it works on all:

Code:
Jul 22 17:12:21 localhost kernel[0]: VoodooPS2Trackpad starting: Synaptics TouchPad reports type 0x46, version 2.0

Normal Synaptics report 0x47.

You can try running in mouse emulation mode:

Code:
sudo rm -Rf /System/Library/Extensions/VoodooPS2Controller.kext/Contents/PlugIns/VoodooPS2Trackpad.kext
sudo touch /System/Library/Extensions

If I had one of these devices, I'd try and fix it, but I don't... If you have some C++ skills perhaps you can try.
 
If I had one of these devices, I'd try and fix it, but I don't... If you have some C++ skills perhaps you can try.

Thank you very much for all your help.
I'm a programmer and I know C++ very well, but I have no experience with Mac development.
If you could help me understand what the bug is about, I'd like to try to fix it.
 
Thank you very much for all your help.
I'm a programmer and I know C++ very well, but I have no experience with Mac development.
If you could help me understand what the bug is about, I'd like to try to fix it.

The problem originates from the way there are essentially two PS2 "mouse" devices in your case (the stick, and the pad). The PS2 interface was originally designed for a keyboard only, and later extended to handle two devices, keyboard and mouse. Handling more than one mouse device is handled in several different ways.

One way is defined in the "Synaptics PS2 interfacing guide": http://www.synaptics.com/sites/default/files/511-000275-01_RevB.pdf. See section 5, "PS/2 pass-through option." In this mode, the main PS2 device is supposed to be the trackpad, and the other device sends its packets, encapsulated as special trackpad packets, so they can be discerned. This is the support that exists in the driver currently.

The other way is described in the "Industry Standard Active PS/2 Multiplexing Specification": http://www.synaptics.com/sites/default/files/ps2-mux.pdf. This document describes a way that multiple PS2 mouse devices can share the PS2 interface. Implementing this would require a fair bit of re-writing of the ApplePS2Controller class (originally written at NeXT) as the controller code is setup to assume only a single keyboard, and single mouse device.

The problem you're seeing is that (I think) the initialization codes are not making it to your trackpad, only to the stick. So your trackpad never enters absolute mode, and therefore most of the data is rejected as they do not conform to the absolute mode packet format.

Probably it would be wise to determine what mode your devices are operating in. It could be that they are operating in pass-through mode and there is just a bug in the existing code. Or it could be they are multiplexed per the second spec. There is a section in the multiplex spec that describes how to determine if the controller supports multiplexing. You could write a little code (after you're familiar with the ApplePS2Controller class) to determine if it is active and if multiple devices respond as described in the spec...
 
Well, I had a look through the code and the PS/2 multiplexing document.
When I tried to reproduce the problem again, it was gone. Removing VoodooPS2Trackpad.kext fixed it.
All buttons and the nub work just fine now. That's good enough for me because I wasn't expecting support for fancy multi-touch gestures anyway.

Thanks again.
 
RehabMan said:
As far as audio, I don't use VoodooHDA. Patched AppleHDA is generally more stable than VoodooHDA...
Sure, but isn't that an option only if the computer has a Realtek sound chip?
 
Sure, but isn't that an option only if the computer has a Realtek sound chip?

No. There are plenty of guides to patch AppleHDA for other chips. For example, the ProBook is IDT and we have patched AppleHDA...
 
No. There are plenty of guides to patch AppleHDA for other chips. For example, the ProBook is IDT and we have patched AppleHDA...
Cool, thanks.
I checked and found some pretty extensive guides.
It looks like a pretty elaborate project so I'll try it a bit later.
 
Status
Not open for further replies.
Back
Top