Contribute
Register

ASUS laptops with optimus. Fan speed issue.

Status
Not open for further replies.
Joined
Jul 8, 2011
Messages
16
Motherboard
ASUS P6T SE
CPU
Xeon X5660
Graphics
GTX 650
Mobile Phone
  1. Android
Hi!
I am using an ASUS FX753 laptop with i7-7700 and nVidia GTX1050 with MacOS High Sierra.
Obviously nVidia GTX1050 optimus is not supported under MacOS, so I shut it OFF in DSDT. And the following issue revealed:
Booting up with nVidia graphics swithed off results in after approx 30sec cooling fan start to blow at 100% speed and shuts off only at shut down. Even reboot doesnt resets fan speed.

It seems that its a popular issue with modern ASUS laptops with nv optimus, I found alot of topics from linux users describing the same issue on ASUS FX753, GL753, GL502 laptops.
Like that:
https://rog.asus.com/forum/showthre...sue-on-GL753VE-nvidia-bumblebee-bbswitch-ACPI

Is anybody facing the same issue? Any thoughts on fixing that?
 
Hi!
I am using an ASUS FX753 laptop with i7-7700 and nVidia GTX1050 with MacOS High Sierra.
Obviously nVidia GTX1050 optimus is not supported under MacOS, so I shut it OFF in DSDT. And the following issue revealed:
Booting up with nVidia graphics swithed off results in after approx 30sec cooling fan start to blow at 100% speed and shuts off only at shut down. Even reboot doesnt resets fan speed.

It seems that its a popular issue with modern ASUS laptops with nv optimus, I found alot of topics from linux users describing the same issue on ASUS FX753, GL753, GL502 laptops.
Like that:
https://rog.asus.com/forum/showthre...sue-on-GL753VE-nvidia-bumblebee-bbswitch-ACPI

Is anybody facing the same issue? Any thoughts on fixing that?
No "Problem Reporting" files attached.
Read FAQ, "Problem Reporting" again. Carefully. Attach all requested files/output.
https://www.tonymacx86.com/threads/faq-read-first-laptop-frequent-questions.164990/
Use the gen_debug.sh tool mentioned in the FAQ, that way it is less likely you'll omit something.

Also, your profile has no hardware details. Please fix as per FAQ.
 
No "Problem Reporting" files attached.

debug_32445 - dump with nVidia enabled, untouched DSDT
debug_2347 - dump with nVidia DISABLED via DSDT patch
 

Attachments

  • debug_32445.zip
    1.9 MB · Views: 66
  • debug_2347.zip
    1.9 MB · Views: 65
Last edited:
I do thing those dumps will provide no help.
The issue itself is very common under linux systems. It might be the case that EC has a somewhat like a watchdog timer, listening to nvidia chip thermal/power status, and when nvidia chips shuts off, EC thinks nv chip "hanged" and applies a safety fan=100% strategy. Only ASUS knows how this works.
There is some posts reporting similar bug appearing randomly under windows.
 
Last edited:
Take a look again, posted a second dump.

You didn't follow the guide.
You should be adding an _INI in SSDT-9 to call _OFF or more likely _PS3.

For example, in SSDT-9 right before _PS3:
Code:
Method(_INI)
{
   OPCE=3
   _PS3()
}
        Method (_PS3, 0, NotSerialized)  // _PS3: Power State 3
        {
...

(setting OPCE=3 and calling _PS3 instead of _OFF has been shown to be needed for proper fan control)
 
You didn't follow the guide.
You should be adding an _INI in SSDT-9 to call _OFF or more likely _PS3.
(setting OPCE=3 and calling _PS3 instead of _OFF has been shown to be needed for proper fan control)

Did that.
Results in same behavior: fan bumps up to limit after ~30secs.
 

Attachments

  • debug_55.zip
    2 MB · Views: 71
Did that.
Results in same behavior: fan bumps up to limit after ~30secs.

ACPI not patched correctly.
Basic patches such as _OSI spoofing are not implemented.
Refer to the guide:
https://www.tonymacx86.com/threads/guide-patching-laptop-dsdt-ssdts.152573/

Also, patching OEM SSDTs requires correct config.plist/ACPI settings. Yours are wrong.
The correct settings (AutoMerge=true, for example) are covered in the guide above, but also the guide plists start out with correct settings:
https://www.tonymacx86.com/threads/guide-booting-the-os-x-installer-on-laptops-with-clover.148093/
 
Basic patches such as _OSI spoofing are not implemented.

AutoMerge=true

Played some with _OSI injecton - no changes.
Decided to do add a debug calls into EC methods using Store("EC method name, Debug") call
As I already said the machine boots up and nVidia successfully shuts off, but after approx 30secs a fan bumps up.
So If do a "log stream" right after boot up, then the following call will be thrown and right in that moment when fan goes crazy:

Code:
Last login: Tue Dec 11 02:48:22 on console
MacBook-Pro-xpbim3:~ xpbim3$ /Users/xpbim3/Desktop/log.sh ; exit;
2018-12-10 21:48:45.448454+0000 0x122      Default     0x0                  0      kernel: (AppleACPIPlatform) [ACPI Debug]
2018-12-10 21:48:45.448471+0000 0x122      Default     0x0                  0      kernel: (AppleACPIPlatform) [ACPI Debug]
2018-12-10 21:48:45.448481+0000 0x122      Default     0x0                  0      kernel: (AppleACPIPlatform) "EC _QB1 enter"
2018-12-10 21:48:45.448483+0000 0x122      Default     0x0                  0      kernel: (AppleACPIPlatform) "EC _QB1 enter"
2018-12-10 21:48:45.448677+0000 0x122      Default     0x0                  0      kernel: (AppleACPIPlatform) [ACPI Debug]
2018-12-10 21:48:45.448679+0000 0x122      Default     0x0                  0      kernel: (AppleACPIPlatform) [ACPI Debug]
2018-12-10 21:48:45.448687+0000 0x122      Default     0x0                  0      kernel: (AppleACPIPlatform) "EC _QB1 exit"
2018-12-10 21:48:45.448690+0000 0x122      Default     0x0                  0      kernel: (AppleACPIPlatform) "EC _QB1 exit"
And vice-versa: no nVidia shutdown and no QB1 call.

The EC method QB1 is:

Code:
        Method (_QB1, 0, NotSerialized)  // _Qxx: EC Query
        {
            Store (0xB1, P80H)
        }

I did commented that line - no luck. Fan does the same.
Looked for _QB1 calls all over entire ACPI code - I though it may be called from somewhere else but nope....
Any help will be appreciated.
 
Last edited:
Status
Not open for further replies.
Back
Top