Jump to content

Asus 1201N Installation and support


  • Please log in to reply
1037 replies to this topic

#1021
theconnactic

theconnactic

    Stubborn AMD user

  • Local Moderators
  • 2,807 posts
  • Gender:Male
No, Walker: some AppleHDA restarting method (restartHDA.app for example) was always needed for audio to work after waking up, at least since Mountain Lion, which was the first system I extensively used with my 1201n. Patching IOAudioFamily.kext used to be of no help, so we simply accommodated with what we had. With my early tests with Mavericks, months before you made it a serious option with your kernel patcher module (thank you, by the way), even restarting AppleHDA wouldn't work, and that's one of the (minor) reasons I stayed with Mountain Lion.
P.S.: And by the way, my Yosemite guide is almost done. I'll post it soon!

#1022
TimeWalker75a

TimeWalker75a

    InsanelyMac Legend

  • Gurus
  • 930 posts
  • Gender:Male
I'm aware, at least from posted in this thread, that IOAudioFamily patched by Dinesh has worked for someone in 10.8.3. Also I'm aware that people who stayed at 10.6.8 also used a patched kext that resumed audio.

From my debugging with codec commander I can wake the audio amp from fugue sleep (when you press Sleep and abort 5 seconds after it starts putting the machine sleep), which normally would just result in no sound if no special measures are applied. With proper sleep cycle the codec amp in node 0x14 remains powered down completely, so it's something internal in either the chip, ACPI problem or improperly patched kext resources that have spread like plague in this thread and this cause problems for everyone.

#1023
theconnactic

theconnactic

    Stubborn AMD user

  • Local Moderators
  • 2,807 posts
  • Gender:Male

Hopefully not a combinations of two or more of the listed causes .:)

 

All the best!



#1024
TimeWalker75a

TimeWalker75a

    InsanelyMac Legend

  • Gurus
  • 930 posts
  • Gender:Male

Just an update .. it's ACPi problem.. I've determined that In 10.8.4 EAPD can be set high with my CodecCommander absolutely flawlessly.. In 10.8.5, unfortunately, it doesn't. Then I recalled something ... 

And the mystery is {censored}*ng solved - same thing as with Lenovo T420 .. it's the stupid _WAK bug introduced with 10.8.5, see https://github.com/R...system_WAK2.txt

Patched my DSDT to resolve it and EAPD now resumes like clockwork every time I wake .. kudos to @RehabMan ! Yet I need to code a workaround behavior in CC for MCP chipset, because it doesn't return proper status from IRR register when EAPD bit is set, so I have no way to monitor whether it will die 30 seconds in after waking  like it does on Intel chipset or it will remain working as is. Long story short - EAPD problem with this machine is resolved. 



#1025
theconnactic

theconnactic

    Stubborn AMD user

  • Local Moderators
  • 2,807 posts
  • Gender:Male
Another good news: your AppleHDA patching method does work with Yosemite. Repairing permissions in Single User Mode was what wasn't doing any good. I copied my S/L/Caches to the Desktop, ran kext utility, put the caches back and restarted in SUM. The system rebuilt the caches and sound started to work after the subsequent reboot.

All the best!

#1026
TimeWalker75a

TimeWalker75a

    InsanelyMac Legend

  • Gurus
  • 930 posts
  • Gender:Male

Well, I've upgraded my 10.8.5 installation to 10.9.4. My EAPD solution still works flawlessly, sound resumes after every kind of sleep. You can even put it to sleep with a youtube video playing and it will play it when you wake the laptop.

 

First thing noticed - the fan runs faster and is more audible, meaning temps are considerably higher with absolutely same setup (kexts, ACPI, config). I've seen same thing happen on my DELL, where with 10.8.5 the fan would hardly ever turn on, while past 10.9.2 it now skyrockets every time I open Chrome, open new tab in a browser or start an app. I consider power management to be clearly broken in 10.9, especially with Atom. I'm not going to jump the Yosemite bandwagon just because there's virtually no point in using pre-release software that has seen only second seed. In short, with 10.9.4 everything except HDMI audio works.

 

I still have to polish many things, the to-do is as follows:

- Solve DisableUsbLegacySupport from Clover issue which doesn't happen with 1014 boot file, but happens with new releases. It causes EHCI handoff fail.

- Get rid of WMIFamily completely, implementing all WMI-related controls through ACPI. Right now it's needed only to toggle radio interfaces and touchpad LED, the rest can be done by pure ACPI+VoodooPS2

- If I succeed with the above, then I plan on inverting media buttons (f-keys only, I don't think reversing arrow keys is a wise idea) and having the ability to switch them to default:

72492532c1.png

This is what I have on my DELL.

- Find custom EDID that has 1360x768 resolution for 12.1 panel in order to get proper GUI resolution with Clover and no white-black-desktop transition when loginwindow starts.

- Implement RehabMan's subtle brightness controls opposed to step-like controls (will tie the CMOS brightness level preservation to touchpad toggle ACPI method from voodoo daemon) 

- Add proper fan sensor or figure out why the heck its not outputting data as is, because the value is indeed correct. Get rid of CPUSensors or make kozlek fix it for Atom. Anyhow with OC enabled it shows garbage, so only CPU and GPU thermocouple registers show relevant information.

- Probably add custom image and text to About this Mac for personalization. 

 

The rest already works .. and should remain working even when 10.10 hits, unless Apple seriously decides to slap us with signature enforcement. 



#1027
theconnactic

theconnactic

    Stubborn AMD user

  • Local Moderators
  • 2,807 posts
  • Gender:Male

Great job!

 

As soon as you share your patch for the no-sound-after-wake issue, I'll post here my beta guide for Yosemite: it's already done. My machine is quite stable now. For me, the sound even resumes after sleep, but only for headphones - no speakers. All other things work as should be. I feel 10.10 to be a little bit faster (but also hotter) than both 10.9. and 10.8.

 

All the best!


P.S.: about the keyboard backlight hack, did you succeed?



#1028
TimeWalker75a

TimeWalker75a

    InsanelyMac Legend

  • Gurus
  • 930 posts
  • Gender:Male

I'm always pushing new changes to my Git, just pushed MCP79 workaround, so you can grab the source and compile Release yourself. 

I guess I'm staying at 10.8.5 because it feels the most snappiest on this little unit, also doesn't take fng forever to enter sleep.

 

I'm still waiting on my keyboard to arrive. And since I wasn't really able to sort out WMI things I wouldn't be able to use the button to illuminate it, because without WMI it doesn't light up. With my setup posted a couple pages back it light up properly and disables trackpad surface, but I can easily remove the notification from ACPI code and it would just light up, acting as +5V source for my keyboard.

 

P.S. Fan speed and temps added successfully ...

b07226c64d.png



#1029
smittiee

smittiee

    InsanelyMac Protégé

  • Members
  • Pip
  • 5 posts

Hallo TimeWalker75a & theconnactic,

 

can one of us write a little tut - how i  can make step by step an Yosmite or Mavericks usb installation stick and install it on hdd with clover. I have never work with clover thats why i need your/ our help. 

 

Thx

smittie B)



#1030
TimeWalker75a

TimeWalker75a

    InsanelyMac Legend

  • Gurus
  • 930 posts
  • Gender:Male

Clover is not yet properly fixed to support our chipset. There are issued related with EHCI port handoff to the OS. There's a workaround to this, but I'd like to see it fixed properly. Using bareBoot in the meantime, but it's absolutely non-user friendly as the name may suggest. I would like to be able to use IDE mode in BIOS and third-party EFI driver from MacBookPro5,5 (MCP too) in order to use Boot Booster, but still have AHCI support in OSX.

 

Referencing my aforementioned To-Do list:

 

- Solve DisableUsbLegacySupport from Clover issue which doesn't happen with 1014 boot file, but happens with new releases. It causes EHCI handoff fail.
- Get rid of WMIFamily completely, implementing all WMI-related controls through ACPI. Right now it's needed only to toggle radio interfaces and touchpad LED, the rest can be done by pure ACPI+VoodooPS2
- If I succeed with the above, then I plan on inverting media buttons (f-keys only, I don't think reversing arrow keys is a wise idea) and having the ability to switch them to default:
This is what I have on my DELL.

- Find custom EDID that has 1360x768 resolution for 12.1 panel in order to get proper GUI resolution with Clover and no white-black-desktop transition when loginwindow starts. Not possible as VBIOS can't be unlocked for patching, no MCP79 datasheet is available, so we are stuck with 1280x720
- Implement RehabMan's subtle brightness controls opposed to step-like controls (will tie the CMOS brightness level preservation to touchpad toggle ACPI method from voodoo daemon)
- Add proper fan sensor or figure out why the heck its not outputting data as is, because the value is indeed correct. Get rid of CPUSensors or make kozlek fix it for Atom. Anyhow with OC enabled it shows garbage, so only CPU and GPU thermocouple registers show relevant information.  Sadly, EC doesn't provide GPU temp register, only CPU heatsing and memory ambient registers are present. Fan speed is readable too. Nothing more to add.
- Probably add custom image and text to About this Mac for personalization. 

 

I have completely got rid of WMIFamily/AsusHotkeys (including WMI related {censored} from DSDT). Everything is implemented via ACPI and RehabMan's voodoops2.

Keys can be used in either special mode (Fn press not required to cause special function) or standard mode (requiring Fn). If someone here is good at following directions I would like someone to try it, because I'm unable to test touchpad toggle properly, since my flex cable is damaged.

 

I implemented subtle brightness transitions which are especially noticeable when waking or plugging/unplugging DC jack.. or adjusting across the entire range of brightness, which is not laggy anymore! 

Perhaps one thing that I need to tweak is the backlight on/off button. When adjusting brightness I should check whether OLBG() returns 0, meaning the panel backlight was off and if it's the case, then write OLBS(One) to let firmware know it's really on now.

 

Attached File  test_backlight_keyboard.zip   158.67KB   1 downloads

 

To test you have to have Clover with NewWay fixes enabled and having it fix MemoryRegions in DSDT in order to prevent PCI Configuration Begin and other problems related to drifted offsets (due to different memory amount in laptops), then:

 

1. Install ACPIBacklight and VoodooPS2Controller with KextUtility (this one, and not something else ..it updates KextExcludeList properly)

2. Install VoodooPS2 Daemon in order to monitor touchpad state across shutdown

sudo cp org.rehabman.voodoo.driver.Daemon.plist /Library/LaunchDaemons

sudo cp VoodooPS2Daemon /usr/bin

3. Put DSDT.aml in ACPI/patched folder on your ESP or USB drive. 

4. Change whichever mode you prefer for keyboard F keys in keyboard preferences.

5. Post results, especially for toggling touchpad with dedicated button or F9/Fn+F9 (depending on the mode you've used)



#1031
theconnactic

theconnactic

    Stubborn AMD user

  • Local Moderators
  • 2,807 posts
  • Gender:Male

Will test it later tonight. 

 

DP3 installed here and working fine, by the way.

 

All the best!



#1032
TimeWalker75a

TimeWalker75a

    InsanelyMac Legend

  • Gurus
  • 930 posts
  • Gender:Male

Will test it later tonight. 

 

DP3 installed here and working fine, by the way.

 

All the best!

If you are still at it, there's a mistake in TPDN method in DSDT... 

If (LAnd (LEqual (Local0, One), LEqual (Arg0, 0xFFFF)))

Change One to Zero.

 

Edit: Actually, no .. it was correct, only I forgot to invert the other thing that comes before that.. so the entire method should be like that. Tested on another Asus - touchpad is properly disabled and enabled in whichever mode (surface and LED correspond). Plugging in USB mouse disabled the touchpad if respective setting is checked under Accessibility preferences. Touchpad state properly resets across reboot.

                    // ACPI Touchpad Disable Notification
                    Method (TPDN, 1, NotSerialized)
                    {
                        // Get current touchpad state from NVS register and invert for comparison
                        XOr (OTPG (), One, Local0)
                        // If state passed from VoodooPS2 is different then update stored state
                        If (LAnd (LNotEqual (Arg0, Local0), LNotEqual (Arg0, 0xFFFF)))
                        {
                            // BIOS treats 0 as disabled and 1 as enabled
                            // VoodooPS2 passes 1 to the method to indicated disabled, need to invert
                            XOr (Arg0, One, Local1)
                            // Update touchpad state in NVS register
                            OTPS (Local1)
                        }
                        // Reset LED if active when VoodooPS2Daemon is telling us shutdown/restart is issued
                        If (LAnd (LEqual (Local0, One), LEqual (Arg0, 0xFFFF)))
                        {
                            OTPS (One)
                        }
                    }


#1033
TimeWalker75a

TimeWalker75a

    InsanelyMac Legend

  • Gurus
  • 930 posts
  • Gender:Male

Tying to figure out how to adapt Asus' Super Hybrid Engine .. there are two methods that Set and Get CPU FSB Value in ACPI...

0x0 argument is Super Performance

0x1 argument is High Performance

0x2 argument is Energy Saver 

 

There's also a 3 profile array in ACPI which defines FSB values for certain profile. So, in a nutshell, one can overclock really easily and to whichever value preferred (up to FSB 667 MHz) just via ACPI.

The odd problem is that even in Windows my FSB doesn't drop below 533 when selecting Energy Saver.. and goes up by 10 Mhz (543) when Super Performance is selected. Adjusting these with OC in BIOS enabled (FSB667) leads to immediate freezes, so need to be adjusting with default clock. Can anyone report (looking with CPU-Z) how SHE acts for them in Windows?



#1034
TimeWalker75a

TimeWalker75a

    InsanelyMac Legend

  • Gurus
  • 930 posts
  • Gender:Male

Well,. I'm not absolutely sure it actually changes the FSB, since HWmonitor keeps reporting same frequencies, but CPU heatsink temperature goes down by 4С when switching to Battery Saver profile from Hight Performance. The fan keeps spinning though. But then again, I never saw fan stopping on this netbook .. but I do recall that selecting Battery Saving profile in Windows reduced the clock to 1.2GHz if I'm not mistaken. 

Here're my temps .. anyone care to report theirs with default profile ?

 

72018c6385.png



#1035
TimeWalker75a

TimeWalker75a

    InsanelyMac Legend

  • Gurus
  • 930 posts
  • Gender:Male

Since VBIOS can't be unlocked for patching due to legacy region address mismatch with regular Intel platforms and there's no MCP79/MCP7A datasheet to obtain the address I had to improvise to go beyond 1280x720 resolution. Based on finding by @Pene I have modified the VBIOS inside BIOS capsule to support 1360x768 resolution, which is good enough (only squashed horizontally by 6px) to render Clover's GUI and Apple logo+spinner in almost-native widescreen resolution. 

0:158  0:057  BiosVideoDriverBindingStart!
0:158  0:000  BiosVideoCheckForVbe
0:158  0:000  EdidOverride not found
0:396  0:237   Edid1-
0:633  0:237   Edid0-
0:679  0:045     0 640x480 attr=3BF - ok, 640x480, working, highest, pref=0
0:680  0:001     1 800x600 attr=3BF - ok, 800x600, working, highest, pref=1
0:682  0:001     2 1024x768 attr=3BF - ok, 1024x768, working, highest, pref=2
0:690  0:007     3 1360x768 attr=3BF - ok
0:690  0:000  BV new mode: 2 1024x768
1:002  0:311   SetMode pref 2 (2) = Success

3:663  0:000  Clover revision: 2758  running on 1201N
3:663  0:000  ... with board 1201N
3:663  0:000  Clover load options size = 0 bytes
3:672  0:008  Using OEM config.plist at path: EFI\CLOVER\OEM\1201N\config.plist
3:673  0:001  EFI\CLOVER\OEM\1201N\config.plist loaded: Success
3:677  0:003  Loading early settings
3:677  0:000  timeout set to 3
3:677  0:000  LoadDrivers() start
3:680  0:002  Loading FSInject-64.efi
3:682  0:002  Loading ION-64.efi
3:684  0:002  LoadDrivers() end
3:684  0:000  SetScreenResolution: 1360x768 - setting Mode 3
3:684  0:000  BV new mode: 3 1360x768
4:025  0:341  Video mode change to mode #3: Success 

If anyone is interested I can post the modified AHCI-unlocked BIOS on request. Not that you need AHCI mode enabled anymore, which takes about twice as long to boot as IDE mode due to controller initialization routine.



#1036
smittiee

smittiee

    InsanelyMac Protégé

  • Members
  • Pip
  • 5 posts

Hallo TimeWalker,

thank you for your little TUT. My1201n works now with 10.9. bootloader clover 2721. Mavericks is little bit faster as ML.

Best Regards

smittie 



#1037
TimeWalker75a

TimeWalker75a

    InsanelyMac Legend

  • Gurus
  • 930 posts
  • Gender:Male

As of Clover r2771 I've corrected memory detection issues on MCP79 chipset (thanks to ASUS for messy SMBIOS).

Set SMBIOS > Trust = No in order to read actual speed/size/type and OEM data from memory modules. 



#1038
TimeWalker75a

TimeWalker75a

    InsanelyMac Legend

  • Gurus
  • 930 posts
  • Gender:Male

Posted everything for 1201N for any OSX version to date in Mavericks *support* thread. 

http://www.insanelym...-mav/?p=2042094

Enjoy ...







0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users

© 2014 InsanelyMac  |   News  |   Forum  |   Downloads  |   OSx86 Wiki  |   Mac Netbook  |   PHP hosting by CatN  |   Designed by Ed Gain  |   Logo by irfan  |   Privacy Policy