Mavericks: HDMI Audio - AppleHDA [Guide]

[Lnerzn]

No HDMI audio on Intel NUC DC3217BY/BIOS.46/Mavericks 10.9.1 after following these guides:

https://github.com/toleda/audio_hdmi_uefi
https://github.com/toleda/audio_hdmi_hd4000

compiling the DSDT.aml by patching the System DSDT with

https://raw.github.com/toleda/audio_hdmi_hd4000/master/Patches/ib1-ami_efi_clean_compile.txt
https://raw.github.com/toleda/audio...-HD4000-NUC-A1/ib8-hdmi_audio_intel_nuc-1.txt

no effect.

---

compiling the DSDT.aml by patching the System DSDT with:

https://raw.github.com/toleda/audio...I-Clean_Compile/ib1-ami_efi_clean_compile.txt
https://raw.github.com/toleda/audio_hdmi_uefi/master/Patches/UEFI-NUC/uefi5-hdmi_audio-1.txt

causes boot -v to get stuck on the dreaded -- Missing Bluetooth Controller Transport! error

---

The SSDT (without DSDT) method from here:
https://github.com/toleda/audio_hdm.../audio_ssdt-hdmi-ami_hd4k-amd-nvidia-3_v4.zip


gets the same -- Missing Bluetooth Controller Transport! error

---

As an aside, the easiest way to recover from getting stuck at the -- Missing Bluetooth Controller Transport! error is to power down and reboot in single user then

rm /Extra/[DS]SDT.asl;reboot

That was the only way I figured out to boot up again after the SSDT method got me stuck. DSDT=null obviously does not help when not using a DSDT.

---

Another note is that the builtin search does not match the raw binary in 10.9.1:

capri06 0A00660100020302000000020000000100000020100700001007000000000000000000000000000000000000000000000205000000040000070100000304000000 0400000701 0800000600 /System/Library/Extensions/AppleIntelFramebufferCapri.kext/Contents/MacOS/AppleIntelFramebufferCapri

The following search string does succeed in patching AppleIntelFramebufferCapri:

capri06 0A00660100020302000000020000000100000040100700001007000000000000000000000000000000000000000000000205000000040000070100000304000000  0400000701 0800000600  /System/Library/Extensions/AppleIntelFramebufferCapri.kext/Contents/MacOS/AppleIntelFramebufferCapri

The old AppleIntelFramebufferCapri had '201007' in the middle of the search string
The 10.9.1 AppleIntelFramebufferCapri has '401007' in the middle of the search string

But since I don't know what I'm doing beyond simple pattern recognition, I have no idea what else may have changed in AppleIntelFramebufferCapri

---

Thanks in advance to any one who takes a look at this for me..

I'm attaching the IOJones output and the extracted unmodified System DSDT.dsl and the DSDT_46.dsl is patched with
https://raw.github.com/toleda/audio_hdmi_hd4000/master/Patches/ib1-ami_efi_clean_compile.txt
https://raw.github.com/toleda/audio...-HD4000-NUC-A1/ib8-hdmi_audio_intel_nuc-1.txt

This is the DSDT which gets the NUC stuck in a -- Missing Bluetooth Controller Transport! error right before the login screen. View attachment DC3217BY.zip

 

[toleda]

No HDMI audio on Intel NUC DC3217BY/BIOS.46/Mavericks 10.9.1 after following these guides:

Fatal cockpit error.

You missed this step (second paragraph in every HDMI Audio Guide)

Configure MaciASL/Preferences/Sources/+ with URL:

https://raw.github.com/toleda/audio_hdmi_uefi/master

See Post #1, [Easy Guide] ML: UEFI HDMI Audio
Download audio_hdmi_uefi/[Easy_Guide]-os_x_hdmi_audio_edits_intel_nuc-v2.0.pdf.zip at master · toleda/audio_hdmi_uefi (Select View Raw)

For MaciASL screenhots, audio_hdmi_uefi/[Guide]-UEFI-hdmi_audio_dsdt_edits_v2.1.pdf.zip at master · toleda/audio_hdmi_uefi (Select View Raw)

 

[freeze]

No idea. The difference between the 1st IOJones and the most recent is a second display with audio. What was connected in both IOJones and what was added in the most recent?

My wife was watching satellite tv when I prepared the first IOJones, so the second display (with audio) did not show up in this one.

Then for the second IOJones I switched the Surround-Receiver with the TV to the input that is connected with the computer.

That is why it then was listed in IOJones.

Sorry for that confusion.

 

[mmmprod]

Hello Toleda,

I have one question I want to use SSDT method instead of DSDT
I'm using Clover you are asking to drop OEM but certain part are needed in order to maintain my computer stable.

Clover is able to Drop a specific SSDT. Which one should I have to drop ?
Is it possible or I may have to use DSDT edit

Thank you for your feedback

Best

I'm on 10.9.1 and Clover

 

[toleda]

Sorry for that confusion.

Is it working correctly or not?

 

[toleda]

I'm using Clover you are asking to drop OEM but certain part are needed in order to maintain my computer stable.

Why? What is not stable?

 

[mmmprod]

I have random Kernel Panic when I drop all SSDT. I was not able to find why but it is coming from everything.
But if I only Drop cpuPm and cpuIOst and that's it I have a stable Hack.

If I may ask: I have one last question regarding your Patch.

I have a GTX 560ti and a i5 4670K.
I have noticed that when I use your HMDI Intel8 series Patch the Memory and Core Clock of my GPU are lowered to the minimum

I have try to replace only POP2 for PEGP without the Patch (everything went fine)
PEGP to GFX0 without the rest of the Patch (everything went fine)
GFX0 to IGPU without the rest of the Patch (everything went fine)

but as soon as I add the DSDT Patch code the CLocks are lowered down.


Another thing strange that I may add.
I'm using FAKEID in Clover to have AGPM working with this awfull GTX 560ti

when I use your full HDMI Patch FAKEID is working and when I dont use your PATCH AGPM is working but AGPM is lost ... LOL

it seems that your patch is able to give me access to AGPM (thx to fakeid) and have my graphics card correctly working with Maverick !... however my Core and Memory are lowered to the minimum. Just wanted to share that

SMBIOS 14,2

My DarwinDump
https://dl.dropboxusercontent.com/u/25204447/DarwinDumper_2.8.8_AMI_X64_2437_Mav_mmmprod.zip

 

[toleda]

I'm using FAKEID in Clover to have AGPM working with this awfull GTX 560ti

With sysdef 14,2, HDMI audio dsdt edits will set the graphics name to GFX0. In AGPM, imac14,2 has GFX0, however, it is Kepler, not good for Fermi. I don't know what fakeid you are using; it should be for Fermi. The correct approach is to add your device ID/PM to AGPM/imac14.2.

 

[mmmprod]

You are right for everything However there is something strange:

As soon as I use your HDMI patch, even without any FAKEID in Clover it is locking down my GPU clock.
The only method to fix that is to remove the HDMI patch from my DSDT... :/



FakeID is just a plus that I wanted to share here but it has nothing to do with my problem.
I'm completely stucked and i don't know what to do.

 

[toleda]

I'm completely stucked and i don't know what to do.

I explained the reason in my previous reply: Kelpler PM on applied to Fermi is the problem; no PM is better than the wrong PM.