DSDT Patching for Dummies - SnowLeo with 100% vanilla /S/L/E

[Silencer]

Actually ... It works for me ... Maybe you should give a try:

I downloaded your info.plist, VoodooHDA(256) from ProjectX and I added a reference to the kext in rc.sleep/rc.wake ... and so far every thing is perfect !

Well, I can only think of two things to try there. I'm using VoodooHDA in /E/E, but might try it in /S/L/E, although I'm against that. Another thing - I disable HDEF device in DSDT to make sure that AppleHDA doesn't load and I don't have to remove it from /S/L/E.

I've tried loading without VoodooHDA at all, and failed to wake from sleep. Then I tried unloading it manually before sleep, and still failed to wake up. It is also possible, that the issue has nothing to do with VoodooHDA at all, and just related to something else, which I can't figure out.

I also don't want unloading/loading VoodooHDA for sleep because in that case I would have to restart all applications working with sound upon waking up.

[Sebinouse]

I also don't want unloading/loading VoodooHDA for sleep because in that case I would have to restart all applications working with sound upon waking up.

I've been using this config from a few time now ... You're right : every application using sound pause on sleep and doesn't play again on wake. And sound is very low on wake.

Good luck with you AppleHDA patch ! (i hope it will be the final solution ...)

[grandflash]

Hi Silencer,
really great job, your post here. First time, i got what is ment by c- p- states. I have a x61 tablet, in fact its quite similar. Of course it has a Intel X3100, but we have everything working, despite of speedstepping. In 10.5, i used a Legacy Plugin from osx86.es and patched my dsdt to got vanilla speedstep working (hpet device, lpc device file working too). In 10.6.2 i assigned irqs 0 10 8 to hpet device to get applehpet.kext working together with bluetooth.

What i wanted to ask you for is, in your dsdt.dsl i found

Store (Zero, RRBF)

/* Fix - Full SSDT tables for Vanilla SpeedStep */
//\_PR.NOTI ()
/* Fix - Enabling proper SMBus */
//\_SB.PCI0.SBUS.ENAB ()
/* Fix End */


Return (WAKI)


Whats that for? Do i need to copy ssdt files to my root dir?

Sorry bothering you, but this is confusing me.

Alex

[Silencer]

What i wanted to ask you for is, in your dsdt.dsl i found

Store (Zero, RRBF)
/* Fix - Full SSDT tables for Vanilla SpeedStep */
//\_PR.NOTI ()
/* Fix - Enabling proper SMBus */
//\_SB.PCI0.SBUS.ENAB ()
/* Fix End */
Return (WAKI)
Whats that for? Do i need to copy ssdt files to my root dir? Sorry bothering you, but this is confusing me.

Well, I never got around to clean my DSDT, so there are a lot of places with old and experimental code which is commented out, like the example you posted. Just ignore it, this is something I've tried before I got my speed step working properly.

[Silencer]

Guys, I need some help with AppleHDA testing. I've come a long way now and, unfortunately, I don't think it will be possible to enable vanilla AppleHDA support without some heavy patching. But I would like some of you (especcially those with ThinkPad models other than T60p) to test what I have. The test should work on all ThinkPads that have AD1981HD codec: R60e, T60, T60p, X60, X60s, X60t, Z60m, Z60t, Z61e, Z61t, Z61p.

If you're lucky (or it is me who's unlucky), you might have working sound!

1. Disable VoodooHDA.kext or AppleAzaliaAudio.kext. I usually do the following:

Code: Select all

sudo -s
cd /Extra/Extensions/
mv VoodooHDA.kext VoodooHDA.kext.noload

2. Apply the DSDT patch. Replace your HDEF device in DSDT with the following:

Code: Select all

            Device (HDEF) 
            {
                Name (_ADR, 0x001B0000)
                Name (_S3D, 0x03)
                Name (RID, Zero)
                Name (_PRW, Package (0x02)
                {
                    0x05, 
                    0x04
                })
                Method (_PSW, 1, NotSerialized)
                {
                    Noop
                }
				Method (_DSM, 4, NotSerialized)
                {
                    Store (Package (0x08)
                    {
							"built-in",
                            Buffer (One) {
								0x00
							},							
							"layout-id", 
                            Buffer (0x04)
                            {
                                0xBD, 0x07, 0x00, 0x00
                            },						
                            "PinConfigurations", 
							Buffer (One)
							{
								0x00								
							}							

                    }, Local0)
                    DTGP (Arg0, Arg1, Arg2, Arg3, RefOf (Local0))
                    Return (Local0)
                }
            }

3. Copy one of the two legacy kexts to your /Extra/Extensions:
LegacyAD1981HDTest1.kext
LegacyAD1981HDTest2.kext

4. Fix permissions with pfix. Grab the 3.0 version, it allows fast /E/E only fix:

Code: Select all

pfix -s -s -t

5. Reboot and check two places. Your System Profiler's Audio device and System Preferences Sound output and input devices. Please also check your system Console and let me know about your "Sound Assertion" errors.

With legacy kext test #1 you should see three connections in System Profiler - Line-Out, Internal Microphone and S/PDIF Out. But there probably won't be any inputs or outputs in System Preferences. But if you're lucky, sound should be fully functional, with two outputs and one input!

With legacy kext test #2 you should see only two connections in System Profiler, and you will see Digital Out and Internal Microphone in Sound preferences. You will also notice, that microphone is fully functional! I've also tested S/PDIF out on external dock I have - it also works.

In both cases you should see plenty of Sound Assertion errors in your log, please copy them here or send me a personal message with them. It might help me to solve the speakers/line-out problem, but as I said before, chances are slim.

[zillac]

Hi, I'm a little curious how do you deal with the fan not automatically adjusting speed.
Do you just set it manually with one of the key combinations?
Doesn't it get overheat while unattended?

[Silencer]

Hi, I'm a little curious how do you deal with the fan not automatically adjusting speed.
Do you just set it manually with one of the key combinations?
Doesn't it get overheat while unattended?

It doesn't overheat for me, but I'm planning to make a fix for this - put Store (0x80, \_SB.PCI0.LPC.EC.HFSP) together with Operating System fix, and it will enable Automatic fan speed, but it will be a bit noisy.

[Silencer]

Hey Silencer - what keyboard layout do you have? I'm in the UK so my bottom row of keys (well, not bottom, the one above the spacebar row) has shift, then \|, then Z, then X, etc. In 10.5.6 i had to use a different ps2 controller kext that set the keyboard type to ISO in order for OSX to recognise the key to the left of Z - i think i then used ukelele to create an english keyboard layout to map the key - have you had any such problem or is your keyboard layout different?

Chris, I've tested VoodooPS2 kext yesterday, and it far more superior to ones I use, the only problem there - trackpoint doesn't work for me. But I'm trying to use old Mouse kext and new keyboard and trackpad kexts from Voodoo to have everything functional. Trackpoint will work, if you disable voodoo trackpad kext.

Anyways, with VoodooPS2 you will have "Change keyboard type" button in Keyboard settings, and it allows system to identify your layout properly. Should solve your problem.

[chris-uk]

Thanks - i hope to have some time to give SL a go in the next couple of weeks - hopefully i will then be able to contribute back to this post and help with the efforts to get everything working..

[Sebinouse]

Guys, I need some help with AppleHDA testing. .

Hey Silencer, I've being trying your kext for hours ... and I can't even get them to load properly !
I patched my DSDT with your HDEF
I removed VoodooHDA.kext
I added one of your kext (test1 or test2)
I ran pfix
... no sound on boot (no device at all)

I got some assertion errors :

Code: Select all

17/03/10 12:01:12	kernel	Sound assertion "0 == pathMap_aDriverInstance" failed in "/SourceCache/AppleHDA/AppleHDA-179.1.4/AppleHDA/AppleHDADriver.cpp" at line 1656 goto Exit

17/03/10 12:01:12	kernel	Sound assertion "0 != createAudioEngines ( fPathMap_aDriverInstance )" failed in "/SourceCache/AppleHDA/AppleHDA-179.1.4/AppleHDA/AppleHDADriver.cpp" at line 203 goto Exit

And then I noticed with lspci that my DeviceId was

Code: Select all

00:1b.0 Audio device [0403]: Intel Corporation 82801G (ICH7 Family) High Definition Audio Controller [8086:27d8] (rev 02)

And yours seems to be 8086:27BD ... Am I right ?
So I updated your DSDT patch with my deviceid ... but so far no improvements ...

I am going to investigate with the "Patch Apple HDA" thread of projectx86 ...

[Silencer]

And yours seems to be 8086:27BD ... Am I right ?
So I updated your DSDT patch with my deviceid ... but so far no improvements ...
I am going to investigate with the "Patch Apple HDA" thread of projectx86 ...

Did you check System Profiler under Audio? You should see some connections there.

The main thing that matters is your codec. I believe you should have exactly the same as me, so the device-id does't matter. I've actually made some progress since I've posted the request, and I now can have vanilla AppleHDA sound after Sleep Getting closer to solving this thing.

One thing that is essential for making audio work - Audio Codec Linux dump. I will post some extended guide on this, but if you can, please try the following:
1. Download Linux Mint installation from http://www.linuxmint.com/
2. Download UNetbootin tool from http://unetbootin.sourceforge.net/
3. Have USB flash at least 2GB prepared
4. Run UNetbootin, choose downloaded Linux image and install it on your USB flash
5. Boot from that USB flash to Linux
6. Run the following command from Terminal: cat /proc/asound/card0/codec#0 > ~/Desktop/codec_dump.txt
7. You will see codec_dump.txt file on your Linux desktop. Save it somewhere and send to me

[Blue Thunder]

Ah.....

Since I've already got Linux Mint installed on my T60p's hard drive, could I just skip to step 6?

EDIT: Booted into Mint and got the codec dump, just not entirely sure how to get it to you. Perhaps one of those file sharing places might do.
Ok got it uploaded here: <gone!> <-- not completely..

[Silencer]

EDIT: Booted into Mint and got the codec dump, just not entirely sure how to get it to you. Perhaps one of those file sharing places might do.
Ok got it uploaded here: http://www.mediafire.com/?zz0jj05mnd2

Great, send it to silencer at parks dot lv, and I will check if your codec is compatible with what I'm doing.

[Blue Thunder]

Hahaha, ok than.

EDIT: Sent.

[Silencer]

Hahaha, ok than.
EDIT: Sent.

I've checked it, and it is 100% like mine, so you will be able to use my AppleHDA legacy kexts without any changes. At least I hope so. I'm very close to getting it work 100%, so will post soon.

[tearme]

Hey,Silencer.I have tested your kext of LegacyAD1981HD.After I placed test1,The icon of volume truned grey.The system log printed result as same as Sebinouse's.
Mar 15 17:03:12 localhost kernel[0]: Kernel is LP64
Mar 15 17:03:12 localhost kernel[0]: Sound assertion "0 == pathMap_aDriverInstance" failed in "/SourceCache/AppleHDA/AppleHDA-179.1.4/AppleHDA/AppleHDADriver.cpp" at line 1656 goto Exit
Mar 15 17:03:12 localhost kernel[0]: Sound assertion "0 != createAudioEngines ( fPathMap_aDriverInstance )" failed in "/SourceCache/AppleHDA/AppleHDA-179.1.4/AppleHDA/AppleHDADriver.cpp" at line 203 goto Exit
In System Profiler I only saw Line-Out and Internal Microphone,not S/PDIF Out.
When I placed test2,I couldn't boot.
My T60 type is 2007E85 which uses AD1981 codec.P.S is my codec dump.Hope you help my sound go well.
http://rapidshare.com/files/364592474/AD1981HD.rar.html

[Silencer]

Hey,Silencer.I have tested your kext of LegacyAD1981HD.After I placed test1,The icon of volume truned grey.The system log printed result as same as Sebinouse's.
Mar 15 17:03:12 localhost kernel[0]: Kernel is LP64

I've checked your dump, and it is again exactly like mine, and you should see different error in your log, and you should be able to see S/PDIF Out. This is really weird. It looks like I'm forgetting something.

[Silencer]

Please try the latest version: LegacyAD1981HD.Test3.zip

Install it to /E/E and run pfix to regenerate all caches (don't use -s param). After you reboot, you should see that sound looks active, there should be Internal Microphone, Internal Speakers and Digital Out in your devices in Sounds settings. But there will be no sound. Then go to Sleep, wake up - test sound again, it should work then. Let me know the results.

[tearme]

Please try the latest version: LegacyAD1981HD.Test3.zip

Install it to /E/E and run pfix to regenerate all caches (don't use -s param). After you reboot, you should see that sound looks active, there should be Internal Microphone, Internal Speakers and Digital Out in your devices in Sounds settings. But there will be no sound. Then go to Sleep, wake up - test sound again, it should work then. Let me know the results.

I'm sorry to trouble you again.For some reasons,maybe as you know,Chinese Internet users hardly visit meidafire.Last time the kexts was transmited from my friend in Japan.Could you upload kext on rapidshare once more, which is easy to download.Thanks a lot.

[Sebinouse]

Here is my X60t audio dump : codec_dump

Hope I can use the same LegacyKext .... even if there are many differences with BT's dump :

Code: Select all

14,17c14,17
<   IO[0]: enable=0, dir=0, wake=0, sticky=0, data=1, unsol=0
<   IO[1]: enable=0, dir=0, wake=0, sticky=0, data=0, unsol=0
<   IO[2]: enable=0, dir=0, wake=0, sticky=0, data=0, unsol=0
<   IO[3]: enable=0, dir=0, wake=0, sticky=0, data=1, unsol=0
---
>   IO[0]: enable=0, dir=0, wake=0, sticky=0, data=1
>   IO[1]: enable=0, dir=0, wake=0, sticky=0, data=1
>   IO[2]: enable=0, dir=0, wake=0, sticky=0, data=1
>   IO[3]: enable=0, dir=0, wake=0, sticky=0, data=1
47c47
<   Amp-Out vals:  [0x33 0x33]
---
>   Amp-Out vals:  [0x1f 0x1f]
51c51
<   Pin Default 0xc4014110: [Both] Line Out at Ext Right
---
>   Pin Default 0xc3014110: [Both] Line Out at Ext Left
86c86
<   Pin Default 0xc4a1502e: [Both] Mic at Ext Right
---
>   Pin Default 0xc3a1502e: [Both] Mic at Ext Left
107c107
<   Pin Default 0x414411f0: [N/A] SPDIF Out at Ext Rear
---
>   Pin Default 0x014411f0: [Jack] SPDIF Out at Ext Rear
150c150
<      0x0d 0x0e 0x0f 0x10 0x13 0x14 0x15 0x16 0x17 0x18 0x19 0x1a 0x1d
---
>      0x0d* 0x0e 0x0f 0x10 0x13 0x14 0x15 0x16 0x17 0x18 0x19 0x1a 0x1d
175c175
<   Amp-Out vals:  [0xbf 0xbf]
---
>   Amp-Out vals:  [0x00 0x00]
209c209
<   Amp-Out vals:  [0x19 0x19]
---
>   Amp-Out vals:  [0x16 0x16]

btw no improvement with your third kext and I only have this on the system profiler :

Code: Select all

Audio haute définition Intel :

  Identifiant du périphérique :	0x17AA2010
  Identifiant audio :	0

Thanks

[Blue Thunder]

Hmmm I've also got an X60T, maybe I'll get an audio dump from that too.

[Silencer]

btw no improvement with your third kext and I only have this on the system profiler

The profiler information is wrong, and this is first point where it breaks for you. Please show me your original HDEF and patched HDEF from DSDT. I think the problem is there.

[Silencer]

Bollocks, I've just spotted the problem with the DSDT patch for HDEF I posted. My fault guys. Here's the correct one:

Code: Select all

Device (HDEF) 
{
    Name (_ADR, 0x001B0000)
    Name (_S3D, 0x03)
    Name (RID, Zero)
    Name (_PRW, Package (0x02)
    {
        0x05, 
        0x04
    })
    Method (_PSW, 1, NotSerialized)
    {
        Noop
    }
    Method (_DSM, 4, NotSerialized)
    {
        Store (Package (0x06)
        {
            "built-in",
            Buffer (One) {
                0x00
            },                     
            "layout-id", 
            Buffer (0x04)
            {
                0xBD, 0x07, 0x00, 0x00
            },                  
            "PinConfigurations", 
            Buffer (One)
            {
                0x00                        
            }                     
        }, Local0)
        DTGP (Arg0, Arg1, Arg2, Arg3, RefOf (Local0))
        Return (Local0)
    }
}

Notice the Store (Package (0x06) as opposed to what I had before Store (Package (0x08). BTW, you only need to add _DSM method to your HDEF device. If you have original code a bit different from the one I have, copy/paste the only _DSM method.

[Blue Thunder]

I keep getting an error when I try to compile my dsdt.aml with the HDEF code that you posted above.
Heres the readout from Terminal:

Code: Select all

dsdt.dsl 14210: [*** iASL: Read error on source code temp file dsdt.src ***]
Error    4095 -^ syntax error, unexpected $end

ASL Input:  dsdt.dsl - 14211 lines, 483640 bytes, 5696 keywords
Compilation complete. 1 Errors, 0 Warnings, 0 Remarks, 0 Optimizations

It only occurs after I add your new code in, whether it be a full copy/paste or just adding the _DSM Method bit.


EDIT: I think I found the problem. I think you've missed a curly bracket at the end of your code. The code below compiled for me without any errors.

Code: Select all

Device (HDEF)
{
    Name (_ADR, 0x001B0000)
    Name (_S3D, 0x03)
    Name (RID, Zero)
    Name (_PRW, Package (0x02)
    {
        0x05,
        0x04
    })
    Method (_PSW, 1, NotSerialized)
    {
        Noop
    }
    Method (_DSM, 4, NotSerialized)
    {
        Store (Package (0x06)
        {
            "built-in",
            Buffer (One) {
                0x00
            },                     
            "layout-id",
            Buffer (0x04)
            {
                0xBD, 0x07, 0x00, 0x00
            },                 
            "PinConfigurations",
            Buffer (One)
            {
                0x00                       
            }                     
        }, Local0)
        DTGP (Arg0, Arg1, Arg2, Arg3, RefOf (Local0))
        Return (Local0)
        }
    }
}

[Sebinouse]

New DSDT works fine for me without compiling error ...

Here is my new System Profile ...

Code: Select all

Audio haute définition Intel :

  Identifiant du périphérique :	0x17AA2010
  Identifiant audio :	1981
  Appareils disponibles :
  Haut-parleur :
  Connexion :	Interne
  Microphone interne :
  Connexion :	Interne
  Sortie S/P-DIF :
  Connexion :	RCA

But sound quality is very poor (lag and distorsion ...)

[Silencer]

New DSDT works fine for me without compiling error ...
But sound quality is very poor (lag and distorsion ...)

Ok, that's already better, but need more details:
1. Does sound work for you straight after you boot your ThinkPad or do you need to go to Sleep before it works?
2. Any changes in sound quality after you go to sleep and wake up?
3. Does microphone work for you?
4. In Sound preferences, which devices you have on your Outputs tab?

[Sebinouse]

1. You're right I need to go to sleep to get sound ! ...
2. No still very poor after several Sleep/Wake
3. No
4. "IntSpeakers DevShortNameOut" and "DigitalOut DevShortNameSpdo" ...

I have no time to investigate this we ... I will try to be more precise on monday ....

[Blue Thunder]

Hmmm well I don't see any input or output devices in the Sound Preferences window.
I'm not entirely sure if my computer is able to sleep either (sleep LED doesn't come on, but it seems to be putting the HDD into park mode).

I also seem to be experiencing the restart error (using AsereBLN) - basically the computer closes itself down for a restart but never starts up.. or properly shutdown. It seems to restart properly when using the boot file from PCEFI v10.6.

[Silencer]

Hmmm well I don't see any input or output devices in the Sound Preferences window.
I'm not entirely sure if my computer is able to sleep either (sleep LED doesn't come on, but it seems to be putting the HDD into park mode).

Here are two basic ways to find out what is currently going wrong:
1. If you don't see any connections in System Profiler for your Intel audio, that means DSDT is wrong OR some info in legace HDA kext is missing. If there's no built-in audio in System Profiler - that means there's no HDEF device in DSDT.
2. If you don't see any inputs or outputs in System Preferences, that means the driver failed to load, and you will see an error in your 'dmesg' or system.log. Something in the lines of "sound assertion, failed to create audio engines". The two lines above that will explain what exactly is wrong. If there are messages about wrong "path", that means legacy kexts needs editing.

I also seem to be experiencing the restart error (using AsereBLN) - basically the computer closes itself down for a restart but never starts up.. or properly shutdown. It seems to restart properly when using the boot file from PCEFI v10.6.

In that case you should stick to PCEFI for now. AsereBLN is very tricky with its restart fix. You can try to disable it in 'com.apple.Boot.plist' with RestartFix=no, it might make it work for you.

[Blue Thunder]

Right, System Profiler detects my "Intel high definition audio" and gives a device ID and audio ID, but no available devices.
System Preferences are empty for both input and output devices.

Found a few sound assertion errors via the "Console" application.

Code: Select all

Sound assertion "32 < fNumProcessingCoefficients" failed in "/SourceCache/AppleHDA/AppleHDA-174.1.1/AppleHDA/AppleHDAWidget.cpp" at line 308 goto Exit

Code: Select all

Sound assertion "0 != result" failed in "/SourceCache/AppleHDA/AppleHDA-174.1.1/AppleHDA/AppleHDAFunctionGroup.cpp" at line 116 goto Exit

Code: Select all

Sound assertion "0 != err" failed in "/SourceCache/AppleHDA/AppleHDA-174.1.1/AppleHDA/AppleHDACodecGeneric.cpp" at line 205 goto ExitError

I've got no idea what that all means. As far as I can tell these are the only errors that show up and they appear to be different from tearme and Sebinouse's errors. I wonder if its because I added an extra curley bracket to the HDEP code...

And it also seems that PCEFI v10.6 won't restart either and has actually gotten worse. It now hangs at a blue screen after I click on the restart button - interestingly AsereBLN does the exact same thing when I disable the restart fix in com.apple.boot.plist. For now I'll use AsereBLN with the restart fix enabled.

EDIT: The 'Console' application reports that sleep does work. It just doesn't turn off the screen, if the lid is open
Oh and another thing, any of you experiencing mouse tearing? (Nevermind, this is a known issue which apparently will involve patching framebuffers for 32bit operation and drivers/new framebuffers for 64bit operation).