OSX on DELL Vostro 3450 / Inspiron N4110 / XPS L702x UEFI Clover

[TimeWalker75a]

Probably same thing as after certain OSX updates - AppleHDA is not included in the kernerlcache and all the patches that are being applied to it therefore fail.

Try running KextUtility (installed from the support package) which will trigger kernelcache to rebuild. For camera use iGlasses demo from the support package.

 

P.S. You don't need to disable anything in regards to graphics in your BIOS as this 1. has no effect and 2. is done on software level.

 

P.S.S. You don't need any supplemental SSDTs for your setup (which for some odd reason is your DSDT, not an actual SSDT), it's no different from the rest of us here. 

[iajheyst]

Probably same thing as after certain OSX updates - AppleHDA is not included in the kernerlcache and all the patches that are being applied to it therefore fail.

Try running KextUtility (installed from the support package) which will trigger kernelcache to rebuild. For camera use iGlasses demo from the support package.

 

P.S. You don't need to disable anything in regards to graphics in your BIOS as this 1. has no effect and 2. is done on software level.

 

P.S.S. You don't need any supplemental SSDTs for your setup (which for some odd reason is your DSDT, not an actual SSDT), it's no different from the rest of us here. 

I solved the problem with the sound. I installed the ALC269 driver from here, and everything works perfectly now on 10.9.2.

Thank you very much for your advice.

[TimeWalker75a]

I solved the problem with the sound. I installed the ALC269 driver from here, and everything works perfectly now on 10.9.2.

Thank you very much for your advice.

You solved it up until next update that updates AppleHDA hits. Then your sound will be broken again.

[iajheyst]

You solved it up until next update that updates AppleHDA hits. Then your sound will be broken again.

I know, so I won't update any more. This kext works perfectly with the 10.9.2 Mavericks image, downloaded from the AppStore 2 days ago.

[TimeWalker75a]

That's an odd direction to pursue. 

[l_zhang]

@TimeWalker75a, could you plz tell me that how did you make your modified BIOS? Did you use PMPatch to make it?

My box is Inspiron 14R 7420 SE, and it also encounters the problem of KP when using native AICPUPM.kext because of the locking MSR 0xE2. Maybe, a modified BIOS will help me, I think, but how can I obtain it? Make it myself with PMPatch or find it somewhere?

[TimeWalker75a]

I've made it manually. And proposed to CodeRush the way to unlock Advanced Setup hidden entries on Tiano 2.0 (second post in PMPatch thread). As your Inspiron is a generation higher the BIOS is RSA protected, so you can't really patch it unless you desolder the chip and reprogram it externally. Better off use AsusAICPM pathc integrated in Clover. If you are not using Clover I suggest you switch to it the sooner the better.

[l_zhang]

@TimeWalker75a, I am using Clover actually now, switching from Chameleon recently.  

I set AsusAICPM to true in my config.plist and restore native AppleIntelCPUPowermanegement.kext and AppleIntelCPUPowermanegementClient.kext in SLE and then reboot, no Kernel Panic any more.

With ioreg, I found that AppleLPC.kext and AppleIntelCPUPowerManagement*.kext are loaded, however, when I try to 'sleep' my box, it seems that the only thing that have happened is the screen turns off. The fan is still running, and the HDD is still running. It this a correct condition?

[TimeWalker75a]

It's not supposed to be like that, your LPC bridge is most likely not supported by default, so you have to mask it with a compatible supported ID. But as this thread is not a support thread for Inspiron 7420 SE I can't comment much here because it's most definitely off topic here. Your laptop is generations above the one discussed, so nothing (except installing the OS) applies.

[akarichn]

Apologies for not being able to test the resolutions much recently, late winter always seems to get quite a bit busier.

 

I've found that, no matter what monitor frequency I use (I've tried 56.9, 55.9, 54.9, 53.9,... all the way down to 45.9) and no matter what, the system produces artifacts. They appear to be coming from the software rather than the hardware₁, since I am able to take screenshots of the artifacts. I've recently bought a second 4GB RAM module and I'm waiting for that to come in, but I'm not sure how much it'll help.

 

Also, sound has been acting odd again. It feels like a game of chance from whenever my system wakes up that the sound card remains awake, too. It will always start out from waking up having sound, but the sound seems to go into sleep mode if there is no sound for a couple seconds. I have CodecCommander.kext installed via Kext Helper, along with the patched AppleHDA.

 

₁ - Not to say I have any knowledge on if this is the case, but it'd make for an interesting hardware issue if it was.

[webcivilian]

Timewalker75a - I'm trying to create a Clover package for L702X similarly as you but I seem to hit a wall. I took your AMD SSDT-2 file and modified it match our system but for some reason:

 

1. It does not seem to attach. Clover boot logs says it added the SSDT but I don't see any of the devices properties available under IORegistery explorer. OSX boots with DSDT patches for L702X but SSDT won't attach. 
2. Is there somethings to do with the external method and device references?

Appreciate your help in advance.

[TimeWalker75a]

Please note that original DSDT retains original device names, however they get renamed via clover acpi patches. Whilst SSDT-2 uses renamed name entries to extend the original DSDT table. This is probably the reason your SSDT has no effect.

[webcivilian]

I will check the config.plist file to see I'm applying the correct patches to ACPI renames. I also noted that Clover patches some of the devices on its own. Do you think that might be a problem?

 

Thanks for your help.

[webcivilian]

@ TimeWalker75a - I found the issue. If I don't drop the nvidia optimus SSDT table, SSDT-2 will not attach. It only works if I drop the optimus SSDT. Any clues as to why?

[TimeWalker75a]

Nvidia table has _DSM defined for IGPU.

[webcivilian]

@ TimeWalker75a - thanks, that fixed it. Had to drop IGPU _DSM from nvidia table and reattach it as SSDT-3. Maybe there's another way to drop the DSM without extracting/removing DSM/reattaching the SSDT.

[TimeWalker75a]

That's the reason I use different SSDT-2 for AMD laptops. Clover's Intel injector in conjunction with Intel HDMI injection is enough for getting the HD3000 up, so _DSM in AMD SG table can stay original.

[webcivilian]

@Timewalker75a - Few things I notice:

 

1. Sleep takes awhile, could be due to hibernate file. I haven't disabled it.

2. Restart seems to be broken. It shuts down every time. Do you know how to fix this?

 

Thanks again.

[TimeWalker75a]

@Timewalker75a - Few things I notice:

 

1. Sleep takes awhile, could be due to hibernate file. I haven't disabled it.

2. Restart seems to be broken. It shuts down every time. Do you know how to fix this?

 

Thanks again.

1. 25 second delay prior to actually entering sleep state is Mavericks' general feature

2. Make sure to include SLPE DSDT fix via Clover patches if you don't want any restart or shutdown related unexplained nonsense. 

[webcivilian]

This was working with patched DSDT:

Method (_PIC, 1, NotSerialized)
    {
        Store (Arg0, GPIC)
        Store (Arg0, PICM)
    }

    Method (_PTS, 1, NotSerialized)
    {
        Store (Zero, P80D)
        P8XH (Zero, Arg0)
        Store (One, MY9F)
        If (LEqual (Arg0, 0x03))
        {
            If (LEqual (USBK, One)) {}
            If (LAnd (DTSE, LGreater (TCNT, One)))
            {
                TRAP (TRTD, 0x1E)
            }
        }

        If (LOr (LEqual (Arg0, 0x03), LEqual (Arg0, 0x04)))
        {
            If (LEqual (Arg0, 0x04))
            {
                Acquire (\_SB.PCI0.LPCB.EC0.MUT0, 0xFFFF)
                \_SB.PHSR (0xB2)
                Release (\_SB.PCI0.LPCB.EC0.MUT0)
            }

            Store (One, SLPO)
        }

        Store (Zero, JPNT)
        If (LOr (LEqual (Arg0, 0x03), LEqual (Arg0, 0x04)))
        {
            If (\_SB.PCI0.RP05.PDSX)
            {
                P8XH (Zero, Arg0)
                Store (0xAA, JPNT)
            }
        }
    }

How is this different that what you are doing in the Clover patch?

[TimeWalker75a]

You weren't booting in UEFI previously, weren't you? You don't have the fix in place.

In SCT Tiano 2.0 UEFI a crucial piece of the puzzle is missing - pp. 529 section 13.8.3.7 SMI_EN—SMI Control and Enable Register

http://www.intel.com/content/dam/www/public/us/en/documents/datasheets/6-chipset-c200-chipset-datasheet.pdf

The PM Base + 30 is even defined in ACPI but never used .. Name (PM30, 0x0430) 

 

Dell's retarded UEFI firmware never resetes it properly, so you have to do it via ACPI. We should be happy that at least legacy BIOS protocol at least performs this on this firmware.

[webcivilian]

I was booting in Clover UEFI but was using a patched DSDT. It was working until I dropped the nvidia table.

 

I added your clover patch "B1B2+PM30+SPLE" seems to be working now.

 

----

 

As everything is working do you want the config.plist and ssdt files? I was hoping you could add those to the package you created for L702X and L502X. The following have changed.

 

1. Config.plist
2. SSDT-ITL-NV
3. SSDT-NV
4. AppleHDA platforms and layout files.
5. CustomPeripheral.

Thats all I think I will double check in case I missed something before sending them to you.

[TimeWalker75a]

That is the correct result you should be seeing in ACPI and it's required for shutdown to work.

You shouldn't be dropping nVidia table, leave it alone and let it live. You need it for nVidia to disable properly. What you shouldn't do is add any custom _DSM injection to IGPU, rely solely on Clover's Intel GFX injection instead.

[webcivilian]

But your SSDT has a _DSM and nvidia table has a _DSM for IGPU. So you are saying that don't use either? I manually removed the _DSM from nvidia table and added the SSDT back. Check my earlier post with SSDT names.

[TimeWalker75a]

I have already pointed out that for laptops with AMD in this guide we use different SSDT-2 that has no _DSM injection on IGPU. My laptop is an Intel-only model which allows me to use _DSM.