[Guide] HP Elite 8300 & 6300 Pro (all form factors) using Clover UEFI hotpatch

[LisaAlterEgo]

Sniki,

Thanks for the quick answers. I've tested some things and have some feedback.

I am successfully using the USBPorts.kext, with a redefined profile for SMBIOS defined as iMacPro1,1.
I changed MacMini6 to iMacPro1 everywhere in the info.plist as you suggested. I also changed the port
type for HP17 to portType 0 and added HP15 portType 0, to correspond to what I see plugging a media
reader in Media 1 or Media 2 on the mobo.

Could I ask, with your original, you have 3 USB2 ports on EHC1, 6 USB2 ports on EHC2, and 8 USB2/USB3
ports on XHC, for a total of 17 ports. Do we need a port limit patch?

I also had difficulty getting the mic to work for Siri. I ended up modifying your earlier SSDT to include only the CodecCommander part and installing it in ACPI patched, and reinstalled CodecCommander.kext to get it working
again. With CodecCommander and the SSDT, IORegistryExplorer shows me an RMCF.cache under HDEF, which
isn't there otherwise.

I also tried installing all the kexts in Clover/Kexts/Other, versus installing most of them in /Library/Extensions.
I see slightly cleaner boot graphic switching with the new preferred method, but I couldn't get the mic working.

The Idiot's Guide to Lilu and its Plugins seems to say that Lilu and WhateverGreen are essential kexts to go in
Clover/Kexts/Other and calls AppleALC a non-essential kext to go in /Library/Extensions, but when I did that
Hackintool would give me an error message that AppleALC's dependencies were missing when I rebuilt the cache.

Finally, SSDTprgen gives an error message when I try to generate a power management SSDT using the iMacPro1,1
definition. I'm not sure which of the flags to invoke, and it didn't seem to be working right with my attempts. I'm still
using XCPM for now.

I like the iMacPro1,1 SMBIOS because VideoProc gives me higher resolution scores, and because I can get HVEC
and AppleTV working with an i5 (HD2500) and AMD RX560 graphics. I dont seem to need the -cdfon or the
-igfxnohdmi boot flag, or the Devices/Properties/PciRoot(0)/Pci(0x02,0) definition of AAPL,ig-platform-id for
headless operation, I simply disable the IGPU in BIOS.


Thanks again,
Lisa

 

[Sniki]

I am successfully using the USBPorts.kext, with a redefined profile for SMBIOS defined as iMacPro1,1.
I changed MacMini6 to iMacPro1 everywhere in the info.plist as you suggested. I also changed the port
type for HP17 to portType 0 and added HP15 portType 0, to correspond to what I see plugging a media
reader in Media 1 or Media 2 on the mobo.

Great job on that.

Could I ask, with your original, you have 3 USB2 ports on EHC1, 6 USB2 ports on EHC2, and 8 USB2/USB3
ports on XHC, for a total of 17 ports. Do we need a port limit patch?

Yes that's true but we don't need a port limit patch for that simply because:
The limit patch is actually only applicable for XHC/XHCI Controllers, so in our case we only have 8 XHC Ports:
SS01 to SS04 = 4 x in two modes: USB2x4 and USB3x4 = 8 XHC Ports.
3 ports are in EHC1 which counts only 3 for that controller and 6 on EHC2 controller which counts only 6.
The total needs to be counted on Haswell+ generations where everything is on XHC Controller as most of these and above do not have EHCI (EHC1 and EHC2) Controllers, so all the ports are routed on XHC but for us we are fine.

I also had difficulty getting the mic to work for Siri. I ended up modifying your earlier SSDT to include only the CodecCommander part and installing it in ACPI patched, and reinstalled CodecCommander.kext to get it working
again. With CodecCommander and the SSDT, IORegistryExplorer shows me an RMCF.cache under HDEF, which
isn't there otherwise.

Ok, thanks for pointing this out, i believe i should be able to fix the mic too without CodecCommander.
That's what has slipped from my side without being tested.
I believe there was input but didn't test the clarity if the mic was getting clear hearing input for Siri to work or not.
Will try by adding SignalProcessing elements into it, add provide a custom AppleALC.kext for you to test as well.

Then if it works fine without CodecCommander.kext i will send a pull request to update the current layout-id 11 patch that exists on AppleALC and in the meantime attach the one i built on the guide or in github so everyone of you can use it before next AppleALC release so it will be included on that.

I also tried installing all the kexts in Clover/Kexts/Other, versus installing most of them in /Library/Extensions.
I see slightly cleaner boot graphic switching with the new preferred method, but I couldn't get the mic working.

The Idiot's Guide to Lilu and its Plugins seems to say that Lilu and WhateverGreen are essential kexts to go in
Clover/Kexts/Other and calls AppleALC a non-essential kext to go in /Library/Extensions, but when I did that
Hackintool would give me an error message that AppleALC's dependencies were missing when I rebuilt the cache.

Sorry Lisa, im not sure if i understand your point correctly here:
So you see better cleaner boot graphic switching but the mic doesn't work with CodecCommander.kext into EFI/Clover/kexts/Other and you had to install it into Library/Extensions ?

Finally, SSDTprgen gives an error message when I try to generate a power management SSDT using the iMacPro1,1
definition. I'm not sure which of the flags to invoke, and it didn't seem to be working right with my attempts. I'm still
using XCPM for now.

Will give it a try for you, i don't have the same CPU but will look into the CPU data and see if SMBIOS is pointed somewhere in the SSDT or not, if it is, i believe just doing the replacement there as well should be enough/the same.

However XCPM only is perfectly fine as well, i just saw better geekbench score with SSDT.aml compared to XCPM, other than that it is perfect that way as well.

I like the iMacPro1,1 SMBIOS because VideoProc gives me higher resolution scores, and because I can get HVEC
and AppleTV working with an i5 (HD2500) and AMD RX560 graphics. I dont seem to need the -cdfon or the
-igfxnohdmi boot flag, or the Devices/Properties/PciRoot(0)/Pci(0x02,0) definition of AAPL,ig-platform-id for
headless operation, I simply disable the IGPU in BIOS.

Thanks for this as well,
Yes you are right, you don't need -cdfon and/or -igfxnohdmi since you don't use it for output and just hardware acceleration (if you were using an older Nvidia Graphics Card that works on Mojave+ as well.
Correct, those features are not needed on AMD cards as they do have those features itself now.

So this brings us to the point where the HD2500 headless may be necessary only for users with Nvidia cards like 710 etc.

 

[LisaAlterEgo]

Sorry Lisa, im not sure if i understand your point correctly here:
So you see better cleaner boot graphic switching but the mic doesn't work with CodecCommander.kext into EFI/Clover/kexts/Other and you had to install it into Library/Extensions ?

I see cleaner graphics switching with all the kexts in Clover/Kexts/Other, but the mic wasn't working that way.
I can recheck again.

 

[Asural]

I can recheck again.

Only when kexts is stored in Clover/Kexts/Other, I installed it newly on SSD and checked audio input/output.
Voice input is possible with Mojave, but voice input is not available with Catalina.

"config.plist" is just the state where SMBIOS is added to "config_HD2500.plist".

 

[Sniki]

I see cleaner graphics switching with all the kexts in Clover/Kexts/Other, but the mic wasn't working that way.
I can recheck again.

@LisaAlterEgo for me, microphones are working fine without codecCommander.kext, maybe it's the AppleALC.kext that i patched for auto switch on Audio Output, just give it a shot without codeccommander.kext and see if microphones work.
If still doesn't recognize, try speaking louder and check if it get's your word. if that works than it shouldn't be hard to increase the input volume in the kext. Give this one a try and let me know what are your results.

I will then build a custom one and send a patch request to Acidanthera AppleALC.

Thank you.

 

[LisaAlterEgo]

@LisaAlterEgo for me, microphones are working fine without codecCommander.kext, maybe it's the AppleALC.kext that i patched for auto switch on Audio Output, just give it a shot without codeccommander.kext and see if microphones work. If still doesn't recognize, try speaking louder and check if it get's your word. if that works than it shouldn't be hard to increase the input volume in the kext. Give this one a try and let me know what are your results.

Sniki,

I tried your revised AppleALC version, but it didn't fix the mic problem for me. I tried it with and
without CodecCommander. However, I found an odd work-around and fixed a problem I had with
copying your new instructions.

In my old config.plist, I had defined the LPC Controller, PciRoot(0x0)/Pci(0x1f,0x0), with

AAPL,slot-name Internal@0,31,0
device_type ISA bridge
model Q77 Express Chipset Chipset Family LPC Controller

It turns out the two new lines from your config.plist:

compatible pci8086,1e42
device-id 421E0000

are essential. Without those two lines, the device-id was set by Clover to be 471E0000, and
AppleLPC would not attach to the LPCB@1F device. As a result, sleep wasn't working.

After adding in the two lines, AppleLPC attaches and sleep works.

I also noticed the mic work-around. When I first turn on my machine, the mic doesn't work.
However, if i put it to sleep, wait til the three blinking lights, and re-awaken the machine,
the mic is working. I'm using CodecCommander 2.7.1, standard AppleALC 1.4.5, and I've
extracted your former CodecCommander SSDT section and put it in Clover/ACPI/patched.
I couldn't get it to work without CodecCommander.

Everything is now apparently working for me with all the kexts installed in Clover/Kexts/Other.
I noticed that it also works after sleep/wake with your revised AppleALC, but LineOut disappeared
as one of my output options when nothing is plugged into the jack.

When the kexts (including CodecCommander) are installed in /Library/Extensions, I don't need
the sleep/wake to get the microphone working.

Is there a modification of the CodecCommander SSDT that will wake it up properly from the
Clover/Kexts/Other folder?

Thanks,
Lisa

 

[Asural]

Is there a modification of the CodecCommander SSDT that will wake it up properly from the
Clover/Kexts/Other folder?

Thank you　@LisaAlterEgo.
With regard to audio input/output, it works properly only with the following modifications.

    <key>Devices</key>
    <dict>
        <key>Audio</key>
        <dict>
            <key>Inject</key>
            <integer>11</integer>
        </dict>
    </dict>

 

[LisaAlterEgo]

Thank you　@LisaAlterEgo.
With regard to audio input/output, it works properly only with the following modifications.

    <key>Devices</key>
    <dict>
        <key>Audio</key>
        <dict>
            <key>Inject</key>
            <integer>11</integer>
        </dict>
    </dict>

Are you sure that makes any difference in the operation of the microphone?

Sniki is already injecting the layout-id 11 in the config.plist in
Devices/Properties/PciRoot(0x0)/Pci(0x1b,0x0)

 

[Asural]

Are you sure that makes any difference in the operation of the microphone?

Sniki is already injecting the layout-id 11 in the config.plist in
Devices/Properties/PciRoot(0x0)/Pci(0x1b,0x0)

Is "Devices / Properties" deleted and does "Audio-> Injext = 11" work?

add ：
It doesn't say "Remove and use Devices->Properties", it's just a fact.
I do not know about detailed settings such as sound quality.

 

[LisaAlterEgo]

Is "Devices / Properties" deleted and does "Audio-> Injext = 11" work?

add ：
It doesn't say "Remove and use Devices->Properties", it's just a fact.
I do not know about detailed settings such as sound quality.

I believe that Devices / Properties is needed, as least as far as the LPC
controller definition goes, or the computer doesn't go to sleep.

To configure AppleALC for layout-id see:

Installation and usage

Native macOS HD audio for not officially supported codecs - acidanthera/AppleALC

github.com

It says:
"It is necessary to set value layout-id in Devices-Properties or boot arguments. ...
Set Devices → Audio → Inject = NO (Necessarily type string!)"

So Sniki's configuration of setting the layout-id in Devices / Properties is one
of the recommended methods. An alternative is to use the boot argument

alcid=11

The method you are talking about is an older method that is no longer recommended.