iMac Pro X299 - Live the Future now with macOS 10.14 Mojave [Successful Build/Extended Guide]

[kgp]

I've been also having sleep issues, apparently caused by GBE and only happening after updating to 10.14.4. I'm waiting for some new monitors to come up for thunderbolt, I'll gladly help debugging once I decide on one

I know, but this is about wake issues with TB devices connected to the TTR during sleep. Everything else works flawless in my case. While on the Deluxe wake from sleep did not work at all with TB devices connected to the TTR during sleep, remaining wake issues with TB devices connected to the TTR during sleep on the Sage 10G rather seem TB device specific/dependent.

Your future help is be appreciated in any case.

 

[vincek8s]

A little update on my side, I just tried to swap out the RX 560 for an old Titan (Kepler) 6GB. Everything work flawlessly with WhateverGreen except for hardware encoding. I'm still not sure if I'm going to keep this for a few more months or go back to the RX. All while waiting for Navi or for Radeon VII prices to normalize.

@kgp, do you have any ideas on why sleep is stopped by GBE?

2019-04-06 09:11:31.451121-0700 localhost kernel[0]: (AppleACPIPlatform) AppleACPIPlatformPower Wake reason: GBE1
2019-04-06 09:11:31.451121-0700 localhost kernel[0]: (AppleACPIPlatform) AppleACPIPlatformPower Wake reason: GBE1

I patched the graphics card and I'm using the official SmallTree driver, not sure if this is also a problem for you, might be worthwhile to check.

 

[kgp]

A little update on my side, I just tried to swap out the RX 560 for an old Titan (Kepler) 6GB. Everything work flawlessly with WhateverGreen except for hardware encoding. I'm still not sure if I'm going to keep this for a few more months or go back to the RX. All while waiting for Navi or for Radeon VII prices to normalize.

@kgp, do you have any ideas on why sleep is stopped by GBE?

2019-04-06 09:11:31.451121-0700 localhost kernel[0]: (AppleACPIPlatform) AppleACPIPlatformPower Wake reason: GBE1
2019-04-06 09:11:31.451121-0700 localhost kernel[0]: (AppleACPIPlatform) AppleACPIPlatformPower Wake reason: GBE1

I patched the graphics card and I'm using the official SmallTree driver, not sure if this is also a problem for you, might be worthwhile to check.

Well.. honestly I am not able to deduce much from the information provided above, sorry.

I suppose that you patched the EEPROM of the X550-AT2 and not your graphics card though and basically your system wakes due to some GBE event and sleeps again subsequently, if I understood correctly.

Maybe a CMOS reset or NVRAM reset might remove this minor issue. Before applying a CMOS reset save your BIOS settings in any case for simply restoring them subsequently.

 

[vincek8s]

Well.. honestly I am not able to deduce much from the information provided above, sorry.

I suppose that you patched the EEPROM of the X550-AT2 and not your graphics card though and basically your system wakes due to some GBE event and sleeps again subsequently, if I understood correctly.

Maybe a CMOS reset or NVRAM reset might remove this minor issue. Before applying a CMOS reset save your BIOS settings in any case for simply restoring them subsequently.

Tried all of the above, after some digging I found out that the Bluetooth USB is the cause of it. Still investigating why, but if you have any ideas lemme know.

@izo1, given that I'm using your same card did you do anything special for sleep to work?

 

[izo1]

Tried all of the above, after some digging I found out that the Bluetooth USB is the cause of it. Still investigating why, but if you have any ideas lemme know.

@izo1, given that I'm using your same card did you do anything special for sleep to work?

Not really doing anything special on my end. Sleep works for me, I just don’t use it because macOS keeps ejecting/mounting my external drives upon wake so I just always leave it on.

How are you doing your USB ports? Port Limit patch/kext/other?

I’m using 255 (internal) for HS12 (internal USB2.0 that the BT USB is connected to) with a custom USB kext.

 

[ModMike]

If any of you are interested in getting Fan RPM readings and more, I just updated the sensors in VirtualSMC. Check it out here and let me know how it works for you.

 

[kgp]

If any of you are interested in getting Fan RPM readings and more, I just updated the sensors in VirtualSMC. Check it out here and let me know how it works for you.

Awesome work, @ModMike! That's almost everything everybody was waiting for two finally switch from FakeSMC.kext to VirtualSMC.kext.

Here are my results for the i9-7980XE and the ASUS WS X299 Sage 10G.

Some things are missing though in my personal opinion:

1.) The interface of iStatMenu needs to be properly adapted by bjango to also properly display all HWSensor.info with iStatMenus, which currently is not the case:

Please also compare with iStatMenus HWSensor Information when using the modified Kozlek/Interferenc FakeSMC.kext and HWSensor-Kext distribution:

You could write to bjango on twitter, ask him to do so and pass him all necessary information he needs. It would be simply awesome if one could also use iStatMenus with VirtualSMC.kext and your HWSensor-Kexts instead of HWMonitorSMC2.

Before doing so, one might however consider points 2.), 3.) and 4.) below at first place:

2.) Additional CPU Frequency and Core Ratio information with VirtualSMC.kext and your HWSensor-Kexts like in case of the modified Kozlek/Interferenc FakeSMC.kext and HW Sensor-Kext distribution would be of benefit for all users.

3.) In line with 2.) we would still also need a GPU HWSensor.kext to display at least the GPU Temps and if possible also GPU frequencies of all available AMD and Nvidia GPUs.

4.) Finally it would be nice to also see some RAM AND PSU Voltage/Watt information when using VirtualSMC.kext and your HWSensor-Kext distribution.

In any case, congrats to this major step ahead in using VirtualSMC.kext without loosing all necessary HWSensor information.

Cheers,

KGP

 

[ModMike]

Awesome work, @ModMike! That's everybody was waiting for two finally switch from FakeSMC to VirtualSMC.kext

Here are my results for the i9-7980XE and the ASUS WS X299 Sage 10G.

View attachment 397944

Two things are missing though in my personal opinion:

1.) The interface of iStatMenu needs to be properly adapted by bjango to also properly display all HWSensor.info with iStatMenus, which currently is not the case:

View attachment 397945

You could write to bjango on twitter, ask him to do so and pass him all necessary information he needs. It would be simply awesome if one could also use iStatMenus with VirtualSMC.kext and your HWSensor-Kexts instead of HWMonitorSMC2.

2.) We also would need a GPU HWSensor.kext to display at least the GPU Temps of all available AMD and Nvidia GPUs.

In any case, congrats to this major step ahead in using VirtualSMC.kext without loosing the necessary HWSensor information.

Cheers,

KGP

Glad it worked! I am a little disappointed we can't get individual core frequencies and a have the kind of flexibility FakeSMC has in configuring HWMonitorSMC2.

Great idea on the graphics cards but I'm not sure I have the chops to add them and I think the IGPU depends on Intel Power Gadget APIs. I will certainly take a look though!

I don't see as single Graphics card in your list, did you install Intel Power Gadgets and configure HWMonitorSMC2 to use them? Here is what I get:

 

[kgp]

Glad it worked! I am a little disappointed we can't get individual core frequencies and a have the kind of flexibility FakeSMC has in configuring HWMonitorSMC2.

Great idea on the graphics cards but I'm not sure I have the chops to add them and I think the IGPU depends on Intel Power Gadget APIs. I will certainly take a look though!

I don't see as single Graphics card in your list, did you install Intel Power Gadgets and configure HWMonitorSMC2 to use them? Here is what I get:

View attachment 397953

1.) To also display CPU frequencies and core ratios might require just a few lines of additional coding. I guess one could easily implement the latter based on @interferenc's modifications in CPUSensor.cpp of the original Kozlek FakeSMC/HWSensor source code distribution. See my post #21 . You could ask @VIT9696 to do so.

2.) Concerning the missing GPU information you mean this? I have IPG installed and now also checked IPG in the HWMonitorSMC2 settings. Do I also need to check "Use the IOAccelerator's monitoring for GPUs" to see the respective AMD and Nvidia GPU information, which is still missing in case of the water blocked Nitro+ RX Vega 64 that I am still employing on my X299 system?

Once the above is clarified, I would like to verify the current Virtual SMC and HWSensor distribution also on my X99 system with the air flow Radeon VII. It would be important to see if the primer also supports X99 and Broadwell-E/Haswell-E HWSensor information. But later, by chance I woke up at 6 am this Sunday morning and it is still just 8 am, when writing the above. X99 and Broadwell-E/Haswell-E needs to be discussed in the respective X99 thread in any case.

 

[kgp]

@ModMike,

Perfect! After checking/enabling "Intel Power Gadget" and "Use the IOAccelerator's monitoring for GPUs" in the HWMonitorSMC2 settings, there is now also the entire water blocked Nitro+ RX Vega 64 GPU information in HWMonitorSMC2 (Total Power info seems wrong though):

Of course, with the above solution no GPU Temps with iStatMenus though: