X299 - Open Core support

[joeblough]

i got mojave up and running on my Prime Deluxe-II. thanks to @djlild7hina for the github repo! made things very easy. the broadcom wifi/bt card swap was successful and although i have not tested handoff, unlocking with an apple watch works, and BT and wifi work for "regular" stuff - sound/network. so it probably works properly. icloud and imessage also seem to be working.

i do have a couple of questions if anyone knows the answers -

sleep works before and after a USB map kext (i guess there are fewer than 15 ports on each controller so USBInjectAll does not cause problems), but in both cases i get the "disk was not ejected properly" thing if a USB thumbdrive is inserted when i sleep/wake. i would have expected that to work after the USB mapping kext (hackintool) because it sets the sleep current. i understand the "eject" thing is a problem even on real macs sometimes, but i wonder if other ASUS x299 users are seeing this too. every single hack i've built in the last month has had this problem, which seems like i'm doing something wrong.

another weird thing is that if i plug in an anker USB-C hub to the internal motherboard port, it shows up as 5gb/sec in Hackintool. however, if i plug that same hub into either of the rear panel USB-C ports, it shows up as 480mbit/sec. i'm not sure if i have any device fast enough to figure out if the hub is actually running faster than that. i wonder if there's a BIOS setting somewhere that i messed up.

the github site seems to say that the smalltree kext is necessary for the 219V intel ethernet, but i found that i need IntelMausiEthernet. does that sound right?

thanks

 

[djlild7hina]

i got mojave up and running on my Prime Deluxe-II. thanks to @djlild7hina for the github repo! made things very easy. the broadcom wifi/bt card swap was successful and although i have not tested handoff, unlocking with an apple watch works, and BT and wifi work for "regular" stuff - sound/network. so it probably works properly. icloud and imessage also seem to be working.

i do have a couple of questions if anyone knows the answers -

sleep works before and after a USB map kext (i guess there are fewer than 15 ports on each controller so USBInjectAll does not cause problems), but in both cases i get the "disk was not ejected properly" thing if a USB thumbdrive is inserted when i sleep/wake. i would have expected that to work after the USB mapping kext (hackintool) because it sets the sleep current. i understand the "eject" thing is a problem even on real macs sometimes, but i wonder if other ASUS x299 users are seeing this too. every single hack i've built in the last month has had this problem, which seems like i'm doing something wrong.

another weird thing is that if i plug in an anker USB-C hub to the internal motherboard port, it shows up as 5gb/sec in Hackintool. however, if i plug that same hub into either of the rear panel USB-C ports, it shows up as 480mbit/sec. i'm not sure if i have any device fast enough to figure out if the hub is actually running faster than that. i wonder if there's a BIOS setting somewhere that i messed up.

the github site seems to say that the smalltree kext is necessary for the 219V intel ethernet, but i found that i need IntelMausiEthernet. does that sound right?

thanks

Intel 219 uses intelmausi, it’s the original deluxe with the 211 that needs the additional smalltree kext.

The usb 3 on the type c ports may be disabled in your USB port mappings? Does it happen with usbinjectall?

I’ve heard using the jettison app may help with the eject issue.

 

[joeblough]

Intel 219 uses intelmausi, it’s the original deluxe with the 211 that needs the additional smalltree kext.

The usb 3 on the type c ports may be disabled in your USB port mappings? Does it happen with usbinjectall?

I’ve heard using the jettison app may help with the eject issue.

ah ok - i keep forgetting that there could be big differences between deluxe and deluxe II.

yeah i had the same behavior when using USBInjectAll. i think i probably don't need the USB-C ports for USB, so it might work out OK even if they are messed up. but i want to try using them with a USB-C display and a DP->DP cable from the RX580 to the back panel. we'll see how that goes when i get the monitor.

i guess the only problem with jettison (i assume) is if you have encrypted volumes then they won't remount unless you store the password in your keychain. anyway i suppose it is worth a try, it only costs $4.

 

[patrickfogle]

If you don't have CPU overclocked, AVX instruction and all tweak parameters must be auto.

Ok, maybe there is some OC tweak going on....

You can set whatever (negative) AVX offsets you want.
Regarding the setting Sync All Cores, just make sure you have set a high enough CPU core voltage. I looked at your crash reports and at least from my own experience, the crashes related to "process name corresponding to current thread: photoanalysisd" were due to CPU not getting enough voltage.

The majority of the crashes you are experiencing are related to "userspace watchdog timeout: no successful checkins from com.apple.WindowServer" which seems to happen to a lot of people - also on genuine Apple machines - and is a PITA to try to fix. I had my share of those a few months ago but somewhere between the OS updates, a BIOS update, and changing my SMBIOS to MacPro7,1 those went away.

Well, I changed my SMBIOS to MacPro7,1 and thankfully, that seems to take care of the reboot/sleep issues!!
So thanks to all especially you, @JimmyS!

 

[patrickfogle]

Ok, maybe there is some OC tweak going on....

Well, I changed my SMBIOS to MacPro7,1 and thankfully, that seems to take care of the reboot/sleep issues!!
So thanks to all especially you, @JimmyS!

But, since I was tinkering around in the EFI anyway, I decided to go ahead and update to 10.15.6 and I am getting this boot hang - (Basically, the Verbose boot hangs on "** In Memory Panic Stackshot Succeeded ** Byte Traced 8000" then the next line says "Attempting system restart...MACH Reboot" and it hangs there). Below is a screenshot for more info (and a shot of the two screens before the final screen). Any help fixing this issue would be appreciated!

 

[patrickfogle]

But, since I was tinkering around in the EFI anyway, I decided to go ahead and update to 10.15.6 and I am getting this boot hang - (Basically, the Verbose boot hangs on "** In Memory Panic Stackshot Succeeded ** Byte Traced 8000" then the next line says "Attempting system restart...MACH Reboot" and it hangs there). Below is a screenshot for more info (and a shot of the two screens before the final screen). Any help fixing this issue would be appreciated!

OK, I am not sure why it worked, but I rebooted with the 10.15.5 Installer USB EFI that as an SMBIOS of iMacPro1,1 and it finished the install (when I ran the update I was under an SMBIOS of MacPro7,1). Now it boots and works fine.

 

[macmaxxx]

.

 

[joeblough]

Intel 219 uses intelmausi, it’s the original deluxe with the 211 that needs the additional smalltree kext.

The usb 3 on the type c ports may be disabled in your USB port mappings? Does it happen with usbinjectall?

I’ve heard using the jettison app may help with the eject issue.

i started digging into this a little bit... something is weird. when USBInjectAll was loaded, there were a bunch of ports that seemed to be behind an internal hub called ASM107x. devices plugged in there didn't show up in hackintool but they were recognized by the system. it is possible though that hackintool is just messed up - i never checked USB ports with IORegistryExplorer when i was mapping them out yesterday. it could be that even with USBInjectAll and overriding the port limit with the quirk, there are just too many for them all to function properly.

so just now i loaded IORegistryExplorer to see if all the sleep/wake currents were injected and i see that the ports on XHCI@14 just cut off at SS01, because that's the 15th port. however, in my USBPorts.kext from hackintool there are 22 ports called out! so i think this means i'm actually over the limit and i need to delete a bunch of ports.

so, i better go back thru this again using USBInjectAll and see what IORegistryExplorer reports when i'm plugging and unplugging USB devices. the hub thing probably faked me out. i think probably what i have to do is the whole "port window" thing with the boot-args to only expose 15 at a time and do the mapping in stages.

edit: well there really is a hub in there, i guess it is an asmedia ASM1074. for some reason it is only showing up as a USB2.0 port. i must have missed an SSDT (SSDT-X299-XHCI.aml) ? that seems to appear in the old threads but i don't know if it's only necessary for < 10.13 or clover or something like that.

 

[JimmyS]

Ok, maybe there is some OC tweak going on....

Well, I changed my SMBIOS to MacPro7,1 and thankfully, that seems to take care of the reboot/sleep issues!!
So thanks to all especially you, @JimmyS!

You're welcome, I'm happy to see you were able to solve the issues.

 

[joeblough]

edit: well there really is a hub in there, i guess it is an asmedia ASM1074. for some reason it is only showing up as a USB2.0 port. i must have missed an SSDT (SSDT-X299-XHCI.aml) ? that seems to appear in the old threads but i don't know if it's only necessary for < 10.13 or clover or something like that.

well - duh, there are just too many ports on XHCI and the SS port for the ASM1074 was past the 15 limit. i deleted one of the motherboard USB3.0 connectors, the USB2.0 motherboard connector and the internal RGB controller and now i'm under the 15 limit.

however now the SS ports on the 2 back panel USB-C connectors are not working. after one reboot with USBInjectAll, i noticed that there were 3 USB controllers showing in hackintool - the USB3.1 motherboard port, the 3.0 XHCI controller, and a titan ridge 3.1 controller, which also seems to be named XHCI. after several reboots i haven't been able to get it to reappear and i guess it might be related to the back panel USB-C SS trouble. is there a need to do a SSDT rename on one (or both) of these controllers? i think iMacPro1,1 calls out XHC1 so the default name of XHCI should be OK, and not sure how to rename just one if both are really called XHCI.