CMOS reset required to boot after wake from sleep, help!

[Pixfix]

Any ideas on why waking from sleep crashes my computer and requires a CMOS reset to boot?

I’ve had intermittent sleep issues ever since I first started in OC 6 era, where often waking from sleep would crash the computer (only in macOS) and require a CMOS reset to boot. This happened so often that I installed a switch to make shorting the pins easier. I eventually figured out that it wouldn’t crash as often if I woke the computer by pressing the power button on the case instead of tapping the keyboard/ mouse. I’ve tried to do USB mapping with several methods, but none seem to help (or I’ve not done it correctly).

After upgrading my monitor and using a different usb hub / switch, I suddenly stopped having these crashes for several months on OC 8.3. Thinking that the new hardware had solved the issue, I decided to upgrade my 3900x to a 5950x and get the machine ready for Ventura/ Sonoma. As soon as I updated my BIOS (and updated all OC-safe bios settings) the crashes came back. I decided to move forward with the CPU and OC update, thinking maybe the newer OC 9.3 would work better with the newer BIOS, but no luck.

I’ve lurked here enough to know that often a single setting can make all the difference (hopefully!). Though my best guess is that I need something more robust than the generic SSDT-EC-USBX-DESKTOP.aml, but I am really struggling to understand the SSDT process. I would be so grateful for any suggestions. Specs and EFI below. Thank you!

Hardware

• ASUS Pro WS X570-Ace
• Ryzen 5950x
• Powercolor Red Dragon 6800
• Corsair Vengeance LPX 3200 64GB (16x4)
• EVGA Supernova 850W Titanium
• WD SN850 1TB (Monterey)
• WD SN770 2TB (Windows 11)
• Intel 660P 2TB in Angelbird PCIE Adapter
• Micron 8TB sata ssd
• Laird BT85x USB Bluetooth module - Cypress CYW20704 A2 chipset (only use occasionally)
• FiiO K3 USB DAC
• Cable Matters USB 2.0 Switch hub (for sharing kb & mouse)

OC 9.3 Built manually from Dortania Guide (used OCAT to update from 8.3->9.3)
Bios version 2.20.1271
Monterey 12.6.8

BIOS Settings - Disabled

• Realtek LAN Controller
• USB power delivery in Soft Off state (S5)
• Serial Port
• Legacy USB Support
• Launch CSM
• Fast Boot
• Resize BAR Support
• Precision Boost Overdrive
• Core Performance Boost (enabled when needed)
• ErP Ready
• Power on by PCI-E
• Power on by RTC

Bios Settings - Enabled

• Intel LAN Controller
• Above 4G Decoding
• OS Type - Windows UEFI mode
• XHCI Hand-off

 

[leesurone]

I

Any ideas on why waking from sleep crashes my computer and requires a CMOS reset to boot?

I’ve had intermittent sleep issues ever since I first started in OC 6 era, where often waking from sleep would crash the computer (only in macOS) and require a CMOS reset to boot. This happened so often that I installed a switch to make shorting the pins easier. I eventually figured out that it wouldn’t crash as often if I woke the computer by pressing the power button on the case instead of tapping the keyboard/ mouse. I’ve tried to do USB mapping with several methods, but none seem to help (or I’ve not done it correctly).

After upgrading my monitor and using a different usb hub / switch, I suddenly stopped having these crashes for several months on OC 8.3. Thinking that the new hardware had solved the issue, I decided to upgrade my 3900x to a 5950x and get the machine ready for Ventura/ Sonoma. As soon as I updated my BIOS (and updated all OC-safe bios settings) the crashes came back. I decided to move forward with the CPU and OC update, thinking maybe the newer OC 9.3 would work better with the newer BIOS, but no luck.

I’ve lurked here enough to know that often a single setting can make all the difference (hopefully!). Though my best guess is that I need something more robust than the generic SSDT-EC-USBX-DESKTOP.aml, but I am really struggling to understand the SSDT process. I would be so grateful for any suggestions. Specs and EFI below. Thank you!

Hardware

• ASUS Pro WS X570-Ace
• Ryzen 5950x
• Powercolor Red Dragon 6800
• Corsair Vengeance LPX 3200 64GB (16x4)
• EVGA Supernova 850W Titanium
• WD SN850 1TB (Monterey)
• WD SN770 2TB (Windows 11)
• Intel 660P 2TB in Angelbird PCIE Adapter
• Micron 8TB sata ssd
• Laird BT85x USB Bluetooth module - Cypress CYW20704 A2 chipset (only use occasionally)
• FiiO K3 USB DAC
• Cable Matters USB 2.0 Switch hub (for sharing kb & mouse)

OC 9.3 Built manually from Dortania Guide (used OCAT to update from 8.3->9.3)
Bios version 2.20.1271
Monterey 12.6.8

BIOS Settings - Disabled

• Realtek LAN Controller
• USB power delivery in Soft Off state (S5)
• Serial Port
• Legacy USB Support
• Launch CSM
• Fast Boot
• Resize BAR Support
• Precision Boost Overdrive
• Core Performance Boost (enabled when needed)
• ErP Ready
• Power on by PCI-E
• Power on by RTC

Bios Settings - Enabled

• Intel LAN Controller
• Above 4G Decoding
• OS Type - Windows UEFI mode
• XHCI Hand-off

After looking at your UTBMap.kext I believe it's as simple an issue as a USB mapping problem. Improperly configured USB mapping can cause havoc similar to what you describe, also I've found different versions of the OS play better or worse with the same USB Map for whatever reason.

There are your USB Ports according to ASUS
3rd and 2nd Gen AMD Ryzen™/2nd and 1st Gen AMD Ryzen™ with Radeon™ Vega Graphics Processors :
3 x port(s) USB 3.2 Gen 1
AMD X570 chipset :
5 x port(s) USB 3.2 Gen 2 (4 Type-A, 1 x Type-C)
AMD X570 chipset :
4 x port(s) USB 2.0

Your Type C port is mapped wrong and you don't have any USB2 ports mapped at all (which if by design that's fine but your total port count is wonky if you aren't using them).
Have you ever tried the USBMap Tool from CorpNewt instead of USBToolBox? It's a little easier to understand how to use and its worked well for me on AMD based systems.

GitHub - corpnewt/USBMap: Python script for mapping USB ports in macOS and creating a custom injector kext.

Python script for mapping USB ports in macOS and creating a custom injector kext. - corpnewt/USBMap

github.com

 

[Pixfix]

Your Type C port is mapped wrong and you don't have any USB2 ports mapped at all (which if by design that's fine but your total port count is wonky if you aren't using them).
Have you ever tried the USBMap Tool from CorpNewt instead of USBToolBox? It's a little easier to understand how to use and its worked well for me on AMD based systems.

GitHub - corpnewt/USBMap: Python script for mapping USB ports in macOS and creating a custom injector kext.

Python script for mapping USB ports in macOS and creating a custom injector kext. - corpnewt/USBMap

github.com

Thank you for your taking a look! Its been a while, but I will certainly try CorpNewt's tool again.

I only use 3 or 4 USB ports, so I tried to disable a few hoping fewer active ports would give me less to worry about, but I likely did it wrong.

 

[leesurone]

Thank you for your taking a look! Its been a while, but I will certainly try CorpNewt's tool again.

I only use 3 or 4 USB ports, so I tried to disable a few hoping fewer active ports would give me less to worry about, but I likely did it wrong.

All good, just keep in mind the USBC port should be identified as ""9" if it shows up as either or both HS or S0. USB2 are "0". I don't think you have any internal ports to map unless you didn't list a wifi solution, which would usually need one.

 

[Pixfix]

Alright, I built a new USBmap, but still getting the same issue, a hard crash when waking up from sleep, need to reset CMOS to boot again. Can you take another look at my USB map?

I only have 8 physical USB ports (6 USB3 on rear I/O, 2 USB3 on front of case, and 1 USBC on rear I/O). No wifi/ bluetooth / internal usb needed. I don't have any physical USB2-only ports on my case or rear Mobo I/O, So that means everything should be a type 3 port except for the USBC which is type 9, right?
I've read that some machines need to reset and force rebuild the USB map at boot. Is that what the RHUB file does? Another issue I sometimes have is my machine restarts 2 or 3 times before a successful boot (sometimes it boots on first try).

Thanks again for all your help!

 

[Edhawk]

The 4 x USB2 ports are served by 2 x Internal motherboard Header connectors, 2 x USB2 ports per connector.

If you have any case front USB2 ports or other USB2 devices connected to these Internal headers you would need to set the ports with Internal (255) connector type. Not USB2 (0), as your motherboard doesn't have any physical USB2 ports. Just these 4 x Internal ports. Plus the companion/Virtual USB2 ports from the USB3 physical ports.

As none of the 3 x USB controllers in your system exceed the 15 port limit, you do not have to worry about disabling any of the ports. Your system will work with all the ports active, as long as they are set with the correct connector type/UsbConnector. Which isn't the case with your current UTBMap.kext.

XHC1 controller:

1. You are disabling HS05 and HS06, this is unnecessary as explained above.
2. Are these two of the Internal Header ports? If yes then they need to be set with UsbConnector - (255), not (3).

XHC1 USB Controller

XHC0 Controller:

1. HS03/SS03 and HS04/SS04 are all disabled, this is unnecessary as explained above. As these are paired ports, I think all of them are set correctly with UsbConnector (3).
2. HS05 and HS06 are also disabled, this is unnecessary as explained above.
3. Are these the other two Internal Header ports? If yes then they need to be set with UsbConnector - (255), not (3).

XHC0 USB Controller

XHC0-1 Controller:

1. The first thing to say about this is the name should be changed to XHC2 , to totally differentiate this controller from the XHC0 controller. You will need to add an ACPI rename patch to your config.plist to do this. Then you would need to change the name in the UTBMap.kext and in the SSDT-USB-Reset.aml table if you are using one.
2. Just HS02/SS02 ports being disabled in this section. Probably have the correct UsbConnector (3).

XHC0-1USB Controller.

Hope this helps.

 

[Pixfix]

Thank you Edhawk!

I think, maybe, hopefully... I've got it. No crashes after several sleep/wake attempts.

I turned all ports back on. Set XHC1 HS05/ HSO6 and XHC0 HS05/ HSO6 to 255. Renamed XHC0-1 to XHC2 with an old .aml file from a previous mapping adventure and created a RHUB.

Thank you both so much!

 

[Pixfix]

Nevermind, crashed again
Did I miss something in the USBmap? Only the 2 internal usb headers need to be set to 255, right?

 

[Edhawk]

Yes, only the 4 x ports from the 2 x USB 2 headers need to be set as 255.

 

[Pixfix]

Hmm, ok, well thanks for confirming.

Any other suggestions for trouble shooting my crashes?
Is there anything in my Config related to usb or power management?
Bios settings?
Could it be a hardware problem? (seems unlikely because windows never crashes like this)

 

[Pixfix]

Is there a crash log that would give me any clues to what is trigging my wake from sleep crash?
I installed the debug version of opencore, but that doesn't seem to help. And I can't find anything relevant when browsing logs in the console app, but I also don't really know what I'm looking for.

 

[Edhawk]

Some of the Bios options you have used are not in accordance with the Dortania guide for AMD Ryzen systems.

1. Legacy USB Support should be Enabled not disabled.
2. Resize Bar Support should be Enabled or Auto not disabled, when Above 4G Decoding is Enabled, along with change to config.plist Resize Bar entry.
3. OS Type should be 'Other' not Windows UEFI mode.

I would recommend you reset your BIOS to 'Default Settings', which can usually be found on the Exit tab in the bios menu.
Then apply the Bios options shown in the screenshot below, and only the options shown on the screenshot.

AMD Bios settings from Dortania guide.

 

[Pixfix]

Thanks for the suggestion, but still crashing
I loaded default bios settings, and only changed the settings called out in the guide.
I had settled on the combination of custom BIOS settings in my first post because they seemed to help minimize the crashes, but I hadn't considered the OS Type as the Dortania guide isn't clear on what motherboards should use the Other setting.
I'll try rebuilding my config again with OC 9.5 and the recommended bios settings. I'm not optimistic though, since I've done this many times since this problem first started years ago with OC6.4. I might try updating to Ventura too now that its recommended for the latest Photoshop.
If that doesn't work, sounds like my only options are to either try a different motherboard or buy a real Mac.

 

[Edhawk]

All motherboards should use the 'Other OS' setting.

Another BIOS setting that can be problematic is 'CSM' (Compatibility Support Module) which should be set as Disabled for 'most' UEFI motherboards. CSM offers backward compatibility by booting the machine as if you were running a legacy BIOS system, which your AMD X570 board shouldn't be using.

I would recommend you temporarily remove these three drives from your system, while you are troubleshooting your macOS installation.

• WD SN770 2TB (Windows 11)
• Intel 660P 2TB in Angelbird PCIE Adapter
• Micron 8TB sata ssd

Which Micron drive are you using? The Micron 2200S drive is known to cause boot issues with macOS.

The drives can be reconnected/reinstalled once you have macOS booting successfully. Just not all in one go, stage their reintroduction so any issues with any particular drive are noticed when it is reinstalled.

 

[Pixfix]

Hey thanks so much for your reply and sticking with me
I’ve kept “Launch CSM” disabled. And already tried removing the intel nvme and micron 5210 ION 2.5 SSD (sata not nvme), but crashes continue. The problem predates my WD SN770 nvme that currently has windows 11 on it, but I’ll try anything at this point haha.

I really wish there was some kind of log file that could give me a clue. Could it be gpu related? Ram? I’ve run memtest with no issues.
Is rx580 still the cheapest metal compatible card these days?

 

[Middleman]

@Pixfix,

Hi there. Looking at your issues I think I might know what is causing the crashing. Instead of using SSDT-EC-USBX-DESKTOP have you tried to use them separately as both SSDT-EC & SSDT-USBX? As I found that worked for me on my Gigabyte B550 Vision D.

 

[Pixfix]

@Pixfix,

Hi there. Looking at your issues I think I might know what is causing the crashing. Instead of using SSDT-EC-USBX-DESKTOP have you tried to use them separately as both SSDT-EC & SSDT-USBX? As I found that worked for me on my Gigabyte B550 Vision D.

I have not, please tell me more!
Did you go through the process of making custom SSDTs? Or were you able to use prebuilt files?

 

[Middleman]

I have not, please tell me more!
Did you go through the process of making custom SSDTs? Or were you able to use prebuilt files?

I just used the prebuilt ones, they worked fine.

 

[Middleman]

@Pixfix,

So I just had a look at your specs very quickly...and I think I might know what the other issue is. It looks like you don't have enough power to be honest.
You have a RX6800 plus a 5950X plus a load of SSDs (3x) and PCIe cards. That should all require at least a 1000W PSU and not a 850W PSU.
So I'd suggest to get it replaced and see how it fairs. In the meantime I'll check out the EFI and see what I find.

 

[Middleman]

@Pixfix,

Okay so I had a further look at your build and found a few things. Firstly is I see you have both AppleALC & AppleALCU enabled. You cannot enable both sadly because that will cause conflicts and crashing. I understand you also have a Fiio USB DAC attached - if you use just the AppleALC that will work along with your onboard Realtek card.
Aside from that your setup was missing a bunch of ACPI settings like PLUG & SBUS-MCHC. In my experience the B550/B570 boards all require these pretty much. So I have added those in for you. I've also enabled the XHC rename ACPIs that I found were needed on my B550. I've enabled it so that you can test and check if it needs improving.
Lastly I have enabled ReleaseUSBOwnership. I found that enabling that helped with some of my setups. I've updated the EFI, give it a try.