Hi everyone,
I'm moving a working Hackintosh (Ryzen 9 5900X + ASUS TUF X570) to new AM5 hardware. Now the boot always freezes at the same point and I can't get past it. I've tested a lot of things (list below) and I'd appreciate any ideas, especially from anyone running Zen 5 on AM5.
HARDWARE
THE PROBLEM
1. OpenCore and boot.efi work normally.
2. The kernel starts (ACM, AppleKeyStore, AppleSSE), then freezes right after:
3. There's no panic (even with debug=0x100), no reboot, and "PCI configuration begin" never appears.
4. At the exact moment of the freeze, USB stops working. The keyboard (rear port, wired) disconnects. If I replug it, it isn't detected. The port still has power, because a phone charges from it.
5. A short press of the power button turns the PC off instantly, as if macOS never took control of ACPI.
6. macOS Recovery freezes at exactly the same point, so it isn't the installed system.
MY CURRENT CONFIG
- Devirtualised: only 0x1080000000
- Also: DisableVariableWrite = true, SetupVirtualMap / RebuildAppleMemoryMap / SyncRuntimePermissions = true
- ACPI:
- SSDT-EC-USBX-DESKTOP. The LPC device is SBRG.
- Custom SSDT-CPUR with 32 Processor() objects. This board defines the CPUs as \_SB.PLTF.C0xx (ACPI0007).
- etorix AM5 patches #1 and #2, with a mask so they work on both BIOS versions. Both apply ("replaced 1 of 1").
- Result: ACPI now loads clean ("15 ACPI AML tables successfully acquired and loaded"). Before, 4 tables failed.
- UEFI: ReleaseUsbOwnership = true.
SOMETHING I FOUND IN OPENCORE (ZEN 5)
In OcCpuLib.c (1.0.7 and 1.0.8), the Family 1Ah code does this:
On the 9950X this gives "FID 204, MaxBR 204", which means FSBFrequency = 21 MHz. That value goes to the kernel (busFreq) through ProvideCurrentCpuInfo.
I binary-patched OpenCore to use MaxBusRatio 43, which gives FSB 99.8 MHz (like Zen 3/4). The log confirms the new value, but the freeze is exactly the same.
I'm mentioning it in case it's useful for someone else, or as a possible bug report.
WHAT I'VE ALREADY TRIED (NONE OF IT CHANGES WHERE IT FREEZES)
- IOMMU off
- Above 4G on, ReBAR off, CSM off
- iGPU off
- Onboard Wi-Fi/BT/LAN off
QUESTIONS
1. Has anyone booted a Ryzen 9000 (Zen 5) on an ASUS B650 or X670 with a recent BIOS (AGESA 1.2.x)? Which BIOS, ACPI patches and OpenCore did you use?
2. USB dies at the same moment as the freeze. Does that point to the chipset (GPP7 / Promontory) being reset during PCI configuration, or to the firmware hanging on the ACPI-enable SMI (SMI_CMD 0xB2, value 0xA0)?
3. Is there an OpenCore build or patch set known to work on Family 1Ah?
ATTACHMENTS
- forum-pack-v27.zip:
- config.plist (serial numbers removed)
- my SSDTs
- full OpenCore DEBUG log
- original ACPI dump from BIOS 3602 (DSDT, SSDTs, FACP, MCFG, APIC)
- Photos of the frozen screen
Thanks in advance for any help.
I'm moving a working Hackintosh (Ryzen 9 5900X + ASUS TUF X570) to new AM5 hardware. Now the boot always freezes at the same point and I can't get past it. I've tested a lot of things (list below) and I'd appreciate any ideas, especially from anyone running Zen 5 on AM5.
HARDWARE
- CPU: AMD Ryzen 9 9950X (Zen 5, Family 1Ah, 16C/32T)
- Motherboard: ASUS ROG STRIX B650-A GAMING WIFI. BIOS 3602 (AGESA 1.2.7.0), and also tested on 3279 (AGESA 1.2.0.3f).
- GPU: ASUS RX 480 8GB (the only GPU; iGPU disabled)
- Disk: Samsung 980 PRO NVMe, on the CPU lanes
- macOS: Sequoia 15.3, already installed (it worked on the 5900X)
- OpenCore: 1.0.7 and 1.0.8 (DEBUG)
THE PROBLEM
1. OpenCore and boot.efi work normally.
2. The kernel starts (ACM, AppleKeyStore, AppleSSE), then freezes right after:
Registered CoreAnalyticsHub functions with xnu ... CoreAnalyticsHub start completed3. There's no panic (even with debug=0x100), no reboot, and "PCI configuration begin" never appears.
4. At the exact moment of the freeze, USB stops working. The keyboard (rear port, wired) disconnects. If I replug it, it isn't detected. The port still has power, because a phone charges from it.
5. A short press of the power button turns the PC off instantly, as if macOS never took control of ACPI.
6. macOS Recovery freezes at exactly the same point, so it isn't the installed system.
MY CURRENT CONFIG
- SMBIOS iMac20,2. AMD_Vanilla patches (core count = 0x10), ProvideCurrentCpuInfo = true.
- Booter: DevirtualiseMmio = true. MmioWhitelist is built from this board's DEBUG log:
- Devirtualised: only 0x1080000000
- Also: DisableVariableWrite = true, SetupVirtualMap / RebuildAppleMemoryMap / SyncRuntimePermissions = true
- ACPI:
- SSDT-EC-USBX-DESKTOP. The LPC device is SBRG.
- Custom SSDT-CPUR with 32 Processor() objects. This board defines the CPUs as \_SB.PLTF.C0xx (ACPI0007).
- etorix AM5 patches #1 and #2, with a mask so they work on both BIOS versions. Both apply ("replaced 1 of 1").
- Result: ACPI now loads clean ("15 ACPI AML tables successfully acquired and loaded"). Before, 4 tables failed.
- UEFI: ReleaseUsbOwnership = true.
SOMETHING I FOUND IN OPENCORE (ZEN 5)
In OcCpuLib.c (1.0.7 and 1.0.8), the Family 1Ah code does this:
Code:
UINT8 CoreFrequencyID = (UINT8)BitFieldRead64 (CofVid, 0, 11); // 12-bit FID stored in 8 bits
if (CoreFrequencyID > 0x0f) CoreFrequencyID *= 5;I binary-patched OpenCore to use MaxBusRatio 43, which gives FSB 99.8 MHz (like Zen 3/4). The log confirms the new value, but the freeze is exactly the same.
I'm mentioning it in case it's useful for someone else, or as a possible bug report.
WHAT I'VE ALREADY TRIED (NONE OF IT CHANGES WHERE IT FREEZES)
- Kexts: -x (safe mode), and only Lilu + VirtualSMC + AppleMCEReporterDisabler
- CPU: cpus=1, SMT off, SSDT-CPUR, FSB 21 MHz vs 99.8 MHz
- MMIO: several whitelists. DevirtualiseMmio = false makes it reboot before the kernel.
- npci=0x2000, npci=0x3000, and no npci
- DisableVariableWrite, ReleaseUsbOwnership, Legacy USB on/off
- BIOS:
- IOMMU off
- Above 4G on, ReBAR off, CSM off
- iGPU off
- Onboard Wi-Fi/BT/LAN off
- BIOS 3279 to 3602. I can't go lower: ASUS blocks rollback below 3263.
- OpenCore 1.0.7 and 1.0.8, both stock and with the FSB patch
- USB hub and front-panel devices disconnected
QUESTIONS
1. Has anyone booted a Ryzen 9000 (Zen 5) on an ASUS B650 or X670 with a recent BIOS (AGESA 1.2.x)? Which BIOS, ACPI patches and OpenCore did you use?
2. USB dies at the same moment as the freeze. Does that point to the chipset (GPP7 / Promontory) being reset during PCI configuration, or to the firmware hanging on the ACPI-enable SMI (SMI_CMD 0xB2, value 0xA0)?
3. Is there an OpenCore build or patch set known to work on Family 1Ah?
ATTACHMENTS
- forum-pack-v27.zip:
- config.plist (serial numbers removed)
- my SSDTs
- full OpenCore DEBUG log
- original ACPI dump from BIOS 3602 (DSDT, SSDTs, FACP, MCFG, APIC)
- Photos of the frozen screen
Thanks in advance for any help.
