[HELP] Ryzen 7 7445HS (Phoenix 2, Zen 4 + Zen 4c hybrid): kernel hangs right after EXITBS:START, no verbose output (14.6 / 15.4 / 15.6)

[Hot Zot]

Hi everyone,

I'm trying to boot macOS on an HP Victus 15 with a Ryzen 7 7445HS. On this chip, OpenCore does everything correctly: all AMD Vanilla patches and kexts report Success. The kernel then freezes immediately after EXITBS:START, before it prints a single verbose line, and no panic is logged. I've worked through Dortania's troubleshooting, tried three macOS versions, and disabled IOMMU in firmware. My suspicion now is the Phoenix 2 hybrid layout (2× Zen 4 + 4× Zen 4c). I'm posting the data in case it helps anyone working on the patches, and any pointers are very welcome.

Hardware​

Laptop	HP Victus 15-fb3xxx (board 8DD0), AMI BIOS F.15 (2026-04-27)
CPU	AMD Ryzen 7 7445HS, 6C/12T, Phoenix 2: 2× Zen 4 + 4× Zen 4c
CPUID	Family 25 (19h), Model 124 (7Ch), Stepping 0, ProcessorId 178BFBFF00A70FC0
iGPU	Radeon 740M 1002:1901. It drives the internal panel (muxless), there is no macOS driver, and disable-gpu is set.
dGPU	RTX 3050 6GB Laptop 10DE:25AC, no driver loaded at this stage
RAM / SSD	16 GB / Micron NVMe 512 GB
Wi-Fi	MediaTek MT7922, disabled via SSDT

Software​

• OpenCore 1.0.8. The EFI was generated by OpCore Simplify. The tests below use the official 1.0.8 DEBUG build.
• Recovery images: macOS 15.6 (24.6.0), 15.4 (24.4.0) and 14.6 (23.6.0). All three hang at the same point.
• SMBIOS: MacBookPro16,1
• AMD Vanilla patches, core count set to 06. Every entry reports Success, including Bypass GenuineIntel, cpuid_set_cpufamily, _i386_init remove rdmsr, Fix PAT 15.0+ and the Zen 4 IOPCIFamily probeBusGated patch.
• ProvideCurrentCpuInfo = true. The log shows Patching MSR 35h to 0006000C and TscFreq = 3194007132Hz.
• DummyPowerManagement = true
• boot-args: -v debug=0x100 keepsyms=1 -vi2c-force-polling amfi=0x80 ...
• Security: SIP 0x0A43, SecureBootModel Disabled

Symptom​

• The last log line is AAPL: #[EB|LOG:EXITBS:START].
• The screen keeps the boot.efi text. The kernel never clears it and never prints anything.
• A short press of the power button powers the laptop off instantly. My read is that the kernel never takes over ACPI. No cursor or GUI appears, even after waiting 5+ minutes.
• The next boot logs OC: Panic log does not exist.
• The DEBUG log shows MAT support is 1. The GOP framebuffer is at 0x7E10000000 (1920x1080), so Above 4G is active.

What I tried (same result every time)​

• SetupVirtualMap true → false
• DevirtualiseMmio true (MMIO devirt was OK), then reverted
• ForceExitBootServices true
• MAT combo: RebuildAppleMemoryMap + SyncRuntimePermissions true, EnableWriteUnprotector false
• BIOS: Secure Boot off, Virtualization (SVM) off
• IOMMU disabled in firmware with Smokeless UMAF (AMD CBS). I confirmed the IVRS table is gone (Failed to drop ACPI IVRS - Not Found), and Windows no longer reports DMA protection.
• Dropped IVRS via ACPI → Delete before that
• Removed a Kernel → Block entry for IONDRVSupport that the generator had added
• macOS 15.6, 15.4 and 14.6 recovery
• Checked the MADT: plain xAPIC entries only (12 enabled, 4 disabled with APIC ID 0), no x2APIC

Why I suspect the Zen 4c (hybrid) cores​

• Same kernel, different CPU. Another user's Ryzen 7 7435H laptop (Zen 3+, family 19h model 44h) also hung at EXITBS with the 26.x installer, but the 15.6 kernel ran and printed verbose output there. With the exact same 15.6 image (24.6.0), my model 7Ch hangs before the first verbose line.
• Full Zen 4 Phoenix chips get further. Threads on this forum show an 8945HS (Minisforum UM890 Pro) reaching Recovery and a 7840HS (Redmibook Pro 15) getting as far as USB init. Both have 8× full Zen 4 cores.
• I found no Phoenix 2 / Zen 4c success at all. I couldn't find a single report of a 7445HS, 7545U, 7540U, 8540U or 8440U getting past EXITBS.

Questions​

1. Has anyone booted macOS on a Phoenix 2 / Hawk Point 2 chip?
2. Is there a known problem with mixed Zen 4 / Zen 4c cores in the AMD kernel patches or in XNU's early CPU init, for example CPUID topology, cache or extended-feature leaves differing between core types?
3. Is there a way to get any output from the kernel this early on a laptop with no serial port?
4. Are there experimental patches or settings worth testing? I'm happy to test anything and post logs.

Attached (logs.zip): OpenCore DEBUG logs for 15.6 (IOMMU off), 15.4, 14.6, and 14.6 with IOMMU off, plus my config.plist with serials removed.

Thanks a lot!