Windows 10 guest hangs intermittently at VirtualBox logo during UEFI boot; when it does boot, fails with 0xc0000001

[edufissure]

Host: Ubuntu 26.04 (Linux 6.x kernel), VirtualBox 7.2.12 r174389
Guest: Windows 10 (build 19041/2004), UEFI firmware, no Guest Additions issue known
Storage: VM files on native ext4 partition (also reproduced on NTFS via ntfs-3g — ruled out as cause)

Problem: Guest intermittently hangs indefinitely at the plain VirtualBox splash screen during UEFI boot — no crash, no reset, no further VBox.log output, just silence forever (left it 5+ minutes with no change). Roughly 1 in 2-3 boot attempts hang this way; the rest proceed normally to Windows.

VBox.log excerpt at the hang point (identical every time it hangs):

Code: Select all

00:00:01.317688 EFI: VBoxDbg> loadimage64 'bootmgfw.efi' 0x10000000 LB 0x1ca000
00:00:01.742217 xHCI: USB Suspended
00:00:01.755025 EFI: debug point DXE_AP
(x7 DXE_AP debug points)
00:00:01.755860 EFI: relocate module to 0xfffff80155306000 from 0xde8ea000
(x8 relocate module lines)
[nothing further — VM state remains RUNNING indefinitely]

VBoxManage showvminfo "Windows10" (relevant lines):

Code: Select all

Chipset:                     piix3
Paravirt. Provider:          Minimal
Effective Paravirt. Prov.:   Minimal
UEFI Secure Boot:            disabled
TPM Type:                    2.0

Troubleshooting already done:

• With Paravirt. Provider on default (Hyper-V), boot failed differently — a rapid ACPI reset loop, with

  Code: Select all

  IEM: wrmsr(0x40000118,0x0`00000001) -> #GP(0)

  logged just before each reset. Switching to KVM stopped the reset loop and got further (to the hang described above). Switching to Minimal did not remove the hang, but seems to make successful boots more frequent than KVM did.
• Disabling TPM entirely: hang occurs at the exact same log offset, so TPM emulation is not the cause.
• With UEFI Secure Boot enabled, the same indefinite silent hang occurred at this exact point on every attempt (100%, not intermittent). Disabling Secure Boot is what turned it into an intermittent issue rather than a permanent one.

When the VM does boot past this point, it now reaches the Windows recovery screen with:

Code: Select all

Your PC/Device needs to be repaired
The Boot Configuration Data file doesn't contain valid information for an operating system.
File: \BCD
Error code: 0xc0000098

which I've since fixed via

Code: Select all

bcdboot C:\Windows /l en-us /s S: /f UEFI

from WinRE. After that fix, successful boots progress to a different stop code:

Code: Select all

Error code: 0xc0000001

which I'm currently isolating via boot logging (

Code: Select all

bcdedit /set {default} bootlog yes

+ ntbtlog.txt inspected from the WinRE command prompt).

ntblog.txt

/home/eduard/Imágenes/Capturas de pantalla/Captura desde 2026-07-13 00-59-00.png

Question: Has anyone seen this specific intermittent silent hang — indefinite, no VBox.log activity, occurring right after

Code: Select all

bootmgfw.efi

loads and DXE modules relocate — on 7.2.12 with UEFI enabled? Is this a known timing/race issue between the EFI firmware and Paravirt Provider = KVM/Minimal, and is there a fix beyond "keep retrying the boot until it succeeds"?

Separately, I'm trying to isolate the driver responsible for the 0xc0000001 stop code by enabling boot logging (

Code: Select all

bcdedit /set {default} bootlog yes

) and reading ntbtlog.txt directly from the WinRE command prompt right after a failed normal boot (not from within a Safe Mode boot, since that overwrites the log with its own successful boot entries) — happy to update this thread once I've identified the specific driver.

[edufissure]

nobody ??