Hacktoberfest 2026: the issues maintainers tagged for October, open and beginner-friendly. Browse Hacktoberfest issues

RTX 3080 / macOS: stuck at Apple logo/loading screen with visible cursor after updates to 1.10, 1.12 and 1.13; recovery/disable attempt did not restore boot

Open
#116 0 comments 0 reactions 0 assignees View on GitHub

Maintainers usually reply within 2 days

Nobody has claimed this yet.

Assessment

Difficulty
5/5
Estimated time
Over a week
Newbie friendliness
20/100
Issue type
Bug
Clarity
Mostly clear
Activity status
Active
Tech stack
macos, rust
Domain
backend

Research direction

Start by reviewing the reported boot setup, update attempts, and recovery configuration; the issue names no source files or tests. The maintainer would need to establish a reproducible failure and determine whether the driver, recovery helper, or storage errors are involved. Done means identifying the cause and providing a supported recovery path or a clearly scoped fix.

Written by the indexing model from the issue text.

Description

Send logs report ID from the working Mac

Unavailable — the affected Mac is stuck at the Apple logo/loading screen with a visible cursor and cannot reach the login screen or desktop, so I cannot run 1401's “Send logs” function. Please advise how to collect logs from macOS Recovery.

Anything you had to do that 1401 did not do for you

Summary

My macOS installation could reach the desktop with an earlier NullMoth NVIDIA driver. After updating, I can no longer reliably reach the desktop.

I tried versions 1.10, 1.12 and 1.13, and encountered the same unresolved failure. These are the version numbers I recorded from my update attempts; I have not yet independently verified the exact manager and installed driver build for each attempt.

Last known working combination (from memory, not yet independently verified): 1401 Mac app 1.0.12 with NVIDIA driver 1.0.8. I believe this was the combination that allowed the same machine to reach the macOS desktop before updating. Please distinguish the app version from the driver version. Whether the macOS build remained unchanged across these attempts has not yet been confirmed.

The current visible symptom is a stall at the Apple logo/loading screen, with a visible mouse cursor. The cursor appears while the Apple boot/loading screen is still displayed, not after reaching the login screen or desktop. Earlier attempts appeared to stop around 30% of the boot progress bar. My earlier description of a “black screen with cursor” was imprecise; this should not be classified as a confirmed post-login or desktop black-screen failure.

Windows still boots. I would appreciate a supported recovery procedure and guidance on collecting diagnostics when the macOS desktop is inaccessible.

Hardware and boot setup

  • CPU: Intel Core i5-12400F (no integrated GPU)
  • GPU: NVIDIA GeForce RTX 3080 10 GB, PCI ID 10DE:2206
  • Motherboard/chipset: H610; exact board model not yet collected
  • RAM: 32 GB
  • macOS: exact version/build not yet collected from the affected installation
  • Boot: disk-based rEFInd -> OpenCore -> macOS (not direct Apple boot.efi)
  • OpenCore version: not yet confirmed
  • OpenCore EFI is on a kimtigo 256 GB SSD shared with the Windows boot files
  • macOS APFS installation is on a WD Green 1 TB SATA drive

Update and symptoms

  1. macOS previously reached the desktop with an older NullMoth driver. From memory, this was 1401 Mac app 1.0.12 / NVIDIA driver 1.0.8, but I have not independently verified that combination.
  2. I updated through the 1401 macOS app. At least one update reported installation success.
  3. After restarting, macOS failed to reach a usable desktop.
  4. I tried versions 1.10, 1.12 and 1.13, without resolving the problem.
  5. The current result is a stall at the Apple logo/loading screen with a visible cursor. Neither the login screen nor the desktop is reached.

I have not collected a complete version-by-version test matrix or logs for every update, so I cannot claim the failure is identical internally across all three releases.

Messages seen in verbose boot photos

Earlier verbose attempts repeatedly showed messages including:

AppleKeyStore: operation failed ...
unexpected session uid: -1 requested by: bluetoothd
unexpected session uid: -1 requested by: AirPlayXPCHelper

One attempt also showed SATA/AHCI errors, with fragments such as:

[AHCI][DISK] WatchdogTimeout
[AHCI][PRT] WaitSecsForNotBusy ... Port 6
AbortCommands ... deviceWithError
NormalCompletion ... Error ...
InitiateDeviceRecovery

These are fragments transcribed from photos, not a full text log. I do not know which physical device corresponds to AHCI Port 6. Disk/controller trouble has not been ruled out, and I do not assume the AppleKeyStore messages themselves are the root cause.

Recovery and disable attempts

OpenCore removal entry

The 1401: Remove NVIDIA driver entry is visible in OpenCore. I selected it twice, but did not observe a successful recovery to the desktop. I have not confirmed whether the tool displayed a confirmation prompt, whether it successfully wrote the removal request, or whether the macOS recovery service completed removal.

During a Windows-side inspection, the existing EFI/OC/Tools/NullMothSafe.efi was 25,088 bytes, with a file timestamp of 2026-10-08. I do not know whether this is the correct tool binary for the installed release.

AMFIPass test

AMFIPass was temporarily disabled as a diagnostic test. This did not resolve the failure, so it was re-enabled. This is not evidence that AMFIPass caused the issue.

Explicit driver-disable boot configuration

An assistant edited the disk OpenCore configuration from Windows, backed it up, and verified the written file. The test configuration used:

boot-args:
nvfb=1 nvaccel=1 nvfbheads=4 -nvkmsnosmooth amfi_get_out_of_my_way=0x1 amfi=0x80 -v -nvoff

Booter -> Quirks -> ResizeAppleGpuBars = 0
UEFI -> Quirks -> ResizeGpuBars = -1
Kernel -> Block -> com.apple.iokit.IONDRVSupport -> Enabled = false

AMFIPass remained enabled. The Apple NVRAM Delete list included both boot-args and nullmoth-remove; nullmoth-remove was not added under NVRAM Add.

An earlier disable configuration had been successfully written, but a subsequent Windows inspection found that it had changed back: -nvoff and -v were missing, BAR settings were back to -1/13, and the IONDRVSupport exclusion was enabled again. The source of that configuration change is unknown; it may have been a restore operation, but this has not been established.

The disable configuration was then reapplied and read back from the actual EFI file to verify it. However, I have not obtained the affected macOS session's actual kern.bootargs or loaded-kext list. Therefore, I cannot yet prove that the stalled Apple-logo boot used -nvoff, that the installed release honored it, or that the driver was successfully uninstalled.

No APFS driver files were manually deleted from Windows, and no complete old EFI was blindly restored.

Expected behavior

  • A supported update should allow macOS to reach a usable desktop, or provide actionable diagnostics if initialization fails.
  • The recovery/disable path should allow recovery using the RTX 3080's firmware/basic display output if supported, since this CPU has no iGPU.

Questions for the maintainer

  1. Is a stall at the Apple logo/loading screen with a visible cursor on an RTX 3080 a known problem in these versions? Could the AppleKeyStore/session-UID messages be secondary symptoms of graphics initialization or another startup failure? I have not reached the login screen or desktop, so a post-login failure is not established.
  2. What is the supported way to verify that -nvoff was honored when the desktop is inaccessible? Does it disable all relevant kernel components, and can installed user-space GPU components still affect login?
  3. What is the safest supported recovery/removal procedure from macOS Recovery, including targeting the installed system/data volume and handling its Auxiliary Kernel Collection? I do not want to run the normal uninstall script against the Recovery environment by mistake.
  4. Can the recovery helper restore OpenCore settings even when removal is incomplete, or otherwise explain the observed configuration reverting? Which logs would distinguish these cases?
  5. How should I verify/update the existing NullMothSafe.efi, and which recovery-service files should be present for the installed release?
  6. Which logs should I collect from Recovery? Please specify paths and safe read-only commands, including the macOS build, installed driver version, recovery log, and relevant WindowServer/kernel diagnostics.
  7. Do the AHCI timeout messages suggest that I should prioritize a storage investigation before further driver tests?

Attachments available

I can attach the verbose boot photos showing AppleKeyStore/session-UID messages and AHCI timeouts. I have not yet collected a full macOS crash report or /Library/NullMoth/recover.log.

Please let me know which additional information would be most useful. I can provide a redacted OpenCore configuration if needed, with SMBIOS serials, MLB, UUID, ROM and other identifying fields removed.

Dominant language
Rust
Stars
2.5k
Forks
233
Avg merge
1h 37m
Merged PRs (30d)
1

Getting set up

This project ships no dev container, Dockerfile or contributing guide, so setting up is up to you: start from its README, and see our first-contribution guide for the general steps.

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

More from nullmoth/nvidia-macos-driver

All issues in nullmoth/nvidia-macos-driver

Similar issues

More Rust issues

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.