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

Display sleep doesn't blank the monitor: the last frame stays on screen, frozen (NVRMFB setPowerState / kConnectionPower are no-ops)

Open
  • Is something wrong?
#90 3 comments 0 reactions 0 assignees View on GitHub

Maintainers usually reply within 1 day

Nobody has claimed this yet.

Assessment

Difficulty
4/5
Estimated time
3-5 days
Newbie friendliness
48/100
Issue type
Bug
Clarity
Mostly clear
Activity status
Active
Tech stack
macos
Domain
backend, desktop

Research direction

Read kexts/NVRMFB/fb/nvrm-fb.cpp and compare its setPowerState and setAttributeForConnection paths with kexts/NVRM/fb/nvrm-fb.cpp. Trace how these paths reach NVKMS, then test display sleep and wake on a DisplayPort monitor. Done means the monitor enters standby or goes dark on power state 0 and restores on power state 1.

Written by the indexing model from the issue text.

Description

When macOS puts the display to sleep (System Settings → Lock Screen → "Turn display off when inactive"), the monitor does not go dark or into standby. It keeps showing the last frame, frozen: the menu-bar clock stops at the time the display "slept". Moving the mouse brings it back normally.

Setup

  • RTX 5080 (10de:2c02), Ryzen 9 9900X, macOS 26.7.1 (25G241), one 5K display (Kuycon G27P, 5120×2880) on DP, OpenCore 1.0.9
  • NullMoth 1.4.0 kexts (NVRMFB 0935a3a0…) with the rebuilt kexts/26/NVAccel; official 1.4.0 NVMTLDriver
  • pmset: displaysleep 1 for the test, sleep 0

What happens (sampled every 30 s with ioreg and pmset -g assertions; log from log show)

  • HID idle time counts up normally and nothing resets it (it reached 431 s in an earlier run), so no input source is keeping the display awake.
  • At 1 min idle, the whole chain runs at the same millisecond:
    • powerd: Display state: Off
    • WindowServer (CoreDisplay): displayWillSleepWithImages
    • kernel (NVRMFB): NVRM-fb: setPowerState 0
  • IODisplayWrangler CurrentPowerState then goes 4 → 2 → 1 → 0.
  • The monitor stays lit and keeps showing the last frame. The clock in the menu bar stays at 20:11 for 8 minutes.
  • On mouse movement: powerd: Display state: On, NVRM-fb: setPowerState 1, and the screen updates again. Wake works.
  • The screen saver never starts either. With the default settings the display goes to sleep first, and after that nothing is drawn.

Cause, from the source (kexts/NVRMFB/fb/nvrm-fb.cpp at v1.9.0)

IOReturn NVRMNVDAFramebuffer::setPowerState(unsigned long ordinal, IOService *device)
{
    FBLOG("setPowerState %lu", (unsigned long)ordinal);
    if (ordinal) handleEvent(kIOFBNotifyDidPowerOn);
    else         handleEvent(kIOFBNotifyWillPowerOff);
    return kIOPMAckImplied;
}

and in setAttributeForConnection:

if (attribute == kConnectionPower) { FBLOG("setAFC '%s' POWER %lu -> success", _s, (unsigned long)value); return kIOReturnSuccess; }

Neither path turns the head or the DP link off (no DPMS / head disable through NVKMS), so scanout continues from the last framebuffer contents while WindowServer has stopped drawing. The same code is in kexts/NVRM/fb/nvrm-fb.cpp.

Expected: on power state 0, blank the output or put the link into DPMS off (so the monitor goes to standby), and restore it on power state 1.

Workaround for now: set "Turn display off when inactive" to Never; use the monitor's power button.

I can test a build.

中文版（Chinese version, same content）

macOS 讓螢幕休眠時（系統設定 → 鎖定畫面 →「閒置時關閉顯示器」），螢幕不會變黑也不會進入待機，而是一直顯示最後一格畫面、凍住不動：選單列的時鐘停在螢幕「休眠」那一刻。動一下滑鼠就恢復正常。

環境

  • RTX 5080（10de:2c02）、Ryzen 9 9900X、macOS 26.7.1（25G241）、一台 5K 螢幕（Kuycon G27P，5120×2880）接 DP、OpenCore 1.0.9
  • NullMoth 1.4.0 kext（NVRMFB 0935a3a0…），kexts/26/NVAccel 重編；NVMTLDriver 為官方 1.4.0
  • pmset：測試時 displaysleep 1、sleep 0

現象（用 ioreg 和 pmset -g assertions 每 30 秒取樣；記錄來自 log show）

  • HID 閒置時間正常累加，沒有被任何東西歸零（之前一次累積到 431 秒），所以不是有輸入裝置讓螢幕保持喚醒。
  • 閒置 1 分鐘時，整條流程在同一毫秒發生：
    • powerd：Display state: Off
    • WindowServer（CoreDisplay）：displayWillSleepWithImages
    • kernel (NVRMFB)：NVRM-fb: setPowerState 0
  • 之後 IODisplayWrangler 的 CurrentPowerState 從 4 → 2 → 1 → 0。
  • 螢幕仍然亮著，持續顯示最後一格畫面。 選單列時鐘停在 20:11 達 8 分鐘。
  • 動滑鼠時：powerd：Display state: On、NVRM-fb: setPowerState 1，畫面恢復更新。喚醒正常。
  • 螢幕保護程式也從未啟動。預設設定下螢幕會先休眠，之後就不再繪製任何東西。

原因（原始碼）（kexts/NVRMFB/fb/nvrm-fb.cpp，v1.9.0）：setPowerState 只送出 kIOFBNotifyWillPowerOff／kIOFBNotifyDidPowerOn 通知；setAttributeForConnection 收到 kConnectionPower 直接回成功。兩條路都沒有關閉 head 或 DP 連線（沒有經 NVKMS 做 DPMS／停用 head），所以 WindowServer 停止繪製後，掃描輸出仍繼續送出 framebuffer 裡最後的內容。kexts/NVRM/fb/nvrm-fb.cpp 裡也是同樣的寫法。

預期：電源狀態 0 時讓輸出變黑，或讓連線進入 DPMS off（螢幕進待機）；電源狀態 1 時恢復。

暫時做法：「閒置時關閉顯示器」設為「永不」，用螢幕本身的電源鍵關螢幕。

有測試版我可以幫忙測。

Dominant language
Rust
Stars
2.5k
Forks
233
No merged PRs in 30d

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.