You've hit 0XC026232D. The full name is ERROR_GRAPHICS_MONITORDESCRIPTOR_ALREADY_IN_SET. Translation: Windows asked the GPU to add a monitor descriptor (usually an EDID block or a CEA-861 extension) to a descriptor set, and the driver said "that one's already in there, stop it."
In my experience this pops up most often when you're swapping monitors on the same DisplayPort or HDMI port without fully power-cycling, or when a driver update tries to re-register an EDID that's still cached from the previous session. It also shows up on docking stations that hand off a monitor to a different GPU mid-session. Real-world trigger: user unplugs a Dell U2720Q, plugs in an identical model on the same DP port, and the app that manages display profiles throws 0XC026232D.
Work top to bottom. Stop as soon as it's gone.
Step 1: The 30-second fix — power-cycle the display chain
Nine times out of ten this is a cached state problem. The GPU still thinks the old descriptor is live.
- Shut down Windows fully. Not sleep, not hibernate — shut down.
- Unplug the monitor's power cable from the wall. Not just the video cable — the AC cord.
- Unplug the video cable (DP/HDMI/USB-C) from both ends.
- Wait 30 seconds. Count it.
- Plug the monitor back into AC first. Give it 5 seconds to boot its internal scaler.
- Reconnect video. Then power on the PC.
The monitor has its own firmware that caches EDID handshakes. Killing its AC power forces a fresh EDID read on the next link training. If the error's still there, the cache is deeper in Windows.
Step 2: The 5-minute fix — purge the monitor cache in Device Manager
Windows keeps a per-user and per-system cache of monitor descriptors. When that cache gets out of sync with what's physically connected, you get 0XC026232D on every call that touches the descriptor set.
- Right-click Start → Device Manager.
- Expand Monitors. You'll likely see entries like Generic PnP Monitor or duplicates of your actual model.
- Right-click each one → Uninstall device. Check Attempt to remove the driver for this device if the box appears.
- Now expand Display adapters, right-click your GPU, and pick Uninstall device. If it offers a checkbox to delete driver software, leave it unchecked. We're clearing state, not nuking your driver.
- Reboot. Windows will re-enumerate the display chain on the next boot.
This works because Windows rebuilds the monitor descriptor set from scratch on first boot after the enumeration cache is cleared. Don't bother with "Scan for hardware changes" — it uses the same broken cache. Reboot is mandatory.
If you're on a laptop with a dock, unplug the dock before step 3 and leave it unplugged through the reboot. Docks are the number-one source of stale descriptor sets.
Step 3: The 15-minute fix — clear the EDID override and configuration keys
If Step 2 didn't do it, the descriptor set is being re-created from a registry key or an INF override. Time to clean it out manually. Back up first.
3a. Remove cached EDID overrides
Open an elevated Command Prompt and run:
reg export "HKLM\SYSTEM\CurrentControlSet\Enum\DISPLAY" %USERPROFILE%\Desktop\display_backup.regThat backs up the display enumeration tree. Now browse to:
HKLM\SYSTEM\CurrentControlSet\Enum\DISPLAYUnder this key you'll see subkeys like Default_Monitor, GenericPnPDevice, or vendor-specific ones matching your monitor. Each contains a child with an EDID binary value. If you have an override installed (common when someone tried to force 4K on a laptop panel or fix a color profile), it lives here.
Delete the specific monitor subkey that matches your problem display. Don't nuke the whole DISPLAY tree — you'll lose unrelated monitor entries.
3b. Clear the per-user display config
The user-side descriptor set cache lives here:
HKCU\Software\Microsoft\Windows\CurrentVersion\Display
HKCU\Software\Microsoft\Windows NT\CurrentVersion\DisplayExport both, then delete the Display subkeys under each. They'll rebuild on next logon. This is safe — these are caches, not source-of-truth.
3c. Kill the graphics driver's private descriptor cache
Nvidia and AMD both keep their own caches outside the registry. Close every app that touches displays (color calibration tools, DisplayFusion, vendor control panels). Then:
- Nvidia: delete
C:\ProgramData\NVIDIA Corporation\NV_Cacheand%LOCALAPPDATA%\NVIDIA\DXCache. - AMD: delete
%LOCALAPPDATA%\AMD\DxCacheand%LOCALAPPDATA%\AMD\DxcCache. - Intel: delete
%LOCALAPPDATA%\Intel\ShaderCache.
Reboot. If the descriptor set was being rebuilt from a shader cache collision (yes, it happens on DisplayPort MST setups), this kills it.
When none of that works
Two possibilities left.
One: the monitor's EDID is genuinely corrupt. Rare, but I've seen it on cheap KVM switches that mangle EDID passthrough. Test the monitor on a different machine, direct cable, no KVM, no dock. If the error follows the monitor, it's the monitor.
Two: a third-party app is calling SetMonitorDescriptor in a loop. Usually a display manager, a streaming tool with virtual monitors (Parsec, Duet, spacedesk), or a color calibration utility. Check Task Manager for anything doing display work, kill it, and retest. If the error stops, you've found your culprit.
Don't waste time on SFC or DISM. I've never once seen them fix 0XC026232D — it's not a system file integrity issue, it's a state synchronization issue between the GPU driver, Windows, and the physical monitor. SFC scans 20 minutes and reports clean every time.
Preventing it next time
If you swap monitors often, always power-cycle the monitor itself when changing cables. Don't hot-swap between two identical models on the same port — Windows can't tell them apart from the descriptor set's point of view. If you're on a dock, use a dock with proper DP alt-mode handling (CalDigit and Lenovo's USB-C docks are solid here; cheap ones aren't). And keep your GPU driver current — Nvidia's 55x.xx and AMD's Adrenalin 24.x branches both shipped fixes for descriptor set handling over the last year.