starred/terminal
1
0
Fork 0
mirror of https://github.com/microsoft/terminal.git synced 2026-07-08 18:16:28 +00:00
Code Issues 19.0k Packages Projects Releases 20 Wiki Activity

[PR #18816] Fix a major stdin wakeup race condition #31613

New Issue
Open
opened 2026-01-31 09:48:24 +00:00 by claunia · 0 comments
claunia commented 2026-01-31 09:48:24 +00:00
Owner
Copy Link

Original Pull Request: https://github.com/microsoft/terminal/pull/18816

State: closed
Merged: Yes

The conhost v2 rewrite from a decade ago introduced a race condition:
Previously, we would acquire and hold the global console lock while servicing
a console API call. If the call cannot be completed a wait task is enqueued,
while the lock is held. The v2 rewrite then split the project up into a
"server" and "host" component (which remain to this day). The "host" would
hold the console lock, while the "server" was responsible for enqueueing wait
tasks outside of the console lock. Without any form of synchronization,
any operations on the waiter list would then of course introduce a race
condition. In conhost this primarily meant keyboard/mouse input, because that
runs on the separate Win32 window thread. For Windows Terminal it primarily
meant the VT input thread.

I do not know why this issue is so extremely noticeable specifically when we
respond to DSC CPR requests, but I'm also not surprised: I suspect that the
overall performance issues that conhost had for a long time, meant that most
things it did were slower than allocating the wait task.
Now that both conhost and Windows Terminal became orders of magnitudes faster
over the last few years, it probably just so happens that the DSC CPR request
takes almost exactly as many cycles to complete as allocating the wait task
does, hence perfectly reproducing the race condition.

There's also a slight chance that this is actually a regression from my ConPTY
rewrite #17510, but I fail to see what that would be. Regardless of that,
I'm 100% certain though, that this is a bug that has existed in v0.1.

Closes #18117
Closes #18800

Validation Steps Performed

  • See repro in #18800. In other words:
    • Continuously emit DSC CPR sequences
    • ...read the response from stdin
    • ...and print the response to stdout
    • Doesn't deadlock randomly anymore ✅
  • Feature & Unit tests ✅
**Original Pull Request:** https://github.com/microsoft/terminal/pull/18816 **State:** closed **Merged:** Yes --- The conhost v2 rewrite from a decade ago introduced a race condition: Previously, we would acquire and hold the global console lock while servicing a console API call. If the call cannot be completed a wait task is enqueued, while the lock is held. The v2 rewrite then split the project up into a "server" and "host" component (which remain to this day). The "host" would hold the console lock, while the "server" was responsible for enqueueing wait tasks _outside of the console lock_. Without any form of synchronization, any operations on the waiter list would then of course introduce a race condition. In conhost this primarily meant keyboard/mouse input, because that runs on the separate Win32 window thread. For Windows Terminal it primarily meant the VT input thread. I do not know why this issue is so extremely noticeable specifically when we respond to DSC CPR requests, but I'm also not surprised: I suspect that the overall performance issues that conhost had for a long time, meant that most things it did were slower than allocating the wait task. Now that both conhost and Windows Terminal became orders of magnitudes faster over the last few years, it probably just so happens that the DSC CPR request takes almost exactly as many cycles to complete as allocating the wait task does, hence perfectly reproducing the race condition. There's also a slight chance that this is actually a regression from my ConPTY rewrite #17510, but I fail to see what that would be. Regardless of that, I'm 100% certain though, that this is a bug that has existed in v0.1. Closes #18117 Closes #18800 ## Validation Steps Performed * See repro in #18800. In other words: * Continuously emit DSC CPR sequences * ...read the response from stdin * ...and print the response to stdout * Doesn't deadlock randomly anymore ✅ * Feature & Unit tests ✅
claunia added the pull-request label 2026-01-31 09:48:24 +00:00
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
⛺ Reserved
A11yCO
A11yMAS
A11ySev1
A11ySev2
A11ySev3
A11yTTValidated
A11yUsable
A11yVoiceAccess
A11yWCAG
Area-Accessibility
Area-AtlasEngine
Area-AzureShell
Area-Build
Area-Build
Area-Chat
Area-CmdPal
Area-CodeHealth
Area-Commandline
Area-CookedRead
Area-DefApp
Area-Extensibility
Area-Fonts
Area-GroupPolicy
Area-i18n
Area-Input
Area-Interaction
Area-Interop
Area-Localization
Area-Output
Area-Performance
Area-Portable
Area-Quality
Area-Remoting
Area-Rendering
Area-Schema
Area-Server
Area-Settings
Area-SettingsUI
Area-ShellExtension
Area-ShellExtension
Area-ShellExtension
Area-Suggestions
Area-Suggestions
Area-TerminalConnection
Area-TerminalControl
Area-Theming
Area-UserInterface
Area-VT
Area-Windowing
Area-WPFControl
AutoMerge
Blocking-Ingestion
Culprit-Centennial
Culprit-WinUI
Disability-All
Disability-Blind
Disability-LowVision
Disability-Mobility
External-Blocked-WinUI3
Fixed
Gathering-Data
good first issue
HCL-E+D
HCL-WindowsTerminal
Help Wanted
Impact-Compatibility
Impact-Compliance
Impact-Correctness
Impact-Visual
In-PR
InclusionBacklog
InclusionBacklog-Windows TerminalWin32
InclusionCommitted-202206
Issue-Bug
Issue-Docs
Issue-Feature
Issue-Feature
Issue-Question
Issue-Samples
Issue-Scenario
Issue-Task
Needs-Attention
Needs-Author-Feedback
Needs-Bisect
Needs-Discussion
Needs-Repro
Needs-Tag-Fix
Needs-Tag-Fix
Needs-Triage
No-Recent-Activity
Priority-0
Priority-1
Priority-2
Priority-3
Product-Cmd.exe
Product-Colortool
Product-Colortool
Product-Colortool
Product-Conhost
Product-Conpty
Product-Meta
Product-Powershell
Product-Terminal
Product-WSL
pull-request
Mirrored from GitHub Pull Request
 Resolution-Answered
Resolution-By-Design
Resolution-Duplicate
Resolution-External
Resolution-Fix-Available
Resolution-Fix-Committed
Resolution-No-Repro
Resolution-Won't-Fix
Severity-Blocking
Severity-Crash
Severity-DataLoss
spam
this-will-be-a-breaking-change
Tracking-External
WindowsTerminal_Win32
Work-Item
zAskModeBug
zInbox-Bug
No Label pull-request
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
claunia
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: starred/terminal#31613
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?