Repository navigation
Enable native Linux editor input and loopback MCP - #6026
Conversation
|
Developer Guide build artifacts are available for download from this workflow run:
Developer Guide quality checks: |
Cloudflare Preview
|
|
Compared 170 screenshots: 170 matched. Native Android coverage
✅ Native Android screenshot tests passed. Native Android coverage
Benchmark ResultsDetailed Performance Metrics
|
|
Compared 206 screenshots: 206 matched. |
✅ Continuous Quality ReportTest & Coverage
Static Analysis
Generated automatically by the PR CI workflow. |
|
Compared 185 screenshots: 185 matched. ParparVM vs HotSpot (JDK 25): Linux x64Runner CPU: INTEL(R) XEON(R) PLATINUM 8573C (baseline 1 performance regression:
Ratios are ParparVM / JDK 25: below 1.00x ParparVM is faster (time) or smaller (RAM). Median of 5 interleaved, paired rounds; every run's output was verified. A ratio more than 15% (time) / 15% (RAM) away from its baseline in
A change that moves performance ON PURPOSE records the new baseline in Result: performance regression |
|
Compared 185 screenshots: 185 matched. ParparVM vs HotSpot (JDK 25): Linux arm64Runner CPU: Neoverse-N2 (baseline Ratios are ParparVM / JDK 25: below 1.00x ParparVM is faster (time) or smaller (RAM). Median of 5 interleaved, paired rounds; every run's output was verified. A ratio more than 15% (time) / 15% (RAM) away from its baseline in
Result: no regression |
|
@codex review |
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 120d445c7d
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
|
@codex review |
|
The failed x64 job completed its native test suite; its final performance gate flagged an improvement in The passing mainline control 38848075973 used the same AMD EPYC 9V74 CPU model and the identical VM tree ( |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: a99c5afa68
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
|
@codex review |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 72c044c94d
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
|
@codex review |
|
Codex Review: Didn't find any major issues. Breezy! Reviewed commit: ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
If Codex has suggestions, it will comment; otherwise it will react with 👍. Codex can also answer questions or update the PR. Try commenting "@codex address that feedback". |
|
Compared 185 screenshots: 185 matched. Benchmark ResultsDetailed Performance Metrics
ParparVM vs HotSpot (JDK 25): Windows x64Runner CPU: AMD64 Family 25 Model 1 Stepping 1, AuthenticAMD (baseline Ratios are ParparVM / JDK 25: below 1.00x ParparVM is faster (time) or smaller (RAM). Median of 5 interleaved, paired rounds; every run's output was verified. A ratio more than 15% (time) / 15% (RAM) away from its baseline in
Result: no regression |
|
Compared 185 screenshots: 185 matched. Benchmark ResultsDetailed Performance Metrics
ParparVM vs HotSpot (JDK 25): Windows arm64Runner CPU: ARMv8 (64-bit) Family 8 Model D49 Revision 0, MICROSOFT CORPORATION (baseline Ratios are ParparVM / JDK 25: below 1.00x ParparVM is faster (time) or smaller (RAM). Median of 5 interleaved, paired rounds; every run's output was verified. A ratio more than 15% (time) / 15% (RAM) away from its baseline in
Result: no regression |
|
Compared 179 screenshots: 179 matched. Benchmark Results
Detailed Performance Metrics
ParparVM vs HotSpot (JDK 25): macOS arm64Runner CPU: Apple M1 (Virtual) (baseline Ratios are ParparVM / JDK 25: below 1.00x ParparVM is faster (time) or smaller (RAM). Median of 5 interleaved, paired rounds; every run's output was verified. A ratio more than 15% (time) / 15% (RAM) away from its baseline in
Result: no regression |
1 participant
Native Linux editor clients could not use
TextInputClient, and native review applications could not host the loopback listener used by built-in MCP. Add the GTK input-method bridge and loopback socket transport so the same editor and debugging APIs work in a ParparVM Linux application.The GTK bridge routes composition, Unicode commits, editing shortcuts and surrounding-text requests to the EDT, guarded by session and event sequence. It binds the owning window, supports initially focused fields before the dialog is mapped, and resumes after temporary native window focus loss without ending the Java input session. Sensitive fields withhold surrounding text and disable spelling/completion requests while preserving their normal input purpose; only password fields request password purpose. Explicit capitalization constraints map to the corresponding GTK hints, with autoCapitalize used as a fallback. Secondary windows now let pointer events through empty native-peer overlay space, matching the main window; previously clicking a different dialog field left typing in the first field. GTK clipboard/native-edit text is decoded as UTF-8 instead of widening UTF-8 bytes. The canvas key handler activates registered GTK menu accelerators before consuming ordinary keys, so shortcuts such as Ctrl-B reach their menu command while native peer widgets retain normal dispatch.
Sockets bind only IPv4 loopback. Retained token handles keep close/read/write races and stale handles from touching freed peers or reused descriptors. Closing wakes blocked I/O, reads retry EINTR, writes suppress SIGPIPE, and stream closure releases its peer. The developer guide documents the input contract and Linux MCP access.
Validation:
Native checks run locally only. Xvfb does not certify window-manager controls, physical input, native TextArea overlay workflows or every IBus/Fcitx engine. Application validation uses merged baseline
008f9eceplus this isolated Linux-port patch; the module tests use current-master baseb5ed7e4e. The review application is a RelWithDebInfo build, not a performance baseline. Outbound connection timeouts retain their pre-existing behavior; the port-based listener lifecycle still shares the core API's stop-before-bind limitation.Review follow-up: local GTK input and socket suites pass with ASan/UBSan. The previous hosted CI failure was a standalone ParparVM hello-memory ratio narrowly crossing the lower improvement threshold; identical VM sources and a passing same-CPU mainline control support measurement variance. Details and retained ratios are in the CI diagnosis comment. No performance baseline or tolerance was changed.