[Windows] Fix for Runtime error when closing external window with WPF Webview Control#34006
Conversation
There was a problem hiding this comment.
Pull request overview
This PR fixes a race condition in the WPF BlazorWebView that occurs when closing a window while messages are still being sent to the WebView2 control. The issue was introduced in PR #31777, which migrated from WebView2 to WebView2CompositionControl.
Changes:
- Implements a three-layer defense mechanism to prevent crashes during disposal
- Adds disposal flag checking in
SendMessageto exit early when disposing - Adds exception handling for COM object disposal errors
- Introduces
MarkAsDisposing()method to allow the owning control to signal disposal start
Reviewed changes
Copilot reviewed 2 out of 2 changed files in this pull request and generated 5 comments.
| File | Description |
|---|---|
src/BlazorWebView/src/Wpf/BlazorWebView.cs |
Calls MarkAsDisposing() in DisposeAsync before setting disposal flag |
src/BlazorWebView/src/SharedSource/WebView2WebViewManager.cs |
Adds disposal flag, modifies SendMessage with defensive checks and exception handling, adds MarkAsDisposing() method |
|
/azp run maui-pr-uitests , maui-pr-devicetests |
|
Azure Pipelines successfully started running 2 pipeline(s). |
kubaflo
left a comment
There was a problem hiding this comment.
Could you please try the AI's summary suggestions?
f333a6f to
df22ecc
Compare
|
🚀 Dogfood this PR with:
curl -fsSL https://raw.githubusercontent.com/dotnet/maui/main/eng/scripts/get-maui-pr.sh | bash -s -- 34006Or
iex "& { $(irm https://raw.githubusercontent.com/dotnet/maui/main/eng/scripts/get-maui-pr.ps1) } 34006" |
I’ve reviewed and updated the changes based on the latest findings. Case 1: Use a simpler volatile bool disposal flag Case 2: COMException handling
Case 3: WinForms parity |
|
/azp run maui-pr-uitests , maui-pr-devicetests |
|
Azure Pipelines successfully started running 2 pipeline(s). |
…otnet#34548) <!-- Please let the below note in for people that find this PR --> > [!NOTE] > Are you waiting for the changes in this PR to be merged? > It would be very helpful if you could [test the resulting artifacts](https://github.com/dotnet/maui/wiki/Testing-PR-Builds) from this PR and let us know in a comment if this change resolves your issue. Thank you! ## Description Adds a [gh-aw (GitHub Agentic Workflows)](https://github.github.com/gh-aw/introduction/overview/) workflow that automatically evaluates test quality on PRs using the `evaluate-pr-tests` skill. ### What it does When a PR adds or modifies test files, this workflow: 1. **Checks out the PR branch** (including fork PRs) in a pre-agent step 2. **Runs the `evaluate-pr-tests` skill** via Copilot CLI in a sandboxed container 3. **Posts the evaluation report** as a PR comment using gh-aw safe-outputs ### Triggers | Trigger | When | Fork PR support | |---------|------|-----------------| | `pull_request` | Automatic on test file changes (`src/**/tests/**`) | ❌ Blocked by `pre_activation` gate | | `workflow_dispatch` | Manual — enter PR number | ✅ Works for all PRs | | `issue_comment` (`/evaluate-tests`) | Comment on PR |⚠️ Same-repo only (see Known Limitations) | ### Security model | Layer | Implementation | |-------|---------------| | **gh-aw sandbox** | Agent runs in container with scrubbed credentials, network firewall | | **Safe outputs** | Max 1 PR comment per run, content-limited | | **Checkout without execution** | `steps:` checks out PR code but never executes workspace scripts | | **Base branch restoration** | `.github/skills/`, `.github/instructions/`, `.github/copilot-instructions.md` restored from base branch after checkout | | **Fork PR activation gate** | `pull_request` events blocked for forks via `head.repo.id == repository_id` | | **Pinned actions** | SHA-pinned `actions/checkout`, `actions/github-script`, etc. | | **Minimal permissions** | Each job declares only what it needs | | **Concurrency** | One evaluation per PR, cancels in-progress | | **Threat detection** | gh-aw built-in threat detection analyzes agent output | ### Files added/modified - `.github/workflows/copilot-evaluate-tests.md` — gh-aw workflow source - `.github/workflows/copilot-evaluate-tests.lock.yml` — Compiled workflow (auto-generated by `gh aw compile`) - `.github/skills/evaluate-pr-tests/scripts/Gather-TestContext.ps1` — Test context gathering script (binary-safe file download, path traversal protection) - `.github/instructions/gh-aw-workflows.instructions.md` — Copilot instructions for gh-aw development ### Known Limitations **Fork PR evaluation via `/evaluate-tests` comment is not supported in v1.** The gh-aw platform inserts a `checkout_pr_branch.cjs` step after all user steps, which may overwrite base-branch skill files restored for fork PRs. This is a known gh-aw platform limitation — user steps always run before platform-generated steps, with no way to insert steps after. **Workaround:** Use `workflow_dispatch` (Actions UI → "Run workflow" → enter PR number) to evaluate fork PRs. This trigger bypasses the platform checkout step entirely and works correctly. **Related upstream issues:** - [github/gh-aw#18481](github/gh-aw#18481) — "Using gh-aw in forks of repositories" - [github/gh-aw#18518](github/gh-aw#18518) — Fork detection and warning in `gh aw init` - [github/gh-aw#18520](github/gh-aw#18520) — Fork context hint in failure messages - [github/gh-aw#18521](github/gh-aw#18521) — Fork support documentation ### Fixes - Fixes dotnet#34602 --------- Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com> Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Jakub Florkowski <kubaflo123@gmail.com>
## Summary Enables the copilot-evaluate-tests gh-aw workflow to run on fork PRs by adding `forks: ["*"]` to the `pull_request` trigger and removing the fork guard from `Checkout-GhAwPr.ps1`. ## Changes 1. **copilot-evaluate-tests.md**: Added `forks: ["*"]` to opt out of gh-aw auto-injected fork activation guard. Scoped `Checkout-GhAwPr.ps1` step to `workflow_dispatch` only (redundant for other triggers since platform handles checkout). 2. **copilot-evaluate-tests.lock.yml**: Recompiled via `gh aw compile` — fork guard removed from activation `if:` conditions. 3. **Checkout-GhAwPr.ps1**: Removed the `isCrossRepository` fork guard. Updated header docs and restore comments to accurately describe behavior for all trigger×fork combinations (including corrected step ordering). 4. **gh-aw-workflows.instructions.md**: Updated all stale references to the removed fork guard. Documented `forks: ["*"]` opt-in, clarified residual risk model for fork PRs, and updated troubleshooting table. ## Security Model Fork PRs are safe because: - Agent runs in **sandboxed container** with all credentials scrubbed - Output limited to **1 comment** via `safe-outputs: add-comment: max: 1` - Agent **prompt comes from base branch** (`runtime-import`) — forks cannot alter instructions - Pre-flight check catches missing `SKILL.md` if fork isn't rebased on `main` - No workspace code is executed with `GITHUB_TOKEN` (checkout without execution) ## Testing - ✅ `workflow_dispatch` tested against fork PR dotnet#34621 - ✅ Lock.yml statically verified — fork guard removed from `if:` conditions - ⏳ `pull_request` trigger on fork PRs can only be verified post-merge (GitHub Actions reads lock.yml from default branch) --------- Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com> Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
🚦 Gate - Test Before and After Fix📊 Expand Full Gate —
|
…taType is compiled (dotnet#34717) ## Description Adds regression tests for dotnet#34713 verifying the XAML source generator correctly handles bindings with `Converter={StaticResource ...}` inside `x:DataType` scopes. Closes dotnet#34713 ## Investigation After thorough investigation of the source generator pipeline (`KnownMarkups.cs`, `CompiledBindingMarkup.cs`, `NodeSGExtensions.cs`): ### When converter IS in page resources (compile-time resolution ✅) `GetResourceNode()` walks the XAML tree, finds the converter resource, and `ProvideValueForStaticResourceExtension` returns the variable directly — **no runtime `ProvideValue` call**. The converter is referenced at compile time. ### When converter is NOT in page resources (runtime resolution ✅) `GetResourceNode()` returns null → falls through to `IsValueProvider` → generates `StaticResourceExtension.ProvideValue(serviceProvider)`. The `SimpleValueTargetProvider` provides the full parent chain, and `TryGetApplicationLevelResource` checks `Application.Current.Resources`. The binding IS still compiled into a `TypedBinding` — only the converter resolution is deferred. ### Verified on both `main` and `net11.0` All tests pass on both branches. ## Tests added | Test | What it verifies | |------|-----------------| | `SourceGenResolvesConverterAtCompileTime_ImplicitResources` | Converter in implicit `<Resources>` → compile-time resolution, no `ProvideValue` | | `SourceGenResolvesConverterAtCompileTime_ExplicitResourceDictionary` | Converter in explicit `<ResourceDictionary>` → compile-time resolution, no `ProvideValue` | | `SourceGenCompilesBindingWithConverterToTypedBinding` | Converter NOT in page resources → still compiled to `TypedBinding`, no raw `Binding` fallback | | `BindingWithConverterFromAppResourcesWorksCorrectly` × 3 | Runtime behavior correct for all inflators (Runtime, XamlC, SourceGen) | Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
<!-- Please let the below note in for people that find this PR --> > [!NOTE] > Are you waiting for the changes in this PR to be merged? > It would be very helpful if you could [test the resulting artifacts](https://github.com/dotnet/maui/wiki/Testing-PR-Builds) from this PR and let us know in a comment if this change resolves your issue. Thank you! ## Description Adds arcade inter-branch merge workflow and configuration to automate merging `net11.0` into the `release/11.0.1xx-preview3` branch. ### Files added | File | Purpose | |------|---------| | `github-merge-flow-release-11.jsonc` | Merge flow config — source `net11.0`, target `release/11.0.1xx-preview3` | | `.github/workflows/merge-net11-to-release.yml` | GitHub Actions workflow — triggers on push to net11.0, daily cron, manual dispatch | ### How it works Uses the shared [dotnet/arcade inter-branch merge infrastructure](https://github.com/dotnet/arcade/blob/main/.github/workflows/inter-branch-merge-base.yml): - **Event-driven**: triggers on push to `net11.0`, with daily cron safety net - **ResetToTargetPaths**: auto-resets `global.json`, `NuGet.config`, `eng/Version.Details.xml`, `eng/Versions.props`, `eng/common/*` to target branch versions - **QuietComments**: reduces GitHub notification noise - Skips PRs when only Maestro bot commits exist ### Incrementing for future releases When cutting a new release (e.g., preview4), update: 1. `github-merge-flow-release-11.jsonc` → change `MergeToBranch` value 2. `.github/workflows/merge-net11-to-release.yml` → update workflow `name` field Follows the same pattern as `merge-main-to-net11.yml` / `github-merge-flow-net11.jsonc`. Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com> Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
<!-- !!!!!!! MAIN IS THE ONLY ACTIVE BRANCH. MAKE SURE THIS PR IS TARGETING MAIN. !!!!!!! --> <!-- Enter description of the fix in this section --> ### Issues Fixed <!-- Please make sure that there is a bug logged for the issue being fixed. The bug should describe the problem and how to reproduce it. --> Fixes dotnet#33355 ### Description of Change This report has the goal to provide a detailed progress on the solution of the memory-leak The test consists doing 100 navigations between 2 pages, as the image below suggest <img width="876" height="502" alt="image" src="https://github.com/user-attachments/assets/e9e80768-dd40-4445-9fc8-90469579236c" /> Running the gc-dump on the desired objects this is what I found. > BaseLine is the dump when the app starts. | | | | | | | ------------------------------------ | ------- | ------------ | -------------- | ------------------------ | | Object Type | Count | Size (Bytes) | Expected Count | Status | | **Page2** | **100** | 84,000 | 1 | ❌ **LEAKED** (99 extra) | | **PageHandler** | **103** | 9,888 | 4 | ❌ **LEAKED** (99 extra) | | **StackNavigationManager** | **102** | 16,320 | 1 | ❌ **LEAKED** (101 extra) | | **StackNavigationManager.Callbacks** | **102** | 5,712 | 1 | ❌ **LEAKED** (101 extra) | | **NavigationViewFragment** | **102** | 5,712 | ~2 | ❌ **LEAKED** (100 extra) | So the first fix was to call `Disconnect` handler, on the `previousDetail` during the `FlyoutPage.Detail_set`. The PageChanges and Navigated events will not see this, since this `set` is not considered a navigation in .Net Maui. After that we see the following data | | | | | | | ------------------------------------ | ------- | ------------ | ---------------------- | ------------ | | Object Type | Count | Size (Bytes) | vs Baseline | Status | | **Page2** | **100** | 84,000 | Same | ❌ **LEAKED** | | **PageHandler** | **103** | 9,888 | Same | ❌ **LEAKED** | | **StackNavigationManager** | **102** | 16,320 | Same | ❌ **LEAKED** | | **StackNavigationManager.Callbacks** | **1** | 56 | ✅ **FIXED!** (was 102) | ✅ **Good!** | | **NavigationViewFragment** | **102** | 5,712 | Same | ❌ **LEAKED** | So, calling the Disconnect handler will fix the leak at `StackNavigationManager.Callbacks`. Next step was to investigate the `StackNavigationManager` and see what's holding it. On `StackNavigationManager` I see lot of object that should be cleaned up in order to release other references from it. After cleaning it up the result is | | | | | | |---|---|---|---|---| |Object Type|Count|Size (Bytes)|vs Baseline|Status| |**Page2**|**1**|840|✅ **FIXED!** (was 100)|✅ **Perfect!**| |**PageHandler**|**4**|384|✅ **FIXED!** (was 103)|✅ **Perfect!**| |**StackNavigationManager**|**102**|16,320|❌ Still leaking (was 102)|❌ **Unchanged**| |**StackNavigationManager.Callbacks**|**1**|56|✅ Fixed (was 102)|✅ **Good!**| |**NavigationViewFragment**|**102**|5,712|❌ Still leaking (was 102)|❌ **Unchanged**| So something is still holding the `StackNavigationManager` and `NavigationViewFragment` so I changed the approach and found that `NavigationViewFragment` is holding everything and after fixing that, cleaning it up on `Destroy` method. here's the result | | | | | | |---|---|---|---|---| |Object Type|Count|Size (Bytes)|vs Previous|Status| |**Page2**|**1**|840|✅ Same|✅ **Perfect!**| |**PageHandler**|**4**|384|✅ Same|✅ **Perfect!**| |**StackNavigationManager**|**1**|160|🎉 **FIXED!** (was 102)|🎉 **FIXED!**| |**StackNavigationManager.Callbacks**|**1**|56|✅ Same|✅ **Perfect!**| |**NavigationViewFragment**|**102**|5,712|⚠️ Still present|⚠️ **Remaining**| With that there's still the leak of `NavigationViewFragment`, looking at the graph the something on Android side is holding it, there's no root into managed objects, as far the gcdump can tell. I tried to cleanup the `FragmentManager`, `NavController` and so on but without success (maybe I did it wrong). There's still one instance of page 2, somehow it lives longer, I don't think it's a leak object because since its value is 1. For reference the Page2 graph is ``` Page2 (1 instance) └── PageHandler └── EventHandler<FocusChangeEventArgs> └── IOnFocusChangeListenerImplementor (Native Android) └── UNDEFINED ``` Looking into www I found that android caches those Fragments, sadly in our case we don't reuse them. The good part is each object has only 56 bytes, so it shouldn't be a big deal, I believe we can take the improvements made by this PR and keep an eye on that, maybe that's fixed when moved to Navigation3 implementation.
…34277) <!-- Please let the below note in for people that find this PR --> > [!NOTE] > Are you waiting for the changes in this PR to be merged? > It would be very helpful if you could [test the resulting artifacts](https://github.com/dotnet/maui/wiki/Testing-PR-Builds) from this PR and let us know in a comment if this change resolves your issue. Thank you! ## Summary Adds the `maui device list` command specification to the existing CLI design document (`docs/design/cli.md`). This command provides unified, cross-platform device enumeration without requiring a project context. See [PR dotnet#33865](dotnet#33865) for the full DevTools spec. ## What is added ### Command: `maui device list [--platform <p>] [--json]` Lists connected devices, running emulators, and available simulators across all platforms in a single call. ### Two approaches analysis The spec analyzes two discovery approaches and recommends the project-free one: | | MSBuild (`dotnet run --list-devices`) | Native CLI (`maui device list`) | |---|---|---| | Project required | Yes | **No** | | Cross-platform | One TFM at a time | All platforms at once | | Speed | Slower (MSBuild eval) | Fast (<2s) | | ID compatibility | Source of truth | Same native IDs | ### Scenarios requiring project-free discovery 1. AI agent bootstrapping (no `.csproj` yet) 2. IDE startup (device picker before project loads) 3. Environment validation ("can I see my phone?") 4. CI pipeline setup (check emulator before build) 5. Multi-project solutions (unified view) 6. Cross-platform overview (all devices at once) ### Recommended approach `maui device list` uses direct native tool invocation (`adb devices`, `simctl list`, `devicectl list`). Device IDs are compatible with `dotnet run --device`, making them complementary: ``` maui device list → "What devices exist on this machine?" dotnet run --list-devices → "What devices can run this project?" ``` ### Other changes - Added references to `ComputeAvailableDevices` MSBuild targets in [dotnet/android](https://github.com/dotnet/android) and [dotnet/macios](https://github.com/dotnet/macios) - Updated AI agent workflow example to include device discovery step - Fixed AppleDev.Tools reference (xcdevice → devicectl) Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
…otnet#34727) ## Description Fixes dotnet#34726 The XAML source generator interpolates `x:Key` values directly into generated C# string literals without escaping special characters. If an `x:Key` contains a double quote (`"`), backslash (`\`), or control character, the generated C# is syntactically invalid. ## Changes - **`SetPropertyHelpers.cs`** — Escape the `x:Key` value via `CSharpExpressionHelpers.EscapeForString()` before interpolating into the generated `resources["..."] = ...` assignment. - **`KnownMarkups.cs`** — Same fix for `DynamicResource` key emission (`new DynamicResource("...")`). - **`CSharpExpressionHelpers.cs`** — Changed `EscapeForString` from `private static` to `internal static` so it can be reused from `SetPropertyHelpers` and `KnownMarkups`. ## Test Added `Maui34726.xaml` / `.xaml.cs` XAML unit test with `x:Key` values containing double quotes and backslashes: - **SourceGen path**: Verifies the generated code compiles without diagnostics - **Runtime/XamlC paths**: Verifies the resources are accessible by their original (unescaped) key names Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
…otnet#34576) > [!NOTE] > Are you waiting for the changes in this PR to be merged? > It would be very helpful if you could [test the resulting artifacts](https://github.com/dotnet/maui/wiki/Testing-PR-Builds) from this PR and let us know in a comment if this change resolves your issue. Thank you! ## Summary Major improvements to the Essentials AI sample app: new Landmark Detail page with AI features, semantic search, streaming response handler improvements, and comprehensive UI polish across all pages for a modern, edge-to-edge experience on iOS and MacCatalyst. ### New Features - **Landmark Detail Page** — New intermediate page between browsing and trip planning with AI-generated travel tips, similar destinations via semantic search, animated hashtag tags, language picker, full-bleed hero image with scrolling gradient overlay, and Plan Trip navigation - **Semantic Search** — Search bar on Landmarks page filters continent groups using semantic similarity with `Timer`-based debounce (300ms) and tracks recent searches for contextual AI descriptions - **ISemanticSearchService abstraction** — Clean interface backed by `EmbeddingSearchService` (Apple NL embeddings + cosine similarity + hybrid keyword boost + sentence chunking) - **StreamingResponseHandler passthrough mode** — Supports streaming without buffering, with new device tests (`DeliversMultipleIncrementalUpdates`, `ConcatenatedText`) - **PromptBasedSchemaClient** — Prompt-based JSON schema middleware for Phi Silica compatibility ### UI Polish - **Edge-to-edge layout** — All pages use `SafeAreaEdges="None"` on root Grid with back buttons wrapped in `SafeAreaEdges="Container"` - **Scrolling gradient pattern** — Fixed gradient overlay + scrolling gradient that transitions to solid background - **Custom search entry** — Replaced `SearchBar` with borderless `Entry` in rounded `Border` with native border/focus ring removed on iOS/MacCatalyst - **Edge-to-edge horizontal scrollers** — Scroll to screen edges, align with page content padding at rest - **Removed broken BoxView global style** — Was breaking gradient overlays - **Added missing Gray700 color** — Prevented silent navigation crash - **Background → Background property migration** — Replaced `BackgroundColor` with `Background` across all pages ### Code Quality - **IDispatcher injection** instead of `MainThread.BeginInvokeOnMainThread` - **Only-once AI initialization** — Guard against re-entry in `LandmarkDetailViewModel` - **Null safety** — Fixed CS8602 warnings in DataService search methods - **Removed debug logging** ### Navigation Flow `LandmarksPage` (browse + search) → `LandmarkDetailPage` (details + AI tips) → `TripPlanningPage` (itinerary generation) ### Deleted Files - `LandmarkDescriptionView`, `LandmarkTripView` — Replaced by new pages - `LanguagePreferenceService` — Refactored into inline language array ### New Test Coverage - `StreamingResponseHandlerTests/Passthrough.cs` — Unit tests for passthrough streaming mode - `ChatClientStreamingTestsBase` — Updated device tests for streaming scenarios --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
…dotnet#34265) <!-- Please let the below note in for people that find this PR --> > [!NOTE] > Are you waiting for the changes in this PR to be merged? > It would be very helpful if you could [test the resulting artifacts](https://github.com/dotnet/maui/wiki/Testing-PR-Builds) from this PR and let us know in a comment if this change resolves your issue. Thank you! ## Summary Adds a standalone `code-review` skill for reviewing PR code changes with MAUI-specific domain knowledge, modeled after [dotnet/android's android-reviewer skill](https://github.com/dotnet/android/tree/main/.github/skills/android-reviewer). ### What's included **`.github/skills/code-review/SKILL.md`** (196 lines) — Review workflow: - Independence-first methodology (read code before PR description to avoid anchoring bias) - 6-step workflow: gather context → load rules → independent assessment → reconcile with PR narrative → check CI → devil's advocate verdict - Three verdicts: `LGTM`, `NEEDS_CHANGES`, `NEEDS_DISCUSSION` - Optional multi-model review for diverse perspectives - Output constraints: don't pile on, don't flag what CI catches **`.github/skills/code-review/references/review-rules.md`** (345 lines) — Domain knowledge: - **22 sections** of review rules distilled from real FTE code reviews - **138 PR citations** linking each rule to the PR where the pattern was identified - Sourced from **366 PRs**: top 142 most-discussed PRs (2,883 review comments), 30 reverted PRs, 50 candidate-branch failures, 64 regression fixes - Covers: handler lifecycle, memory leaks, safe area, layout, platform code, Android/iOS/Windows specifics, navigation, CollectionView, threading, XAML, testing, performance, error handling, API design, images, gestures, build, accessibility - §21 Regression Prevention — lessons from reverts and candidate failures - §22 Component Risk Table — ranking the most regression-prone components ### Design decisions - **Separation of concerns**: SKILL.md = workflow, review-rules.md = knowledge (follows Android's pattern) - **Standalone**: Can be invoked directly by users or by other agents. No coupling to the PR agent pipeline. - **Data-driven**: Every rule traces to a real PR where a senior maintainer flagged the pattern. No generic advice. ### Changes to copilot-instructions.md - Added "Opening PRs" section with required NOTE block template - Added code-review skill to the skills listing - Minor reformatting of existing documentation sections --------- Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com> Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Jakub Florkowski <kubaflo123@gmail.com>
🤖 AI Summary📊 Expand Full Review —
|
| # | Source | Approach | Test Result | Files Changed | Notes |
|---|---|---|---|---|---|
| PR | PR #34006 | Three-layer defense: volatile bool disposal flag + CoreWebView2 null check + try/catch InvalidOperationException + self-heal + MarkAsDisposing() | WebView2WebViewManager.cs, Wpf/BlazorWebView.cs |
Author verified fix manually on reproduction; WinForms parity intentionally omitted |
🔧 Fix — Analysis & Comparison
Fix Candidates
| # | Source | Approach | Test Result | Files Changed | Notes |
|---|---|---|---|---|---|
| 1 | try-fix (claude-opus-4.6) | Pure try/catch in SharedSource: CoreWebView2 null-check + catch (InvalidOperationException | ObjectDisposedException | COMException); no flag, no MarkAsDisposing() | 1 file | Simplest; single-file; catches broader exception set; no coupling to owning control | |
| 2 | try-fix (claude-sonnet-4.6) | WPF-specific subclass WpfWebView2WebViewManager overrides SendMessage() with guard + try/catch; SharedSource unchanged |
2 files (new file + Wpf/BlazorWebView.cs) | Cleanest architecture; keeps SharedSource pure; requires subclass wiring | |
| 3 | try-fix (gpt-5.3-codex) | CancellationTokenSource-based shutdown via StopAcceptingMessages() in SharedSource; idiomatic .NET cancellation pattern |
2 files | Most idiomatic .NET; more overhead; better composability with async patterns | |
| 4 | try-fix (gpt-4.1) | Interlocked.Exchange with int state field (0=active, 1=disposing) instead of volatile bool; stronger memory ordering | ❌ FAIL (build command error — infrastructure issue, not code) | 2 files | Stronger memory semantics than volatile bool; code is logically correct |
| PR | PR #34006 | Three-layer defense: volatile bool _isDisposing + CoreWebView2 null-check + try/catch InvalidOperationException + self-heal + MarkAsDisposing() called from WPF DisposeAsync() | 2 files | Author verified manually on reproduction; comprehensive defense-in-depth |
Cross-Pollination
| Model | Round | New Ideas? | Details |
|---|---|---|---|
| claude-opus-4.6 | 2 | NO NEW IDEAS | All viable strategies covered: guard flags (volatile/Interlocked), exception handling, CancellationToken, subclass override. PR's fix is most comprehensive. |
Exhausted: Yes — all 4 models queried; cross-pollination produced no new viable approaches
Selected Fix: PR's fix — All alternatives compile (Attempt 4 had a build command error, not a code error) but all are BLOCKED (no runtime test). The PR's three-layer approach is defense-in-depth appropriate for COM disposal races. Attempt 1 is the simplest alternative (broader exception catch, single file, no state), but the PR's approach is more explicit about lifecycle semantics and was manually verified by the author on the exact reproduction case.
📋 Report — Final Recommendation
⚠️ Final Recommendation: REQUEST CHANGES
Phase Status
| Phase | Status | Notes |
|---|---|---|
| Pre-Flight | ✅ COMPLETE | WPF BlazorWebView race condition; Windows only; no tests |
| Gate | No tests detected in PR | |
| Try-Fix | ✅ COMPLETE | 4 attempts, all BLOCKED (no runtime test for race condition); cross-pollination exhausted |
| Report | ✅ COMPLETE |
Summary
PR #34006 fixes a real regression (introduced in PR #31777) where closing a WPF window while Blazor messages are queued causes an InvalidOperationException on a disposed CoreWebView2 COM object. The fix is technically sound and the author manually verified it on the reproduction case. The prior agent review (on the same commit 23434b1) recommended REQUEST CHANGES for one outstanding item: a code comment explaining why WinForms does not call MarkAsDisposing(). The author responded without making that change, stating "No changes required for WinForms because the issue is limited to the WPF Blazor flow."
The single remaining issue is a documentation gap: MarkAsDisposing()'s XML doc comment says "This should be called by the owning control before starting disposal", but WinForms does not call it — with no comment in the code explaining why. This is a maintainability concern for future contributors who may not have access to the PR discussion.
Root Cause
PR #31777 migrated WPF BlazorWebView from WebView2 to WebView2CompositionControl, changing disposal timing. When a WPF window closes, queued Blazor messages on the dispatcher thread may still execute SendMessage() → PostWebMessageAsString() after CoreWebView2 has been disposed, throwing InvalidOperationException.
Fix Quality
What's good:
volatile bool _isDisposingis correctly typed for cross-thread read/writeSendMessage()correctly reads_webview.CoreWebView2(no dead null check on_webviewitself) and null-checksCoreWebView2which CAN be null before initialization- Try/catch around
PostWebMessageAsString()is the documented pattern for COM objects without exposed disposal state - Self-healing behavior (
_isDisposing = trueon exception) prevents repeated exception overhead MarkAsDisposing()called before_isDisposed = truein WPFDisposeAsync()provides a proactive fast-path
Issues found:
-
WinForms disposal difference undocumented (minor, but maintaining REQUEST CHANGES):
MarkAsDisposing()'s doc comment says "This should be called by the owning control before starting disposal", yet WinFormsBlazorWebView.Dispose(bool)does not call it. The code contains no explanation of why. A future maintainer reading this code will not know if this is intentional or an oversight.Requested change: Add one of these (pick either):
- A comment on
MarkAsDisposing()noting the WinForms exception: e.g.,// Note: WinForms BlazorWebView.Dispose() dispatches async disposal via ComponentsDispatcher.InvokeAsync(), which serializes disposal differently and does not exhibit this race condition. - OR a comment in WinForms
BlazorWebView.Dispose(bool disposing)explaining whyMarkAsDisposing()is not called
- A comment on
-
No tests (acknowledged): The author explicitly stated tests cannot be added for this race condition scenario, which is understandable given the timing-dependent COM disposal race.
Try-Fix Findings
All 4 alternative approaches compiled successfully but were BLOCKED (no runtime test). The simplest alternative (Attempt 1: pure try/catch with broader exception types, single file) demonstrates the fix can be achieved without MarkAsDisposing() — but the PR's approach is more explicit about lifecycle intent and was directly verified by the author on the reproduction case.
Selected Fix: PR's fix — Selected Fix: PR
kubaflo
left a comment
There was a problem hiding this comment.
Could you please check the AI's summary?
…on preventing InvalidOperationException when CoreWebView2 is disposed during SendMessage.
a2e423b to
d880b77
Compare
I’ve reviewed and updated the changes based on the latest findings:
|
Code Review — PR #34006Independent AssessmentWhat this changes: Adds defensive disposal handling to Inferred motivation: Race condition where Blazor messages queued on the WPF dispatcher execute Reconciliation with PR NarrativeAuthor claims: Race condition in WPF BlazorWebView from PR #31777 regression (WebView2 → WebView2CompositionControl migration changed disposal timing). Findings✅ Good — Defense-in-depth is appropriate for COM disposal races
💡 Suggestion — Document why WinForms and MAUI Windows don't call
|
… Webview Control (#34006) <!-- Please let the below note in for people that find this PR --> > [!NOTE] > Are you waiting for the changes in this PR to be merged? > It would be very helpful if you could [test the resulting artifacts](https://github.com/dotnet/maui/wiki/Testing-PR-Builds) from this PR and let us know in a comment if this change resolves your issue. Thank you! ### Root cause The issue occurs due to a race condition in the WPF BlazorWebView during window closure. When a Blazor component sends messages to the WebView, those messages are queued on the WPF dispatcher thread. If the window is closed while messages are still pending, WPF begins disposing of the WebView2CompositionControl along with its underlying CoreWebView2 COM object. Once disposal starts, the queued messages may still execute. When they attempt to call PostWebMessageAsString(), the method tries to access an already disposed COM object, which results in an InvalidOperationException. **Regression PR:** #31777 ### Description of Issue Fix The fix introduces a three-layer defense mechanism to safely handle the disposal race condition. The first layer adds a disposal flag that is checked before sending any message. If disposal has already started, the method exits immediately to prevent further operations. The second layer wraps the PostWebMessageAsString() call in a try-catch block to safely handle cases where the COM object is disposed externally before the disposal flag is updated. The third layer sets the disposal flag when an exception is caught. This creates a self-healing behavior that prevents repeated attempts after disposal is detected. The solution ensures safe execution during both normal disposal and race scenarios while maintaining minimal performance overhead. Tested the behavior in the following platforms. - [x] Windows - [ ] Mac - [ ] iOS - [ ] Android **Testcase:** Unable to add a test case for this scenario due to the reported issue with the WPF Blazor WebView. ### Issues Fixed <!-- Please make sure that there is a bug logged for the issue being fixed. The bug should describe the problem and how to reproduce it. --> Fixes #32944 <!-- Are you targeting main? All PRs should target the main branch unless otherwise noted. --> ### Output Before Issue Fix | After Issue Fix | |----------|----------| |<video width="100" height="100" alt="Before Fix" src="https://github.com/user-attachments/assets/865a8076-865a-466b-9aa2-44b7a338b112">|<video width="100" height="100" alt="After Fix" src="https://github.com/user-attachments/assets/c352506d-8ccc-4bb9-99b3-ea07509e3620">| --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Jakub Florkowski <kubaflo123@gmail.com>
… Webview Control (#34006) <!-- Please let the below note in for people that find this PR --> > [!NOTE] > Are you waiting for the changes in this PR to be merged? > It would be very helpful if you could [test the resulting artifacts](https://github.com/dotnet/maui/wiki/Testing-PR-Builds) from this PR and let us know in a comment if this change resolves your issue. Thank you! ### Root cause The issue occurs due to a race condition in the WPF BlazorWebView during window closure. When a Blazor component sends messages to the WebView, those messages are queued on the WPF dispatcher thread. If the window is closed while messages are still pending, WPF begins disposing of the WebView2CompositionControl along with its underlying CoreWebView2 COM object. Once disposal starts, the queued messages may still execute. When they attempt to call PostWebMessageAsString(), the method tries to access an already disposed COM object, which results in an InvalidOperationException. **Regression PR:** #31777 ### Description of Issue Fix The fix introduces a three-layer defense mechanism to safely handle the disposal race condition. The first layer adds a disposal flag that is checked before sending any message. If disposal has already started, the method exits immediately to prevent further operations. The second layer wraps the PostWebMessageAsString() call in a try-catch block to safely handle cases where the COM object is disposed externally before the disposal flag is updated. The third layer sets the disposal flag when an exception is caught. This creates a self-healing behavior that prevents repeated attempts after disposal is detected. The solution ensures safe execution during both normal disposal and race scenarios while maintaining minimal performance overhead. Tested the behavior in the following platforms. - [x] Windows - [ ] Mac - [ ] iOS - [ ] Android **Testcase:** Unable to add a test case for this scenario due to the reported issue with the WPF Blazor WebView. ### Issues Fixed <!-- Please make sure that there is a bug logged for the issue being fixed. The bug should describe the problem and how to reproduce it. --> Fixes #32944 <!-- Are you targeting main? All PRs should target the main branch unless otherwise noted. --> ### Output Before Issue Fix | After Issue Fix | |----------|----------| |<video width="100" height="100" alt="Before Fix" src="https://github.com/user-attachments/assets/865a8076-865a-466b-9aa2-44b7a338b112">|<video width="100" height="100" alt="After Fix" src="https://github.com/user-attachments/assets/c352506d-8ccc-4bb9-99b3-ea07509e3620">| --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Jakub Florkowski <kubaflo123@gmail.com>
… Webview Control (dotnet#34006) <!-- Please let the below note in for people that find this PR --> > [!NOTE] > Are you waiting for the changes in this PR to be merged? > It would be very helpful if you could [test the resulting artifacts](https://github.com/dotnet/maui/wiki/Testing-PR-Builds) from this PR and let us know in a comment if this change resolves your issue. Thank you! ### Root cause The issue occurs due to a race condition in the WPF BlazorWebView during window closure. When a Blazor component sends messages to the WebView, those messages are queued on the WPF dispatcher thread. If the window is closed while messages are still pending, WPF begins disposing of the WebView2CompositionControl along with its underlying CoreWebView2 COM object. Once disposal starts, the queued messages may still execute. When they attempt to call PostWebMessageAsString(), the method tries to access an already disposed COM object, which results in an InvalidOperationException. **Regression PR:** dotnet#31777 ### Description of Issue Fix The fix introduces a three-layer defense mechanism to safely handle the disposal race condition. The first layer adds a disposal flag that is checked before sending any message. If disposal has already started, the method exits immediately to prevent further operations. The second layer wraps the PostWebMessageAsString() call in a try-catch block to safely handle cases where the COM object is disposed externally before the disposal flag is updated. The third layer sets the disposal flag when an exception is caught. This creates a self-healing behavior that prevents repeated attempts after disposal is detected. The solution ensures safe execution during both normal disposal and race scenarios while maintaining minimal performance overhead. Tested the behavior in the following platforms. - [x] Windows - [ ] Mac - [ ] iOS - [ ] Android **Testcase:** Unable to add a test case for this scenario due to the reported issue with the WPF Blazor WebView. ### Issues Fixed <!-- Please make sure that there is a bug logged for the issue being fixed. The bug should describe the problem and how to reproduce it. --> Fixes dotnet#32944 <!-- Are you targeting main? All PRs should target the main branch unless otherwise noted. --> ### Output Before Issue Fix | After Issue Fix | |----------|----------| |<video width="100" height="100" alt="Before Fix" src="https://github.com/user-attachments/assets/865a8076-865a-466b-9aa2-44b7a338b112">|<video width="100" height="100" alt="After Fix" src="https://github.com/user-attachments/assets/c352506d-8ccc-4bb9-99b3-ea07509e3620">| --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Jakub Florkowski <kubaflo123@gmail.com>
… Webview Control (#34006) <!-- Please let the below note in for people that find this PR --> > [!NOTE] > Are you waiting for the changes in this PR to be merged? > It would be very helpful if you could [test the resulting artifacts](https://github.com/dotnet/maui/wiki/Testing-PR-Builds) from this PR and let us know in a comment if this change resolves your issue. Thank you! ### Root cause The issue occurs due to a race condition in the WPF BlazorWebView during window closure. When a Blazor component sends messages to the WebView, those messages are queued on the WPF dispatcher thread. If the window is closed while messages are still pending, WPF begins disposing of the WebView2CompositionControl along with its underlying CoreWebView2 COM object. Once disposal starts, the queued messages may still execute. When they attempt to call PostWebMessageAsString(), the method tries to access an already disposed COM object, which results in an InvalidOperationException. **Regression PR:** #31777 ### Description of Issue Fix The fix introduces a three-layer defense mechanism to safely handle the disposal race condition. The first layer adds a disposal flag that is checked before sending any message. If disposal has already started, the method exits immediately to prevent further operations. The second layer wraps the PostWebMessageAsString() call in a try-catch block to safely handle cases where the COM object is disposed externally before the disposal flag is updated. The third layer sets the disposal flag when an exception is caught. This creates a self-healing behavior that prevents repeated attempts after disposal is detected. The solution ensures safe execution during both normal disposal and race scenarios while maintaining minimal performance overhead. Tested the behavior in the following platforms. - [x] Windows - [ ] Mac - [ ] iOS - [ ] Android **Testcase:** Unable to add a test case for this scenario due to the reported issue with the WPF Blazor WebView. ### Issues Fixed <!-- Please make sure that there is a bug logged for the issue being fixed. The bug should describe the problem and how to reproduce it. --> Fixes #32944 <!-- Are you targeting main? All PRs should target the main branch unless otherwise noted. --> ### Output Before Issue Fix | After Issue Fix | |----------|----------| |<video width="100" height="100" alt="Before Fix" src="https://github.com/user-attachments/assets/865a8076-865a-466b-9aa2-44b7a338b112">|<video width="100" height="100" alt="After Fix" src="https://github.com/user-attachments/assets/c352506d-8ccc-4bb9-99b3-ea07509e3620">| --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Jakub Florkowski <kubaflo123@gmail.com>
… Webview Control (#34006) <!-- Please let the below note in for people that find this PR --> > [!NOTE] > Are you waiting for the changes in this PR to be merged? > It would be very helpful if you could [test the resulting artifacts](https://github.com/dotnet/maui/wiki/Testing-PR-Builds) from this PR and let us know in a comment if this change resolves your issue. Thank you! ### Root cause The issue occurs due to a race condition in the WPF BlazorWebView during window closure. When a Blazor component sends messages to the WebView, those messages are queued on the WPF dispatcher thread. If the window is closed while messages are still pending, WPF begins disposing of the WebView2CompositionControl along with its underlying CoreWebView2 COM object. Once disposal starts, the queued messages may still execute. When they attempt to call PostWebMessageAsString(), the method tries to access an already disposed COM object, which results in an InvalidOperationException. **Regression PR:** #31777 ### Description of Issue Fix The fix introduces a three-layer defense mechanism to safely handle the disposal race condition. The first layer adds a disposal flag that is checked before sending any message. If disposal has already started, the method exits immediately to prevent further operations. The second layer wraps the PostWebMessageAsString() call in a try-catch block to safely handle cases where the COM object is disposed externally before the disposal flag is updated. The third layer sets the disposal flag when an exception is caught. This creates a self-healing behavior that prevents repeated attempts after disposal is detected. The solution ensures safe execution during both normal disposal and race scenarios while maintaining minimal performance overhead. Tested the behavior in the following platforms. - [x] Windows - [ ] Mac - [ ] iOS - [ ] Android **Testcase:** Unable to add a test case for this scenario due to the reported issue with the WPF Blazor WebView. ### Issues Fixed <!-- Please make sure that there is a bug logged for the issue being fixed. The bug should describe the problem and how to reproduce it. --> Fixes #32944 <!-- Are you targeting main? All PRs should target the main branch unless otherwise noted. --> ### Output Before Issue Fix | After Issue Fix | |----------|----------| |<video width="100" height="100" alt="Before Fix" src="https://github.com/user-attachments/assets/865a8076-865a-466b-9aa2-44b7a338b112">|<video width="100" height="100" alt="After Fix" src="https://github.com/user-attachments/assets/c352506d-8ccc-4bb9-99b3-ea07509e3620">| --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Jakub Florkowski <kubaflo123@gmail.com>
… Webview Control (#34006) <!-- Please let the below note in for people that find this PR --> > [!NOTE] > Are you waiting for the changes in this PR to be merged? > It would be very helpful if you could [test the resulting artifacts](https://github.com/dotnet/maui/wiki/Testing-PR-Builds) from this PR and let us know in a comment if this change resolves your issue. Thank you! ### Root cause The issue occurs due to a race condition in the WPF BlazorWebView during window closure. When a Blazor component sends messages to the WebView, those messages are queued on the WPF dispatcher thread. If the window is closed while messages are still pending, WPF begins disposing of the WebView2CompositionControl along with its underlying CoreWebView2 COM object. Once disposal starts, the queued messages may still execute. When they attempt to call PostWebMessageAsString(), the method tries to access an already disposed COM object, which results in an InvalidOperationException. **Regression PR:** #31777 ### Description of Issue Fix The fix introduces a three-layer defense mechanism to safely handle the disposal race condition. The first layer adds a disposal flag that is checked before sending any message. If disposal has already started, the method exits immediately to prevent further operations. The second layer wraps the PostWebMessageAsString() call in a try-catch block to safely handle cases where the COM object is disposed externally before the disposal flag is updated. The third layer sets the disposal flag when an exception is caught. This creates a self-healing behavior that prevents repeated attempts after disposal is detected. The solution ensures safe execution during both normal disposal and race scenarios while maintaining minimal performance overhead. Tested the behavior in the following platforms. - [x] Windows - [ ] Mac - [ ] iOS - [ ] Android **Testcase:** Unable to add a test case for this scenario due to the reported issue with the WPF Blazor WebView. ### Issues Fixed <!-- Please make sure that there is a bug logged for the issue being fixed. The bug should describe the problem and how to reproduce it. --> Fixes #32944 <!-- Are you targeting main? All PRs should target the main branch unless otherwise noted. --> ### Output Before Issue Fix | After Issue Fix | |----------|----------| |<video width="100" height="100" alt="Before Fix" src="https://github.com/user-attachments/assets/865a8076-865a-466b-9aa2-44b7a338b112">|<video width="100" height="100" alt="After Fix" src="https://github.com/user-attachments/assets/c352506d-8ccc-4bb9-99b3-ea07509e3620">| --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: Jakub Florkowski <kubaflo123@gmail.com>
## Blazor - Fix: Filter precompressed RCL assets from MAUI Blazor Hybrid APKs by @mattleibow in #33917 <details> <summary>🔧 Fixes</summary> - [.NET MAUI Blazor Hybrid App should not precompress assets](#33773) </details> - [Windows] Fix for Runtime error when closing external window with WPF Webview Control by @BagavathiPerumal in #34006 <details> <summary>🔧 Fixes</summary> - [Runtime error when closing external window with WPF Webview Control](#32944) </details> ## Button - [Android] ImageButton CornerRadius not being applied - fix by @kubaflo in #30074 <details> <summary>🔧 Fixes</summary> - [ImageButton CornerRadius not being applied on Android](#23854) </details> - Fix Disabled visual state ignored when Button has locally-set BackgroundColor/TextColor by @Dhivya-SF4094 in #34444 <details> <summary>🔧 Fixes</summary> - [[regression/9.0] VisualState "Disabled" is not properly applied for Button with custom appearance](#34363) </details> ## CollectionView - Fix CollectionView grid spacing updates for first row and column by @KarthikRajaKalaimani in #34527 <details> <summary>🔧 Fixes</summary> - [[MAUI] I2_Vertical grid for horizontal Item Spacing and Vertical Item Spacing - horizontally updating the spacing only applies to the second column](#34257) </details> - Fix CollectionView record struct selection on Windows by @jeremy-visionaid in #33488 - [Android] Ensure disconnected ItemsViewHandler doesn't hold onto the items source by @filipnavara in #24610 <details> <summary>🔧 Fixes</summary> - [Crash on NullReferenceException with measurement cells in CollectionView](#24304) </details> - [Windows] Fixed VisualState Setters not working properly for CollectionView by @Dhivya-SF4094 in #27230 <details> <summary>🔧 Fixes</summary> - [VisualState Setters not working properly on Windows for a CollectionView](#27086) - [[regression/8.0.3] [Windows][CollectionView]Label Disappear when set Style in ContentPage.Resources](#19209) - [[Windows] Label style defined as ContentPage Resource doesn't propagate to CollectionView](#18701) </details> - [Windows] Fixed Margin doesn't work inside CollectionView EmptyView by @Dhivya-SF4094 in #29897 <details> <summary>🔧 Fixes</summary> - [Margin doesn't work inside CollectionView EmptyView](#8494) </details> - [Android, Windows] Fix CarouselView PreviousPosition/PreviousItem incorrect during animated ScrollTo() by @praveenkumarkarunanithi in #34570 <details> <summary>🔧 Fixes</summary> - [[Android] CurrentItemChangedEventArgs.PreviousItem and PositionChangedEventArgs.PreviousPosition Not Updating Correctly When Using ScrollTo or Setting Position](#29544) </details> - [iOS] CarouselView2: Update internal scroll indicators for compositional layout by @SubhikshaSf4851 in #33639 <details> <summary>🔧 Fixes</summary> - [[iOS] Horizontal Scroll Bar Not Visible on CarouselView (CV2)](#29390) </details> - [CarouselViewHandler2] Fir fox CurrentItem does not work when ItemSpacing is set by @SyedAbdulAzeemSF4852 in #32135 <details> <summary>🔧 Fixes</summary> - [[CarouselViewHandler2] CurrentItem does not work when ItemSpacing is set](#32048) </details> - [iOS] Fix for Incorrect Scroll in Loop Mode When CurrentItem Is Not Found in ItemsSource by @SyedAbdulAzeemSF4852 in #32141 <details> <summary>🔧 Fixes</summary> - [[Android & iOS] Setting an invalid CurrentItem causes scroll to last item in looped CarouselView](#32139) </details> - [Android] IndicatorView: Add TalkBack accessibility descriptions for indicators by @praveenkumarkarunanithi in #31775 <details> <summary>🔧 Fixes</summary> - [[Android] IndicatorView does not convey correct accessibility information](#31446) </details> - [iOS, macOS] Fixed CollectionView KeepLastItemInView Not Updating Correctly When Items Are Added Dynamically by @NanthiniMahalingam in #32191 <details> <summary>🔧 Fixes</summary> - [[.NET10] I9 - Scroll_Position - "KeepLastItemInView" does not keep the last item at the end of the displayed list when adding new items.](#31825) </details> - [Windows, Android] Resolved issue with dynamic Header/Footer reassignment in CollectionView. by @prakashKannanSf3972 in #28403 <details> <summary>🔧 Fixes</summary> - [[Windows, Android] Toggling Header/Footer in CollectionView Dynamically is not working](#27959) - [CollectionView HeaderTemplate and FooterTemplate are not displayed when ItemsSource is initially set to null](#28337) - [[Android] Header and Footer Not Visible in CollectionView When EmptyView is Selected First](#28351) </details> - [Android] Fix CollectionView inside disabled RefreshView blocks scroll by @Vignesh-SF3580 in #34702 <details> <summary>🔧 Fixes</summary> - [C6-The C6 page cannot scroll on Windows and Android platforms.](#34666) </details> - [Android] CollectionView: Fix SelectedItem visual state not applying when re-selecting same item by @KarthikRajaKalaimani in #31591 <details> <summary>🔧 Fixes</summary> - [CollectionView - SelectedItem visual state manager not working](#20062) </details> - [Windows] Fixed CollectionView.EmptyView can not be removed by setting it to Null by @Dhivya-SF4094 in #29487 <details> <summary>🔧 Fixes</summary> - [[Windows] CollectionView.EmptyView can not be removed by setting it to Null](#18657) - [[Windows] EmptyViewTemplate Not Working in CarouselView](#29463) - [EmptyViewTemplate does not do anything](#18551) - [[MAUI] I5_EmptyView - The data template selector cannot display the correct string.](#23330) </details> - [iOS] Support for IsSwipeEnabled on CarouselView2 by @kubaflo in #29996 <details> <summary>🔧 Fixes</summary> - [[iOS] IsSwipeEnabled Not Working on CarouselView (CV2)](#29391) </details> - [iOS, MacOS] Fixed FlowDirection not working on Header/Footer in CollectionView by @Dhivya-SF4094 in #32775 <details> <summary>🔧 Fixes</summary> - [[iOS, MacOS] FlowDirection not working on Header/Footer in CollectionView](#32771) </details> - [iOS] CollectionView: Fix drag-and-drop reordering into empty groups by @SuthiYuvaraj in #34151 <details> <summary>🔧 Fixes</summary> - [CollectionView Drag and Drop Reordering Can't Drop in Empty Group](#12008) </details> - [Android] CollectionView: Fix drag-and-drop reordering into empty groups by @SuthiYuvaraj in #31867 <details> <summary>🔧 Fixes</summary> - [CollectionView Drag and Drop Reordering Can't Drop in Empty Group](#12008) </details> - [iOS] Fix vertical CarouselView MandatorySingle snapping on iOS by @Vignesh-SF3580 in #34700 <details> <summary>🔧 Fixes</summary> - [CarouselView vertical snap points ignored on iOS with Microsoft.Maui.Controls v10.0.20 (regression from v9.0.120)](#33308) </details> - [iOS26] Fix CarouselView scrolling to wrong item when navigating to last item by @Vignesh-SF3580 in #34013 <details> <summary>🔧 Fixes</summary> - [[iOS 26] CarouselView does not scroll to the correct last item](#33770) </details> - Fixed the OnPlatform does not work for header property in Collection view by @NanthiniMahalingam in #28935 <details> <summary>🔧 Fixes</summary> - [OnPlatform does not work in Header of CollectionView](#25124) </details> - [Android] [Candidate branch] Fix VerifySelectedItemClearsOnNullAssignment, CollectionViewSelectionShouldClear, SelectedItemVisualIsCleared UI test failure on Android by @KarthikRajaKalaimani in #34928 ## DateTimePicker - [iOS] Fix for DatePicker FlowDirection Not Working on iOS by @SyedAbdulAzeemSF4852 in #30193 <details> <summary>🔧 Fixes</summary> - [[iOS] DatePicker FlowDirection Not Working on iOS](#30065) </details> ## Drawing - [Shapes] Line: Fix asymmetric Stretch.None path translation when right/bottom edge overflows by @NirmalKumarYuvaraj in #34385 <details> <summary>🔧 Fixes</summary> - [Line coordinates not computed correctly](#11404) - [Lines not drawing correctly](#26961) </details> - [Android] Fixed GraphicsView drawable is visible outside the canvas by @NirmalKumarYuvaraj in #28353 <details> <summary>🔧 Fixes</summary> - [[Android] GraphicsView, The drawn image can also be visible outside the canvas](#20834) </details> - Fixed Custom Drawable does not support binding by @NirmalKumarYuvaraj in #29442 <details> <summary>🔧 Fixes</summary> - [Custom IDrawable control does not databind to a model property when used inside a CollectionView ItemTemplate](#20991) </details> - Added a support for GradientBrushes on Shape.Stroke by @kubaflo in #22208 <details> <summary>🔧 Fixes</summary> - [GradientBrushes are not supported on Shape.Stroke](#21983) </details> ## Editor - Fixed Editor HorizontalTextAlignment does not update at run time by @NirmalKumarYuvaraj in #25129 <details> <summary>🔧 Fixes</summary> - [Editor HorizontalTextAlignment Does not Works.](#10987) - [[iOS/MacOs] Right-To-Left (RTL) alignment is not applied to Editor placeholder](#30052) </details> - [Windows] Fixed Entry Editor placeholder Text CharacterSpacing by @SubhikshaSf4851 in #30324 <details> <summary>🔧 Fixes</summary> - [[Windows] CharacterSpacing not applied to Placeholder text in Entry and Editor controls](#30071) </details> ## Entry - [Windows] Fix fo setting an Entry's Keyboard to Date causes it to be interpreted as a password input by @SyedAbdulAzeemSF4852 in #29344 <details> <summary>🔧 Fixes</summary> - [[Windows] Entry Keyboad-Type "Date" results in Password-Entry](#28975) </details> - [Android] Exception thrown when give more than 5000 characters to the Text property of Entry. by @KarthikRajaKalaimani in #30242 <details> <summary>🔧 Fixes</summary> - [Android crash when Entry has >5000 characters](#30144) </details> ## Essentials - Bump MonoApiToolsMSBuildTasksPackageVersion to 0.5.0 and ship Essentials.AI public APIs by @mattleibow via @Copilot in #34574 - [Mac] DeviceDisplay.KeepScreenOn not being respected on Mac OS by @HarishwaranVijayakumar in #32708 <details> <summary>🔧 Fixes</summary> - [[Mac Catalyst] DeviceDisplay.KeepScreenOn not being respected on Mac OS](#26059) </details> ## Flyoutpage - [Windows] FlyoutPage: update CollapseStyle at runtime by @devanathan-vaithiyanathan in #29927 <details> <summary>🔧 Fixes</summary> - [Flyout Page SetCollapseStyle doesn't have any change](#18200) </details> ## Gestures - [Android] Fix for TapGestureRecognizer doesn't fire by @HarishwaranVijayakumar in #34497 <details> <summary>🔧 Fixes</summary> - [[Android] TapGestureRecognizer doesn't fire](#5825) </details> ## Image - [Android] Fix Share.RequestAsync SecurityException on Android 10+ caused by missing ClipData by @HarishwaranVijayakumar in #34417 <details> <summary>🔧 Fixes</summary> - [[Bug] Share.RequestAsync throws java.lang.SecurityException (uid=1000) on Android 10+ due to missing intent.ClipData](#34370) </details> - [Windows]Fixed the MauiImage with logical name containing path issue by @sheiksyedm in #32864 <details> <summary>🔧 Fixes</summary> - [MauiImage with LogicalName containing path - is not working on Windows](#32356) </details> - [Android, Windows & iOS] Fix Downsize/ScaleImage to maintain aspect ratio and prevent upscaling by @SyedAbdulAzeemSF4852 in #30808 <details> <summary>🔧 Fixes</summary> - [[Android & Windows] In GraphicsView, the aspect ratio is not maintained when Downsize is called with both maxWidth and maxHeight](#30803) </details> ## Label - [iOS , macOS] Fixed Label text cropping when a width request is specified on the label inside a VerticalStackLayout with specified width request by @NanthiniMahalingam in #29166 <details> <summary>🔧 Fixes</summary> - [Label text gets cropped when a width request is specified on the label inside a VerticalStackLayout](#28660) - [[iOS] Label with a fixed WidthRequest has wrong height](#26644) </details> - [Android] Fix Label word wrapping clips text depending on alignment and layout options by @Dhivya-SF4094 in #34533 <details> <summary>🔧 Fixes</summary> - [Bug: Android Label word wrapping clips text depending on alignment and layout options](#34459) </details> - LineHeight and decorations for HTML Label - fix by @kubaflo in #31202 <details> <summary>🔧 Fixes</summary> - [LineHeight with HTML Label not working](#22193) - [lineheight is broken ](#22197) </details> - [iOS] Fix Label with TailTruncation not rendering after empty-to-non-empty text transition by @kubaflo in #34812 <details> <summary>🔧 Fixes</summary> - [Label with LineBreakMode="TailTruncation" does not render text if initial Text is null or empty on first render (iOS)](#34591) </details> ## Layout - [Android] Fix overflowing children clipped when parent Opacity < 1 by @SyedAbdulAzeemSF4852 in #34565 <details> <summary>🔧 Fixes</summary> - [Maui Android parent view inappropriately creates clipping mask when its opacity is less than 1, cropping out children](#22038) </details> - Fixed the FlexLayout reverse issue with the AlignContent by @Ahamed-Ali in #32134 <details> <summary>🔧 Fixes</summary> - [FlexLayout alignment issue when Wrap is set to Reverse and AlignContent is set to SpaceAround, SpaceBetween or SpaceEvenly](#31565) </details> - [iOS/Mac] Fixed BoxView in AbsoluteLayout did not return to its default AutoSize for Height and Width after reset by @Dhivya-SF4094 in #31648 <details> <summary>🔧 Fixes</summary> - [[iOS, Catalyst] BoxView in AbsoluteLayout does not return to default AutoSize for Height/Width after reset](#31496) </details> ## Map - [Windows] Implement WinUI 3 MapControl handler using Azure Maps by @jfversluis in #34138 ## Modal - [Android] PopToRootAsync for modal pages - improvements by @kubaflo in #26851 <details> <summary>🔧 Fixes</summary> - [Shell PopToRootAsync doesn't happen instantly - previous pages flash quickly. Only happens in NET 9](#26846) </details> - [Android] Fix HideSoftInputOnTapped doesn't work on Modal Pages by @HarishwaranVijayakumar in #34770 <details> <summary>🔧 Fixes</summary> - [HideSoftInputOnTapped doesn't work on Modal Pages](#34730) </details> ## Navigation - [iOS] Alert popup may be displayed on wrong window when modal page navigation is in progress - fix by @kubaflo in #31016 <details> <summary>🔧 Fixes</summary> - [Alert popup may be displayed on wrong window when modal page navigation is in progress on iOS/MacOS](#30970) </details> - [Android] Page: Fix OnNavigatedTo called twice when NavigationPage is FlyoutPage Detail by @KarthikRajaKalaimani in #31931 <details> <summary>🔧 Fixes</summary> - [NavigationPage and FlyoutPage both call OnNavigatedTo, so it is called twice](#23902) </details> ## Picker - Fixed the Picker didn't dismiss it when tapping outside on iOS and MacCatalyst platform. by @KarthikRajaKalaimani in #30067 <details> <summary>🔧 Fixes</summary> - [[regression/8.0.3] iOS Picker dismiss does not work when clicking outside of the Picker](#19168) </details> - [Windows] Fixed Picker items width wont resize back by @SubhikshaSf4851 in #33042 <details> <summary>🔧 Fixes</summary> - [Picker items width won't resize back when its container window gets resized down.](#32984) </details> ## RadioButton - Fix TalkBack not correctly narrating RadioButtons with Content by @SubhikshaSf4851 in #34521 <details> <summary>🔧 Fixes</summary> - [[Android] TalkBack does not correctly narrate RadioButtons with Content](#34322) </details> ## SafeArea - [Android] Fix SafeAreaShouldWorkOnAllShellTabs test failure on API 36 by @praveenkumarkarunanithi in #34239 ## ScrollView - [iOS] Preserve ScrollView offsets when Orientation changes to Neither by @Vignesh-SF3580 in #34672 <details> <summary>🔧 Fixes</summary> - [Incorrect implementation of ScrollView.Orientation](#34583) </details> ## Searchbar - [Android] Fix SearchBar text bleeding between instances after navigation by @SyedAbdulAzeemSF4852 in #34703 <details> <summary>🔧 Fixes</summary> - [MAUI Android: SearchBar copies content from one to the other](#20348) </details> - Fixed SearchBar CursorPosition and SelectionLength not updating when typing by @Dhivya-SF4094 in #34347 <details> <summary>🔧 Fixes</summary> - [SearchBar - CursorPosition and SelectionLength are not updated when the user types](#30779) </details> ## SearchBar - [Windows] Fixed SearchHandler issues by @Tamilarasan-Paranthaman in #29520 <details> <summary>🔧 Fixes</summary> - [[Windows] SearchHandler APIs are not functioning properly](#29493) </details> ## Shell - [iOS, Mac] Fix for Background set to Transparent doesn't have the same behavior as BackgroundColor Transparent by @HarishwaranVijayakumar in #32245 <details> <summary>🔧 Fixes</summary> - [Background set to Transparent doesn't have the same behavior as BackgroundColor = Transparent](#22769) </details> - [iOS] Fix App crash with NullReferenceException in ShellSectionRenderer by @devanathan-vaithiyanathan in #32109 <details> <summary>🔧 Fixes</summary> - [[iOS] App crash with NullReferenceException in ShellSectionRenderer](#31961) </details> - [Android] Fixed back button icon selection logic in ShellToolbarTracker by @kubaflo in #32080 <details> <summary>🔧 Fixes</summary> - [IconOverride in Shell.BackButtonBehavior does not work.](#32050) </details> - Fix TabBarIsVisible Not Updating Dynamically When Set on ShellContent by @Vignesh-SF3580 in #33090 <details> <summary>🔧 Fixes</summary> - [Shell.TabBarIsVisible is not updated dynamically at runtime](#32994) </details> - [iOS, macOS] Shell: Fix RTL flow direction for flyout, menu cells, tab bar, and Locked flyout position by @NanthiniMahalingam in #32701 <details> <summary>🔧 Fixes</summary> - [[iOS, Mac Catalyst] Shell Flyout and Content Do Not Fully Support RightToLeft (RTL)](#32419) </details> - [IOS] Inconsistent Resize Behavior for Header/Footer - fix by @kubaflo in #28713 <details> <summary>🔧 Fixes</summary> - [[IOS, Mac] Inconsistent Resize Behavior for Header/Footer](#26397) - [Enable Shell Flyout Header/Footer resize tests on iOS/Catalyst](#33501) </details> - [Android] Fix for SearchHandler retaining previous page SearchView data in pages within Shell sections by @BagavathiPerumal in #29545 <details> <summary>🔧 Fixes</summary> - [[Shell][Android] The truth is out there...but not on top tab search handlers](#8716) </details> - [Android] Fix empty space above TabBar after navigating back when TabBar visibility is toggled by @praveenkumarkarunanithi in #34324 <details> <summary>🔧 Fixes</summary> - [Empty space appears above TabBar after navigating back when TabBar visibility is toggled](#33703) - [Grid with SafeAreaEdges=Container has incorrect size when tab bar appears](#34256) </details> ## SwipeView - [Android] SwipeView: Use MeasureSpecMode.Exactly for SwipeItem layout to fix text visibility by @Ahamed-Ali in #27399 <details> <summary>🔧 Fixes</summary> - [[Android] Right SwipeView items are not visible in the SwipeView.](#27367) </details> - [Android] Prevent the tap that closes an open SwipeView from being propagated to children by @sjordanGSS in #24275 <details> <summary>🔧 Fixes</summary> - [Tapping to close a SwipeView will activate TapGestureRecognizers on .Content](#23921) </details> ## Switch - [iOS & Mac] Fix for SearchHandler retains previous page state when switching top tabs by @BagavathiPerumal in #34735 <details> <summary>🔧 Fixes</summary> - [[Shell] [iOS & Mac] SearchHandler retains previous page state when switching top tabs](#34693) </details> ## TabbedPage - [Android] Fixed NullReferenceException in app with TabBar after returning from minimized state by @NirmalKumarYuvaraj in #34779 <details> <summary>🔧 Fixes</summary> - [NullReferenceException in app with TabBar after returning from minimized state](#34720) </details> ## Titlebar - Fixed BindingContext of the Window TitleBar is not being passed on to its child content. by @NirmalKumarYuvaraj in #30080 <details> <summary>🔧 Fixes</summary> - [The BindingContext of the Window TitleBar is not being passed on to its child content.](#24831) </details> - [Windows/Mac] Fix RTL FlowDirection causes overlap with native window control buttons in TitleBar by @devanathan-vaithiyanathan in #30400 <details> <summary>🔧 Fixes</summary> - [[Windows, Mac] RTL FlowDirection causes overlap with native window control buttons in TitleBar](#30399) </details> ## WebView - [Windows] Fix WebView background color not being applied by @SubhikshaSf4851 in #34599 <details> <summary>🔧 Fixes</summary> - [WebView background color has changed after update, can't override.](#34518) </details> - [Android] Fix for WebView/HybridWebView briefly flashes full screen before layout completes by @praveenkumarkarunanithi in #33207 <details> <summary>🔧 Fixes</summary> - [[Android] HybridWebView briefly resizes to full screen when page is opened before snapping back to correct size](#31475) </details> ## Xaml - Improved style inheritance by @kubaflo in #31317 <details> <summary>🔧 Fixes</summary> - [Styles based on a style that is based on another style that uses AppThemeBinding do not inherit properties correctly.](#31280) </details> - Fix for VisualStateManager Setter.TargetName failing when ControlTemplate is applied by @BagavathiPerumal in #33208 <details> <summary>🔧 Fixes</summary> - [Setter.TargetName + ControlTemplate crash](#26977) </details> <details> <summary>🧪 Testing (4)</summary> - [Testing] Additional Feature Matrix Event Test Cases for Slider and ScrollView by @nivetha-nagalingam in #34352 - [Testing] Fixed Build error on inflight/ candidate PR 34885 by @NafeelaNazhir in #34891 - [Testing] Fixed UI test image failure in PR 34885 - [13/4/2026] by @NafeelaNazhir in #34933 - Fixed test failure - CursorPositionUpdatesWhenSearchBarGainsFocus by @Dhivya-SF4094 in #34938 </details> <details> <summary>📦 Other (3)</summary> - Fix Loaded event not called for MAUI View added to native View by @NirmalKumarYuvaraj in #34345 <details> <summary>🔧 Fixes</summary> - [Loaded event not called for MAUI View added to native View](#34310) </details> - Add public IAlertManager and IAlertManagerSubscription interfaces by @Redth in #34228 <details> <summary>🔧 Fixes</summary> - [Alert/Dialog system (`DisplayAlert`, `DisplayActionSheet`, `DisplayPromptAsync`) needs a public extensibility point](#34104) </details> - Fix crash when displaying alerts on unloaded pages by @kubaflo in #33288 </details> <details> <summary>📝 Issue References</summary> Fixes #5825, Fixes #8494, Fixes #8716, Fixes #10987, Fixes #11404, Fixes #12008, Fixes #18200, Fixes #18551, Fixes #18657, Fixes #18701, Fixes #19168, Fixes #19209, Fixes #20062, Fixes #20348, Fixes #20834, Fixes #20991, Fixes #21983, Fixes #22038, Fixes #22193, Fixes #22197, Fixes #22769, Fixes #23330, Fixes #23854, Fixes #23902, Fixes #23921, Fixes #24304, Fixes #24831, Fixes #25124, Fixes #26059, Fixes #26397, Fixes #26644, Fixes #26846, Fixes #26961, Fixes #26977, Fixes #27086, Fixes #27367, Fixes #27959, Fixes #28337, Fixes #28351, Fixes #28660, Fixes #28975, Fixes #29390, Fixes #29391, Fixes #29463, Fixes #29493, Fixes #29544, Fixes #30052, Fixes #30065, Fixes #30071, Fixes #30144, Fixes #30399, Fixes #30779, Fixes #30803, Fixes #30970, Fixes #31280, Fixes #31446, Fixes #31475, Fixes #31496, Fixes #31565, Fixes #31825, Fixes #31961, Fixes #32048, Fixes #32050, Fixes #32139, Fixes #32356, Fixes #32419, Fixes #32771, Fixes #32944, Fixes #32984, Fixes #32994, Fixes #33308, Fixes #33501, Fixes #33703, Fixes #33770, Fixes #33773, Fixes #34104, Fixes #34256, Fixes #34257, Fixes #34310, Fixes #34322, Fixes #34363, Fixes #34370, Fixes #34459, Fixes #34518, Fixes #34583, Fixes #34591, Fixes #34666, Fixes #34693, Fixes #34720, Fixes #34730 </details> **Full Changelog**: main...inflight/candidate
Note
Are you waiting for the changes in this PR to be merged?
It would be very helpful if you could test the resulting artifacts from this PR and let us know in a comment if this change resolves your issue. Thank you!
Root cause
The issue occurs due to a race condition in the WPF BlazorWebView during window closure. When a Blazor component sends messages to the WebView, those messages are queued on the WPF dispatcher thread. If the window is closed while messages are still pending, WPF begins disposing of the WebView2CompositionControl along with its underlying CoreWebView2 COM object.
Once disposal starts, the queued messages may still execute. When they attempt to call PostWebMessageAsString(), the method tries to access an already disposed COM object, which results in an InvalidOperationException.
Regression PR: #31777
Description of Issue Fix
The fix introduces a three-layer defense mechanism to safely handle the disposal race condition.
The first layer adds a disposal flag that is checked before sending any message. If disposal has already started, the method exits immediately to prevent further operations.
The second layer wraps the PostWebMessageAsString() call in a try-catch block to safely handle cases where the COM object is disposed externally before the disposal flag is updated.
The third layer sets the disposal flag when an exception is caught. This creates a self-healing behavior that prevents repeated attempts after disposal is detected.
The solution ensures safe execution during both normal disposal and race scenarios while maintaining minimal performance overhead.
Tested the behavior in the following platforms.
Testcase: Unable to add a test case for this scenario due to the reported issue with the WPF Blazor WebView.
Issues Fixed
Fixes #32944
Output
32944-BeforeFix.mp4
32944-AfterFix.mp4