Stop AppHost debug session before Aspire parent

[adamint]

Description

Related to #18524. This intentionally does not auto-close the failing-test tracking issue until the normal tracking policy says it is safe to close.

The VS Code extension E2E global debug stop removes running apphost has been timing out on rolling main: the E2E state file shows debugSessions: [], but the global AppHost registry still contains the stopped AppHost PID/dashboard URL until after the test times out.

This makes the global stop path stop the real AppHost debug session before stopping the synthetic Aspire parent debug session. That matches the existing dispose ordering, but applies it to the explicit stopDebugging() path used by the E2E bridge and global debug stop. The parent session is still stopped in a finally block if the AppHost stop request fails.

I also made the tracked resource stop callback return the VS Code stop promise, so callers can await the child stop request, and added unit coverage for both the ordering and parent-stop fallback cases.

Validation:

• ./restore.sh
• dotnet build src/Aspire.Cli/Aspire.Cli.csproj /p:SkipNativeBuild=true --verbosity minimal
• cd extension && COREPACK_HOME="$PWD/.corepack-cache" corepack yarn run compile-tests
• cd extension && COREPACK_HOME="$PWD/.corepack-cache" corepack yarn run unit-test --config .vscode-test-short.mjs --run out/test/aspireDebugSession.test.js (20 passing)
• cd extension && COREPACK_HOME="$PWD/.corepack-cache" corepack yarn eslint src/debugger/AspireDebugSession.ts src/dcp/types.ts src/test/aspireDebugSession.test.ts

I attempted the targeted E2E locally with ASPIRE_EXTENSION_E2E_SPEC=out/test-e2e/test-e2e/debugDashboard.e2e.test.js. It did not reach the Aspire test: local macOS ExTester failed during VS Code/Chrome session creation with SessionNotCreatedError: Chrome instance exited. Before that, the local fixture restore also needed WarningsNotAsErrors=NU1507 because this checkout has multiple NuGet feeds without source mapping in the generated E2E workspace.

Checklist

• Is this feature complete?
  • Yes. Ready to ship.
  • No. Follow-up changes expected.
• Are you including unit tests for the changes and scenario tests if relevant?
  • Yes
  • No
• Did you add public API?
  • Yes
    • If yes, did you have an API Review for it?
      • Yes
      • No
    • Did you add <remarks /> and <code /> elements on your triple slash comments?
      • Yes
      • No
  • No
• Does the change make any security assumptions or guarantees?
  • Yes
    • If yes, have you done a threat model and had a security review?
      • Yes
      • No
  • No

[github-actions]

🚀 Dogfood this PR with:

⚠️ WARNING: Do not do this without first carefully reviewing the code of this PR to satisfy yourself it is safe.

curl -fsSL https://raw.githubusercontent.com/microsoft/aspire/main/eng/scripts/get-aspire-cli-pr.sh | bash -s -- 18561

Or

• Run remotely in PowerShell:

iex "& { $(irm https://raw.githubusercontent.com/microsoft/aspire/main/eng/scripts/get-aspire-cli-pr.ps1) } 18561"

[Copilot]

Pull request overview

This PR fixes flaky VS Code extension E2E test global debug stop removes running apphost (#18524), which was turning rolling main red. The test timed out because, after a global/E2E debug stop, the synthetic Aspire parent debug session was stopped but the real AppHost child session (and its global registry entry) lingered until the test timed out. The fix reorders AspireDebugSession.stopDebugging() to stop the real AppHost session before the synthetic Aspire parent — mirroring the existing dispose() child-before-parent ordering — so the AppHost registry refresh observes the stopped AppHost promptly instead of relying on VS Code cascading termination.

Changes:

• stopDebugging() now stops the AppHost child session first, then stops the parent in a finally block so the parent is stopped even if the AppHost stop request fails.
• The tracked resource stop callback (disposalFunction) now returns the vscode.debug.stopDebugging(session) promise (with stopSession() retyped to Thenable<void>) so callers can await the child stop, while cleanupRun still runs synchronously.
• Added two unit tests covering the new stop ordering and the parent-stop fallback when the AppHost stop fails.

Show a summary per file

File	Description
extension/src/debugger/AspireDebugSession.ts	Reorders stopDebugging() to stop AppHost before parent; returns the child stop promise from the disposal callback.
extension/src/dcp/types.ts	Changes AspireResourceDebugSession.stopSession() return type from void to Thenable<void> so the child stop can be awaited.
extension/src/test/aspireDebugSession.test.ts	Adds unit tests verifying stop ordering and parent-stop fallback on AppHost stop failure.

I verified the only non-test stopSession() call site (AspireDcpServer.ts DELETE /run_session/:id) remains a deliberate fire-and-forget call, and confirmed the extension's ESLint config does not enable no-floating-promises, so the return-type change introduces no new lint failures. The new unit tests are consistent with existing patterns in the file, and the ordering/fallback assertions are correct. I found no concrete issues to comment on.

Review details

• Files reviewed: 3/3 changed files
• Comments generated: 0
• Review effort level: Medium

[github-actions]

Retrying the failed CI jobs for this pull request from the CI run attempt. The rerun is being tracked in the rerun attempt.

[Copilot]

Review details

• Files reviewed: 3/3 changed files
• Comments generated: 0 new
• Review effort level: Medium