[mobile] Skip DnsGetHostEntry_LocalHost test on Android

[github-actions]

Note

This content was generated by GitHub Copilot.

Summary

Fixes System.Net.NameResolution.Functional.Tests.GetHostEntryTest.DnsGetHostEntry_LocalHost_ReturnsFqdnAndLoopbackIPs test failure on Android by skipping the test on that platform.

Background

The test assumes that resolving "localhost" returns only loopback IP addresses (127.0.0.1 and ::1). However, on Android Helix machines (emulators), the DNS resolution for "localhost" returns actual network interfaces in addition to loopback addresses, causing the test to fail with assertions like:

Assert.All() Failure: 5 out of 6 items in the collection did not pass.
[1]: Item:  fe80::e05a:f2ff:fee9:6ae6%47
     Error: Not a loopback address: fe80::e05a:f2ff:fee9:6ae6%47
[2]: Item:  2001:4898:502:2:e05a:f2ff:fee9:6ae6
     Error: Not a loopback address: 2001:4898:502:2:e05a:f2ff:fee9:6ae6
[5]: Item:  10.64.32.211
     Error: Not a loopback address: 10.64.32.211
````

## Changes

Added `[SkipOnPlatform(TestPlatforms.Android, ...)]` attribute to the failing test, consistent with existing skips for iOS, tvOS, and MacCatalyst on the same test (which have similar issues).

## Failure Details

**Build:** https://dev.azure.com/dnceng-public/public/_build/results?buildId=1396313

**Job:** `android-arm64 Release AllSubsets_Mono`

**Helix Work Item:** System.Net.NameResolution.Functional.Tests
- Job ID: `93091357-214b-436d-a5c4-a31ee456a876`
- Console log: https://helix.dot.net/api/2019-06-17/jobs/93091357-214b-436d-a5c4-a31ee456a876/workitems/System.Net.NameResolution.Functional.Tests/console

**Log excerpt (sanitized):**
````
04-25 00:03:35.029  DOTNET  : [FAIL] System.Net.NameResolution.Tests.GetHostEntryTest.DnsGetHostEntry_LocalHost_ReturnsFqdnAndLoopbackIPs(mode: 0)
04-25 00:03:35.029  DOTNET  : Assert.All() Failure: 5 out of 6 items in the collection did not pass.
04-25 00:03:35.029  DOTNET  : [5]: Item:  10.64.32.211
04-25 00:03:35.029  DOTNET  :      Error: Not a loopback address: 10.64.32.211

Related

• Existing skip on iOS/tvOS/MacCatalyst: TcpReceiveSendGetsCanceledByDispose test timeouts on Apple mobile and Android #124079
• Similar Wasi skip: [WASI] failing System.Net.NameResolution tests #107339

Note

🔒 Integrity filter blocked 1 item

The following item were blocked because they don't meet the GitHub integrity level.

• #126805 search_pull_requests: has lower integrity than agent requires. The agent cannot read data with integrity below "approved".

To allow these resources, lower min-integrity in your GitHub frontmatter:

tools:
  github:
    min-integrity: approved  # merged | approved | unapproved | none

Generated by Mobile Platform Failure Scanner · ● 4.1M · ◷

[dotnet-policy-service]

Tagging subscribers to 'arch-android': @vitek-karas, @simonrozsival, @steveisok, @akoeplinger
See info in area-owners.md if you want to be subscribed.

[github-actions]

🤖 Copilot Code Review — PR #127420

Note

This review was generated by Copilot.

Holistic Assessment

Motivation: The test DnsGetHostEntry_LocalHost_ReturnsFqdnAndLoopbackIPs asserts that localhost DNS resolution returns only loopback addresses, but Android Helix machines return non-loopback addresses. Skipping this test on Android is a reasonable response to a platform behavior difference that causes CI failures.

Approach: The change adds a single [SkipOnPlatform] attribute. The skip reason is descriptive and the attribute is correctly placed alongside the existing platform-specific attributes.

Summary: ⚠️ Needs Human Review. The code change itself is correct, but two points warrant human judgment: (1) the choice of [SkipOnPlatform] vs [ActiveIssue], and (2) whether sibling tests asserting IsLoopback for localhost resolution also need the same skip on Android.

---

Detailed Findings

⚠️ Convention — [SkipOnPlatform] vs [ActiveIssue] for this failure

The existing convention in this file for Android-specific test failures is [ActiveIssue] with a tracking GitHub issue:

• Line 418: [ActiveIssue("https://github.com/dotnet/runtime/issues/124751", TestPlatforms.Android)]
• Line 453: [ActiveIssue("https://github.com/dotnet/runtime/issues/126456", ... | TestPlatforms.Android)]

The same test already uses [ActiveIssue] for iOS/tvOS/MacCatalyst (line 263) for what appears to be the same category of platform-specific localhost resolution behavior.

[SkipOnPlatform] is used in this file for permanent platform limitations (e.g., "WASI has no getnameinfo") — cases where the platform fundamentally cannot support the tested functionality. An [ActiveIssue] with a tracking issue would be more consistent here unless this is considered a permanent, unfixable Android behavior rather than a Helix environment issue or a resolvable platform quirk. Without a tracking issue, this failure becomes invisible to future triage.

Suggestion: Consider using [ActiveIssue("https://github.com/dotnet/runtime/issues/<new_issue>", TestPlatforms.Android)] with a filed tracking issue, consistent with how similar platform-specific failures are tracked in this file. If [SkipOnPlatform] is intentional (i.e., Android is expected to return non-loopback for localhost and this will never change), then a brief comment in the PR description explaining that rationale would be helpful.

⚠️ Cross-cutting — Sibling tests that also assert IsLoopback for localhost

Several other tests in this file and in GetHostAddressesTest.cs assert that localhost-related resolution returns only loopback addresses:

• DnsGetHostEntry_LocalhostSubdomain_ReturnsLoopback (line 357) — resolves foo.localhost etc., falls back to localhost
• DnsGetHostEntry_LocalhostWithTrailingDot_ReturnsLoopback (line 402) — resolves localhost.
• DnsGetHostAddresses_LocalhostSubdomain_ReturnsLoopback (line 234 in GetHostAddressesTest.cs) — same pattern

If Android DNS resolution returns non-loopback addresses for plain localhost, these related tests may also be affected (especially the trailing-dot variant, which is equivalent to plain localhost). If they are already passing on Android, no action is needed — but it's worth verifying.

✅ Attribute placement and skip message are correct

The [SkipOnPlatform] is placed correctly in the attribute list and the skip message clearly explains the reason for skipping.

Generated by Code Review for issue #127420 · ◷

[kotlarmilos]

@copilot extend [ActiveIssue("https://github.com/dotnet/runtime/issues/124079", TestPlatforms.iOS | TestPlatforms.tvOS | TestPlatforms.MacCatalyst)] instead of SkipOnPlatform.

[Copilot]

@copilot extend [ActiveIssue("https://github.com/dotnet/runtime/issues/124079", TestPlatforms.iOS | TestPlatforms.tvOS | TestPlatforms.MacCatalyst)] instead of SkipOnPlatform.

Done in 88e4d2f. Replaced [SkipOnPlatform(TestPlatforms.Android, ...)] with extending the existing [ActiveIssue] attribute to include TestPlatforms.Android.

[Copilot]

Pull request overview

Updates System.Net.NameResolution functional tests to avoid Android-specific failures in DnsGetHostEntry_LocalHost_ReturnsFqdnAndLoopbackIPs by extending an existing platform-conditional suppression to include Android.

Changes:

• Extend the existing [ActiveIssue(...)] platform mask for DnsGetHostEntry_LocalHost_ReturnsFqdnAndLoopbackIPs to include TestPlatforms.Android.

Show a summary per file

File	Description
src/libraries/System.Net.NameResolution/tests/FunctionalTests/GetHostEntryTest.cs	Adds Android to the platform list for an ActiveIssue-based skip on a localhost/loopback assertion test.

Copilot's findings

• Files reviewed: 1/1 changed files
• Comments generated: 1