[release/5.0] accept empty realm for digest auth (#56369)

[github-actions]

Backport of #56455 to release/5.0

Fixes #56369

/cc @karelz @camillo-toselli

Customer Impact

This breaks web app managing telephone system of the University of Bologna (10K phone lines) - see details in #56369 (comment)
Upgrade to .NET 6.0 (where the issue is fixed) is blocked by missing Oracle EF Core 6 provider.

Note: It is similar to backport of similar issue about empty domain for digest auth - see #50598 which brought that one into 5.0.6 release.

Testing

New targeted test case was added.
Customer validated on Windows10/x64 (on private build locally built).

Risk

Small, because:

• It is unlikely anyone depends on empty realm to report error.
• We serviced 5.0 with similar problem with empty domain in [release/5.0] ignore empty domain for digest auth #50598 (we build on that PR).
• Change is simple and straightforward, so risk of unexpected side effects is small.

[ghost]

Tagging subscribers to this area: @dotnet/ncl
See info in area-owners.md if you want to be subscribed.

Issue Details

---

Backport of #56455 to release/5.0

/cc @karelz @camillo-toselli

Customer Impact

Testing

Risk

Author:	github-actions[bot]
Assignees:	-
Labels:	area-System.Net
Milestone:	-

[karelz]

@danmoseley can you please check the request? Any concerns?

[karelz]

Test failures are unrelated (re-running them):

• System.Diagnostics.Tests.EventLogRecordTests.ExceptionOnce - https://helixre8s23ayyeko0k025g8.blob.core.windows.net/dotnet-runtime-refs-pull-61203-merge-e8c7db19d6d54a0691/System.Diagnostics.EventLog.Tests/1/console.eac6a880.log?sv=2019-07-07&se=2021-11-24T11%3A41%3A20Z&sr=c&sp=rl&sig=jgQGnBEA%2BnvRgxM%2BIAghQ0RSG7TPoZ44zDz7eCEcgIA%3D

    System.Diagnostics.Tests.EventLogRecordTests.ExceptionOnce [FAIL]
      System.Diagnostics.Eventing.Reader.EventLogProviderDisabledException : The publisher has been disabled and its resource is not available. This usually occurs when the publisher is in the process of being uninstalled or upgraded.
      Stack Trace:
        /_/src/libraries/System.Diagnostics.EventLog/src/System/Diagnostics/Reader/EventLogException.cs(36,0): at System.Diagnostics.Eventing.Reader.EventLogException.Throw(Int32 errorCode)
        /_/src/libraries/System.Diagnostics.EventLog/src/System/Diagnostics/Reader/NativeWrapper.cs(124,0): at System.Diagnostics.Eventing.Reader.NativeWrapper.EvtOpenProviderMetadata(EventLogHandle session, String ProviderId, String logFilePath, Int32 locale, Int32 flags)
        /_/src/libraries/System.Diagnostics.EventLog/src/System/Diagnostics/Reader/ProviderMetadata.cs(69,0): at System.Diagnostics.Eventing.Reader.ProviderMetadata..ctor(String providerName, EventLogSession session, CultureInfo targetCultureInfo, String logFilePath)
        /_/src/libraries/System.Diagnostics.EventLog/src/System/Diagnostics/Reader/ProviderMetadataCachedInformation.cs(139,0): at System.Diagnostics.Eventing.Reader.ProviderMetadataCachedInformation.GetProviderMetadata(ProviderMetadataId key)
        /_/src/libraries/System.Diagnostics.EventLog/src/System/Diagnostics/Reader/ProviderMetadataCachedInformation.cs(218,0): at System.Diagnostics.Eventing.Reader.ProviderMetadataCachedInformation.GetLevelDisplayName(String ProviderName, EventLogHandle eventHandle)
        /_/src/libraries/System.Diagnostics.EventLog/src/System/Diagnostics/Reader/EventLogRecord.cs(319,0): at System.Diagnostics.Eventing.Reader.EventLogRecord.get_LevelDisplayName()
        /_/src/libraries/System.Diagnostics.EventLog/tests/System/Diagnostics/Reader/EventLogRecordTests.cs(102,0): at System.Diagnostics.Tests.EventLogRecordTests.<>c__DisplayClass4_1.<ExceptionOnce>b__0()
        /_/src/libraries/System.Diagnostics.EventLog/tests/System/Diagnostics/Reader/EventLogRecordTests.cs(116,0): at System.Diagnostics.Tests.EventLogRecordTests.ThrowsMaxOnce[T](Action action)
        /_/src/libraries/System.Diagnostics.EventLog/tests/System/Diagnostics/Reader/EventLogRecordTests.cs(102,0): at System.Diagnostics.Tests.EventLogRecordTests.ExceptionOnce()

• JIT arm64 test run didn't finish execution (only 1 line in log) - https://helixre8s23ayyeko0k025g8.blob.core.windows.net/dotnet-runtime-refs-pull-61203-merge-8ac834ff904e4ab396/JIT.Regression.CLR-x86-JIT.V1-M12-M13/1/console.99c21a76.log?sv=2019-07-07&se=2021-11-24T13%3A10%3A17Z&sr=c&sp=rl&sig=KwjQCY%2BqS2RbCH77WhS5aLtwMawDfvMrBr9i62LNY5U%3D

Console log: 'JIT.Regression.CLR-x86-JIT.V1-M12-M13' from job 8ac834ff-904e-4ab3-96b9-a96d6f30e073 workitem 8ff53c12-fb10-41d0-9568-e320d2ce8f35 (windows.10.arm64v8.open) executed on machine DDARM64-179

[danmoseley]

Approved once customer verifies fix.

Curious why this didn't show up when customer used #50598 fix?

[wfurt]

Curious why this didn't show up when customer used #50598 fix?

It was different reporter so perhaps different server with different behavior @danmoseley. When the original fix was done, we really just fix the reported issue. AFAIK we did not do any extensive search through RFC for similar cases.

[karelz]

(marking it NO MERGE to block on customer validation)

[karelz]

Rerun test failures are again unrelated:

• JIT test run on arm64 didn't finish execution (only 1 line in log) -- https://helix.dot.net/api/2019-06-17/jobs/815d9fa9-2c56-435c-8400-8b4dfb82a69f/workitems/JIT.Methodical.a-dA-D/console

Console log: 'JIT.Methodical.a-dA-D' from job 815d9fa9-2c56-435c-8400-8b4dfb82a69f workitem e007438e-7141-4761-ba97-987e1412c1ad (windows.10.arm64v8.open) executed on machine DDARM64-179

It is the same machine DDARM64-179 as above in #61203 (comment) - @dotnet/dncenghot is that a known machine problem perhaps?

[karelz]

Removing NO MERGE as customer validated the bits, see #56369 (comment)

[lpatalas]

It is the same machine DDARM64-179 as above in #61203 (comment) - @dotnet/dncenghot is that a known machine problem perhaps?

I'll take a look at that machine.

[lpatalas]

I have put that machine offline and created an issue to track further work: https://github.com/dotnet/core-eng/issues/14861

[wfurt]

LGTM

[karelz]

The remaining CI failure is unrelated -- Build Analysis warns that image of Ubuntu 16.04 is not available for "runtime-libraries enterprise-linux":

runtime-libraries enterprise-linux 20211104.1 / Job
❌ An image label with the label ubuntu-16.04 does not exist.

It is not related to this PR and would not help if it worked.
@wfurt is it something we should look into?

@danmoseley the PR is ready for merge now.

[MattGal]

The remaining CI failure is unrelated -- Build Analysis warns that image of Ubuntu 16.04 is not available for "runtime-libraries enterprise-linux":

runtime-libraries enterprise-linux 20211104.1 / Job
❌ An image label with the label ubuntu-16.04 does not exist.

It is not related to this PR and would not help if it worked. @wfurt is it something we should look into?

@danmoseley the PR is ready for merge now.

This is definitely worth fixing, since that error means an entire pipeline didn't run.

Images are listed here: https://github.com/actions/virtual-environments and Ubuntu 16.04 was removed because it is past its end-of-life support window.

Simply changing ubuntu-16.04 to ubuntu-18.04, ubuntu-20.04, or ubuntu-latest will bring the pipeline back online.

[danmoseley]

@Anipik please see Matt's note above. Seems like we need a yaml tweak in this branch