Cleanup some XML tests

[danmoseley]

Fix #65235 ... hopefully.

The XML tests are very old, and thus messy. Making significant changes is a non goal and I'm not certain what caused this failure.

In separated commit order

• fixed cases where a MemoryStream would get disposed at an arbitrary point by a wrapping StreamWriter, causing ODE. Instead, deterministically dispose the wrapper, without closing the inner stream. Remove redundant flushes. I don't believe this is the cause of the test failure.
• cleaned up the elaborate static cache of MemoryStreams, since it was locking around accessing the cache, but generally handing out the same streams to all comers. Potentially could have caused the test failure seen, if two threads are meddling with the offset of the stream, and copying is happening during: the copy could end up with nulls. The complexity of these tests is such that I can't reason whether this could happen. Instead, always clone to a new stream before handing it out.
• fixed the test exception to not crash in some failure cases, especially as it happens on a threadpool thread
• fixed a test to test what it claims to test per its comment
• fixed a case where a test expected to receive the same stream from the cache that it inserted - broken by the 2nd change

[ghost]

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

Issue Details

---

Fix #65235 ... hopefully.

The XML tests are very old, and thus messy. Making significant changes is a non goal and I'm not certain what caused this failure.

In separated commit order

• fixed cases where a MemoryStream would get disposed at an arbitrary point by a wrapping StreamWriter, causing ODE. Instead, deterministically dispose the wrapper, without closing the inner stream. Remove redundant flushes. I don't believe this is the cause of the test failure.
• cleaned up the elaborate static cache of MemoryStreams, since it was locking around accessing the cache, but generally handing out the same streams to all comers. Potentially could have caused the test failure seen, if two threads are meddling with the offset of the stream, and copying is happening during: the copy could end up with nulls. The complexity of these tests is such that I can't reason whether this could happen. Instead, always clone to a new stream before handing it out.
• fixed the test exception to not crash in some failure cases, especially as it happens on a threadpool thread
• fixed a test to test what it claims to test per its comment
• fixed a case where a test expected to receive the same stream from the cache that it inserted - broken by the 2nd change

Author:	danmoseley
Assignees:	-
Labels:	area-System.Xml
Milestone:	-

[krwq]

nit: should we put space after :?

[danmoseley]

I actually don't know what our prevailing style is. I will have to check. @sharwell is there any StyleCop or editorconfig linting of this available?

[krwq]

Positive changes, looks like removing some potential culprits, just a single nit around style (foo:bar vs foo: bar), thanks for the PR!

[danmoseley]

@dotnet/dncenghot can you please help with these?

2022-02-14T03:12:09.1609457Z D:\a\_work\1\s\.packages\microsoft.dotnet.helix.sdk\7.0.0-beta.22110.7\tools\Microsoft.DotNet.Helix.Sdk.MultiQueue.targets(55,5): error : Helix encountered job-level error(s) for this job (Exception thrown while: sending work items to Service Bus).  Please contact dnceng with this information. [D:\a\_work\1\s\src\libraries\sendtohelixhelp.proj]
2022-02-14T03:12:09.1894357Z ##[error].packages\microsoft.dotnet.helix.sdk\7.0.0-beta.22110.7\tools\Microsoft.DotNet.Helix.Sdk.MultiQueue.targets(55,5): error : (NETCORE_ENGINEERING_TELEMETRY=Build) Helix encountered job-level error(s) for this job (Exception thrown while: sending work items to Service Bus).  Please contact dnceng with this information.

[MattGal]

@dotnet/ d n ce n ghot can you please help with these?

2022-02-14T03:12:09.1609457Z D:\a\_work\1\s\.packages\microsoft.dotnet.helix.sdk\7.0.0-beta.22110.7\tools\Microsoft.DotNet.Helix.Sdk.MultiQueue.targets(55,5): error : Helix encountered job-level error(s) for this job (Exception thrown while: sending work items to Service Bus).  Please contact dnceng with this information. [D:\a\_work\1\s\src\libraries\sendtohelixhelp.proj]
2022-02-14T03:12:09.1894357Z ##[error].packages\microsoft.dotnet.helix.sdk\7.0.0-beta.22110.7\tools\Microsoft.DotNet.Helix.Sdk.MultiQueue.targets(55,5): error : (NETCORE_ENGINEERING_TELEMETRY=Build) Helix encountered job-level error(s) for this job (Exception thrown while: sending work items to Service Bus).  Please contact dnceng with this information.

@lpatalas I think was looking earlier but I'll take a look too to make sure it's not ongoing. I'd expect this to be a burst of trouble with the Service Bus endpoint we were talking to and not a persistent issue since work has continued to be enqueued this whole time.

[MattGal]

@danmoseley this was an outage from Service bus clustered between 2/13/2022 6:12:08.241 PM -> 2/13/2022, 8:47:12.666 PM PST. It self resolved and was external to us, so hopefully just a blip.