[release/6.0] The signal enum in the native library should match the managed code.

[github-actions]

Backport of #58658 to release/6.0

/cc @stephentoub @JamesWTruher

Customer Impact

Trying to register for SIGWINCH (window resize) will actually register for SIGCONT (making a suspend process continue), and vice versa.

Testing

CI (a new test was added separately/subsequently to validate our mappings).

Risk

Minimal. It's obvious from code inspection that the code was wrong and is now correct.

[stephentoub]

For reference, these values map to

runtime/src/libraries/System.Private.CoreLib/src/System/Runtime/InteropServices/PosixSignal.cs

Lines 28 to 33 in 721a802

	[UnsupportedOSPlatform("windows")]
	SIGCONT = -6,
	/// <summary>Window resized</summary>
	[UnsupportedOSPlatform("windows")]
	SIGWINCH = -7,

.

[stephentoub]

cc: @danmoseley

[danmoseley]

@stephentoub @tmds should we pull #58711 into this also?

[stephentoub]

@stephentoub @tmds should we pull #58711 into this also?

Either way is fine. I wouldn't block fixing the bug on getting that test pulled in as well.

[danmoseley]

would you mind cherry picking it in? it will only hold us up a couple hours.

[danmoseley]

Approved otherwise -- localized fix for self evidently wrong code.

[stephentoub]

would you mind cherry picking it in?

Done

[danmoseley]

@mdh1418 could you please port into 6.0 the fix for The PKCS#12 Exportable flag is not supported.. and any other deterministic failures in runtime-staging? We shouldn't have any known failures here.

@steveisok I'm not sure what to make of the test jobs that just terminated on iOSSimulator and tvOSSimulator. how can we make these diagnosable? I see no dump or relevant info in any of the logs -- is there any more info we can gather than:

[19:59:16] dbug: Test run completed
[19:59:16] dbug: Test run crashed before it started (no log file produced)
[19:59:16] dbug: No crash reports, waiting 30 seconds for the crash report service...
[19:59:47] fail: Application test run crashed
                 No test log file was produced

I'm going to merge this though as I can't see how it could have failures specific to just these.

[steveisok]

@danmoseley We have a few commits that @mdh1418 is working to backport that will clean these up. And to answer your question, there are about 7-8 suites that crash immediately for unknown reasons and seemingly only on CI. We will be skipping them for now until we're able to get a better understanding as to why.

[danmoseley]

Sounds good thanks @steveisok

[mdh1418]

#58841 Backport PR is open :) !