JitStress Fixes

[CarolEidt]

Initialize method info early so that jitStress can use that information to determine when to enable stress.

Fix #8889

[CarolEidt]

@dotnet/dnceng - AFAICT the 'Performance Linux x64 release mono JIT net5.0' leg completed without errors, but reported failing.

[CarolEidt]

@dotnet/jit-contrib PTAL
I've done some sanity checks that I can now use COMPlus_JitStressOnly and see stress only applied to the given method.
If this looks like a reasonable approach, I will run stress legs to validate that we see the same failures as in recent runs.

[AndyAyersMS]

Seems reasonable.

Note that when trying to "simplify" a stress failure down to one method you may run into the issues described in #1983 (comment), in particular changing inlining by toggling stress can then change the methods that get jitted.

[MattGal]

@dotnet/dnceng - AFAICT the 'Performance Linux x64 release mono JIT net5.0' leg completed without errors, but reported failing.

fascinating, I'm taking a look

[MattGal]

@CarolEidt I think we can resolve this one as "the UI is funky"

https://dev.azure.com/dnceng/public/_build/results?buildId=694463&view=logs&jobId=26ef8efa-a13f-5927-0bd6-f52d59ca1b8a&j=26ef8efa-a13f-5927-0bd6-f52d59ca1b8a&t=327979ff-79d9-5fe2-303c-2056212ed821
was the failing build.

And the failure is a hang timeout where shutting down msbuild stuff makes the entry script not exit. There's a vigorous conversation going on about this in Teams I'll @ you on.
https://helix.dot.net/api/2019-06-17/jobs/3d8ace93-5826-4540-9c74-9002da994549/workitems/x64.micro_mono.net5.0.Partition24/console

[CarolEidt]

The stress failures are the same as in master.