PGO considerations for finally cloning

[AndyAyersMS]

Current finally cloning algorithm is roughly:

   foreach try-finally in the method
       if the finally is not too big, not rare, and is not must throw
           find the "normal" exit path from the try-finally (lexically last exit)
           move the normal exit path to be the fall through path if necessary
           clone the finally and reroute the normal exit and any other exit that can share the normal path to the clone

With PGO data, we should reconsider:

• We might want to adjust the size threshold based on relative profile counts
• We should pick the hottest exit path, not the "normal" path (JIT: profile updates for finally optimizations #49139)
• We might consider multiple clones
• We should rescale the profile counts in the clone to reflect how much of the finally execution now flows there (JIT: profile updates for finally optimizations #49139)

At a minimum, we should do the rescaling.

category:cq
theme:profile-feedback

[dotnet-issue-labeler]

I couldn't figure out the best area label to add to this issue. If you have write-permissions please help me learn by adding exactly one area label.

[AndyAyersMS]

See also #48850 (comment)

[AndyAyersMS]

Will defer the rest for now.