JIT: Switch tailcall implicit byref optimization to use last use information

[jakobbotsch]

The last use information is more general in most cases (for example, the
old logic cannot allow the optimization when there are previous field
writes to the implicit byref). In addition the old logic can interact
with the new last-use copy elision optimization in an incorrect way.

Fix#81019

[ghost]

Tagging subscribers to this area: @JulieLeeMSFT, @jakobbotsch, @kunalspathak
See info in area-owners.md if you want to be subscribed.

Issue Details

---

The last use information is more general in most cases (for example, the
old logic cannot allow the optimization when there are previous field
writes to the implicit byref). In addition the old logic can interact
with the new last-use copy elision optimization in an incorrect way.

Fix #81019

Author:	jakobbotsch
Assignees:	jakobbotsch
Labels:	area-CodeGen-coreclr
Milestone:	-

[jakobbotsch]

SuperPMI jobs seemingly did not trigger..?

[jakobbotsch]

This is blocked on #81083 for now.

[jakobbotsch]

cc @dotnet/jit-contrib PTAL @AndyAyersMS

Diffs. The new misses are tailcall candidates (that hit a new updateEntryForTailcall) -- these would presumably add some mostly positive mix of diffs.

The regressions are in similar situations as the case inside #81019 -- i.e. the old logic prefers omitting the copy at some non-last use because it believes a subsequent use won't alias. That allows tailcalling, so it is usually better, but the case is quite rare as can be seen in the diffs.

[jakobbotsch]

Ping @AndyAyersMS

[AndyAyersMS]

LGTM.

Sorry to see all that fabulous alias analysis get deleted, but this approach is better and more obviously correct.

[jakobbotsch]

Sorry to see all that fabulous alias analysis get deleted

Yes, I felt both good and bad removing that code... :-)