Reduce calli address spilling

[MichalPetryka]

Avoid spilling FTN_ADDR params, removes spilling of address on 0 arg default cconv callis.

Split off from #85197.

[ghost]

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

Issue Details

---

Avoid spilling FTN_ADDR params, removes spilling of address on 0 arg default cconv callis.

Split off from #85197.

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

[jakobbotsch]

/azp run jit-cfg

[azure-pipelines]

Azure Pipelines successfully started running 1 pipeline(s).

[MichalPetryka]

This seems to trigger huge diffs due to delegate constructors taking less of the inliner budget.

[AndyAyersMS]

Seems like the huge diffs are mostly in what are likely unrealistic test cases.

I'd like to see the ASP.NET diffs but unfortunately I can't recollect this right now as the machine I use is offline. So let's hold onto this change until I can get that collection going again.

[jakobbotsch]

/azp run runtime-coreclr jitstress, runtime-coreclr libraries-jitstress

[azure-pipelines]

Azure Pipelines successfully started running 2 pipeline(s).

[AndyAyersMS]

I have a new ASP.NET SPMI collection so am going to rerun that part of normal CI.

@jakobbotsch do you know if the stress failures here were mostly from promotion?

[AndyAyersMS]

Seems like since SPMI passed the first time the only rerun option is to bounce the PR.

@MichalPetryka why don't we merge this up to main and trigger a new round of CI that way? Would prefer you to do this but I can do it if you like.

[MichalPetryka]

do you know if the stress failures here were mostly from promotion?

I'd like to rerun stress after @jakobbotsch fixes the promotion issues with #85343.

[jakobbotsch]

/azp run runtime-coreclr jitstress, runtime-coreclr libraries-jitstress

[azure-pipelines]

Azure Pipelines successfully started running 2 pipeline(s).

[AndyAyersMS]

FYI you can generally just merge and not rebase -- that makes it a bit easier for us to track what is going on.

[jakobbotsch]

@jakobbotsch do you know if the stress failures here were mostly from promotion?

Looks like it, the remaining failures are #85449, #85426 and #48798 as far as I can tell.

[AndyAyersMS]

Seems like the various failures are known, so let's merge.

[AndyAyersMS]

@MichalPetryka thanks!