Fix totalILArgs for explicithis

[MichalPetryka]

Fixes totalILArgs counting this twice with explicithis.

Split off from #85197.

[AndyAyersMS]

What about:

runtime/src/coreclr/tools/Common/JitInterface/CorInfoTypes.cs

Lines 117 to 119 in 5191a75

	private bool hasThis() { return ((callConv & CorInfoCallConv.CORINFO_CALLCONV_HASTHIS) != 0); }
	private bool hasExplicitThis() { return ((callConv & CorInfoCallConv.CORINFO_CALLCONV_EXPLICITTHIS) != 0); }
	private uint totalILArgs() { return (uint)(numArgs + (hasThis() ? 1 : 0)); }

I suspect there are a few more places that will need changes, would suggest you review the places where totalILArgs is used and make sure they're ok. It may be hard to surface these during testing as explicit this signatures are restricted to cases we don't optimize (yet).

[AndyAyersMS]

@markples perhaps you can review this one?

[markples]

It looks like hasthis is also a thing to audit, though that would uncover separate bugs and doesn't need to be part of this PR. Unfortunately, there are a lot of them. In general, it seems to be documented that explicit-call is only for signatures, not methods themselves (assuming that this isn't one of those things were the spec isn't followed), so places that check hascall and ignore explicitcall might be fine anyway.

[markples]

My hasthis comment doesn't seem like it should block anything here. Thanks!

[AndyAyersMS]

@MichalPetryka thanks -- and thanks for splitting this off from your other PR.