[clang] [Sema] When checking for constraint equivalence, do not calculate satisfaction (PR #74490)

[Erich Keane via cfe-commits]

================
@@ -4077,6 +4084,7 @@ Sema::SubstExpr(Expr *E, const MultiLevelTemplateArgumentList &TemplateArgs) {
   TemplateInstantiator Instantiator(*this, TemplateArgs,
                                     SourceLocation(),
                                     DeclarationName());
+  Instantiator.setEvaluateConstraints(false);
----------------
erichkeane wrote:

Oh, absolutely!  I'm reassured somewhat by  that fact of course, I was as comprehensive as I could be at the time.  BUT when doing this the first time, I ended up with about a dozen reverts trying to work through it, so there is definitely possibility that I missed some.  I'm reasonably confident/ok with this patch, just not until the release branch is taken.

https://github.com/llvm/llvm-project/pull/74490