ClangIR emits functions in the opposite order as OG sometimes

[smeenai]

https://godbolt.org/z/zxqshqdcv shows an example. This seems to be intentional on CIR's part, per:

clangir/clang/lib/CIR/CodeGen/CIRGenModule.cpp

Lines 2396 to 2402 in c76c138

	// Some global emissions are triggered while emitting a function, e.g.
	// void s() { x.method() }
	//
	// Be sure to insert a new function before a current one.
	auto *curCGF = getCurrCIRGenFun();
	if (curCGF)
	builder.setInsertionPoint(curCGF->CurFn);

On the other hand, LLVM always adds a new function to the end of the module's function list:

clangir/llvm/lib/IR/Function.cpp

Line 506 in c76c138

ParentModule->getFunctionList().push_back(this);

It's not a huge deal, but it'll make sharing tests with OG more complicated, and I'm wondering if we can match OG instead.