[test][NewPM] Make dead-uses.ll work under NPM

[aeubanks]

Summary

This one is weird...

globals-aa needs to be already computed at licm, or else a function pass
can't run a module analysis and won't have access to globals-aa.
But the globals-aa result is impacted by instcombine in a way that
affects what the test is expecting. If globals-aa is computed before
instcombine, it is cached and globals-aa used in licm won't contain the
necessary info provided by instcombine.
Another catch is that if we don't invalidate AAManager, it will use the
cached AAManager that instcombine requested, which may not contain
globals-aa. So we have to invalidate<aa> so that licm can recompute
an AAManager with the globals-aa created by the require<globals-aa>.

This is essentially the problem described in https://reviews.llvm.org/D84259.

[asbirlea]

This is fine if we really want to keep this test. I'm not sure we'll see this type of expected behavior in typical pipeline, but I guess this can be used as model for how to achieve something like this.

llvm/test/Analysis/GlobalsModRef/dead-uses.ll

-; RUN: opt < %s -instcombine -globals-aa -licm -S | FileCheck %s
+; RUN: opt < %s -instcombine -globals-aa -licm -enable-new-pm=0 -S | FileCheck %s
+; RUN: opt < %s -aa-pipeline=basic-aa,globals-aa -passes='function(instcombine),require<globals-aa>,function(invalidate<aa>,loop(licm))' -S | FileCheck %s
 
 ; Make sure -globals-aa ignores dead uses of globals.
 
 @a = internal global i32 0, align 4
 @c = common global i32 0, align 4
 
 ; Function Attrs: nounwind
 define i32 @g() {