[llvm] [AMDGPU][NewPM] Port "GCNRewritePartialRegUses" pass to NPM (PR #126024)

[Akshat Oke via llvm-commits]

================
@@ -34,6 +34,7 @@
 #include "GCNIterativeScheduler.h"
 #include "GCNPreRALongBranchReg.h"
 #include "GCNPreRAOptimizations.h"
+#include "GCNRewritePartialRegUses.h"
----------------
optimisan wrote:

Oh yes

https://github.com/llvm/llvm-project/pull/126024