[PATCH] D143886: [DAG] Handle build_vector with all undefs in reduceBuildVecTruncToBitCast

[Matt Arsenault via Phabricator via llvm-commits]

arsenm accepted this revision.
arsenm added inline comments.
This revision is now accepted and ready to land.


================
Comment at: llvm/test/CodeGen/AMDGPU/amdgcn.bitcast.ll:875
+  %8 = phi i64 [ 0, %entry ], [ %16, %if ]
+  %9 = phi i64 [ 0, %entry ], [ %17, %if ]
+  %10 = phi i64 [ 0, %entry ], [ %18, %if ]
----------------
Shouldn't use anonymous values in tests


================
Comment at: llvm/test/CodeGen/AMDGPU/amdgcn.bitcast.ll:914
+  %42 = call <16 x i16> @llvm.vector.insert.v16i16.v2i16(<16 x i16> %41, <2 x i16> %34, i64 12)
+  %43 = call <16 x i16> @llvm.vector.insert.v16i16.v2i16(<16 x i16> %42, <2 x i16> %35, i64 14)
+  store <16 x i16> %43, ptr addrspace(1) %out, align 32
----------------
Can this not reduce at all?


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D143886/new/

https://reviews.llvm.org/D143886