[PATCH] D109243: [GlobalISel] Build_vector_trunc artifact combine into bitcast

[Petar Avramovic via Phabricator via llvm-commits]

Petar.Avramovic added a comment.

Using merge/unmerge for more elements together with build_vector_combines in this series of patches gives optimal code for add <N x s16> with various vector sizes as far as I can tell (look at tests in add.vNi16.ll.mir).


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D109243/new/

https://reviews.llvm.org/D109243