[PATCH] [CodeGen] Combine small-element shuffles of scalar_to_vector in terms of the wider scalar.

[Simon Pilgrim]

There is already an optimization that attempts to shuffle 'scalar source' input operands into a single BUILD_VECTOR just above this code. Might it make more sense to add the ability to peek through bitcasts there?


http://reviews.llvm.org/D8885

EMAIL PREFERENCES
  http://reviews.llvm.org/settings/panel/emailpreferences/