[PATCH] D38735: [ScheduleDAGInstrs] fix behavior of getUnderlyingObjectsForCodeGen when no identifiable object found

[Matthias Braun via Phabricator via llvm-commits]

MatzeB added a comment.

This looks good, but makes me wonder if we shouldn't do the same change (returning a bool instead of looking at the vector containing 0 or more elements) to `getUnderlyingObjectsForInstr()` for consistency. Could you try doing that?


https://reviews.llvm.org/D38735