[PATCH] D129999: [TableGen] Mark incomplete MachineInstr as not compressible

[Luís Marques via Phabricator via llvm-commits]

luismarques added a comment.

This seems to make sense but can you please provide more information about the scenario where this is called and the instruction still has a FrameIndex? I understand you want to run this earlier but more information might be useful.

Being conservative about the size seems correct (e.g. for the branch relaxation pass we want to be conservative), but if anyone has any info to the contrary please indicate.


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D129999/new/

https://reviews.llvm.org/D129999