[PATCH] D43005: [ARM] Error out on .arm assembler directives on windows

[Peter Smith via Phabricator via llvm-commits]

peter.smith added a comment.

My initial thought is that there must be a better way than hard-coding a windows OS choice in the hasARM() field, particularly as all the others work entirely from the feature bits. Do you happen to know why clearing the feature bits is insufficient? I haven't touched this part of the code for a long time so I haven't got an answer for why myself. I can see that the assembler might recalculate them from the CPU description when particular directives are used.

I'd also say that there is a more generally useful feature here of "While my processor supports both Arm and Thumb state I choose to enforce that either Arm or Thumb state is used." this could be then enabled for Thumb2 by default for the Windows platform.

So overall I'd like to think that there was no possible alternative before doing this.


Repository:
  rL LLVM

https://reviews.llvm.org/D43005