[PATCH] [AArch64] Add v8.1a "Rounding Doubling Multiply Add/Subtract" extension

[James Molloy]

Hi Vladimir,

Thanks for doing that. It looks a lot better without the nasty selection logic, and we can just fix up vector instructions to scalar ones in the AdvSIMD pass if needed, as Tim suggested.

LGTM.

Cheers,

James


REPOSITORY
  rL LLVM

http://reviews.llvm.org/D8502

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