[PATCH] D74619: [ARM] Enabling range checks on Neon intrinsics' lane arguments

Leonard Chan via Phabricator via cfe-commits cfe-commits at lists.llvm.org
Fri Mar 20 16:18:08 PDT 2020


leonardchan added a comment.

Actually this may not need a reproducer since the error seems straightforward with just calling a builtin function. The main issue is that before this patch, something like

  #include <arm_neon.h>
  float64x2_t func(int8x8_t x) {
    return __builtin_neon_splatq_lane_v(x, 1, 10);
  }

would've compiled, but we get the error with this patch. I see that the intent of this patch is to do range checking, but does it make sense for the range shown in the error to be `[0, 0]`?


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D74619/new/

https://reviews.llvm.org/D74619





More information about the cfe-commits mailing list