[PATCH] D144163: [X86] Support load/store for bf16 vector in avx

[LuoYuanke via Phabricator via llvm-commits]

LuoYuanke added inline comments.


================
Comment at: llvm/test/CodeGen/X86/avx512bf16-mov.ll:36
+  store <8 x bfloat> %0, ptr %dst, align 1
+  %1 = load <8 x bfloat>, ptr %src, align 32
+  store <8 x bfloat> %1, ptr %dst, align 32
----------------
align 16 ?


================
Comment at: llvm/test/CodeGen/X86/avx512bf16-mov.ll:37
+  %1 = load <8 x bfloat>, ptr %src, align 32
+  store <8 x bfloat> %1, ptr %dst, align 32
+  %2 = load <16 x bfloat>, ptr %src, align 1
----------------
ditto.


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D144163/new/

https://reviews.llvm.org/D144163