[PATCH] D113107: Support of expression granularity for _Float16.

[Zahira Ammarguellat via Phabricator via cfe-commits]

zahiraam added inline comments.


================
Comment at: clang/test/CodeGen/X86/avx512fp16-complex.c:123
   // X86-NOT: fdiv
-  // X86: call {{.*}} @__divhc3(
+  // X86: call {{.*}} @__divsc3(
   // X86: ret
----------------
zahiraam wrote:
> pengfei wrote:
> > andrew.w.kaylor wrote:
> > > Why did this change happen?
> > This shouldn't be affect either.
> Still working on this.  
Still not sure what the reason but I think it is because of the coerce.
The IR generated for div_half_rc is :
%retval = alloca { half, half }, align 2
  %b = alloca { half, half }, align 2
  %a.addr = alloca half, align 2
  **%coerce = alloca { float, float }, align 4**
  %0 = bitcast { half, half }* %b to <2 x half>*
....
I would have expected to have %coerce = alloca { half, half }, align 2 
Still not sure what the issue is.


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D113107/new/

https://reviews.llvm.org/D113107