[PATCH] D39931: AMDGPU: Don't try to enable fp64 denormals on <SI

[Matt Arsenault via Phabricator via llvm-commits]

arsenm added a comment.

Fp64 denormals are always supported

In https://reviews.llvm.org/D39931#922846, @jvesely wrote:

> In https://reviews.llvm.org/D39931#922825, @arsenm wrote:
>
> > FP64 support without denormals doesn't make sense. They are mandatory in the OpenCL spec
>
>
> so what? this has clearly been disabled before, and it stays disabled after this change, just with less damage done in the process..
>
> enabling fp64 denormals can be considered when fp64 support is added to eg/ni targets.


The Evergreen manual lists double_denorm_flush_input and double_denorm_force_underflow_to_zero modes, so that implies they are supported. I don't care that much because the f64 support is missing, but I'm not sure what problem you are trying to solve here



================
Comment at: lib/Target/AMDGPU/AMDGPUSubtarget.cpp:62
+  } else {
+    FullFS += "-fp64-fp16-denormals,-fp32-denormals,";
+  }
----------------
You shouldn't explicitly disable them either if it's not a recognized feature for r600


Repository:
  rL LLVM

https://reviews.llvm.org/D39931