[PATCH] D63452: AMDGPU: Support some GDS atomics

[Nicolai Hähnle via Phabricator via llvm-commits]

nhaehnle added inline comments.


================
Comment at: lib/Target/AMDGPU/SIISelLowering.cpp:5750-5751
+
+    if (Subtarget->getGeneration() >= AMDGPUSubtarget::GFX10)
+      Offset1 |= (CountDw - 1) << 6;
+
----------------
arsenm wrote:
> This is a bit magical
Yes, but it's just hardware encoding. In any case, I'm pulling it into a separate patch since it's a logically separate change.


Repository:
  rL LLVM

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D63452/new/

https://reviews.llvm.org/D63452