[PATCH] D51925: [AMDGPU] Fix issue for zext of f16 to i32

[David Stuttard via Phabricator via llvm-commits]

dstuttard added inline comments.


================
Comment at: lib/Target/AMDGPU/SIInstructions.td:1353
+def : GCNPat <
+  (i32 (AMDGPUfp16_zext f16:$src)),
+  (V_AND_B32_e64 $src, (V_MOV_B32_e32 (i32 0x0000ffff)))
----------------
arsenm wrote:
> IIRC this node is only supposed to be emitted if the high bits are known zero, so something is wrong upstream if it’s gotten here 
Ok - thanks. I'll take another look.


Repository:
  rL LLVM

https://reviews.llvm.org/D51925