[PATCH] D69275: Add constrained int->FP intrinsics

[Craig Topper via Phabricator via llvm-commits]

craig.topper added inline comments.


================
Comment at: llvm/lib/Target/X86/X86ISelLowering.cpp:18766
+                                  Op.getSimpleValueType())
+        .first;
+  }
----------------
kpn wrote:
> craig.topper wrote:
> > Don't we need to propagate the chain result too?
> When given a strict FP node LowerOperation() will require that the resulting new node has the chain in the same place in the values. So, no, I don't think so unless we want to thread the std::pair all the way back up.
Oh this works because the node is either a STRICT_FP_EXTEND or STRICT_FP_ROUND so the node in first and second is the same. But we should probably use a MERGE_VALUES here to not rely on that.


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D69275/new/

https://reviews.llvm.org/D69275