[PATCH] Fix TODO above x86 FeatureCallRegIndirect

[David Zarzycki via llvm-commits]

The attached patch fixes the TODO above FeatureCallRegIndirect. The underlying problem is that two memory operand instructions are slow on some processors, namely Atom and Atom derived processors like Knights Landing.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20170501/341c70e5/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: SlowTwoMemOps.diff
Type: application/octet-stream
Size: 6175 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20170501/341c70e5/attachment.obj>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20170501/341c70e5/attachment-0001.html>