WidenVectorOperand() bug?

[Jonas Paulsson via llvm-commits]

> This is a little confusing, but I wouldn't call it a bug.  We 
> consistently call CustomLowerNode with the illegal type, whether it's 
> the result or an operand, and existing targets depend on that.

Aha - that's a bit confusing, but thanks for explaining! I think a 
comment on this would be useful -- see my attached proposal for it.

/Jonas

-------------- next part --------------
A non-text attachment was scrubbed...
Name: setOperationAction_comment.patch
Type: text/x-patch
Size: 745 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20170119/a81b83c2/attachment.bin>