StringFromSingleCharCode is a better name because the intrinsic it represents only applies when we are converting from a single char code.
Created attachment 268548 [details] proposed patch.
Comment on attachment 268548 [details] proposed patch. Mark Lam told me on IRC that the StringFromCharCode logic is called when the DFG detects that "String.fromCharCode is only passed a single value". For completeness the JavaScript API String.fromCharCode() takes more than one char code.
(In reply to comment #2) > Comment on attachment 268548 [details] > proposed patch. > > Mark Lam told me on IRC that the StringFromCharCode logic is called when the > DFG detects that "String.fromCharCode is only passed a single value". For > completeness the JavaScript API String.fromCharCode() takes more than one > char code. To clarify, the StringFromCharCode logic here refers to a DFG intrinsic implementation of String.fromCharCode(). See DFGByteCodeParser.cpp for where it will only emit this intrinsic node if the number of arguments is 2 (this + the single char code).
Thanks for the review. Landed in r194767: <http://trac.webkit.org/r194767>.
I think that it should take more than a non-JSC reviewer to rename parts of DFG IR.
Fil has told me offline that it is common practice to name DFG intrinsic nodes based on the functions they replace, even if the intrinsic does not fully implement the most generic form of said functions. For example, ArithMax, ArithMin, and NeewArray. I will roll out the patch.
The patch has been rolled out in r194770: <http://trac.webkit.org/r194770>.