In the Clang language extensions documentation, there’s a heading for a builtin function called __builtin_shufflevector but no description. Have this feature been implemented and could someone help me with a description?

Thank you,

Van To

Yes, it's implemented; it's roughly the equivalent of the LLVM
shufflevector instruction. The first two arguments are vectors, then
the remaining arguments are the indices used for the returned vector.


thank you for the information


I just checked in some documentation for this, here: