#pragma nosimd

In IBM® XL C/C++ for AIX® 16.1.0 or earlier releases, #pragma nosimd disables the automatic generation of vector instructions. This pragma needs to be specified on a per-loop basis.

IBM Open XL C/C++ for AIX 17.1.4 accepts #pragma nosimd and maps it to the #pragma clang loop vectorize(disable) Clang pragma. If you used #pragma nosimd in your program, you are recommended to replace it with #pragma clang loop vectorize(disable) when you migrate the program to IBM Open XL C/C++ for AIX 17.1.4.

Related information

• -qsimd
• "Clang Compiler User's Manual" in the Clang documentation