efriedma added a comment. I think we should assume libm complies with the C99 standard (specifically, the appendix specifying the behavior of IEEE floating-point). Otherwise, we have no basis to predict the behavior of any standard library call.
From the standard: "Functions with a NaN argument return a NaN result and raise no floating-point exception, except where stated otherwise." https://reviews.llvm.org/D39481 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits