2 ms·
That was just a clang bug, and as I read the comments they fixed it quietly without exactly admitting as much. That kind of local bitsmithery on floating point
by ajross 8y ago
That was just a clang bug, and as I read the comments they fixed it quietly without exactly admitting as much. That kind of local bitsmithery on floating point values is a pervasive idiom, clearly explained by the syntax of the language, and the compiler had no business getting smart with it.