On 2012-04-24 22:13:11 +1200, Michael Kerrisk (man-pages) wrote: > Vincent, > > On Thu, Jan 6, 2011 at 5:38 AM, Vincent Lefevre <vinc...@vinc17.net> wrote: > > The isgreater(3) man page says: > > > > The normal relation operations (like <, "less than") will fail if one > > of the operands is NaN. This will cause an exception. To avoid this, > > C99 defines these macros. The macros are guaranteed to evaluate their > > operands only once. The operands can be of any real floating-point > > type. > > > > It should probably be emphasised that the operands must not be of > > integer types. > > The man page already says: "The operands can be of any real > floating-point type." Is that not emphasis enough?
This is a bit ambiguous, as "can" is not necessarily exclusive. Macros often have a function-like interface: for instance, if a real floating-point type is expected and the user provides an integer, then the integer will automatically be converted to a real floating-point type. This is not the case here. Also, the beginning of the paragraph is a bit like saying that "isgreater" is a safe replacement for ">" to avoid a possible exception. However, while with ">" the user doesn't need to take care of the types, this is not the case for "isgreater". For instance, in order to avoid a possible exception, one should do the following replacements: With double x int i; replace x > i x > 0 by isgreater (x, (double) i) isgreater (x, 0.0) Users may easily forget to do the change on the second argument. -- Vincent Lefèvre <vinc...@vinc17.net> - Web: <http://www.vinc17.net/> 100% accessible validated (X)HTML - Blog: <http://www.vinc17.net/blog/> Work: CR INRIA - computer arithmetic / AriC project (LIP, ENS-Lyon) -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
