fmax.3: Add caveat for going beyond C std requirements

libm's fmax and fmin family of functions treat +0.0 as greater than
-0.0.  This is not required by the C standard, so the user may not see
this behaviour due to compiler optimization.

PR:		294214
Reviewed by:	fuz
Sponsored by:	The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D56230

1 files changed, 8 insertions, 5 deletions

diff --git a/lib/msun/man/fmax.3 b/lib/msun/man/fmax.3
index 25fc9b6d518a..873a19375ce5 100644
--- a/lib/msun/man/fmax.3
+++ b/lib/msun/man/fmax.3
@@ -22,7 +22,7 @@
 .\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
 .\" SUCH DAMAGE.
 .\"
-.Dd June 29, 2004
+.Dd April 6, 2026
 .Dt FMAX 3
 .Os
 .Sh NAME
@@ -68,10 +68,6 @@ functions return the smaller of
 .Fa x
 and
 .Fa y .
-They treat
-.Li +0.0
-as being larger than
-.Li -0.0 .
 If one argument is an \*(Na, then the other argument is returned.
 If both arguments are \*(Nas, then the result is an \*(Na.
 These routines do not raise any floating-point exceptions.
@@ -90,6 +86,13 @@ and
 .Fn fminl
 functions conform to
 .St -isoC-99 .
+.Sh CAVEATS
+The library implementations of these functions treat
+.Li +0.0
+as being larger than
+.Li -0.0 .
+This behavior is not specified by the C standard, is not portable,
+and may not occur in light of compiler optimizations.
 .Sh HISTORY
 These routines first appeared in
 .Fx 5.3 .