From fe26ea7a7a097749b8db45e67f5f5bad304c06d5 Mon Sep 17 00:00:00 2001 From: christos Date: Thu, 5 Feb 2015 22:05:33 +0000 Subject: Remove incorrect comment about size; mention that the malloc ambiguity is part of the C standard. --- lib/libc/stdlib/reallocarray.3 | 15 ++++----------- 1 file changed, 4 insertions(+), 11 deletions(-) (limited to 'lib/libc/stdlib/reallocarray.3') diff --git a/lib/libc/stdlib/reallocarray.3 b/lib/libc/stdlib/reallocarray.3 index 71459d1bc0d..e570abb56b9 100644 --- a/lib/libc/stdlib/reallocarray.3 +++ b/lib/libc/stdlib/reallocarray.3 @@ -1,4 +1,4 @@ -.\" $NetBSD: reallocarray.3,v 1.1 2015/02/05 20:02:28 christos Exp $ +.\" $NetBSD: reallocarray.3,v 1.2 2015/02/05 22:05:33 christos Exp $ .\" .Dd February 5, 2015 .Dt REALLOCARRAY 3 @@ -84,17 +84,10 @@ functions by centralizing the overflow check in the multiplication of and .Fa size . .Pp -Implementation issues prevent the function from being used correctly (a -.Dv 0 -.Fa size -parameter will return -.Dv ENOMEM -in the -.Ox -implementation), while there are still portability issues (it does not solve +There are still portability issues (it does not solve the .Dv 0 -sized allocation return ambiguity: does +sized allocation return ambiguity in the C standard: does .Fn reallocarray return .Dv NULL @@ -104,7 +97,7 @@ mean that an error occurred, and can someone check .Dv errno in that case to find out what happened?). .Pp -For those reasons +For this reason .Nx decided to go with an alternative implementation, and created .Xr reallocarr 3 . -- cgit