From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Michael Kerrisk (man-pages)" Subject: Re: [PATCH] mm: msync: require either MS_ASYNC or MS_SYNC Date: Fri, 04 Apr 2014 09:11:19 +0200 Message-ID: <533E5B17.8010804@gmail.com> References: <533B04A9.6090405@bbn.com> <20140402111032.GA27551@infradead.org> <1396439119.2726.29.camel@menhir> <533CA0F6.2070100@bbn.com> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Return-path: In-Reply-To: Sender: owner-linux-mm@kvack.org To: Greg Troxel Cc: mtk.manpages@gmail.com, Richard Hansen , Steven Whitehouse , Christoph Hellwig , "linux-mm@kvack.org" , lkml , Linux API , Peter Zijlstra List-Id: linux-api@vger.kernel.org Hi Greg, On 04/03/2014 02:57 PM, Greg Troxel wrote: >=20 > "Michael Kerrisk (man-pages)" writes: >=20 >> I think the only reasonable solution is to better document existing >> behavior and what the programmer should do. With that in mind, I've >> drafted the following text for the msync(2) man page: >> >> NOTES >> According to POSIX, exactly one of MS_SYNC and MS_ASYNC must = be >> specified in flags. However, Linux permits a call to msync= () >> that specifies neither of these flags, with semantics that a= re >> (currently) equivalent to specifying MS_ASYNC. (Since Lin= ux >> 2.6.19, MS_ASYNC is in fact a no-op, since the kernel proper= ly >> tracks dirty pages and flushes them to storage as necessary= .) >> Notwithstanding the Linux behavior, portable, future-proof appl= i=E2=80=90 >> cations should ensure that they specify exactly one of MS_SY= NC >> and MS_ASYNC in flags. >> >> Comments on this draft welcome. >=20 > I think it's a step backwards to document unspecified behavior. If > anything, the man page should make it clear that providing neither flag > results in undefined behavior and will lead to failure on systems on > than Linux. While I can see the point of not changing the previous > behavior to protect buggy code, there's no need to document it in the > man page and further enshrine it. The Linux behavior is what it is. For the reasons I mentioned already, it's unlikely to change. And, when the man pages omit to explain what Linux actually does, there will one day come a request to actually document the behavior. So, I don't think it's quite enough to say the=20 behavior is undefined. On the other hand, it does not hurt to further expand the portability warning. I made the text now: NOTES According to POSIX, either MS_SYNC or MS_ASYNC must be specified in flags, and indeed failure to include one of these flags will cause msync() to fail on some systems. However, Linux permits a call to msync() that specifies neither of these flags, with semantics that are (currently) equivalent to specifying MS_ASYNC. (Since Linux 2.6.19, MS_ASYNC is in fact a no-op, since the ker=E2= =80=90 nel properly tracks dirty pages and flushes them to storage as necessary.) Notwithstanding the Linux behavior, portable, future-proof applications should ensure that they specify either MS_SYNC or MS_ASYNC in flags. --=20 Michael Kerrisk Linux man-pages maintainer; http://www.kernel.org/doc/man-pages/ Linux/UNIX System Programming Training: http://man7.org/training/ -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: email@kvack.org