From nokubi@ff.iij4u.or.jp  Wed Jun 14 07:31:53 2000
Return-Path: <nokubi@ff.iij4u.or.jp>
Received: from mfi01.iij.ad.jp (mfi01.iij.ad.jp [202.232.2.116])
	by hub.freebsd.org (Postfix) with ESMTP id 4431037C21D
	for <FreeBSD-gnats-submit@freebsd.org>; Wed, 14 Jun 2000 07:31:51 -0700 (PDT)
	(envelope-from nokubi@ff.iij4u.or.jp)
Received: from ppp-client.ff.iij4u.or.jp (h141.p029.iij4u.or.jp [210.130.29.141])
	by mfi01.iij.ad.jp (8.8.8/MFI1.3) with ESMTP id XAA04483
	for <FreeBSD-gnats-submit@freebsd.org>; Wed, 14 Jun 2000 23:31:48 +0900 (JST)
Received: (from h-nokubi@localhost) by ppp-client.ff.iij4u.or.jp (8.9.3/3.5Wpl7-ppp) id XAA52526; Wed, 14 Jun 2000 23:06:12 +0900 (JST)
Message-Id: <200006141431.XAA04483@mfi01.iij.ad.jp>
Date: Wed, 14 Jun 2000 23:06:12 +0900 (JST)
From: nokubi@ff.iij4u.or.jp
Reply-To: nokubi@ff.iij4u.or.jp
To: FreeBSD-gnats-submit@freebsd.org
Subject: Is fsync.2 NAME section wrong? 
X-Send-Pr-Version: 3.2

>Number:         19262
>Category:       docs
>Synopsis:       Is fsync.2 NAME section wrong?
>Confidential:   no
>Severity:       non-critical
>Priority:       low
>Responsible:    ben
>State:          closed
>Quarter:        
>Keywords:       
>Date-Required:  
>Class:          doc-bug
>Submitter-Id:   current-users
>Arrival-Date:   Wed Jun 14 07:40:03 PDT 2000
>Closed-Date:    Sat Jul 15 14:02:15 BST 2000
>Last-Modified:  Sat Jul 15 14:06:06 BST 2000
>Originator:     NOKUBI Hirotaka
>Release:        FreeBSD 3.4-STABLE i386
>Organization:
>Environment:

	FreeBSD fifthluna.nokubi.or.jp 3.4-STABLE FreeBSD 3.4-STABLE #13: Mon Jan 31 00:17:41 JST 2000     root@sassaby.nokubi.or.jp:/usr/src/sys/compile/FIFTHLUNA  i386

>Description:

	Because I'm not native English speaker, I'm not sure about this.
	But I have strange feeling about fsync.2 and sync.2 NAME section.
	They say like this:
	  fsync - synchronize a file's in-core state with that on disk
	  sync - synchronize disk block in-core status with that on disk

	I think that fsync(2) and sync(2) write changes from memory to disk.
	But these sentences make me feel like that they read from disk
	to memory and discard changes.

	If it's my misunderstanding, please forget about this PR.

>How-To-Repeat:

	man 2 fsync
	man 2 sync

>Fix:
	
	SUSV2 saids like this:
	  fsync - synchronise changes to a file
	  sync - schedule filesystem updates


>Release-Note:
>Audit-Trail:

From: Ben Smithurst <ben@FreeBSD.org>
To: nokubi@ff.iij4u.or.jp
Cc: FreeBSD-gnats-submit@freebsd.org
Subject: Re: docs/19262: Is fsync.2 NAME section wrong?
Date: Tue, 11 Jul 2000 01:15:00 +0100

 nokubi@ff.iij4u.or.jp wrote:
 
 > 	Because I'm not native English speaker, I'm not sure about this.
 > 	But I have strange feeling about fsync.2 and sync.2 NAME section.
 > 	They say like this:
 > 	  fsync - synchronize a file's in-core state with that on disk
 > 	  sync - synchronize disk block in-core status with that on disk
 > 
 > 	I think that fsync(2) and sync(2) write changes from memory to disk.
 > 	But these sentences make me feel like that they read from disk
 > 	to memory and discard changes.
 
 Yes, I see what you mean...
 
 > 	SUSV2 saids like this:
 > 	  fsync - synchronise changes to a file
 > 	  sync - schedule filesystem updates
 
 They seem a bit short to me, but I can't seem to come up with any wording
 that I like. :-(
 
 fsync - write all in-core changes of a file to disk
 sync - schedule an update of all modified filesystem data to disk
 
 ?? Are they better?  Anyone care to comment?
 
 -- 
 Ben Smithurst / ben@FreeBSD.org / PGP: 0x99392F7D
 
Responsible-Changed-From-To: freebsd-doc->ben 
Responsible-Changed-By: ben 
Responsible-Changed-When: Thu Jul 13 13:17:36 BST 2000 
Responsible-Changed-Why:  
I'm looking at this one. 

http://www.freebsd.org/cgi/query-pr.cgi?pr=19262 

From: NOKUBI Hirotaka <nokubi@ff.iij4u.or.jp>
To: Ben Smithurst <ben@FreeBSD.org>
Cc: nokubi@ff.iij4u.or.jp, FreeBSD-gnats-submit@freebsd.org
Subject: Re: docs/19262: Is fsync.2 NAME section wrong? 
Date: Fri, 14 Jul 2000 20:39:53 +0900

 In message <20000711011500.A15649@magnesium.scientia.demon.co.uk>, Ben Smithurs
 t writes:
 
 >> 	Because I'm not native English speaker, I'm not sure about this.
 >> 	But I have strange feeling about fsync.2 and sync.2 NAME section.
 >> 	They say like this:
 >> 	  fsync - synchronize a file's in-core state with that on disk
 >> 	  sync - synchronize disk block in-core status with that on disk
 >> 
 >> 	I think that fsync(2) and sync(2) write changes from memory to disk.
 >> 	But these sentences make me feel like that they read from disk
 >> 	to memory and discard changes.
 >
 >Yes, I see what you mean...
 
 I'm relieved to here that. Because these NAME sections are remain
 unchanged for 15 years (at least, from 4.3BSD), I think it might be
 my misunderstanding.
 
 >> 	SUSV2 saids like this:
 >> 	  fsync - synchronise changes to a file
 >> 	  sync - schedule filesystem updates
 >
 >They seem a bit short to me, but I can't seem to come up with any wording
 >that I like. :-(
 >
 >fsync - write all in-core changes of a file to disk
 >sync - schedule an update of all modified filesystem data to disk
 >
 >?? Are they better?  Anyone care to comment?
 
 They sound very good for me. Please commit them.
 
 ----
 NOKUBI Hirotaka
 Fingerprint20 = DEBC 0793 7CD6 92F1 0A1F  A792 9E2F EEEE A41B 171D
 

From: Ben Smithurst <ben@FreeBSD.org>
To: NOKUBI Hirotaka <nokubi@ff.iij4u.or.jp>
Cc: FreeBSD-gnats-submit@freebsd.org
Subject: Re: docs/19262: Is fsync.2 NAME section wrong?
Date: Fri, 14 Jul 2000 14:58:34 +0100

 NOKUBI Hirotaka wrote:
 
 >>> 	SUSV2 saids like this:
 >>> 	  fsync - synchronise changes to a file
 >>> 	  sync - schedule filesystem updates
 >>
 >> fsync - write all in-core changes of a file to disk
 >> sync - schedule an update of all modified filesystem data to disk
 >>
 >> ?? Are they better?  Anyone care to comment?
 > 
 > They sound very good for me. Please commit them.
 
 Sheldon thinks the shorter SUSv2 versions above are better.  Would those
 be ok?
 
 -- 
 Ben Smithurst                 / ben@FreeBSD.org / PGP: 0x99392F7D
 FreeBSD Documentation Project /
 

From: NOKUBI Hirotaka <nokubi@ff.iij4u.or.jp>
To: Ben Smithurst <ben@FreeBSD.org>
Cc: NOKUBI Hirotaka <nokubi@ff.iij4u.or.jp>,
	FreeBSD-gnats-submit@freebsd.org
Subject: Re: docs/19262: Is fsync.2 NAME section wrong? 
Date: Sat, 15 Jul 2000 21:12:04 +0900

 In message <20000714145834.V48641@strontium.scientia.demon.co.uk>, Ben Smithurs
 t writes:
 >NOKUBI Hirotaka wrote:
 >
 >>>> 	SUSV2 saids like this:
 >>>> 	  fsync - synchronise changes to a file
 >>>> 	  sync - schedule filesystem updates
 >>>
 >>> fsync - write all in-core changes of a file to disk
 >>> sync - schedule an update of all modified filesystem data to disk
 >>>
 >>> ?? Are they better?  Anyone care to comment?
 >> 
 >> They sound very good for me. Please commit them.
 >
 >Sheldon thinks the shorter SUSv2 versions above are better.  Would those
 >be ok?
 
 Although I prefer more informative one, SUSv2 versions are acceptable.
 Thanks for your confirmation.
 ----
 NOKUBI Hirotaka
 Fingerprint20 = DEBC 0793 7CD6 92F1 0A1F  A792 9E2F EEEE A41B 171D
 
State-Changed-From-To: open->closed 
State-Changed-By: ben 
State-Changed-When: Sat Jul 15 14:02:15 BST 2000 
State-Changed-Why:  
Committed in -current and 4-stable, thanks! 

http://www.freebsd.org/cgi/query-pr.cgi?pr=19262 
>Unformatted:
