--- Begin Message ---
Subject: |
doc: Adds a bug section to sync.1 page to draw attention to unexpected behavior allowed by POSIX. |
Date: |
Wed, 4 Mar 2015 01:06:01 +0100 |
From: Stéphane Aulery <address@hidden>
Almost identical copy of sync.2.
Debian bug #507085 reported by address@hidden
Signed-off-by: Stéphane Aulery <address@hidden>
---
man/sync.x | 5 +++++
1 file changed, 5 insertions(+)
diff --git a/man/sync.x b/man/sync.x
index 79fee22..5b85bba 100644
--- a/man/sync.x
+++ b/man/sync.x
@@ -2,5 +2,10 @@
sync \- Synchronize cached writes to persistent storage
[DESCRIPTION]
.\" Add any additional description here
+[BUGS]
+According to the standard specification (e.g., POSIX.1-2001), sync()
+syscall schedules the writes, but may return before the actual writing is
+done. However, since version 1.3.20 Linux does actually wait. (This still
+does not guarantee data integrity: modern disks have large caches.)
[SEE ALSO]
fdatasync(2), fsync(2), sync(2), syncfs(2)
--
2.1.4
--- End Message ---
--- Begin Message ---
Subject: |
Re: bug#19995: [patch] sync.1: prevent data loss (reposte of my previous merged message) |
Date: |
Thu, 05 Mar 2015 12:23:44 +0000 |
User-agent: |
Mozilla/5.0 (X11; Linux x86_64; rv:31.0) Gecko/20100101 Thunderbird/31.4.0 |
On 05/03/15 11:45, Stéphane Aulery wrote:
> Hello,
>
> A Debian user reported [1] that sync.1 man page is incomplete unlike sync.2
> There is no mention of the unexpected behavior allowed by POSIX, which can
> cause data loss.
>
> This patch provides a possible correction.
>
> [1] https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=507085
How about I apply this in your name instead?
diff --git a/man/sync.x b/man/sync.x
index 79fee22..18ee3fc 100644
--- a/man/sync.x
+++ b/man/sync.x
@@ -2,5 +2,8 @@
sync \- Synchronize cached writes to persistent storage
[DESCRIPTION]
.\" Add any additional description here
+[BUGS]
+Persistence guarantees vary per system.
+See the system calls below for more details.
[SEE ALSO]
fdatasync(2), fsync(2), sync(2), syncfs(2)
--- End Message ---