Home » Mailing list archives » linux-kernel » 2011 » January » 4

[PATCH resend] Update atime from future.

Previous thread: [PATCH] fs/9p: Use the correct dentry operations by Aneesh Kumar K.V on Tuesday, January 4, 2011 - 1:44 am. (1 message)

Next thread: [PATCH] Update atime from future. by yangsheng on Tuesday, January 4, 2011 - 2:08 am. (1 message)
[view original message]
From: yangsheng
Subject: [PATCH resend] Update atime from future.
Date: Tuesday, January 4, 2011 - 1:56 am

If atime has been wrong set to future, then it cannot
be updated back to current time.

CC: swhiteho@redhat.com
Signed-off-by: sickadm@gmail.com
Reviewed-by: adilger@dilger.ca
---
 fs/inode.c |   12 ++++++++++--
 1 files changed, 10 insertions(+), 2 deletions(-)

diff --git a/fs/inode.c b/fs/inode.c
index da85e56..9cf7375 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -1446,6 +1446,8 @@ sector_t bmap(struct inode *inode, sector_t block)
 }
 EXPORT_SYMBOL(bmap);
 
+#define RELATIME_MARGIN (24 * 60 * 60)
+
 /*
  * With relative atime, only update atime if the previous atime is
  * earlier than either the ctime or mtime or if at least a day has
@@ -1469,10 +1471,16 @@ static int relatime_need_update(struct vfsmount *mnt, struct inode *inode,
 		return 1;
 
 	/*
-	 * Is the previous atime value older than a day? If yes,
+	 * Is the previous atime value in future? If yes,
+	 * update atime:
+	 */
+	if ((long)(now.tv_sec - inode->i_atime.tv_sec) < -RELATIME_MARGIN)
+		return 1;
+	/*
+	 * Is the previous atime value old than a day? If yes,
 	 * update atime:
 	 */
-	if ((long)(now.tv_sec - inode->i_atime.tv_sec) >= 24*60*60)
+	if ((long)(now.tv_sec - inode->i_atime.tv_sec) >= RELATIME_MARGIN)
 		return 1;
 	/*
 	 * Good, we can skip the atime update:
-- 
1.7.2.3

--


[ message continues ]
[view original message]
From: Andreas Schwab
Subject: Re: [PATCH resend] Update atime from future.
Date: Tuesday, January 4, 2011 - 2:02 am

Why did you change that comment?

Andreas.

-- 
Andreas Schwab, schwab@linux-m68k.org
GPG Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."
--


[ message continues ]
[view original message]
From: YangSheng
Subject: Re: [PATCH resend] Update atime from future.
Date: Tuesday, January 4, 2011 - 2:05 am

--


[ message continues ]
[view original message]
From: Valdis.Kletnieks
Subject: Re: [PATCH resend] Update atime from future.
Date: Tuesday, January 4, 2011 - 11:21 am

Nice patch overall.  Should this be a #define, or a CONFIG_ variable,
or a tweakable /proc/sys/fs variable?  Or am I senile and we thrashed
all this out once before when the relatime code landed?

[ message continues ]
[view original message]
From: Andreas Dilger
Subject: Re: [PATCH resend] Update atime from future.
Date: Tuesday, January 4, 2011 - 12:13 pm

I recall the consensus was that a /proc tunable was "too much" for the initial patch.  An atime update interval of 1 day is sufficient for most applications, since they run daily to do file access scanning.  The #define was added because I dislike having multiple hard-coded values in any code.

I haven't heard of any complaints about the relatime update frequency, except for this "atime in the future" problem, so until that happens we may as well leave it as-is.

Cheers, Andreas





--


[ message continues ]
[view original message]
From: Valdis.Kletnieks
Subject: Re: [PATCH resend] Update atime from future.
Date: Tuesday, January 4, 2011 - 1:31 pm

OK, in that case yangsheng's patch is probably good to go.

[ message continues ]
Previous thread: [PATCH] fs/9p: Use the correct dentry operations by Aneesh Kumar K.V on Tuesday, January 4, 2011 - 1:44 am. (1 message)

Next thread: [PATCH] Update atime from future. by yangsheng on Tuesday, January 4, 2011 - 2:08 am. (1 message)