linux-fsdevel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Boaz Harrosh <bharrosh@panasas.com>
To: Christoph Hellwig <hch@lst.de>
Cc: linux-fsdevel@vger.kernel.org
Subject: Re: exofs_file_fsync
Date: Mon, 31 May 2010 16:46:33 +0300	[thread overview]
Message-ID: <4C03BDB9.30904@panasas.com> (raw)
In-Reply-To: <4C03BD14.5030004@panasas.com>

On 05/31/2010 04:43 PM, Boaz Harrosh wrote:
> On 05/31/2010 01:33 PM, Christoph Hellwig wrote:
>> On Mon, May 31, 2010 at 01:31:01PM +0300, Boaz Harrosh wrote:
>>> OK, I was just looking at that. thanks you saved me some digging.
>>> should I just open-code the generic_file_fsync minus the blocks
>>> thing then?
>>
>> sync_mapping_buffers is a no-op for you so you can keep it.
>> The other difference is that you sync out the superblock at the
>> end of your fsync implementation.  That is rather unusual, but I don't
>> know enough about exofs if you really need to update data in the
>> superblock to commit file data to disk.
>>
>>> I'm busy with the truncate stuff, but I'll do this next.
>>> Do you need this ASAP?
>>
>> I just noticed it while walking through the fsync implementations.
>>
> 
> OK Chritoff I would need your help Please.
> 
> It looks like what I need exactly is:
> 	write_inode_now(inode, sync)
> 
> But write_inode_now() has one extra hunk over generic_file_fsync:
> 	if (sync)
> 		inode_sync_wait(inode);
> 
> Do you think I can get in trouble calling it from ->fsync
> 
> I don't like generic_file_fsync because it does not write my
> data since I don't have buffer_heads.
> 
> OK, I'm totally lost what does ->fsync need to do? only write
> the inode or the pages as well?
> 
> Boaz
> ---
> git diff --stat -p -M fs/exofs/file.c
>  fs/exofs/file.c |    8 ++------
>  1 files changed, 2 insertions(+), 6 deletions(-)
> 
> diff --git a/fs/exofs/file.c b/fs/exofs/file.c
> index f9bfe2b..9b3555e 100644
> --- a/fs/exofs/file.c
> +++ b/fs/exofs/file.c
> @@ -47,18 +47,14 @@ static int exofs_file_fsync(struct file *filp, int datasync)
>  	struct inode *inode = mapping->host;
>  	struct super_block *sb;
>  
> -	ret = filemap_write_and_wait(mapping);
> -	if (ret)
> -		return ret;
> -

You think I also need a:
	if (!(inode->i_state & I_DIRTY))
		return ret;
	if (datasync && !(inode->i_state & I_DIRTY_DATASYNC))
		return ret;
before the  write_inode_now call ?

>  	/* sync the inode attributes */
> -	ret = write_inode_now(inode, 1);
> +	ret = write_inode_now(inode, datasync);
>  
>  	/* This is a good place to write the sb */
>  	/* TODO: Sechedule an sb-sync on create */
>  	sb = inode->i_sb;
>  	if (sb->s_dirt)
> -		exofs_sync_fs(sb, 1);
> +		exofs_sync_fs(sb, datasync);
>  
>  	return ret;
>  }


  reply	other threads:[~2010-05-31 13:46 UTC|newest]

Thread overview: 20+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2010-05-31 10:09 exofs_file_fsync Christoph Hellwig
2010-05-31 10:23 ` exofs_file_fsync Boaz Harrosh
2010-05-31 10:27   ` exofs_file_fsync Christoph Hellwig
2010-05-31 10:31     ` exofs_file_fsync Boaz Harrosh
2010-05-31 10:33       ` exofs_file_fsync Christoph Hellwig
2010-05-31 13:43         ` exofs_file_fsync Boaz Harrosh
2010-05-31 13:46           ` Boaz Harrosh [this message]
2010-06-01 10:05             ` exofs_file_fsync Christoph Hellwig
2010-06-01 10:04           ` exofs_file_fsync Christoph Hellwig
2010-06-01 15:29             ` exofs_file_fsync Boaz Harrosh
2010-06-01 15:30 ` [PATCH] exofs: exofs_file_fsync correctness Boaz Harrosh
2010-06-01 15:33   ` [osd-dev] " Boaz Harrosh
2010-06-01 15:34   ` Christoph Hellwig
2010-06-01 15:40     ` Boaz Harrosh
2010-06-01 15:47       ` Christoph Hellwig
2010-06-01 16:10         ` Boaz Harrosh
2010-06-01 16:11   ` [PATCH ver2] exofs: exofs_file_fsync and exofs_file_flush correctness Boaz Harrosh
2010-06-01 16:17     ` Christoph Hellwig
2010-06-01 16:36       ` Boaz Harrosh
2010-06-01 17:03   ` [PATCH ver3] " Boaz Harrosh

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=4C03BDB9.30904@panasas.com \
    --to=bharrosh@panasas.com \
    --cc=hch@lst.de \
    --cc=linux-fsdevel@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).