linux-next: manual merge of the logfs tree with Linus' tree

linux-next: manual merge of the logfs tree with Linus' tree

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 1268 bytes --]

Hi all,

Today's linux-next merge of the logfs tree got a conflict in
fs/logfs/file.c between commit 02c24a82187d ("fs: push i_mutex and
filemap_write_and_wait down into ->fsync() handlers") from Linus' tree
and commit 39da12ef4bbe ("logfs: take write mutex lock during fsync and
sync") from the logfs tree.

I have no idea what needs to be done here.  I fixed it like below to make
it build, but a better fix is needed.
-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au

diff --cc fs/logfs/file.c
index b548c87,f85d603..0000000
--- a/fs/logfs/file.c
+++ b/fs/logfs/file.c
@@@ -219,20 -219,13 +219,22 @@@ long logfs_ioctl(struct file *file, uns
  	}
  }
  
 -int logfs_fsync(struct file *file, int datasync)
 +int logfs_fsync(struct file *file, loff_t start, loff_t end, int datasync)
  {
  	struct super_block *sb = file->f_mapping->host->i_sb;
 +	struct inode *inode = file->f_mapping->host;
 +	int ret;
 +
 +	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
 +	if (ret)
 +		return ret;
  
 +	mutex_lock(&inode->i_mutex);
+ 	logfs_get_wblocks(sb, NULL, WF_LOCK);
  	logfs_write_anchor(sb);
+ 	logfs_put_wblocks(sb, NULL, WF_LOCK);
 +	mutex_unlock(&inode->i_mutex);
 +
  	return 0;
  }
  

[-- Attachment #2: Type: application/pgp-signature, Size: 836 bytes --]

Re: linux-next: manual merge of the logfs tree with Linus' tree

[Jörn Engel]

On Tue, 1 November 2011 14:10:00 +1100, Stephen Rothwell wrote:
> 
> Hi all,
> 
> Today's linux-next merge of the logfs tree got a conflict in
> fs/logfs/file.c between commit 02c24a82187d ("fs: push i_mutex and
> filemap_write_and_wait down into ->fsync() handlers") from Linus' tree
> and commit 39da12ef4bbe ("logfs: take write mutex lock during fsync and
> sync") from the logfs tree.
> 
> I have no idea what needs to be done here.  I fixed it like below to make
> it build, but a better fix is needed.

From a code perspective your fix below is correct, to the best of my
judgement.  I'm less sure what to do from a git perspective.
Explicitly tell Linus about it in the logfs pull request?

> diff --cc fs/logfs/file.c
> index b548c87,f85d603..0000000
> --- a/fs/logfs/file.c
> +++ b/fs/logfs/file.c
> @@@ -219,20 -219,13 +219,22 @@@ long logfs_ioctl(struct file *file, uns
>   	}
>   }
>   
>  -int logfs_fsync(struct file *file, int datasync)
>  +int logfs_fsync(struct file *file, loff_t start, loff_t end, int datasync)
>   {
>   	struct super_block *sb = file->f_mapping->host->i_sb;
>  +	struct inode *inode = file->f_mapping->host;
>  +	int ret;
>  +
>  +	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
>  +	if (ret)
>  +		return ret;
>   
>  +	mutex_lock(&inode->i_mutex);
> + 	logfs_get_wblocks(sb, NULL, WF_LOCK);
>   	logfs_write_anchor(sb);
> + 	logfs_put_wblocks(sb, NULL, WF_LOCK);
>  +	mutex_unlock(&inode->i_mutex);
>  +
>   	return 0;
>   }
>   



Jörn

-- 
You can take my soul, but not my lack of enthusiasm.
-- Wally

Re: linux-next: manual merge of the logfs tree with Linus' tree

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 1028 bytes --]

Hi Jörn,

On Wed, 2 Nov 2011 20:00:46 +0100 Jörn Engel <joern@logfs.org> wrote:
>
> On Tue, 1 November 2011 14:10:00 +1100, Stephen Rothwell wrote:
> > 
> > Today's linux-next merge of the logfs tree got a conflict in
> > fs/logfs/file.c between commit 02c24a82187d ("fs: push i_mutex and
> > filemap_write_and_wait down into ->fsync() handlers") from Linus' tree
> > and commit 39da12ef4bbe ("logfs: take write mutex lock during fsync and
> > sync") from the logfs tree.
> > 
> > I have no idea what needs to be done here.  I fixed it like below to make
> > it build, but a better fix is needed.
> 
> From a code perspective your fix below is correct, to the best of my
> judgement.  I'm less sure what to do from a git perspective.
> Explicitly tell Linus about it in the logfs pull request?

I was concered about the locking order (or if both locks were needed at
all).  And, yes, tell Linus.

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/

[-- Attachment #2: Type: application/pgp-signature, Size: 836 bytes --]

Re: linux-next: manual merge of the logfs tree with Linus' tree

[Jörn Engel]

On Thu, 3 November 2011 12:02:57 +1100, Stephen Rothwell wrote:
> On Wed, 2 Nov 2011 20:00:46 +0100 Jörn Engel <joern@logfs.org> wrote:
> > On Tue, 1 November 2011 14:10:00 +1100, Stephen Rothwell wrote:
> > > 
> > > Today's linux-next merge of the logfs tree got a conflict in
> > > fs/logfs/file.c between commit 02c24a82187d ("fs: push i_mutex and
> > > filemap_write_and_wait down into ->fsync() handlers") from Linus' tree
> > > and commit 39da12ef4bbe ("logfs: take write mutex lock during fsync and
> > > sync") from the logfs tree.
> > > 
> > > I have no idea what needs to be done here.  I fixed it like below to make
> > > it build, but a better fix is needed.
> > 
> > From a code perspective your fix below is correct, to the best of my
> > judgement.  I'm less sure what to do from a git perspective.
> > Explicitly tell Linus about it in the logfs pull request?
> 
> I was concered about the locking order (or if both locks were needed at
> all).  And, yes, tell Linus.

Locking order should be fine.  Whether both locks are needed is a
valid question.  I suspect the answer is yes.

Jörn

-- 
Simplicity is prerequisite for reliability.
-- Edsger W. Dijkstra