[Ocfs2-devel] [PATCH v2] ocfs2: Try to free truncate log when meeting ENOSPC in write.

[Joel Becker]

On Thu, Nov 04, 2010 at 03:14:11PM +0800, Tao Ma wrote:
> Recently, one of our colleagues meet with a problem that if we
> write/delete a 32mb files repeatly, we will get an ENOSPC in
> the end. And the corresponding bug is 1288.
> http://oss.oracle.com/bugzilla/show_bug.cgi?id=1288
> 
> The real problem is that although we have freed the clusters,
> they are in truncate log and they will be summed up so that
> we can free them once in a whole.
> 
> So this patch just try to resolve it. In case we see -ENOSPC
> in ocfs2_write_begin_no_lock, we will check whether the truncate
> log has enough clusters for our need, if yes, we will try to
> flush the truncate log at that point and try again. This method
> is inspired by Mark Fasheh <mfasheh at suse.com>. Thanks.
> 
> Cc: Mark Fasheh <mfasheh at suse.com>
> Signed-off-by: Tao Ma <tao.ma at oracle.com>

This is now in the merge-window branch of ocfs2.git.

Joel

> ---
>  fs/ocfs2/alloc.c |    3 ++
>  fs/ocfs2/aops.c  |   59 +++++++++++++++++++++++++++++++++++++++++++++++++++++-
>  fs/ocfs2/ocfs2.h |    5 ++++
>  3 files changed, 66 insertions(+), 1 deletions(-)
> 
> diff --git a/fs/ocfs2/alloc.c b/fs/ocfs2/alloc.c
> index 592fae5..8ec418d 100644
> --- a/fs/ocfs2/alloc.c
> +++ b/fs/ocfs2/alloc.c
> @@ -5858,6 +5858,7 @@ int ocfs2_truncate_log_append(struct ocfs2_super *osb,
>  
>  	ocfs2_journal_dirty(handle, tl_bh);
>  
> +	osb->truncated_clusters += num_clusters;
>  bail:
>  	mlog_exit(status);
>  	return status;
> @@ -5929,6 +5930,8 @@ static int ocfs2_replay_truncate_records(struct ocfs2_super *osb,
>  		i--;
>  	}
>  
> +	osb->truncated_clusters = 0;
> +
>  bail:
>  	mlog_exit(status);
>  	return status;
> diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c
> index 5cfeee1..108743f 100644
> --- a/fs/ocfs2/aops.c
> +++ b/fs/ocfs2/aops.c
> @@ -1642,6 +1642,43 @@ static int ocfs2_zero_tail(struct inode *inode, struct buffer_head *di_bh,
>  	return ret;
>  }
>  
> +/*
> + * Try to flush truncate logs if we can free enough clusters from it.
> + * As for return value, "< 0" means error, "0" no space and "1" means
> + * we have freed enough spaces and let the caller try to allocate again.
> + */
> +static int ocfs2_try_to_free_truncate_log(struct ocfs2_super *osb,
> +					  unsigned int needed)
> +{
> +	tid_t target;
> +	int ret = 0;
> +	unsigned int truncated_clusters;
> +
> +	mutex_lock(&osb->osb_tl_inode->i_mutex);
> +	truncated_clusters = osb->truncated_clusters;
> +	mutex_unlock(&osb->osb_tl_inode->i_mutex);
> +
> +	/*
> +	 * Check whether we can succeed in allocating if we free
> +	 * the truncate log.
> +	 */
> +	if (truncated_clusters < needed)
> +		goto out;
> +
> +	ret = ocfs2_flush_truncate_log(osb);
> +	if (ret) {
> +		mlog_errno(ret);
> +		goto out;
> +	}
> +
> +	if (jbd2_journal_start_commit(osb->journal->j_journal, &target)) {
> +		jbd2_log_wait_commit(osb->journal->j_journal, target);
> +		ret = 1;
> +	}
> +out:
> +	return ret;
> +}
> +
>  int ocfs2_write_begin_nolock(struct file *filp,
>  			     struct address_space *mapping,
>  			     loff_t pos, unsigned len, unsigned flags,
> @@ -1649,7 +1686,7 @@ int ocfs2_write_begin_nolock(struct file *filp,
>  			     struct buffer_head *di_bh, struct page *mmap_page)
>  {
>  	int ret, cluster_of_pages, credits = OCFS2_INODE_UPDATE_CREDITS;
> -	unsigned int clusters_to_alloc, extents_to_split;
> +	unsigned int clusters_to_alloc, extents_to_split, clusters_need = 0;
>  	struct ocfs2_write_ctxt *wc;
>  	struct inode *inode = mapping->host;
>  	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
> @@ -1658,7 +1695,9 @@ int ocfs2_write_begin_nolock(struct file *filp,
>  	struct ocfs2_alloc_context *meta_ac = NULL;
>  	handle_t *handle;
>  	struct ocfs2_extent_tree et;
> +	int try_free = 1, ret1;
>  
> +try_again:
>  	ret = ocfs2_alloc_write_ctxt(&wc, osb, pos, len, di_bh);
>  	if (ret) {
>  		mlog_errno(ret);
> @@ -1693,6 +1732,7 @@ int ocfs2_write_begin_nolock(struct file *filp,
>  		mlog_errno(ret);
>  		goto out;
>  	} else if (ret == 1) {
> +		clusters_need = wc->w_clen;
>  		ret = ocfs2_refcount_cow(inode, filp, di_bh,
>  					 wc->w_cpos, wc->w_clen, UINT_MAX);
>  		if (ret) {
> @@ -1707,6 +1747,7 @@ int ocfs2_write_begin_nolock(struct file *filp,
>  		mlog_errno(ret);
>  		goto out;
>  	}
> +	clusters_need += clusters_to_alloc;
>  
>  	di = (struct ocfs2_dinode *)wc->w_di_bh->b_data;
>  
> @@ -1829,6 +1870,22 @@ out:
>  		ocfs2_free_alloc_context(data_ac);
>  	if (meta_ac)
>  		ocfs2_free_alloc_context(meta_ac);
> +
> +	if (ret == -ENOSPC && try_free) {
> +		/*
> +		 * Try to free some truncate log so that we can have enough
> +		 * clusters to allocate.
> +		 */
> +		try_free = 0;
> +
> +		ret1 = ocfs2_try_to_free_truncate_log(osb, clusters_need);
> +		if (ret1 == 1)
> +			goto try_again;
> +
> +		if (ret1 < 0)
> +			mlog_errno(ret1);
> +	}
> +
>  	return ret;
>  }
>  
> diff --git a/fs/ocfs2/ocfs2.h b/fs/ocfs2/ocfs2.h
> index d840821..f641b27 100644
> --- a/fs/ocfs2/ocfs2.h
> +++ b/fs/ocfs2/ocfs2.h
> @@ -420,6 +420,11 @@ struct ocfs2_super
>  	struct inode			*osb_tl_inode;
>  	struct buffer_head		*osb_tl_bh;
>  	struct delayed_work		osb_truncate_log_wq;
> +	/*
> +	 * How many clusters in our truncate log.
> +	 * It must be protected by osb_tl_inode->i_mutex.
> +	 */
> +	unsigned int truncated_clusters;
>  
>  	struct ocfs2_node_map		osb_recovering_orphan_dirs;
>  	unsigned int			*osb_orphan_wipes;
> -- 
> 1.7.1.571.gba4d01
> 

-- 

"Egotist: a person more interested in himself than in me."
         - Ambrose Bierce 

Joel Becker
Senior Development Manager
Oracle
E-mail: joel.becker at oracle.com
Phone: (650) 506-8127