On Tue 06-02-07 14:23:33, Christoph Hellwig wrote:
> Iterate over sb->s_inodes instead of sb->s_files in add_dquot_ref.
> This reduces list search and lock hold time aswell as getting rid of
> one of the few uses of file_list_lock which Ingo identified as a
> scalability problem.
>
> Previously we called dq_op->initialize for every inode handing of
> a writeable file that wasn't initialized before. Now we're calling
> it for every inode that has a non-zero i_writecount, aka a writeable
> file descriptor refering to it.
>
> Thanks a lot to Jan Kara for running this patch through his quota
> test harness.
>
> Note: this is ontop of '[PATCH] move remove_dquot_ref to dqout.c'
> which I sent out yesterday.
>
>
> Signed-off-by: Christoph Hellwig <hch@xxxxxx>
Signed-off-by: Jan Kara <jack@xxxxxxx>
> Index: linux-2.6/fs/dquot.c
> ===================================================================
> --- linux-2.6.orig/fs/dquot.c 2007-02-05 18:54:36.000000000 +0100
> +++ linux-2.6/fs/dquot.c 2007-02-05 18:59:48.000000000 +0100
> @@ -689,23 +689,27 @@
> /* This routine is guarded by dqonoff_mutex mutex */
> static void add_dquot_ref(struct super_block *sb, int type)
> {
> - struct list_head *p;
> + struct inode *inode;
>
> restart:
> - file_list_lock();
> - list_for_each(p, &sb->s_files) {
> - struct file *filp = list_entry(p, struct file, f_u.fu_list);
> - struct inode *inode = filp->f_path.dentry->d_inode;
> - if (filp->f_mode & FMODE_WRITE && dqinit_needed(inode, type)) {
> - struct dentry *dentry = dget(filp->f_path.dentry);
> - file_list_unlock();
> - sb->dq_op->initialize(inode, type);
> - dput(dentry);
> - /* As we may have blocked we had better restart... */
> - goto restart;
> - }
> + spin_lock(&inode_lock);
> + list_for_each_entry(inode, &sb->s_inodes, i_sb_list) {
> + if (!atomic_read(&inode->i_writecount))
> + continue;
> + if (!dqinit_needed(inode, type))
> + continue;
> + if (inode->i_state & (I_FREEING|I_WILL_FREE))
> + continue;
> +
> + __iget(inode);
> + spin_unlock(&inode_lock);
> +
> + sb->dq_op->initialize(inode, type);
> + iput(inode);
> + /* As we may have blocked we had better restart... */
> + goto restart;
> }
> - file_list_unlock();
> + spin_unlock(&inode_lock);
> }
>
> /* Return 0 if dqput() won't block (note that 1 doesn't necessarily mean blocking) */
Honza
--
Jan Kara <jack@xxxxxxx>
SuSE CR Labs
-
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
