[PATCH] Btrfs: remove wait_ordered_range in btrfs_evict_inode

[PATCH] Btrfs: remove wait_ordered_range in btrfs_evict_inode

[Liu Bo]

As VFS has called ->invalidatepage() to get all ordered extents done
and truncated all page cache pages, wait_ordered_range() is just a
noop.

Signed-off-by: Liu Bo <bo.liu@linux.alibaba.com>
---
 fs/btrfs/inode.c | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index ff1d2ed2dc94..d3febc3a6bc0 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -5390,9 +5390,6 @@ void btrfs_evict_inode(struct inode *inode)
 
 	if (is_bad_inode(inode))
 		goto no_delete;
-	/* do we really want it for ->i_nlink > 0 and zero btrfs_root_refs? */
-	if (!special_file(inode->i_mode))
-		btrfs_wait_ordered_range(inode, 0, (u64)-1);
 
 	btrfs_free_io_failure_record(BTRFS_I(inode), 0, (u64)-1);
 
-- 
1.8.3.1

Re: [PATCH] Btrfs: remove wait_ordered_range in btrfs_evict_inode

[David Sterba]

On Wed, Sep 12, 2018 at 06:06:22AM +0800, Liu Bo wrote:
> As VFS has called ->invalidatepage() to get all ordered extents done
> and truncated all page cache pages, wait_ordered_range() is just a
> noop.

Agreed, though looking up the exact points when there are no pages to be
waited for took me some time. More references and hints in the changelog
would be good. I'll add the patch to misc-next so it can be tested, but
please send me an update if you have idea how to rephrase the changelog.
Thanks.

Reviewed-by: David Sterba <dsterba@suse.com>

Re: [PATCH] Btrfs: remove wait_ordered_range in btrfs_evict_inode

[Liu Bo]

On Fri, Sep 14, 2018 at 03:14:33PM +0200, David Sterba wrote:
> On Wed, Sep 12, 2018 at 06:06:22AM +0800, Liu Bo wrote:
> > As VFS has called ->invalidatepage() to get all ordered extents done
> > and truncated all page cache pages, wait_ordered_range() is just a
> > noop.
> 
> Agreed, though looking up the exact points when there are no pages to be
> waited for took me some time. More references and hints in the changelog
> would be good. I'll add the patch to misc-next so it can be tested, but
> please send me an update if you have idea how to rephrase the changelog.
> Thanks.
> 
> Reviewed-by: David Sterba <dsterba@suse.com>

OK, I'll add some call stack, hopefully can make it better.

thanks,
-liubo

[PATCH v2] Btrfs: remove wait_ordered_rane in btrfs_evict_inode

[Liu Bo]

When we delete an inode,

btrfs_evict_inode() {
    truncate_inode_pages_final()
        truncate_inode_pages_range()
            lock_page()
            truncate_cleanup_page()
                 btrfs_invalidatepage()
                      wait_on_page_writeback
                           btrfs_lookup_ordered_range()
                 cancel_dirty_page()
           unlock_page()
     ...    
     btrfs_wait_ordered_range()
     ...

As VFS has called ->invalidatepage() to get all ordered extents done
(if there is any) and truncated all page cache pages (no dirty pages
to writeback after this step), wait_ordered_range() is just a noop.

Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: Liu Bo <bo.liu@linux.alibaba.com>
---
v2: More details in the description.

 fs/btrfs/inode.c | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index ff1d2ed2dc94..d3febc3a6bc0 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -5390,9 +5390,6 @@ void btrfs_evict_inode(struct inode *inode)
 
 	if (is_bad_inode(inode))
 		goto no_delete;
-	/* do we really want it for ->i_nlink > 0 and zero btrfs_root_refs? */
-	if (!special_file(inode->i_mode))
-		btrfs_wait_ordered_range(inode, 0, (u64)-1);
 
 	btrfs_free_io_failure_record(BTRFS_I(inode), 0, (u64)-1);
 
-- 
1.8.3.1

Re: [PATCH v2] Btrfs: remove wait_ordered_rane in btrfs_evict_inode

[David Sterba]

On Sat, Sep 15, 2018 at 05:03:28AM +0800, Liu Bo wrote:
> When we delete an inode,
> 
> btrfs_evict_inode() {
>     truncate_inode_pages_final()
>         truncate_inode_pages_range()
>             lock_page()
>             truncate_cleanup_page()
>                  btrfs_invalidatepage()
>                       wait_on_page_writeback
>                            btrfs_lookup_ordered_range()
>                  cancel_dirty_page()
>            unlock_page()
>      ...    
>      btrfs_wait_ordered_range()
>      ...
> 
> As VFS has called ->invalidatepage() to get all ordered extents done
> (if there is any) and truncated all page cache pages (no dirty pages
> to writeback after this step), wait_ordered_range() is just a noop.
> 
> Reviewed-by: David Sterba <dsterba@suse.com>
> Signed-off-by: Liu Bo <bo.liu@linux.alibaba.com>
> ---
> v2: More details in the description.

Thanks, patch updated.