* [PATCH] Btrfs: allow compressed extents to be merged during defragment
@ 2013-08-02 6:49 Liu Bo
2013-08-02 8:11 ` Miao Xie
0 siblings, 1 reply; 3+ messages in thread
From: Liu Bo @ 2013-08-02 6:49 UTC (permalink / raw)
To: linux-btrfs
The rule originally comes from nocow writing, but snapshot-aware
defrag is a different case, the extent has been writen and we're
not going to change the extent but add a reference on the data.
So we're able to allow such compressed extents to be merged into
one bigger extent if they're pointing to the same data.
Signed-off-by: Liu Bo <bo.li.liu@oracle.com>
---
fs/btrfs/inode.c | 13 ++++++++-----
1 file changed, 8 insertions(+), 5 deletions(-)
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 55dda87..a7aeecc 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -2229,7 +2229,7 @@ static noinline bool record_extent_backrefs(struct btrfs_path *path,
static int relink_is_mergable(struct extent_buffer *leaf,
struct btrfs_file_extent_item *fi,
- u64 disk_bytenr)
+ u64 disk_bytenr, u8 compress)
{
if (btrfs_file_extent_disk_bytenr(leaf, fi) != disk_bytenr)
return 0;
@@ -2237,8 +2237,10 @@ static int relink_is_mergable(struct extent_buffer *leaf,
if (btrfs_file_extent_type(leaf, fi) != BTRFS_FILE_EXTENT_REG)
return 0;
- if (btrfs_file_extent_compression(leaf, fi) ||
- btrfs_file_extent_encryption(leaf, fi) ||
+ if (btrfs_file_extent_compression(leaf, fi) != compress)
+ return 0;
+
+ if (btrfs_file_extent_encryption(leaf, fi) ||
btrfs_file_extent_other_encoding(leaf, fi))
return 0;
@@ -2382,8 +2384,9 @@ again:
struct btrfs_file_extent_item);
extent_len = btrfs_file_extent_num_bytes(leaf, fi);
- if (relink_is_mergable(leaf, fi, new->bytenr) &&
- extent_len + found_key.offset == start) {
+ if (extent_len + found_key.offset == start &&
+ relink_is_mergable(leaf, fi, new->bytenr,
+ new->compress_type)) {
btrfs_set_file_extent_num_bytes(leaf, fi,
extent_len + len);
btrfs_mark_buffer_dirty(leaf);
--
1.8.1.4
^ permalink raw reply related [flat|nested] 3+ messages in thread* Re: [PATCH] Btrfs: allow compressed extents to be merged during defragment
2013-08-02 6:49 [PATCH] Btrfs: allow compressed extents to be merged during defragment Liu Bo
@ 2013-08-02 8:11 ` Miao Xie
2013-08-02 8:17 ` Liu Bo
0 siblings, 1 reply; 3+ messages in thread
From: Miao Xie @ 2013-08-02 8:11 UTC (permalink / raw)
To: Liu Bo; +Cc: linux-btrfs
On fri, 2 Aug 2013 14:49:54 +0800, Liu Bo wrote:
> The rule originally comes from nocow writing, but snapshot-aware
> defrag is a different case, the extent has been writen and we're
> not going to change the extent but add a reference on the data.
>
> So we're able to allow such compressed extents to be merged into
> one bigger extent if they're pointing to the same data.
>
> Signed-off-by: Liu Bo <bo.li.liu@oracle.com>
> ---
> fs/btrfs/inode.c | 13 ++++++++-----
> 1 file changed, 8 insertions(+), 5 deletions(-)
>
> diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
> index 55dda87..a7aeecc 100644
> --- a/fs/btrfs/inode.c
> +++ b/fs/btrfs/inode.c
> @@ -2229,7 +2229,7 @@ static noinline bool record_extent_backrefs(struct btrfs_path *path,
>
> static int relink_is_mergable(struct extent_buffer *leaf,
> struct btrfs_file_extent_item *fi,
> - u64 disk_bytenr)
> + u64 disk_bytenr, u8 compress)
> {
> if (btrfs_file_extent_disk_bytenr(leaf, fi) != disk_bytenr)
> return 0;
> @@ -2237,8 +2237,10 @@ static int relink_is_mergable(struct extent_buffer *leaf,
> if (btrfs_file_extent_type(leaf, fi) != BTRFS_FILE_EXTENT_REG)
> return 0;
>
> - if (btrfs_file_extent_compression(leaf, fi) ||
> - btrfs_file_extent_encryption(leaf, fi) ||
> + if (btrfs_file_extent_compression(leaf, fi) != compress)
> + return 0;
> +
> + if (btrfs_file_extent_encryption(leaf, fi) ||
> btrfs_file_extent_other_encoding(leaf, fi))
> return 0;
>
> @@ -2382,8 +2384,9 @@ again:
> struct btrfs_file_extent_item);
> extent_len = btrfs_file_extent_num_bytes(leaf, fi);
>
> - if (relink_is_mergable(leaf, fi, new->bytenr) &&
> - extent_len + found_key.offset == start) {
> + if (extent_len + found_key.offset == start &&
> + relink_is_mergable(leaf, fi, new->bytenr,
> + new->compress_type)) {
There is a petty comment: Why not pass "new" to relink_is_mergable() directly?
The other code is OK.
Reviewed-by: Miao Xie <miaox@cn.fujitsu.com>
> btrfs_set_file_extent_num_bytes(leaf, fi,
> extent_len + len);
> btrfs_mark_buffer_dirty(leaf);
>
^ permalink raw reply [flat|nested] 3+ messages in thread* Re: [PATCH] Btrfs: allow compressed extents to be merged during defragment
2013-08-02 8:11 ` Miao Xie
@ 2013-08-02 8:17 ` Liu Bo
0 siblings, 0 replies; 3+ messages in thread
From: Liu Bo @ 2013-08-02 8:17 UTC (permalink / raw)
To: Miao Xie; +Cc: linux-btrfs
On Fri, Aug 02, 2013 at 04:11:16PM +0800, Miao Xie wrote:
> On fri, 2 Aug 2013 14:49:54 +0800, Liu Bo wrote:
> > The rule originally comes from nocow writing, but snapshot-aware
> > defrag is a different case, the extent has been writen and we're
> > not going to change the extent but add a reference on the data.
> >
> > So we're able to allow such compressed extents to be merged into
> > one bigger extent if they're pointing to the same data.
> >
> > Signed-off-by: Liu Bo <bo.li.liu@oracle.com>
> > ---
> > fs/btrfs/inode.c | 13 ++++++++-----
> > 1 file changed, 8 insertions(+), 5 deletions(-)
> >
> > diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
> > index 55dda87..a7aeecc 100644
> > --- a/fs/btrfs/inode.c
> > +++ b/fs/btrfs/inode.c
> > @@ -2229,7 +2229,7 @@ static noinline bool record_extent_backrefs(struct btrfs_path *path,
> >
> > static int relink_is_mergable(struct extent_buffer *leaf,
> > struct btrfs_file_extent_item *fi,
> > - u64 disk_bytenr)
> > + u64 disk_bytenr, u8 compress)
> > {
> > if (btrfs_file_extent_disk_bytenr(leaf, fi) != disk_bytenr)
> > return 0;
> > @@ -2237,8 +2237,10 @@ static int relink_is_mergable(struct extent_buffer *leaf,
> > if (btrfs_file_extent_type(leaf, fi) != BTRFS_FILE_EXTENT_REG)
> > return 0;
> >
> > - if (btrfs_file_extent_compression(leaf, fi) ||
> > - btrfs_file_extent_encryption(leaf, fi) ||
> > + if (btrfs_file_extent_compression(leaf, fi) != compress)
> > + return 0;
> > +
> > + if (btrfs_file_extent_encryption(leaf, fi) ||
> > btrfs_file_extent_other_encoding(leaf, fi))
> > return 0;
> >
> > @@ -2382,8 +2384,9 @@ again:
> > struct btrfs_file_extent_item);
> > extent_len = btrfs_file_extent_num_bytes(leaf, fi);
> >
> > - if (relink_is_mergable(leaf, fi, new->bytenr) &&
> > - extent_len + found_key.offset == start) {
> > + if (extent_len + found_key.offset == start &&
> > + relink_is_mergable(leaf, fi, new->bytenr,
> > + new->compress_type)) {
>
> There is a petty comment: Why not pass "new" to relink_is_mergable() directly?
I was thinking that there're only two args here, but the idea is nice,
I'll adopt it in v2, thanks,
-liubo
>
> The other code is OK.
>
> Reviewed-by: Miao Xie <miaox@cn.fujitsu.com>
>
> > btrfs_set_file_extent_num_bytes(leaf, fi,
> > extent_len + len);
> > btrfs_mark_buffer_dirty(leaf);
> >
>
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2013-08-02 8:17 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2013-08-02 6:49 [PATCH] Btrfs: allow compressed extents to be merged during defragment Liu Bo
2013-08-02 8:11 ` Miao Xie
2013-08-02 8:17 ` Liu Bo
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).