From: Jaegeuk Kim <jaegeuk@kernel.org>
To: Chao Yu <chao@kernel.org>
Cc: zhangshiming@oppo.com, linux-f2fs-devel@lists.sourceforge.net
Subject: Re: [f2fs-dev] [PATCH] f2fs: allocate consective blkaddrs for defragment
Date: Fri, 19 Aug 2022 16:42:03 -0700 [thread overview]
Message-ID: <YwAfyz3kR4mG5h7f@google.com> (raw)
In-Reply-To: <42dd80bb-c130-2408-7605-770b70a70f2e@kernel.org>
On 08/15, Chao Yu wrote:
> Hi Jaegeuk,
>
> On 2022/8/13 1:38, Jaegeuk Kim wrote:
> > Hi Weichao,
> >
> > On 08/12, Weichao Guo wrote:
> > > When we try to defrag a file, its data blocks may mess with others if there
> > > are lots of concurrent writes. This causes the file is still fragmented
> > > after defrag. So It's better to isolate defrag writes from others.
> >
> > I really don't want to add more log like this. What about using
>
> Any concern about adding more in-mem logs?
It gives more complexity and abuses free sections.
>
> > wb_sync_req[DATA] to prevent async writes at least?
>
> __should_serialize_io() says for synchronous write or small writes, it won't
> use writepages lock to serialize IO, so it may cause fragmentation if such
> writes were mixed w/ writes from defragment ioctl.
I think that, if you want to make a perfect layout, it'd be better to freeze
f2fs. Can we measure the realistic overhead on our current best-effort approach?
>
> Thanks,
>
> >
> > >
> > > Signed-off-by: Weichao Guo <guoweichao@oppo.com>
> > > Signed-off-by: Chao Yu <chao@kernel.org>
> > > ---
> > > fs/f2fs/debug.c | 4 ++++
> > > fs/f2fs/f2fs.h | 2 ++
> > > fs/f2fs/file.c | 5 +++++
> > > fs/f2fs/segment.c | 7 +++++++
> > > fs/f2fs/segment.h | 5 ++++-
> > > 5 files changed, 22 insertions(+), 1 deletion(-)
> > >
> > > diff --git a/fs/f2fs/debug.c b/fs/f2fs/debug.c
> > > index c014715..d85dc17 100644
> > > --- a/fs/f2fs/debug.c
> > > +++ b/fs/f2fs/debug.c
> > > @@ -442,6 +442,10 @@ static int stat_show(struct seq_file *s, void *v)
> > > si->curseg[CURSEG_ALL_DATA_ATGC],
> > > si->cursec[CURSEG_ALL_DATA_ATGC],
> > > si->curzone[CURSEG_ALL_DATA_ATGC]);
> > > + seq_printf(s, " - DEFRAG data: %8d %8d %8d\n",
> > > + si->curseg[CURSEG_COLD_DATA_DEFRAG],
> > > + si->cursec[CURSEG_COLD_DATA_DEFRAG],
> > > + si->curzone[CURSEG_COLD_DATA_DEFRAG]);
> > > seq_printf(s, "\n - Valid: %d\n - Dirty: %d\n",
> > > si->main_area_segs - si->dirty_count -
> > > si->prefree_count - si->free_segs,
> > > diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h
> > > index 3c7cdb7..5f9a185 100644
> > > --- a/fs/f2fs/f2fs.h
> > > +++ b/fs/f2fs/f2fs.h
> > > @@ -760,6 +760,7 @@ enum {
> > > FI_COMPRESS_RELEASED, /* compressed blocks were released */
> > > FI_ALIGNED_WRITE, /* enable aligned write */
> > > FI_COW_FILE, /* indicate COW file */
> > > + FI_DEFRAG_WRITE, /* indicate defragment writes need consective blkaddrs*/
> > > FI_MAX, /* max flag, never be used */
> > > };
> > > @@ -1017,6 +1018,7 @@ enum {
> > > CURSEG_COLD_DATA_PINNED = NR_PERSISTENT_LOG,
> > > /* pinned file that needs consecutive block address */
> > > CURSEG_ALL_DATA_ATGC, /* SSR alloctor in hot/warm/cold data area */
> > > + CURSEG_COLD_DATA_DEFRAG,/* used for defragment which needs consective blkaddrs */
> > > NO_CHECK_TYPE, /* number of persistent & inmem log */
> > > };
> > > diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c
> > > index ce4905a0..f104e2e 100644
> > > --- a/fs/f2fs/file.c
> > > +++ b/fs/f2fs/file.c
> > > @@ -2626,7 +2626,12 @@ static int f2fs_defragment_range(struct f2fs_sb_info *sbi,
> > > clear_inode_flag(inode, FI_SKIP_WRITES);
> > > + set_inode_flag(inode, FI_DEFRAG_WRITE);
> > > + f2fs_lock_op(sbi);
> > > + f2fs_allocate_new_section(sbi, CURSEG_COLD_DATA_DEFRAG, false);
> > > + f2fs_unlock_op(sbi);
> > > err = filemap_fdatawrite(inode->i_mapping);
> > > + clear_inode_flag(inode, FI_DEFRAG_WRITE);
> > > if (err)
> > > goto out;
> > > }
> > > diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c
> > > index 0de21f8..17e7d14 100644
> > > --- a/fs/f2fs/segment.c
> > > +++ b/fs/f2fs/segment.c
> > > @@ -2749,6 +2749,7 @@ static void __f2fs_save_inmem_curseg(struct f2fs_sb_info *sbi, int type)
> > > void f2fs_save_inmem_curseg(struct f2fs_sb_info *sbi)
> > > {
> > > __f2fs_save_inmem_curseg(sbi, CURSEG_COLD_DATA_PINNED);
> > > + __f2fs_save_inmem_curseg(sbi, CURSEG_COLD_DATA_DEFRAG);
> > > if (sbi->am.atgc_enabled)
> > > __f2fs_save_inmem_curseg(sbi, CURSEG_ALL_DATA_ATGC);
> > > @@ -2774,6 +2775,7 @@ static void __f2fs_restore_inmem_curseg(struct f2fs_sb_info *sbi, int type)
> > > void f2fs_restore_inmem_curseg(struct f2fs_sb_info *sbi)
> > > {
> > > __f2fs_restore_inmem_curseg(sbi, CURSEG_COLD_DATA_PINNED);
> > > + __f2fs_restore_inmem_curseg(sbi, CURSEG_COLD_DATA_DEFRAG);
> > > if (sbi->am.atgc_enabled)
> > > __f2fs_restore_inmem_curseg(sbi, CURSEG_ALL_DATA_ATGC);
> > > @@ -3161,6 +3163,9 @@ static int __get_segment_type_6(struct f2fs_io_info *fio)
> > > if (is_inode_flag_set(inode, FI_ALIGNED_WRITE))
> > > return CURSEG_COLD_DATA_PINNED;
> > > + if (is_inode_flag_set(inode, FI_DEFRAG_WRITE))
> > > + return CURSEG_COLD_DATA_DEFRAG;
> > > +
> > > if (page_private_gcing(fio->page)) {
> > > if (fio->sbi->am.atgc_enabled &&
> > > (fio->io_type == FS_DATA_IO) &&
> > > @@ -4332,6 +4337,8 @@ static int build_curseg(struct f2fs_sb_info *sbi)
> > > array[i].seg_type = CURSEG_COLD_DATA;
> > > else if (i == CURSEG_ALL_DATA_ATGC)
> > > array[i].seg_type = CURSEG_COLD_DATA;
> > > + else if (i == CURSEG_COLD_DATA_DEFRAG)
> > > + array[i].seg_type = CURSEG_COLD_DATA;
> > > array[i].segno = NULL_SEGNO;
> > > array[i].next_blkoff = 0;
> > > array[i].inited = false;
> > > diff --git a/fs/f2fs/segment.h b/fs/f2fs/segment.h
> > > index d1d6376..75e2aa8 100644
> > > --- a/fs/f2fs/segment.h
> > > +++ b/fs/f2fs/segment.h
> > > @@ -44,7 +44,8 @@ static inline void sanity_check_seg_type(struct f2fs_sb_info *sbi,
> > > ((seg) == CURSEG_I(sbi, CURSEG_WARM_NODE)->segno) || \
> > > ((seg) == CURSEG_I(sbi, CURSEG_COLD_NODE)->segno) || \
> > > ((seg) == CURSEG_I(sbi, CURSEG_COLD_DATA_PINNED)->segno) || \
> > > - ((seg) == CURSEG_I(sbi, CURSEG_ALL_DATA_ATGC)->segno))
> > > + ((seg) == CURSEG_I(sbi, CURSEG_ALL_DATA_ATGC)->segno) || \
> > > + ((seg) == CURSEG_I(sbi, CURSEG_COLD_DATA_DEFRAG)->segno))
> > > #define IS_CURSEC(sbi, secno) \
> > > (((secno) == CURSEG_I(sbi, CURSEG_HOT_DATA)->segno / \
> > > @@ -62,6 +63,8 @@ static inline void sanity_check_seg_type(struct f2fs_sb_info *sbi,
> > > ((secno) == CURSEG_I(sbi, CURSEG_COLD_DATA_PINNED)->segno / \
> > > (sbi)->segs_per_sec) || \
> > > ((secno) == CURSEG_I(sbi, CURSEG_ALL_DATA_ATGC)->segno / \
> > > + (sbi)->segs_per_sec) || \
> > > + ((secno) == CURSEG_I(sbi, CURSEG_COLD_DATA_DEFRAG)->segno / \
> > > (sbi)->segs_per_sec))
> > > #define MAIN_BLKADDR(sbi) \
> > > --
> > > 2.7.4
_______________________________________________
Linux-f2fs-devel mailing list
Linux-f2fs-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linux-f2fs-devel
next prev parent reply other threads:[~2022-08-19 23:42 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-08-12 3:56 [f2fs-dev] [PATCH] f2fs: allocate consective blkaddrs for defragment Weichao Guo via Linux-f2fs-devel
2022-08-12 17:38 ` Jaegeuk Kim
2022-08-15 3:30 ` Chao Yu
2022-08-19 23:42 ` Jaegeuk Kim [this message]
2022-08-20 2:30 ` Chao Yu
[not found] ` <DM5PR17MB095307935ABD5D5A6EEDD050C6699@DM5PR17MB0953.namprd17.prod.outlook.com>
2022-08-19 23:45 ` Jaegeuk Kim
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=YwAfyz3kR4mG5h7f@google.com \
--to=jaegeuk@kernel.org \
--cc=chao@kernel.org \
--cc=linux-f2fs-devel@lists.sourceforge.net \
--cc=zhangshiming@oppo.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.