Skip to content

periodic write-back timer optimization

Artem Bityutskiy linux-fsdevel lkml 2 versions – In review

Get series with b4
b4 am [email protected]

Versions newest first · linked by author and title

  1. v3 Latest18 patches 7 repliesIn review
  2. v119 patches 8 repliesSuperseded

Patches in v3 No applies recorded yet

#Subject and filesSizeRepliesApplied
1/18VFS: introduce helpers for manipulation s_dirty flaginclude/linux/fs.h+17 -00 replies
2/18AFFS: do not manipulate s_dirt directlyfs/affs/bitmap.c fs/affs/super.c+11 -70 replies
3/18BFS: do not manipulate s_dirt directlyfs/bfs/inode.c+4 -40 replies
4/18EXOFS: do not manipulate s_dirt directlyfs/exofs/inode.c fs/exofs/super.c+4 -42 replies
5/18EXT2: do not manipulate s_dirt directlyfs/ext2/balloc.c fs/ext2/ialloc.c fs/ext2/super.c +1 more+10 -100 replies
6/18EXT4: do not manipulate s_dirt directlyfs/ext4/balloc.c fs/ext4/file.c fs/ext4/ialloc.c +5 more+13 -130 replies
7/18FAT: do not manipulate s_dirt directlyfs/fat/fatent.c fs/fat/inode.c+7 -70 replies
8/18HFS: do not manipulate s_dirt directlyfs/hfs/extent.c fs/hfs/hfs_fs.h fs/hfs/inode.c +1 more+8 -80 replies
9/18HFSPLUS: do not manipulate s_dirt directlyfs/hfsplus/bitmap.c fs/hfsplus/dir.c fs/hfsplus/inode.c +1 more+14 -140 replies
10/18JFFS2: do not manipulate s_dirt directlyfs/jffs2/os-linux.h fs/jffs2/super.c+3 -30 replies
11/18NILFS: do not manipulate s_dirt directlyfs/nilfs2/segment.c fs/nilfs2/super.c fs/nilfs2/the_nilfs.c+11 -110 replies
12/18reiserfs: do not manipulate s_dirt directlyfs/reiserfs/journal.c fs/reiserfs/resize.c fs/reiserfs/super.c+9 -90 replies
13/18SYSV: do not manipulate s_dirt directlyfs/sysv/inode.c fs/sysv/super.c fs/sysv/sysv.h+5 -50 replies
14/18UDF: do not manipulate s_dirt directlyfs/udf/super.c fs/udf/udfdecl.h+4 -40 replies
15/18UFS: do not manipulate s_dirt directlyfs/ufs/balloc.c fs/ufs/ialloc.c fs/ufs/super.c+12 -120 replies
16/18VFS: use is_sb_dirty helperfs/super.c fs/sync.c+3 -30 replies
17/18rename s_dirt to s_dirtyinclude/linux/fs.h+4 -40 replies
18/18writeback: optimize periodic sync_supersinclude/linux/fs.h mm/backing-dev.c+46 -50 replies

Reviews trailers per patch

3 of 18 patches have a Reviewed-by or Acked-by.

PatchBoaz HarroshDavid WoodhouseRyusuke Konishi
1/18
2/18
3/18
4/18A-b
5/18
6/18
7/18
8/18
9/18
10/18A-b
11/18A-b
12/18
13/18
14/18
15/18
16/18
17/18
18/18

Trailers given in replies are filled; those carried in the patches from earlier versions are outlined with dashes. Scroll the table sideways for all 3 reviewers.

Read the v3 thread

Cover letter [PATCH v3 00/18] periodic write-back timer optimization

Hi Al, Jens,

this is the third or fourth attempt. Al, would you please take a look at this? Would it be possible to get your ack and then merge this via Jens' tree?

The patches attempt to optimize the periodic write-back and stop it when there are no dirty data. IOW, we do not want the thread to wake up every 5 seconds (by default), find there is nothing to do, and so on. We want to save power instead.

The patches are against Jens Axboe's per-bdi write-back tree. Namely, against:

git://git.kernel.dk/linux-2.6-block.git writeback-v12

Patches 01-16 introduce 2 VFS helpers and wrap all 'sb->s_dirt' manipulations.

Patch 17 re-names s_dirt to s_dirty in order to make sure we catch all direct s_dirt accesses.

Patch 18 optimizes the periodic 'sync_supers' thread. This used to be 'pdflush', but Jens re-worked this area and now it is called 'sync_supers'. By the way, with Jens' changes the former kupdated stuff stuff looks a lot cleaner and it is easier to work on top of that.

Signature · 6 lines
Best Regards,
Artem Bityutskiy (�ртём Битюцкий)
--
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html