sync(2) is performed in two stages: the WB_SYNC_NONE sync and the WB_SYNC_ALL sync. Tag the first stage with wbc.tagged_sync and do livelock prevention for it, too. Note that writeback_inodes_sb() is called by not only sync(), they are treated the same because the other callers need also need livelock prevention. Impacts: - it changes the order in which pages/inodes are synced to disk. Now in the WB_SYNC_NONE stage, it won't proceed to write the next inode until finished with the current inode. - this adds a new field to the writeback trace events and may possibly break some scripts. CC: Jan Kara CC: Dave Chinner Signed-off-by: Wu Fengguang --- fs/ext4/inode.c | 4 ++-- fs/fs-writeback.c | 9 +++++---- include/linux/writeback.h | 1 + mm/page-writeback.c | 4 ++-- 4 files changed, 10 insertions(+), 8 deletions(-) --- linux-next.orig/fs/fs-writeback.c 2011-05-02 11:16:12.000000000 +0800 +++ linux-next/fs/fs-writeback.c 2011-05-02 11:17:33.000000000 +0800 @@ -36,6 +36,7 @@ struct wb_writeback_work { long nr_pages; struct super_block *sb; enum writeback_sync_modes sync_mode; + unsigned int tagged_sync:1; unsigned int for_kupdate:1; unsigned int range_cyclic:1; unsigned int for_background:1; @@ -644,6 +645,7 @@ static long wb_writeback(struct bdi_writ { struct writeback_control wbc = { .sync_mode = work->sync_mode, + .tagged_sync = work->tagged_sync, .older_than_this = NULL, .for_kupdate = work->for_kupdate, .for_background = work->for_background, @@ -651,7 +653,7 @@ static long wb_writeback(struct bdi_writ }; unsigned long oldest_jif; long wrote = 0; - long write_chunk; + long write_chunk = MAX_WRITEBACK_PAGES; struct inode *inode; if (!wbc.range_cyclic) { @@ -672,9 +674,7 @@ static long wb_writeback(struct bdi_writ * (quickly) tag currently dirty pages * (maybe slowly) sync all tagged pages */ - if (wbc.sync_mode == WB_SYNC_NONE) - write_chunk = MAX_WRITEBACK_PAGES; - else + if (wbc.sync_mode == WB_SYNC_ALL || wbc.tagged_sync) write_chunk = LONG_MAX; wbc.wb_start = jiffies; /* livelock avoidance */ @@ -1209,6 +1209,7 @@ void writeback_inodes_sb_nr(struct super struct wb_writeback_work work = { .sb = sb, .sync_mode = WB_SYNC_NONE, + .tagged_sync = 1, .done = &done, .nr_pages = nr, }; --- linux-next.orig/include/linux/writeback.h 2011-05-02 11:16:12.000000000 +0800 +++ linux-next/include/linux/writeback.h 2011-05-02 11:17:31.000000000 +0800 @@ -46,6 +46,7 @@ struct writeback_control { unsigned encountered_congestion:1; /* An output: a queue is full */ unsigned for_kupdate:1; /* A kupdate writeback */ unsigned for_background:1; /* A background writeback */ + unsigned tagged_sync:1; /* do livelock prevention for sync */ unsigned for_reclaim:1; /* Invoked from the page allocator */ unsigned range_cyclic:1; /* range_start is cyclic */ unsigned more_io:1; /* more io to be dispatched */ --- linux-next.orig/mm/page-writeback.c 2011-05-02 11:07:57.000000000 +0800 +++ linux-next/mm/page-writeback.c 2011-05-02 11:16:12.000000000 +0800 @@ -892,12 +892,12 @@ int write_cache_pages(struct address_spa range_whole = 1; cycled = 1; /* ignore range_cyclic tests */ } - if (wbc->sync_mode == WB_SYNC_ALL) + if (wbc->sync_mode == WB_SYNC_ALL || wbc->tagged_sync) tag = PAGECACHE_TAG_TOWRITE; else tag = PAGECACHE_TAG_DIRTY; retry: - if (wbc->sync_mode == WB_SYNC_ALL) + if (wbc->sync_mode == WB_SYNC_ALL || wbc->tagged_sync) tag_pages_for_writeback(mapping, index, end); done_index = index; while (!done && (index <= end)) { --- linux-next.orig/fs/ext4/inode.c 2011-05-02 11:07:57.000000000 +0800 +++ linux-next/fs/ext4/inode.c 2011-05-02 11:16:12.000000000 +0800 @@ -2741,7 +2741,7 @@ static int write_cache_pages_da(struct a index = wbc->range_start >> PAGE_CACHE_SHIFT; end = wbc->range_end >> PAGE_CACHE_SHIFT; - if (wbc->sync_mode == WB_SYNC_ALL) + if (wbc->sync_mode == WB_SYNC_ALL || wbc->tagged_sync) tag = PAGECACHE_TAG_TOWRITE; else tag = PAGECACHE_TAG_DIRTY; @@ -2975,7 +2975,7 @@ static int ext4_da_writepages(struct add } retry: - if (wbc->sync_mode == WB_SYNC_ALL) + if (wbc->sync_mode == WB_SYNC_ALL || wbc->tagged_sync) tag_pages_for_writeback(mapping, index, end); while (!ret && wbc->nr_to_write > 0) { -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/