2.6.31-stable review patch. If anyone has any objections, please let us know. ------------------ (cherry picked from commit 55138e0bc29c0751e2152df9ad35deea542f29b3) Work around problems in the writeback code to force out writebacks in larger chunks than just 4mb, which is just too small. This also works around limitations in the ext4 block allocator, which can't allocate more than 2048 blocks at a time. So we need to defeat the round-robin characteristics of the writeback code and try to write out as many blocks in one inode before allowing the writeback code to move on to another inode. We add a a new per-filesystem tunable, max_writeback_mb_bump, which caps this to a default of 128mb per inode. Signed-off-by: "Theodore Ts'o" Signed-off-by: Greg Kroah-Hartman --- fs/ext4/ext4.h | 17 ++++++ fs/ext4/inode.c | 121 +++++++++++++++++++++++++++++++++----------- fs/ext4/super.c | 3 + include/trace/events/ext4.h | 54 +++++++++++++++++-- 4 files changed, 161 insertions(+), 34 deletions(-) --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -114,6 +114,22 @@ struct ext4_allocation_request { }; /* + * Delayed allocation stuff + */ + +struct mpage_da_data { + struct inode *inode; + sector_t b_blocknr; /* start block number of extent */ + size_t b_size; /* size of extent */ + unsigned long b_state; /* state of the extent */ + unsigned long first_page, next_page; /* extent of pages */ + struct writeback_control *wbc; + int io_done; + int pages_written; + int retval; +}; + +/* * Special inodes numbers */ #define EXT4_BAD_INO 1 /* Bad blocks inode */ @@ -929,6 +945,7 @@ struct ext4_sb_info { unsigned int s_mb_stats; unsigned int s_mb_order2_reqs; unsigned int s_mb_group_prealloc; + unsigned int s_max_writeback_mb_bump; /* where last allocation was done - for stream allocation */ unsigned long s_mb_last_group; unsigned long s_mb_last_start; --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -1146,6 +1146,64 @@ static int check_block_validity(struct i } /* + * Return the number of dirty pages in the given inode starting at + * page frame idx. + */ +static pgoff_t ext4_num_dirty_pages(struct inode *inode, pgoff_t idx, + unsigned int max_pages) +{ + struct address_space *mapping = inode->i_mapping; + pgoff_t index; + struct pagevec pvec; + pgoff_t num = 0; + int i, nr_pages, done = 0; + + if (max_pages == 0) + return 0; + pagevec_init(&pvec, 0); + while (!done) { + index = idx; + nr_pages = pagevec_lookup_tag(&pvec, mapping, &index, + PAGECACHE_TAG_DIRTY, + (pgoff_t)PAGEVEC_SIZE); + if (nr_pages == 0) + break; + for (i = 0; i < nr_pages; i++) { + struct page *page = pvec.pages[i]; + struct buffer_head *bh, *head; + + lock_page(page); + if (unlikely(page->mapping != mapping) || + !PageDirty(page) || + PageWriteback(page) || + page->index != idx) { + done = 1; + unlock_page(page); + break; + } + head = page_buffers(page); + bh = head; + do { + if (!buffer_delay(bh) && + !buffer_unwritten(bh)) { + done = 1; + break; + } + } while ((bh = bh->b_this_page) != head); + unlock_page(page); + if (done) + break; + idx++; + num++; + if (num >= max_pages) + break; + } + pagevec_release(&pvec); + } + return num; +} + +/* * The ext4_get_blocks() function tries to look up the requested blocks, * and returns if the blocks are already mapped. * @@ -1881,22 +1939,6 @@ static void ext4_da_page_release_reserva } /* - * Delayed allocation stuff - */ - -struct mpage_da_data { - struct inode *inode; - sector_t b_blocknr; /* start block number of extent */ - size_t b_size; /* size of extent */ - unsigned long b_state; /* state of the extent */ - unsigned long first_page, next_page; /* extent of pages */ - struct writeback_control *wbc; - int io_done; - int pages_written; - int retval; -}; - -/* * mpage_da_submit_io - walks through extent of pages and try to write * them with writepage() call back * @@ -2756,8 +2798,10 @@ static int ext4_da_writepages(struct add int no_nrwrite_index_update; int pages_written = 0; long pages_skipped; + unsigned int max_pages; int range_cyclic, cycled = 1, io_done = 0; - int needed_blocks, ret = 0, nr_to_writebump = 0; + int needed_blocks, ret = 0; + long desired_nr_to_write, nr_to_writebump = 0; loff_t range_start = wbc->range_start; struct ext4_sb_info *sbi = EXT4_SB(mapping->host->i_sb); @@ -2784,16 +2828,6 @@ static int ext4_da_writepages(struct add if (unlikely(sbi->s_mount_flags & EXT4_MF_FS_ABORTED)) return -EROFS; - /* - * Make sure nr_to_write is >= sbi->s_mb_stream_request - * This make sure small files blocks are allocated in - * single attempt. This ensure that small files - * get less fragmented. - */ - if (wbc->nr_to_write < sbi->s_mb_stream_request) { - nr_to_writebump = sbi->s_mb_stream_request - wbc->nr_to_write; - wbc->nr_to_write = sbi->s_mb_stream_request; - } if (wbc->range_start == 0 && wbc->range_end == LLONG_MAX) range_whole = 1; @@ -2808,6 +2842,36 @@ static int ext4_da_writepages(struct add } else index = wbc->range_start >> PAGE_CACHE_SHIFT; + /* + * This works around two forms of stupidity. The first is in + * the writeback code, which caps the maximum number of pages + * written to be 1024 pages. This is wrong on multiple + * levels; different architectues have a different page size, + * which changes the maximum amount of data which gets + * written. Secondly, 4 megabytes is way too small. XFS + * forces this value to be 16 megabytes by multiplying + * nr_to_write parameter by four, and then relies on its + * allocator to allocate larger extents to make them + * contiguous. Unfortunately this brings us to the second + * stupidity, which is that ext4's mballoc code only allocates + * at most 2048 blocks. So we force contiguous writes up to + * the number of dirty blocks in the inode, or + * sbi->max_writeback_mb_bump whichever is smaller. + */ + max_pages = sbi->s_max_writeback_mb_bump << (20 - PAGE_CACHE_SHIFT); + if (!range_cyclic && range_whole) + desired_nr_to_write = wbc->nr_to_write * 8; + else + desired_nr_to_write = ext4_num_dirty_pages(inode, index, + max_pages); + if (desired_nr_to_write > max_pages) + desired_nr_to_write = max_pages; + + if (wbc->nr_to_write < desired_nr_to_write) { + nr_to_writebump = desired_nr_to_write - wbc->nr_to_write; + wbc->nr_to_write = desired_nr_to_write; + } + mpd.wbc = wbc; mpd.inode = mapping->host; @@ -2926,7 +2990,8 @@ retry: out_writepages: if (!no_nrwrite_index_update) wbc->no_nrwrite_index_update = 0; - wbc->nr_to_write -= nr_to_writebump; + if (wbc->nr_to_write > nr_to_writebump) + wbc->nr_to_write -= nr_to_writebump; wbc->range_start = range_start; trace_ext4_da_writepages_result(inode, wbc, ret, pages_written); return ret; --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -2199,6 +2199,7 @@ EXT4_RW_ATTR_SBI_UI(mb_min_to_scan, s_mb EXT4_RW_ATTR_SBI_UI(mb_order2_req, s_mb_order2_reqs); EXT4_RW_ATTR_SBI_UI(mb_stream_req, s_mb_stream_request); EXT4_RW_ATTR_SBI_UI(mb_group_prealloc, s_mb_group_prealloc); +EXT4_RW_ATTR_SBI_UI(max_writeback_mb_bump, s_max_writeback_mb_bump); static struct attribute *ext4_attrs[] = { ATTR_LIST(delayed_allocation_blocks), @@ -2212,6 +2213,7 @@ static struct attribute *ext4_attrs[] = ATTR_LIST(mb_order2_req), ATTR_LIST(mb_stream_req), ATTR_LIST(mb_group_prealloc), + ATTR_LIST(max_writeback_mb_bump), NULL, }; @@ -2681,6 +2683,7 @@ static int ext4_fill_super(struct super_ } sbi->s_stripe = ext4_get_stripe_size(sbi); + sbi->s_max_writeback_mb_bump = 128; /* * set up enough so that it can read an inode --- a/include/trace/events/ext4.h +++ b/include/trace/events/ext4.h @@ -231,6 +231,7 @@ TRACE_EVENT(ext4_da_writepages, __field( char, for_reclaim ) __field( char, for_writepages ) __field( char, range_cyclic ) + __field( pgoff_t, writeback_index ) ), TP_fast_assign( @@ -245,14 +246,51 @@ TRACE_EVENT(ext4_da_writepages, __entry->for_reclaim = wbc->for_reclaim; __entry->for_writepages = wbc->for_writepages; __entry->range_cyclic = wbc->range_cyclic; + __entry->writeback_index = inode->i_mapping->writeback_index; ), - TP_printk("dev %s ino %lu nr_t_write %ld pages_skipped %ld range_start %llu range_end %llu nonblocking %d for_kupdate %d for_reclaim %d for_writepages %d range_cyclic %d", - jbd2_dev_to_name(__entry->dev), __entry->ino, __entry->nr_to_write, + TP_printk("dev %s ino %lu nr_to_write %ld pages_skipped %ld range_start %llu range_end %llu nonblocking %d for_kupdate %d for_reclaim %d for_writepages %d range_cyclic %d writeback_index %lu", + jbd2_dev_to_name(__entry->dev), + (unsigned long) __entry->ino, __entry->nr_to_write, __entry->pages_skipped, __entry->range_start, __entry->range_end, __entry->nonblocking, __entry->for_kupdate, __entry->for_reclaim, - __entry->for_writepages, __entry->range_cyclic) + __entry->for_writepages, __entry->range_cyclic, + (unsigned long) __entry->writeback_index) +); + +TRACE_EVENT(ext4_da_write_pages, + TP_PROTO(struct inode *inode, struct mpage_da_data *mpd), + + TP_ARGS(inode, mpd), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( __u64, b_blocknr ) + __field( __u32, b_size ) + __field( __u32, b_state ) + __field( unsigned long, first_page ) + __field( int, io_done ) + __field( int, pages_written ) + ), + + TP_fast_assign( + __entry->dev = inode->i_sb->s_dev; + __entry->ino = inode->i_ino; + __entry->b_blocknr = mpd->b_blocknr; + __entry->b_size = mpd->b_size; + __entry->b_state = mpd->b_state; + __entry->first_page = mpd->first_page; + __entry->io_done = mpd->io_done; + __entry->pages_written = mpd->pages_written; + ), + + TP_printk("dev %s ino %lu b_blocknr %llu b_size %u b_state 0x%04x first_page %lu io_done %d pages_written %d", + jbd2_dev_to_name(__entry->dev), (unsigned long) __entry->ino, + __entry->b_blocknr, __entry->b_size, + __entry->b_state, __entry->first_page, + __entry->io_done, __entry->pages_written) ); TRACE_EVENT(ext4_da_writepages_result, @@ -270,6 +308,7 @@ TRACE_EVENT(ext4_da_writepages_result, __field( char, encountered_congestion ) __field( char, more_io ) __field( char, no_nrwrite_index_update ) + __field( pgoff_t, writeback_index ) ), TP_fast_assign( @@ -281,13 +320,16 @@ TRACE_EVENT(ext4_da_writepages_result, __entry->encountered_congestion = wbc->encountered_congestion; __entry->more_io = wbc->more_io; __entry->no_nrwrite_index_update = wbc->no_nrwrite_index_update; + __entry->writeback_index = inode->i_mapping->writeback_index; ), - TP_printk("dev %s ino %lu ret %d pages_written %d pages_skipped %ld congestion %d more_io %d no_nrwrite_index_update %d", - jbd2_dev_to_name(__entry->dev), __entry->ino, __entry->ret, + TP_printk("dev %s ino %lu ret %d pages_written %d pages_skipped %ld congestion %d more_io %d no_nrwrite_index_update %d writeback_index %lu", + jbd2_dev_to_name(__entry->dev), + (unsigned long) __entry->ino, __entry->ret, __entry->pages_written, __entry->pages_skipped, __entry->encountered_congestion, __entry->more_io, - __entry->no_nrwrite_index_update) + __entry->no_nrwrite_index_update, + (unsigned long) __entry->writeback_index) ); TRACE_EVENT(ext4_da_write_begin, -- 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/