From f73bee49855fe968e87af18e44f2d17e1e675e74 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Wed, 18 Aug 2010 15:56:56 +0200 Subject: jbd2: Modify ASYNC_COMMIT code to not rely on queue draining on barrier Currently JBD2 relies blkdev_issue_flush() draining the queue when ASYNC_COMMIT feature is set. This property is going away so make JBD2 wait for buffers it needs on its own before submitting the cache flush. Signed-off-by: Jan Kara Signed-off-by: Tejun Heo Signed-off-by: Jens Axboe --- fs/jbd2/commit.c | 29 ++++++++++++++++------------- 1 file changed, 16 insertions(+), 13 deletions(-) (limited to 'fs/jbd2/commit.c') diff --git a/fs/jbd2/commit.c b/fs/jbd2/commit.c index 7c068c189d8..d6aeb1f6cfe 100644 --- a/fs/jbd2/commit.c +++ b/fs/jbd2/commit.c @@ -701,6 +701,16 @@ start_journal_io: } } + err = journal_finish_inode_data_buffers(journal, commit_transaction); + if (err) { + printk(KERN_WARNING + "JBD2: Detected IO errors while flushing file data " + "on %s\n", journal->j_devname); + if (journal->j_flags & JBD2_ABORT_ON_SYNCDATA_ERR) + jbd2_journal_abort(journal, err); + err = 0; + } + /* * If the journal is not located on the file system device, * then we must flush the file system device before we issue @@ -719,19 +729,6 @@ start_journal_io: &cbh, crc32_sum); if (err) __jbd2_journal_abort_hard(journal); - if (journal->j_flags & JBD2_BARRIER) - blkdev_issue_flush(journal->j_dev, GFP_KERNEL, NULL, - BLKDEV_IFL_WAIT); - } - - err = journal_finish_inode_data_buffers(journal, commit_transaction); - if (err) { - printk(KERN_WARNING - "JBD2: Detected IO errors while flushing file data " - "on %s\n", journal->j_devname); - if (journal->j_flags & JBD2_ABORT_ON_SYNCDATA_ERR) - jbd2_journal_abort(journal, err); - err = 0; } /* Lo and behold: we have just managed to send a transaction to @@ -845,6 +842,12 @@ wait_for_iobuf: } if (!err && !is_journal_aborted(journal)) err = journal_wait_on_commit_record(journal, cbh); + if (JBD2_HAS_INCOMPAT_FEATURE(journal, + JBD2_FEATURE_INCOMPAT_ASYNC_COMMIT) && + journal->j_flags & JBD2_BARRIER) { + blkdev_issue_flush(journal->j_dev, GFP_KERNEL, NULL, + BLKDEV_IFL_WAIT); + } if (err) jbd2_journal_abort(journal, err); -- cgit v1.2.3 From 9c35575bbe6b1dd4914a5323c8df8b3159edcc75 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Wed, 18 Aug 2010 05:29:17 -0400 Subject: jbd2: replace barriers with explicit flush / FUA usage Switch to the WRITE_FLUSH_FUA flag for journal commits and remove the EOPNOTSUPP detection for barriers. Signed-off-by: Christoph Hellwig Acked-by: Jan Kara Signed-off-by: Tejun Heo Signed-off-by: Jens Axboe --- fs/jbd2/commit.c | 43 ++++--------------------------------------- 1 file changed, 4 insertions(+), 39 deletions(-) (limited to 'fs/jbd2/commit.c') diff --git a/fs/jbd2/commit.c b/fs/jbd2/commit.c index d6aeb1f6cfe..f204e27f44d 100644 --- a/fs/jbd2/commit.c +++ b/fs/jbd2/commit.c @@ -134,25 +134,11 @@ static int journal_submit_commit_record(journal_t *journal, if (journal->j_flags & JBD2_BARRIER && !JBD2_HAS_INCOMPAT_FEATURE(journal, - JBD2_FEATURE_INCOMPAT_ASYNC_COMMIT)) { - ret = submit_bh(WRITE_SYNC_PLUG | WRITE_BARRIER, bh); - if (ret == -EOPNOTSUPP) { - printk(KERN_WARNING - "JBD2: Disabling barriers on %s, " - "not supported by device\n", journal->j_devname); - write_lock(&journal->j_state_lock); - journal->j_flags &= ~JBD2_BARRIER; - write_unlock(&journal->j_state_lock); - - /* And try again, without the barrier */ - lock_buffer(bh); - set_buffer_uptodate(bh); - clear_buffer_dirty(bh); - ret = submit_bh(WRITE_SYNC_PLUG, bh); - } - } else { + JBD2_FEATURE_INCOMPAT_ASYNC_COMMIT)) + ret = submit_bh(WRITE_SYNC_PLUG | WRITE_FLUSH_FUA, bh); + else ret = submit_bh(WRITE_SYNC_PLUG, bh); - } + *cbh = bh; return ret; } @@ -166,29 +152,8 @@ static int journal_wait_on_commit_record(journal_t *journal, { int ret = 0; -retry: clear_buffer_dirty(bh); wait_on_buffer(bh); - if (buffer_eopnotsupp(bh) && (journal->j_flags & JBD2_BARRIER)) { - printk(KERN_WARNING - "JBD2: %s: disabling barries on %s - not supported " - "by device\n", __func__, journal->j_devname); - write_lock(&journal->j_state_lock); - journal->j_flags &= ~JBD2_BARRIER; - write_unlock(&journal->j_state_lock); - - lock_buffer(bh); - clear_buffer_dirty(bh); - set_buffer_uptodate(bh); - bh->b_end_io = journal_end_buffer_io_sync; - - ret = submit_bh(WRITE_SYNC_PLUG, bh); - if (ret) { - unlock_buffer(bh); - return ret; - } - goto retry; - } if (unlikely(!buffer_uptodate(bh))) ret = -EIO; -- cgit v1.2.3 From dd3932eddf428571762596e17b65f5dc92ca361b Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Thu, 16 Sep 2010 20:51:46 +0200 Subject: block: remove BLKDEV_IFL_WAIT All the blkdev_issue_* helpers can only sanely be used for synchronous caller. To issue cache flushes or barriers asynchronously the caller needs to set up a bio by itself with a completion callback to move the asynchronous state machine ahead. So drop the BLKDEV_IFL_WAIT flag that is always specified when calling blkdev_issue_* and also remove the now unused flags argument to blkdev_issue_flush and blkdev_issue_zeroout. For blkdev_issue_discard we need to keep it for the secure discard flag, which gains a more descriptive name and loses the bitops vs flag confusion. Signed-off-by: Christoph Hellwig Signed-off-by: Jens Axboe --- block/blk-flush.c | 25 +++++++++++-------------- block/blk-lib.c | 21 ++++++++------------- block/ioctl.c | 4 ++-- drivers/block/drbd/drbd_int.h | 3 +-- drivers/block/drbd/drbd_receiver.c | 2 +- fs/block_dev.c | 2 +- fs/btrfs/extent-tree.c | 3 +-- fs/ext3/fsync.c | 3 +-- fs/ext4/fsync.c | 5 ++--- fs/ext4/mballoc.c | 3 +-- fs/fat/fatent.c | 3 +-- fs/gfs2/rgrp.c | 5 ++--- fs/jbd2/checkpoint.c | 3 +-- fs/jbd2/commit.c | 6 ++---- fs/nilfs2/the_nilfs.c | 4 ++-- fs/reiserfs/file.c | 3 +-- fs/xfs/linux-2.6/xfs_super.c | 3 +-- include/linux/blkdev.h | 14 +++++--------- mm/swapfile.c | 6 +++--- 19 files changed, 47 insertions(+), 71 deletions(-) (limited to 'fs/jbd2/commit.c') diff --git a/block/blk-flush.c b/block/blk-flush.c index 62b7df9bca9..54b123d6563 100644 --- a/block/blk-flush.c +++ b/block/blk-flush.c @@ -205,7 +205,6 @@ static void bio_end_flush(struct bio *bio, int err) * @bdev: blockdev to issue flush for * @gfp_mask: memory allocation flags (for bio_alloc) * @error_sector: error sector - * @flags: BLKDEV_IFL_* flags to control behaviour * * Description: * Issue a flush for the block device in question. Caller can supply @@ -214,7 +213,7 @@ static void bio_end_flush(struct bio *bio, int err) * request was pushed in some internal queue for later handling. */ int blkdev_issue_flush(struct block_device *bdev, gfp_t gfp_mask, - sector_t *error_sector, unsigned long flags) + sector_t *error_sector) { DECLARE_COMPLETION_ONSTACK(wait); struct request_queue *q; @@ -240,21 +239,19 @@ int blkdev_issue_flush(struct block_device *bdev, gfp_t gfp_mask, bio = bio_alloc(gfp_mask, 0); bio->bi_end_io = bio_end_flush; bio->bi_bdev = bdev; - if (test_bit(BLKDEV_WAIT, &flags)) - bio->bi_private = &wait; + bio->bi_private = &wait; bio_get(bio); submit_bio(WRITE_FLUSH, bio); - if (test_bit(BLKDEV_WAIT, &flags)) { - wait_for_completion(&wait); - /* - * The driver must store the error location in ->bi_sector, if - * it supports it. For non-stacked drivers, this should be - * copied from blk_rq_pos(rq). - */ - if (error_sector) - *error_sector = bio->bi_sector; - } + wait_for_completion(&wait); + + /* + * The driver must store the error location in ->bi_sector, if + * it supports it. For non-stacked drivers, this should be + * copied from blk_rq_pos(rq). + */ + if (error_sector) + *error_sector = bio->bi_sector; if (!bio_flagged(bio, BIO_UPTODATE)) ret = -EIO; diff --git a/block/blk-lib.c b/block/blk-lib.c index fe2e6ed0f51..1a320d2406b 100644 --- a/block/blk-lib.c +++ b/block/blk-lib.c @@ -61,7 +61,7 @@ int blkdev_issue_discard(struct block_device *bdev, sector_t sector, max_discard_sectors &= ~(disc_sects - 1); } - if (flags & BLKDEV_IFL_SECURE) { + if (flags & BLKDEV_DISCARD_SECURE) { if (!blk_queue_secdiscard(q)) return -EOPNOTSUPP; type |= REQ_SECURE; @@ -77,8 +77,7 @@ int blkdev_issue_discard(struct block_device *bdev, sector_t sector, bio->bi_sector = sector; bio->bi_end_io = blkdev_discard_end_io; bio->bi_bdev = bdev; - if (flags & BLKDEV_IFL_WAIT) - bio->bi_private = &wait; + bio->bi_private = &wait; if (nr_sects > max_discard_sectors) { bio->bi_size = max_discard_sectors << 9; @@ -92,8 +91,7 @@ int blkdev_issue_discard(struct block_device *bdev, sector_t sector, bio_get(bio); submit_bio(type, bio); - if (flags & BLKDEV_IFL_WAIT) - wait_for_completion(&wait); + wait_for_completion(&wait); if (bio_flagged(bio, BIO_EOPNOTSUPP)) ret = -EOPNOTSUPP; @@ -139,7 +137,6 @@ static void bio_batch_end_io(struct bio *bio, int err) * @sector: start sector * @nr_sects: number of sectors to write * @gfp_mask: memory allocation flags (for bio_alloc) - * @flags: BLKDEV_IFL_* flags to control behaviour * * Description: * Generate and issue number of bios with zerofiled pages. @@ -148,7 +145,7 @@ static void bio_batch_end_io(struct bio *bio, int err) */ int blkdev_issue_zeroout(struct block_device *bdev, sector_t sector, - sector_t nr_sects, gfp_t gfp_mask, unsigned long flags) + sector_t nr_sects, gfp_t gfp_mask) { int ret; struct bio *bio; @@ -174,8 +171,7 @@ submit: bio->bi_sector = sector; bio->bi_bdev = bdev; bio->bi_end_io = bio_batch_end_io; - if (flags & BLKDEV_IFL_WAIT) - bio->bi_private = &bb; + bio->bi_private = &bb; while (nr_sects != 0) { sz = min((sector_t) PAGE_SIZE >> 9 , nr_sects); @@ -193,10 +189,9 @@ submit: submit_bio(WRITE, bio); } - if (flags & BLKDEV_IFL_WAIT) - /* Wait for bios in-flight */ - while ( issued != atomic_read(&bb.done)) - wait_for_completion(&wait); + /* Wait for bios in-flight */ + while (issued != atomic_read(&bb.done)) + wait_for_completion(&wait); if (!test_bit(BIO_UPTODATE, &bb.flags)) /* One of bios in the batch was completed with error.*/ diff --git a/block/ioctl.c b/block/ioctl.c index d8052f0dabd..cb2b9099862 100644 --- a/block/ioctl.c +++ b/block/ioctl.c @@ -116,7 +116,7 @@ static int blkdev_reread_part(struct block_device *bdev) static int blk_ioctl_discard(struct block_device *bdev, uint64_t start, uint64_t len, int secure) { - unsigned long flags = BLKDEV_IFL_WAIT; + unsigned long flags = 0; if (start & 511) return -EINVAL; @@ -128,7 +128,7 @@ static int blk_ioctl_discard(struct block_device *bdev, uint64_t start, if (start + len > (bdev->bd_inode->i_size >> 9)) return -EINVAL; if (secure) - flags |= BLKDEV_IFL_SECURE; + flags |= BLKDEV_DISCARD_SECURE; return blkdev_issue_discard(bdev, start, len, GFP_KERNEL, flags); } diff --git a/drivers/block/drbd/drbd_int.h b/drivers/block/drbd/drbd_int.h index 352441b0f92..c2ef476f571 100644 --- a/drivers/block/drbd/drbd_int.h +++ b/drivers/block/drbd/drbd_int.h @@ -2321,8 +2321,7 @@ static inline void drbd_md_flush(struct drbd_conf *mdev) if (test_bit(MD_NO_BARRIER, &mdev->flags)) return; - r = blkdev_issue_flush(mdev->ldev->md_bdev, GFP_KERNEL, NULL, - BLKDEV_IFL_WAIT); + r = blkdev_issue_flush(mdev->ldev->md_bdev, GFP_KERNEL, NULL); if (r) { set_bit(MD_NO_BARRIER, &mdev->flags); dev_err(DEV, "meta data flush failed with status %d, disabling md-flushes\n", r); diff --git a/drivers/block/drbd/drbd_receiver.c b/drivers/block/drbd/drbd_receiver.c index 081522d3c74..df15e7f0e7b 100644 --- a/drivers/block/drbd/drbd_receiver.c +++ b/drivers/block/drbd/drbd_receiver.c @@ -975,7 +975,7 @@ static enum finish_epoch drbd_flush_after_epoch(struct drbd_conf *mdev, struct d if (mdev->write_ordering >= WO_bdev_flush && get_ldev(mdev)) { rv = blkdev_issue_flush(mdev->ldev->backing_bdev, GFP_KERNEL, - NULL, BLKDEV_IFL_WAIT); + NULL); if (rv) { dev_err(DEV, "local disk flush failed with status %d\n", rv); /* would rather check on EOPNOTSUPP, but that is not reliable. diff --git a/fs/block_dev.c b/fs/block_dev.c index 50e8c8582fa..b737451e2e9 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -370,7 +370,7 @@ int blkdev_fsync(struct file *filp, int datasync) */ mutex_unlock(&bd_inode->i_mutex); - error = blkdev_issue_flush(bdev, GFP_KERNEL, NULL, BLKDEV_IFL_WAIT); + error = blkdev_issue_flush(bdev, GFP_KERNEL, NULL); if (error == -EOPNOTSUPP) error = 0; diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 43dc9ea9aef..0b81ecdb101 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -1695,8 +1695,7 @@ static int remove_extent_backref(struct btrfs_trans_handle *trans, static void btrfs_issue_discard(struct block_device *bdev, u64 start, u64 len) { - blkdev_issue_discard(bdev, start >> 9, len >> 9, GFP_KERNEL, - BLKDEV_IFL_WAIT); + blkdev_issue_discard(bdev, start >> 9, len >> 9, GFP_KERNEL, 0); } static int btrfs_discard_extent(struct btrfs_root *root, u64 bytenr, diff --git a/fs/ext3/fsync.c b/fs/ext3/fsync.c index d7e9f74dc3a..09b13bb34c9 100644 --- a/fs/ext3/fsync.c +++ b/fs/ext3/fsync.c @@ -90,7 +90,6 @@ int ext3_sync_file(struct file *file, int datasync) * storage */ if (needs_barrier) - blkdev_issue_flush(inode->i_sb->s_bdev, GFP_KERNEL, NULL, - BLKDEV_IFL_WAIT); + blkdev_issue_flush(inode->i_sb->s_bdev, GFP_KERNEL, NULL); return ret; } diff --git a/fs/ext4/fsync.c b/fs/ext4/fsync.c index 592adf2e546..3f3ff5ee8f9 100644 --- a/fs/ext4/fsync.c +++ b/fs/ext4/fsync.c @@ -128,10 +128,9 @@ int ext4_sync_file(struct file *file, int datasync) (journal->j_fs_dev != journal->j_dev) && (journal->j_flags & JBD2_BARRIER)) blkdev_issue_flush(inode->i_sb->s_bdev, GFP_KERNEL, - NULL, BLKDEV_IFL_WAIT); + NULL); ret = jbd2_log_wait_commit(journal, commit_tid); } else if (journal->j_flags & JBD2_BARRIER) - blkdev_issue_flush(inode->i_sb->s_bdev, GFP_KERNEL, NULL, - BLKDEV_IFL_WAIT); + blkdev_issue_flush(inode->i_sb->s_bdev, GFP_KERNEL, NULL); return ret; } diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index a22bfef3da9..19aa0d44d82 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -2566,8 +2566,7 @@ static inline void ext4_issue_discard(struct super_block *sb, discard_block = block + ext4_group_first_block_no(sb, block_group); trace_ext4_discard_blocks(sb, (unsigned long long) discard_block, count); - ret = sb_issue_discard(sb, discard_block, count, GFP_NOFS, - BLKDEV_IFL_WAIT); + ret = sb_issue_discard(sb, discard_block, count, GFP_NOFS, 0); if (ret == EOPNOTSUPP) { ext4_warning(sb, "discard not supported, disabling"); clear_opt(EXT4_SB(sb)->s_mount_opt, DISCARD); diff --git a/fs/fat/fatent.c b/fs/fat/fatent.c index f9a0b7ae864..b47d2c9f4fa 100644 --- a/fs/fat/fatent.c +++ b/fs/fat/fatent.c @@ -578,8 +578,7 @@ int fat_free_clusters(struct inode *inode, int cluster) sb_issue_discard(sb, fat_clus_to_blknr(sbi, first_cl), nr_clus * sbi->sec_per_clus, - GFP_NOFS, - BLKDEV_IFL_WAIT); + GFP_NOFS, 0); first_cl = cluster; } diff --git a/fs/gfs2/rgrp.c b/fs/gfs2/rgrp.c index 37931647291..38b3ea1abac 100644 --- a/fs/gfs2/rgrp.c +++ b/fs/gfs2/rgrp.c @@ -854,7 +854,7 @@ static void gfs2_rgrp_send_discards(struct gfs2_sbd *sdp, u64 offset, if ((start + nr_sects) != blk) { rv = blkdev_issue_discard(bdev, start, nr_sects, GFP_NOFS, - BLKDEV_IFL_WAIT); + 0); if (rv) goto fail; nr_sects = 0; @@ -868,8 +868,7 @@ start_new_extent: } } if (nr_sects) { - rv = blkdev_issue_discard(bdev, start, nr_sects, GFP_NOFS, - BLKDEV_IFL_WAIT); + rv = blkdev_issue_discard(bdev, start, nr_sects, GFP_NOFS, 0); if (rv) goto fail; } diff --git a/fs/jbd2/checkpoint.c b/fs/jbd2/checkpoint.c index 5247e7ffdcb..6571a056e55 100644 --- a/fs/jbd2/checkpoint.c +++ b/fs/jbd2/checkpoint.c @@ -532,8 +532,7 @@ int jbd2_cleanup_journal_tail(journal_t *journal) */ if ((journal->j_fs_dev != journal->j_dev) && (journal->j_flags & JBD2_BARRIER)) - blkdev_issue_flush(journal->j_fs_dev, GFP_KERNEL, NULL, - BLKDEV_IFL_WAIT); + blkdev_issue_flush(journal->j_fs_dev, GFP_KERNEL, NULL); if (!(journal->j_flags & JBD2_ABORT)) jbd2_journal_update_superblock(journal, 1); return 0; diff --git a/fs/jbd2/commit.c b/fs/jbd2/commit.c index f204e27f44d..cb43c605cfa 100644 --- a/fs/jbd2/commit.c +++ b/fs/jbd2/commit.c @@ -684,8 +684,7 @@ start_journal_io: if (commit_transaction->t_flushed_data_blocks && (journal->j_fs_dev != journal->j_dev) && (journal->j_flags & JBD2_BARRIER)) - blkdev_issue_flush(journal->j_fs_dev, GFP_KERNEL, NULL, - BLKDEV_IFL_WAIT); + blkdev_issue_flush(journal->j_fs_dev, GFP_KERNEL, NULL); /* Done it all: now write the commit record asynchronously. */ if (JBD2_HAS_INCOMPAT_FEATURE(journal, @@ -810,8 +809,7 @@ wait_for_iobuf: if (JBD2_HAS_INCOMPAT_FEATURE(journal, JBD2_FEATURE_INCOMPAT_ASYNC_COMMIT) && journal->j_flags & JBD2_BARRIER) { - blkdev_issue_flush(journal->j_dev, GFP_KERNEL, NULL, - BLKDEV_IFL_WAIT); + blkdev_issue_flush(journal->j_dev, GFP_KERNEL, NULL); } if (err) diff --git a/fs/nilfs2/the_nilfs.c b/fs/nilfs2/the_nilfs.c index 400b2caef4d..d97310f07be 100644 --- a/fs/nilfs2/the_nilfs.c +++ b/fs/nilfs2/the_nilfs.c @@ -774,7 +774,7 @@ int nilfs_discard_segments(struct the_nilfs *nilfs, __u64 *segnump, ret = blkdev_issue_discard(nilfs->ns_bdev, start * sects_per_block, nblocks * sects_per_block, - GFP_NOFS, BLKDEV_IFL_WAIT); + GFP_NOFS, 0); if (ret < 0) return ret; nblocks = 0; @@ -784,7 +784,7 @@ int nilfs_discard_segments(struct the_nilfs *nilfs, __u64 *segnump, ret = blkdev_issue_discard(nilfs->ns_bdev, start * sects_per_block, nblocks * sects_per_block, - GFP_NOFS, BLKDEV_IFL_WAIT); + GFP_NOFS, 0); return ret; } diff --git a/fs/reiserfs/file.c b/fs/reiserfs/file.c index 6846371498b..91f080cc76c 100644 --- a/fs/reiserfs/file.c +++ b/fs/reiserfs/file.c @@ -152,8 +152,7 @@ static int reiserfs_sync_file(struct file *filp, int datasync) barrier_done = reiserfs_commit_for_inode(inode); reiserfs_write_unlock(inode->i_sb); if (barrier_done != 1 && reiserfs_barrier_flush(inode->i_sb)) - blkdev_issue_flush(inode->i_sb->s_bdev, GFP_KERNEL, NULL, - BLKDEV_IFL_WAIT); + blkdev_issue_flush(inode->i_sb->s_bdev, GFP_KERNEL, NULL); if (barrier_done < 0) return barrier_done; return (err < 0) ? -EIO : 0; diff --git a/fs/xfs/linux-2.6/xfs_super.c b/fs/xfs/linux-2.6/xfs_super.c index 15c35b62ff1..5fa7a30cc3f 100644 --- a/fs/xfs/linux-2.6/xfs_super.c +++ b/fs/xfs/linux-2.6/xfs_super.c @@ -693,8 +693,7 @@ void xfs_blkdev_issue_flush( xfs_buftarg_t *buftarg) { - blkdev_issue_flush(buftarg->bt_bdev, GFP_KERNEL, NULL, - BLKDEV_IFL_WAIT); + blkdev_issue_flush(buftarg->bt_bdev, GFP_KERNEL, NULL); } STATIC void diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index cfcb3a61060..accbd0e5c89 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -867,18 +867,14 @@ static inline struct request *blk_map_queue_find_tag(struct blk_queue_tag *bqt, return NULL; return bqt->tag_index[tag]; } -enum{ - BLKDEV_WAIT, /* wait for completion */ - BLKDEV_SECURE, /* secure discard */ -}; -#define BLKDEV_IFL_WAIT (1 << BLKDEV_WAIT) -#define BLKDEV_IFL_SECURE (1 << BLKDEV_SECURE) -extern int blkdev_issue_flush(struct block_device *, gfp_t, sector_t *, - unsigned long); + +#define BLKDEV_DISCARD_SECURE 0x01 /* secure discard */ + +extern int blkdev_issue_flush(struct block_device *, gfp_t, sector_t *); extern int blkdev_issue_discard(struct block_device *bdev, sector_t sector, sector_t nr_sects, gfp_t gfp_mask, unsigned long flags); extern int blkdev_issue_zeroout(struct block_device *bdev, sector_t sector, - sector_t nr_sects, gfp_t gfp_mask, unsigned long flags); + sector_t nr_sects, gfp_t gfp_mask); static inline int sb_issue_discard(struct super_block *sb, sector_t block, sector_t nr_blocks, gfp_t gfp_mask, unsigned long flags) { diff --git a/mm/swapfile.c b/mm/swapfile.c index 68cda164dff..e132e1708ac 100644 --- a/mm/swapfile.c +++ b/mm/swapfile.c @@ -141,7 +141,7 @@ static int discard_swap(struct swap_info_struct *si) nr_blocks = ((sector_t)se->nr_pages - 1) << (PAGE_SHIFT - 9); if (nr_blocks) { err = blkdev_issue_discard(si->bdev, start_block, - nr_blocks, GFP_KERNEL, BLKDEV_IFL_WAIT); + nr_blocks, GFP_KERNEL, 0); if (err) return err; cond_resched(); @@ -152,7 +152,7 @@ static int discard_swap(struct swap_info_struct *si) nr_blocks = (sector_t)se->nr_pages << (PAGE_SHIFT - 9); err = blkdev_issue_discard(si->bdev, start_block, - nr_blocks, GFP_KERNEL, BLKDEV_IFL_WAIT); + nr_blocks, GFP_KERNEL, 0); if (err) break; @@ -191,7 +191,7 @@ static void discard_swap_cluster(struct swap_info_struct *si, start_block <<= PAGE_SHIFT - 9; nr_blocks <<= PAGE_SHIFT - 9; if (blkdev_issue_discard(si->bdev, start_block, - nr_blocks, GFP_NOIO, BLKDEV_IFL_WAIT)) + nr_blocks, GFP_NOIO, 0)) break; } -- cgit v1.2.3 From 749ef9f8423054e326f3a246327ed2db4b6d395f Mon Sep 17 00:00:00 2001 From: Corrado Zoccolo Date: Mon, 20 Sep 2010 15:24:50 +0200 Subject: cfq: improve fsync performance for small files Fsync performance for small files achieved by cfq on high-end disks is lower than what deadline can achieve, due to idling introduced between the sync write happening in process context and the journal commit. Moreover, when competing with a sequential reader, a process writing small files and fsync-ing them is starved. This patch fixes the two problems by: - marking journal commits as WRITE_SYNC, so that they get the REQ_NOIDLE flag set, - force all queues that have REQ_NOIDLE requests to be put in the noidle tree. Having the queue associated to the fsync-ing process and the one associated to journal commits in the noidle tree allows: - switching between them without idling, - fairness vs. competing idling queues, since they will be serviced only after the noidle tree expires its slice. Acked-by: Vivek Goyal Reviewed-by: Jeff Moyer Tested-by: Jeff Moyer Signed-off-by: Corrado Zoccolo Signed-off-by: Jens Axboe --- block/cfq-iosched.c | 18 ++++-------------- fs/jbd/commit.c | 2 +- fs/jbd2/commit.c | 2 +- 3 files changed, 6 insertions(+), 16 deletions(-) (limited to 'fs/jbd2/commit.c') diff --git a/block/cfq-iosched.c b/block/cfq-iosched.c index b9f86190763..68459262173 100644 --- a/block/cfq-iosched.c +++ b/block/cfq-iosched.c @@ -216,7 +216,6 @@ struct cfq_data { enum wl_type_t serving_type; unsigned long workload_expires; struct cfq_group *serving_group; - bool noidle_tree_requires_idle; /* * Each priority tree is sorted by next_request position. These @@ -2126,7 +2125,6 @@ static void choose_service_tree(struct cfq_data *cfqd, struct cfq_group *cfqg) slice = max_t(unsigned, slice, CFQ_MIN_TT); cfq_log(cfqd, "workload slice:%d", slice); cfqd->workload_expires = jiffies + slice; - cfqd->noidle_tree_requires_idle = false; } static struct cfq_group *cfq_get_next_cfqg(struct cfq_data *cfqd) @@ -3108,7 +3106,9 @@ cfq_update_idle_window(struct cfq_data *cfqd, struct cfq_queue *cfqq, if (cfqq->queued[0] + cfqq->queued[1] >= 4) cfq_mark_cfqq_deep(cfqq); - if (!atomic_read(&cic->ioc->nr_tasks) || !cfqd->cfq_slice_idle || + if (cfqq->next_rq && (cfqq->next_rq->cmd_flags & REQ_NOIDLE)) + enable_idle = 0; + else if (!atomic_read(&cic->ioc->nr_tasks) || !cfqd->cfq_slice_idle || (!cfq_cfqq_deep(cfqq) && CFQQ_SEEKY(cfqq))) enable_idle = 0; else if (sample_valid(cic->ttime_samples)) { @@ -3421,17 +3421,7 @@ static void cfq_completed_request(struct request_queue *q, struct request *rq) cfq_slice_expired(cfqd, 1); else if (sync && cfqq_empty && !cfq_close_cooperator(cfqd, cfqq)) { - cfqd->noidle_tree_requires_idle |= - !(rq->cmd_flags & REQ_NOIDLE); - /* - * Idling is enabled for SYNC_WORKLOAD. - * SYNC_NOIDLE_WORKLOAD idles at the end of the tree - * only if we processed at least one !REQ_NOIDLE request - */ - if (cfqd->serving_type == SYNC_WORKLOAD - || cfqd->noidle_tree_requires_idle - || cfqq->cfqg->nr_cfqq == 1) - cfq_arm_slice_timer(cfqd); + cfq_arm_slice_timer(cfqd); } } diff --git a/fs/jbd/commit.c b/fs/jbd/commit.c index 95d8c11c929..3f030e9efea 100644 --- a/fs/jbd/commit.c +++ b/fs/jbd/commit.c @@ -318,7 +318,7 @@ void journal_commit_transaction(journal_t *journal) int first_tag = 0; int tag_flag; int i; - int write_op = WRITE; + int write_op = WRITE_SYNC; /* * First job: lock down the current transaction and wait for diff --git a/fs/jbd2/commit.c b/fs/jbd2/commit.c index 7c068c189d8..80910f51d4b 100644 --- a/fs/jbd2/commit.c +++ b/fs/jbd2/commit.c @@ -360,7 +360,7 @@ void jbd2_journal_commit_transaction(journal_t *journal) int tag_bytes = journal_tag_bytes(journal); struct buffer_head *cbh = NULL; /* For transactional checksums */ __u32 crc32_sum = ~0; - int write_op = WRITE; + int write_op = WRITE_SYNC; /* * First job: lock down the current transaction and wait for -- cgit v1.2.3 From 39e3ac2599a5f9aba499b5f8af809108e70a6163 Mon Sep 17 00:00:00 2001 From: Brian King Date: Wed, 27 Oct 2010 21:25:12 -0400 Subject: jbd2: Fix I/O hang in jbd2_journal_release_jbd_inode This fixes a hang seen in jbd2_journal_release_jbd_inode on a lot of Power 6 systems running with ext4. When we get in the hung state, all I/O to the disk in question gets blocked where we stay indefinitely. Looking at the task list, I can see we are stuck in jbd2_journal_release_jbd_inode waiting on a wake up. I added some debug code to detect this scenario and dump additional data if we were stuck in jbd2_journal_release_jbd_inode for longer than 30 minutes. When it hit, I was able to see that i_flags was 0, suggesting we missed the wake up. This patch changes i_flags to be an unsigned long, uses bit operators to access it, and adds barriers around the accesses. Prior to applying this patch, we were regularly hitting this hang on numerous systems in our test environment. After applying the patch, the hangs no longer occur. Signed-off-by: Brian King Signed-off-by: "Theodore Ts'o" --- fs/jbd2/commit.c | 12 ++++++++---- fs/jbd2/journal.c | 4 +++- include/linux/jbd2.h | 2 +- 3 files changed, 12 insertions(+), 6 deletions(-) (limited to 'fs/jbd2/commit.c') diff --git a/fs/jbd2/commit.c b/fs/jbd2/commit.c index 7c068c189d8..6494c81e3b0 100644 --- a/fs/jbd2/commit.c +++ b/fs/jbd2/commit.c @@ -26,7 +26,9 @@ #include #include #include +#include #include +#include /* * Default IO end handler for temporary BJ_IO buffer_heads. @@ -236,7 +238,7 @@ static int journal_submit_data_buffers(journal_t *journal, spin_lock(&journal->j_list_lock); list_for_each_entry(jinode, &commit_transaction->t_inode_list, i_list) { mapping = jinode->i_vfs_inode->i_mapping; - jinode->i_flags |= JI_COMMIT_RUNNING; + set_bit(__JI_COMMIT_RUNNING, &jinode->i_flags); spin_unlock(&journal->j_list_lock); /* * submit the inode data buffers. We use writepage @@ -251,7 +253,8 @@ static int journal_submit_data_buffers(journal_t *journal, spin_lock(&journal->j_list_lock); J_ASSERT(jinode->i_transaction == commit_transaction); commit_transaction->t_flushed_data_blocks = 1; - jinode->i_flags &= ~JI_COMMIT_RUNNING; + clear_bit(__JI_COMMIT_RUNNING, &jinode->i_flags); + smp_mb__after_clear_bit(); wake_up_bit(&jinode->i_flags, __JI_COMMIT_RUNNING); } spin_unlock(&journal->j_list_lock); @@ -272,7 +275,7 @@ static int journal_finish_inode_data_buffers(journal_t *journal, /* For locking, see the comment in journal_submit_data_buffers() */ spin_lock(&journal->j_list_lock); list_for_each_entry(jinode, &commit_transaction->t_inode_list, i_list) { - jinode->i_flags |= JI_COMMIT_RUNNING; + set_bit(__JI_COMMIT_RUNNING, &jinode->i_flags); spin_unlock(&journal->j_list_lock); err = filemap_fdatawait(jinode->i_vfs_inode->i_mapping); if (err) { @@ -288,7 +291,8 @@ static int journal_finish_inode_data_buffers(journal_t *journal, ret = err; } spin_lock(&journal->j_list_lock); - jinode->i_flags &= ~JI_COMMIT_RUNNING; + clear_bit(__JI_COMMIT_RUNNING, &jinode->i_flags); + smp_mb__after_clear_bit(); wake_up_bit(&jinode->i_flags, __JI_COMMIT_RUNNING); } diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index 0e8014ea6b9..75e1b5a0bc2 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -42,12 +42,14 @@ #include #include #include +#include #define CREATE_TRACE_POINTS #include #include #include +#include EXPORT_SYMBOL(jbd2_journal_extend); EXPORT_SYMBOL(jbd2_journal_stop); @@ -2206,7 +2208,7 @@ void jbd2_journal_release_jbd_inode(journal_t *journal, restart: spin_lock(&journal->j_list_lock); /* Is commit writing out inode - we have to wait */ - if (jinode->i_flags & JI_COMMIT_RUNNING) { + if (test_bit(__JI_COMMIT_RUNNING, &jinode->i_flags)) { wait_queue_head_t *wq; DEFINE_WAIT_BIT(wait, &jinode->i_flags, __JI_COMMIT_RUNNING); wq = bit_waitqueue(&jinode->i_flags, __JI_COMMIT_RUNNING); diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h index 0b52924a0cb..2ae86aa21fc 100644 --- a/include/linux/jbd2.h +++ b/include/linux/jbd2.h @@ -395,7 +395,7 @@ struct jbd2_inode { struct inode *i_vfs_inode; /* Flags of inode [j_list_lock] */ - unsigned int i_flags; + unsigned long i_flags; }; struct jbd2_revoke_table_s; -- cgit v1.2.3