From 2c575026fae6e63771bd2a4c1d407214a8096a89 Mon Sep 17 00:00:00 2001 From: Hong Zhiguo Date: Wed, 20 Nov 2013 10:35:05 -0700 Subject: Update of blkg_stat and blkg_rwstat may happen in bh context. While u64_stats_fetch_retry is only preempt_disable on 32bit UP system. This is not enough to avoid preemption by bh and may read strange 64 bit value. Signed-off-by: Hong Zhiguo Acked-by: Tejun Heo Cc: stable@kernel.org Signed-off-by: Jens Axboe --- block/blk-cgroup.h | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'block') diff --git a/block/blk-cgroup.h b/block/blk-cgroup.h index 1610b22edf09..86154eab9523 100644 --- a/block/blk-cgroup.h +++ b/block/blk-cgroup.h @@ -435,9 +435,9 @@ static inline uint64_t blkg_stat_read(struct blkg_stat *stat) uint64_t v; do { - start = u64_stats_fetch_begin(&stat->syncp); + start = u64_stats_fetch_begin_bh(&stat->syncp); v = stat->cnt; - } while (u64_stats_fetch_retry(&stat->syncp, start)); + } while (u64_stats_fetch_retry_bh(&stat->syncp, start)); return v; } @@ -508,9 +508,9 @@ static inline struct blkg_rwstat blkg_rwstat_read(struct blkg_rwstat *rwstat) struct blkg_rwstat tmp; do { - start = u64_stats_fetch_begin(&rwstat->syncp); + start = u64_stats_fetch_begin_bh(&rwstat->syncp); tmp = *rwstat; - } while (u64_stats_fetch_retry(&rwstat->syncp, start)); + } while (u64_stats_fetch_retry_bh(&rwstat->syncp, start)); return tmp; } -- cgit v1.2.3 From c170bbb45febc03ac4d34ba2b8bb55e06104b7e7 Mon Sep 17 00:00:00 2001 From: Kent Overstreet Date: Sun, 24 Nov 2013 16:32:22 -0700 Subject: block: submit_bio_wait() conversions It was being open coded in a few places. Signed-off-by: Kent Overstreet Cc: Jens Axboe Cc: Joern Engel Cc: Prasad Joshi Cc: Neil Brown Cc: Chris Mason Acked-by: NeilBrown Signed-off-by: Jens Axboe --- block/blk-flush.c | 19 +------------------ drivers/md/md.c | 12 +----------- fs/btrfs/check-integrity.c | 32 +++++++++++++------------------- fs/btrfs/check-integrity.h | 2 ++ fs/btrfs/extent_io.c | 12 +----------- fs/btrfs/scrub.c | 33 ++++----------------------------- fs/hfsplus/wrapper.c | 17 +---------------- fs/logfs/dev_bdev.c | 13 +------------ 8 files changed, 24 insertions(+), 116 deletions(-) (limited to 'block') diff --git a/block/blk-flush.c b/block/blk-flush.c index 331e627301ea..fb6f3c0ffa49 100644 --- a/block/blk-flush.c +++ b/block/blk-flush.c @@ -502,15 +502,6 @@ void blk_abort_flushes(struct request_queue *q) } } -static void bio_end_flush(struct bio *bio, int err) -{ - if (err) - clear_bit(BIO_UPTODATE, &bio->bi_flags); - if (bio->bi_private) - complete(bio->bi_private); - bio_put(bio); -} - /** * blkdev_issue_flush - queue a flush * @bdev: blockdev to issue flush for @@ -526,7 +517,6 @@ static void bio_end_flush(struct bio *bio, int err) int blkdev_issue_flush(struct block_device *bdev, gfp_t gfp_mask, sector_t *error_sector) { - DECLARE_COMPLETION_ONSTACK(wait); struct request_queue *q; struct bio *bio; int ret = 0; @@ -548,13 +538,9 @@ int blkdev_issue_flush(struct block_device *bdev, gfp_t gfp_mask, return -ENXIO; bio = bio_alloc(gfp_mask, 0); - bio->bi_end_io = bio_end_flush; bio->bi_bdev = bdev; - bio->bi_private = &wait; - bio_get(bio); - submit_bio(WRITE_FLUSH, bio); - wait_for_completion_io(&wait); + ret = submit_bio_wait(WRITE_FLUSH, bio); /* * The driver must store the error location in ->bi_sector, if @@ -564,9 +550,6 @@ int blkdev_issue_flush(struct block_device *bdev, gfp_t gfp_mask, if (error_sector) *error_sector = bio->bi_sector; - if (!bio_flagged(bio, BIO_UPTODATE)) - ret = -EIO; - bio_put(bio); return ret; } diff --git a/drivers/md/md.c b/drivers/md/md.c index b6b7a2866c9e..8700de376876 100644 --- a/drivers/md/md.c +++ b/drivers/md/md.c @@ -776,16 +776,10 @@ void md_super_wait(struct mddev *mddev) finish_wait(&mddev->sb_wait, &wq); } -static void bi_complete(struct bio *bio, int error) -{ - complete((struct completion*)bio->bi_private); -} - int sync_page_io(struct md_rdev *rdev, sector_t sector, int size, struct page *page, int rw, bool metadata_op) { struct bio *bio = bio_alloc_mddev(GFP_NOIO, 1, rdev->mddev); - struct completion event; int ret; rw |= REQ_SYNC; @@ -801,11 +795,7 @@ int sync_page_io(struct md_rdev *rdev, sector_t sector, int size, else bio->bi_sector = sector + rdev->data_offset; bio_add_page(bio, page, size, 0); - init_completion(&event); - bio->bi_private = &event; - bio->bi_end_io = bi_complete; - submit_bio(rw, bio); - wait_for_completion(&event); + submit_bio_wait(rw, bio); ret = test_bit(BIO_UPTODATE, &bio->bi_flags); bio_put(bio); diff --git a/fs/btrfs/check-integrity.c b/fs/btrfs/check-integrity.c index e0aab4456974..f85b1c409003 100644 --- a/fs/btrfs/check-integrity.c +++ b/fs/btrfs/check-integrity.c @@ -323,7 +323,6 @@ static void btrfsic_release_block_ctx(struct btrfsic_block_data_ctx *block_ctx); static int btrfsic_read_block(struct btrfsic_state *state, struct btrfsic_block_data_ctx *block_ctx); static void btrfsic_dump_database(struct btrfsic_state *state); -static void btrfsic_complete_bio_end_io(struct bio *bio, int err); static int btrfsic_test_for_metadata(struct btrfsic_state *state, char **datav, unsigned int num_pages); static void btrfsic_process_written_block(struct btrfsic_dev_state *dev_state, @@ -1677,7 +1676,6 @@ static int btrfsic_read_block(struct btrfsic_state *state, for (i = 0; i < num_pages;) { struct bio *bio; unsigned int j; - DECLARE_COMPLETION_ONSTACK(complete); bio = btrfs_io_bio_alloc(GFP_NOFS, num_pages - i); if (!bio) { @@ -1688,8 +1686,6 @@ static int btrfsic_read_block(struct btrfsic_state *state, } bio->bi_bdev = block_ctx->dev->bdev; bio->bi_sector = dev_bytenr >> 9; - bio->bi_end_io = btrfsic_complete_bio_end_io; - bio->bi_private = &complete; for (j = i; j < num_pages; j++) { ret = bio_add_page(bio, block_ctx->pagev[j], @@ -1702,12 +1698,7 @@ static int btrfsic_read_block(struct btrfsic_state *state, "btrfsic: error, failed to add a single page!\n"); return -1; } - submit_bio(READ, bio); - - /* this will also unplug the queue */ - wait_for_completion(&complete); - - if (!test_bit(BIO_UPTODATE, &bio->bi_flags)) { + if (submit_bio_wait(READ, bio)) { printk(KERN_INFO "btrfsic: read error at logical %llu dev %s!\n", block_ctx->start, block_ctx->dev->name); @@ -1730,11 +1721,6 @@ static int btrfsic_read_block(struct btrfsic_state *state, return block_ctx->len; } -static void btrfsic_complete_bio_end_io(struct bio *bio, int err) -{ - complete((struct completion *)bio->bi_private); -} - static void btrfsic_dump_database(struct btrfsic_state *state) { struct list_head *elem_all; @@ -2998,14 +2984,12 @@ int btrfsic_submit_bh(int rw, struct buffer_head *bh) return submit_bh(rw, bh); } -void btrfsic_submit_bio(int rw, struct bio *bio) +static void __btrfsic_submit_bio(int rw, struct bio *bio) { struct btrfsic_dev_state *dev_state; - if (!btrfsic_is_initialized) { - submit_bio(rw, bio); + if (!btrfsic_is_initialized) return; - } mutex_lock(&btrfsic_mutex); /* since btrfsic_submit_bio() is also called before @@ -3097,10 +3081,20 @@ void btrfsic_submit_bio(int rw, struct bio *bio) } leave: mutex_unlock(&btrfsic_mutex); +} +void btrfsic_submit_bio(int rw, struct bio *bio) +{ + __btrfsic_submit_bio(rw, bio); submit_bio(rw, bio); } +int btrfsic_submit_bio_wait(int rw, struct bio *bio) +{ + __btrfsic_submit_bio(rw, bio); + return submit_bio_wait(rw, bio); +} + int btrfsic_mount(struct btrfs_root *root, struct btrfs_fs_devices *fs_devices, int including_extent_data, u32 print_mask) diff --git a/fs/btrfs/check-integrity.h b/fs/btrfs/check-integrity.h index 8b59175cc502..13b8566c97ab 100644 --- a/fs/btrfs/check-integrity.h +++ b/fs/btrfs/check-integrity.h @@ -22,9 +22,11 @@ #ifdef CONFIG_BTRFS_FS_CHECK_INTEGRITY int btrfsic_submit_bh(int rw, struct buffer_head *bh); void btrfsic_submit_bio(int rw, struct bio *bio); +int btrfsic_submit_bio_wait(int rw, struct bio *bio); #else #define btrfsic_submit_bh submit_bh #define btrfsic_submit_bio submit_bio +#define btrfsic_submit_bio_wait submit_bio_wait #endif int btrfsic_mount(struct btrfs_root *root, diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index 856bc2b2192c..014beaa9458c 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -1952,11 +1952,6 @@ static int free_io_failure(struct inode *inode, struct io_failure_record *rec, return err; } -static void repair_io_failure_callback(struct bio *bio, int err) -{ - complete(bio->bi_private); -} - /* * this bypasses the standard btrfs submit functions deliberately, as * the standard behavior is to write all copies in a raid setup. here we only @@ -1973,7 +1968,6 @@ int repair_io_failure(struct btrfs_fs_info *fs_info, u64 start, { struct bio *bio; struct btrfs_device *dev; - DECLARE_COMPLETION_ONSTACK(compl); u64 map_length = 0; u64 sector; struct btrfs_bio *bbio = NULL; @@ -1989,8 +1983,6 @@ int repair_io_failure(struct btrfs_fs_info *fs_info, u64 start, bio = btrfs_io_bio_alloc(GFP_NOFS, 1); if (!bio) return -EIO; - bio->bi_private = &compl; - bio->bi_end_io = repair_io_failure_callback; bio->bi_size = 0; map_length = length; @@ -2011,10 +2003,8 @@ int repair_io_failure(struct btrfs_fs_info *fs_info, u64 start, } bio->bi_bdev = dev->bdev; bio_add_page(bio, page, length, start - page_offset(page)); - btrfsic_submit_bio(WRITE_SYNC, bio); - wait_for_completion(&compl); - if (!test_bit(BIO_UPTODATE, &bio->bi_flags)) { + if (btrfsic_submit_bio_wait(WRITE_SYNC, bio)) { /* try to remap that extent elsewhere? */ bio_put(bio); btrfs_dev_stat_inc_and_print(dev, BTRFS_DEV_STAT_WRITE_ERRS); diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c index 2544805544f0..3214ebe593bd 100644 --- a/fs/btrfs/scrub.c +++ b/fs/btrfs/scrub.c @@ -208,7 +208,6 @@ static void scrub_recheck_block_checksum(struct btrfs_fs_info *fs_info, int is_metadata, int have_csum, const u8 *csum, u64 generation, u16 csum_size); -static void scrub_complete_bio_end_io(struct bio *bio, int err); static int scrub_repair_block_from_good_copy(struct scrub_block *sblock_bad, struct scrub_block *sblock_good, int force_write); @@ -1292,7 +1291,6 @@ static void scrub_recheck_block(struct btrfs_fs_info *fs_info, for (page_num = 0; page_num < sblock->page_count; page_num++) { struct bio *bio; struct scrub_page *page = sblock->pagev[page_num]; - DECLARE_COMPLETION_ONSTACK(complete); if (page->dev->bdev == NULL) { page->io_error = 1; @@ -1309,18 +1307,11 @@ static void scrub_recheck_block(struct btrfs_fs_info *fs_info, } bio->bi_bdev = page->dev->bdev; bio->bi_sector = page->physical >> 9; - bio->bi_end_io = scrub_complete_bio_end_io; - bio->bi_private = &complete; bio_add_page(bio, page->page, PAGE_SIZE, 0); - btrfsic_submit_bio(READ, bio); - - /* this will also unplug the queue */ - wait_for_completion(&complete); - - page->io_error = !test_bit(BIO_UPTODATE, &bio->bi_flags); - if (!test_bit(BIO_UPTODATE, &bio->bi_flags)) + if (btrfsic_submit_bio_wait(READ, bio)) sblock->no_io_error_seen = 0; + bio_put(bio); } @@ -1389,11 +1380,6 @@ static void scrub_recheck_block_checksum(struct btrfs_fs_info *fs_info, sblock->checksum_error = 1; } -static void scrub_complete_bio_end_io(struct bio *bio, int err) -{ - complete((struct completion *)bio->bi_private); -} - static int scrub_repair_block_from_good_copy(struct scrub_block *sblock_bad, struct scrub_block *sblock_good, int force_write) @@ -1428,7 +1414,6 @@ static int scrub_repair_page_from_good_copy(struct scrub_block *sblock_bad, sblock_bad->checksum_error || page_bad->io_error) { struct bio *bio; int ret; - DECLARE_COMPLETION_ONSTACK(complete); if (!page_bad->dev->bdev) { printk_ratelimited(KERN_WARNING @@ -1441,19 +1426,14 @@ static int scrub_repair_page_from_good_copy(struct scrub_block *sblock_bad, return -EIO; bio->bi_bdev = page_bad->dev->bdev; bio->bi_sector = page_bad->physical >> 9; - bio->bi_end_io = scrub_complete_bio_end_io; - bio->bi_private = &complete; ret = bio_add_page(bio, page_good->page, PAGE_SIZE, 0); if (PAGE_SIZE != ret) { bio_put(bio); return -EIO; } - btrfsic_submit_bio(WRITE, bio); - /* this will also unplug the queue */ - wait_for_completion(&complete); - if (!bio_flagged(bio, BIO_UPTODATE)) { + if (btrfsic_submit_bio_wait(WRITE, bio)) { btrfs_dev_stat_inc_and_print(page_bad->dev, BTRFS_DEV_STAT_WRITE_ERRS); btrfs_dev_replace_stats_inc( @@ -3373,7 +3353,6 @@ static int write_page_nocow(struct scrub_ctx *sctx, struct bio *bio; struct btrfs_device *dev; int ret; - DECLARE_COMPLETION_ONSTACK(compl); dev = sctx->wr_ctx.tgtdev; if (!dev) @@ -3390,8 +3369,6 @@ static int write_page_nocow(struct scrub_ctx *sctx, spin_unlock(&sctx->stat_lock); return -ENOMEM; } - bio->bi_private = &compl; - bio->bi_end_io = scrub_complete_bio_end_io; bio->bi_size = 0; bio->bi_sector = physical_for_dev_replace >> 9; bio->bi_bdev = dev->bdev; @@ -3402,10 +3379,8 @@ leave_with_eio: btrfs_dev_stat_inc_and_print(dev, BTRFS_DEV_STAT_WRITE_ERRS); return -EIO; } - btrfsic_submit_bio(WRITE_SYNC, bio); - wait_for_completion(&compl); - if (!test_bit(BIO_UPTODATE, &bio->bi_flags)) + if (btrfsic_submit_bio_wait(WRITE_SYNC, bio)) goto leave_with_eio; bio_put(bio); diff --git a/fs/hfsplus/wrapper.c b/fs/hfsplus/wrapper.c index b51a6079108d..e9a97a0d4314 100644 --- a/fs/hfsplus/wrapper.c +++ b/fs/hfsplus/wrapper.c @@ -24,13 +24,6 @@ struct hfsplus_wd { u16 embed_count; }; -static void hfsplus_end_io_sync(struct bio *bio, int err) -{ - if (err) - clear_bit(BIO_UPTODATE, &bio->bi_flags); - complete(bio->bi_private); -} - /* * hfsplus_submit_bio - Perfrom block I/O * @sb: super block of volume for I/O @@ -53,7 +46,6 @@ static void hfsplus_end_io_sync(struct bio *bio, int err) int hfsplus_submit_bio(struct super_block *sb, sector_t sector, void *buf, void **data, int rw) { - DECLARE_COMPLETION_ONSTACK(wait); struct bio *bio; int ret = 0; u64 io_size; @@ -73,8 +65,6 @@ int hfsplus_submit_bio(struct super_block *sb, sector_t sector, bio = bio_alloc(GFP_NOIO, 1); bio->bi_sector = sector; bio->bi_bdev = sb->s_bdev; - bio->bi_end_io = hfsplus_end_io_sync; - bio->bi_private = &wait; if (!(rw & WRITE) && data) *data = (u8 *)buf + offset; @@ -93,12 +83,7 @@ int hfsplus_submit_bio(struct super_block *sb, sector_t sector, buf = (u8 *)buf + len; } - submit_bio(rw, bio); - wait_for_completion(&wait); - - if (!bio_flagged(bio, BIO_UPTODATE)) - ret = -EIO; - + ret = submit_bio_wait(rw, bio); out: bio_put(bio); return ret < 0 ? ret : 0; diff --git a/fs/logfs/dev_bdev.c b/fs/logfs/dev_bdev.c index 550475ca6a0e..0f95f0d0b313 100644 --- a/fs/logfs/dev_bdev.c +++ b/fs/logfs/dev_bdev.c @@ -14,16 +14,10 @@ #define PAGE_OFS(ofs) ((ofs) & (PAGE_SIZE-1)) -static void request_complete(struct bio *bio, int err) -{ - complete((struct completion *)bio->bi_private); -} - static int sync_request(struct page *page, struct block_device *bdev, int rw) { struct bio bio; struct bio_vec bio_vec; - struct completion complete; bio_init(&bio); bio.bi_max_vecs = 1; @@ -35,13 +29,8 @@ static int sync_request(struct page *page, struct block_device *bdev, int rw) bio.bi_size = PAGE_SIZE; bio.bi_bdev = bdev; bio.bi_sector = page->index * (PAGE_SIZE >> 9); - init_completion(&complete); - bio.bi_private = &complete; - bio.bi_end_io = request_complete; - submit_bio(rw, &bio); - wait_for_completion(&complete); - return test_bit(BIO_UPTODATE, &bio.bi_flags) ? 0 : -EIO; + return submit_bio_wait(rw, &bio); } static int bdev_readpage(void *_sb, struct page *page) -- cgit v1.2.3 From 959a35f13eb785f982d79b1aaa75872d05c821da Mon Sep 17 00:00:00 2001 From: Jeff Moyer Date: Tue, 3 Dec 2013 14:23:00 -0700 Subject: blk-mq: fix dereference of rq->mq_ctx if allocation fails If __GFP_WAIT isn't set and we fail allocating, when we go to drop the reference on the ctx, we will attempt to dereference the NULL rq. Fix that. Signed-off-by: Jeff Moyer Signed-off-by: Jens Axboe --- block/blk-mq.c | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-) (limited to 'block') diff --git a/block/blk-mq.c b/block/blk-mq.c index cdc629cf075b..70fd6f996600 100644 --- a/block/blk-mq.c +++ b/block/blk-mq.c @@ -202,10 +202,12 @@ static struct request *blk_mq_alloc_request_pinned(struct request_queue *q, if (rq) { blk_mq_rq_ctx_init(q, ctx, rq, rw); break; - } else if (!(gfp & __GFP_WAIT)) - break; + } blk_mq_put_ctx(ctx); + if (!(gfp & __GFP_WAIT)) + break; + __blk_mq_run_hw_queue(hctx); blk_mq_wait_for_tags(hctx->tags); } while (1); @@ -222,7 +224,8 @@ struct request *blk_mq_alloc_request(struct request_queue *q, int rw, return NULL; rq = blk_mq_alloc_request_pinned(q, rw, gfp, reserved); - blk_mq_put_ctx(rq->mq_ctx); + if (rq) + blk_mq_put_ctx(rq->mq_ctx); return rq; } @@ -235,7 +238,8 @@ struct request *blk_mq_alloc_reserved_request(struct request_queue *q, int rw, return NULL; rq = blk_mq_alloc_request_pinned(q, rw, gfp, true); - blk_mq_put_ctx(rq->mq_ctx); + if (rq) + blk_mq_put_ctx(rq->mq_ctx); return rq; } EXPORT_SYMBOL(blk_mq_alloc_reserved_request); -- cgit v1.2.3 From 0d11e6aca396e679c07b2dd6af5dc8b7f041fbbd Mon Sep 17 00:00:00 2001 From: Ming Lei Date: Thu, 5 Dec 2013 10:50:39 -0700 Subject: blk-mq: fix use-after-free of request If accounting is on, we will do the IO completion accounting after we have freed the request. Fix that by moving it sooner instead. Signed-off-by: Jens Axboe --- block/blk-mq.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'block') diff --git a/block/blk-mq.c b/block/blk-mq.c index 70fd6f996600..c79126e11030 100644 --- a/block/blk-mq.c +++ b/block/blk-mq.c @@ -312,12 +312,12 @@ void blk_mq_complete_request(struct request *rq, int error) blk_account_io_completion(rq, bytes); + blk_account_io_done(rq); + if (rq->end_io) rq->end_io(rq, error); else blk_mq_free_request(rq); - - blk_account_io_done(rq); } void __blk_mq_end_io(struct request *rq, int error) -- cgit v1.2.3