From f6ff53d3611b564661896be23369b54d84941a0e Mon Sep 17 00:00:00 2001 From: Paolo Bonzini Date: Thu, 2 Aug 2012 09:48:49 +0200 Subject: block: reorganize rounding of max_discard_sectors Mostly a preparation for the next patch. In principle this fixes an infinite loop if max_discard_sectors < granularity, but that really shouldn't happen. Signed-off-by: Paolo Bonzini Acked-by: Vivek Goyal Tested-by: Mike Snitzer Signed-off-by: Jens Axboe --- block/blk-lib.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/block/blk-lib.c b/block/blk-lib.c index 2b461b496a78..16b06f62e68c 100644 --- a/block/blk-lib.c +++ b/block/blk-lib.c @@ -44,6 +44,7 @@ int blkdev_issue_discard(struct block_device *bdev, sector_t sector, struct request_queue *q = bdev_get_queue(bdev); int type = REQ_WRITE | REQ_DISCARD; unsigned int max_discard_sectors; + unsigned int granularity; struct bio_batch bb; struct bio *bio; int ret = 0; @@ -54,18 +55,18 @@ int blkdev_issue_discard(struct block_device *bdev, sector_t sector, if (!blk_queue_discard(q)) return -EOPNOTSUPP; + /* Zero-sector (unknown) and one-sector granularities are the same. */ + granularity = max(q->limits.discard_granularity >> 9, 1U); + /* * Ensure that max_discard_sectors is of the proper * granularity */ max_discard_sectors = min(q->limits.max_discard_sectors, UINT_MAX >> 9); + max_discard_sectors = round_down(max_discard_sectors, granularity); if (unlikely(!max_discard_sectors)) { /* Avoid infinite loop below. Being cautious never hurts. */ return -EOPNOTSUPP; - } else if (q->limits.discard_granularity) { - unsigned int disc_sects = q->limits.discard_granularity >> 9; - - max_discard_sectors &= ~(disc_sects - 1); } if (flags & BLKDEV_DISCARD_SECURE) { -- cgit v1.2.3-58-ga151 From c6e666345e1b79c62ba82339cc7d55a89cb73f88 Mon Sep 17 00:00:00 2001 From: Paolo Bonzini Date: Thu, 2 Aug 2012 09:48:50 +0200 Subject: block: split discard into aligned requests When a disk has large discard_granularity and small max_discard_sectors, discards are not split with optimal alignment. In the limit case of discard_granularity == max_discard_sectors, no request could be aligned correctly, so in fact you might end up with no discarded logical blocks at all. Another example that helps showing the condition in the patch is with discard_granularity == 64, max_discard_sectors == 128. A request that is submitted for 256 sectors 2..257 will be split in two: 2..129, 130..257. However, only 2 aligned blocks out of 3 are included in the request; 128..191 may be left intact and not discarded. With this patch, the first request will be truncated to ensure good alignment of what's left, and the split will be 2..127, 128..255, 256..257. The patch will also take into account the discard_alignment. At most one extra request will be introduced, because the first request will be reduced by at most granularity-1 sectors, and granularity must be less than max_discard_sectors. Subsequent requests will run on round_down(max_discard_sectors, granularity) sectors, as in the current code. Signed-off-by: Paolo Bonzini Acked-by: Vivek Goyal Tested-by: Mike Snitzer Signed-off-by: Jens Axboe --- block/blk-lib.c | 34 ++++++++++++++++++++++++---------- include/linux/blkdev.h | 10 ++++++++++ 2 files changed, 34 insertions(+), 10 deletions(-) diff --git a/block/blk-lib.c b/block/blk-lib.c index 16b06f62e68c..19cc761cacb2 100644 --- a/block/blk-lib.c +++ b/block/blk-lib.c @@ -44,7 +44,7 @@ int blkdev_issue_discard(struct block_device *bdev, sector_t sector, struct request_queue *q = bdev_get_queue(bdev); int type = REQ_WRITE | REQ_DISCARD; unsigned int max_discard_sectors; - unsigned int granularity; + unsigned int granularity, alignment, mask; struct bio_batch bb; struct bio *bio; int ret = 0; @@ -57,10 +57,12 @@ int blkdev_issue_discard(struct block_device *bdev, sector_t sector, /* Zero-sector (unknown) and one-sector granularities are the same. */ granularity = max(q->limits.discard_granularity >> 9, 1U); + mask = granularity - 1; + alignment = (bdev_discard_alignment(bdev) >> 9) & mask; /* * Ensure that max_discard_sectors is of the proper - * granularity + * granularity, so that requests stay aligned after a split. */ max_discard_sectors = min(q->limits.max_discard_sectors, UINT_MAX >> 9); max_discard_sectors = round_down(max_discard_sectors, granularity); @@ -80,25 +82,37 @@ int blkdev_issue_discard(struct block_device *bdev, sector_t sector, bb.wait = &wait; while (nr_sects) { + unsigned int req_sects; + sector_t end_sect; + bio = bio_alloc(gfp_mask, 1); if (!bio) { ret = -ENOMEM; break; } + req_sects = min_t(sector_t, nr_sects, max_discard_sectors); + + /* + * If splitting a request, and the next starting sector would be + * misaligned, stop the discard at the previous aligned sector. + */ + end_sect = sector + req_sects; + if (req_sects < nr_sects && (end_sect & mask) != alignment) { + end_sect = + round_down(end_sect - alignment, granularity) + + alignment; + req_sects = end_sect - sector; + } + bio->bi_sector = sector; bio->bi_end_io = bio_batch_end_io; bio->bi_bdev = bdev; bio->bi_private = &bb; - if (nr_sects > max_discard_sectors) { - bio->bi_size = max_discard_sectors << 9; - nr_sects -= max_discard_sectors; - sector += max_discard_sectors; - } else { - bio->bi_size = nr_sects << 9; - nr_sects = 0; - } + bio->bi_size = req_sects << 9; + nr_sects -= req_sects; + sector = end_sect; atomic_inc(&bb.done); submit_bio(type, bio); diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index 4e72a9d48232..281516ae8b4e 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -1139,6 +1139,16 @@ static inline int queue_limit_discard_alignment(struct queue_limits *lim, sector & (lim->discard_granularity - 1); } +static inline int bdev_discard_alignment(struct block_device *bdev) +{ + struct request_queue *q = bdev_get_queue(bdev); + + if (bdev != bdev->bd_contains) + return bdev->bd_part->discard_alignment; + + return q->limits.discard_alignment; +} + static inline unsigned int queue_discard_zeroes_data(struct request_queue *q) { if (q->limits.max_discard_sectors && q->limits.discard_zeroes_data == 1) -- cgit v1.2.3-58-ga151 From 53362a05ae683e12a20d9ffdf58a88094a0bed9d Mon Sep 17 00:00:00 2001 From: Jianpeng Ma Date: Thu, 2 Aug 2012 09:50:39 +0200 Subject: fs/block-dev.c:fix performance regression in O_DIRECT writes to md block devices For regular file, write operaion used blk_plug function.But for block file,write operation did not use blk_plug. This patch is also for write-cache mode for block-device. Signed-off-by: Jianpeng Ma Reviewed-by: NeilBrown Signed-off-by: Jens Axboe --- fs/block_dev.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/fs/block_dev.c b/fs/block_dev.c index 1e519195d45b..38e721b35d45 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -1578,10 +1578,12 @@ ssize_t blkdev_aio_write(struct kiocb *iocb, const struct iovec *iov, unsigned long nr_segs, loff_t pos) { struct file *file = iocb->ki_filp; + struct blk_plug plug; ssize_t ret; BUG_ON(iocb->ki_pos != pos); + blk_start_plug(&plug); ret = __generic_file_aio_write(iocb, iov, nr_segs, &iocb->ki_pos); if (ret > 0 || ret == -EIOCBQUEUED) { ssize_t err; @@ -1590,6 +1592,7 @@ ssize_t blkdev_aio_write(struct kiocb *iocb, const struct iovec *iov, if (err < 0 && ret > 0) ret = err; } + blk_finish_plug(&plug); return ret; } EXPORT_SYMBOL_GPL(blkdev_aio_write); -- cgit v1.2.3-58-ga151 From 963ab9e5da95c654bb3ab937cc478de4f7088a96 Mon Sep 17 00:00:00 2001 From: Asias He Date: Thu, 2 Aug 2012 23:42:03 +0200 Subject: block: Introduce __blk_segment_map_sg() helper Split the mapping code in blk_rq_map_sg() to a helper __blk_segment_map_sg(), so that other mapping function, e.g. blk_bio_map_sg(), can share the code. Cc: Rusty Russell Cc: Christoph Hellwig Cc: Tejun Heo Cc: Shaohua Li Cc: "Michael S. Tsirkin" Cc: kvm@vger.kernel.org Cc: linux-kernel@vger.kernel.org Cc: virtualization@lists.linux-foundation.org Suggested-by: Jens Axboe Suggested-by: Tejun Heo Signed-off-by: Asias He Signed-off-by: Jens Axboe --- block/blk-merge.c | 80 +++++++++++++++++++++++++++++++------------------------ 1 file changed, 45 insertions(+), 35 deletions(-) diff --git a/block/blk-merge.c b/block/blk-merge.c index 160035f54882..576b68e79248 100644 --- a/block/blk-merge.c +++ b/block/blk-merge.c @@ -110,6 +110,49 @@ static int blk_phys_contig_segment(struct request_queue *q, struct bio *bio, return 0; } +static void +__blk_segment_map_sg(struct request_queue *q, struct bio_vec *bvec, + struct scatterlist *sglist, struct bio_vec **bvprv, + struct scatterlist **sg, int *nsegs, int *cluster) +{ + + int nbytes = bvec->bv_len; + + if (*bvprv && *cluster) { + if ((*sg)->length + nbytes > queue_max_segment_size(q)) + goto new_segment; + + if (!BIOVEC_PHYS_MERGEABLE(*bvprv, bvec)) + goto new_segment; + if (!BIOVEC_SEG_BOUNDARY(q, *bvprv, bvec)) + goto new_segment; + + (*sg)->length += nbytes; + } else { +new_segment: + if (!*sg) + *sg = sglist; + else { + /* + * If the driver previously mapped a shorter + * list, we could see a termination bit + * prematurely unless it fully inits the sg + * table on each mapping. We KNOW that there + * must be more entries here or the driver + * would be buggy, so force clear the + * termination bit to avoid doing a full + * sg_init_table() in drivers for each command. + */ + (*sg)->page_link &= ~0x02; + *sg = sg_next(*sg); + } + + sg_set_page(*sg, bvec->bv_page, nbytes, bvec->bv_offset); + (*nsegs)++; + } + *bvprv = bvec; +} + /* * map a request to scatterlist, return number of sg entries setup. Caller * must make sure sg can hold rq->nr_phys_segments entries @@ -131,41 +174,8 @@ int blk_rq_map_sg(struct request_queue *q, struct request *rq, bvprv = NULL; sg = NULL; rq_for_each_segment(bvec, rq, iter) { - int nbytes = bvec->bv_len; - - if (bvprv && cluster) { - if (sg->length + nbytes > queue_max_segment_size(q)) - goto new_segment; - - if (!BIOVEC_PHYS_MERGEABLE(bvprv, bvec)) - goto new_segment; - if (!BIOVEC_SEG_BOUNDARY(q, bvprv, bvec)) - goto new_segment; - - sg->length += nbytes; - } else { -new_segment: - if (!sg) - sg = sglist; - else { - /* - * If the driver previously mapped a shorter - * list, we could see a termination bit - * prematurely unless it fully inits the sg - * table on each mapping. We KNOW that there - * must be more entries here or the driver - * would be buggy, so force clear the - * termination bit to avoid doing a full - * sg_init_table() in drivers for each command. - */ - sg->page_link &= ~0x02; - sg = sg_next(sg); - } - - sg_set_page(sg, bvec->bv_page, nbytes, bvec->bv_offset); - nsegs++; - } - bvprv = bvec; + __blk_segment_map_sg(q, bvec, sglist, &bvprv, &sg, + &nsegs, &cluster); } /* segments in rq */ -- cgit v1.2.3-58-ga151 From 85b9f66a41eb8ee3f1dfc95707412705463cdd97 Mon Sep 17 00:00:00 2001 From: Asias He Date: Thu, 2 Aug 2012 23:42:04 +0200 Subject: block: Add blk_bio_map_sg() helper Add a helper to map a bio to a scatterlist, modelled after blk_rq_map_sg. This helper is useful for any driver that wants to create a scatterlist from its ->make_request_fn method. Changes in v2: - Use __blk_segment_map_sg to avoid duplicated code - Add cocbook style function comment Cc: Rusty Russell Cc: Christoph Hellwig Cc: Tejun Heo Cc: Shaohua Li Cc: "Michael S. Tsirkin" Cc: kvm@vger.kernel.org Cc: linux-kernel@vger.kernel.org Cc: virtualization@lists.linux-foundation.org Signed-off-by: Christoph Hellwig Signed-off-by: Minchan Kim Signed-off-by: Asias He Signed-off-by: Jens Axboe --- block/blk-merge.c | 37 +++++++++++++++++++++++++++++++++++++ include/linux/blkdev.h | 2 ++ 2 files changed, 39 insertions(+) diff --git a/block/blk-merge.c b/block/blk-merge.c index 576b68e79248..e76279e41162 100644 --- a/block/blk-merge.c +++ b/block/blk-merge.c @@ -209,6 +209,43 @@ int blk_rq_map_sg(struct request_queue *q, struct request *rq, } EXPORT_SYMBOL(blk_rq_map_sg); +/** + * blk_bio_map_sg - map a bio to a scatterlist + * @q: request_queue in question + * @bio: bio being mapped + * @sglist: scatterlist being mapped + * + * Note: + * Caller must make sure sg can hold bio->bi_phys_segments entries + * + * Will return the number of sg entries setup + */ +int blk_bio_map_sg(struct request_queue *q, struct bio *bio, + struct scatterlist *sglist) +{ + struct bio_vec *bvec, *bvprv; + struct scatterlist *sg; + int nsegs, cluster; + unsigned long i; + + nsegs = 0; + cluster = blk_queue_cluster(q); + + bvprv = NULL; + sg = NULL; + bio_for_each_segment(bvec, bio, i) { + __blk_segment_map_sg(q, bvec, sglist, &bvprv, &sg, + &nsegs, &cluster); + } /* segments in bio */ + + if (sg) + sg_mark_end(sg); + + BUG_ON(bio->bi_phys_segments && nsegs > bio->bi_phys_segments); + return nsegs; +} +EXPORT_SYMBOL(blk_bio_map_sg); + static inline int ll_new_hw_segment(struct request_queue *q, struct request *req, struct bio *bio) diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index 281516ae8b4e..dc632975d54f 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -894,6 +894,8 @@ extern void blk_queue_flush_queueable(struct request_queue *q, bool queueable); extern struct backing_dev_info *blk_get_backing_dev_info(struct block_device *bdev); extern int blk_rq_map_sg(struct request_queue *, struct request *, struct scatterlist *); +extern int blk_bio_map_sg(struct request_queue *q, struct bio *bio, + struct scatterlist *sglist); extern void blk_dump_rq_flags(struct request *, char *); extern long nr_blockdev_pages(void); -- cgit v1.2.3-58-ga151 From 0676806707281e27b13d44323bed580a8160b7a4 Mon Sep 17 00:00:00 2001 From: Jianpeng Ma Date: Fri, 3 Aug 2012 10:42:00 +0200 Subject: block: Don't use static to define "void *p" in show_partition_start() I met a odd prblem:read /proc/partitions may return zero. I wrote a file test.c: int main() { char buff[4096]; int ret; int fd; printf("pid=%d\n",getpid()); while (1) { fd = open("/proc/partitions", O_RDONLY); if (fd < 0) { printf("open error %s\n", strerror(errno)); return 0; } ret = read(fd, buff, 4096); if (ret <= 0) printf("ret=%d, %s, %ld\n", ret, strerror(errno), lseek(fd,0,SEEK_CUR)); close(fd); } exit(0); } You can reproduce by: 1:while true;do cat /proc/partitions > /dev/null ;done 2:./test I reviewed the code and found: >> static void *show_partition_start(struct seq_file *seqf, loff_t *pos) >> { >> static void *p; >> >> p = disk_seqf_start(seqf, pos); >> if (!IS_ERR_OR_NULL(p) && !*pos) >> seq_puts(seqf, "major minor #blocks name\n\n"); >> return p; >> } test cat /proc/partitions p = disk_seqf_start()(Not NULL) p = disk_seqf_start()(NULL because pos) if (!IS_ERR_OR_NULL(p) && !*pos) Signed-off-by: Jianpeng Ma Signed-off-by: Jens Axboe --- block/genhd.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/block/genhd.c b/block/genhd.c index cac7366957c3..d839723303c8 100644 --- a/block/genhd.c +++ b/block/genhd.c @@ -835,7 +835,7 @@ static void disk_seqf_stop(struct seq_file *seqf, void *v) static void *show_partition_start(struct seq_file *seqf, loff_t *pos) { - static void *p; + void *p; p = disk_seqf_start(seqf, pos); if (!IS_ERR_OR_NULL(p) && !*pos) -- cgit v1.2.3-58-ga151 From 389d7b26d9e4f78b17366c23a3aa16b3c5cb3bde Mon Sep 17 00:00:00 2001 From: Alexey Khoroshilov Date: Thu, 9 Aug 2012 15:19:25 +0200 Subject: bio: Fix potential memory leak in bio_find_or_create_slab() Do not leak memory by updating pointer with potentially NULL realloc return value. Found by Linux Driver Verification project (linuxtesting.org). Signed-off-by: Alexey Khoroshilov Acked-by: Jeff Moyer Signed-off-by: Jens Axboe --- fs/bio.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/fs/bio.c b/fs/bio.c index 73922abba832..fed1f799cb56 100644 --- a/fs/bio.c +++ b/fs/bio.c @@ -73,7 +73,7 @@ static struct kmem_cache *bio_find_or_create_slab(unsigned int extra_size) { unsigned int sz = sizeof(struct bio) + extra_size; struct kmem_cache *slab = NULL; - struct bio_slab *bslab; + struct bio_slab *bslab, *new_bio_slabs; unsigned int i, entry = -1; mutex_lock(&bio_slab_lock); @@ -97,11 +97,12 @@ static struct kmem_cache *bio_find_or_create_slab(unsigned int extra_size) if (bio_slab_nr == bio_slab_max && entry == -1) { bio_slab_max <<= 1; - bio_slabs = krealloc(bio_slabs, - bio_slab_max * sizeof(struct bio_slab), - GFP_KERNEL); - if (!bio_slabs) + new_bio_slabs = krealloc(bio_slabs, + bio_slab_max * sizeof(struct bio_slab), + GFP_KERNEL); + if (!new_bio_slabs) goto out_unlock; + bio_slabs = new_bio_slabs; } if (entry == -1) entry = bio_slab_nr++; -- cgit v1.2.3-58-ga151 From 276f0f5d157bb4a816053f4f3a941dbcd4f76556 Mon Sep 17 00:00:00 2001 From: Shaohua Li Date: Thu, 9 Aug 2012 15:20:23 +0200 Subject: block: disable discard request merge temporarily The SCSI discard request merge never worked, and looks no solution for in future, let's disable it temporarily. Signed-off-by: Shaohua Li Reviewed-by: Christoph Hellwig Signed-off-by: Jens Axboe --- include/linux/blkdev.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index dc632975d54f..4a2ab7c85393 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -601,7 +601,7 @@ static inline void blk_clear_rl_full(struct request_list *rl, bool sync) * it already be started by driver. */ #define RQ_NOMERGE_FLAGS \ - (REQ_NOMERGE | REQ_STARTED | REQ_SOFTBARRIER | REQ_FLUSH | REQ_FUA) + (REQ_NOMERGE | REQ_STARTED | REQ_SOFTBARRIER | REQ_FLUSH | REQ_FUA | REQ_DISCARD) #define rq_mergeable(rq) \ (!((rq)->cmd_flags & RQ_NOMERGE_FLAGS) && \ (((rq)->cmd_flags & REQ_DISCARD) || \ -- cgit v1.2.3-58-ga151 From 8430f9772f9ac305211f177b0fa56a312175da11 Mon Sep 17 00:00:00 2001 From: Fengguang Wu Date: Thu, 9 Aug 2012 15:23:07 +0200 Subject: block: remove plugging at buffered write time Buffered write(2) is not directly tied to IO, so it's not suitable to handle plug in generic_file_aio_write(). Note that plugging for O_SYNC writes is also removed. The user may pass arbitrary @size arguments, which may be much larger than the preferable I/O size, or may cross extent/device boundaries. Let the lower layers handle the plugging. The plugging code here actually turns them into no-ops. CC: Li Shaohua Signed-off-by: Wu Fengguang Signed-off-by: Jens Axboe --- mm/filemap.c | 3 --- 1 file changed, 3 deletions(-) diff --git a/mm/filemap.c b/mm/filemap.c index fa5ca304148e..2b0952974cb9 100644 --- a/mm/filemap.c +++ b/mm/filemap.c @@ -2527,14 +2527,12 @@ ssize_t generic_file_aio_write(struct kiocb *iocb, const struct iovec *iov, { struct file *file = iocb->ki_filp; struct inode *inode = file->f_mapping->host; - struct blk_plug plug; ssize_t ret; BUG_ON(iocb->ki_pos != pos); sb_start_write(inode->i_sb); mutex_lock(&inode->i_mutex); - blk_start_plug(&plug); ret = __generic_file_aio_write(iocb, iov, nr_segs, &iocb->ki_pos); mutex_unlock(&inode->i_mutex); @@ -2545,7 +2543,6 @@ ssize_t generic_file_aio_write(struct kiocb *iocb, const struct iovec *iov, if (err < 0 && ret > 0) ret = err; } - blk_finish_plug(&plug); sb_end_write(inode->i_sb); return ret; } -- cgit v1.2.3-58-ga151 From 647d1e4c5235763b83fbfe74a09d148edc6ca152 Mon Sep 17 00:00:00 2001 From: Fengguang Wu Date: Thu, 9 Aug 2012 15:23:09 +0200 Subject: block: move down direct IO plugging Move unplugging for direct I/O from around ->direct_IO() down to do_blockdev_direct_IO(). This implicitly adds plugging for direct writes. CC: Li Shaohua Acked-by: Jeff Moyer Signed-off-by: Wu Fengguang Signed-off-by: Jens Axboe --- fs/direct-io.c | 5 +++++ mm/filemap.c | 4 ---- 2 files changed, 5 insertions(+), 4 deletions(-) diff --git a/fs/direct-io.c b/fs/direct-io.c index 1faf4cb56f39..f86c720dba0e 100644 --- a/fs/direct-io.c +++ b/fs/direct-io.c @@ -1062,6 +1062,7 @@ do_blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode, unsigned long user_addr; size_t bytes; struct buffer_head map_bh = { 0, }; + struct blk_plug plug; if (rw & WRITE) rw = WRITE_ODIRECT; @@ -1177,6 +1178,8 @@ do_blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode, PAGE_SIZE - user_addr / PAGE_SIZE); } + blk_start_plug(&plug); + for (seg = 0; seg < nr_segs; seg++) { user_addr = (unsigned long)iov[seg].iov_base; sdio.size += bytes = iov[seg].iov_len; @@ -1235,6 +1238,8 @@ do_blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode, if (sdio.bio) dio_bio_submit(dio, &sdio); + blk_finish_plug(&plug); + /* * It is possible that, we return short IO due to end of file. * In that case, we need to release all the pages we got hold on. diff --git a/mm/filemap.c b/mm/filemap.c index 2b0952974cb9..384344575c37 100644 --- a/mm/filemap.c +++ b/mm/filemap.c @@ -1412,12 +1412,8 @@ generic_file_aio_read(struct kiocb *iocb, const struct iovec *iov, retval = filemap_write_and_wait_range(mapping, pos, pos + iov_length(iov, nr_segs) - 1); if (!retval) { - struct blk_plug plug; - - blk_start_plug(&plug); retval = mapping->a_ops->direct_IO(READ, iocb, iov, pos, nr_segs); - blk_finish_plug(&plug); } if (retval > 0) { *ppos = pos + retval; -- cgit v1.2.3-58-ga151 From 865826644e23c329297459b62ef9cc9904abf61a Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Thu, 9 Aug 2012 15:27:28 +0200 Subject: Documentation: update missing index files in block/00-INDEX Update missing index files in block/00-INDEX. Signed-off-by: Namjae Jeon Signed-off-by: Jens Axboe --- Documentation/block/00-INDEX | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) diff --git a/Documentation/block/00-INDEX b/Documentation/block/00-INDEX index d111e3b23db0..d18ecd827c40 100644 --- a/Documentation/block/00-INDEX +++ b/Documentation/block/00-INDEX @@ -3,15 +3,21 @@ biodoc.txt - Notes on the Generic Block Layer Rewrite in Linux 2.5 capability.txt - - Generic Block Device Capability (/sys/block//capability) + - Generic Block Device Capability (/sys/block//capability) +cfq-iosched.txt + - CFQ IO scheduler tunables +data-integrity.txt + - Block data integrity deadline-iosched.txt - Deadline IO scheduler tunables ioprio.txt - Block io priorities (in CFQ scheduler) +queue-sysfs.txt + - Queue's sysfs entries request.txt - The members of struct request (in include/linux/blkdev.h) stat.txt - - Block layer statistics in /sys/block//stat + - Block layer statistics in /sys/block//stat switching-sched.txt - Switching I/O schedulers at runtime writeback_cache_control.txt -- cgit v1.2.3-58-ga151 From 2792d8719393a518a847769d36a080ade4273e6b Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Thu, 9 Aug 2012 15:27:29 +0200 Subject: Documentation: update tunable options in block/cfq-iosched.txt Update tunable options in block/cfq-iosched.txt. Signed-off-by: Namjae Jeon Signed-off-by: Jens Axboe --- Documentation/block/cfq-iosched.txt | 77 +++++++++++++++++++++++++++++++++++++ 1 file changed, 77 insertions(+) diff --git a/Documentation/block/cfq-iosched.txt b/Documentation/block/cfq-iosched.txt index 6d670f570451..d89b4fe724d7 100644 --- a/Documentation/block/cfq-iosched.txt +++ b/Documentation/block/cfq-iosched.txt @@ -1,3 +1,14 @@ +CFQ (Complete Fairness Queueing) +=============================== + +The main aim of CFQ scheduler is to provide a fair allocation of the disk +I/O bandwidth for all the processes which requests an I/O operation. + +CFQ maintains the per process queue for the processes which request I/O +operation(syncronous requests). In case of asynchronous requests, all the +requests from all the processes are batched together according to their +process's I/O priority. + CFQ ioscheduler tunables ======================== @@ -25,6 +36,72 @@ there are multiple spindles behind single LUN (Host based hardware RAID controller or for storage arrays), setting slice_idle=0 might end up in better throughput and acceptable latencies. +back_seek_max +------------- +This specifies, given in Kbytes, the maximum "distance" for backward seeking. +The distance is the amount of space from the current head location to the +sectors that are backward in terms of distance. + +This parameter allows the scheduler to anticipate requests in the "backward" +direction and consider them as being the "next" if they are within this +distance from the current head location. + +back_seek_penalty +----------------- +This parameter is used to compute the cost of backward seeking. If the +backward distance of request is just 1/back_seek_penalty from a "front" +request, then the seeking cost of two requests is considered equivalent. + +So scheduler will not bias toward one or the other request (otherwise scheduler +will bias toward front request). Default value of back_seek_penalty is 2. + +fifo_expire_async +----------------- +This parameter is used to set the timeout of asynchronous requests. Default +value of this is 248ms. + +fifo_expire_sync +---------------- +This parameter is used to set the timeout of synchronous requests. Default +value of this is 124ms. In case to favor synchronous requests over asynchronous +one, this value should be decreased relative to fifo_expire_async. + +slice_async +----------- +This parameter is same as of slice_sync but for asynchronous queue. The +default value is 40ms. + +slice_async_rq +-------------- +This parameter is used to limit the dispatching of asynchronous request to +device request queue in queue's slice time. The maximum number of request that +are allowed to be dispatched also depends upon the io priority. Default value +for this is 2. + +slice_sync +---------- +When a queue is selected for execution, the queues IO requests are only +executed for a certain amount of time(time_slice) before switching to another +queue. This parameter is used to calculate the time slice of synchronous +queue. + +time_slice is computed using the below equation:- +time_slice = slice_sync + (slice_sync/5 * (4 - prio)). To increase the +time_slice of synchronous queue, increase the value of slice_sync. Default +value is 100ms. + +quantum +------- +This specifies the number of request dispatched to the device queue. In a +queue's time slice, a request will not be dispatched if the number of request +in the device exceeds this parameter. This parameter is used for synchronous +request. + +In case of storage with several disk, this setting can limit the parallel +processing of request. Therefore, increasing the value can imporve the +performace although this can cause the latency of some I/O to increase due +to more number of requests. + CFQ IOPS Mode for group scheduling =================================== Basic CFQ design is to provide priority based time slices. Higher priority -- cgit v1.2.3-58-ga151 From 4004e90cfa4488cf99798b1138183158b64f4a42 Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Thu, 9 Aug 2012 15:28:05 +0200 Subject: Documentation: update tunable options in block/cfq-iosched.txt Update tunable options in block/cfq-iosched.txt. Signed-off-by: Namjae Jeon Signed-off-by: Jens Axboe --- Documentation/block/queue-sysfs.txt | 64 +++++++++++++++++++++++++++++++++++++ 1 file changed, 64 insertions(+) diff --git a/Documentation/block/queue-sysfs.txt b/Documentation/block/queue-sysfs.txt index 6518a55273e7..e54ac1d53403 100644 --- a/Documentation/block/queue-sysfs.txt +++ b/Documentation/block/queue-sysfs.txt @@ -9,20 +9,71 @@ These files are the ones found in the /sys/block/xxx/queue/ directory. Files denoted with a RO postfix are readonly and the RW postfix means read-write. +add_random (RW) +---------------- +This file allows to trun off the disk entropy contribution. Default +value of this file is '1'(on). + +discard_granularity (RO) +----------------------- +This shows the size of internal allocation of the device in bytes, if +reported by the device. A value of '0' means device does not support +the discard functionality. + +discard_max_bytes (RO) +---------------------- +Devices that support discard functionality may have internal limits on +the number of bytes that can be trimmed or unmapped in a single operation. +The discard_max_bytes parameter is set by the device driver to the maximum +number of bytes that can be discarded in a single operation. Discard +requests issued to the device must not exceed this limit. A discard_max_bytes +value of 0 means that the device does not support discard functionality. + +discard_zeroes_data (RO) +------------------------ +When read, this file will show if the discarded block are zeroed by the +device or not. If its value is '1' the blocks are zeroed otherwise not. + hw_sector_size (RO) ------------------- This is the hardware sector size of the device, in bytes. +iostats (RW) +------------- +This file is used to control (on/off) the iostats accounting of the +disk. + +logical_block_size (RO) +----------------------- +This is the logcal block size of the device, in bytes. + max_hw_sectors_kb (RO) ---------------------- This is the maximum number of kilobytes supported in a single data transfer. +max_integrity_segments (RO) +--------------------------- +When read, this file shows the max limit of integrity segments as +set by block layer which a hardware controller can handle. + max_sectors_kb (RW) ------------------- This is the maximum number of kilobytes that the block layer will allow for a filesystem request. Must be smaller than or equal to the maximum size allowed by the hardware. +max_segments (RO) +----------------- +Maximum number of segments of the device. + +max_segment_size (RO) +--------------------- +Maximum segment size of the device. + +minimum_io_size (RO) +-------------------- +This is the smallest preferred io size reported by the device. + nomerges (RW) ------------- This enables the user to disable the lookup logic involved with IO @@ -45,11 +96,24 @@ per-block-cgroup request pool. IOW, if there are N block cgroups, each request queue may have upto N request pools, each independently regulated by nr_requests. +optimal_io_size (RO) +-------------------- +This is the optimal io size reported by the device. + +physical_block_size (RO) +------------------------ +This is the physical block size of device, in bytes. + read_ahead_kb (RW) ------------------ Maximum number of kilobytes to read-ahead for filesystems on this block device. +rotational (RW) +--------------- +This file is used to stat if the device is of rotational type or +non-rotational type. + rq_affinity (RW) ---------------- If this option is '1', the block layer will migrate request completions to the -- cgit v1.2.3-58-ga151 From 227f052f4711caf432b9a7dbcfe1a2857d3c0def Mon Sep 17 00:00:00 2001 From: Lars Ellenberg Date: Tue, 31 Jul 2012 09:31:11 +0200 Subject: drbd: fix drbd wire compatibility for empty flushes DRBD has a concept of request epochs or reorder-domains, which are separated on the wire by P_BARRIER packets. Older DRBD is not able to handle zero-sized requests at all, so we need to map empty flushes to these drbd barriers. These are the equivalent of empty flushes, and by default trigger flushes on the receiving side anyways (unless not supported or explicitly disabled), so there is no need to handle this differently in newer drbd either. Signed-off-by: Philipp Reisner Signed-off-by: Lars Ellenberg --- drivers/block/drbd/drbd_req.c | 30 ++++++++++++++++++++++++++---- 1 file changed, 26 insertions(+), 4 deletions(-) diff --git a/drivers/block/drbd/drbd_req.c b/drivers/block/drbd/drbd_req.c index 910335c30927..0bb1e41f136f 100644 --- a/drivers/block/drbd/drbd_req.c +++ b/drivers/block/drbd/drbd_req.c @@ -834,7 +834,15 @@ static int drbd_make_request_common(struct drbd_conf *mdev, struct bio *bio, uns req->private_bio = NULL; } if (rw == WRITE) { - remote = 1; + /* Need to replicate writes. Unless it is an empty flush, + * which is better mapped to a DRBD P_BARRIER packet, + * also for drbd wire protocol compatibility reasons. */ + if (unlikely(size == 0)) { + /* The only size==0 bios we expect are empty flushes. */ + D_ASSERT(bio->bi_rw & REQ_FLUSH); + remote = 0; + } else + remote = 1; } else { /* READ || READA */ if (local) { @@ -870,8 +878,11 @@ static int drbd_make_request_common(struct drbd_conf *mdev, struct bio *bio, uns * extent. This waits for any resync activity in the corresponding * resync extent to finish, and, if necessary, pulls in the target * extent into the activity log, which involves further disk io because - * of transactional on-disk meta data updates. */ - if (rw == WRITE && local && !test_bit(AL_SUSPENDED, &mdev->flags)) { + * of transactional on-disk meta data updates. + * Empty flushes don't need to go into the activity log, they can only + * flush data for pending writes which are already in there. */ + if (rw == WRITE && local && size + && !test_bit(AL_SUSPENDED, &mdev->flags)) { req->rq_state |= RQ_IN_ACT_LOG; drbd_al_begin_io(mdev, sector); } @@ -994,7 +1005,10 @@ allocate_barrier: if (rw == WRITE && _req_conflicts(req)) goto fail_conflicting; - list_add_tail(&req->tl_requests, &mdev->newest_tle->requests); + /* no point in adding empty flushes to the transfer log, + * they are mapped to drbd barriers already. */ + if (likely(size!=0)) + list_add_tail(&req->tl_requests, &mdev->newest_tle->requests); /* NOTE remote first: to get the concurrent write detection right, * we must register the request before start of local IO. */ @@ -1014,6 +1028,14 @@ allocate_barrier: mdev->net_conf->on_congestion != OC_BLOCK && mdev->agreed_pro_version >= 96) maybe_pull_ahead(mdev); + /* If this was a flush, queue a drbd barrier/start a new epoch. + * Unless the current epoch was empty anyways, or we are not currently + * replicating, in which case there is no point. */ + if (unlikely(bio->bi_rw & REQ_FLUSH) + && mdev->newest_tle->n_writes + && drbd_should_do_remote(mdev->state)) + queue_barrier(mdev); + spin_unlock_irq(&mdev->req_lock); kfree(b); /* if someone else has beaten us to it... */ -- cgit v1.2.3-58-ga151 From 509fc019e534bdf5f3969d78c53184db4cf7ff48 Mon Sep 17 00:00:00 2001 From: Philipp Reisner Date: Tue, 31 Jul 2012 11:22:58 +0200 Subject: drbd: Finish requests that completed while IO was frozen Requests of an acked epoch are stored on the barrier_acked_requests list. In case the private bio of such a request completes while IO on the drbd device is suspended [req_mod(completed_ok)] then the request stays there. When thawing IO because the fence_peer handler returned, then we use tl_clear() to apply the connection_lost_while_pending event to all requests on the transfer-log and the barrier_acked_requests list. Up to now the connection_lost_while_pending event was not applied on requests on the barrier_acked_requests list. Fixed that. I.e. now the connection_lost_while_pending and resend events are applied to requests on the barrier_acked_requests list. For that it is necessary that the resend event finishes (local only) READS correctly. Signed-off-by: Philipp Reisner Signed-off-by: Lars Ellenberg --- drivers/block/drbd/drbd_main.c | 28 ++++++++++++---------------- drivers/block/drbd/drbd_req.c | 6 ++++++ 2 files changed, 18 insertions(+), 16 deletions(-) diff --git a/drivers/block/drbd/drbd_main.c b/drivers/block/drbd/drbd_main.c index 2e0e7fc1dbba..136db95b212d 100644 --- a/drivers/block/drbd/drbd_main.c +++ b/drivers/block/drbd/drbd_main.c @@ -79,6 +79,7 @@ static int w_md_sync(struct drbd_conf *mdev, struct drbd_work *w, int unused); static void md_sync_timer_fn(unsigned long data); static int w_bitmap_io(struct drbd_conf *mdev, struct drbd_work *w, int unused); static int w_go_diskless(struct drbd_conf *mdev, struct drbd_work *w, int unused); +static void _tl_clear(struct drbd_conf *mdev); MODULE_AUTHOR("Philipp Reisner , " "Lars Ellenberg "); @@ -432,19 +433,10 @@ static void _tl_restart(struct drbd_conf *mdev, enum drbd_req_event what) /* Actions operating on the disk state, also want to work on requests that got barrier acked. */ - switch (what) { - case fail_frozen_disk_io: - case restart_frozen_disk_io: - list_for_each_safe(le, tle, &mdev->barrier_acked_requests) { - req = list_entry(le, struct drbd_request, tl_requests); - _req_mod(req, what); - } - case connection_lost_while_pending: - case resend: - break; - default: - dev_err(DEV, "what = %d in _tl_restart()\n", what); + list_for_each_safe(le, tle, &mdev->barrier_acked_requests) { + req = list_entry(le, struct drbd_request, tl_requests); + _req_mod(req, what); } } @@ -458,12 +450,17 @@ static void _tl_restart(struct drbd_conf *mdev, enum drbd_req_event what) * receiver thread and the worker thread. */ void tl_clear(struct drbd_conf *mdev) +{ + spin_lock_irq(&mdev->req_lock); + _tl_clear(mdev); + spin_unlock_irq(&mdev->req_lock); +} + +static void _tl_clear(struct drbd_conf *mdev) { struct list_head *le, *tle; struct drbd_request *r; - spin_lock_irq(&mdev->req_lock); - _tl_restart(mdev, connection_lost_while_pending); /* we expect this list to be empty. */ @@ -482,7 +479,6 @@ void tl_clear(struct drbd_conf *mdev) memset(mdev->app_reads_hash, 0, APP_R_HSIZE*sizeof(void *)); - spin_unlock_irq(&mdev->req_lock); } void tl_restart(struct drbd_conf *mdev, enum drbd_req_event what) @@ -1476,12 +1472,12 @@ static void after_state_ch(struct drbd_conf *mdev, union drbd_state os, if (ns.susp_fen) { /* case1: The outdate peer handler is successful: */ if (os.pdsk > D_OUTDATED && ns.pdsk <= D_OUTDATED) { - tl_clear(mdev); if (test_bit(NEW_CUR_UUID, &mdev->flags)) { drbd_uuid_new_current(mdev); clear_bit(NEW_CUR_UUID, &mdev->flags); } spin_lock_irq(&mdev->req_lock); + _tl_clear(mdev); _drbd_set_state(_NS(mdev, susp_fen, 0), CS_VERBOSE, NULL); spin_unlock_irq(&mdev->req_lock); } diff --git a/drivers/block/drbd/drbd_req.c b/drivers/block/drbd/drbd_req.c index 0bb1e41f136f..01b2ac641c7b 100644 --- a/drivers/block/drbd/drbd_req.c +++ b/drivers/block/drbd/drbd_req.c @@ -695,6 +695,12 @@ int __req_mod(struct drbd_request *req, enum drbd_req_event what, break; case resend: + /* Simply complete (local only) READs. */ + if (!(req->rq_state & RQ_WRITE) && !req->w.cb) { + _req_may_be_done(req, m); + break; + } + /* If RQ_NET_OK is already set, we got a P_WRITE_ACK or P_RECV_ACK before the connection loss (B&C only); only P_BARRIER_ACK was missing. Trowing them out of the TL here by pretending we got a BARRIER_ACK -- cgit v1.2.3-58-ga151 From d1aa4d04da8de5c89d73859e077d89c4c71d8ed1 Mon Sep 17 00:00:00 2001 From: Philipp Reisner Date: Wed, 8 Aug 2012 21:19:09 +0200 Subject: drbd: Write all pages of the bitmap after an online resize We need to write the whole bitmap after we moved the meta data due to an online resize operation. With the support for one peta byte devices bitmap IO was optimized to only write out touched pages. This optimization must be turned off when writing the bitmap after an online resize. This issue was introduced with drbd-8.3.10. The impact of this bug is that after an online resize, the next resync could become larger than expected. Signed-off-by: Philipp Reisner Signed-off-by: Lars Ellenberg --- drivers/block/drbd/drbd_bitmap.c | 15 ++++++++++++++- drivers/block/drbd/drbd_int.h | 1 + drivers/block/drbd/drbd_nl.c | 4 ++-- 3 files changed, 17 insertions(+), 3 deletions(-) diff --git a/drivers/block/drbd/drbd_bitmap.c b/drivers/block/drbd/drbd_bitmap.c index ba91b408abad..d84566496746 100644 --- a/drivers/block/drbd/drbd_bitmap.c +++ b/drivers/block/drbd/drbd_bitmap.c @@ -889,6 +889,7 @@ struct bm_aio_ctx { unsigned int done; unsigned flags; #define BM_AIO_COPY_PAGES 1 +#define BM_WRITE_ALL_PAGES 2 int error; struct kref kref; }; @@ -1059,7 +1060,8 @@ static int bm_rw(struct drbd_conf *mdev, int rw, unsigned flags, unsigned lazy_w if (lazy_writeout_upper_idx && i == lazy_writeout_upper_idx) break; if (rw & WRITE) { - if (bm_test_page_unchanged(b->bm_pages[i])) { + if (!(flags & BM_WRITE_ALL_PAGES) && + bm_test_page_unchanged(b->bm_pages[i])) { dynamic_dev_dbg(DEV, "skipped bm write for idx %u\n", i); continue; } @@ -1140,6 +1142,17 @@ int drbd_bm_write(struct drbd_conf *mdev) __must_hold(local) return bm_rw(mdev, WRITE, 0, 0); } +/** + * drbd_bm_write_all() - Write the whole bitmap to its on disk location. + * @mdev: DRBD device. + * + * Will write all pages. + */ +int drbd_bm_write_all(struct drbd_conf *mdev) __must_hold(local) +{ + return bm_rw(mdev, WRITE, BM_WRITE_ALL_PAGES, 0); +} + /** * drbd_bm_lazy_write_out() - Write bitmap pages 0 to @upper_idx-1, if they have changed. * @mdev: DRBD device. diff --git a/drivers/block/drbd/drbd_int.h b/drivers/block/drbd/drbd_int.h index b2ca143d0053..b953cc7c9c00 100644 --- a/drivers/block/drbd/drbd_int.h +++ b/drivers/block/drbd/drbd_int.h @@ -1469,6 +1469,7 @@ extern int drbd_bm_e_weight(struct drbd_conf *mdev, unsigned long enr); extern int drbd_bm_write_page(struct drbd_conf *mdev, unsigned int idx) __must_hold(local); extern int drbd_bm_read(struct drbd_conf *mdev) __must_hold(local); extern int drbd_bm_write(struct drbd_conf *mdev) __must_hold(local); +extern int drbd_bm_write_all(struct drbd_conf *mdev) __must_hold(local); extern int drbd_bm_write_copy_pages(struct drbd_conf *mdev) __must_hold(local); extern unsigned long drbd_bm_ALe_set_all(struct drbd_conf *mdev, unsigned long al_enr); diff --git a/drivers/block/drbd/drbd_nl.c b/drivers/block/drbd/drbd_nl.c index fb9dce8daa24..edb490aad8b4 100644 --- a/drivers/block/drbd/drbd_nl.c +++ b/drivers/block/drbd/drbd_nl.c @@ -674,8 +674,8 @@ enum determine_dev_size drbd_determine_dev_size(struct drbd_conf *mdev, enum dds la_size_changed && md_moved ? "size changed and md moved" : la_size_changed ? "size changed" : "md moved"); /* next line implicitly does drbd_suspend_io()+drbd_resume_io() */ - err = drbd_bitmap_io(mdev, &drbd_bm_write, - "size changed", BM_LOCKED_MASK); + err = drbd_bitmap_io(mdev, md_moved ? &drbd_bm_write_all : &drbd_bm_write, + "size changed", BM_LOCKED_MASK); if (err) { rv = dev_size_error; goto out; -- cgit v1.2.3-58-ga151 From 676ce6d5ca3098339c028d44fe0427d1566a4d2d Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Thu, 23 Aug 2012 12:17:36 +0200 Subject: block: replace __getblk_slow misfix by grow_dev_page fix Commit 91f68c89d8f3 ("block: fix infinite loop in __getblk_slow") is not good: a successful call to grow_buffers() cannot guarantee that the page won't be reclaimed before the immediate next call to __find_get_block(), which is why there was always a loop there. Yesterday I got "EXT4-fs error (device loop0): __ext4_get_inode_loc:3595: inode #19278: block 664: comm cc1: unable to read itable block" on console, which pointed to this commit. I've been trying to bisect for weeks, why kbuild-on-ext4-on-loop-on-tmpfs sometimes fails from a missing header file, under memory pressure on ppc G5. I've never seen this on x86, and I've never seen it on 3.5-rc7 itself, despite that commit being in there: bisection pointed to an irrelevant pinctrl merge, but hard to tell when failure takes between 18 minutes and 38 hours (but so far it's happened quicker on 3.6-rc2). (I've since found such __ext4_get_inode_loc errors in /var/log/messages from previous weeks: why the message never appeared on console until yesterday morning is a mystery for another day.) Revert 91f68c89d8f3, restoring __getblk_slow() to how it was (plus a checkpatch nitfix). Simplify the interface between grow_buffers() and grow_dev_page(), and avoid the infinite loop beyond end of device by instead checking init_page_buffers()'s end_block there (I presume that's more efficient than a repeated call to blkdev_max_block()), returning -ENXIO to __getblk_slow() in that case. And remove akpm's ten-year-old "__getblk() cannot fail ... weird" comment, but that is worrying: are all users of __getblk() really now prepared for a NULL bh beyond end of device, or will some oops?? Signed-off-by: Hugh Dickins Cc: stable@vger.kernel.org # 3.0 3.2 3.4 3.5 Signed-off-by: Jens Axboe --- fs/buffer.c | 66 ++++++++++++++++++++++++++++--------------------------------- 1 file changed, 30 insertions(+), 36 deletions(-) diff --git a/fs/buffer.c b/fs/buffer.c index 9f6d2e41281d..58e2e7b77372 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -914,7 +914,7 @@ link_dev_buffers(struct page *page, struct buffer_head *head) /* * Initialise the state of a blockdev page's buffers. */ -static void +static sector_t init_page_buffers(struct page *page, struct block_device *bdev, sector_t block, int size) { @@ -936,33 +936,41 @@ init_page_buffers(struct page *page, struct block_device *bdev, block++; bh = bh->b_this_page; } while (bh != head); + + /* + * Caller needs to validate requested block against end of device. + */ + return end_block; } /* * Create the page-cache page that contains the requested block. * - * This is user purely for blockdev mappings. + * This is used purely for blockdev mappings. */ -static struct page * +static int grow_dev_page(struct block_device *bdev, sector_t block, - pgoff_t index, int size) + pgoff_t index, int size, int sizebits) { struct inode *inode = bdev->bd_inode; struct page *page; struct buffer_head *bh; + sector_t end_block; + int ret = 0; /* Will call free_more_memory() */ page = find_or_create_page(inode->i_mapping, index, (mapping_gfp_mask(inode->i_mapping) & ~__GFP_FS)|__GFP_MOVABLE); if (!page) - return NULL; + return ret; BUG_ON(!PageLocked(page)); if (page_has_buffers(page)) { bh = page_buffers(page); if (bh->b_size == size) { - init_page_buffers(page, bdev, block, size); - return page; + end_block = init_page_buffers(page, bdev, + index << sizebits, size); + goto done; } if (!try_to_free_buffers(page)) goto failed; @@ -982,14 +990,14 @@ grow_dev_page(struct block_device *bdev, sector_t block, */ spin_lock(&inode->i_mapping->private_lock); link_dev_buffers(page, bh); - init_page_buffers(page, bdev, block, size); + end_block = init_page_buffers(page, bdev, index << sizebits, size); spin_unlock(&inode->i_mapping->private_lock); - return page; - +done: + ret = (block < end_block) ? 1 : -ENXIO; failed: unlock_page(page); page_cache_release(page); - return NULL; + return ret; } /* @@ -999,7 +1007,6 @@ failed: static int grow_buffers(struct block_device *bdev, sector_t block, int size) { - struct page *page; pgoff_t index; int sizebits; @@ -1023,22 +1030,14 @@ grow_buffers(struct block_device *bdev, sector_t block, int size) bdevname(bdev, b)); return -EIO; } - block = index << sizebits; + /* Create a page with the proper size buffers.. */ - page = grow_dev_page(bdev, block, index, size); - if (!page) - return 0; - unlock_page(page); - page_cache_release(page); - return 1; + return grow_dev_page(bdev, block, index, size, sizebits); } static struct buffer_head * __getblk_slow(struct block_device *bdev, sector_t block, int size) { - int ret; - struct buffer_head *bh; - /* Size must be multiple of hard sectorsize */ if (unlikely(size & (bdev_logical_block_size(bdev)-1) || (size < 512 || size > PAGE_SIZE))) { @@ -1051,21 +1050,20 @@ __getblk_slow(struct block_device *bdev, sector_t block, int size) return NULL; } -retry: - bh = __find_get_block(bdev, block, size); - if (bh) - return bh; + for (;;) { + struct buffer_head *bh; + int ret; - ret = grow_buffers(bdev, block, size); - if (ret == 0) { - free_more_memory(); - goto retry; - } else if (ret > 0) { bh = __find_get_block(bdev, block, size); if (bh) return bh; + + ret = grow_buffers(bdev, block, size); + if (ret < 0) + return NULL; + if (ret == 0) + free_more_memory(); } - return NULL; } /* @@ -1321,10 +1319,6 @@ EXPORT_SYMBOL(__find_get_block); * which corresponds to the passed block_device, block and size. The * returned buffer has its reference count incremented. * - * __getblk() cannot fail - it just keeps trying. If you pass it an - * illegal block number, __getblk() will happily return a buffer_head - * which represents the non-existent block. Very weird. - * * __getblk() will lock up the machine if grow_dev_page's try_to_free_buffers() * attempt is failing. FIXME, perhaps? */ -- cgit v1.2.3-58-ga151