From 9283b42e46c2646dff1bec47e2dd683add7f9972 Mon Sep 17 00:00:00 2001 From: Keith Busch Date: Thu, 23 Apr 2015 10:24:45 -0600 Subject: NVMe: Fix VPD B0 max sectors translation Use the namespace's block format for reporting the max transfer length. Max unmap count is left as-is since NVMe doesn't provide a max, so the value the driver provided the block layer is valid for any format. Reported-by: Martin K. Petersen Signed-off-by: Keith Busch Signed-off-by: Jens Axboe --- drivers/block/nvme-scsi.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'drivers/block') diff --git a/drivers/block/nvme-scsi.c b/drivers/block/nvme-scsi.c index 6b736b00f63e..88f13c525712 100644 --- a/drivers/block/nvme-scsi.c +++ b/drivers/block/nvme-scsi.c @@ -944,7 +944,8 @@ static int nvme_trans_ext_inq_page(struct nvme_ns *ns, struct sg_io_hdr *hdr, static int nvme_trans_bdev_limits_page(struct nvme_ns *ns, struct sg_io_hdr *hdr, u8 *inq_response, int alloc_len) { - __be32 max_sectors = cpu_to_be32(queue_max_hw_sectors(ns->queue)); + __be32 max_sectors = cpu_to_be32( + nvme_block_nr(ns, queue_max_hw_sectors(ns->queue))); __be32 max_discard = cpu_to_be32(ns->queue->limits.max_discard_sectors); __be32 discard_desc_count = cpu_to_be32(0x100); -- cgit v1.2.3 From 325d73bf8fea8af2227240b7305253fb052d3a68 Mon Sep 17 00:00:00 2001 From: Bob Liu Date: Fri, 3 Apr 2015 14:42:58 +0800 Subject: xen/blkback: safely unmap purge persistent grants MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Commit c43cf3ea8385 ("xen-blkback: safely unmap grants in case they are still in use") use gnttab_unmap_refs_async() to wait until the mapped pages are no longer in use before unmapping them, but that commit missed the persistent case. Purge persistent pages can't be unmapped either unless no longer in use. Signed-off-by: Bob Liu Acked-by: Roger Pau Monné Signed-off-by: David Vrabel --- drivers/block/xen-blkback/blkback.c | 24 ++++++++++++++++++------ 1 file changed, 18 insertions(+), 6 deletions(-) (limited to 'drivers/block') diff --git a/drivers/block/xen-blkback/blkback.c b/drivers/block/xen-blkback/blkback.c index bd2b3bbbb22c..48e98f2712b5 100644 --- a/drivers/block/xen-blkback/blkback.c +++ b/drivers/block/xen-blkback/blkback.c @@ -329,8 +329,18 @@ void xen_blkbk_unmap_purged_grants(struct work_struct *work) struct gnttab_unmap_grant_ref unmap[BLKIF_MAX_SEGMENTS_PER_REQUEST]; struct page *pages[BLKIF_MAX_SEGMENTS_PER_REQUEST]; struct persistent_gnt *persistent_gnt; - int ret, segs_to_unmap = 0; + int segs_to_unmap = 0; struct xen_blkif *blkif = container_of(work, typeof(*blkif), persistent_purge_work); + struct gntab_unmap_queue_data unmap_data; + struct completion unmap_completion; + + init_completion(&unmap_completion); + + unmap_data.data = &unmap_completion; + unmap_data.done = &free_persistent_gnts_unmap_callback; + unmap_data.pages = pages; + unmap_data.unmap_ops = unmap; + unmap_data.kunmap_ops = NULL; while(!list_empty(&blkif->persistent_purge_list)) { persistent_gnt = list_first_entry(&blkif->persistent_purge_list, @@ -346,17 +356,19 @@ void xen_blkbk_unmap_purged_grants(struct work_struct *work) pages[segs_to_unmap] = persistent_gnt->page; if (++segs_to_unmap == BLKIF_MAX_SEGMENTS_PER_REQUEST) { - ret = gnttab_unmap_refs(unmap, NULL, pages, - segs_to_unmap); - BUG_ON(ret); + unmap_data.count = segs_to_unmap; + gnttab_unmap_refs_async(&unmap_data); + wait_for_completion(&unmap_completion); + put_free_pages(blkif, pages, segs_to_unmap); segs_to_unmap = 0; } kfree(persistent_gnt); } if (segs_to_unmap > 0) { - ret = gnttab_unmap_refs(unmap, NULL, pages, segs_to_unmap); - BUG_ON(ret); + unmap_data.count = segs_to_unmap; + gnttab_unmap_refs_async(&unmap_data); + wait_for_completion(&unmap_completion); put_free_pages(blkif, pages, segs_to_unmap); } } -- cgit v1.2.3 From b44166cd46e28dd608d5baa5873047a40f32919c Mon Sep 17 00:00:00 2001 From: Bob Liu Date: Fri, 3 Apr 2015 14:42:59 +0800 Subject: xen/grant: introduce func gnttab_unmap_refs_sync() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit There are several place using gnttab async unmap and wait for completion, so move the common code to a function gnttab_unmap_refs_sync(). Signed-off-by: Bob Liu Acked-by: Roger Pau Monné Acked-by: Konrad Rzeszutek Wilk Signed-off-by: David Vrabel --- drivers/block/xen-blkback/blkback.c | 31 +++---------------------------- 1 file changed, 3 insertions(+), 28 deletions(-) (limited to 'drivers/block') diff --git a/drivers/block/xen-blkback/blkback.c b/drivers/block/xen-blkback/blkback.c index 48e98f2712b5..713fc9ff1149 100644 --- a/drivers/block/xen-blkback/blkback.c +++ b/drivers/block/xen-blkback/blkback.c @@ -265,17 +265,6 @@ static void put_persistent_gnt(struct xen_blkif *blkif, atomic_dec(&blkif->persistent_gnt_in_use); } -static void free_persistent_gnts_unmap_callback(int result, - struct gntab_unmap_queue_data *data) -{ - struct completion *c = data->data; - - /* BUG_ON used to reproduce existing behaviour, - but is this the best way to deal with this? */ - BUG_ON(result); - complete(c); -} - static void free_persistent_gnts(struct xen_blkif *blkif, struct rb_root *root, unsigned int num) { @@ -285,12 +274,7 @@ static void free_persistent_gnts(struct xen_blkif *blkif, struct rb_root *root, struct rb_node *n; int segs_to_unmap = 0; struct gntab_unmap_queue_data unmap_data; - struct completion unmap_completion; - - init_completion(&unmap_completion); - unmap_data.data = &unmap_completion; - unmap_data.done = &free_persistent_gnts_unmap_callback; unmap_data.pages = pages; unmap_data.unmap_ops = unmap; unmap_data.kunmap_ops = NULL; @@ -310,8 +294,7 @@ static void free_persistent_gnts(struct xen_blkif *blkif, struct rb_root *root, !rb_next(&persistent_gnt->node)) { unmap_data.count = segs_to_unmap; - gnttab_unmap_refs_async(&unmap_data); - wait_for_completion(&unmap_completion); + BUG_ON(gnttab_unmap_refs_sync(&unmap_data)); put_free_pages(blkif, pages, segs_to_unmap); segs_to_unmap = 0; @@ -332,12 +315,7 @@ void xen_blkbk_unmap_purged_grants(struct work_struct *work) int segs_to_unmap = 0; struct xen_blkif *blkif = container_of(work, typeof(*blkif), persistent_purge_work); struct gntab_unmap_queue_data unmap_data; - struct completion unmap_completion; - init_completion(&unmap_completion); - - unmap_data.data = &unmap_completion; - unmap_data.done = &free_persistent_gnts_unmap_callback; unmap_data.pages = pages; unmap_data.unmap_ops = unmap; unmap_data.kunmap_ops = NULL; @@ -357,9 +335,7 @@ void xen_blkbk_unmap_purged_grants(struct work_struct *work) if (++segs_to_unmap == BLKIF_MAX_SEGMENTS_PER_REQUEST) { unmap_data.count = segs_to_unmap; - gnttab_unmap_refs_async(&unmap_data); - wait_for_completion(&unmap_completion); - + BUG_ON(gnttab_unmap_refs_sync(&unmap_data)); put_free_pages(blkif, pages, segs_to_unmap); segs_to_unmap = 0; } @@ -367,8 +343,7 @@ void xen_blkbk_unmap_purged_grants(struct work_struct *work) } if (segs_to_unmap > 0) { unmap_data.count = segs_to_unmap; - gnttab_unmap_refs_async(&unmap_data); - wait_for_completion(&unmap_completion); + BUG_ON(gnttab_unmap_refs_sync(&unmap_data)); put_free_pages(blkif, pages, segs_to_unmap); } } -- cgit v1.2.3 From 6cd18e711dd8075da9d78cfc1239f912ff28968a Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Mon, 27 Apr 2015 14:12:22 +1000 Subject: block: destroy bdi before blockdev is unregistered. Because of the peculiar way that md devices are created (automatically when the device node is opened), a new device can be created and registered immediately after the blk_unregister_region(disk_devt(disk), disk->minors); call in del_gendisk(). Therefore it is important that all visible artifacts of the previous device are removed before this call. In particular, the 'bdi'. Since: commit c4db59d31e39ea067c32163ac961e9c80198fd37 Author: Christoph Hellwig fs: don't reassign dirty inodes to default_backing_dev_info moved the device_unregister(bdi->dev); call from bdi_unregister() to bdi_destroy() it has been quite easy to lose a race and have a new (e.g.) "md127" be created after the blk_unregister_region() call and before bdi_destroy() is ultimately called by the final 'put_disk', which must come after del_gendisk(). The new device finds that the bdi name is already registered in sysfs and complains > [ 9627.630029] WARNING: CPU: 18 PID: 3330 at fs/sysfs/dir.c:31 sysfs_warn_dup+0x5a/0x70() > [ 9627.630032] sysfs: cannot create duplicate filename '/devices/virtual/bdi/9:127' We can fix this by moving the bdi_destroy() call out of blk_release_queue() (which can happen very late when a refcount reaches zero) and into blk_cleanup_queue() - which happens exactly when the md device driver calls it. Then it is only necessary for md to call blk_cleanup_queue() before del_gendisk(). As loop.c devices are also created on demand by opening the device node, we make the same change there. Fixes: c4db59d31e39ea067c32163ac961e9c80198fd37 Reported-by: Azat Khuzhin Cc: Christoph Hellwig Cc: stable@vger.kernel.org (v4.0) Signed-off-by: NeilBrown Reviewed-by: Christoph Hellwig Signed-off-by: Jens Axboe --- drivers/block/loop.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/block') diff --git a/drivers/block/loop.c b/drivers/block/loop.c index ae3fcb4199e9..d7173cb1ea76 100644 --- a/drivers/block/loop.c +++ b/drivers/block/loop.c @@ -1620,8 +1620,8 @@ out: static void loop_remove(struct loop_device *lo) { - del_gendisk(lo->lo_disk); blk_cleanup_queue(lo->lo_queue); + del_gendisk(lo->lo_disk); blk_mq_free_tag_set(&lo->tag_set); put_disk(lo->lo_disk); kfree(lo); -- cgit v1.2.3 From 082a75dad84d79d1c15ea9e50f31cb4bb4fa7fd6 Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Sat, 25 Apr 2015 15:56:15 +0300 Subject: rbd: end I/O the entire obj_request on error When we end I/O struct request with error, we need to pass obj_request->length as @nr_bytes so that the entire obj_request worth of bytes is completed. Otherwise block layer ends up confused and we trip on rbd_assert(more ^ (which == img_request->obj_request_count)); in rbd_img_obj_callback() due to more being true no matter what. We already do it in most cases but we are missing some, in particular those where we don't even get a chance to submit any obj_requests, due to an early -ENOMEM for example. A number of obj_request->xferred assignments seem to be redundant but I haven't touched any of obj_request->xferred stuff to keep this small and isolated. Cc: Alex Elder Cc: stable@vger.kernel.org # 3.10+ Reported-by: Shawn Edwards Reviewed-by: Sage Weil Signed-off-by: Ilya Dryomov --- drivers/block/rbd.c | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'drivers/block') diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index 812523330a78..ec6c5c6e1ac9 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -2264,6 +2264,11 @@ static bool rbd_img_obj_end_request(struct rbd_obj_request *obj_request) result, xferred); if (!img_request->result) img_request->result = result; + /* + * Need to end I/O on the entire obj_request worth of + * bytes in case of error. + */ + xferred = obj_request->length; } /* Image object requests don't own their page array */ -- cgit v1.2.3 From 99ebbd30e3640f6addb37f222b4d6ad4b807d9ea Mon Sep 17 00:00:00 2001 From: Andrew Morton Date: Tue, 5 May 2015 16:23:25 -0700 Subject: revert "zram: move compact_store() to sysfs functions area" Revert commit c72c6160d967ed26a0b136dbab337f821d233509 It was intended to be a cosmetic change that w/o any functional change and was part of a bigger change: http://lkml.iu.edu/hypermail/linux/kernel/1503.1/01818.html Sergey Senozhatsky Cc: Linus Torvalds Cc: Minchan Kim Cc: Nitin Gupta Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/block/zram/zram_drv.c | 23 +++++++++++++++++++++++ 1 file changed, 23 insertions(+) (limited to 'drivers/block') diff --git a/drivers/block/zram/zram_drv.c b/drivers/block/zram/zram_drv.c index c94386aa563d..8dcbced0eafd 100644 --- a/drivers/block/zram/zram_drv.c +++ b/drivers/block/zram/zram_drv.c @@ -74,6 +74,27 @@ static inline struct zram *dev_to_zram(struct device *dev) return (struct zram *)dev_to_disk(dev)->private_data; } +static ssize_t compact_store(struct device *dev, + struct device_attribute *attr, const char *buf, size_t len) +{ + unsigned long nr_migrated; + struct zram *zram = dev_to_zram(dev); + struct zram_meta *meta; + + down_read(&zram->init_lock); + if (!init_done(zram)) { + up_read(&zram->init_lock); + return -EINVAL; + } + + meta = zram->meta; + nr_migrated = zs_compact(meta->mem_pool); + atomic64_add(nr_migrated, &zram->stats.num_migrated); + up_read(&zram->init_lock); + + return len; +} + static ssize_t disksize_show(struct device *dev, struct device_attribute *attr, char *buf) { @@ -1038,6 +1059,7 @@ static const struct block_device_operations zram_devops = { .owner = THIS_MODULE }; +static DEVICE_ATTR_WO(compact); static DEVICE_ATTR_RW(disksize); static DEVICE_ATTR_RO(initstate); static DEVICE_ATTR_WO(reset); @@ -1114,6 +1136,7 @@ static struct attribute *zram_disk_attrs[] = { &dev_attr_num_writes.attr, &dev_attr_failed_reads.attr, &dev_attr_failed_writes.attr, + &dev_attr_compact.attr, &dev_attr_invalid_io.attr, &dev_attr_notify_free.attr, &dev_attr_zero_pages.attr, -- cgit v1.2.3