From 90519d66abbccc251d14719ac76f191f70826e40 Mon Sep 17 00:00:00 2001 From: Arne Jansen Date: Mon, 23 May 2011 14:30:00 +0200 Subject: btrfs: state information for readahead Add state information for readahead to btrfs_fs_info and btrfs_device Changes v2: - don't wait in radix_trees - add own set of workers for readahead Reviewed-by: Josef Bacik Signed-off-by: Arne Jansen --- fs/btrfs/ctree.h | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'fs/btrfs/ctree.h') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 03912c5c6f49..f71fd24cc152 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -1036,6 +1036,7 @@ struct btrfs_fs_info { struct btrfs_workers endio_freespace_worker; struct btrfs_workers submit_workers; struct btrfs_workers caching_workers; + struct btrfs_workers readahead_workers; /* * fixup workers take dirty pages that didn't properly go through @@ -1119,6 +1120,10 @@ struct btrfs_fs_info { u64 fs_state; struct btrfs_delayed_root *delayed_root; + + /* readahead tree */ + spinlock_t reada_lock; + struct radix_tree_root reada_tree; }; /* -- cgit v1.2.3 From 7414a03fbf9e75fbbf2a3c16828cd862e572aa44 Mon Sep 17 00:00:00 2001 From: Arne Jansen Date: Mon, 23 May 2011 14:33:49 +0200 Subject: btrfs: initial readahead code and prototypes This is the implementation for the generic read ahead framework. To trigger a readahead, btrfs_reada_add must be called. It will start a read ahead for the given range [start, end) on tree root. The returned handle can either be used to wait on the readahead to finish (btrfs_reada_wait), or to send it to the background (btrfs_reada_detach). The read ahead works as follows: On btrfs_reada_add, the root of the tree is inserted into a radix_tree. reada_start_machine will then search for extents to prefetch and trigger some reads. When a read finishes for a node, all contained node/leaf pointers that lie in the given range will also be enqueued. The reads will be triggered in sequential order, thus giving a big win over a naive enumeration. It will also make use of multi-device layouts. Each disk will have its on read pointer and all disks will by utilized in parallel. Also will no two disks read both sides of a mirror simultaneously, as this would waste seeking capacity. Instead both disks will read different parts of the filesystem. Any number of readaheads can be started in parallel. The read order will be determined globally, i.e. 2 parallel readaheads will normally finish faster than the 2 started one after another. Changes v2: - protect root->node by transaction instead of node_lock - fix missed branches: The readahead had a too simple check to determine if a branch from a node should be checked or not. It now also records the upper bound of each node to see if the requested RA range lies within. - use KERN_CONT to debug output, to avoid line breaks - defer reada_start_machine to worker to avoid deadlock Changes v3: - protect root->node by rcu Changes v5: - changed EIO-semantics of reada_tree_block_flagged - remove spin_lock from reada_control and make elems an atomic_t - remove unused read_total from reada_control - kill reada_key_cmp, use btrfs_comp_cpu_keys instead - use kref-style release functions where possible - return struct reada_control * instead of void * from btrfs_reada_add Signed-off-by: Arne Jansen --- fs/btrfs/ctree.h | 16 ++++++++++++++++ 1 file changed, 16 insertions(+) (limited to 'fs/btrfs/ctree.h') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index f71fd24cc152..370af767440d 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -2702,4 +2702,20 @@ int btrfs_scrub_cancel_devid(struct btrfs_root *root, u64 devid); int btrfs_scrub_progress(struct btrfs_root *root, u64 devid, struct btrfs_scrub_progress *progress); +/* reada.c */ +struct reada_control { + struct btrfs_root *root; /* tree to prefetch */ + struct btrfs_key key_start; + struct btrfs_key key_end; /* exclusive */ + atomic_t elems; + struct kref refcnt; + wait_queue_head_t wait; +}; +struct reada_control *btrfs_reada_add(struct btrfs_root *root, + struct btrfs_key *start, struct btrfs_key *end); +int btrfs_reada_wait(void *handle); +void btrfs_reada_detach(void *handle); +int btree_readahead_hook(struct btrfs_root *root, struct extent_buffer *eb, + u64 start, int err); + #endif -- cgit v1.2.3 From fb25e9141ab843794d5cdef3936ccb58435e2371 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Tue, 26 Jul 2011 17:00:46 -0400 Subject: Btrfs: use bytes_may_use for all ENOSPC reservations We have been using bytes_reserved for metadata reservations, which is wrong since we use that to keep track of outstanding reservations from the allocator. This resulted in us doing a lot of silly things to make sure we don't allocate a bunch of metadata chunks since we never had a real view of how much space was actually in use by metadata. This passes Arne's enospc test and xfstests as well as my own enospc tests. Hopefully this will get us moving in the right direction. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/ctree.h | 2 -- 1 file changed, 2 deletions(-) (limited to 'fs/btrfs/ctree.h') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 03912c5c6f49..332cbdc86ad4 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -2196,8 +2196,6 @@ int btrfs_free_extent(struct btrfs_trans_handle *trans, u64 root_objectid, u64 owner, u64 offset); int btrfs_free_reserved_extent(struct btrfs_root *root, u64 start, u64 len); -int btrfs_update_reserved_bytes(struct btrfs_block_group_cache *cache, - u64 num_bytes, int reserve, int sinfo); int btrfs_prepare_extent_commit(struct btrfs_trans_handle *trans, struct btrfs_root *root); int btrfs_finish_extent_commit(struct btrfs_trans_handle *trans, -- cgit v1.2.3 From dba68306f3fae681b1005137f130f5bcfdfed34a Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Thu, 4 Aug 2011 15:34:57 -0400 Subject: Btrfs: kill the orphan space calculation for snapshots This patch kills off the calculation for the amount of space needed for the orphan operations during a snapshot. The thing is we only do snapshots on commit, so any space that is in the block_rsv->freed[] isn't going to be in the new snapshot anyway, so there isn't any reason to require that space to be reserved for the snapshot to occur. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/ctree.h | 5 ----- 1 file changed, 5 deletions(-) (limited to 'fs/btrfs/ctree.h') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 332cbdc86ad4..4ef777e85c89 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -2577,11 +2577,6 @@ int btrfs_update_inode(struct btrfs_trans_handle *trans, int btrfs_orphan_add(struct btrfs_trans_handle *trans, struct inode *inode); int btrfs_orphan_del(struct btrfs_trans_handle *trans, struct inode *inode); int btrfs_orphan_cleanup(struct btrfs_root *root); -void btrfs_orphan_pre_snapshot(struct btrfs_trans_handle *trans, - struct btrfs_pending_snapshot *pending, - u64 *bytes_to_reserve); -void btrfs_orphan_post_snapshot(struct btrfs_trans_handle *trans, - struct btrfs_pending_snapshot *pending); void btrfs_orphan_commit_root(struct btrfs_trans_handle *trans, struct btrfs_root *root); int btrfs_cont_expand(struct inode *inode, loff_t oldsize, loff_t size); -- cgit v1.2.3 From 37be25bcb6d731914e126f8de59c4367f0d66b80 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Fri, 5 Aug 2011 10:25:38 -0400 Subject: Btrfs: kill the durable block rsv stuff This is confusing code and isn't used by anything anymore, so delete it. Signed-off-by: Josef Bacik --- fs/btrfs/ctree.h | 11 ----------- 1 file changed, 11 deletions(-) (limited to 'fs/btrfs/ctree.h') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 4ef777e85c89..c5ceba4078cc 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -772,13 +772,10 @@ struct btrfs_space_info { struct btrfs_block_rsv { u64 size; u64 reserved; - u64 freed[2]; struct btrfs_space_info *space_info; - struct list_head list; spinlock_t lock; atomic_t usage; unsigned int priority:8; - unsigned int durable:1; unsigned int refill_used:1; unsigned int full:1; }; @@ -840,7 +837,6 @@ struct btrfs_block_group_cache { spinlock_t lock; u64 pinned; u64 reserved; - u64 reserved_pinned; u64 bytes_super; u64 flags; u64 sectorsize; @@ -919,11 +915,6 @@ struct btrfs_fs_info { struct btrfs_block_rsv empty_block_rsv; - /* list of block reservations that cross multiple transactions */ - struct list_head durable_block_rsv_list; - - struct mutex durable_block_rsv_mutex; - u64 generation; u64 last_trans_committed; @@ -2238,8 +2229,6 @@ void btrfs_init_block_rsv(struct btrfs_block_rsv *rsv); struct btrfs_block_rsv *btrfs_alloc_block_rsv(struct btrfs_root *root); void btrfs_free_block_rsv(struct btrfs_root *root, struct btrfs_block_rsv *rsv); -void btrfs_add_durable_block_rsv(struct btrfs_fs_info *fs_info, - struct btrfs_block_rsv *rsv); int btrfs_block_rsv_add(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct btrfs_block_rsv *block_rsv, -- cgit v1.2.3 From dabdb6408cb801644fa613c7432da012640b348c Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Mon, 8 Aug 2011 12:50:18 -0400 Subject: Btrfs: kill unused parts of block_rsv The priority and refill_used flags are not used anymore, and neither is the usage counter, so just remove them from btrfs_block_rsv. Signed-off-by: Josef Bacik --- fs/btrfs/ctree.h | 3 --- 1 file changed, 3 deletions(-) (limited to 'fs/btrfs/ctree.h') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index c5ceba4078cc..58a06dea4791 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -774,9 +774,6 @@ struct btrfs_block_rsv { u64 reserved; struct btrfs_space_info *space_info; spinlock_t lock; - atomic_t usage; - unsigned int priority:8; - unsigned int refill_used:1; unsigned int full:1; }; -- cgit v1.2.3 From 5e962c7850c273b483acc747b41bd5cddf631049 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Mon, 8 Aug 2011 14:03:37 -0400 Subject: Btrfs: kill btrfs_truncate_reserve_metadata Since we've optimized the truncate path, we no longer require this function. Signed-off-by: Josef Bacik --- fs/btrfs/ctree.h | 3 --- 1 file changed, 3 deletions(-) (limited to 'fs/btrfs/ctree.h') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 58a06dea4791..22a9347a3908 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -2240,9 +2240,6 @@ int btrfs_block_rsv_migrate(struct btrfs_block_rsv *src_rsv, void btrfs_block_rsv_release(struct btrfs_root *root, struct btrfs_block_rsv *block_rsv, u64 num_bytes); -int btrfs_truncate_reserve_metadata(struct btrfs_trans_handle *trans, - struct btrfs_root *root, - struct btrfs_block_rsv *rsv); int btrfs_set_block_group_ro(struct btrfs_root *root, struct btrfs_block_group_cache *cache); int btrfs_set_block_group_rw(struct btrfs_root *root, -- cgit v1.2.3 From 07127184efb629f1336c0592bfdacec258cab731 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Fri, 19 Aug 2011 10:29:59 -0400 Subject: Btrfs: reduce the amount of space needed for truncates With btrfs_truncate_inode_items we always return if we have to go to another leaf, which makes us do our reservation again. This means we will only ever modify one leaf at a time, so we only need 1 items worth of slack space. Also, since we are deleting we will not be creating nodes as we go down, if anything we'll be free'ing them as we merge them together, so make a different calculation for truncate which will only have the worst case useage of COW'ing the entire path down to the leaf. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/ctree.h | 11 +++++++++++ 1 file changed, 11 insertions(+) (limited to 'fs/btrfs/ctree.h') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 22a9347a3908..2e18b068841b 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -2125,6 +2125,17 @@ static inline u64 btrfs_calc_trans_metadata_size(struct btrfs_root *root, 3 * num_items; } +/* + * Doing a truncate won't result in new nodes or leaves, just what we need for + * COW. + */ +static inline u64 btrfs_calc_trunc_metadata_size(struct btrfs_root *root, + unsigned num_items) +{ + return (root->leafsize + root->nodesize * (BTRFS_MAX_LEVEL - 1)) * + num_items; +} + void btrfs_put_block_group(struct btrfs_block_group_cache *cache); int btrfs_run_delayed_refs(struct btrfs_trans_handle *trans, struct btrfs_root *root, unsigned long count); -- cgit v1.2.3 From 482e6dc5261406fdb921946e70b51467b0305bad Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Fri, 19 Aug 2011 10:31:56 -0400 Subject: Btrfs: allow callers to specify if flushing can occur for btrfs_block_rsv_check If you run xfstest 224 it you will get lots of messages about not being able to delete inodes and that they will be cleaned up next mount. This is because btrfs_block_rsv_check was not calling reserve_metadata_bytes with the ability to flush, so if there was not enough space, it simply failed. But in truncate and evict case we could easily flush space to try and get enough space to do our work, so make btrfs_block_rsv_check take a flush argument to pass down to reserve_metadata_bytes. Now xfstests 224 runs fine without all those complaints. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/ctree.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs/btrfs/ctree.h') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 2e18b068841b..caa73cd8c00a 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -2244,7 +2244,7 @@ int btrfs_block_rsv_add(struct btrfs_trans_handle *trans, int btrfs_block_rsv_check(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct btrfs_block_rsv *block_rsv, - u64 min_reserved, int min_factor); + u64 min_reserved, int min_factor, int flush); int btrfs_block_rsv_migrate(struct btrfs_block_rsv *src_rsv, struct btrfs_block_rsv *dst_rsv, u64 num_bytes); -- cgit v1.2.3 From 4a92b1b8d2810db4ea0c34616b94c0b3810fa027 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Tue, 30 Aug 2011 12:34:28 -0400 Subject: Btrfs: stop passing a trans handle all around the reservation code The only thing that we need to have a trans handle for is in reserve_metadata_bytes and thats to know how much flushing we can do. So instead of passing it around, just check current->journal_info for a trans_handle so we know if we can commit a transaction to try and free up space or not. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/ctree.h | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) (limited to 'fs/btrfs/ctree.h') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index caa73cd8c00a..a5faf8e33baa 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -2237,12 +2237,10 @@ void btrfs_init_block_rsv(struct btrfs_block_rsv *rsv); struct btrfs_block_rsv *btrfs_alloc_block_rsv(struct btrfs_root *root); void btrfs_free_block_rsv(struct btrfs_root *root, struct btrfs_block_rsv *rsv); -int btrfs_block_rsv_add(struct btrfs_trans_handle *trans, - struct btrfs_root *root, +int btrfs_block_rsv_add(struct btrfs_root *root, struct btrfs_block_rsv *block_rsv, u64 num_bytes); -int btrfs_block_rsv_check(struct btrfs_trans_handle *trans, - struct btrfs_root *root, +int btrfs_block_rsv_check(struct btrfs_root *root, struct btrfs_block_rsv *block_rsv, u64 min_reserved, int min_factor, int flush); int btrfs_block_rsv_migrate(struct btrfs_block_rsv *src_rsv, -- cgit v1.2.3 From 3b16a4e3c355ee3c790473decfcf83d4faeb8ce0 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Wed, 21 Sep 2011 15:05:58 -0400 Subject: Btrfs: use the inode's mapping mask for allocating pages Johannes pointed out we were allocating only kernel pages for doing writes, which is kind of a big deal if you are on 32bit and have more than a gig of ram. So fix our allocations to use the mapping's gfp but still clear __GFP_FS so we don't re-enter. Thanks, Reported-by: Johannes Weiner Signed-off-by: Josef Bacik --- fs/btrfs/ctree.h | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'fs/btrfs/ctree.h') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index a5faf8e33baa..47dea7118e0e 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -30,6 +30,7 @@ #include #include #include +#include #include "extent_io.h" #include "extent_map.h" #include "async-thread.h" @@ -2117,6 +2118,11 @@ static inline bool btrfs_mixed_space_info(struct btrfs_space_info *space_info) (space_info->flags & BTRFS_BLOCK_GROUP_DATA)); } +static inline gfp_t btrfs_alloc_write_mask(struct address_space *mapping) +{ + return mapping_gfp_mask(mapping) & ~__GFP_FS; +} + /* extent-tree.c */ static inline u64 btrfs_calc_trans_metadata_size(struct btrfs_root *root, unsigned num_items) -- cgit v1.2.3 From 2bf64758fd6290797a5ce97d4b9c698a4ed1cbad Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Mon, 26 Sep 2011 17:12:22 -0400 Subject: Btrfs: allow us to overcommit our enospc reservations One of the things that kills us is the fact that our ENOSPC reservations are horribly over the top in most normal cases. There isn't too much that can be done about this because when we are completely full we really need them to work like this so we don't under reserve. However if there is plenty of unallocated chunks on the disk we can use that to gauge how much we can overcommit. So this patch adds chunk free space accounting so we always know how much unallocated space we have. Then if we fail to make a reservation within our allocated space, check to see if we can overcommit. In the normal flushing case (like with delalloc metadata reservations) we'll take the free space and divide it by 2 if our metadata profile is setup for DUP or any of those, and then divide it by 8 to make sure we don't overcommit too much. Then if we're in a non-flushing case (we really need this reservation now!) we only limit ourselves to half of the free space. This makes this fio test [torrent] filename=torrent-test rw=randwrite size=4g ioengine=sync directory=/mnt/btrfs-test go from taking around 45 minutes to 10 seconds on my freshly formatted 3 TiB file system. This doesn't seem to break my other enospc tests, but could really use some more testing as this is a super scary change. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/ctree.h | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'fs/btrfs/ctree.h') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 47dea7118e0e..1eafccb162ee 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -893,6 +893,10 @@ struct btrfs_fs_info { spinlock_t block_group_cache_lock; struct rb_root block_group_cache_tree; + /* keep track of unallocated space */ + spinlock_t free_chunk_lock; + u64 free_chunk_space; + struct extent_io_tree freed_extents[2]; struct extent_io_tree *pinned_extents; -- cgit v1.2.3 From 5b0e95bf607ddd59b39f52d3d55e6581c817b530 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Thu, 6 Oct 2011 08:58:24 -0400 Subject: Btrfs: inline checksums into the disk free space cache Yeah yeah I know this is how we used to do it and then I changed it, but damnit I'm changing it back. The fact is that writing out checksums will modify metadata, which could cause us to dirty a block group we've already written out, so we have to truncate it and all of it's checksums and re-write it which will write new checksums which could dirty a blockg roup that has already been written and you see where I'm going with this? This can cause unmount or really anything that depends on a transaction to commit to take it's sweet damned time to happen. So go back to the way it was, only this time we're specifically setting NODATACOW because we can't go through the COW pathway anyway and we're doing our own built-in cow'ing by truncating the free space cache. The other new thing is once we truncate the old cache and preallocate the new space, we don't need to do that song and dance at all for the rest of the transaction, we can just overwrite the existing space with the new cache if the block group changes for whatever reason, and the NODATACOW will let us do this fine. So keep track of which transaction we last cleared our cache in and if we cleared it in this transaction just say we're all setup and carry on. This survives xfstests and stress.sh. The inode cache will continue to use the normal csum infrastructure since it only gets written once and there will be no more modifications to the fs tree in a transaction commit. Signed-off-by: Josef Bacik --- fs/btrfs/ctree.h | 1 + 1 file changed, 1 insertion(+) (limited to 'fs/btrfs/ctree.h') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 1eafccb162ee..ea60897a9171 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -838,6 +838,7 @@ struct btrfs_block_group_cache { u64 bytes_super; u64 flags; u64 sectorsize; + u64 cache_generation; unsigned int ro:1; unsigned int dirty:1; unsigned int iref:1; -- cgit v1.2.3 From 36ba022ac0b748dd543f43430b03198e899426c9 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Tue, 18 Oct 2011 12:15:48 -0400 Subject: Btrfs: seperate out btrfs_block_rsv_check out into 2 different functions Currently btrfs_block_rsv_check does 2 things, it will either refill a block reserve like in the truncate or refill case, or it will check to see if there is enough space in the global reserve and possibly refill it. However because of overcommit we could be well overcommitting ourselves just to try and refill the global reserve, when really we should just be committing the transaction. So breack this out into btrfs_block_rsv_refill and btrfs_block_rsv_check. Refill will try to reserve more metadata if it can and btrfs_block_rsv_check will not, it will only tell you if the factor of the total space is still reserved. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/ctree.h | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'fs/btrfs/ctree.h') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index ea60897a9171..227620993bce 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -2252,8 +2252,10 @@ int btrfs_block_rsv_add(struct btrfs_root *root, struct btrfs_block_rsv *block_rsv, u64 num_bytes); int btrfs_block_rsv_check(struct btrfs_root *root, + struct btrfs_block_rsv *block_rsv, int min_factor); +int btrfs_block_rsv_refill(struct btrfs_root *root, struct btrfs_block_rsv *block_rsv, - u64 min_reserved, int min_factor, int flush); + u64 min_reserved); int btrfs_block_rsv_migrate(struct btrfs_block_rsv *src_rsv, struct btrfs_block_rsv *dst_rsv, u64 num_bytes); -- cgit v1.2.3 From e688b7252f784c2479d559f9f70ca8354752c5e7 Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Mon, 31 Oct 2011 20:52:39 -0400 Subject: Btrfs: fix extent pinning bugs in the tree log The tree log had two important bugs that could cause corruptions after a crash. Sometimes we were allowing tree log blocks to be reused after the tree log was committed but before the transaction commit was done. This allowed a future metadata write to overwrite the tree log data. It is fixed by adding a new variant of freeing reserved extents that always pins them. Credit goes to Stefan Behrens and Arne Jansen for many many hours spent tracking this bug down. During tree log replay, we do a pass through the tree log and pin all the extents we find. This makes sure the replay code won't go in and use any of those blocks for new allocations during replay. The problem is the free space cache isn't honoring these pinned extents. So the allocator can end up handing them out, leading to all kinds of problems during replay. The fix here is to force any free space cache to load while we pin the extents, and then to make sure we remove the pinned extents from the free space rbtree. Signed-off-by: Chris Mason Reported-by: Stefan Behrens --- fs/btrfs/ctree.h | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'fs/btrfs/ctree.h') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 227620993bce..f63c9b3f6e08 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -2156,6 +2156,9 @@ int btrfs_lookup_extent_info(struct btrfs_trans_handle *trans, u64 num_bytes, u64 *refs, u64 *flags); int btrfs_pin_extent(struct btrfs_root *root, u64 bytenr, u64 num, int reserved); +int btrfs_pin_extent_for_log_replay(struct btrfs_trans_handle *trans, + struct btrfs_root *root, + u64 bytenr, u64 num_bytes); int btrfs_cross_ref_exist(struct btrfs_trans_handle *trans, struct btrfs_root *root, u64 objectid, u64 offset, u64 bytenr); @@ -2206,6 +2209,8 @@ int btrfs_free_extent(struct btrfs_trans_handle *trans, u64 root_objectid, u64 owner, u64 offset); int btrfs_free_reserved_extent(struct btrfs_root *root, u64 start, u64 len); +int btrfs_free_and_pin_reserved_extent(struct btrfs_root *root, + u64 start, u64 len); int btrfs_prepare_extent_commit(struct btrfs_trans_handle *trans, struct btrfs_root *root); int btrfs_finish_extent_commit(struct btrfs_trans_handle *trans, -- cgit v1.2.3 From 6c41761fc6efe1503103a1afe03a6635c0b5d4ec Mon Sep 17 00:00:00 2001 From: David Sterba Date: Wed, 13 Apr 2011 15:41:04 +0200 Subject: btrfs: separate superblock items out of fs_info fs_info has now ~9kb, more than fits into one page. This will cause mount failure when memory is too fragmented. Top space consumers are super block structures super_copy and super_for_commit, ~2.8kb each. Allocate them dynamically. fs_info will be ~3.5kb. (measured on x86_64) Add a wrapper for freeing fs_info and all of it's dynamically allocated members. Signed-off-by: David Sterba --- fs/btrfs/ctree.h | 16 ++++++++++++++-- 1 file changed, 14 insertions(+), 2 deletions(-) (limited to 'fs/btrfs/ctree.h') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index f63c9b3f6e08..5181c53c1124 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -936,8 +936,8 @@ struct btrfs_fs_info { wait_queue_head_t transaction_blocked_wait; wait_queue_head_t async_submit_wait; - struct btrfs_super_block super_copy; - struct btrfs_super_block super_for_commit; + struct btrfs_super_block *super_copy; + struct btrfs_super_block *super_for_commit; struct block_device *__bdev; struct super_block *sb; struct inode *btree_inode; @@ -2387,6 +2387,18 @@ static inline int btrfs_fs_closing(struct btrfs_fs_info *fs_info) smp_mb(); return fs_info->closing; } +static inline void free_fs_info(struct btrfs_fs_info *fs_info) +{ + kfree(fs_info->delayed_root); + kfree(fs_info->extent_root); + kfree(fs_info->tree_root); + kfree(fs_info->chunk_root); + kfree(fs_info->dev_root); + kfree(fs_info->csum_root); + kfree(fs_info->super_copy); + kfree(fs_info->super_for_commit); + kfree(fs_info); +} /* root-item.c */ int btrfs_find_root_ref(struct btrfs_root *tree_root, -- cgit v1.2.3 From af31f5e5b84b5bf2bcec464153a5130b170b2770 Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Thu, 3 Nov 2011 15:17:42 -0400 Subject: Btrfs: add a log of past tree roots This takes some of the free space in the btrfs super block to record information about most of the roots in the last four commits. It also adds a -o recovery to use the root history log when we're not able to read the tree of tree roots, the extent tree root, the device tree root or the csum root. Signed-off-by: Chris Mason --- fs/btrfs/ctree.h | 95 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 95 insertions(+) (limited to 'fs/btrfs/ctree.h') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 5181c53c1124..78f43d1102a0 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -360,6 +360,47 @@ struct btrfs_header { #define BTRFS_SYSTEM_CHUNK_ARRAY_SIZE 2048 #define BTRFS_LABEL_SIZE 256 +/* + * just in case we somehow lose the roots and are not able to mount, + * we store an array of the roots from previous transactions + * in the super. + */ +#define BTRFS_NUM_BACKUP_ROOTS 4 +struct btrfs_root_backup { + __le64 tree_root; + __le64 tree_root_gen; + + __le64 chunk_root; + __le64 chunk_root_gen; + + __le64 extent_root; + __le64 extent_root_gen; + + __le64 fs_root; + __le64 fs_root_gen; + + __le64 dev_root; + __le64 dev_root_gen; + + __le64 csum_root; + __le64 csum_root_gen; + + __le64 total_bytes; + __le64 bytes_used; + __le64 num_devices; + /* future */ + __le64 unsed_64[4]; + + u8 tree_root_level; + u8 chunk_root_level; + u8 extent_root_level; + u8 fs_root_level; + u8 dev_root_level; + u8 csum_root_level; + /* future and to align */ + u8 unused_8[10]; +} __attribute__ ((__packed__)); + /* * the super block basically lists the main trees of the FS * it currently lacks any block count etc etc @@ -406,6 +447,7 @@ struct btrfs_super_block { /* future expansion */ __le64 reserved[31]; u8 sys_chunk_array[BTRFS_SYSTEM_CHUNK_ARRAY_SIZE]; + struct btrfs_root_backup super_roots[BTRFS_NUM_BACKUP_ROOTS]; } __attribute__ ((__packed__)); /* @@ -1113,6 +1155,9 @@ struct btrfs_fs_info { u64 fs_state; struct btrfs_delayed_root *delayed_root; + + /* next backup root to be overwritten */ + int backup_root_index; }; /* @@ -1357,6 +1402,7 @@ struct btrfs_ioctl_defrag_range_args { #define BTRFS_MOUNT_ENOSPC_DEBUG (1 << 15) #define BTRFS_MOUNT_AUTO_DEFRAG (1 << 16) #define BTRFS_MOUNT_INODE_MAP_CACHE (1 << 17) +#define BTRFS_MOUNT_RECOVERY (1 << 18) #define btrfs_clear_opt(o, opt) ((o) &= ~BTRFS_MOUNT_##opt) #define btrfs_set_opt(o, opt) ((o) |= BTRFS_MOUNT_##opt) @@ -1972,6 +2018,55 @@ static inline bool btrfs_root_readonly(struct btrfs_root *root) return root->root_item.flags & BTRFS_ROOT_SUBVOL_RDONLY; } +/* struct btrfs_root_backup */ +BTRFS_SETGET_STACK_FUNCS(backup_tree_root, struct btrfs_root_backup, + tree_root, 64); +BTRFS_SETGET_STACK_FUNCS(backup_tree_root_gen, struct btrfs_root_backup, + tree_root_gen, 64); +BTRFS_SETGET_STACK_FUNCS(backup_tree_root_level, struct btrfs_root_backup, + tree_root_level, 8); + +BTRFS_SETGET_STACK_FUNCS(backup_chunk_root, struct btrfs_root_backup, + chunk_root, 64); +BTRFS_SETGET_STACK_FUNCS(backup_chunk_root_gen, struct btrfs_root_backup, + chunk_root_gen, 64); +BTRFS_SETGET_STACK_FUNCS(backup_chunk_root_level, struct btrfs_root_backup, + chunk_root_level, 8); + +BTRFS_SETGET_STACK_FUNCS(backup_extent_root, struct btrfs_root_backup, + extent_root, 64); +BTRFS_SETGET_STACK_FUNCS(backup_extent_root_gen, struct btrfs_root_backup, + extent_root_gen, 64); +BTRFS_SETGET_STACK_FUNCS(backup_extent_root_level, struct btrfs_root_backup, + extent_root_level, 8); + +BTRFS_SETGET_STACK_FUNCS(backup_fs_root, struct btrfs_root_backup, + fs_root, 64); +BTRFS_SETGET_STACK_FUNCS(backup_fs_root_gen, struct btrfs_root_backup, + fs_root_gen, 64); +BTRFS_SETGET_STACK_FUNCS(backup_fs_root_level, struct btrfs_root_backup, + fs_root_level, 8); + +BTRFS_SETGET_STACK_FUNCS(backup_dev_root, struct btrfs_root_backup, + dev_root, 64); +BTRFS_SETGET_STACK_FUNCS(backup_dev_root_gen, struct btrfs_root_backup, + dev_root_gen, 64); +BTRFS_SETGET_STACK_FUNCS(backup_dev_root_level, struct btrfs_root_backup, + dev_root_level, 8); + +BTRFS_SETGET_STACK_FUNCS(backup_csum_root, struct btrfs_root_backup, + csum_root, 64); +BTRFS_SETGET_STACK_FUNCS(backup_csum_root_gen, struct btrfs_root_backup, + csum_root_gen, 64); +BTRFS_SETGET_STACK_FUNCS(backup_csum_root_level, struct btrfs_root_backup, + csum_root_level, 8); +BTRFS_SETGET_STACK_FUNCS(backup_total_bytes, struct btrfs_root_backup, + total_bytes, 64); +BTRFS_SETGET_STACK_FUNCS(backup_bytes_used, struct btrfs_root_backup, + bytes_used, 64); +BTRFS_SETGET_STACK_FUNCS(backup_num_devices, struct btrfs_root_backup, + num_devices, 64); + /* struct btrfs_super_block */ BTRFS_SETGET_STACK_FUNCS(super_bytenr, struct btrfs_super_block, bytenr, 64); -- cgit v1.2.3 From 6d668dda0caec537fbf28c4d91e6d18181af3cff Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Thu, 3 Nov 2011 22:54:25 -0400 Subject: Btrfs: make a delayed_block_rsv for the delayed item insertion I've been hitting warnings in use_block_rsv when running the delayed insertion stuff. It's because we will readjust global block rsv based on what is in use, which means we could end up discarding reservations that are for the delayed insertion stuff. So instead create a seperate block rsv for the delayed insertion stuff. This will also make it easier to debug problems with the delayed insertion reservations since we will know that only the delayed insertion code touches this block_rsv. Thanks, Signed-off-by: Josef Bacik Signed-off-by: Chris Mason --- fs/btrfs/ctree.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'fs/btrfs/ctree.h') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 78f43d1102a0..3002e5d4da0b 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -957,6 +957,8 @@ struct btrfs_fs_info { struct btrfs_block_rsv trans_block_rsv; /* block reservation for chunk tree */ struct btrfs_block_rsv chunk_block_rsv; + /* block reservation for delayed operations */ + struct btrfs_block_rsv delayed_block_rsv; struct btrfs_block_rsv empty_block_rsv; -- cgit v1.2.3 From c06a0e120a4e381a1c291c1fce3c6155c5791cae Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Fri, 4 Nov 2011 19:56:02 -0400 Subject: Btrfs: fix delayed insertion reservation We all keep getting those stupid warnings from use_block_rsv when running stress.sh, and it's because the delayed insertion stuff is being stupid. It's not the delayed insertion stuffs fault, it's all just stupid. When marking an inode dirty for oh say updating the time on it, we just do a btrfs_join_transaction, which doesn't reserve any space. This is stupid because we're going to have to have space reserve to make this change, but we do it because it's fast because chances are we're going to call it over and over again and it doesn't matter. Well thanks to the delayed insertion stuff this is mostly the case, so we do actually need to make this reservation. So if trans->bytes_reserved is 0 then try to do a normal reservation. If not return ENOSPC which will make the btrfs_dirty_inode start a proper transaction which will let it do the whole ENOSPC dance and reserve enough space for the delayed insertion to steal the reservation from the transaction. The other stupid thing we do is not reserve space for the inode when writing to the thing. Usually this is ok since we have to update the time so we'd have already done all this work before we get to the endio stuff, so it doesn't matter. But this is stupid because we could write the data after the transaction commits where we changed the mtime of the inode so we have to cow all the way down to the inode anyway. This used to be masked by the delalloc reservation stuff, but because we delay the update it doesn't get masked in this case. So again the delayed insertion stuff bites us in the ass. So if our trans->block_rsv is delalloc, just steal the reservation from the delalloc reserve. Hopefully this won't bite us in the ass, but I've said that before. With this patch stress.sh no longer spits out those stupid warnings (famous last words). Thanks, Signed-off-by: Josef Bacik Signed-off-by: Chris Mason --- fs/btrfs/ctree.h | 3 +++ 1 file changed, 3 insertions(+) (limited to 'fs/btrfs/ctree.h') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 3002e5d4da0b..6bb34fc1ff22 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -2353,6 +2353,9 @@ void btrfs_free_block_rsv(struct btrfs_root *root, int btrfs_block_rsv_add(struct btrfs_root *root, struct btrfs_block_rsv *block_rsv, u64 num_bytes); +int btrfs_block_rsv_add_noflush(struct btrfs_root *root, + struct btrfs_block_rsv *block_rsv, + u64 num_bytes); int btrfs_block_rsv_check(struct btrfs_root *root, struct btrfs_block_rsv *block_rsv, int min_factor); int btrfs_block_rsv_refill(struct btrfs_root *root, -- cgit v1.2.3