diff options
Diffstat (limited to 'fs')
176 files changed, 7321 insertions, 4109 deletions
diff --git a/fs/afs/cell.c b/fs/afs/cell.c index fd5133e26a38..78ba5f932287 100644 --- a/fs/afs/cell.c +++ b/fs/afs/cell.c @@ -134,8 +134,17 @@ static struct afs_cell *afs_alloc_cell(struct afs_net *net, _leave(" = -ENAMETOOLONG"); return ERR_PTR(-ENAMETOOLONG); } - if (namelen == 5 && memcmp(name, "@cell", 5) == 0) + + /* Prohibit cell names that contain unprintable chars, '/' and '@' or + * that begin with a dot. This also precludes "@cell". + */ + if (name[0] == '.') return ERR_PTR(-EINVAL); + for (i = 0; i < namelen; i++) { + char ch = name[i]; + if (!isprint(ch) || ch == '/' || ch == '@') + return ERR_PTR(-EINVAL); + } _enter("%*.*s,%s", namelen, namelen, name, addresses); diff --git a/fs/afs/dir.c b/fs/afs/dir.c index 497f979018c2..5c794f4b051a 100644 --- a/fs/afs/dir.c +++ b/fs/afs/dir.c @@ -908,6 +908,7 @@ static struct dentry *afs_lookup(struct inode *dir, struct dentry *dentry, unsigned int flags) { struct afs_vnode *dvnode = AFS_FS_I(dir); + struct afs_fid fid = {}; struct inode *inode; struct dentry *d; struct key *key; @@ -951,21 +952,18 @@ static struct dentry *afs_lookup(struct inode *dir, struct dentry *dentry, afs_stat_v(dvnode, n_lookup); inode = afs_do_lookup(dir, dentry, key); key_put(key); - if (inode == ERR_PTR(-ENOENT)) { + if (inode == ERR_PTR(-ENOENT)) inode = afs_try_auto_mntpt(dentry, dir); - } else { - dentry->d_fsdata = - (void *)(unsigned long)dvnode->status.data_version; - } + + if (!IS_ERR_OR_NULL(inode)) + fid = AFS_FS_I(inode)->fid; + d = d_splice_alias(inode, dentry); if (!IS_ERR_OR_NULL(d)) { d->d_fsdata = dentry->d_fsdata; - trace_afs_lookup(dvnode, &d->d_name, - inode ? AFS_FS_I(inode) : NULL); + trace_afs_lookup(dvnode, &d->d_name, &fid); } else { - trace_afs_lookup(dvnode, &dentry->d_name, - IS_ERR_OR_NULL(inode) ? NULL - : AFS_FS_I(inode)); + trace_afs_lookup(dvnode, &dentry->d_name, &fid); } return d; } diff --git a/fs/afs/dynroot.c b/fs/afs/dynroot.c index 4150280509ff..7503899c0a1b 100644 --- a/fs/afs/dynroot.c +++ b/fs/afs/dynroot.c @@ -136,6 +136,9 @@ static struct dentry *afs_dynroot_lookup(struct inode *dir, struct dentry *dentr ASSERTCMP(d_inode(dentry), ==, NULL); + if (flags & LOOKUP_CREATE) + return ERR_PTR(-EOPNOTSUPP); + if (dentry->d_name.len >= AFSNAMEMAX) { _leave(" = -ENAMETOOLONG"); return ERR_PTR(-ENAMETOOLONG); diff --git a/fs/afs/mntpt.c b/fs/afs/mntpt.c index f532d6d3bd28..79bc5f1338ed 100644 --- a/fs/afs/mntpt.c +++ b/fs/afs/mntpt.c @@ -126,7 +126,7 @@ static int afs_mntpt_set_params(struct fs_context *fc, struct dentry *mntpt) if (src_as->cell) ctx->cell = afs_get_cell(src_as->cell); - if (size > PAGE_SIZE - 1) + if (size < 2 || size > PAGE_SIZE - 1) return -EINVAL; page = read_mapping_page(d_inode(mntpt)->i_mapping, 0, NULL); @@ -140,7 +140,9 @@ static int afs_mntpt_set_params(struct fs_context *fc, struct dentry *mntpt) } buf = kmap(page); - ret = vfs_parse_fs_string(fc, "source", buf, size); + ret = -EINVAL; + if (buf[size - 1] == '.') + ret = vfs_parse_fs_string(fc, "source", buf, size - 1); kunmap(page); put_page(page); if (ret < 0) diff --git a/fs/afs/proc.c b/fs/afs/proc.c index fba2ec3a3a9c..468e1713bce1 100644 --- a/fs/afs/proc.c +++ b/fs/afs/proc.c @@ -213,13 +213,14 @@ static int afs_proc_cell_volumes_show(struct seq_file *m, void *v) /* Display header on line 1 */ if (v == &cell->proc_volumes) { - seq_puts(m, "USE VID TY\n"); + seq_puts(m, "USE VID TY NAME\n"); return 0; } - seq_printf(m, "%3d %08llx %s\n", + seq_printf(m, "%3d %08llx %s %s\n", atomic_read(&vol->usage), vol->vid, - afs_vol_types[vol->type]); + afs_vol_types[vol->type], + vol->name); return 0; } diff --git a/fs/afs/server.c b/fs/afs/server.c index 1686bf188ccd..b7f3cb2130ca 100644 --- a/fs/afs/server.c +++ b/fs/afs/server.c @@ -32,18 +32,11 @@ static void afs_dec_servers_outstanding(struct afs_net *net) struct afs_server *afs_find_server(struct afs_net *net, const struct sockaddr_rxrpc *srx) { - const struct sockaddr_in6 *a = &srx->transport.sin6, *b; const struct afs_addr_list *alist; struct afs_server *server = NULL; unsigned int i; - bool ipv6 = true; int seq = 0, diff; - if (srx->transport.sin6.sin6_addr.s6_addr32[0] == 0 || - srx->transport.sin6.sin6_addr.s6_addr32[1] == 0 || - srx->transport.sin6.sin6_addr.s6_addr32[2] == htonl(0xffff)) - ipv6 = false; - rcu_read_lock(); do { @@ -52,7 +45,8 @@ struct afs_server *afs_find_server(struct afs_net *net, server = NULL; read_seqbegin_or_lock(&net->fs_addr_lock, &seq); - if (ipv6) { + if (srx->transport.family == AF_INET6) { + const struct sockaddr_in6 *a = &srx->transport.sin6, *b; hlist_for_each_entry_rcu(server, &net->fs_addresses6, addr6_link) { alist = rcu_dereference(server->addresses); for (i = alist->nr_ipv4; i < alist->nr_addrs; i++) { @@ -68,15 +62,16 @@ struct afs_server *afs_find_server(struct afs_net *net, } } } else { + const struct sockaddr_in *a = &srx->transport.sin, *b; hlist_for_each_entry_rcu(server, &net->fs_addresses4, addr4_link) { alist = rcu_dereference(server->addresses); for (i = 0; i < alist->nr_ipv4; i++) { - b = &alist->addrs[i].transport.sin6; - diff = ((u16 __force)a->sin6_port - - (u16 __force)b->sin6_port); + b = &alist->addrs[i].transport.sin; + diff = ((u16 __force)a->sin_port - + (u16 __force)b->sin_port); if (diff == 0) - diff = ((u32 __force)a->sin6_addr.s6_addr32[3] - - (u32 __force)b->sin6_addr.s6_addr32[3]); + diff = ((u32 __force)a->sin_addr.s_addr - + (u32 __force)b->sin_addr.s_addr); if (diff == 0) goto found; } diff --git a/fs/afs/super.c b/fs/afs/super.c index 488641b1a418..7f8a9b3137bf 100644 --- a/fs/afs/super.c +++ b/fs/afs/super.c @@ -404,6 +404,7 @@ static int afs_test_super(struct super_block *sb, struct fs_context *fc) return (as->net_ns == fc->net_ns && as->volume && as->volume->vid == ctx->volume->vid && + as->cell == ctx->cell && !as->dyn_root); } @@ -448,7 +449,6 @@ static int afs_fill_super(struct super_block *sb, struct afs_fs_context *ctx) /* allocate the root inode and dentry */ if (as->dyn_root) { inode = afs_iget_pseudo_dir(sb, true); - sb->s_flags |= SB_RDONLY; } else { sprintf(sb->s_id, "%llu", as->volume->vid); afs_activate_volume(as->volume); diff --git a/fs/btrfs/Kconfig b/fs/btrfs/Kconfig index 75b6d10c9845..575636f6491e 100644 --- a/fs/btrfs/Kconfig +++ b/fs/btrfs/Kconfig @@ -7,6 +7,7 @@ config BTRFS_FS select LIBCRC32C select CRYPTO_XXHASH select CRYPTO_SHA256 + select CRYPTO_BLAKE2B select ZLIB_INFLATE select ZLIB_DEFLATE select LZO_COMPRESS diff --git a/fs/btrfs/Makefile b/fs/btrfs/Makefile index 82200dbca5ac..9a0ff3384381 100644 --- a/fs/btrfs/Makefile +++ b/fs/btrfs/Makefile @@ -11,7 +11,7 @@ btrfs-y += super.o ctree.o extent-tree.o print-tree.o root-tree.o dir-item.o \ compression.o delayed-ref.o relocation.o delayed-inode.o scrub.o \ reada.o backref.o ulist.o qgroup.o send.o dev-replace.o raid56.o \ uuid-tree.o props.o free-space-tree.o tree-checker.o space-info.o \ - block-rsv.o delalloc-space.o block-group.o + block-rsv.o delalloc-space.o block-group.o discard.o btrfs-$(CONFIG_BTRFS_FS_POSIX_ACL) += acl.o btrfs-$(CONFIG_BTRFS_FS_CHECK_INTEGRITY) += check-integrity.o diff --git a/fs/btrfs/block-group.c b/fs/btrfs/block-group.c index 6934a5b8708f..14851584e245 100644 --- a/fs/btrfs/block-group.c +++ b/fs/btrfs/block-group.c @@ -14,6 +14,8 @@ #include "sysfs.h" #include "tree-log.h" #include "delalloc-space.h" +#include "discard.h" +#include "raid56.h" /* * Return target flags in extended format or 0 if restripe for this chunk_type @@ -95,7 +97,7 @@ static u64 btrfs_reduce_alloc_profile(struct btrfs_fs_info *fs_info, u64 flags) return extended_to_chunk(flags | allowed); } -static u64 get_alloc_profile(struct btrfs_fs_info *fs_info, u64 orig_flags) +u64 btrfs_get_alloc_profile(struct btrfs_fs_info *fs_info, u64 orig_flags) { unsigned seq; u64 flags; @@ -115,11 +117,6 @@ static u64 get_alloc_profile(struct btrfs_fs_info *fs_info, u64 orig_flags) return btrfs_reduce_alloc_profile(fs_info, flags); } -u64 btrfs_get_alloc_profile(struct btrfs_fs_info *fs_info, u64 orig_flags) -{ - return get_alloc_profile(fs_info, orig_flags); -} - void btrfs_get_block_group(struct btrfs_block_group *cache) { atomic_inc(&cache->count); @@ -132,6 +129,15 @@ void btrfs_put_block_group(struct btrfs_block_group *cache) WARN_ON(cache->reserved > 0); /* + * A block_group shouldn't be on the discard_list anymore. + * Remove the block_group from the discard_list to prevent us + * from causing a panic due to NULL pointer dereference. + */ + if (WARN_ON(!list_empty(&cache->discard_list))) + btrfs_discard_cancel_work(&cache->fs_info->discard_ctl, + cache); + + /* * If not empty, someone is still holding mutex of * full_stripe_lock, which can only be released by caller. * And it will definitely cause use-after-free when caller @@ -466,8 +472,8 @@ u64 add_new_free_space(struct btrfs_block_group *block_group, u64 start, u64 end } else if (extent_start > start && extent_start < end) { size = extent_start - start; total_added += size; - ret = btrfs_add_free_space(block_group, start, - size); + ret = btrfs_add_free_space_async_trimmed(block_group, + start, size); BUG_ON(ret); /* -ENOMEM or logic error */ start = extent_end + 1; } else { @@ -478,7 +484,8 @@ u64 add_new_free_space(struct btrfs_block_group *block_group, u64 start, u64 end if (start < end) { size = end - start; total_added += size; - ret = btrfs_add_free_space(block_group, start, size); + ret = btrfs_add_free_space_async_trimmed(block_group, start, + size); BUG_ON(ret); /* -ENOMEM or logic error */ } @@ -1185,21 +1192,8 @@ static int inc_block_group_ro(struct btrfs_block_group *cache, int force) struct btrfs_space_info *sinfo = cache->space_info; u64 num_bytes; u64 sinfo_used; - u64 min_allocable_bytes; int ret = -ENOSPC; - /* - * We need some metadata space and system metadata space for - * allocating chunks in some corner cases until we force to set - * it to be readonly. - */ - if ((sinfo->flags & - (BTRFS_BLOCK_GROUP_SYSTEM | BTRFS_BLOCK_GROUP_METADATA)) && - !force) - min_allocable_bytes = SZ_1M; - else - min_allocable_bytes = 0; - spin_lock(&sinfo->lock); spin_lock(&cache->lock); @@ -1217,10 +1211,9 @@ static int inc_block_group_ro(struct btrfs_block_group *cache, int force) * sinfo_used + num_bytes should always <= sinfo->total_bytes. * * Here we make sure if we mark this bg RO, we still have enough - * free space as buffer (if min_allocable_bytes is not 0). + * free space as buffer. */ - if (sinfo_used + num_bytes + min_allocable_bytes <= - sinfo->total_bytes) { + if (sinfo_used + num_bytes <= sinfo->total_bytes) { sinfo->bytes_readonly += num_bytes; cache->ro++; list_add_tail(&cache->ro_list, &sinfo->ro_bgs); @@ -1233,8 +1226,8 @@ out: btrfs_info(cache->fs_info, "unable to make block group %llu ro", cache->start); btrfs_info(cache->fs_info, - "sinfo_used=%llu bg_num_bytes=%llu min_allocable=%llu", - sinfo_used, num_bytes, min_allocable_bytes); + "sinfo_used=%llu bg_num_bytes=%llu", + sinfo_used, num_bytes); btrfs_dump_space_info(cache->fs_info, cache->space_info, 0, 0); } return ret; @@ -1249,6 +1242,7 @@ void btrfs_delete_unused_bgs(struct btrfs_fs_info *fs_info) struct btrfs_block_group *block_group; struct btrfs_space_info *space_info; struct btrfs_trans_handle *trans; + const bool async_trim_enabled = btrfs_test_opt(fs_info, DISCARD_ASYNC); int ret = 0; if (!test_bit(BTRFS_FS_OPEN, &fs_info->flags)) @@ -1272,10 +1266,28 @@ void btrfs_delete_unused_bgs(struct btrfs_fs_info *fs_info) } spin_unlock(&fs_info->unused_bgs_lock); + btrfs_discard_cancel_work(&fs_info->discard_ctl, block_group); + mutex_lock(&fs_info->delete_unused_bgs_mutex); /* Don't want to race with allocators so take the groups_sem */ down_write(&space_info->groups_sem); + + /* + * Async discard moves the final block group discard to be prior + * to the unused_bgs code path. Therefore, if it's not fully + * trimmed, punt it back to the async discard lists. + */ + if (btrfs_test_opt(fs_info, DISCARD_ASYNC) && + !btrfs_is_free_space_trimmed(block_group)) { + trace_btrfs_skip_unused_block_group(block_group); + up_write(&space_info->groups_sem); + /* Requeue if we failed because of async discard */ + btrfs_discard_queue_work(&fs_info->discard_ctl, + block_group); + goto next; + } + spin_lock(&block_group->lock); if (block_group->reserved || block_group->pinned || block_group->used || block_group->ro || @@ -1347,6 +1359,23 @@ void btrfs_delete_unused_bgs(struct btrfs_fs_info *fs_info) } mutex_unlock(&fs_info->unused_bg_unpin_mutex); + /* + * At this point, the block_group is read only and should fail + * new allocations. However, btrfs_finish_extent_commit() can + * cause this block_group to be placed back on the discard + * lists because now the block_group isn't fully discarded. + * Bail here and try again later after discarding everything. + */ + spin_lock(&fs_info->discard_ctl.lock); + if (!list_empty(&block_group->discard_list)) { + spin_unlock(&fs_info->discard_ctl.lock); + btrfs_dec_block_group_ro(block_group); + btrfs_discard_queue_work(&fs_info->discard_ctl, + block_group); + goto end_trans; + } + spin_unlock(&fs_info->discard_ctl.lock); + /* Reset pinned so btrfs_put_block_group doesn't complain */ spin_lock(&space_info->lock); spin_lock(&block_group->lock); @@ -1362,8 +1391,18 @@ void btrfs_delete_unused_bgs(struct btrfs_fs_info *fs_info) spin_unlock(&block_group->lock); spin_unlock(&space_info->lock); + /* + * The normal path here is an unused block group is passed here, + * then trimming is handled in the transaction commit path. + * Async discard interposes before this to do the trimming + * before coming down the unused block group path as trimming + * will no longer be done later in the transaction commit path. + */ + if (!async_trim_enabled && btrfs_test_opt(fs_info, DISCARD_ASYNC)) + goto flip_async; + /* DISCARD can flip during remount */ - trimming = btrfs_test_opt(fs_info, DISCARD); + trimming = btrfs_test_opt(fs_info, DISCARD_SYNC); /* Implicit trim during transaction commit. */ if (trimming) @@ -1406,6 +1445,13 @@ next: spin_lock(&fs_info->unused_bgs_lock); } spin_unlock(&fs_info->unused_bgs_lock); + return; + +flip_async: + btrfs_end_transaction(trans); + mutex_unlock(&fs_info->delete_unused_bgs_mutex); + btrfs_put_block_group(block_group); + btrfs_discard_punt_unused_bgs_list(fs_info); } void btrfs_mark_bg_unused(struct btrfs_block_group *bg) @@ -1516,6 +1562,102 @@ static void set_avail_alloc_bits(struct btrfs_fs_info *fs_info, u64 flags) write_sequnlock(&fs_info->profiles_lock); } +/** + * btrfs_rmap_block - Map a physical disk address to a list of logical addresses + * @chunk_start: logical address of block group + * @physical: physical address to map to logical addresses + * @logical: return array of logical addresses which map to @physical + * @naddrs: length of @logical + * @stripe_len: size of IO stripe for the given block group + * + * Maps a particular @physical disk address to a list of @logical addresses. + * Used primarily to exclude those portions of a block group that contain super + * block copies. + */ +EXPORT_FOR_TESTS +int btrfs_rmap_block(struct btrfs_fs_info *fs_info, u64 chunk_start, + u64 physical, u64 **logical, int *naddrs, int *stripe_len) +{ + struct extent_map *em; + struct map_lookup *map; + u64 *buf; + u64 bytenr; + u64 data_stripe_length; + u64 io_stripe_size; + int i, nr = 0; + int ret = 0; + + em = btrfs_get_chunk_map(fs_info, chunk_start, 1); + if (IS_ERR(em)) + return -EIO; + + map = em->map_lookup; + data_stripe_length = em->len; + io_stripe_size = map->stripe_len; + + if (map->type & BTRFS_BLOCK_GROUP_RAID10) + data_stripe_length = div_u64(data_stripe_length, + map->num_stripes / map->sub_stripes); + else if (map->type & BTRFS_BLOCK_GROUP_RAID0) + data_stripe_length = div_u64(data_stripe_length, map->num_stripes); + else if (map->type & BTRFS_BLOCK_GROUP_RAID56_MASK) { + data_stripe_length = div_u64(data_stripe_length, + nr_data_stripes(map)); + io_stripe_size = map->stripe_len * nr_data_stripes(map); + } + + buf = kcalloc(map->num_stripes, sizeof(u64), GFP_NOFS); + if (!buf) { + ret = -ENOMEM; + goto out; + } + + for (i = 0; i < map->num_stripes; i++) { + bool already_inserted = false; + u64 stripe_nr; + int j; + + if (!in_range(physical, map->stripes[i].physical, + data_stripe_length)) + continue; + + stripe_nr = physical - map->stripes[i].physical; + stripe_nr = div64_u64(stripe_nr, map->stripe_len); + + if (map->type & BTRFS_BLOCK_GROUP_RAID10) { + stripe_nr = stripe_nr * map->num_stripes + i; + stripe_nr = div_u64(stripe_nr, map->sub_stripes); + } else if (map->type & BTRFS_BLOCK_GROUP_RAID0) { + stripe_nr = stripe_nr * map->num_stripes + i; + } + /* + * The remaining case would be for RAID56, multiply by + * nr_data_stripes(). Alternatively, just use rmap_len below + * instead of map->stripe_len + */ + + bytenr = chunk_start + stripe_nr * io_stripe_size; + + /* Ensure we don't add duplicate addresses */ + for (j = 0; j < nr; j++) { + if (buf[j] == bytenr) { + already_inserted = true; + break; + } + } + + if (!already_inserted) + buf[nr++] = bytenr; + } + + *logical = buf; + *naddrs = nr; + *stripe_len = io_stripe_size; +out: + free_extent_map(em); + return ret; +} + static int exclude_super_stripes(struct btrfs_block_group *cache) { struct btrfs_fs_info *fs_info = cache->fs_info; @@ -1610,6 +1752,8 @@ static struct btrfs_block_group *btrfs_create_block_group_cache( cache->full_stripe_len = btrfs_full_stripe_len(fs_info, start); set_free_space_tree_thresholds(cache); + cache->discard_index = BTRFS_DISCARD_INDEX_UNUSED; + atomic_set(&cache->count, 1); spin_lock_init(&cache->lock); init_rwsem(&cache->data_rwsem); @@ -1617,6 +1761,7 @@ static struct btrfs_block_group *btrfs_create_block_group_cache( INIT_LIST_HEAD(&cache->cluster_list); INIT_LIST_HEAD(&cache->bg_list); INIT_LIST_HEAD(&cache->ro_list); + INIT_LIST_HEAD(&cache->discard_list); INIT_LIST_HEAD(&cache->dirty_list); INIT_LIST_HEAD(&cache->io_list); btrfs_init_free_space_ctl(cache); @@ -1775,7 +1920,10 @@ static int read_one_block_group(struct btrfs_fs_info *info, inc_block_group_ro(cache, 1); } else if (cache->used == 0) { ASSERT(list_empty(&cache->bg_list)); - btrfs_mark_bg_unused(cache); + if (btrfs_test_opt(info, DISCARD_ASYNC)) + btrfs_discard_queue_work(&info->discard_ctl, cache); + else + btrfs_mark_bg_unused(cache); } return 0; error: @@ -2738,8 +2886,10 @@ int btrfs_update_block_group(struct btrfs_trans_handle *trans, * dirty list to avoid races between cleaner kthread and space * cache writeout. */ - if (!alloc && old_val == 0) - btrfs_mark_bg_unused(cache); + if (!alloc && old_val == 0) { + if (!btrfs_test_opt(info, DISCARD_ASYNC)) + btrfs_mark_bg_unused(cache); + } btrfs_put_block_group(cache); total -= num_bytes; diff --git a/fs/btrfs/block-group.h b/fs/btrfs/block-group.h index 9b409676c4b2..107bb557ca8d 100644 --- a/fs/btrfs/block-group.h +++ b/fs/btrfs/block-group.h @@ -13,6 +13,19 @@ enum btrfs_disk_cache_state { }; /* + * This describes the state of the block_group for async discard. This is due + * to the two pass nature of it where extent discarding is prioritized over + * bitmap discarding. BTRFS_DISCARD_RESET_CURSOR is set when we are resetting + * between lists to prevent contention for discard state variables + * (eg. discard_cursor). + */ +enum btrfs_discard_state { + BTRFS_DISCARD_EXTENTS, + BTRFS_DISCARD_BITMAPS, + BTRFS_DISCARD_RESET_CURSOR, +}; + +/* * Control flags for do_chunk_alloc's force field CHUNK_ALLOC_NO_FORCE means to * only allocate a chunk if we really need one. * @@ -116,7 +129,13 @@ struct btrfs_block_group { /* For read-only block groups */ struct list_head ro_list; + /* For discard operations */ atomic_t trimming; + struct list_head discard_list; + int discard_index; + u64 discard_eligible_time; + u64 discard_cursor; + enum btrfs_discard_state discard_state; /* For dirty block groups */ struct list_head dirty_list; @@ -158,6 +177,22 @@ struct btrfs_block_group { struct btrfs_full_stripe_locks_tree full_stripe_locks_root; }; +static inline u64 btrfs_block_group_end(struct btrfs_block_group *block_group) +{ + return (block_group->start + block_group->length); +} + +static inline bool btrfs_is_block_group_data_only( + struct btrfs_block_group *block_group) +{ + /* + * In mixed mode the fragmentation is expected to be high, lowering the + * efficiency, so only proper data block groups are considered. + */ + return (block_group->flags & BTRFS_BLOCK_GROUP_DATA) && + !(block_group->flags & BTRFS_BLOCK_GROUP_METADATA); +} + #ifdef CONFIG_BTRFS_DEBUG static inline int btrfs_should_fragment_free_space( struct btrfs_block_group *block_group) @@ -248,4 +283,9 @@ static inline int btrfs_block_group_done(struct btrfs_block_group *cache) cache->cached == BTRFS_CACHE_ERROR; } +#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS +int btrfs_rmap_block(struct btrfs_fs_info *fs_info, u64 chunk_start, + u64 physical, u64 **logical, int *naddrs, int *stripe_len); +#endif + #endif /* BTRFS_BLOCK_GROUP_H */ diff --git a/fs/btrfs/check-integrity.c b/fs/btrfs/check-integrity.c index 0b52ab4cb964..a0ce69f2d27c 100644 --- a/fs/btrfs/check-integrity.c +++ b/fs/btrfs/check-integrity.c @@ -629,7 +629,6 @@ static struct btrfsic_dev_state *btrfsic_dev_state_hashtable_lookup(dev_t dev, static int btrfsic_process_superblock(struct btrfsic_state *state, struct btrfs_fs_devices *fs_devices) { - struct btrfs_fs_info *fs_info = state->fs_info; struct btrfs_super_block *selected_super; struct list_head *dev_head = &fs_devices->devices; struct btrfs_device *device; @@ -637,7 +636,6 @@ static int btrfsic_process_superblock(struct btrfsic_state *state, int ret = 0; int pass; - BUG_ON(NULL == state); selected_super = kzalloc(sizeof(*selected_super), GFP_NOFS); if (NULL == selected_super) { pr_info("btrfsic: error, kmalloc failed!\n"); @@ -700,7 +698,7 @@ static int btrfsic_process_superblock(struct btrfsic_state *state, break; } - num_copies = btrfs_num_copies(fs_info, next_bytenr, + num_copies = btrfs_num_copies(state->fs_info, next_bytenr, state->metablock_size); if (state->print_mask & BTRFSIC_PRINT_MASK_NUM_COPIES) pr_info("num_copies(log_bytenr=%llu) = %d\n", diff --git a/fs/btrfs/compression.c b/fs/btrfs/compression.c index ee834ef7beb4..de95ad27722f 100644 --- a/fs/btrfs/compression.c +++ b/fs/btrfs/compression.c @@ -447,7 +447,7 @@ blk_status_t btrfs_submit_compressed_write(struct inode *inode, u64 start, if (blkcg_css) { bio->bi_opf |= REQ_CGROUP_PUNT; - bio_associate_blkg_from_css(bio, blkcg_css); + kthread_associate_blkcg(blkcg_css); } refcount_set(&cb->pending_bios, 1); @@ -491,6 +491,8 @@ blk_status_t btrfs_submit_compressed_write(struct inode *inode, u64 start, bio->bi_opf = REQ_OP_WRITE | write_flags; bio->bi_private = cb; bio->bi_end_io = end_compressed_bio_write; + if (blkcg_css) + bio->bi_opf |= REQ_CGROUP_PUNT; bio_add_page(bio, page, PAGE_SIZE, 0); } if (bytes_left < PAGE_SIZE) { @@ -517,6 +519,9 @@ blk_status_t btrfs_submit_compressed_write(struct inode *inode, u64 start, bio_endio(bio); } + if (blkcg_css) + kthread_associate_blkcg(NULL); + return 0; } @@ -758,7 +763,7 @@ blk_status_t btrfs_submit_compressed_read(struct inode *inode, struct bio *bio, if (!(BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM)) { ret = btrfs_lookup_bio_sums(inode, comp_bio, - sums); + (u64)-1, sums); BUG_ON(ret); /* -ENOMEM */ } @@ -786,7 +791,7 @@ blk_status_t btrfs_submit_compressed_read(struct inode *inode, struct bio *bio, BUG_ON(ret); /* -ENOMEM */ if (!(BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM)) { - ret = btrfs_lookup_bio_sums(inode, comp_bio, sums); + ret = btrfs_lookup_bio_sums(inode, comp_bio, (u64)-1, sums); BUG_ON(ret); /* -ENOMEM */ } diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c index 5b6e86aaf2e1..24658b5a5787 100644 --- a/fs/btrfs/ctree.c +++ b/fs/btrfs/ctree.c @@ -379,7 +379,7 @@ void btrfs_put_tree_mod_seq(struct btrfs_fs_info *fs_info, for (node = rb_first(tm_root); node; node = next) { next = rb_next(node); tm = rb_entry(node, struct tree_mod_elem, node); - if (tm->seq > min_seq) + if (tm->seq >= min_seq) continue; rb_erase(node, tm_root); kfree(tm); diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index b2e8fd8a8e59..f90b82050d2d 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -101,6 +101,14 @@ struct btrfs_ref; #define BTRFS_MAX_EXTENT_SIZE SZ_128M +/* + * Deltas are an effective way to populate global statistics. Give macro names + * to make it clear what we're doing. An example is discard_extents in + * btrfs_free_space_ctl. + */ +#define BTRFS_STAT_NR_ENTRIES 2 +#define BTRFS_STAT_CURR 0 +#define BTRFS_STAT_PREV 1 /* * Count how many BTRFS_MAX_EXTENT_SIZE cover the @size @@ -440,6 +448,36 @@ struct btrfs_full_stripe_locks_tree { struct mutex lock; }; +/* Discard control. */ +/* + * Async discard uses multiple lists to differentiate the discard filter + * parameters. Index 0 is for completely free block groups where we need to + * ensure the entire block group is trimmed without being lossy. Indices + * afterwards represent monotonically decreasing discard filter sizes to + * prioritize what should be discarded next. + */ +#define BTRFS_NR_DISCARD_LISTS 3 +#define BTRFS_DISCARD_INDEX_UNUSED 0 +#define BTRFS_DISCARD_INDEX_START 1 + +struct btrfs_discard_ctl { + struct workqueue_struct *discard_workers; + struct delayed_work work; + spinlock_t lock; + struct btrfs_block_group *block_group; + struct list_head discard_list[BTRFS_NR_DISCARD_LISTS]; + u64 prev_discard; + atomic_t discardable_extents; + atomic64_t discardable_bytes; + u64 max_discard_size; + unsigned long delay; + u32 iops_limit; + u32 kbps_limit; + u64 discard_extent_bytes; + u64 discard_bitmap_bytes; + atomic64_t discard_bytes_saved; +}; + /* delayed seq elem */ struct seq_list { struct list_head list; @@ -526,6 +564,9 @@ enum { * so we don't need to offload checksums to workqueues. */ BTRFS_FS_CSUM_IMPL_FAST, + + /* Indicate that the discard workqueue can service discards. */ + BTRFS_FS_DISCARD_RUNNING, }; struct btrfs_fs_info { @@ -816,6 +857,8 @@ struct btrfs_fs_info { struct btrfs_workqueue *scrub_wr_completion_workers; struct btrfs_workqueue *scrub_parity_workers; + struct btrfs_discard_ctl discard_ctl; + #ifdef CONFIG_BTRFS_FS_CHECK_INTEGRITY u32 check_integrity_print_mask; #endif @@ -902,6 +945,11 @@ struct btrfs_fs_info { spinlock_t ref_verify_lock; struct rb_root block_tree; #endif + +#ifdef CONFIG_BTRFS_DEBUG + struct kobject *debug_kobj; + struct kobject *discard_debug_kobj; +#endif }; static inline struct btrfs_fs_info *btrfs_sb(struct super_block *sb) @@ -1170,7 +1218,7 @@ static inline u32 BTRFS_MAX_XATTR_SIZE(const struct btrfs_fs_info *info) #define BTRFS_MOUNT_FLUSHONCOMMIT (1 << 7) #define BTRFS_MOUNT_SSD_SPREAD (1 << 8) #define BTRFS_MOUNT_NOSSD (1 << 9) -#define BTRFS_MOUNT_DISCARD (1 << 10) +#define BTRFS_MOUNT_DISCARD_SYNC (1 << 10) #define BTRFS_MOUNT_FORCE_COMPRESS (1 << 11) #define BTRFS_MOUNT_SPACE_CACHE (1 << 12) #define BTRFS_MOUNT_CLEAR_CACHE (1 << 13) @@ -1189,6 +1237,7 @@ static inline u32 BTRFS_MAX_XATTR_SIZE(const struct btrfs_fs_info *info) #define BTRFS_MOUNT_FREE_SPACE_TREE (1 << 26) #define BTRFS_MOUNT_NOLOGREPLAY (1 << 27) #define BTRFS_MOUNT_REF_VERIFY (1 << 28) +#define BTRFS_MOUNT_DISCARD_ASYNC (1 << 29) #define BTRFS_DEFAULT_COMMIT_INTERVAL (30) #define BTRFS_DEFAULT_MAX_INLINE (2048) @@ -2449,8 +2498,8 @@ int btrfs_free_extent(struct btrfs_trans_handle *trans, struct btrfs_ref *ref); int btrfs_free_reserved_extent(struct btrfs_fs_info *fs_info, u64 start, u64 len, int delalloc); -int btrfs_free_and_pin_reserved_extent(struct btrfs_fs_info *fs_info, - u64 start, u64 len); +int btrfs_pin_reserved_extent(struct btrfs_fs_info *fs_info, u64 start, + u64 len); void btrfs_prepare_extent_commit(struct btrfs_fs_info *fs_info); int btrfs_finish_extent_commit(struct btrfs_trans_handle *trans); int btrfs_inc_extent_ref(struct btrfs_trans_handle *trans, @@ -2787,11 +2836,9 @@ struct btrfs_inode_extref *btrfs_find_name_in_ext_backref( /* file-item.c */ struct btrfs_dio_private; int btrfs_del_csums(struct btrfs_trans_handle *trans, - struct btrfs_fs_info *fs_info, u64 bytenr, u64 len); + struct btrfs_root *root, u64 bytenr, u64 len); blk_status_t btrfs_lookup_bio_sums(struct inode *inode, struct bio *bio, - u8 *dst); -blk_status_t btrfs_lookup_bio_sums_dio(struct inode *inode, struct bio *bio, - u64 logical_offset); + u64 offset, u8 *dst); int btrfs_insert_file_extent(struct btrfs_trans_handle *trans, struct btrfs_root *root, u64 objectid, u64 pos, @@ -2877,7 +2924,7 @@ struct inode *btrfs_iget(struct super_block *s, struct btrfs_key *location, struct btrfs_root *root); struct extent_map *btrfs_get_extent(struct btrfs_inode *inode, struct page *page, size_t pg_offset, - u64 start, u64 end, int create); + u64 start, u64 end); int btrfs_update_inode(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct inode *inode); @@ -3110,17 +3157,21 @@ do { \ rcu_read_unlock(); \ } while (0) -__cold -static inline void assfail(const char *expr, const char *file, int line) +#ifdef CONFIG_BTRFS_ASSERT +__cold __noreturn +static inline void assertfail(const char *expr, const char *file, int line) { - if (IS_ENABLED(CONFIG_BTRFS_ASSERT)) { - pr_err("assertion failed: %s, in %s:%d\n", expr, file, line); - BUG(); - } + pr_err("assertion failed: %s, in %s:%d\n", expr, file, line); + BUG(); } -#define ASSERT(expr) \ - (likely(expr) ? (void)0 : assfail(#expr, __FILE__, __LINE__)) +#define ASSERT(expr) \ + (likely(expr) ? (void)0 : assertfail(#expr, __FILE__, __LINE__)) + +#else +static inline void assertfail(const char *expr, const char* file, int line) { } +#define ASSERT(expr) (void)(expr) +#endif /* * Use that for functions that are conditionally exported for sanity tests but diff --git a/fs/btrfs/dev-replace.c b/fs/btrfs/dev-replace.c index f639dde2a679..2ca2a09d0e23 100644 --- a/fs/btrfs/dev-replace.c +++ b/fs/btrfs/dev-replace.c @@ -500,11 +500,8 @@ static int btrfs_dev_replace_start(struct btrfs_fs_info *fs_info, &dev_replace->scrub_progress, 0, 1); ret = btrfs_dev_replace_finishing(fs_info, ret); - if (ret == -EINPROGRESS) { + if (ret == -EINPROGRESS) ret = BTRFS_IOCTL_DEV_REPLACE_RESULT_SCRUB_INPROGRESS; - } else if (ret != -ECANCELED) { - WARN_ON(ret); - } return ret; @@ -707,6 +704,7 @@ static int btrfs_dev_replace_finishing(struct btrfs_fs_info *fs_info, /* replace the sysfs entry */ btrfs_sysfs_rm_device_link(fs_info->fs_devices, src_device); + btrfs_sysfs_update_devid(tgt_device); btrfs_rm_dev_replace_free_srcdev(src_device); /* write back the superblocks */ diff --git a/fs/btrfs/discard.c b/fs/btrfs/discard.c new file mode 100644 index 000000000000..5615320fa659 --- /dev/null +++ b/fs/btrfs/discard.c @@ -0,0 +1,702 @@ +// SPDX-License-Identifier: GPL-2.0 + +#include <linux/jiffies.h> +#include <linux/kernel.h> +#include <linux/ktime.h> +#include <linux/list.h> +#include <linux/math64.h> +#include <linux/sizes.h> +#include <linux/workqueue.h> +#include "ctree.h" +#include "block-group.h" +#include "discard.h" +#include "free-space-cache.h" + +/* + * This contains the logic to handle async discard. + * + * Async discard manages trimming of free space outside of transaction commit. + * Discarding is done by managing the block_groups on a LRU list based on free + * space recency. Two passes are used to first prioritize discarding extents + * and then allow for trimming in the bitmap the best opportunity to coalesce. + * The block_groups are maintained on multiple lists to allow for multiple + * passes with different discard filter requirements. A delayed work item is + * used to manage discarding with timeout determined by a max of the delay + * incurred by the iops rate limit, the byte rate limit, and the max delay of + * BTRFS_DISCARD_MAX_DELAY. + * + * Note, this only keeps track of block_groups that are explicitly for data. + * Mixed block_groups are not supported. + * + * The first list is special to manage discarding of fully free block groups. + * This is necessary because we issue a final trim for a full free block group + * after forgetting it. When a block group becomes unused, instead of directly + * being added to the unused_bgs list, we add it to this first list. Then + * from there, if it becomes fully discarded, we place it onto the unused_bgs + * list. + * + * The in-memory free space cache serves as the backing state for discard. + * Consequently this means there is no persistence. We opt to load all the + * block groups in as not discarded, so the mount case degenerates to the + * crashing case. + * + * As the free space cache uses bitmaps, there exists a tradeoff between + * ease/efficiency for find_free_extent() and the accuracy of discard state. + * Here we opt to let untrimmed regions merge with everything while only letting + * trimmed regions merge with other trimmed regions. This can cause + * overtrimming, but the coalescing benefit seems to be worth it. Additionally, + * bitmap state is tracked as a whole. If we're able to fully trim a bitmap, + * the trimmed flag is set on the bitmap. Otherwise, if an allocation comes in, + * this resets the state and we will retry trimming the whole bitmap. This is a + * tradeoff between discard state accuracy and the cost of accounting. + */ + +/* This is an initial delay to give some chance for block reuse */ +#define BTRFS_DISCARD_DELAY (120ULL * NSEC_PER_SEC) +#define BTRFS_DISCARD_UNUSED_DELAY (10ULL * NSEC_PER_SEC) + +/* Target completion latency of discarding all discardable extents */ +#define BTRFS_DISCARD_TARGET_MSEC (6 * 60 * 60UL * MSEC_PER_SEC) +#define BTRFS_DISCARD_MIN_DELAY_MSEC (1UL) +#define BTRFS_DISCARD_MAX_DELAY_MSEC (1000UL) +#define BTRFS_DISCARD_MAX_IOPS (10U) + +/* Montonically decreasing minimum length filters after index 0 */ +static int discard_minlen[BTRFS_NR_DISCARD_LISTS] = { + 0, + BTRFS_ASYNC_DISCARD_MAX_FILTER, + BTRFS_ASYNC_DISCARD_MIN_FILTER +}; + +static struct list_head *get_discard_list(struct btrfs_discard_ctl *discard_ctl, + struct btrfs_block_group *block_group) +{ + return &discard_ctl->discard_list[block_group->discard_index]; +} + +static void __add_to_discard_list(struct btrfs_discard_ctl *discard_ctl, + struct btrfs_block_group *block_group) +{ + if (!btrfs_run_discard_work(discard_ctl)) + return; + + if (list_empty(&block_group->discard_list) || + block_group->discard_index == BTRFS_DISCARD_INDEX_UNUSED) { + if (block_group->discard_index == BTRFS_DISCARD_INDEX_UNUSED) + block_group->discard_index = BTRFS_DISCARD_INDEX_START; + block_group->discard_eligible_time = (ktime_get_ns() + + BTRFS_DISCARD_DELAY); + block_group->discard_state = BTRFS_DISCARD_RESET_CURSOR; + } + + list_move_tail(&block_group->discard_list, + get_discard_list(discard_ctl, block_group)); +} + +static void add_to_discard_list(struct btrfs_discard_ctl *discard_ctl, + struct btrfs_block_group *block_group) +{ + if (!btrfs_is_block_group_data_only(block_group)) + return; + + spin_lock(&discard_ctl->lock); + __add_to_discard_list(discard_ctl, block_group); + spin_unlock(&discard_ctl->lock); +} + +static void add_to_discard_unused_list(struct btrfs_discard_ctl *discard_ctl, + struct btrfs_block_group *block_group) +{ + spin_lock(&discard_ctl->lock); + + if (!btrfs_run_discard_work(discard_ctl)) { + spin_unlock(&discard_ctl->lock); + return; + } + + list_del_init(&block_group->discard_list); + + block_group->discard_index = BTRFS_DISCARD_INDEX_UNUSED; + block_group->discard_eligible_time = (ktime_get_ns() + + BTRFS_DISCARD_UNUSED_DELAY); + block_group->discard_state = BTRFS_DISCARD_RESET_CURSOR; + list_add_tail(&block_group->discard_list, + &discard_ctl->discard_list[BTRFS_DISCARD_INDEX_UNUSED]); + + spin_unlock(&discard_ctl->lock); +} + +static bool remove_from_discard_list(struct btrfs_discard_ctl *discard_ctl, + struct btrfs_block_group *block_group) +{ + bool running = false; + + spin_lock(&discard_ctl->lock); + + if (block_group == discard_ctl->block_group) { + running = true; + discard_ctl->block_group = NULL; + } + + block_group->discard_eligible_time = 0; + list_del_init(&block_group->discard_list); + + spin_unlock(&discard_ctl->lock); + + return running; +} + +/** + * find_next_block_group - find block_group that's up next for discarding + * @discard_ctl: discard control + * @now: current time + * + * Iterate over the discard lists to find the next block_group up for + * discarding checking the discard_eligible_time of block_group. + */ +static struct btrfs_block_group *find_next_block_group( + struct btrfs_discard_ctl *discard_ctl, + u64 now) +{ + struct btrfs_block_group *ret_block_group = NULL, *block_group; + int i; + + for (i = 0; i < BTRFS_NR_DISCARD_LISTS; i++) { + struct list_head *discard_list = &discard_ctl->discard_list[i]; + + if (!list_empty(discard_list)) { + block_group = list_first_entry(discard_list, + struct btrfs_block_group, + discard_list); + + if (!ret_block_group) + ret_block_group = block_group; + + if (ret_block_group->discard_eligible_time < now) + break; + + if (ret_block_group->discard_eligible_time > + block_group->discard_eligible_time) + ret_block_group = block_group; + } + } + + return ret_block_group; +} + +/** + * peek_discard_list - wrap find_next_block_group() + * @discard_ctl: discard control + * @discard_state: the discard_state of the block_group after state management + * @discard_index: the discard_index of the block_group after state management + * + * This wraps find_next_block_group() and sets the block_group to be in use. + * discard_state's control flow is managed here. Variables related to + * discard_state are reset here as needed (eg discard_cursor). @discard_state + * and @discard_index are remembered as it may change while we're discarding, + * but we want the discard to execute in the context determined here. + */ +static struct btrfs_block_group *peek_discard_list( + struct btrfs_discard_ctl *discard_ctl, + enum btrfs_discard_state *discard_state, + int *discard_index) +{ + struct btrfs_block_group *block_group; + const u64 now = ktime_get_ns(); + + spin_lock(&discard_ctl->lock); +again: + block_group = find_next_block_group(discard_ctl, now); + + if (block_group && now > block_group->discard_eligible_time) { + if (block_group->discard_index == BTRFS_DISCARD_INDEX_UNUSED && + block_group->used != 0) { + if (btrfs_is_block_group_data_only(block_group)) + __add_to_discard_list(discard_ctl, block_group); + else + list_del_init(&block_group->discard_list); + goto again; + } + if (block_group->discard_state == BTRFS_DISCARD_RESET_CURSOR) { + block_group->discard_cursor = block_group->start; + block_group->discard_state = BTRFS_DISCARD_EXTENTS; + } + discard_ctl->block_group = block_group; + *discard_state = block_group->discard_state; + *discard_index = block_group->discard_index; + } else { + block_group = NULL; + } + + spin_unlock(&discard_ctl->lock); + + return block_group; +} + +/** + * btrfs_discard_check_filter - updates a block groups filters + * @block_group: block group of interest + * @bytes: recently freed region size after coalescing + * + * Async discard maintains multiple lists with progressively smaller filters + * to prioritize discarding based on size. Should a free space that matches + * a larger filter be returned to the free_space_cache, prioritize that discard + * by moving @block_group to the proper filter. + */ +void btrfs_discard_check_filter(struct btrfs_block_group *block_group, + u64 bytes) +{ + struct btrfs_discard_ctl *discard_ctl; + + if (!block_group || + !btrfs_test_opt(block_group->fs_info, DISCARD_ASYNC)) + return; + + discard_ctl = &block_group->fs_info->discard_ctl; + + if (block_group->discard_index > BTRFS_DISCARD_INDEX_START && + bytes >= discard_minlen[block_group->discard_index - 1]) { + int i; + + remove_from_discard_list(discard_ctl, block_group); + + for (i = BTRFS_DISCARD_INDEX_START; i < BTRFS_NR_DISCARD_LISTS; + i++) { + if (bytes >= discard_minlen[i]) { + block_group->discard_index = i; + add_to_discard_list(discard_ctl, block_group); + break; + } + } + } +} + +/** + * btrfs_update_discard_index - moves a block group along the discard lists + * @discard_ctl: discard control + * @block_group: block_group of interest + * + * Increment @block_group's discard_index. If it falls of the list, let it be. + * Otherwise add it back to the appropriate list. + */ +static void btrfs_update_discard_index(struct btrfs_discard_ctl *discard_ctl, + struct btrfs_block_group *block_group) +{ + block_group->discard_index++; + if (block_group->discard_index == BTRFS_NR_DISCARD_LISTS) { + block_group->discard_index = 1; + return; + } + + add_to_discard_list(discard_ctl, block_group); +} + +/** + * btrfs_discard_cancel_work - remove a block_group from the discard lists + * @discard_ctl: discard control + * @block_group: block_group of interest + * + * This removes @block_group from the discard lists. If necessary, it waits on + * the current work and then reschedules the delayed work. + */ +void btrfs_discard_cancel_work(struct btrfs_discard_ctl *discard_ctl, + struct btrfs_block_group *block_group) +{ + if (remove_from_discard_list(discard_ctl, block_group)) { + cancel_delayed_work_sync(&discard_ctl->work); + btrfs_discard_schedule_work(discard_ctl, true); + } +} + +/** + * btrfs_discard_queue_work - handles queuing the block_groups + * @discard_ctl: discard control + * @block_group: block_group of interest + * + * This maintains the LRU order of the discard lists. + */ +void btrfs_discard_queue_work(struct btrfs_discard_ctl *discard_ctl, + struct btrfs_block_group *block_group) +{ + if (!block_group || !btrfs_test_opt(block_group->fs_info, DISCARD_ASYNC)) + return; + + if (block_group->used == 0) + add_to_discard_unused_list(discard_ctl, block_group); + else + add_to_discard_list(discard_ctl, block_group); + + if (!delayed_work_pending(&discard_ctl->work)) + btrfs_discard_schedule_work(discard_ctl, false); +} + +/** + * btrfs_discard_schedule_work - responsible for scheduling the discard work + * @discard_ctl: discard control + * @override: override the current timer + * + * Discards are issued by a delayed workqueue item. @override is used to + * update the current delay as the baseline delay interval is reevaluated on + * transaction commit. This is also maxed with any other rate limit. + */ +void btrfs_discard_schedule_work(struct btrfs_discard_ctl *discard_ctl, + bool override) +{ + struct btrfs_block_group *block_group; + const u64 now = ktime_get_ns(); + + spin_lock(&discard_ctl->lock); + + if (!btrfs_run_discard_work(discard_ctl)) + goto out; + + if (!override && delayed_work_pending(&discard_ctl->work)) + goto out; + + block_group = find_next_block_group(discard_ctl, now); + if (block_group) { + unsigned long delay = discard_ctl->delay; + u32 kbps_limit = READ_ONCE(discard_ctl->kbps_limit); + + /* + * A single delayed workqueue item is responsible for + * discarding, so we can manage the bytes rate limit by keeping + * track of the previous discard. + */ + if (kbps_limit && discard_ctl->prev_discard) { + u64 bps_limit = ((u64)kbps_limit) * SZ_1K; + u64 bps_delay = div64_u64(discard_ctl->prev_discard * + MSEC_PER_SEC, bps_limit); + + delay = max(delay, msecs_to_jiffies(bps_delay)); + } + + /* + * This timeout is to hopefully prevent immediate discarding + * in a recently allocated block group. + */ + if (now < block_group->discard_eligible_time) { + u64 bg_timeout = block_group->discard_eligible_time - now; + + delay = max(delay, nsecs_to_jiffies(bg_timeout)); + } + + mod_delayed_work(discard_ctl->discard_workers, + &discard_ctl->work, delay); + } +out: + spin_unlock(&discard_ctl->lock); +} + +/** + * btrfs_finish_discard_pass - determine next step of a block_group + * @discard_ctl: discard control + * @block_group: block_group of interest + * + * This determines the next step for a block group after it's finished going + * through a pass on a discard list. If it is unused and fully trimmed, we can + * mark it unused and send it to the unused_bgs path. Otherwise, pass it onto + * the appropriate filter list or let it fall off. + */ +static void btrfs_finish_discard_pass(struct btrfs_discard_ctl *discard_ctl, + struct btrfs_block_group *block_group) +{ + remove_from_discard_list(discard_ctl, block_group); + + if (block_group->used == 0) { + if (btrfs_is_free_space_trimmed(block_group)) + btrfs_mark_bg_unused(block_group); + else + add_to_discard_unused_list(discard_ctl, block_group); + } else { + btrfs_update_discard_index(discard_ctl, block_group); + } +} + +/** + * btrfs_discard_workfn - discard work function + * @work: work + * + * This finds the next block_group to start discarding and then discards a + * single region. It does this in a two-pass fashion: first extents and second + * bitmaps. Completely discarded block groups are sent to the unused_bgs path. + */ +static void btrfs_discard_workfn(struct work_struct *work) +{ + struct btrfs_discard_ctl *discard_ctl; + struct btrfs_block_group *block_group; + enum btrfs_discard_state discard_state; + int discard_index = 0; + u64 trimmed = 0; + u64 minlen = 0; + + discard_ctl = container_of(work, struct btrfs_discard_ctl, work.work); + + block_group = peek_discard_list(discard_ctl, &discard_state, + &discard_index); + if (!block_group || !btrfs_run_discard_work(discard_ctl)) + return; + + /* Perform discarding */ + minlen = discard_minlen[discard_index]; + + if (discard_state == BTRFS_DISCARD_BITMAPS) { + u64 maxlen = 0; + + /* + * Use the previous levels minimum discard length as the max + * length filter. In the case something is added to make a + * region go beyond the max filter, the entire bitmap is set + * back to BTRFS_TRIM_STATE_UNTRIMMED. + */ + if (discard_index != BTRFS_DISCARD_INDEX_UNUSED) + maxlen = discard_minlen[discard_index - 1]; + + btrfs_trim_block_group_bitmaps(block_group, &trimmed, + block_group->discard_cursor, + btrfs_block_group_end(block_group), + minlen, maxlen, true); + discard_ctl->discard_bitmap_bytes += trimmed; + } else { + btrfs_trim_block_group_extents(block_group, &trimmed, + block_group->discard_cursor, + btrfs_block_group_end(block_group), + minlen, true); + discard_ctl->discard_extent_bytes += trimmed; + } + + discard_ctl->prev_discard = trimmed; + + /* Determine next steps for a block_group */ + if (block_group->discard_cursor >= btrfs_block_group_end(block_group)) { + if (discard_state == BTRFS_DISCARD_BITMAPS) { + btrfs_finish_discard_pass(discard_ctl, block_group); + } else { + block_group->discard_cursor = block_group->start; + spin_lock(&discard_ctl->lock); + if (block_group->discard_state != + BTRFS_DISCARD_RESET_CURSOR) + block_group->discard_state = + BTRFS_DISCARD_BITMAPS; + spin_unlock(&discard_ctl->lock); + } + } + + spin_lock(&discard_ctl->lock); + discard_ctl->block_group = NULL; + spin_unlock(&discard_ctl->lock); + + btrfs_discard_schedule_work(discard_ctl, false); +} + +/** + * btrfs_run_discard_work - determines if async discard should be running + * @discard_ctl: discard control + * + * Checks if the file system is writeable and BTRFS_FS_DISCARD_RUNNING is set. + */ +bool btrfs_run_discard_work(struct btrfs_discard_ctl *discard_ctl) +{ + struct btrfs_fs_info *fs_info = container_of(discard_ctl, + struct btrfs_fs_info, + discard_ctl); + + return (!(fs_info->sb->s_flags & SB_RDONLY) && + test_bit(BTRFS_FS_DISCARD_RUNNING, &fs_info->flags)); +} + +/** + * btrfs_discard_calc_delay - recalculate the base delay + * @discard_ctl: discard control + * + * Recalculate the base delay which is based off the total number of + * discardable_extents. Clamp this between the lower_limit (iops_limit or 1ms) + * and the upper_limit (BTRFS_DISCARD_MAX_DELAY_MSEC). + */ +void btrfs_discard_calc_delay(struct btrfs_discard_ctl *discard_ctl) +{ + s32 discardable_extents; + s64 discardable_bytes; + u32 iops_limit; + unsigned long delay; + unsigned long lower_limit = BTRFS_DISCARD_MIN_DELAY_MSEC; + + discardable_extents = atomic_read(&discard_ctl->discardable_extents); + if (!discardable_extents) + return; + + spin_lock(&discard_ctl->lock); + + /* + * The following is to fix a potential -1 discrepenancy that we're not + * sure how to reproduce. But given that this is the only place that + * utilizes these numbers and this is only called by from + * btrfs_finish_extent_commit() which is synchronized, we can correct + * here. + */ + if (discardable_extents < 0) + atomic_add(-discardable_extents, + &discard_ctl->discardable_extents); + + discardable_bytes = atomic64_read(&discard_ctl->discardable_bytes); + if (discardable_bytes < 0) + atomic64_add(-discardable_bytes, + &discard_ctl->discardable_bytes); + + if (discardable_extents <= 0) { + spin_unlock(&discard_ctl->lock); + return; + } + + iops_limit = READ_ONCE(discard_ctl->iops_limit); + if (iops_limit) + lower_limit = max_t(unsigned long, lower_limit, + MSEC_PER_SEC / iops_limit); + + delay = BTRFS_DISCARD_TARGET_MSEC / discardable_extents; + delay = clamp(delay, lower_limit, BTRFS_DISCARD_MAX_DELAY_MSEC); + discard_ctl->delay = msecs_to_jiffies(delay); + + spin_unlock(&discard_ctl->lock); +} + +/** + * btrfs_discard_update_discardable - propagate discard counters + * @block_group: block_group of interest + * @ctl: free_space_ctl of @block_group + * + * This propagates deltas of counters up to the discard_ctl. It maintains a + * current counter and a previous counter passing the delta up to the global + * stat. Then the current counter value becomes the previous counter value. + */ +void btrfs_discard_update_discardable(struct btrfs_block_group *block_group, + struct btrfs_free_space_ctl *ctl) +{ + struct btrfs_discard_ctl *discard_ctl; + s32 extents_delta; + s64 bytes_delta; + + if (!block_group || + !btrfs_test_opt(block_group->fs_info, DISCARD_ASYNC) || + !btrfs_is_block_group_data_only(block_group)) + return; + + discard_ctl = &block_group->fs_info->discard_ctl; + + extents_delta = ctl->discardable_extents[BTRFS_STAT_CURR] - + ctl->discardable_extents[BTRFS_STAT_PREV]; + if (extents_delta) { + atomic_add(extents_delta, &discard_ctl->discardable_extents); + ctl->discardable_extents[BTRFS_STAT_PREV] = + ctl->discardable_extents[BTRFS_STAT_CURR]; + } + + bytes_delta = ctl->discardable_bytes[BTRFS_STAT_CURR] - + ctl->discardable_bytes[BTRFS_STAT_PREV]; + if (bytes_delta) { + atomic64_add(bytes_delta, &discard_ctl->discardable_bytes); + ctl->discardable_bytes[BTRFS_STAT_PREV] = + ctl->discardable_bytes[BTRFS_STAT_CURR]; + } +} + +/** + * btrfs_discard_punt_unused_bgs_list - punt unused_bgs list to discard lists + * @fs_info: fs_info of interest + * + * The unused_bgs list needs to be punted to the discard lists because the + * order of operations is changed. In the normal sychronous discard path, the + * block groups are trimmed via a single large trim in transaction commit. This + * is ultimately what we are trying to avoid with asynchronous discard. Thus, + * it must be done before going down the unused_bgs path. + */ +void btrfs_discard_punt_unused_bgs_list(struct btrfs_fs_info *fs_info) +{ + struct btrfs_block_group *block_group, *next; + + spin_lock(&fs_info->unused_bgs_lock); + /* We enabled async discard, so punt all to the queue */ + list_for_each_entry_safe(block_group, next, &fs_info->unused_bgs, + bg_list) { + list_del_init(&block_group->bg_list); + btrfs_discard_queue_work(&fs_info->discard_ctl, block_group); + } + spin_unlock(&fs_info->unused_bgs_lock); +} + +/** + * btrfs_discard_purge_list - purge discard lists + * @discard_ctl: discard control + * + * If we are disabling async discard, we may have intercepted block groups that + * are completely free and ready for the unused_bgs path. As discarding will + * now happen in transaction commit or not at all, we can safely mark the + * corresponding block groups as unused and they will be sent on their merry + * way to the unused_bgs list. + */ +static void btrfs_discard_purge_list(struct btrfs_discard_ctl *discard_ctl) +{ + struct btrfs_block_group *block_group, *next; + int i; + + spin_lock(&discard_ctl->lock); + for (i = 0; i < BTRFS_NR_DISCARD_LISTS; i++) { + list_for_each_entry_safe(block_group, next, + &discard_ctl->discard_list[i], + discard_list) { + list_del_init(&block_group->discard_list); + spin_unlock(&discard_ctl->lock); + if (block_group->used == 0) + btrfs_mark_bg_unused(block_group); + spin_lock(&discard_ctl->lock); + } + } + spin_unlock(&discard_ctl->lock); +} + +void btrfs_discard_resume(struct btrfs_fs_info *fs_info) +{ + if (!btrfs_test_opt(fs_info, DISCARD_ASYNC)) { + btrfs_discard_cleanup(fs_info); + return; + } + + btrfs_discard_punt_unused_bgs_list(fs_info); + + set_bit(BTRFS_FS_DISCARD_RUNNING, &fs_info->flags); +} + +void btrfs_discard_stop(struct btrfs_fs_info *fs_info) +{ + clear_bit(BTRFS_FS_DISCARD_RUNNING, &fs_info->flags); +} + +void btrfs_discard_init(struct btrfs_fs_info *fs_info) +{ + struct btrfs_discard_ctl *discard_ctl = &fs_info->discard_ctl; + int i; + + spin_lock_init(&discard_ctl->lock); + INIT_DELAYED_WORK(&discard_ctl->work, btrfs_discard_workfn); + + for (i = 0; i < BTRFS_NR_DISCARD_LISTS; i++) + INIT_LIST_HEAD(&discard_ctl->discard_list[i]); + + discard_ctl->prev_discard = 0; + atomic_set(&discard_ctl->discardable_extents, 0); + atomic64_set(&discard_ctl->discardable_bytes, 0); + discard_ctl->max_discard_size = BTRFS_ASYNC_DISCARD_DEFAULT_MAX_SIZE; + discard_ctl->delay = BTRFS_DISCARD_MAX_DELAY_MSEC; + discard_ctl->iops_limit = BTRFS_DISCARD_MAX_IOPS; + discard_ctl->kbps_limit = 0; + discard_ctl->discard_extent_bytes = 0; + discard_ctl->discard_bitmap_bytes = 0; + atomic64_set(&discard_ctl->discard_bytes_saved, 0); +} + +void btrfs_discard_cleanup(struct btrfs_fs_info *fs_info) +{ + btrfs_discard_stop(fs_info); + cancel_delayed_work_sync(&fs_info->discard_ctl.work); + btrfs_discard_purge_list(&fs_info->discard_ctl); +} diff --git a/fs/btrfs/discard.h b/fs/btrfs/discard.h new file mode 100644 index 000000000000..21a15776dac4 --- /dev/null +++ b/fs/btrfs/discard.h @@ -0,0 +1,41 @@ +// SPDX-License-Identifier: GPL-2.0 + +#ifndef BTRFS_DISCARD_H +#define BTRFS_DISCARD_H + +#include <linux/sizes.h> + +struct btrfs_fs_info; +struct btrfs_discard_ctl; +struct btrfs_block_group; + +/* Discard size limits */ +#define BTRFS_ASYNC_DISCARD_DEFAULT_MAX_SIZE (SZ_64M) +#define BTRFS_ASYNC_DISCARD_MAX_FILTER (SZ_1M) +#define BTRFS_ASYNC_DISCARD_MIN_FILTER (SZ_32K) + +/* List operations */ +void btrfs_discard_check_filter(struct btrfs_block_group *block_group, u64 bytes); + +/* Work operations */ +void btrfs_discard_cancel_work(struct btrfs_discard_ctl *discard_ctl, + struct btrfs_block_group *block_group); +void btrfs_discard_queue_work(struct btrfs_discard_ctl *discard_ctl, + struct btrfs_block_group *block_group); +void btrfs_discard_schedule_work(struct btrfs_discard_ctl *discard_ctl, + bool override); +bool btrfs_run_discard_work(struct btrfs_discard_ctl *discard_ctl); + +/* Update operations */ +void btrfs_discard_calc_delay(struct btrfs_discard_ctl *discard_ctl); +void btrfs_discard_update_discardable(struct btrfs_block_group *block_group, + struct btrfs_free_space_ctl *ctl); + +/* Setup/cleanup operations */ +void btrfs_discard_punt_unused_bgs_list(struct btrfs_fs_info *fs_info); +void btrfs_discard_resume(struct btrfs_fs_info *fs_info); +void btrfs_discard_stop(struct btrfs_fs_info *fs_info); +void btrfs_discard_init(struct btrfs_fs_info *fs_info); +void btrfs_discard_cleanup(struct btrfs_fs_info *fs_info); + +#endif diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index e0edfdc9c82b..aea48d6ddc0c 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -41,6 +41,7 @@ #include "tree-checker.h" #include "ref-verify.h" #include "block-group.h" +#include "discard.h" #define BTRFS_SUPER_FLAG_SUPP (BTRFS_HEADER_FLAG_WRITTEN |\ BTRFS_HEADER_FLAG_RELOC |\ @@ -202,8 +203,8 @@ void btrfs_set_buffer_lockdep_class(u64 objectid, struct extent_buffer *eb, * that covers the entire device */ struct extent_map *btree_get_extent(struct btrfs_inode *inode, - struct page *page, size_t pg_offset, u64 start, u64 len, - int create) + struct page *page, size_t pg_offset, + u64 start, u64 len) { struct extent_map_tree *em_tree = &inode->extent_tree; struct extent_map *em; @@ -1953,6 +1954,8 @@ static void btrfs_stop_all_workers(struct btrfs_fs_info *fs_info) btrfs_destroy_workqueue(fs_info->readahead_workers); btrfs_destroy_workqueue(fs_info->flush_workers); btrfs_destroy_workqueue(fs_info->qgroup_rescan_workers); + if (fs_info->discard_ctl.discard_workers) + destroy_workqueue(fs_info->discard_ctl.discard_workers); /* * Now that all other work queues are destroyed, we can safely destroy * the queues used for metadata I/O, since tasks from those other work @@ -2148,6 +2151,8 @@ static int btrfs_init_workqueues(struct btrfs_fs_info *fs_info, max_active, 2); fs_info->qgroup_rescan_workers = btrfs_alloc_workqueue(fs_info, "qgroup-rescan", flags, 1, 0); + fs_info->discard_ctl.discard_workers = + alloc_workqueue("btrfs_discard", WQ_UNBOUND | WQ_FREEZABLE, 1); if (!(fs_info->workers && fs_info->delalloc_workers && fs_info->flush_workers && @@ -2158,7 +2163,8 @@ static int btrfs_init_workqueues(struct btrfs_fs_info *fs_info, fs_info->endio_freespace_worker && fs_info->rmw_workers && fs_info->caching_workers && fs_info->readahead_workers && fs_info->fixup_workers && fs_info->delayed_workers && - fs_info->qgroup_rescan_workers)) { + fs_info->qgroup_rescan_workers && + fs_info->discard_ctl.discard_workers)) { return -ENOMEM; } @@ -2792,6 +2798,7 @@ int __cold open_ctree(struct super_block *sb, btrfs_init_dev_replace_locks(fs_info); btrfs_init_qgroup(fs_info); + btrfs_discard_init(fs_info); btrfs_init_free_cluster(&fs_info->meta_alloc_cluster); btrfs_init_free_cluster(&fs_info->data_alloc_cluster); @@ -3082,20 +3089,13 @@ int __cold open_ctree(struct super_block *sb, btrfs_free_extra_devids(fs_devices, 1); - ret = btrfs_sysfs_add_fsid(fs_devices, NULL); + ret = btrfs_sysfs_add_fsid(fs_devices); if (ret) { btrfs_err(fs_info, "failed to init sysfs fsid interface: %d", ret); goto fail_block_groups; } - ret = btrfs_sysfs_add_device(fs_devices); - if (ret) { - btrfs_err(fs_info, "failed to init sysfs device interface: %d", - ret); - goto fail_fsdev_sysfs; - } - ret = btrfs_sysfs_add_mounted(fs_info); if (ret) { btrfs_err(fs_info, "failed to init sysfs interface: %d", ret); @@ -3262,6 +3262,7 @@ int __cold open_ctree(struct super_block *sb, } btrfs_qgroup_rescan_resume(fs_info); + btrfs_discard_resume(fs_info); if (!fs_info->uuid_root) { btrfs_info(fs_info, "creating UUID tree"); @@ -3978,6 +3979,9 @@ void __cold close_ctree(struct btrfs_fs_info *fs_info) cancel_work_sync(&fs_info->async_reclaim_work); + /* Cancel or finish ongoing discard work */ + btrfs_discard_cleanup(fs_info); + if (!sb_rdonly(fs_info->sb)) { /* * The cleaner kthread is stopped, so do one final pass over @@ -4026,11 +4030,18 @@ void __cold close_ctree(struct btrfs_fs_info *fs_info) invalidate_inode_pages2(fs_info->btree_inode->i_mapping); btrfs_stop_all_workers(fs_info); - btrfs_free_block_groups(fs_info); - clear_bit(BTRFS_FS_OPEN, &fs_info->flags); free_root_pointers(fs_info, true); + /* + * We must free the block groups after dropping the fs_roots as we could + * have had an IO error and have left over tree log blocks that aren't + * cleaned up until the fs roots are freed. This makes the block group + * accounting appear to be wrong because there's pending reserved bytes, + * so make sure we do the block group cleanup afterwards. + */ + btrfs_free_block_groups(fs_info); + iput(fs_info->btree_inode); #ifdef CONFIG_BTRFS_FS_CHECK_INTEGRITY diff --git a/fs/btrfs/disk-io.h b/fs/btrfs/disk-io.h index 76f123ebb292..8c2d6cf1ce59 100644 --- a/fs/btrfs/disk-io.h +++ b/fs/btrfs/disk-io.h @@ -134,8 +134,8 @@ struct btrfs_root *btrfs_create_tree(struct btrfs_trans_handle *trans, int btree_lock_page_hook(struct page *page, void *data, void (*flush_fn)(void *)); struct extent_map *btree_get_extent(struct btrfs_inode *inode, - struct page *page, size_t pg_offset, u64 start, u64 len, - int create); + struct page *page, size_t pg_offset, + u64 start, u64 len); int btrfs_get_num_tolerated_disk_barrier_failures(u64 flags); int __init btrfs_end_io_wq_init(void); void __cold btrfs_end_io_wq_exit(void); diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 153f71a5bba9..0163fdd59f8f 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -32,6 +32,7 @@ #include "block-rsv.h" #include "delalloc-space.h" #include "block-group.h" +#include "discard.h" #undef SCRAMBLE_DELAYED_REFS @@ -1869,8 +1870,8 @@ static int cleanup_ref_head(struct btrfs_trans_handle *trans, btrfs_pin_extent(fs_info, head->bytenr, head->num_bytes, 1); if (head->is_data) { - ret = btrfs_del_csums(trans, fs_info, head->bytenr, - head->num_bytes); + ret = btrfs_del_csums(trans, fs_info->csum_root, + head->bytenr, head->num_bytes); } } @@ -2923,7 +2924,7 @@ int btrfs_finish_extent_commit(struct btrfs_trans_handle *trans) break; } - if (btrfs_test_opt(fs_info, DISCARD)) + if (btrfs_test_opt(fs_info, DISCARD_SYNC)) ret = btrfs_discard_extent(fs_info, start, end + 1 - start, NULL); @@ -2934,6 +2935,11 @@ int btrfs_finish_extent_commit(struct btrfs_trans_handle *trans) cond_resched(); } + if (btrfs_test_opt(fs_info, DISCARD_ASYNC)) { + btrfs_discard_calc_delay(&fs_info->discard_ctl); + btrfs_discard_schedule_work(&fs_info->discard_ctl, true); + } + /* * Transaction is finished. We don't need the lock anymore. We * do need to clean up the block groups in case of a transaction @@ -3175,7 +3181,8 @@ static int __btrfs_free_extent(struct btrfs_trans_handle *trans, btrfs_release_path(path); if (is_data) { - ret = btrfs_del_csums(trans, info, bytenr, num_bytes); + ret = btrfs_del_csums(trans, info->csum_root, bytenr, + num_bytes); if (ret) { btrfs_abort_transaction(trans, ret); goto out; @@ -3437,7 +3444,6 @@ btrfs_release_block_group(struct btrfs_block_group *cache, */ struct find_free_extent_ctl { /* Basic allocation info */ - u64 ram_bytes; u64 num_bytes; u64 empty_size; u64 flags; @@ -3799,6 +3805,7 @@ static noinline int find_free_extent(struct btrfs_fs_info *fs_info, u64 flags, int delalloc) { int ret = 0; + int cache_block_group_error = 0; struct btrfs_free_cluster *last_ptr = NULL; struct btrfs_block_group *block_group = NULL; struct find_free_extent_ctl ffe_ctl = {0}; @@ -3808,7 +3815,6 @@ static noinline int find_free_extent(struct btrfs_fs_info *fs_info, WARN_ON(num_bytes < fs_info->sectorsize); - ffe_ctl.ram_bytes = ram_bytes; ffe_ctl.num_bytes = num_bytes; ffe_ctl.empty_size = empty_size; ffe_ctl.flags = flags; @@ -3958,7 +3964,20 @@ have_block_group: if (unlikely(!ffe_ctl.cached)) { ffe_ctl.have_caching_bg = true; ret = btrfs_cache_block_group(block_group, 0); - BUG_ON(ret < 0); + + /* + * If we get ENOMEM here or something else we want to + * try other block groups, because it may not be fatal. + * However if we can't find anything else we need to + * save our return here so that we return the actual + * error that caused problems, not ENOSPC. + */ + if (ret < 0) { + if (!cache_block_group_error) + cache_block_group_error = ret; + ret = 0; + goto loop; + } ret = 0; } @@ -4045,7 +4064,7 @@ loop: if (ret > 0) goto search; - if (ret == -ENOSPC) { + if (ret == -ENOSPC && !cache_block_group_error) { /* * Use ffe_ctl->total_free_space as fallback if we can't find * any contiguous hole. @@ -4056,6 +4075,8 @@ loop: space_info->max_extent_size = ffe_ctl.max_extent_size; spin_unlock(&space_info->lock); ins->offset = ffe_ctl.max_extent_size; + } else if (ret == -ENOSPC) { + ret = cache_block_group_error; } return ret; } @@ -4148,12 +4169,10 @@ again: return ret; } -static int __btrfs_free_reserved_extent(struct btrfs_fs_info *fs_info, - u64 start, u64 len, - int pin, int delalloc) +int btrfs_free_reserved_extent(struct btrfs_fs_info *fs_info, + u64 start, u64 len, int delalloc) { struct btrfs_block_group *cache; - int ret = 0; cache = btrfs_lookup_block_group(fs_info, start); if (!cache) { @@ -4162,30 +4181,28 @@ static int __btrfs_free_reserved_extent(struct btrfs_fs_info *fs_info, return -ENOSPC; } - if (pin) - pin_down_extent(cache, start, len, 1); - else { - if (btrfs_test_opt(fs_info, DISCARD)) - ret = btrfs_discard_extent(fs_info, start, len, NULL); - btrfs_add_free_space(cache, start, len); - btrfs_free_reserved_bytes(cache, len, delalloc); - trace_btrfs_reserved_extent_free(fs_info, start, len); - } + btrfs_add_free_space(cache, start, len); + btrfs_free_reserved_bytes(cache, len, delalloc); + trace_btrfs_reserved_extent_free(fs_info, start, len); btrfs_put_block_group(cache); - return ret; + return 0; } -int btrfs_free_reserved_extent(struct btrfs_fs_info *fs_info, - u64 start, u64 len, int delalloc) +int btrfs_pin_reserved_extent(struct btrfs_fs_info *fs_info, u64 start, u64 len) { - return __btrfs_free_reserved_extent(fs_info, start, len, 0, delalloc); -} + struct btrfs_block_group *cache; + int ret = 0; -int btrfs_free_and_pin_reserved_extent(struct btrfs_fs_info *fs_info, - u64 start, u64 len) -{ - return __btrfs_free_reserved_extent(fs_info, start, len, 1, 0); + cache = btrfs_lookup_block_group(fs_info, start); + if (!cache) { + btrfs_err(fs_info, "unable to find block group for %llu", start); + return -ENOSPC; + } + + ret = pin_down_extent(cache, start, len, 1); + btrfs_put_block_group(cache); + return ret; } static int alloc_reserved_file_extent(struct btrfs_trans_handle *trans, diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index eb8bd0258360..e2d30287e2d5 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -3043,7 +3043,7 @@ __get_extent_map(struct inode *inode, struct page *page, size_t pg_offset, *em_cached = NULL; } - em = get_extent(BTRFS_I(inode), page, pg_offset, start, len, 0); + em = get_extent(BTRFS_I(inode), page, pg_offset, start, len); if (em_cached && !IS_ERR_OR_NULL(em)) { BUG_ON(*em_cached); refcount_inc(&em->refs); @@ -3455,11 +3455,6 @@ static noinline_for_stack int __extent_writepage_io(struct inode *inode, update_nr_written(wbc, nr_written + 1); end = page_end; - if (i_size <= start) { - btrfs_writepage_endio_finish_ordered(page, start, page_end, 1); - goto done; - } - blocksize = inode->i_sb->s_blocksize; while (cur <= end) { @@ -3471,8 +3466,8 @@ static noinline_for_stack int __extent_writepage_io(struct inode *inode, page_end, 1); break; } - em = btrfs_get_extent(BTRFS_I(inode), page, pg_offset, cur, - end - cur + 1, 1); + em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, cur, + end - cur + 1); if (IS_ERR_OR_NULL(em)) { SetPageError(page); ret = PTR_ERR_OR_ZERO(em); @@ -3497,22 +3492,11 @@ static noinline_for_stack int __extent_writepage_io(struct inode *inode, */ if (compressed || block_start == EXTENT_MAP_HOLE || block_start == EXTENT_MAP_INLINE) { - /* - * end_io notification does not happen here for - * compressed extents - */ - if (!compressed) - btrfs_writepage_endio_finish_ordered(page, cur, - cur + iosize - 1, - 1); - else if (compressed) { - /* we don't want to end_page_writeback on - * a compressed extent. this happens - * elsewhere - */ + if (compressed) nr++; - } - + else + btrfs_writepage_endio_finish_ordered(page, cur, + cur + iosize - 1, 1); cur += iosize; pg_offset += iosize; continue; @@ -3540,7 +3524,6 @@ static noinline_for_stack int __extent_writepage_io(struct inode *inode, pg_offset += iosize; nr++; } -done: *nr_ret = nr; return ret; } @@ -3562,7 +3545,7 @@ static int __extent_writepage(struct page *page, struct writeback_control *wbc, u64 page_end = start + PAGE_SIZE - 1; int ret; int nr = 0; - size_t pg_offset = 0; + size_t pg_offset; loff_t i_size = i_size_read(inode); unsigned long end_index = i_size >> PAGE_SHIFT; unsigned long nr_written = 0; @@ -3591,14 +3574,12 @@ static int __extent_writepage(struct page *page, struct writeback_control *wbc, flush_dcache_page(page); } - pg_offset = 0; - set_page_extent_mapped(page); if (!epd->extent_locked) { ret = writepage_delalloc(inode, page, wbc, start, &nr_written); if (ret == 1) - goto done_unlocked; + return 0; if (ret) goto done; } @@ -3606,7 +3587,7 @@ static int __extent_writepage(struct page *page, struct writeback_control *wbc, ret = __extent_writepage_io(inode, page, wbc, epd, i_size, nr_written, &nr); if (ret == 1) - goto done_unlocked; + return 0; done: if (nr == 0) { @@ -3621,9 +3602,6 @@ done: unlock_page(page); ASSERT(ret <= 0); return ret; - -done_unlocked: - return 0; } void wait_on_extent_buffer_writeback(struct extent_buffer *eb) @@ -3941,6 +3919,11 @@ int btree_write_cache_pages(struct address_space *mapping, if (wbc->range_cyclic) { index = mapping->writeback_index; /* Start from prev offset */ end = -1; + /* + * Start from the beginning does not need to cycle over the + * range, mark it as scanned. + */ + scanned = (index == 0); } else { index = wbc->range_start >> PAGE_SHIFT; end = wbc->range_end >> PAGE_SHIFT; @@ -3958,7 +3941,6 @@ retry: tag))) { unsigned i; - scanned = 1; for (i = 0; i < nr_pages; i++) { struct page *page = pvec.pages[i]; @@ -4087,6 +4069,11 @@ static int extent_write_cache_pages(struct address_space *mapping, if (wbc->range_cyclic) { index = mapping->writeback_index; /* Start from prev offset */ end = -1; + /* + * Start from the beginning does not need to cycle over the + * range, mark it as scanned. + */ + scanned = (index == 0); } else { index = wbc->range_start >> PAGE_SHIFT; end = wbc->range_end >> PAGE_SHIFT; @@ -4120,7 +4107,6 @@ retry: &index, end, tag))) { unsigned i; - scanned = 1; for (i = 0; i < nr_pages; i++) { struct page *page = pvec.pages[i]; @@ -5074,12 +5060,14 @@ struct extent_buffer *alloc_test_extent_buffer(struct btrfs_fs_info *fs_info, return eb; eb = alloc_dummy_extent_buffer(fs_info, start); if (!eb) - return NULL; + return ERR_PTR(-ENOMEM); eb->fs_info = fs_info; again: ret = radix_tree_preload(GFP_NOFS); - if (ret) + if (ret) { + exists = ERR_PTR(ret); goto free_eb; + } spin_lock(&fs_info->buffer_lock); ret = radix_tree_insert(&fs_info->buffer_radix, start >> PAGE_SHIFT, eb); diff --git a/fs/btrfs/extent_io.h b/fs/btrfs/extent_io.h index a8551a1f56e2..5d205bbaafdc 100644 --- a/fs/btrfs/extent_io.h +++ b/fs/btrfs/extent_io.h @@ -183,10 +183,8 @@ static inline int extent_compress_type(unsigned long bio_flags) struct extent_map_tree; typedef struct extent_map *(get_extent_t)(struct btrfs_inode *inode, - struct page *page, - size_t pg_offset, - u64 start, u64 len, - int create); + struct page *page, size_t pg_offset, + u64 start, u64 len); int try_release_extent_mapping(struct page *page, gfp_t mask); int try_release_extent_buffer(struct page *page); diff --git a/fs/btrfs/file-item.c b/fs/btrfs/file-item.c index 3270a40b0777..c2f365662d55 100644 --- a/fs/btrfs/file-item.c +++ b/fs/btrfs/file-item.c @@ -148,8 +148,19 @@ int btrfs_lookup_file_extent(struct btrfs_trans_handle *trans, return ret; } -static blk_status_t __btrfs_lookup_bio_sums(struct inode *inode, struct bio *bio, - u64 logical_offset, u8 *dst, int dio) +/** + * btrfs_lookup_bio_sums - Look up checksums for a bio. + * @inode: inode that the bio is for. + * @bio: bio embedded in btrfs_io_bio. + * @offset: Unless (u64)-1, look up checksums for this offset in the file. + * If (u64)-1, use the page offsets from the bio instead. + * @dst: Buffer of size btrfs_super_csum_size() used to return checksum. If + * NULL, the checksum is returned in btrfs_io_bio(bio)->csum instead. + * + * Return: BLK_STS_RESOURCE if allocating memory fails, BLK_STS_OK otherwise. + */ +blk_status_t btrfs_lookup_bio_sums(struct inode *inode, struct bio *bio, + u64 offset, u8 *dst) { struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb); struct bio_vec bvec; @@ -158,8 +169,8 @@ static blk_status_t __btrfs_lookup_bio_sums(struct inode *inode, struct bio *bio struct btrfs_csum_item *item = NULL; struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree; struct btrfs_path *path; + const bool page_offsets = (offset == (u64)-1); u8 *csum; - u64 offset = 0; u64 item_start_offset = 0; u64 item_last_offset = 0; u64 disk_bytenr; @@ -205,15 +216,13 @@ static blk_status_t __btrfs_lookup_bio_sums(struct inode *inode, struct bio *bio } disk_bytenr = (u64)bio->bi_iter.bi_sector << 9; - if (dio) - offset = logical_offset; bio_for_each_segment(bvec, bio, iter) { page_bytes_left = bvec.bv_len; if (count) goto next; - if (!dio) + if (page_offsets) offset = page_offset(bvec.bv_page) + bvec.bv_offset; count = btrfs_find_ordered_sum(inode, offset, disk_bytenr, csum, nblocks); @@ -274,7 +283,8 @@ found: csum += count * csum_size; nblocks -= count; next: - while (count--) { + while (count > 0) { + count--; disk_bytenr += fs_info->sectorsize; offset += fs_info->sectorsize; page_bytes_left -= fs_info->sectorsize; @@ -285,18 +295,7 @@ next: WARN_ON_ONCE(count); btrfs_free_path(path); - return 0; -} - -blk_status_t btrfs_lookup_bio_sums(struct inode *inode, struct bio *bio, - u8 *dst) -{ - return __btrfs_lookup_bio_sums(inode, bio, 0, dst, 0); -} - -blk_status_t btrfs_lookup_bio_sums_dio(struct inode *inode, struct bio *bio, u64 offset) -{ - return __btrfs_lookup_bio_sums(inode, bio, offset, NULL, 1); + return BLK_STS_OK; } int btrfs_lookup_csums_range(struct btrfs_root *root, u64 start, u64 end, @@ -483,8 +482,8 @@ blk_status_t btrfs_csum_one_bio(struct inode *inode, struct bio *bio, - 1); for (i = 0; i < nr_sectors; i++) { - if (offset >= ordered->file_offset + ordered->len || - offset < ordered->file_offset) { + if (offset >= ordered->file_offset + ordered->num_bytes || + offset < ordered->file_offset) { unsigned long bytes_left; sums->len = this_sum_bytes; @@ -590,9 +589,9 @@ static noinline void truncate_one_csum(struct btrfs_fs_info *fs_info, * range of bytes. */ int btrfs_del_csums(struct btrfs_trans_handle *trans, - struct btrfs_fs_info *fs_info, u64 bytenr, u64 len) + struct btrfs_root *root, u64 bytenr, u64 len) { - struct btrfs_root *root = fs_info->csum_root; + struct btrfs_fs_info *fs_info = trans->fs_info; struct btrfs_path *path; struct btrfs_key key; u64 end_byte = bytenr + len; @@ -602,6 +601,9 @@ int btrfs_del_csums(struct btrfs_trans_handle *trans, u16 csum_size = btrfs_super_csum_size(fs_info->super_copy); int blocksize_bits = fs_info->sb->s_blocksize_bits; + ASSERT(root == fs_info->csum_root || + root->root_key.objectid == BTRFS_TREE_LOG_OBJECTID); + path = btrfs_alloc_path(); if (!path) return -ENOMEM; diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 0cb43b682789..a16da274c9aa 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -477,8 +477,7 @@ static int btrfs_find_new_delalloc_bytes(struct btrfs_inode *inode, u64 em_len; int ret = 0; - em = btrfs_get_extent(inode, NULL, 0, search_start, - search_len, 0); + em = btrfs_get_extent(inode, NULL, 0, search_start, search_len); if (IS_ERR(em)) return PTR_ERR(em); @@ -1501,7 +1500,7 @@ lock_and_cleanup_extent_if_need(struct btrfs_inode *inode, struct page **pages, ordered = btrfs_lookup_ordered_range(inode, start_pos, last_pos - start_pos + 1); if (ordered && - ordered->file_offset + ordered->len > start_pos && + ordered->file_offset + ordered->num_bytes > start_pos && ordered->file_offset <= last_pos) { unlock_extent_cached(&inode->io_tree, start_pos, last_pos, cached_state); @@ -2390,7 +2389,7 @@ static int find_first_non_hole(struct inode *inode, u64 *start, u64 *len) em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, round_down(*start, fs_info->sectorsize), - round_up(*len, fs_info->sectorsize), 0); + round_up(*len, fs_info->sectorsize)); if (IS_ERR(em)) return PTR_ERR(em); @@ -2426,7 +2425,7 @@ static int btrfs_punch_hole_lock_range(struct inode *inode, * we need to try again. */ if ((!ordered || - (ordered->file_offset + ordered->len <= lockstart || + (ordered->file_offset + ordered->num_bytes <= lockstart || ordered->file_offset > lockend)) && !filemap_range_has_page(inode->i_mapping, lockstart, lockend)) { @@ -2599,8 +2598,8 @@ int btrfs_punch_hole_range(struct inode *inode, struct btrfs_path *path, } } - if (clone_info) { - u64 clone_len = drop_end - cur_offset; + if (clone_info && drop_end > clone_info->file_offset) { + u64 clone_len = drop_end - clone_info->file_offset; ret = btrfs_insert_clone_extent(trans, inode, path, clone_info, clone_len); @@ -2957,7 +2956,7 @@ static int btrfs_zero_range_check_range_boundary(struct inode *inode, int ret; offset = round_down(offset, sectorsize); - em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize, 0); + em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize); if (IS_ERR(em)) return PTR_ERR(em); @@ -2990,8 +2989,8 @@ static int btrfs_zero_range(struct inode *inode, inode_dio_wait(inode); - em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, - alloc_start, alloc_end - alloc_start, 0); + em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, alloc_start, + alloc_end - alloc_start); if (IS_ERR(em)) { ret = PTR_ERR(em); goto out; @@ -3034,8 +3033,8 @@ static int btrfs_zero_range(struct inode *inode, if (BTRFS_BYTES_TO_BLKS(fs_info, offset) == BTRFS_BYTES_TO_BLKS(fs_info, offset + len - 1)) { - em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, - alloc_start, sectorsize, 0); + em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, alloc_start, + sectorsize); if (IS_ERR(em)) { ret = PTR_ERR(em); goto out; @@ -3248,7 +3247,7 @@ static long btrfs_fallocate(struct file *file, int mode, ordered = btrfs_lookup_first_ordered_extent(inode, locked_end); if (ordered && - ordered->file_offset + ordered->len > alloc_start && + ordered->file_offset + ordered->num_bytes > alloc_start && ordered->file_offset < alloc_end) { btrfs_put_ordered_extent(ordered); unlock_extent_cached(&BTRFS_I(inode)->io_tree, @@ -3273,7 +3272,7 @@ static long btrfs_fallocate(struct file *file, int mode, INIT_LIST_HEAD(&reserve_list); while (cur_offset < alloc_end) { em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, cur_offset, - alloc_end - cur_offset, 0); + alloc_end - cur_offset); if (IS_ERR(em)) { ret = PTR_ERR(em); break; diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c index 3283da419200..0598fd3c6e3f 100644 --- a/fs/btrfs/free-space-cache.c +++ b/fs/btrfs/free-space-cache.c @@ -21,9 +21,11 @@ #include "space-info.h" #include "delalloc-space.h" #include "block-group.h" +#include "discard.h" #define BITS_PER_BITMAP (PAGE_SIZE * 8UL) -#define MAX_CACHE_BYTES_PER_GIG SZ_32K +#define MAX_CACHE_BYTES_PER_GIG SZ_64K +#define FORCE_EXTENT_THRESHOLD SZ_1M struct btrfs_trim_range { u64 start; @@ -31,6 +33,8 @@ struct btrfs_trim_range { struct list_head list; }; +static int count_bitmap_extents(struct btrfs_free_space_ctl *ctl, + struct btrfs_free_space *bitmap_info); static int link_free_space(struct btrfs_free_space_ctl *ctl, struct btrfs_free_space *info); static void unlink_free_space(struct btrfs_free_space_ctl *ctl, @@ -752,6 +756,16 @@ static int __load_free_space_cache(struct btrfs_root *root, struct inode *inode, goto free_cache; } + /* + * Sync discard ensures that the free space cache is always + * trimmed. So when reading this in, the state should reflect + * that. We also do this for async as a stop gap for lack of + * persistence. + */ + if (btrfs_test_opt(fs_info, DISCARD_SYNC) || + btrfs_test_opt(fs_info, DISCARD_ASYNC)) + e->trim_state = BTRFS_TRIM_STATE_TRIMMED; + if (!e->bytes) { kmem_cache_free(btrfs_free_space_cachep, e); goto free_cache; @@ -805,12 +819,19 @@ static int __load_free_space_cache(struct btrfs_root *root, struct inode *inode, ret = io_ctl_read_bitmap(&io_ctl, e); if (ret) goto free_cache; + e->bitmap_extents = count_bitmap_extents(ctl, e); + if (!btrfs_free_space_trimmed(e)) { + ctl->discardable_extents[BTRFS_STAT_CURR] += + e->bitmap_extents; + ctl->discardable_bytes[BTRFS_STAT_CURR] += e->bytes; + } } io_ctl_drop_pages(&io_ctl); merge_space_tree(ctl); ret = 1; out: + btrfs_discard_update_discardable(ctl->private, ctl); io_ctl_free(&io_ctl); return ret; free_cache: @@ -1624,6 +1645,11 @@ __unlink_free_space(struct btrfs_free_space_ctl *ctl, { rb_erase(&info->offset_index, &ctl->free_space_offset); ctl->free_extents--; + + if (!info->bitmap && !btrfs_free_space_trimmed(info)) { + ctl->discardable_extents[BTRFS_STAT_CURR]--; + ctl->discardable_bytes[BTRFS_STAT_CURR] -= info->bytes; + } } static void unlink_free_space(struct btrfs_free_space_ctl *ctl, @@ -1644,6 +1670,11 @@ static int link_free_space(struct btrfs_free_space_ctl *ctl, if (ret) return ret; + if (!info->bitmap && !btrfs_free_space_trimmed(info)) { + ctl->discardable_extents[BTRFS_STAT_CURR]++; + ctl->discardable_bytes[BTRFS_STAT_CURR] += info->bytes; + } + ctl->free_space += info->bytes; ctl->free_extents++; return ret; @@ -1664,26 +1695,17 @@ static void recalculate_thresholds(struct btrfs_free_space_ctl *ctl) ASSERT(ctl->total_bitmaps <= max_bitmaps); /* - * The goal is to keep the total amount of memory used per 1gb of space - * at or below 32k, so we need to adjust how much memory we allow to be - * used by extent based free space tracking + * We are trying to keep the total amount of memory used per 1GiB of + * space to be MAX_CACHE_BYTES_PER_GIG. However, with a reclamation + * mechanism of pulling extents >= FORCE_EXTENT_THRESHOLD out of + * bitmaps, we may end up using more memory than this. */ if (size < SZ_1G) max_bytes = MAX_CACHE_BYTES_PER_GIG; else max_bytes = MAX_CACHE_BYTES_PER_GIG * div_u64(size, SZ_1G); - /* - * we want to account for 1 more bitmap than what we have so we can make - * sure we don't go over our overall goal of MAX_CACHE_BYTES_PER_GIG as - * we add more bitmaps. - */ - bitmap_bytes = (ctl->total_bitmaps + 1) * ctl->unit; - - if (bitmap_bytes >= max_bytes) { - ctl->extents_thresh = 0; - return; - } + bitmap_bytes = ctl->total_bitmaps * ctl->unit; /* * we want the extent entry threshold to always be at most 1/2 the max @@ -1700,17 +1722,31 @@ static inline void __bitmap_clear_bits(struct btrfs_free_space_ctl *ctl, struct btrfs_free_space *info, u64 offset, u64 bytes) { - unsigned long start, count; + unsigned long start, count, end; + int extent_delta = -1; start = offset_to_bit(info->offset, ctl->unit, offset); count = bytes_to_bits(bytes, ctl->unit); - ASSERT(start + count <= BITS_PER_BITMAP); + end = start + count; + ASSERT(end <= BITS_PER_BITMAP); bitmap_clear(info->bitmap, start, count); info->bytes -= bytes; if (info->max_extent_size > ctl->unit) info->max_extent_size = 0; + + if (start && test_bit(start - 1, info->bitmap)) + extent_delta++; + + if (end < BITS_PER_BITMAP && test_bit(end, info->bitmap)) + extent_delta++; + + info->bitmap_extents += extent_delta; + if (!btrfs_free_space_trimmed(info)) { + ctl->discardable_extents[BTRFS_STAT_CURR] += extent_delta; + ctl->discardable_bytes[BTRFS_STAT_CURR] -= bytes; + } } static void bitmap_clear_bits(struct btrfs_free_space_ctl *ctl, @@ -1725,16 +1761,30 @@ static void bitmap_set_bits(struct btrfs_free_space_ctl *ctl, struct btrfs_free_space *info, u64 offset, u64 bytes) { - unsigned long start, count; + unsigned long start, count, end; + int extent_delta = 1; start = offset_to_bit(info->offset, ctl->unit, offset); count = bytes_to_bits(bytes, ctl->unit); - ASSERT(start + count <= BITS_PER_BITMAP); + end = start + count; + ASSERT(end <= BITS_PER_BITMAP); bitmap_set(info->bitmap, start, count); info->bytes += bytes; ctl->free_space += bytes; + + if (start && test_bit(start - 1, info->bitmap)) + extent_delta--; + + if (end < BITS_PER_BITMAP && test_bit(end, info->bitmap)) + extent_delta--; + + info->bitmap_extents += extent_delta; + if (!btrfs_free_space_trimmed(info)) { + ctl->discardable_extents[BTRFS_STAT_CURR] += extent_delta; + ctl->discardable_bytes[BTRFS_STAT_CURR] += bytes; + } } /* @@ -1870,11 +1920,35 @@ out: return NULL; } +static int count_bitmap_extents(struct btrfs_free_space_ctl *ctl, + struct btrfs_free_space *bitmap_info) +{ + struct btrfs_block_group *block_group = ctl->private; + u64 bytes = bitmap_info->bytes; + unsigned int rs, re; + int count = 0; + + if (!block_group || !bytes) + return count; + + bitmap_for_each_set_region(bitmap_info->bitmap, rs, re, 0, + BITS_PER_BITMAP) { + bytes -= (rs - re) * ctl->unit; + count++; + + if (!bytes) + break; + } + + return count; +} + static void add_new_bitmap(struct btrfs_free_space_ctl *ctl, struct btrfs_free_space *info, u64 offset) { info->offset = offset_to_bitmap(ctl, offset); info->bytes = 0; + info->bitmap_extents = 0; INIT_LIST_HEAD(&info->list); link_free_space(ctl, info); ctl->total_bitmaps++; @@ -1885,6 +1959,18 @@ static void add_new_bitmap(struct btrfs_free_space_ctl *ctl, static void free_bitmap(struct btrfs_free_space_ctl *ctl, struct btrfs_free_space *bitmap_info) { + /* + * Normally when this is called, the bitmap is completely empty. However, + * if we are blowing up the free space cache for one reason or another + * via __btrfs_remove_free_space_cache(), then it may not be freed and + * we may leave stats on the table. + */ + if (bitmap_info->bytes && !btrfs_free_space_trimmed(bitmap_info)) { + ctl->discardable_extents[BTRFS_STAT_CURR] -= + bitmap_info->bitmap_extents; + ctl->discardable_bytes[BTRFS_STAT_CURR] -= bitmap_info->bytes; + + } unlink_free_space(ctl, bitmap_info); kmem_cache_free(btrfs_free_space_bitmap_cachep, bitmap_info->bitmap); kmem_cache_free(btrfs_free_space_cachep, bitmap_info); @@ -1971,11 +2057,24 @@ again: static u64 add_bytes_to_bitmap(struct btrfs_free_space_ctl *ctl, struct btrfs_free_space *info, u64 offset, - u64 bytes) + u64 bytes, enum btrfs_trim_state trim_state) { u64 bytes_to_set = 0; u64 end; + /* + * This is a tradeoff to make bitmap trim state minimal. We mark the + * whole bitmap untrimmed if at any point we add untrimmed regions. + */ + if (trim_state == BTRFS_TRIM_STATE_UNTRIMMED) { + if (btrfs_free_space_trimmed(info)) { + ctl->discardable_extents[BTRFS_STAT_CURR] += + info->bitmap_extents; + ctl->discardable_bytes[BTRFS_STAT_CURR] += info->bytes; + } + info->trim_state = BTRFS_TRIM_STATE_UNTRIMMED; + } + end = info->offset + (u64)(BITS_PER_BITMAP * ctl->unit); bytes_to_set = min(end - offset, bytes); @@ -2004,6 +2103,10 @@ static bool use_bitmap(struct btrfs_free_space_ctl *ctl, forced = true; #endif + /* This is a way to reclaim large regions from the bitmaps. */ + if (!forced && info->bytes >= FORCE_EXTENT_THRESHOLD) + return false; + /* * If we are below the extents threshold then we can add this as an * extent, and don't have to deal with the bitmap @@ -2016,8 +2119,8 @@ static bool use_bitmap(struct btrfs_free_space_ctl *ctl, * of cache left then go ahead an dadd them, no sense in adding * the overhead of a bitmap if we don't have to. */ - if (info->bytes <= fs_info->sectorsize * 4) { - if (ctl->free_extents * 2 <= ctl->extents_thresh) + if (info->bytes <= fs_info->sectorsize * 8) { + if (ctl->free_extents * 3 <= ctl->extents_thresh) return false; } else { return false; @@ -2050,10 +2153,12 @@ static int insert_into_bitmap(struct btrfs_free_space_ctl *ctl, struct btrfs_block_group *block_group = NULL; int added = 0; u64 bytes, offset, bytes_added; + enum btrfs_trim_state trim_state; int ret; bytes = info->bytes; offset = info->offset; + trim_state = info->trim_state; if (!ctl->op->use_bitmap(ctl, info)) return 0; @@ -2088,8 +2193,8 @@ again: } if (entry->offset == offset_to_bitmap(ctl, offset)) { - bytes_added = add_bytes_to_bitmap(ctl, entry, - offset, bytes); + bytes_added = add_bytes_to_bitmap(ctl, entry, offset, + bytes, trim_state); bytes -= bytes_added; offset += bytes_added; } @@ -2108,7 +2213,8 @@ no_cluster_bitmap: goto new_bitmap; } - bytes_added = add_bytes_to_bitmap(ctl, bitmap_info, offset, bytes); + bytes_added = add_bytes_to_bitmap(ctl, bitmap_info, offset, bytes, + trim_state); bytes -= bytes_added; offset += bytes_added; added = 0; @@ -2142,6 +2248,7 @@ new_bitmap: /* allocate the bitmap */ info->bitmap = kmem_cache_zalloc(btrfs_free_space_bitmap_cachep, GFP_NOFS); + info->trim_state = BTRFS_TRIM_STATE_TRIMMED; spin_lock(&ctl->tree_lock); if (!info->bitmap) { ret = -ENOMEM; @@ -2161,6 +2268,22 @@ out: return ret; } +/* + * Free space merging rules: + * 1) Merge trimmed areas together + * 2) Let untrimmed areas coalesce with trimmed areas + * 3) Always pull neighboring regions from bitmaps + * + * The above rules are for when we merge free space based on btrfs_trim_state. + * Rules 2 and 3 are subtle because they are suboptimal, but are done for the + * same reason: to promote larger extent regions which makes life easier for + * find_free_extent(). Rule 2 enables coalescing based on the common path + * being returning free space from btrfs_finish_extent_commit(). So when free + * space is trimmed, it will prevent aggregating trimmed new region and + * untrimmed regions in the rb_tree. Rule 3 is purely to obtain larger extents + * and provide find_free_extent() with the largest extents possible hoping for + * the reuse path. + */ static bool try_merge_free_space(struct btrfs_free_space_ctl *ctl, struct btrfs_free_space *info, bool update_stat) { @@ -2169,6 +2292,7 @@ static bool try_merge_free_space(struct btrfs_free_space_ctl *ctl, bool merged = false; u64 offset = info->offset; u64 bytes = info->bytes; + const bool is_trimmed = btrfs_free_space_trimmed(info); /* * first we want to see if there is free space adjacent to the range we @@ -2182,7 +2306,9 @@ static bool try_merge_free_space(struct btrfs_free_space_ctl *ctl, else left_info = tree_search_offset(ctl, offset - 1, 0, 0); - if (right_info && !right_info->bitmap) { + /* See try_merge_free_space() comment. */ + if (right_info && !right_info->bitmap && + (!is_trimmed || btrfs_free_space_trimmed(right_info))) { if (update_stat) unlink_free_space(ctl, right_info); else @@ -2192,8 +2318,10 @@ static bool try_merge_free_space(struct btrfs_free_space_ctl *ctl, merged = true; } + /* See try_merge_free_space() comment. */ if (left_info && !left_info->bitmap && - left_info->offset + left_info->bytes == offset) { + left_info->offset + left_info->bytes == offset && + (!is_trimmed || btrfs_free_space_trimmed(left_info))) { if (update_stat) unlink_free_space(ctl, left_info); else @@ -2229,6 +2357,10 @@ static bool steal_from_bitmap_to_end(struct btrfs_free_space_ctl *ctl, bytes = (j - i) * ctl->unit; info->bytes += bytes; + /* See try_merge_free_space() comment. */ + if (!btrfs_free_space_trimmed(bitmap)) + info->trim_state = BTRFS_TRIM_STATE_UNTRIMMED; + if (update_stat) bitmap_clear_bits(ctl, bitmap, end, bytes); else @@ -2282,6 +2414,10 @@ static bool steal_from_bitmap_to_front(struct btrfs_free_space_ctl *ctl, info->offset -= bytes; info->bytes += bytes; + /* See try_merge_free_space() comment. */ + if (!btrfs_free_space_trimmed(bitmap)) + info->trim_state = BTRFS_TRIM_STATE_UNTRIMMED; + if (update_stat) bitmap_clear_bits(ctl, bitmap, info->offset, bytes); else @@ -2331,10 +2467,13 @@ static void steal_from_bitmap(struct btrfs_free_space_ctl *ctl, int __btrfs_add_free_space(struct btrfs_fs_info *fs_info, struct btrfs_free_space_ctl *ctl, - u64 offset, u64 bytes) + u64 offset, u64 bytes, + enum btrfs_trim_state trim_state) { + struct btrfs_block_group *block_group = ctl->private; struct btrfs_free_space *info; int ret = 0; + u64 filter_bytes = bytes; info = kmem_cache_zalloc(btrfs_free_space_cachep, GFP_NOFS); if (!info) @@ -2342,6 +2481,7 @@ int __btrfs_add_free_space(struct btrfs_fs_info *fs_info, info->offset = offset; info->bytes = bytes; + info->trim_state = trim_state; RB_CLEAR_NODE(&info->offset_index); spin_lock(&ctl->tree_lock); @@ -2370,10 +2510,13 @@ link: */ steal_from_bitmap(ctl, info, true); + filter_bytes = max(filter_bytes, info->bytes); + ret = link_free_space(ctl, info); if (ret) kmem_cache_free(btrfs_free_space_cachep, info); out: + btrfs_discard_update_discardable(block_group, ctl); spin_unlock(&ctl->tree_lock); if (ret) { @@ -2381,15 +2524,44 @@ out: ASSERT(ret != -EEXIST); } + if (trim_state != BTRFS_TRIM_STATE_TRIMMED) { + btrfs_discard_check_filter(block_group, filter_bytes); + btrfs_discard_queue_work(&fs_info->discard_ctl, block_group); + } + return ret; } int btrfs_add_free_space(struct btrfs_block_group *block_group, u64 bytenr, u64 size) { + enum btrfs_trim_state trim_state = BTRFS_TRIM_STATE_UNTRIMMED; + + if (btrfs_test_opt(block_group->fs_info, DISCARD_SYNC)) + trim_state = BTRFS_TRIM_STATE_TRIMMED; + + return __btrfs_add_free_space(block_group->fs_info, + block_group->free_space_ctl, + bytenr, size, trim_state); +} + +/* + * This is a subtle distinction because when adding free space back in general, + * we want it to be added as untrimmed for async. But in the case where we add + * it on loading of a block group, we want to consider it trimmed. + */ +int btrfs_add_free_space_async_trimmed(struct btrfs_block_group *block_group, + u64 bytenr, u64 size) +{ + enum btrfs_trim_state trim_state = BTRFS_TRIM_STATE_UNTRIMMED; + + if (btrfs_test_opt(block_group->fs_info, DISCARD_SYNC) || + btrfs_test_opt(block_group->fs_info, DISCARD_ASYNC)) + trim_state = BTRFS_TRIM_STATE_TRIMMED; + return __btrfs_add_free_space(block_group->fs_info, block_group->free_space_ctl, - bytenr, size); + bytenr, size, trim_state); } int btrfs_remove_free_space(struct btrfs_block_group *block_group, @@ -2464,8 +2636,10 @@ again: } spin_unlock(&ctl->tree_lock); - ret = btrfs_add_free_space(block_group, offset + bytes, - old_end - (offset + bytes)); + ret = __btrfs_add_free_space(block_group->fs_info, ctl, + offset + bytes, + old_end - (offset + bytes), + info->trim_state); WARN_ON(ret); goto out; } @@ -2477,6 +2651,7 @@ again: goto again; } out_lock: + btrfs_discard_update_discardable(block_group, ctl); spin_unlock(&ctl->tree_lock); out: return ret; @@ -2562,8 +2737,22 @@ __btrfs_return_cluster_to_free_space( bitmap = (entry->bitmap != NULL); if (!bitmap) { + /* Merging treats extents as if they were new */ + if (!btrfs_free_space_trimmed(entry)) { + ctl->discardable_extents[BTRFS_STAT_CURR]--; + ctl->discardable_bytes[BTRFS_STAT_CURR] -= + entry->bytes; + } + try_merge_free_space(ctl, entry, false); steal_from_bitmap(ctl, entry, false); + + /* As we insert directly, update these statistics */ + if (!btrfs_free_space_trimmed(entry)) { + ctl->discardable_extents[BTRFS_STAT_CURR]++; + ctl->discardable_bytes[BTRFS_STAT_CURR] += + entry->bytes; + } } tree_insert_offset(&ctl->free_space_offset, entry->offset, &entry->offset_index, bitmap); @@ -2599,6 +2788,8 @@ void __btrfs_remove_free_space_cache(struct btrfs_free_space_ctl *ctl) { spin_lock(&ctl->tree_lock); __btrfs_remove_free_space_cache_locked(ctl); + if (ctl->private) + btrfs_discard_update_discardable(ctl->private, ctl); spin_unlock(&ctl->tree_lock); } @@ -2620,20 +2811,55 @@ void btrfs_remove_free_space_cache(struct btrfs_block_group *block_group) cond_resched_lock(&ctl->tree_lock); } __btrfs_remove_free_space_cache_locked(ctl); + btrfs_discard_update_discardable(block_group, ctl); spin_unlock(&ctl->tree_lock); } +/** + * btrfs_is_free_space_trimmed - see if everything is trimmed + * @block_group: block_group of interest + * + * Walk @block_group's free space rb_tree to determine if everything is trimmed. + */ +bool btrfs_is_free_space_trimmed(struct btrfs_block_group *block_group) +{ + struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; + struct btrfs_free_space *info; + struct rb_node *node; + bool ret = true; + + spin_lock(&ctl->tree_lock); + node = rb_first(&ctl->free_space_offset); + + while (node) { + info = rb_entry(node, struct btrfs_free_space, offset_index); + + if (!btrfs_free_space_trimmed(info)) { + ret = false; + break; + } + + node = rb_next(node); + } + + spin_unlock(&ctl->tree_lock); + return ret; +} + u64 btrfs_find_space_for_alloc(struct btrfs_block_group *block_group, u64 offset, u64 bytes, u64 empty_size, u64 *max_extent_size) { struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; + struct btrfs_discard_ctl *discard_ctl = + &block_group->fs_info->discard_ctl; struct btrfs_free_space *entry = NULL; u64 bytes_search = bytes + empty_size; u64 ret = 0; u64 align_gap = 0; u64 align_gap_len = 0; + enum btrfs_trim_state align_gap_trim_state = BTRFS_TRIM_STATE_UNTRIMMED; spin_lock(&ctl->tree_lock); entry = find_free_space(ctl, &offset, &bytes_search, @@ -2644,12 +2870,20 @@ u64 btrfs_find_space_for_alloc(struct btrfs_block_group *block_group, ret = offset; if (entry->bitmap) { bitmap_clear_bits(ctl, entry, offset, bytes); + + if (!btrfs_free_space_trimmed(entry)) + atomic64_add(bytes, &discard_ctl->discard_bytes_saved); + if (!entry->bytes) free_bitmap(ctl, entry); } else { unlink_free_space(ctl, entry); align_gap_len = offset - entry->offset; align_gap = entry->offset; + align_gap_trim_state = entry->trim_state; + + if (!btrfs_free_space_trimmed(entry)) + atomic64_add(bytes, &discard_ctl->discard_bytes_saved); entry->offset = offset + bytes; WARN_ON(entry->bytes < bytes + align_gap_len); @@ -2661,11 +2895,13 @@ u64 btrfs_find_space_for_alloc(struct btrfs_block_group *block_group, link_free_space(ctl, entry); } out: + btrfs_discard_update_discardable(block_group, ctl); spin_unlock(&ctl->tree_lock); if (align_gap_len) __btrfs_add_free_space(block_group->fs_info, ctl, - align_gap, align_gap_len); + align_gap, align_gap_len, + align_gap_trim_state); return ret; } @@ -2707,6 +2943,8 @@ int btrfs_return_cluster_to_free_space( ret = __btrfs_return_cluster_to_free_space(block_group, cluster); spin_unlock(&ctl->tree_lock); + btrfs_discard_queue_work(&block_group->fs_info->discard_ctl, block_group); + /* finally drop our ref */ btrfs_put_block_group(block_group); return ret; @@ -2750,6 +2988,8 @@ u64 btrfs_alloc_from_cluster(struct btrfs_block_group *block_group, u64 min_start, u64 *max_extent_size) { struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; + struct btrfs_discard_ctl *discard_ctl = + &block_group->fs_info->discard_ctl; struct btrfs_free_space *entry = NULL; struct rb_node *node; u64 ret = 0; @@ -2814,7 +3054,12 @@ out: spin_lock(&ctl->tree_lock); + if (!btrfs_free_space_trimmed(entry)) + atomic64_add(bytes, &discard_ctl->discard_bytes_saved); + ctl->free_space -= bytes; + if (!entry->bitmap && !btrfs_free_space_trimmed(entry)) + ctl->discardable_bytes[BTRFS_STAT_CURR] -= bytes; if (entry->bytes == 0) { ctl->free_extents--; if (entry->bitmap) { @@ -2822,6 +3067,8 @@ out: entry->bitmap); ctl->total_bitmaps--; ctl->op->recalc_thresholds(ctl); + } else if (!btrfs_free_space_trimmed(entry)) { + ctl->discardable_extents[BTRFS_STAT_CURR]--; } kmem_cache_free(btrfs_free_space_cachep, entry); } @@ -3148,6 +3395,7 @@ void btrfs_init_free_cluster(struct btrfs_free_cluster *cluster) static int do_trimming(struct btrfs_block_group *block_group, u64 *total_trimmed, u64 start, u64 bytes, u64 reserved_start, u64 reserved_bytes, + enum btrfs_trim_state reserved_trim_state, struct btrfs_trim_range *trim_entry) { struct btrfs_space_info *space_info = block_group->space_info; @@ -3155,6 +3403,9 @@ static int do_trimming(struct btrfs_block_group *block_group, struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; int ret; int update = 0; + const u64 end = start + bytes; + const u64 reserved_end = reserved_start + reserved_bytes; + enum btrfs_trim_state trim_state = BTRFS_TRIM_STATE_UNTRIMMED; u64 trimmed = 0; spin_lock(&space_info->lock); @@ -3168,11 +3419,20 @@ static int do_trimming(struct btrfs_block_group *block_group, spin_unlock(&space_info->lock); ret = btrfs_discard_extent(fs_info, start, bytes, &trimmed); - if (!ret) + if (!ret) { *total_trimmed += trimmed; + trim_state = BTRFS_TRIM_STATE_TRIMMED; + } mutex_lock(&ctl->cache_writeout_mutex); - btrfs_add_free_space(block_group, reserved_start, reserved_bytes); + if (reserved_start < start) + __btrfs_add_free_space(fs_info, ctl, reserved_start, + start - reserved_start, + reserved_trim_state); + if (start + bytes < reserved_start + reserved_bytes) + __btrfs_add_free_space(fs_info, ctl, end, reserved_end - end, + reserved_trim_state); + __btrfs_add_free_space(fs_info, ctl, start, bytes, trim_state); list_del(&trim_entry->list); mutex_unlock(&ctl->cache_writeout_mutex); @@ -3190,16 +3450,24 @@ static int do_trimming(struct btrfs_block_group *block_group, return ret; } +/* + * If @async is set, then we will trim 1 region and return. + */ static int trim_no_bitmap(struct btrfs_block_group *block_group, - u64 *total_trimmed, u64 start, u64 end, u64 minlen) + u64 *total_trimmed, u64 start, u64 end, u64 minlen, + bool async) { + struct btrfs_discard_ctl *discard_ctl = + &block_group->fs_info->discard_ctl; struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; struct btrfs_free_space *entry; struct rb_node *node; int ret = 0; u64 extent_start; u64 extent_bytes; + enum btrfs_trim_state extent_trim_state; u64 bytes; + const u64 max_discard_size = READ_ONCE(discard_ctl->max_discard_size); while (start < end) { struct btrfs_trim_range trim_entry; @@ -3207,49 +3475,66 @@ static int trim_no_bitmap(struct btrfs_block_group *block_group, mutex_lock(&ctl->cache_writeout_mutex); spin_lock(&ctl->tree_lock); - if (ctl->free_space < minlen) { - spin_unlock(&ctl->tree_lock); - mutex_unlock(&ctl->cache_writeout_mutex); - break; - } + if (ctl->free_space < minlen) + goto out_unlock; entry = tree_search_offset(ctl, start, 0, 1); - if (!entry) { - spin_unlock(&ctl->tree_lock); - mutex_unlock(&ctl->cache_writeout_mutex); - break; - } + if (!entry) + goto out_unlock; - /* skip bitmaps */ - while (entry->bitmap) { + /* Skip bitmaps and if async, already trimmed entries */ + while (entry->bitmap || + (async && btrfs_free_space_trimmed(entry))) { node = rb_next(&entry->offset_index); - if (!node) { - spin_unlock(&ctl->tree_lock); - mutex_unlock(&ctl->cache_writeout_mutex); - goto out; - } + if (!node) + goto out_unlock; entry = rb_entry(node, struct btrfs_free_space, offset_index); } - if (entry->offset >= end) { - spin_unlock(&ctl->tree_lock); - mutex_unlock(&ctl->cache_writeout_mutex); - break; - } + if (entry->offset >= end) + goto out_unlock; extent_start = entry->offset; extent_bytes = entry->bytes; - start = max(start, extent_start); - bytes = min(extent_start + extent_bytes, end) - start; - if (bytes < minlen) { - spin_unlock(&ctl->tree_lock); - mutex_unlock(&ctl->cache_writeout_mutex); - goto next; - } + extent_trim_state = entry->trim_state; + if (async) { + start = entry->offset; + bytes = entry->bytes; + if (bytes < minlen) { + spin_unlock(&ctl->tree_lock); + mutex_unlock(&ctl->cache_writeout_mutex); + goto next; + } + unlink_free_space(ctl, entry); + /* + * Let bytes = BTRFS_MAX_DISCARD_SIZE + X. + * If X < BTRFS_ASYNC_DISCARD_MIN_FILTER, we won't trim + * X when we come back around. So trim it now. + */ + if (max_discard_size && + bytes >= (max_discard_size + + BTRFS_ASYNC_DISCARD_MIN_FILTER)) { + bytes = max_discard_size; + extent_bytes = max_discard_size; + entry->offset += max_discard_size; + entry->bytes -= max_discard_size; + link_free_space(ctl, entry); + } else { + kmem_cache_free(btrfs_free_space_cachep, entry); + } + } else { + start = max(start, extent_start); + bytes = min(extent_start + extent_bytes, end) - start; + if (bytes < minlen) { + spin_unlock(&ctl->tree_lock); + mutex_unlock(&ctl->cache_writeout_mutex); + goto next; + } - unlink_free_space(ctl, entry); - kmem_cache_free(btrfs_free_space_cachep, entry); + unlink_free_space(ctl, entry); + kmem_cache_free(btrfs_free_space_cachep, entry); + } spin_unlock(&ctl->tree_lock); trim_entry.start = extent_start; @@ -3258,11 +3543,17 @@ static int trim_no_bitmap(struct btrfs_block_group *block_group, mutex_unlock(&ctl->cache_writeout_mutex); ret = do_trimming(block_group, total_trimmed, start, bytes, - extent_start, extent_bytes, &trim_entry); - if (ret) + extent_start, extent_bytes, extent_trim_state, + &trim_entry); + if (ret) { + block_group->discard_cursor = start + bytes; break; + } next: start += bytes; + block_group->discard_cursor = start; + if (async && *total_trimmed) + break; if (fatal_signal_pending(current)) { ret = -ERESTARTSYS; @@ -3271,19 +3562,76 @@ next: cond_resched(); } -out: + + return ret; + +out_unlock: + block_group->discard_cursor = btrfs_block_group_end(block_group); + spin_unlock(&ctl->tree_lock); + mutex_unlock(&ctl->cache_writeout_mutex); + return ret; } +/* + * If we break out of trimming a bitmap prematurely, we should reset the + * trimming bit. In a rather contrieved case, it's possible to race here so + * reset the state to BTRFS_TRIM_STATE_UNTRIMMED. + * + * start = start of bitmap + * end = near end of bitmap + * + * Thread 1: Thread 2: + * trim_bitmaps(start) + * trim_bitmaps(end) + * end_trimming_bitmap() + * reset_trimming_bitmap() + */ +static void reset_trimming_bitmap(struct btrfs_free_space_ctl *ctl, u64 offset) +{ + struct btrfs_free_space *entry; + + spin_lock(&ctl->tree_lock); + entry = tree_search_offset(ctl, offset, 1, 0); + if (entry) { + if (btrfs_free_space_trimmed(entry)) { + ctl->discardable_extents[BTRFS_STAT_CURR] += + entry->bitmap_extents; + ctl->discardable_bytes[BTRFS_STAT_CURR] += entry->bytes; + } + entry->trim_state = BTRFS_TRIM_STATE_UNTRIMMED; + } + + spin_unlock(&ctl->tree_lock); +} + +static void end_trimming_bitmap(struct btrfs_free_space_ctl *ctl, + struct btrfs_free_space *entry) +{ + if (btrfs_free_space_trimming_bitmap(entry)) { + entry->trim_state = BTRFS_TRIM_STATE_TRIMMED; + ctl->discardable_extents[BTRFS_STAT_CURR] -= + entry->bitmap_extents; + ctl->discardable_bytes[BTRFS_STAT_CURR] -= entry->bytes; + } +} + +/* + * If @async is set, then we will trim 1 region and return. + */ static int trim_bitmaps(struct btrfs_block_group *block_group, - u64 *total_trimmed, u64 start, u64 end, u64 minlen) + u64 *total_trimmed, u64 start, u64 end, u64 minlen, + u64 maxlen, bool async) { + struct btrfs_discard_ctl *discard_ctl = + &block_group->fs_info->discard_ctl; struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; struct btrfs_free_space *entry; int ret = 0; int ret2; u64 bytes; u64 offset = offset_to_bitmap(ctl, start); + const u64 max_discard_size = READ_ONCE(discard_ctl->max_discard_size); while (offset < end) { bool next_bitmap = false; @@ -3293,35 +3641,84 @@ static int trim_bitmaps(struct btrfs_block_group *block_group, spin_lock(&ctl->tree_lock); if (ctl->free_space < minlen) { + block_group->discard_cursor = + btrfs_block_group_end(block_group); spin_unlock(&ctl->tree_lock); mutex_unlock(&ctl->cache_writeout_mutex); break; } entry = tree_search_offset(ctl, offset, 1, 0); - if (!entry) { + /* + * Bitmaps are marked trimmed lossily now to prevent constant + * discarding of the same bitmap (the reason why we are bound + * by the filters). So, retrim the block group bitmaps when we + * are preparing to punt to the unused_bgs list. This uses + * @minlen to determine if we are in BTRFS_DISCARD_INDEX_UNUSED + * which is the only discard index which sets minlen to 0. + */ + if (!entry || (async && minlen && start == offset && + btrfs_free_space_trimmed(entry))) { spin_unlock(&ctl->tree_lock); mutex_unlock(&ctl->cache_writeout_mutex); next_bitmap = true; goto next; } + /* + * Async discard bitmap trimming begins at by setting the start + * to be key.objectid and the offset_to_bitmap() aligns to the + * start of the bitmap. This lets us know we are fully + * scanning the bitmap rather than only some portion of it. + */ + if (start == offset) + entry->trim_state = BTRFS_TRIM_STATE_TRIMMING; + bytes = minlen; ret2 = search_bitmap(ctl, entry, &start, &bytes, false); if (ret2 || start >= end) { + /* + * We lossily consider a bitmap trimmed if we only skip + * over regions <= BTRFS_ASYNC_DISCARD_MIN_FILTER. + */ + if (ret2 && minlen <= BTRFS_ASYNC_DISCARD_MIN_FILTER) + end_trimming_bitmap(ctl, entry); + else + entry->trim_state = BTRFS_TRIM_STATE_UNTRIMMED; spin_unlock(&ctl->tree_lock); mutex_unlock(&ctl->cache_writeout_mutex); next_bitmap = true; goto next; } + /* + * We already trimmed a region, but are using the locking above + * to reset the trim_state. + */ + if (async && *total_trimmed) { + spin_unlock(&ctl->tree_lock); + mutex_unlock(&ctl->cache_writeout_mutex); + goto out; + } + bytes = min(bytes, end - start); - if (bytes < minlen) { + if (bytes < minlen || (async && maxlen && bytes > maxlen)) { spin_unlock(&ctl->tree_lock); mutex_unlock(&ctl->cache_writeout_mutex); goto next; } + /* + * Let bytes = BTRFS_MAX_DISCARD_SIZE + X. + * If X < @minlen, we won't trim X when we come back around. + * So trim it now. We differ here from trimming extents as we + * don't keep individual state per bit. + */ + if (async && + max_discard_size && + bytes > (max_discard_size + minlen)) + bytes = max_discard_size; + bitmap_clear_bits(ctl, entry, start, bytes); if (entry->bytes == 0) free_bitmap(ctl, entry); @@ -3333,19 +3730,25 @@ static int trim_bitmaps(struct btrfs_block_group *block_group, mutex_unlock(&ctl->cache_writeout_mutex); ret = do_trimming(block_group, total_trimmed, start, bytes, - start, bytes, &trim_entry); - if (ret) + start, bytes, 0, &trim_entry); + if (ret) { + reset_trimming_bitmap(ctl, offset); + block_group->discard_cursor = + btrfs_block_group_end(block_group); break; + } next: if (next_bitmap) { offset += BITS_PER_BITMAP * ctl->unit; + start = offset; } else { start += bytes; - if (start >= offset + BITS_PER_BITMAP * ctl->unit) - offset += BITS_PER_BITMAP * ctl->unit; } + block_group->discard_cursor = start; if (fatal_signal_pending(current)) { + if (start != offset) + reset_trimming_bitmap(ctl, offset); ret = -ERESTARTSYS; break; } @@ -3353,6 +3756,10 @@ next: cond_resched(); } + if (offset >= end) + block_group->discard_cursor = end; + +out: return ret; } @@ -3399,7 +3806,9 @@ void btrfs_put_block_group_trimming(struct btrfs_block_group *block_group) int btrfs_trim_block_group(struct btrfs_block_group *block_group, u64 *trimmed, u64 start, u64 end, u64 minlen) { + struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; int ret; + u64 rem = 0; *trimmed = 0; @@ -3411,16 +3820,66 @@ int btrfs_trim_block_group(struct btrfs_block_group *block_group, btrfs_get_block_group_trimming(block_group); spin_unlock(&block_group->lock); - ret = trim_no_bitmap(block_group, trimmed, start, end, minlen); + ret = trim_no_bitmap(block_group, trimmed, start, end, minlen, false); if (ret) goto out; - ret = trim_bitmaps(block_group, trimmed, start, end, minlen); + ret = trim_bitmaps(block_group, trimmed, start, end, minlen, 0, false); + div64_u64_rem(end, BITS_PER_BITMAP * ctl->unit, &rem); + /* If we ended in the middle of a bitmap, reset the trimming flag */ + if (rem) + reset_trimming_bitmap(ctl, offset_to_bitmap(ctl, end)); out: btrfs_put_block_group_trimming(block_group); return ret; } +int btrfs_trim_block_group_extents(struct btrfs_block_group *block_group, + u64 *trimmed, u64 start, u64 end, u64 minlen, + bool async) +{ + int ret; + + *trimmed = 0; + + spin_lock(&block_group->lock); + if (block_group->removed) { + spin_unlock(&block_group->lock); + return 0; + } + btrfs_get_block_group_trimming(block_group); + spin_unlock(&block_group->lock); + + ret = trim_no_bitmap(block_group, trimmed, start, end, minlen, async); + btrfs_put_block_group_trimming(block_group); + + return ret; +} + +int btrfs_trim_block_group_bitmaps(struct btrfs_block_group *block_group, + u64 *trimmed, u64 start, u64 end, u64 minlen, + u64 maxlen, bool async) +{ + int ret; + + *trimmed = 0; + + spin_lock(&block_group->lock); + if (block_group->removed) { + spin_unlock(&block_group->lock); + return 0; + } + btrfs_get_block_group_trimming(block_group); + spin_unlock(&block_group->lock); + + ret = trim_bitmaps(block_group, trimmed, start, end, minlen, maxlen, + async); + + btrfs_put_block_group_trimming(block_group); + + return ret; +} + /* * Find the left-most item in the cache tree, and then return the * smallest inode number in the item. @@ -3600,6 +4059,7 @@ int test_add_free_space_entry(struct btrfs_block_group *cache, struct btrfs_free_space_ctl *ctl = cache->free_space_ctl; struct btrfs_free_space *info = NULL, *bitmap_info; void *map = NULL; + enum btrfs_trim_state trim_state = BTRFS_TRIM_STATE_TRIMMED; u64 bytes_added; int ret; @@ -3641,7 +4101,8 @@ again: info = NULL; } - bytes_added = add_bytes_to_bitmap(ctl, bitmap_info, offset, bytes); + bytes_added = add_bytes_to_bitmap(ctl, bitmap_info, offset, bytes, + trim_state); bytes -= bytes_added; offset += bytes_added; diff --git a/fs/btrfs/free-space-cache.h b/fs/btrfs/free-space-cache.h index ba9a23241101..2e0a8077aa74 100644 --- a/fs/btrfs/free-space-cache.h +++ b/fs/btrfs/free-space-cache.h @@ -6,6 +6,20 @@ #ifndef BTRFS_FREE_SPACE_CACHE_H #define BTRFS_FREE_SPACE_CACHE_H +/* + * This is the trim state of an extent or bitmap. + * + * BTRFS_TRIM_STATE_TRIMMING is special and used to maintain the state of a + * bitmap as we may need several trims to fully trim a single bitmap entry. + * This is reset should any free space other than trimmed space be added to the + * bitmap. + */ +enum btrfs_trim_state { + BTRFS_TRIM_STATE_UNTRIMMED, + BTRFS_TRIM_STATE_TRIMMED, + BTRFS_TRIM_STATE_TRIMMING, +}; + struct btrfs_free_space { struct rb_node offset_index; u64 offset; @@ -13,8 +27,21 @@ struct btrfs_free_space { u64 max_extent_size; unsigned long *bitmap; struct list_head list; + enum btrfs_trim_state trim_state; + s32 bitmap_extents; }; +static inline bool btrfs_free_space_trimmed(struct btrfs_free_space *info) +{ + return (info->trim_state == BTRFS_TRIM_STATE_TRIMMED); +} + +static inline bool btrfs_free_space_trimming_bitmap( + struct btrfs_free_space *info) +{ + return (info->trim_state == BTRFS_TRIM_STATE_TRIMMING); +} + struct btrfs_free_space_ctl { spinlock_t tree_lock; struct rb_root free_space_offset; @@ -24,6 +51,8 @@ struct btrfs_free_space_ctl { int total_bitmaps; int unit; u64 start; + s32 discardable_extents[BTRFS_STAT_NR_ENTRIES]; + s64 discardable_bytes[BTRFS_STAT_NR_ENTRIES]; const struct btrfs_free_space_op *op; void *private; struct mutex cache_writeout_mutex; @@ -83,13 +112,17 @@ int btrfs_write_out_ino_cache(struct btrfs_root *root, void btrfs_init_free_space_ctl(struct btrfs_block_group *block_group); int __btrfs_add_free_space(struct btrfs_fs_info *fs_info, struct btrfs_free_space_ctl *ctl, - u64 bytenr, u64 size); + u64 bytenr, u64 size, + enum btrfs_trim_state trim_state); int btrfs_add_free_space(struct btrfs_block_group *block_group, u64 bytenr, u64 size); +int btrfs_add_free_space_async_trimmed(struct btrfs_block_group *block_group, + u64 bytenr, u64 size); int btrfs_remove_free_space(struct btrfs_block_group *block_group, u64 bytenr, u64 size); void __btrfs_remove_free_space_cache(struct btrfs_free_space_ctl *ctl); void btrfs_remove_free_space_cache(struct btrfs_block_group *block_group); +bool btrfs_is_free_space_trimmed(struct btrfs_block_group *block_group); u64 btrfs_find_space_for_alloc(struct btrfs_block_group *block_group, u64 offset, u64 bytes, u64 empty_size, u64 *max_extent_size); @@ -108,6 +141,12 @@ int btrfs_return_cluster_to_free_space( struct btrfs_free_cluster *cluster); int btrfs_trim_block_group(struct btrfs_block_group *block_group, u64 *trimmed, u64 start, u64 end, u64 minlen); +int btrfs_trim_block_group_extents(struct btrfs_block_group *block_group, + u64 *trimmed, u64 start, u64 end, u64 minlen, + bool async); +int btrfs_trim_block_group_bitmaps(struct btrfs_block_group *block_group, + u64 *trimmed, u64 start, u64 end, u64 minlen, + u64 maxlen, bool async); /* Support functions for running our sanity tests */ #ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS diff --git a/fs/btrfs/inode-map.c b/fs/btrfs/inode-map.c index 37345fb6191d..d5c9c69d8263 100644 --- a/fs/btrfs/inode-map.c +++ b/fs/btrfs/inode-map.c @@ -107,7 +107,7 @@ again: if (last != (u64)-1 && last + 1 != key.objectid) { __btrfs_add_free_space(fs_info, ctl, last + 1, - key.objectid - last - 1); + key.objectid - last - 1, 0); wake_up(&root->ino_cache_wait); } @@ -118,7 +118,7 @@ next: if (last < root->highest_objectid - 1) { __btrfs_add_free_space(fs_info, ctl, last + 1, - root->highest_objectid - last - 1); + root->highest_objectid - last - 1, 0); } spin_lock(&root->ino_cache_lock); @@ -175,7 +175,8 @@ static void start_caching(struct btrfs_root *root) ret = btrfs_find_free_objectid(root, &objectid); if (!ret && objectid <= BTRFS_LAST_FREE_OBJECTID) { __btrfs_add_free_space(fs_info, ctl, objectid, - BTRFS_LAST_FREE_OBJECTID - objectid + 1); + BTRFS_LAST_FREE_OBJECTID - objectid + 1, + 0); wake_up(&root->ino_cache_wait); } @@ -221,7 +222,7 @@ void btrfs_return_ino(struct btrfs_root *root, u64 objectid) return; again: if (root->ino_cache_state == BTRFS_CACHE_FINISHED) { - __btrfs_add_free_space(fs_info, pinned, objectid, 1); + __btrfs_add_free_space(fs_info, pinned, objectid, 1, 0); } else { down_write(&fs_info->commit_root_sem); spin_lock(&root->ino_cache_lock); @@ -234,7 +235,7 @@ again: start_caching(root); - __btrfs_add_free_space(fs_info, pinned, objectid, 1); + __btrfs_add_free_space(fs_info, pinned, objectid, 1, 0); up_write(&fs_info->commit_root_sem); } @@ -281,7 +282,7 @@ void btrfs_unpin_free_ino(struct btrfs_root *root) spin_unlock(rbroot_lock); if (count) __btrfs_add_free_space(root->fs_info, ctl, - info->offset, count); + info->offset, count, 0); kmem_cache_free(btrfs_free_space_cachep, info); } } diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 56032c518b26..6d2bb58d277a 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -44,7 +44,6 @@ #include "locking.h" #include "free-space-cache.h" #include "inode-map.h" -#include "backref.h" #include "props.h" #include "qgroup.h" #include "delalloc-space.h" @@ -64,7 +63,6 @@ struct btrfs_dio_data { static const struct inode_operations btrfs_dir_inode_operations; static const struct inode_operations btrfs_symlink_inode_operations; -static const struct inode_operations btrfs_dir_ro_inode_operations; static const struct inode_operations btrfs_special_inode_operations; static const struct inode_operations btrfs_file_inode_operations; static const struct address_space_operations btrfs_aops; @@ -1479,10 +1477,10 @@ next_slot: disk_num_bytes = btrfs_file_extent_disk_num_bytes(leaf, fi); /* - * If extent we got ends before our range starts, skip - * to next extent + * If the extent we got ends before our current offset, + * skip to the next extent. */ - if (extent_end <= start) { + if (extent_end <= cur_offset) { path->slots[0]++; goto next_slot; } @@ -2128,7 +2126,7 @@ static blk_status_t btrfs_submit_bio_hook(struct inode *inode, struct bio *bio, bio_flags); goto out; } else if (!skip_sum) { - ret = btrfs_lookup_bio_sums(inode, bio, NULL); + ret = btrfs_lookup_bio_sums(inode, bio, (u64)-1, NULL); if (ret) goto out; } @@ -2394,649 +2392,6 @@ out: return ret; } -/* snapshot-aware defrag */ -struct sa_defrag_extent_backref { - struct rb_node node; - struct old_sa_defrag_extent *old; - u64 root_id; - u64 inum; - u64 file_pos; - u64 extent_offset; - u64 num_bytes; - u64 generation; -}; - -struct old_sa_defrag_extent { - struct list_head list; - struct new_sa_defrag_extent *new; - - u64 extent_offset; - u64 bytenr; - u64 offset; - u64 len; - int count; -}; - -struct new_sa_defrag_extent { - struct rb_root root; - struct list_head head; - struct btrfs_path *path; - struct inode *inode; - u64 file_pos; - u64 len; - u64 bytenr; - u64 disk_len; - u8 compress_type; -}; - -static int backref_comp(struct sa_defrag_extent_backref *b1, - struct sa_defrag_extent_backref *b2) -{ - if (b1->root_id < b2->root_id) - return -1; - else if (b1->root_id > b2->root_id) - return 1; - - if (b1->inum < b2->inum) - return -1; - else if (b1->inum > b2->inum) - return 1; - - if (b1->file_pos < b2->file_pos) - return -1; - else if (b1->file_pos > b2->file_pos) - return 1; - - /* - * [------------------------------] ===> (a range of space) - * |<--->| |<---->| =============> (fs/file tree A) - * |<---------------------------->| ===> (fs/file tree B) - * - * A range of space can refer to two file extents in one tree while - * refer to only one file extent in another tree. - * - * So we may process a disk offset more than one time(two extents in A) - * and locate at the same extent(one extent in B), then insert two same - * backrefs(both refer to the extent in B). - */ - return 0; -} - -static void backref_insert(struct rb_root *root, - struct sa_defrag_extent_backref *backref) -{ - struct rb_node **p = &root->rb_node; - struct rb_node *parent = NULL; - struct sa_defrag_extent_backref *entry; - int ret; - - while (*p) { - parent = *p; - entry = rb_entry(parent, struct sa_defrag_extent_backref, node); - - ret = backref_comp(backref, entry); - if (ret < 0) - p = &(*p)->rb_left; - else - p = &(*p)->rb_right; - } - - rb_link_node(&backref->node, parent, p); - rb_insert_color(&backref->node, root); -} - -/* - * Note the backref might has changed, and in this case we just return 0. - */ -static noinline int record_one_backref(u64 inum, u64 offset, u64 root_id, - void *ctx) -{ - struct btrfs_file_extent_item *extent; - struct old_sa_defrag_extent *old = ctx; - struct new_sa_defrag_extent *new = old->new; - struct btrfs_path *path = new->path; - struct btrfs_key key; - struct btrfs_root *root; - struct sa_defrag_extent_backref *backref; - struct extent_buffer *leaf; - struct inode *inode = new->inode; - struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb); - int slot; - int ret; - u64 extent_offset; - u64 num_bytes; - - if (BTRFS_I(inode)->root->root_key.objectid == root_id && - inum == btrfs_ino(BTRFS_I(inode))) - return 0; - - key.objectid = root_id; - key.type = BTRFS_ROOT_ITEM_KEY; - key.offset = (u64)-1; - - root = btrfs_read_fs_root_no_name(fs_info, &key); - if (IS_ERR(root)) { - if (PTR_ERR(root) == -ENOENT) - return 0; - WARN_ON(1); - btrfs_debug(fs_info, "inum=%llu, offset=%llu, root_id=%llu", - inum, offset, root_id); - return PTR_ERR(root); - } - - key.objectid = inum; - key.type = BTRFS_EXTENT_DATA_KEY; - if (offset > (u64)-1 << 32) - key.offset = 0; - else - key.offset = offset; - - ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); - if (WARN_ON(ret < 0)) - return ret; - ret = 0; - - while (1) { - cond_resched(); - - leaf = path->nodes[0]; - slot = path->slots[0]; - - if (slot >= btrfs_header_nritems(leaf)) { - ret = btrfs_next_leaf(root, path); - if (ret < 0) { - goto out; - } else if (ret > 0) { - ret = 0; - goto out; - } - continue; - } - - path->slots[0]++; - - btrfs_item_key_to_cpu(leaf, &key, slot); - - if (key.objectid > inum) - goto out; - - if (key.objectid < inum || key.type != BTRFS_EXTENT_DATA_KEY) - continue; - - extent = btrfs_item_ptr(leaf, slot, - struct btrfs_file_extent_item); - - if (btrfs_file_extent_disk_bytenr(leaf, extent) != old->bytenr) - continue; - - /* - * 'offset' refers to the exact key.offset, - * NOT the 'offset' field in btrfs_extent_data_ref, ie. - * (key.offset - extent_offset). - */ - if (key.offset != offset) - continue; - - extent_offset = btrfs_file_extent_offset(leaf, extent); - num_bytes = btrfs_file_extent_num_bytes(leaf, extent); - - if (extent_offset >= old->extent_offset + old->offset + - old->len || extent_offset + num_bytes <= - old->extent_offset + old->offset) - continue; - break; - } - - backref = kmalloc(sizeof(*backref), GFP_NOFS); - if (!backref) { - ret = -ENOENT; - goto out; - } - - backref->root_id = root_id; - backref->inum = inum; - backref->file_pos = offset; - backref->num_bytes = num_bytes; - backref->extent_offset = extent_offset; - backref->generation = btrfs_file_extent_generation(leaf, extent); - backref->old = old; - backref_insert(&new->root, backref); - old->count++; -out: - btrfs_release_path(path); - WARN_ON(ret); - return ret; -} - -static noinline bool record_extent_backrefs(struct btrfs_path *path, - struct new_sa_defrag_extent *new) -{ - struct btrfs_fs_info *fs_info = btrfs_sb(new->inode->i_sb); - struct old_sa_defrag_extent *old, *tmp; - int ret; - - new->path = path; - - list_for_each_entry_safe(old, tmp, &new->head, list) { - ret = iterate_inodes_from_logical(old->bytenr + - old->extent_offset, fs_info, - path, record_one_backref, - old, false); - if (ret < 0 && ret != -ENOENT) - return false; - - /* no backref to be processed for this extent */ - if (!old->count) { - list_del(&old->list); - kfree(old); - } - } - - if (list_empty(&new->head)) - return false; - - return true; -} - -static int relink_is_mergable(struct extent_buffer *leaf, - struct btrfs_file_extent_item *fi, - struct new_sa_defrag_extent *new) -{ - if (btrfs_file_extent_disk_bytenr(leaf, fi) != new->bytenr) - return 0; - - if (btrfs_file_extent_type(leaf, fi) != BTRFS_FILE_EXTENT_REG) - return 0; - - if (btrfs_file_extent_compression(leaf, fi) != new->compress_type) - return 0; - - if (btrfs_file_extent_encryption(leaf, fi) || - btrfs_file_extent_other_encoding(leaf, fi)) - return 0; - - return 1; -} - -/* - * Note the backref might has changed, and in this case we just return 0. - */ -static noinline int relink_extent_backref(struct btrfs_path *path, - struct sa_defrag_extent_backref *prev, - struct sa_defrag_extent_backref *backref) -{ - struct btrfs_file_extent_item *extent; - struct btrfs_file_extent_item *item; - struct btrfs_ordered_extent *ordered; - struct btrfs_trans_handle *trans; - struct btrfs_ref ref = { 0 }; - struct btrfs_root *root; - struct btrfs_key key; - struct extent_buffer *leaf; - struct old_sa_defrag_extent *old = backref->old; - struct new_sa_defrag_extent *new = old->new; - struct btrfs_fs_info *fs_info = btrfs_sb(new->inode->i_sb); - struct inode *inode; - struct extent_state *cached = NULL; - int ret = 0; - u64 start; - u64 len; - u64 lock_start; - u64 lock_end; - bool merge = false; - int index; - - if (prev && prev->root_id == backref->root_id && - prev->inum == backref->inum && - prev->file_pos + prev->num_bytes == backref->file_pos) - merge = true; - - /* step 1: get root */ - key.objectid = backref->root_id; - key.type = BTRFS_ROOT_ITEM_KEY; - key.offset = (u64)-1; - - index = srcu_read_lock(&fs_info->subvol_srcu); - - root = btrfs_read_fs_root_no_name(fs_info, &key); - if (IS_ERR(root)) { - srcu_read_unlock(&fs_info->subvol_srcu, index); - if (PTR_ERR(root) == -ENOENT) - return 0; - return PTR_ERR(root); - } - - if (btrfs_root_readonly(root)) { - srcu_read_unlock(&fs_info->subvol_srcu, index); - return 0; - } - - /* step 2: get inode */ - key.objectid = backref->inum; - key.type = BTRFS_INODE_ITEM_KEY; - key.offset = 0; - - inode = btrfs_iget(fs_info->sb, &key, root); - if (IS_ERR(inode)) { - srcu_read_unlock(&fs_info->subvol_srcu, index); - return 0; - } - - srcu_read_unlock(&fs_info->subvol_srcu, index); - - /* step 3: relink backref */ - lock_start = backref->file_pos; - lock_end = backref->file_pos + backref->num_bytes - 1; - lock_extent_bits(&BTRFS_I(inode)->io_tree, lock_start, lock_end, - &cached); - - ordered = btrfs_lookup_first_ordered_extent(inode, lock_end); - if (ordered) { - btrfs_put_ordered_extent(ordered); - goto out_unlock; - } - - trans = btrfs_join_transaction(root); - if (IS_ERR(trans)) { - ret = PTR_ERR(trans); - goto out_unlock; - } - - key.objectid = backref->inum; - key.type = BTRFS_EXTENT_DATA_KEY; - key.offset = backref->file_pos; - - ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); - if (ret < 0) { - goto out_free_path; - } else if (ret > 0) { - ret = 0; - goto out_free_path; - } - - extent = btrfs_item_ptr(path->nodes[0], path->slots[0], - struct btrfs_file_extent_item); - - if (btrfs_file_extent_generation(path->nodes[0], extent) != - backref->generation) - goto out_free_path; - - btrfs_release_path(path); - - start = backref->file_pos; - if (backref->extent_offset < old->extent_offset + old->offset) - start += old->extent_offset + old->offset - - backref->extent_offset; - - len = min(backref->extent_offset + backref->num_bytes, - old->extent_offset + old->offset + old->len); - len -= max(backref->extent_offset, old->extent_offset + old->offset); - - ret = btrfs_drop_extents(trans, root, inode, start, - start + len, 1); - if (ret) - goto out_free_path; -again: - key.objectid = btrfs_ino(BTRFS_I(inode)); - key.type = BTRFS_EXTENT_DATA_KEY; - key.offset = start; - - path->leave_spinning = 1; - if (merge) { - struct btrfs_file_extent_item *fi; - u64 extent_len; - struct btrfs_key found_key; - - ret = btrfs_search_slot(trans, root, &key, path, 0, 1); - if (ret < 0) - goto out_free_path; - - path->slots[0]--; - leaf = path->nodes[0]; - btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]); - - fi = btrfs_item_ptr(leaf, path->slots[0], - struct btrfs_file_extent_item); - extent_len = btrfs_file_extent_num_bytes(leaf, fi); - - if (extent_len + found_key.offset == start && - relink_is_mergable(leaf, fi, new)) { - btrfs_set_file_extent_num_bytes(leaf, fi, - extent_len + len); - btrfs_mark_buffer_dirty(leaf); - inode_add_bytes(inode, len); - - ret = 1; - goto out_free_path; - } else { - merge = false; - btrfs_release_path(path); - goto again; - } - } - - ret = btrfs_insert_empty_item(trans, root, path, &key, - sizeof(*extent)); - if (ret) { - btrfs_abort_transaction(trans, ret); - goto out_free_path; - } - - leaf = path->nodes[0]; - item = btrfs_item_ptr(leaf, path->slots[0], - struct btrfs_file_extent_item); - btrfs_set_file_extent_disk_bytenr(leaf, item, new->bytenr); - btrfs_set_file_extent_disk_num_bytes(leaf, item, new->disk_len); - btrfs_set_file_extent_offset(leaf, item, start - new->file_pos); - btrfs_set_file_extent_num_bytes(leaf, item, len); - btrfs_set_file_extent_ram_bytes(leaf, item, new->len); - btrfs_set_file_extent_generation(leaf, item, trans->transid); - btrfs_set_file_extent_type(leaf, item, BTRFS_FILE_EXTENT_REG); - btrfs_set_file_extent_compression(leaf, item, new->compress_type); - btrfs_set_file_extent_encryption(leaf, item, 0); - btrfs_set_file_extent_other_encoding(leaf, item, 0); - - btrfs_mark_buffer_dirty(leaf); - inode_add_bytes(inode, len); - btrfs_release_path(path); - - btrfs_init_generic_ref(&ref, BTRFS_ADD_DELAYED_REF, new->bytenr, - new->disk_len, 0); - btrfs_init_data_ref(&ref, backref->root_id, backref->inum, - new->file_pos); /* start - extent_offset */ - ret = btrfs_inc_extent_ref(trans, &ref); - if (ret) { - btrfs_abort_transaction(trans, ret); - goto out_free_path; - } - - ret = 1; -out_free_path: - btrfs_release_path(path); - path->leave_spinning = 0; - btrfs_end_transaction(trans); -out_unlock: - unlock_extent_cached(&BTRFS_I(inode)->io_tree, lock_start, lock_end, - &cached); - iput(inode); - return ret; -} - -static void free_sa_defrag_extent(struct new_sa_defrag_extent *new) -{ - struct old_sa_defrag_extent *old, *tmp; - - if (!new) - return; - - list_for_each_entry_safe(old, tmp, &new->head, list) { - kfree(old); - } - kfree(new); -} - -static void relink_file_extents(struct new_sa_defrag_extent *new) -{ - struct btrfs_fs_info *fs_info = btrfs_sb(new->inode->i_sb); - struct btrfs_path *path; - struct sa_defrag_extent_backref *backref; - struct sa_defrag_extent_backref *prev = NULL; - struct rb_node *node; - int ret; - - path = btrfs_alloc_path(); - if (!path) - return; - - if (!record_extent_backrefs(path, new)) { - btrfs_free_path(path); - goto out; - } - btrfs_release_path(path); - - while (1) { - node = rb_first(&new->root); - if (!node) - break; - rb_erase(node, &new->root); - - backref = rb_entry(node, struct sa_defrag_extent_backref, node); - - ret = relink_extent_backref(path, prev, backref); - WARN_ON(ret < 0); - - kfree(prev); - - if (ret == 1) - prev = backref; - else - prev = NULL; - cond_resched(); - } - kfree(prev); - - btrfs_free_path(path); -out: - free_sa_defrag_extent(new); - - atomic_dec(&fs_info->defrag_running); - wake_up(&fs_info->transaction_wait); -} - -static struct new_sa_defrag_extent * -record_old_file_extents(struct inode *inode, - struct btrfs_ordered_extent *ordered) -{ - struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb); - struct btrfs_root *root = BTRFS_I(inode)->root; - struct btrfs_path *path; - struct btrfs_key key; - struct old_sa_defrag_extent *old; - struct new_sa_defrag_extent *new; - int ret; - - new = kmalloc(sizeof(*new), GFP_NOFS); - if (!new) - return NULL; - - new->inode = inode; - new->file_pos = ordered->file_offset; - new->len = ordered->len; - new->bytenr = ordered->start; - new->disk_len = ordered->disk_len; - new->compress_type = ordered->compress_type; - new->root = RB_ROOT; - INIT_LIST_HEAD(&new->head); - - path = btrfs_alloc_path(); - if (!path) - goto out_kfree; - - key.objectid = btrfs_ino(BTRFS_I(inode)); - key.type = BTRFS_EXTENT_DATA_KEY; - key.offset = new->file_pos; - - ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); - if (ret < 0) - goto out_free_path; - if (ret > 0 && path->slots[0] > 0) - path->slots[0]--; - - /* find out all the old extents for the file range */ - while (1) { - struct btrfs_file_extent_item *extent; - struct extent_buffer *l; - int slot; - u64 num_bytes; - u64 offset; - u64 end; - u64 disk_bytenr; - u64 extent_offset; - - l = path->nodes[0]; - slot = path->slots[0]; - - if (slot >= btrfs_header_nritems(l)) { - ret = btrfs_next_leaf(root, path); - if (ret < 0) - goto out_free_path; - else if (ret > 0) - break; - continue; - } - - btrfs_item_key_to_cpu(l, &key, slot); - - if (key.objectid != btrfs_ino(BTRFS_I(inode))) - break; - if (key.type != BTRFS_EXTENT_DATA_KEY) - break; - if (key.offset >= new->file_pos + new->len) - break; - - extent = btrfs_item_ptr(l, slot, struct btrfs_file_extent_item); - - num_bytes = btrfs_file_extent_num_bytes(l, extent); - if (key.offset + num_bytes < new->file_pos) - goto next; - - disk_bytenr = btrfs_file_extent_disk_bytenr(l, extent); - if (!disk_bytenr) - goto next; - - extent_offset = btrfs_file_extent_offset(l, extent); - - old = kmalloc(sizeof(*old), GFP_NOFS); - if (!old) - goto out_free_path; - - offset = max(new->file_pos, key.offset); - end = min(new->file_pos + new->len, key.offset + num_bytes); - - old->bytenr = disk_bytenr; - old->extent_offset = extent_offset; - old->offset = offset - key.offset; - old->len = end - offset; - old->new = new; - old->count = 0; - list_add_tail(&old->list, &new->head); -next: - path->slots[0]++; - cond_resched(); - } - - btrfs_free_path(path); - atomic_inc(&fs_info->defrag_running); - - return new; - -out_free_path: - btrfs_free_path(path); -out_kfree: - free_sa_defrag_extent(new); - return NULL; -} - static void btrfs_release_delalloc_bytes(struct btrfs_fs_info *fs_info, u64 start, u64 len) { @@ -3064,15 +2419,19 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent) struct btrfs_trans_handle *trans = NULL; struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree; struct extent_state *cached_state = NULL; - struct new_sa_defrag_extent *new = NULL; + u64 start, end; int compress_type = 0; int ret = 0; - u64 logical_len = ordered_extent->len; + u64 logical_len = ordered_extent->num_bytes; bool freespace_inode; bool truncated = false; bool range_locked = false; bool clear_new_delalloc_bytes = false; bool clear_reserved_extent = true; + unsigned int clear_bits; + + start = ordered_extent->file_offset; + end = start + ordered_extent->num_bytes - 1; if (!test_bit(BTRFS_ORDERED_NOCOW, &ordered_extent->flags) && !test_bit(BTRFS_ORDERED_PREALLOC, &ordered_extent->flags) && @@ -3086,10 +2445,7 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent) goto out; } - btrfs_free_io_failure_record(BTRFS_I(inode), - ordered_extent->file_offset, - ordered_extent->file_offset + - ordered_extent->len - 1); + btrfs_free_io_failure_record(BTRFS_I(inode), start, end); if (test_bit(BTRFS_ORDERED_TRUNCATED, &ordered_extent->flags)) { truncated = true; @@ -3107,8 +2463,8 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent) * space for NOCOW range. * As NOCOW won't cause a new delayed ref, just free the space */ - btrfs_qgroup_free_data(inode, NULL, ordered_extent->file_offset, - ordered_extent->len); + btrfs_qgroup_free_data(inode, NULL, start, + ordered_extent->num_bytes); btrfs_ordered_update_i_size(inode, 0, ordered_extent); if (freespace_inode) trans = btrfs_join_transaction_spacecache(root); @@ -3127,23 +2483,7 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent) } range_locked = true; - lock_extent_bits(io_tree, ordered_extent->file_offset, - ordered_extent->file_offset + ordered_extent->len - 1, - &cached_state); - - ret = test_range_bit(io_tree, ordered_extent->file_offset, - ordered_extent->file_offset + ordered_extent->len - 1, - EXTENT_DEFRAG, 0, cached_state); - if (ret) { - u64 last_snapshot = btrfs_root_last_snapshot(&root->root_item); - if (0 && last_snapshot >= BTRFS_I(inode)->generation) - /* the inode is shared */ - new = record_old_file_extents(inode, ordered_extent); - - clear_extent_bit(io_tree, ordered_extent->file_offset, - ordered_extent->file_offset + ordered_extent->len - 1, - EXTENT_DEFRAG, 0, 0, &cached_state); - } + lock_extent_bits(io_tree, start, end, &cached_state); if (freespace_inode) trans = btrfs_join_transaction_spacecache(root); @@ -3161,31 +2501,30 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent) compress_type = ordered_extent->compress_type; if (test_bit(BTRFS_ORDERED_PREALLOC, &ordered_extent->flags)) { BUG_ON(compress_type); - btrfs_qgroup_free_data(inode, NULL, ordered_extent->file_offset, - ordered_extent->len); + btrfs_qgroup_free_data(inode, NULL, start, + ordered_extent->num_bytes); ret = btrfs_mark_extent_written(trans, BTRFS_I(inode), ordered_extent->file_offset, ordered_extent->file_offset + logical_len); } else { BUG_ON(root == fs_info->tree_root); - ret = insert_reserved_file_extent(trans, inode, - ordered_extent->file_offset, - ordered_extent->start, - ordered_extent->disk_len, + ret = insert_reserved_file_extent(trans, inode, start, + ordered_extent->disk_bytenr, + ordered_extent->disk_num_bytes, logical_len, logical_len, compress_type, 0, 0, BTRFS_FILE_EXTENT_REG); if (!ret) { clear_reserved_extent = false; btrfs_release_delalloc_bytes(fs_info, - ordered_extent->start, - ordered_extent->disk_len); + ordered_extent->disk_bytenr, + ordered_extent->disk_num_bytes); } } unpin_extent_cache(&BTRFS_I(inode)->extent_tree, - ordered_extent->file_offset, ordered_extent->len, - trans->transid); + ordered_extent->file_offset, + ordered_extent->num_bytes, trans->transid); if (ret < 0) { btrfs_abort_transaction(trans, ret); goto out; @@ -3205,37 +2544,27 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent) } ret = 0; out: - if (range_locked || clear_new_delalloc_bytes) { - unsigned int clear_bits = 0; - - if (range_locked) - clear_bits |= EXTENT_LOCKED; - if (clear_new_delalloc_bytes) - clear_bits |= EXTENT_DELALLOC_NEW; - clear_extent_bit(&BTRFS_I(inode)->io_tree, - ordered_extent->file_offset, - ordered_extent->file_offset + - ordered_extent->len - 1, - clear_bits, - (clear_bits & EXTENT_LOCKED) ? 1 : 0, - 0, &cached_state); - } + clear_bits = EXTENT_DEFRAG; + if (range_locked) + clear_bits |= EXTENT_LOCKED; + if (clear_new_delalloc_bytes) + clear_bits |= EXTENT_DELALLOC_NEW; + clear_extent_bit(&BTRFS_I(inode)->io_tree, start, end, clear_bits, + (clear_bits & EXTENT_LOCKED) ? 1 : 0, 0, + &cached_state); if (trans) btrfs_end_transaction(trans); if (ret || truncated) { - u64 start, end; + u64 unwritten_start = start; if (truncated) - start = ordered_extent->file_offset + logical_len; - else - start = ordered_extent->file_offset; - end = ordered_extent->file_offset + ordered_extent->len - 1; - clear_extent_uptodate(io_tree, start, end, NULL); + unwritten_start += logical_len; + clear_extent_uptodate(io_tree, unwritten_start, end, NULL); /* Drop the cache for the part of the extent we didn't write. */ - btrfs_drop_extent_cache(BTRFS_I(inode), start, end, 0); + btrfs_drop_extent_cache(BTRFS_I(inode), unwritten_start, end, 0); /* * If the ordered extent had an IOERR or something else went @@ -3250,29 +2579,28 @@ out: if ((ret || !logical_len) && clear_reserved_extent && !test_bit(BTRFS_ORDERED_NOCOW, &ordered_extent->flags) && - !test_bit(BTRFS_ORDERED_PREALLOC, &ordered_extent->flags)) + !test_bit(BTRFS_ORDERED_PREALLOC, &ordered_extent->flags)) { + /* + * Discard the range before returning it back to the + * free space pool + */ + if (ret && btrfs_test_opt(fs_info, DISCARD_SYNC)) + btrfs_discard_extent(fs_info, + ordered_extent->disk_bytenr, + ordered_extent->disk_num_bytes, + NULL); btrfs_free_reserved_extent(fs_info, - ordered_extent->start, - ordered_extent->disk_len, 1); + ordered_extent->disk_bytenr, + ordered_extent->disk_num_bytes, 1); + } } - /* * This needs to be done to make sure anybody waiting knows we are done * updating everything for this ordered extent. */ btrfs_remove_ordered_extent(inode, ordered_extent); - /* for snapshot-aware defrag */ - if (new) { - if (ret) { - free_sa_defrag_extent(new); - atomic_dec(&fs_info->defrag_running); - } else { - relink_file_extents(new); - } - } - /* once for us */ btrfs_put_ordered_extent(ordered_extent); /* once for the tree */ @@ -4238,18 +3566,30 @@ out: } static int btrfs_unlink_subvol(struct btrfs_trans_handle *trans, - struct inode *dir, u64 objectid, - const char *name, int name_len) + struct inode *dir, struct dentry *dentry) { struct btrfs_root *root = BTRFS_I(dir)->root; + struct btrfs_inode *inode = BTRFS_I(d_inode(dentry)); struct btrfs_path *path; struct extent_buffer *leaf; struct btrfs_dir_item *di; struct btrfs_key key; + const char *name = dentry->d_name.name; + int name_len = dentry->d_name.len; u64 index; int ret; + u64 objectid; u64 dir_ino = btrfs_ino(BTRFS_I(dir)); + if (btrfs_ino(inode) == BTRFS_FIRST_FREE_OBJECTID) { + objectid = inode->root->root_key.objectid; + } else if (btrfs_ino(inode) == BTRFS_EMPTY_SUBVOL_DIR_OBJECTID) { + objectid = inode->location.objectid; + } else { + WARN_ON(1); + return -EINVAL; + } + path = btrfs_alloc_path(); if (!path) return -ENOMEM; @@ -4271,13 +3611,16 @@ static int btrfs_unlink_subvol(struct btrfs_trans_handle *trans, } btrfs_release_path(path); - ret = btrfs_del_root_ref(trans, objectid, root->root_key.objectid, - dir_ino, &index, name, name_len); - if (ret < 0) { - if (ret != -ENOENT) { - btrfs_abort_transaction(trans, ret); - goto out; - } + /* + * This is a placeholder inode for a subvolume we didn't have a + * reference to at the time of the snapshot creation. In the meantime + * we could have renamed the real subvol link into our snapshot, so + * depending on btrfs_del_root_ref to return -ENOENT here is incorret. + * Instead simply lookup the dir_index_item for this entry so we can + * remove it. Otherwise we know we have a ref to the root and we can + * call btrfs_del_root_ref, and it _shouldn't_ fail. + */ + if (btrfs_ino(inode) == BTRFS_EMPTY_SUBVOL_DIR_OBJECTID) { di = btrfs_search_dir_index_item(root, path, dir_ino, name, name_len); if (IS_ERR_OR_NULL(di)) { @@ -4292,8 +3635,16 @@ static int btrfs_unlink_subvol(struct btrfs_trans_handle *trans, leaf = path->nodes[0]; btrfs_item_key_to_cpu(leaf, &key, path->slots[0]); index = key.offset; + btrfs_release_path(path); + } else { + ret = btrfs_del_root_ref(trans, objectid, + root->root_key.objectid, dir_ino, + &index, name, name_len); + if (ret) { + btrfs_abort_transaction(trans, ret); + goto out; + } } - btrfs_release_path(path); ret = btrfs_delete_delayed_dir_index(trans, BTRFS_I(dir), index); if (ret) { @@ -4487,8 +3838,7 @@ int btrfs_delete_subvolume(struct inode *dir, struct dentry *dentry) btrfs_record_snapshot_destroy(trans, BTRFS_I(dir)); - ret = btrfs_unlink_subvol(trans, dir, dest->root_key.objectid, - dentry->d_name.name, dentry->d_name.len); + ret = btrfs_unlink_subvol(trans, dir, dentry); if (ret) { err = ret; btrfs_abort_transaction(trans, ret); @@ -4583,10 +3933,7 @@ static int btrfs_rmdir(struct inode *dir, struct dentry *dentry) return PTR_ERR(trans); if (unlikely(btrfs_ino(BTRFS_I(inode)) == BTRFS_EMPTY_SUBVOL_DIR_OBJECTID)) { - err = btrfs_unlink_subvol(trans, dir, - BTRFS_I(inode)->location.objectid, - dentry->d_name.name, - dentry->d_name.len); + err = btrfs_unlink_subvol(trans, dir, dentry); goto out; } @@ -5157,7 +4504,7 @@ int btrfs_cont_expand(struct inode *inode, loff_t oldsize, loff_t size) cur_offset = hole_start; while (1) { em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, cur_offset, - block_end - cur_offset, 0); + block_end - cur_offset); if (IS_ERR(em)) { err = PTR_ERR(em); em = NULL; @@ -5728,7 +5075,6 @@ static void inode_tree_add(struct inode *inode) static void inode_tree_del(struct inode *inode) { - struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb); struct btrfs_root *root = BTRFS_I(inode)->root; int empty = 0; @@ -5741,7 +5087,6 @@ static void inode_tree_del(struct inode *inode) spin_unlock(&root->inode_lock); if (empty && btrfs_root_refs(&root->root_item) == 0) { - synchronize_srcu(&fs_info->subvol_srcu); spin_lock(&root->inode_lock); empty = RB_EMPTY_ROOT(&root->inode_tree); spin_unlock(&root->inode_lock); @@ -5843,7 +5188,11 @@ static struct inode *new_simple_dir(struct super_block *s, set_bit(BTRFS_INODE_DUMMY, &BTRFS_I(inode)->runtime_flags); inode->i_ino = BTRFS_EMPTY_SUBVOL_DIR_OBJECTID; - inode->i_op = &btrfs_dir_ro_inode_operations; + /* + * We only need lookup, the rest is read-only and there's no inode + * associated with the dentry + */ + inode->i_op = &simple_dir_inode_operations; inode->i_opflags &= ~IOP_XATTR; inode->i_fop = &simple_dir_operations; inode->i_mode = S_IFDIR | S_IRUGO | S_IWUSR | S_IXUGO; @@ -6934,18 +6283,27 @@ static noinline int uncompress_inline(struct btrfs_path *path, return ret; } -/* - * a bit scary, this does extent mapping from logical file offset to the disk. - * the ugly parts come from merging extents from the disk with the in-ram - * representation. This gets more complex because of the data=ordered code, - * where the in-ram extents might be locked pending data=ordered completion. +/** + * btrfs_get_extent - Lookup the first extent overlapping a range in a file. + * @inode: file to search in + * @page: page to read extent data into if the extent is inline + * @pg_offset: offset into @page to copy to + * @start: file offset + * @len: length of range starting at @start * - * This also copies inline extents directly into the page. + * This returns the first &struct extent_map which overlaps with the given + * range, reading it from the B-tree and caching it if necessary. Note that + * there may be more extents which overlap the given range after the returned + * extent_map. + * + * If @page is not NULL and the extent is inline, this also reads the extent + * data directly into the page and marks the extent up to date in the io_tree. + * + * Return: ERR_PTR on error, non-NULL extent_map on success. */ struct extent_map *btrfs_get_extent(struct btrfs_inode *inode, - struct page *page, - size_t pg_offset, u64 start, u64 len, - int create) + struct page *page, size_t pg_offset, + u64 start, u64 len) { struct btrfs_fs_info *fs_info = inode->root->fs_info; int ret; @@ -6962,7 +6320,6 @@ struct extent_map *btrfs_get_extent(struct btrfs_inode *inode, struct extent_map *em = NULL; struct extent_map_tree *em_tree = &inode->extent_tree; struct extent_io_tree *io_tree = &inode->io_tree; - const bool new_inline = !page || create; read_lock(&em_tree->lock); em = lookup_extent_mapping(em_tree, start, len); @@ -7085,8 +6442,7 @@ next: goto insert; } - btrfs_extent_item_to_extent_map(inode, path, item, - new_inline, em); + btrfs_extent_item_to_extent_map(inode, path, item, !page, em); if (extent_type == BTRFS_FILE_EXTENT_REG || extent_type == BTRFS_FILE_EXTENT_PREALLOC) { @@ -7098,7 +6454,7 @@ next: size_t extent_offset; size_t copy_size; - if (new_inline) + if (!page) goto out; size = btrfs_file_extent_ram_bytes(leaf, item); @@ -7181,7 +6537,7 @@ struct extent_map *btrfs_get_extent_fiemap(struct btrfs_inode *inode, u64 delalloc_end; int err = 0; - em = btrfs_get_extent(inode, NULL, 0, start, len, 0); + em = btrfs_get_extent(inode, NULL, 0, start, len); if (IS_ERR(em)) return em; /* @@ -7806,7 +7162,7 @@ static int btrfs_get_blocks_direct(struct inode *inode, sector_t iblock, goto err; } - em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, start, len, 0); + em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, start, len); if (IS_ERR(em)) { ret = PTR_ERR(em); goto unlock_err; @@ -8358,8 +7714,8 @@ static inline blk_status_t btrfs_lookup_and_bind_dio_csum(struct inode *inode, * contention. */ if (dip->logical_offset == file_offset) { - ret = btrfs_lookup_bio_sums_dio(inode, dip->orig_bio, - file_offset); + ret = btrfs_lookup_bio_sums(inode, dip->orig_bio, file_offset, + NULL); if (ret) return ret; } @@ -8872,7 +8228,8 @@ again: ordered = btrfs_lookup_ordered_range(BTRFS_I(inode), start, page_end - start + 1); if (ordered) { - end = min(page_end, ordered->file_offset + ordered->len - 1); + end = min(page_end, + ordered->file_offset + ordered->num_bytes - 1); /* * IO on this page will never be started, so we need * to account for any ordered extents now @@ -9073,7 +8430,6 @@ again: ret = VM_FAULT_SIGBUS; goto out_unlock; } - ret2 = 0; /* page is wholly or partially inside EOF */ if (page_start + PAGE_SIZE > size) @@ -9097,12 +8453,10 @@ again: unlock_extent_cached(io_tree, page_start, page_end, &cached_state); - if (!ret2) { - btrfs_delalloc_release_extents(BTRFS_I(inode), PAGE_SIZE); - sb_end_pagefault(inode->i_sb); - extent_changeset_free(data_reserved); - return VM_FAULT_LOCKED; - } + btrfs_delalloc_release_extents(BTRFS_I(inode), PAGE_SIZE); + sb_end_pagefault(inode->i_sb); + extent_changeset_free(data_reserved); + return VM_FAULT_LOCKED; out_unlock: unlock_page(page); @@ -9400,7 +8754,7 @@ void btrfs_destroy_inode(struct inode *inode) else { btrfs_err(fs_info, "found ordered extent %llu %llu on inode cleanup", - ordered->file_offset, ordered->len); + ordered->file_offset, ordered->num_bytes); btrfs_remove_ordered_extent(inode, ordered); btrfs_put_ordered_extent(ordered); btrfs_put_ordered_extent(ordered); @@ -9538,7 +8892,6 @@ static int btrfs_rename_exchange(struct inode *old_dir, u64 new_ino = btrfs_ino(BTRFS_I(new_inode)); u64 old_idx = 0; u64 new_idx = 0; - u64 root_objectid; int ret; bool root_log_pinned = false; bool dest_log_pinned = false; @@ -9556,9 +8909,8 @@ static int btrfs_rename_exchange(struct inode *old_dir, btrfs_init_log_ctx(&ctx_dest, new_inode); /* close the race window with snapshot create/destroy ioctl */ - if (old_ino == BTRFS_FIRST_FREE_OBJECTID) - down_read(&fs_info->subvol_sem); - if (new_ino == BTRFS_FIRST_FREE_OBJECTID) + if (old_ino == BTRFS_FIRST_FREE_OBJECTID || + new_ino == BTRFS_FIRST_FREE_OBJECTID) down_read(&fs_info->subvol_sem); /* @@ -9645,10 +8997,7 @@ static int btrfs_rename_exchange(struct inode *old_dir, /* src is a subvolume */ if (old_ino == BTRFS_FIRST_FREE_OBJECTID) { - root_objectid = BTRFS_I(old_inode)->root->root_key.objectid; - ret = btrfs_unlink_subvol(trans, old_dir, root_objectid, - old_dentry->d_name.name, - old_dentry->d_name.len); + ret = btrfs_unlink_subvol(trans, old_dir, old_dentry); } else { /* src is an inode */ ret = __btrfs_unlink_inode(trans, root, BTRFS_I(old_dir), BTRFS_I(old_dentry->d_inode), @@ -9664,10 +9013,7 @@ static int btrfs_rename_exchange(struct inode *old_dir, /* dest is a subvolume */ if (new_ino == BTRFS_FIRST_FREE_OBJECTID) { - root_objectid = BTRFS_I(new_inode)->root->root_key.objectid; - ret = btrfs_unlink_subvol(trans, new_dir, root_objectid, - new_dentry->d_name.name, - new_dentry->d_name.len); + ret = btrfs_unlink_subvol(trans, new_dir, new_dentry); } else { /* dest is an inode */ ret = __btrfs_unlink_inode(trans, dest, BTRFS_I(new_dir), BTRFS_I(new_dentry->d_inode), @@ -9792,9 +9138,8 @@ out_fail: ret = ret ? ret : ret2; } out_notrans: - if (new_ino == BTRFS_FIRST_FREE_OBJECTID) - up_read(&fs_info->subvol_sem); - if (old_ino == BTRFS_FIRST_FREE_OBJECTID) + if (new_ino == BTRFS_FIRST_FREE_OBJECTID || + old_ino == BTRFS_FIRST_FREE_OBJECTID) up_read(&fs_info->subvol_sem); ASSERT(list_empty(&ctx_root.list)); @@ -9866,7 +9211,6 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry, struct inode *new_inode = d_inode(new_dentry); struct inode *old_inode = d_inode(old_dentry); u64 index = 0; - u64 root_objectid; int ret; u64 old_ino = btrfs_ino(BTRFS_I(old_inode)); bool log_pinned = false; @@ -9974,10 +9318,7 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry, BTRFS_I(old_inode), 1); if (unlikely(old_ino == BTRFS_FIRST_FREE_OBJECTID)) { - root_objectid = BTRFS_I(old_inode)->root->root_key.objectid; - ret = btrfs_unlink_subvol(trans, old_dir, root_objectid, - old_dentry->d_name.name, - old_dentry->d_name.len); + ret = btrfs_unlink_subvol(trans, old_dir, old_dentry); } else { ret = __btrfs_unlink_inode(trans, root, BTRFS_I(old_dir), BTRFS_I(d_inode(old_dentry)), @@ -9996,10 +9337,7 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry, new_inode->i_ctime = current_time(new_inode); if (unlikely(btrfs_ino(BTRFS_I(new_inode)) == BTRFS_EMPTY_SUBVOL_DIR_OBJECTID)) { - root_objectid = BTRFS_I(new_inode)->location.objectid; - ret = btrfs_unlink_subvol(trans, new_dir, root_objectid, - new_dentry->d_name.name, - new_dentry->d_name.len); + ret = btrfs_unlink_subvol(trans, new_dir, new_dentry); BUG_ON(new_inode->i_nlink == 0); } else { ret = btrfs_unlink_inode(trans, dest, BTRFS_I(new_dir), @@ -10835,7 +10173,7 @@ static int btrfs_swap_activate(struct swap_info_struct *sis, struct file *file, struct btrfs_block_group *bg; u64 len = isize - start; - em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, start, len, 0); + em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, start, len); if (IS_ERR(em)) { ret = PTR_ERR(em); goto out; @@ -11003,11 +10341,6 @@ static const struct inode_operations btrfs_dir_inode_operations = { .update_time = btrfs_update_time, .tmpfile = btrfs_tmpfile, }; -static const struct inode_operations btrfs_dir_ro_inode_operations = { - .lookup = btrfs_lookup, - .permission = btrfs_permission, - .update_time = btrfs_update_time, -}; static const struct file_operations btrfs_dir_file_operations = { .llseek = generic_file_llseek, diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index a1ee0b775e65..4f4b13830b25 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -704,11 +704,17 @@ static noinline int create_subvol(struct inode *dir, btrfs_i_size_write(BTRFS_I(dir), dir->i_size + namelen * 2); ret = btrfs_update_inode(trans, root, dir); - BUG_ON(ret); + if (ret) { + btrfs_abort_transaction(trans, ret); + goto fail; + } ret = btrfs_add_root_ref(trans, objectid, root->root_key.objectid, btrfs_ino(BTRFS_I(dir)), index, name, namelen); - BUG_ON(ret); + if (ret) { + btrfs_abort_transaction(trans, ret); + goto fail; + } ret = btrfs_uuid_tree_add(trans, root_item->uuid, BTRFS_UUID_KEY_SUBVOL, objectid); @@ -1122,7 +1128,7 @@ static struct extent_map *defrag_lookup_extent(struct inode *inode, u64 start) /* get the big lock and read metadata off disk */ lock_extent_bits(io_tree, start, end, &cached); - em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, start, len, 0); + em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, start, len); unlock_extent_cached(io_tree, start, end, &cached); if (IS_ERR(em)) @@ -3237,6 +3243,7 @@ static void btrfs_double_extent_lock(struct inode *inode1, u64 loff1, static int btrfs_extent_same_range(struct inode *src, u64 loff, u64 len, struct inode *dst, u64 dst_loff) { + const u64 bs = BTRFS_I(src)->root->fs_info->sb->s_blocksize; int ret; /* @@ -3244,7 +3251,7 @@ static int btrfs_extent_same_range(struct inode *src, u64 loff, u64 len, * source range to serialize with relocation. */ btrfs_double_extent_lock(src, loff, dst, dst_loff, len); - ret = btrfs_clone(src, dst, loff, len, len, dst_loff, 1); + ret = btrfs_clone(src, dst, loff, len, ALIGN(len, bs), dst_loff, 1); btrfs_double_extent_unlock(src, loff, dst, dst_loff, len); return ret; @@ -3720,24 +3727,18 @@ process_slot: ret = 0; if (last_dest_end < destoff + len) { - struct btrfs_clone_extent_info clone_info = { 0 }; /* - * We have an implicit hole (NO_HOLES feature is enabled) that - * fully or partially overlaps our cloning range at its end. + * We have an implicit hole that fully or partially overlaps our + * cloning range at its end. This means that we either have the + * NO_HOLES feature enabled or the implicit hole happened due to + * mixing buffered and direct IO writes against this file. */ btrfs_release_path(path); path->leave_spinning = 0; - /* - * We are dealing with a hole and our clone_info already has a - * disk_offset of 0, we only need to fill the data length and - * file offset. - */ - clone_info.data_len = destoff + len - last_dest_end; - clone_info.file_offset = last_dest_end; ret = btrfs_punch_hole_range(inode, path, last_dest_end, destoff + len - 1, - &clone_info, &trans); + NULL, &trans); if (ret) goto out; @@ -4252,7 +4253,19 @@ static long btrfs_ioctl_scrub(struct file *file, void __user *arg) &sa->progress, sa->flags & BTRFS_SCRUB_READONLY, 0); - if (ret == 0 && copy_to_user(arg, sa, sizeof(*sa))) + /* + * Copy scrub args to user space even if btrfs_scrub_dev() returned an + * error. This is important as it allows user space to know how much + * progress scrub has done. For example, if scrub is canceled we get + * -ECANCELED from btrfs_scrub_dev() and return that error back to user + * space. Later user space can inspect the progress from the structure + * btrfs_ioctl_scrub_args and resume scrub from where it left off + * previously (btrfs-progs does this). + * If we fail to copy the btrfs_ioctl_scrub_args structure to user space + * then return -EFAULT to signal the structure was not copied or it may + * be corrupt and unreliable due to a partial copy. + */ + if (copy_to_user(arg, sa, sizeof(*sa))) ret = -EFAULT; if (!(sa->flags & BTRFS_SCRUB_READONLY)) diff --git a/fs/btrfs/ordered-data.c b/fs/btrfs/ordered-data.c index fb09bc2f8e4d..ecb9fb6a6fe0 100644 --- a/fs/btrfs/ordered-data.c +++ b/fs/btrfs/ordered-data.c @@ -20,9 +20,9 @@ static struct kmem_cache *btrfs_ordered_extent_cache; static u64 entry_end(struct btrfs_ordered_extent *entry) { - if (entry->file_offset + entry->len < entry->file_offset) + if (entry->file_offset + entry->num_bytes < entry->file_offset) return (u64)-1; - return entry->file_offset + entry->len; + return entry->file_offset + entry->num_bytes; } /* returns NULL if the insertion worked, or it returns the node it did find @@ -52,14 +52,6 @@ static struct rb_node *tree_insert(struct rb_root *root, u64 file_offset, return NULL; } -static void ordered_data_tree_panic(struct inode *inode, int errno, - u64 offset) -{ - struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb); - btrfs_panic(fs_info, errno, - "Inconsistency in ordered tree at offset %llu", offset); -} - /* * look for a given offset in the tree, and if it can't be found return the * first lesser offset @@ -120,7 +112,7 @@ static struct rb_node *__tree_search(struct rb_root *root, u64 file_offset, static int offset_in_entry(struct btrfs_ordered_extent *entry, u64 file_offset) { if (file_offset < entry->file_offset || - entry->file_offset + entry->len <= file_offset) + entry->file_offset + entry->num_bytes <= file_offset) return 0; return 1; } @@ -129,7 +121,7 @@ static int range_overlaps(struct btrfs_ordered_extent *entry, u64 file_offset, u64 len) { if (file_offset + len <= entry->file_offset || - entry->file_offset + entry->len <= file_offset) + entry->file_offset + entry->num_bytes <= file_offset) return 0; return 1; } @@ -161,19 +153,14 @@ static inline struct rb_node *tree_search(struct btrfs_ordered_inode_tree *tree, } /* allocate and add a new ordered_extent into the per-inode tree. - * file_offset is the logical offset in the file - * - * start is the disk block number of an extent already reserved in the - * extent allocation tree - * - * len is the length of the extent * * The tree is given a single reference on the ordered extent that was * inserted. */ static int __btrfs_add_ordered_extent(struct inode *inode, u64 file_offset, - u64 start, u64 len, u64 disk_len, - int type, int dio, int compress_type) + u64 disk_bytenr, u64 num_bytes, + u64 disk_num_bytes, int type, int dio, + int compress_type) { struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb); struct btrfs_root *root = BTRFS_I(inode)->root; @@ -187,10 +174,10 @@ static int __btrfs_add_ordered_extent(struct inode *inode, u64 file_offset, return -ENOMEM; entry->file_offset = file_offset; - entry->start = start; - entry->len = len; - entry->disk_len = disk_len; - entry->bytes_left = len; + entry->disk_bytenr = disk_bytenr; + entry->num_bytes = num_bytes; + entry->disk_num_bytes = disk_num_bytes; + entry->bytes_left = num_bytes; entry->inode = igrab(inode); entry->compress_type = compress_type; entry->truncated_len = (u64)-1; @@ -198,7 +185,7 @@ static int __btrfs_add_ordered_extent(struct inode *inode, u64 file_offset, set_bit(type, &entry->flags); if (dio) { - percpu_counter_add_batch(&fs_info->dio_bytes, len, + percpu_counter_add_batch(&fs_info->dio_bytes, num_bytes, fs_info->delalloc_batch); set_bit(BTRFS_ORDERED_DIRECT, &entry->flags); } @@ -219,7 +206,9 @@ static int __btrfs_add_ordered_extent(struct inode *inode, u64 file_offset, node = tree_insert(&tree->tree, file_offset, &entry->rb_node); if (node) - ordered_data_tree_panic(inode, -EEXIST, file_offset); + btrfs_panic(fs_info, -EEXIST, + "inconsistency in ordered tree at offset %llu", + file_offset); spin_unlock_irq(&tree->lock); spin_lock(&root->ordered_extent_lock); @@ -247,27 +236,30 @@ static int __btrfs_add_ordered_extent(struct inode *inode, u64 file_offset, } int btrfs_add_ordered_extent(struct inode *inode, u64 file_offset, - u64 start, u64 len, u64 disk_len, int type) + u64 disk_bytenr, u64 num_bytes, u64 disk_num_bytes, + int type) { - return __btrfs_add_ordered_extent(inode, file_offset, start, len, - disk_len, type, 0, + return __btrfs_add_ordered_extent(inode, file_offset, disk_bytenr, + num_bytes, disk_num_bytes, type, 0, BTRFS_COMPRESS_NONE); } int btrfs_add_ordered_extent_dio(struct inode *inode, u64 file_offset, - u64 start, u64 len, u64 disk_len, int type) + u64 disk_bytenr, u64 num_bytes, + u64 disk_num_bytes, int type) { - return __btrfs_add_ordered_extent(inode, file_offset, start, len, - disk_len, type, 1, + return __btrfs_add_ordered_extent(inode, file_offset, disk_bytenr, + num_bytes, disk_num_bytes, type, 1, BTRFS_COMPRESS_NONE); } int btrfs_add_ordered_extent_compress(struct inode *inode, u64 file_offset, - u64 start, u64 len, u64 disk_len, - int type, int compress_type) + u64 disk_bytenr, u64 num_bytes, + u64 disk_num_bytes, int type, + int compress_type) { - return __btrfs_add_ordered_extent(inode, file_offset, start, len, - disk_len, type, 0, + return __btrfs_add_ordered_extent(inode, file_offset, disk_bytenr, + num_bytes, disk_num_bytes, type, 0, compress_type); } @@ -328,8 +320,8 @@ int btrfs_dec_test_first_ordered_pending(struct inode *inode, } dec_start = max(*file_offset, entry->file_offset); - dec_end = min(*file_offset + io_size, entry->file_offset + - entry->len); + dec_end = min(*file_offset + io_size, + entry->file_offset + entry->num_bytes); *file_offset = dec_end; if (dec_start > dec_end) { btrfs_crit(fs_info, "bad ordering dec_start %llu end %llu", @@ -471,10 +463,11 @@ void btrfs_remove_ordered_extent(struct inode *inode, btrfs_mod_outstanding_extents(btrfs_inode, -1); spin_unlock(&btrfs_inode->lock); if (root != fs_info->tree_root) - btrfs_delalloc_release_metadata(btrfs_inode, entry->len, false); + btrfs_delalloc_release_metadata(btrfs_inode, entry->num_bytes, + false); if (test_bit(BTRFS_ORDERED_DIRECT, &entry->flags)) - percpu_counter_add_batch(&fs_info->dio_bytes, -entry->len, + percpu_counter_add_batch(&fs_info->dio_bytes, -entry->num_bytes, fs_info->delalloc_batch); tree = &btrfs_inode->ordered_tree; @@ -534,8 +527,8 @@ u64 btrfs_wait_ordered_extents(struct btrfs_root *root, u64 nr, ordered = list_first_entry(&splice, struct btrfs_ordered_extent, root_extent_list); - if (range_end <= ordered->start || - ordered->start + ordered->disk_len <= range_start) { + if (range_end <= ordered->disk_bytenr || + ordered->disk_bytenr + ordered->disk_num_bytes <= range_start) { list_move_tail(&ordered->root_extent_list, &skipped); cond_resched_lock(&root->ordered_extent_lock); continue; @@ -619,7 +612,7 @@ void btrfs_start_ordered_extent(struct inode *inode, int wait) { u64 start = entry->file_offset; - u64 end = start + entry->len - 1; + u64 end = start + entry->num_bytes - 1; trace_btrfs_ordered_extent_start(inode, entry); @@ -680,7 +673,7 @@ int btrfs_wait_ordered_range(struct inode *inode, u64 start, u64 len) btrfs_put_ordered_extent(ordered); break; } - if (ordered->file_offset + ordered->len <= start) { + if (ordered->file_offset + ordered->num_bytes <= start) { btrfs_put_ordered_extent(ordered); break; } diff --git a/fs/btrfs/ordered-data.h b/fs/btrfs/ordered-data.h index 4eb0319a86d7..3beb4da4ab41 100644 --- a/fs/btrfs/ordered-data.h +++ b/fs/btrfs/ordered-data.h @@ -67,14 +67,13 @@ struct btrfs_ordered_extent { /* logical offset in the file */ u64 file_offset; - /* disk byte number */ - u64 start; - - /* ram length of the extent in bytes */ - u64 len; - - /* extent length on disk */ - u64 disk_len; + /* + * These fields directly correspond to the same fields in + * btrfs_file_extent_item. + */ + u64 disk_bytenr; + u64 num_bytes; + u64 disk_num_bytes; /* number of bytes that still need writing */ u64 bytes_left; @@ -161,12 +160,15 @@ int btrfs_dec_test_first_ordered_pending(struct inode *inode, u64 *file_offset, u64 io_size, int uptodate); int btrfs_add_ordered_extent(struct inode *inode, u64 file_offset, - u64 start, u64 len, u64 disk_len, int type); + u64 disk_bytenr, u64 num_bytes, u64 disk_num_bytes, + int type); int btrfs_add_ordered_extent_dio(struct inode *inode, u64 file_offset, - u64 start, u64 len, u64 disk_len, int type); + u64 disk_bytenr, u64 num_bytes, + u64 disk_num_bytes, int type); int btrfs_add_ordered_extent_compress(struct inode *inode, u64 file_offset, - u64 start, u64 len, u64 disk_len, - int type, int compress_type); + u64 disk_bytenr, u64 num_bytes, + u64 disk_num_bytes, int type, + int compress_type); void btrfs_add_ordered_sum(struct btrfs_ordered_extent *entry, struct btrfs_ordered_sum *sum); struct btrfs_ordered_extent *btrfs_lookup_ordered_extent(struct inode *inode, diff --git a/fs/btrfs/print-tree.c b/fs/btrfs/print-tree.c index 873b6b694107..61f44e78e3c9 100644 --- a/fs/btrfs/print-tree.c +++ b/fs/btrfs/print-tree.c @@ -317,7 +317,7 @@ void btrfs_print_leaf(struct extent_buffer *l) print_uuid_item(l, btrfs_item_ptr_offset(l, i), btrfs_item_size_nr(l, i)); break; - }; + } } } diff --git a/fs/btrfs/qgroup.c b/fs/btrfs/qgroup.c index 93aeb2e539a4..98d9a50352d6 100644 --- a/fs/btrfs/qgroup.c +++ b/fs/btrfs/qgroup.c @@ -1243,7 +1243,6 @@ int btrfs_add_qgroup_relation(struct btrfs_trans_handle *trans, u64 src, u64 dst) { struct btrfs_fs_info *fs_info = trans->fs_info; - struct btrfs_root *quota_root; struct btrfs_qgroup *parent; struct btrfs_qgroup *member; struct btrfs_qgroup_list *list; @@ -1259,9 +1258,8 @@ int btrfs_add_qgroup_relation(struct btrfs_trans_handle *trans, u64 src, return -ENOMEM; mutex_lock(&fs_info->qgroup_ioctl_lock); - quota_root = fs_info->quota_root; - if (!quota_root) { - ret = -EINVAL; + if (!fs_info->quota_root) { + ret = -ENOTCONN; goto out; } member = find_qgroup_rb(fs_info, src); @@ -1307,7 +1305,6 @@ static int __del_qgroup_relation(struct btrfs_trans_handle *trans, u64 src, u64 dst) { struct btrfs_fs_info *fs_info = trans->fs_info; - struct btrfs_root *quota_root; struct btrfs_qgroup *parent; struct btrfs_qgroup *member; struct btrfs_qgroup_list *list; @@ -1320,9 +1317,8 @@ static int __del_qgroup_relation(struct btrfs_trans_handle *trans, u64 src, if (!tmp) return -ENOMEM; - quota_root = fs_info->quota_root; - if (!quota_root) { - ret = -EINVAL; + if (!fs_info->quota_root) { + ret = -ENOTCONN; goto out; } @@ -1387,11 +1383,11 @@ int btrfs_create_qgroup(struct btrfs_trans_handle *trans, u64 qgroupid) int ret = 0; mutex_lock(&fs_info->qgroup_ioctl_lock); - quota_root = fs_info->quota_root; - if (!quota_root) { - ret = -EINVAL; + if (!fs_info->quota_root) { + ret = -ENOTCONN; goto out; } + quota_root = fs_info->quota_root; qgroup = find_qgroup_rb(fs_info, qgroupid); if (qgroup) { ret = -EEXIST; @@ -1416,15 +1412,13 @@ out: int btrfs_remove_qgroup(struct btrfs_trans_handle *trans, u64 qgroupid) { struct btrfs_fs_info *fs_info = trans->fs_info; - struct btrfs_root *quota_root; struct btrfs_qgroup *qgroup; struct btrfs_qgroup_list *list; int ret = 0; mutex_lock(&fs_info->qgroup_ioctl_lock); - quota_root = fs_info->quota_root; - if (!quota_root) { - ret = -EINVAL; + if (!fs_info->quota_root) { + ret = -ENOTCONN; goto out; } @@ -1465,7 +1459,6 @@ int btrfs_limit_qgroup(struct btrfs_trans_handle *trans, u64 qgroupid, struct btrfs_qgroup_limit *limit) { struct btrfs_fs_info *fs_info = trans->fs_info; - struct btrfs_root *quota_root; struct btrfs_qgroup *qgroup; int ret = 0; /* Sometimes we would want to clear the limit on this qgroup. @@ -1475,9 +1468,8 @@ int btrfs_limit_qgroup(struct btrfs_trans_handle *trans, u64 qgroupid, const u64 CLEAR_VALUE = -1; mutex_lock(&fs_info->qgroup_ioctl_lock); - quota_root = fs_info->quota_root; - if (!quota_root) { - ret = -EINVAL; + if (!fs_info->quota_root) { + ret = -ENOTCONN; goto out; } @@ -2423,8 +2415,12 @@ int btrfs_qgroup_account_extent(struct btrfs_trans_handle *trans, u64 bytenr, u64 nr_old_roots = 0; int ret = 0; + /* + * If quotas get disabled meanwhile, the resouces need to be freed and + * we can't just exit here. + */ if (!test_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags)) - return 0; + goto out_free; if (new_roots) { if (!maybe_fs_roots(new_roots)) @@ -2578,10 +2574,9 @@ cleanup: int btrfs_run_qgroups(struct btrfs_trans_handle *trans) { struct btrfs_fs_info *fs_info = trans->fs_info; - struct btrfs_root *quota_root = fs_info->quota_root; int ret = 0; - if (!quota_root) + if (!fs_info->quota_root) return ret; spin_lock(&fs_info->qgroup_lock); @@ -2875,7 +2870,6 @@ static bool qgroup_check_limits(struct btrfs_fs_info *fs_info, static int qgroup_reserve(struct btrfs_root *root, u64 num_bytes, bool enforce, enum btrfs_qgroup_rsv_type type) { - struct btrfs_root *quota_root; struct btrfs_qgroup *qgroup; struct btrfs_fs_info *fs_info = root->fs_info; u64 ref_root = root->root_key.objectid; @@ -2894,8 +2888,7 @@ static int qgroup_reserve(struct btrfs_root *root, u64 num_bytes, bool enforce, enforce = false; spin_lock(&fs_info->qgroup_lock); - quota_root = fs_info->quota_root; - if (!quota_root) + if (!fs_info->quota_root) goto out; qgroup = find_qgroup_rb(fs_info, ref_root); @@ -2962,7 +2955,6 @@ void btrfs_qgroup_free_refroot(struct btrfs_fs_info *fs_info, u64 ref_root, u64 num_bytes, enum btrfs_qgroup_rsv_type type) { - struct btrfs_root *quota_root; struct btrfs_qgroup *qgroup; struct ulist_node *unode; struct ulist_iterator uiter; @@ -2980,8 +2972,7 @@ void btrfs_qgroup_free_refroot(struct btrfs_fs_info *fs_info, } spin_lock(&fs_info->qgroup_lock); - quota_root = fs_info->quota_root; - if (!quota_root) + if (!fs_info->quota_root) goto out; qgroup = find_qgroup_rb(fs_info, ref_root); @@ -3232,12 +3223,12 @@ qgroup_rescan_init(struct btrfs_fs_info *fs_info, u64 progress_objectid, if (!(fs_info->qgroup_flags & BTRFS_QGROUP_STATUS_FLAG_RESCAN)) { btrfs_warn(fs_info, - "qgroup rescan init failed, qgroup is not enabled"); + "qgroup rescan init failed, qgroup rescan is not queued"); ret = -EINVAL; } else if (!(fs_info->qgroup_flags & BTRFS_QGROUP_STATUS_FLAG_ON)) { btrfs_warn(fs_info, - "qgroup rescan init failed, qgroup rescan is not queued"); + "qgroup rescan init failed, qgroup is not enabled"); ret = -EINVAL; } @@ -3681,7 +3672,6 @@ void __btrfs_qgroup_free_meta(struct btrfs_root *root, int num_bytes, static void qgroup_convert_meta(struct btrfs_fs_info *fs_info, u64 ref_root, int num_bytes) { - struct btrfs_root *quota_root = fs_info->quota_root; struct btrfs_qgroup *qgroup; struct ulist_node *unode; struct ulist_iterator uiter; @@ -3689,7 +3679,7 @@ static void qgroup_convert_meta(struct btrfs_fs_info *fs_info, u64 ref_root, if (num_bytes == 0) return; - if (!quota_root) + if (!fs_info->quota_root) return; spin_lock(&fs_info->qgroup_lock); diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c index d897a8e5e430..995d4b8b1cfd 100644 --- a/fs/btrfs/relocation.c +++ b/fs/btrfs/relocation.c @@ -517,6 +517,34 @@ static int update_backref_cache(struct btrfs_trans_handle *trans, return 1; } +static bool reloc_root_is_dead(struct btrfs_root *root) +{ + /* + * Pair with set_bit/clear_bit in clean_dirty_subvols and + * btrfs_update_reloc_root. We need to see the updated bit before + * trying to access reloc_root + */ + smp_rmb(); + if (test_bit(BTRFS_ROOT_DEAD_RELOC_TREE, &root->state)) + return true; + return false; +} + +/* + * Check if this subvolume tree has valid reloc tree. + * + * Reloc tree after swap is considered dead, thus not considered as valid. + * This is enough for most callers, as they don't distinguish dead reloc root + * from no reloc root. But should_ignore_root() below is a special case. + */ +static bool have_reloc_root(struct btrfs_root *root) +{ + if (reloc_root_is_dead(root)) + return false; + if (!root->reloc_root) + return false; + return true; +} static int should_ignore_root(struct btrfs_root *root) { @@ -525,6 +553,10 @@ static int should_ignore_root(struct btrfs_root *root) if (!test_bit(BTRFS_ROOT_REF_COWS, &root->state)) return 0; + /* This root has been merged with its reloc tree, we can ignore it */ + if (reloc_root_is_dead(root)) + return 1; + reloc_root = root->reloc_root; if (!reloc_root) return 0; @@ -1439,7 +1471,7 @@ int btrfs_init_reloc_root(struct btrfs_trans_handle *trans, * The subvolume has reloc tree but the swap is finished, no need to * create/update the dead reloc tree */ - if (test_bit(BTRFS_ROOT_DEAD_RELOC_TREE, &root->state)) + if (reloc_root_is_dead(root)) return 0; if (root->reloc_root) { @@ -1478,8 +1510,7 @@ int btrfs_update_reloc_root(struct btrfs_trans_handle *trans, struct btrfs_root_item *root_item; int ret; - if (test_bit(BTRFS_ROOT_DEAD_RELOC_TREE, &root->state) || - !root->reloc_root) + if (!have_reloc_root(root)) goto out; reloc_root = root->reloc_root; @@ -1489,6 +1520,11 @@ int btrfs_update_reloc_root(struct btrfs_trans_handle *trans, if (fs_info->reloc_ctl->merge_reloc_tree && btrfs_root_refs(root_item) == 0) { set_bit(BTRFS_ROOT_DEAD_RELOC_TREE, &root->state); + /* + * Mark the tree as dead before we change reloc_root so + * have_reloc_root will not touch it from now on. + */ + smp_wmb(); __del_reloc_root(reloc_root); } @@ -2201,6 +2237,11 @@ static int clean_dirty_subvols(struct reloc_control *rc) if (ret2 < 0 && !ret) ret = ret2; } + /* + * Need barrier to ensure clear_bit() only happens after + * root->reloc_root = NULL. Pairs with have_reloc_root. + */ + smp_wmb(); clear_bit(BTRFS_ROOT_DEAD_RELOC_TREE, &root->state); btrfs_put_fs_root(root); } else { @@ -4291,6 +4332,15 @@ static void describe_relocation(struct btrfs_fs_info *fs_info, block_group->start, buf); } +static const char *stage_to_string(int stage) +{ + if (stage == MOVE_DATA_EXTENTS) + return "move data extents"; + if (stage == UPDATE_DATA_PTRS) + return "update data pointers"; + return "unknown"; +} + /* * function to relocate all extents in a block group. */ @@ -4365,12 +4415,15 @@ int btrfs_relocate_block_group(struct btrfs_fs_info *fs_info, u64 group_start) rc->block_group->length); while (1) { + int finishes_stage; + mutex_lock(&fs_info->cleaner_mutex); ret = relocate_block_group(rc); mutex_unlock(&fs_info->cleaner_mutex); if (ret < 0) err = ret; + finishes_stage = rc->stage; /* * We may have gotten ENOSPC after we already dirtied some * extents. If writeout happens while we're relocating a @@ -4396,8 +4449,8 @@ int btrfs_relocate_block_group(struct btrfs_fs_info *fs_info, u64 group_start) if (rc->extents_found == 0) break; - btrfs_info(fs_info, "found %llu extents", rc->extents_found); - + btrfs_info(fs_info, "found %llu extents, stage: %s", + rc->extents_found, stage_to_string(finishes_stage)); } WARN_ON(rc->block_group->pinned > 0); @@ -4552,6 +4605,7 @@ int btrfs_recover_relocation(struct btrfs_root *root) fs_root = read_fs_root(fs_info, reloc_root->root_key.offset); if (IS_ERR(fs_root)) { err = PTR_ERR(fs_root); + list_add_tail(&reloc_root->root_list, &reloc_roots); goto out_free; } @@ -4614,7 +4668,7 @@ int btrfs_reloc_clone_csums(struct inode *inode, u64 file_pos, u64 len) LIST_HEAD(list); ordered = btrfs_lookup_ordered_extent(inode, file_pos); - BUG_ON(ordered->file_offset != file_pos || ordered->len != len); + BUG_ON(ordered->file_offset != file_pos || ordered->num_bytes != len); disk_bytenr = file_pos + BTRFS_I(inode)->index_cnt; ret = btrfs_lookup_csums_range(fs_info->csum_root, disk_bytenr, @@ -4638,7 +4692,7 @@ int btrfs_reloc_clone_csums(struct inode *inode, u64 file_pos, u64 len) * disk_len vs real len like with real inodes since it's all * disk length. */ - new_bytenr = ordered->start + (sums->bytenr - disk_bytenr); + new_bytenr = ordered->disk_bytenr + sums->bytenr - disk_bytenr; sums->bytenr = new_bytenr; btrfs_add_ordered_sum(ordered, sums); @@ -4717,7 +4771,7 @@ void btrfs_reloc_pre_snapshot(struct btrfs_pending_snapshot *pending, struct btrfs_root *root = pending->root; struct reloc_control *rc = root->fs_info->reloc_ctl; - if (!root->reloc_root || !rc) + if (!rc || !have_reloc_root(root)) return; if (!rc->merge_reloc_tree) @@ -4751,7 +4805,7 @@ int btrfs_reloc_post_snapshot(struct btrfs_trans_handle *trans, struct reloc_control *rc = root->fs_info->reloc_ctl; int ret; - if (!root->reloc_root || !rc) + if (!rc || !have_reloc_root(root)) return 0; rc = root->fs_info->reloc_ctl; diff --git a/fs/btrfs/root-tree.c b/fs/btrfs/root-tree.c index 3b17b647d002..612411c74550 100644 --- a/fs/btrfs/root-tree.c +++ b/fs/btrfs/root-tree.c @@ -376,11 +376,13 @@ again: leaf = path->nodes[0]; ref = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_root_ref); - - WARN_ON(btrfs_root_ref_dirid(leaf, ref) != dirid); - WARN_ON(btrfs_root_ref_name_len(leaf, ref) != name_len); ptr = (unsigned long)(ref + 1); - WARN_ON(memcmp_extent_buffer(leaf, name, ptr, name_len)); + if ((btrfs_root_ref_dirid(leaf, ref) != dirid) || + (btrfs_root_ref_name_len(leaf, ref) != name_len) || + memcmp_extent_buffer(leaf, name, ptr, name_len)) { + err = -ENOENT; + goto out; + } *sequence = btrfs_root_ref_sequence(leaf, ref); ret = btrfs_del_item(trans, tree_root, path); diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c index 21de630b0730..61b37c56a7fb 100644 --- a/fs/btrfs/scrub.c +++ b/fs/btrfs/scrub.c @@ -8,6 +8,7 @@ #include <linux/sched/mm.h> #include <crypto/hash.h> #include "ctree.h" +#include "discard.h" #include "volumes.h" #include "disk-io.h" #include "ordered-data.h" @@ -3577,17 +3578,27 @@ int scrub_enumerate_chunks(struct scrub_ctx *sctx, * This can easily boost the amount of SYSTEM chunks if cleaner * thread can't be triggered fast enough, and use up all space * of btrfs_super_block::sys_chunk_array + * + * While for dev replace, we need to try our best to mark block + * group RO, to prevent race between: + * - Write duplication + * Contains latest data + * - Scrub copy + * Contains data from commit tree + * + * If target block group is not marked RO, nocow writes can + * be overwritten by scrub copy, causing data corruption. + * So for dev-replace, it's not allowed to continue if a block + * group is not RO. */ - ret = btrfs_inc_block_group_ro(cache, false); - scrub_pause_off(fs_info); - + ret = btrfs_inc_block_group_ro(cache, sctx->is_dev_replace); if (ret == 0) { ro_set = 1; - } else if (ret == -ENOSPC) { + } else if (ret == -ENOSPC && !sctx->is_dev_replace) { /* * btrfs_inc_block_group_ro return -ENOSPC when it * failed in creating new chunk for metadata. - * It is not a problem for scrub/replace, because + * It is not a problem for scrub, because * metadata are always cowed, and our scrub paused * commit_transactions. */ @@ -3596,9 +3607,22 @@ int scrub_enumerate_chunks(struct scrub_ctx *sctx, btrfs_warn(fs_info, "failed setting block group ro: %d", ret); btrfs_put_block_group(cache); + scrub_pause_off(fs_info); break; } + /* + * Now the target block is marked RO, wait for nocow writes to + * finish before dev-replace. + * COW is fine, as COW never overwrites extents in commit tree. + */ + if (sctx->is_dev_replace) { + btrfs_wait_nocow_writers(cache); + btrfs_wait_ordered_roots(fs_info, U64_MAX, cache->start, + cache->length); + } + + scrub_pause_off(fs_info); down_write(&dev_replace->rwsem); dev_replace->cursor_right = found_key.offset + length; dev_replace->cursor_left = found_key.offset; @@ -3659,7 +3683,11 @@ int scrub_enumerate_chunks(struct scrub_ctx *sctx, if (!cache->removed && !cache->ro && cache->reserved == 0 && cache->used == 0) { spin_unlock(&cache->lock); - btrfs_mark_bg_unused(cache); + if (btrfs_test_opt(fs_info, DISCARD_ASYNC)) + btrfs_discard_queue_work(&fs_info->discard_ctl, + cache); + else + btrfs_mark_bg_unused(cache); } else { spin_unlock(&cache->lock); } diff --git a/fs/btrfs/send.c b/fs/btrfs/send.c index ae2db5eb1549..091e5bc8c7ea 100644 --- a/fs/btrfs/send.c +++ b/fs/btrfs/send.c @@ -7084,12 +7084,6 @@ long btrfs_ioctl_send(struct file *mnt_file, struct btrfs_ioctl_send_args *arg) spin_unlock(&send_root->root_item_lock); /* - * This is done when we lookup the root, it should already be complete - * by the time we get here. - */ - WARN_ON(send_root->orphan_cleanup_state != ORPHAN_CLEANUP_DONE); - - /* * Userspace tools do the checks and warn the user if it's * not RO. */ diff --git a/fs/btrfs/space-info.c b/fs/btrfs/space-info.c index f09aa6ee9113..537bc310a673 100644 --- a/fs/btrfs/space-info.c +++ b/fs/btrfs/space-info.c @@ -161,8 +161,7 @@ static inline u64 calc_global_rsv_need_space(struct btrfs_block_rsv *global) static int can_overcommit(struct btrfs_fs_info *fs_info, struct btrfs_space_info *space_info, u64 bytes, - enum btrfs_reserve_flush_enum flush, - bool system_chunk) + enum btrfs_reserve_flush_enum flush) { u64 profile; u64 avail; @@ -173,7 +172,7 @@ static int can_overcommit(struct btrfs_fs_info *fs_info, if (space_info->flags & BTRFS_BLOCK_GROUP_DATA) return 0; - if (system_chunk) + if (space_info->flags & BTRFS_BLOCK_GROUP_SYSTEM) profile = btrfs_system_alloc_profile(fs_info); else profile = btrfs_metadata_alloc_profile(fs_info); @@ -227,8 +226,7 @@ again: /* Check and see if our ticket can be satisified now. */ if ((used + ticket->bytes <= space_info->total_bytes) || - can_overcommit(fs_info, space_info, ticket->bytes, flush, - false)) { + can_overcommit(fs_info, space_info, ticket->bytes, flush)) { btrfs_space_info_update_bytes_may_use(fs_info, space_info, ticket->bytes); @@ -626,8 +624,7 @@ static void flush_space(struct btrfs_fs_info *fs_info, static inline u64 btrfs_calc_reclaim_metadata_size(struct btrfs_fs_info *fs_info, - struct btrfs_space_info *space_info, - bool system_chunk) + struct btrfs_space_info *space_info) { struct reserve_ticket *ticket; u64 used; @@ -643,13 +640,12 @@ btrfs_calc_reclaim_metadata_size(struct btrfs_fs_info *fs_info, to_reclaim = min_t(u64, num_online_cpus() * SZ_1M, SZ_16M); if (can_overcommit(fs_info, space_info, to_reclaim, - BTRFS_RESERVE_FLUSH_ALL, system_chunk)) + BTRFS_RESERVE_FLUSH_ALL)) return 0; used = btrfs_space_info_used(space_info, true); - if (can_overcommit(fs_info, space_info, SZ_1M, - BTRFS_RESERVE_FLUSH_ALL, system_chunk)) + if (can_overcommit(fs_info, space_info, SZ_1M, BTRFS_RESERVE_FLUSH_ALL)) expected = div_factor_fine(space_info->total_bytes, 95); else expected = div_factor_fine(space_info->total_bytes, 90); @@ -665,7 +661,7 @@ btrfs_calc_reclaim_metadata_size(struct btrfs_fs_info *fs_info, static inline int need_do_async_reclaim(struct btrfs_fs_info *fs_info, struct btrfs_space_info *space_info, - u64 used, bool system_chunk) + u64 used) { u64 thresh = div_factor_fine(space_info->total_bytes, 98); @@ -673,8 +669,7 @@ static inline int need_do_async_reclaim(struct btrfs_fs_info *fs_info, if ((space_info->bytes_used + space_info->bytes_reserved) >= thresh) return 0; - if (!btrfs_calc_reclaim_metadata_size(fs_info, space_info, - system_chunk)) + if (!btrfs_calc_reclaim_metadata_size(fs_info, space_info)) return 0; return (used >= thresh && !btrfs_fs_closing(fs_info) && @@ -765,8 +760,7 @@ static void btrfs_async_reclaim_metadata_space(struct work_struct *work) space_info = btrfs_find_space_info(fs_info, BTRFS_BLOCK_GROUP_METADATA); spin_lock(&space_info->lock); - to_reclaim = btrfs_calc_reclaim_metadata_size(fs_info, space_info, - false); + to_reclaim = btrfs_calc_reclaim_metadata_size(fs_info, space_info); if (!to_reclaim) { space_info->flush = 0; spin_unlock(&space_info->lock); @@ -785,8 +779,7 @@ static void btrfs_async_reclaim_metadata_space(struct work_struct *work) return; } to_reclaim = btrfs_calc_reclaim_metadata_size(fs_info, - space_info, - false); + space_info); if (last_tickets_id == space_info->tickets_id) { flush_state++; } else { @@ -858,8 +851,7 @@ static void priority_reclaim_metadata_space(struct btrfs_fs_info *fs_info, int flush_state; spin_lock(&space_info->lock); - to_reclaim = btrfs_calc_reclaim_metadata_size(fs_info, space_info, - false); + to_reclaim = btrfs_calc_reclaim_metadata_size(fs_info, space_info); if (!to_reclaim) { spin_unlock(&space_info->lock); return; @@ -990,8 +982,7 @@ static int handle_reserve_ticket(struct btrfs_fs_info *fs_info, static int __reserve_metadata_bytes(struct btrfs_fs_info *fs_info, struct btrfs_space_info *space_info, u64 orig_bytes, - enum btrfs_reserve_flush_enum flush, - bool system_chunk) + enum btrfs_reserve_flush_enum flush) { struct reserve_ticket ticket; u64 used; @@ -1013,8 +1004,7 @@ static int __reserve_metadata_bytes(struct btrfs_fs_info *fs_info, */ if (!pending_tickets && ((used + orig_bytes <= space_info->total_bytes) || - can_overcommit(fs_info, space_info, orig_bytes, flush, - system_chunk))) { + can_overcommit(fs_info, space_info, orig_bytes, flush))) { btrfs_space_info_update_bytes_may_use(fs_info, space_info, orig_bytes); ret = 0; @@ -1054,8 +1044,7 @@ static int __reserve_metadata_bytes(struct btrfs_fs_info *fs_info, * the async reclaim as we will panic. */ if (!test_bit(BTRFS_FS_LOG_RECOVERING, &fs_info->flags) && - need_do_async_reclaim(fs_info, space_info, - used, system_chunk) && + need_do_async_reclaim(fs_info, space_info, used) && !work_busy(&fs_info->async_reclaim_work)) { trace_btrfs_trigger_flush(fs_info, space_info->flags, orig_bytes, flush, "preempt"); @@ -1092,10 +1081,9 @@ int btrfs_reserve_metadata_bytes(struct btrfs_root *root, struct btrfs_fs_info *fs_info = root->fs_info; struct btrfs_block_rsv *global_rsv = &fs_info->global_block_rsv; int ret; - bool system_chunk = (root == fs_info->chunk_root); ret = __reserve_metadata_bytes(fs_info, block_rsv->space_info, - orig_bytes, flush, system_chunk); + orig_bytes, flush); if (ret == -ENOSPC && unlikely(root->orphan_cleanup_state == ORPHAN_CLEANUP_STARTED)) { if (block_rsv != global_rsv && diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index f452a94abdc3..a906315efd19 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -46,6 +46,7 @@ #include "sysfs.h" #include "tests/btrfs-tests.h" #include "block-group.h" +#include "discard.h" #include "qgroup.h" #define CREATE_TRACE_POINTS @@ -146,6 +147,8 @@ void __btrfs_handle_fs_error(struct btrfs_fs_info *fs_info, const char *function if (sb_rdonly(sb)) return; + btrfs_discard_stop(fs_info); + /* btrfs handle error by forcing the filesystem readonly */ sb->s_flags |= SB_RDONLY; btrfs_info(fs_info, "forced readonly"); @@ -313,6 +316,7 @@ enum { Opt_datasum, Opt_nodatasum, Opt_defrag, Opt_nodefrag, Opt_discard, Opt_nodiscard, + Opt_discard_mode, Opt_nologreplay, Opt_norecovery, Opt_ratio, @@ -375,6 +379,7 @@ static const match_table_t tokens = { {Opt_defrag, "autodefrag"}, {Opt_nodefrag, "noautodefrag"}, {Opt_discard, "discard"}, + {Opt_discard_mode, "discard=%s"}, {Opt_nodiscard, "nodiscard"}, {Opt_nologreplay, "nologreplay"}, {Opt_norecovery, "norecovery"}, @@ -695,12 +700,26 @@ int btrfs_parse_options(struct btrfs_fs_info *info, char *options, info->metadata_ratio); break; case Opt_discard: - btrfs_set_and_info(info, DISCARD, - "turning on discard"); + case Opt_discard_mode: + if (token == Opt_discard || + strcmp(args[0].from, "sync") == 0) { + btrfs_clear_opt(info->mount_opt, DISCARD_ASYNC); + btrfs_set_and_info(info, DISCARD_SYNC, + "turning on sync discard"); + } else if (strcmp(args[0].from, "async") == 0) { + btrfs_clear_opt(info->mount_opt, DISCARD_SYNC); + btrfs_set_and_info(info, DISCARD_ASYNC, + "turning on async discard"); + } else { + ret = -EINVAL; + goto out; + } break; case Opt_nodiscard: - btrfs_clear_and_info(info, DISCARD, + btrfs_clear_and_info(info, DISCARD_SYNC, "turning off discard"); + btrfs_clear_and_info(info, DISCARD_ASYNC, + "turning off async discard"); break; case Opt_space_cache: case Opt_space_cache_version: @@ -1322,8 +1341,10 @@ static int btrfs_show_options(struct seq_file *seq, struct dentry *dentry) seq_puts(seq, ",nologreplay"); if (btrfs_test_opt(info, FLUSHONCOMMIT)) seq_puts(seq, ",flushoncommit"); - if (btrfs_test_opt(info, DISCARD)) + if (btrfs_test_opt(info, DISCARD_SYNC)) seq_puts(seq, ",discard"); + if (btrfs_test_opt(info, DISCARD_ASYNC)) + seq_puts(seq, ",discard=async"); if (!(info->sb->s_flags & SB_POSIXACL)) seq_puts(seq, ",noacl"); if (btrfs_test_opt(info, SPACE_CACHE)) @@ -1713,6 +1734,14 @@ static inline void btrfs_remount_cleanup(struct btrfs_fs_info *fs_info, btrfs_cleanup_defrag_inodes(fs_info); } + /* If we toggled discard async */ + if (!btrfs_raw_test_opt(old_opts, DISCARD_ASYNC) && + btrfs_test_opt(fs_info, DISCARD_ASYNC)) + btrfs_discard_resume(fs_info); + else if (btrfs_raw_test_opt(old_opts, DISCARD_ASYNC) && + !btrfs_test_opt(fs_info, DISCARD_ASYNC)) + btrfs_discard_cleanup(fs_info); + clear_bit(BTRFS_FS_STATE_REMOUNTING, &fs_info->fs_state); } @@ -1760,6 +1789,8 @@ static int btrfs_remount(struct super_block *sb, int *flags, char *data) */ cancel_work_sync(&fs_info->async_reclaim_work); + btrfs_discard_cleanup(fs_info); + /* wait for the uuid_scan task to finish */ down(&fs_info->uuid_tree_rescan_sem); /* avoid complains from lockdep et al. */ diff --git a/fs/btrfs/sysfs.c b/fs/btrfs/sysfs.c index 5ebbe8a5ee76..7436422194da 100644 --- a/fs/btrfs/sysfs.c +++ b/fs/btrfs/sysfs.c @@ -12,6 +12,7 @@ #include <crypto/hash.h> #include "ctree.h" +#include "discard.h" #include "disk-io.h" #include "transaction.h" #include "sysfs.h" @@ -339,11 +340,177 @@ static const struct attribute_group btrfs_static_feature_attr_group = { #ifdef CONFIG_BTRFS_DEBUG /* + * Discard statistics and tunables + */ +#define discard_to_fs_info(_kobj) to_fs_info((_kobj)->parent->parent) + +static ssize_t btrfs_discardable_bytes_show(struct kobject *kobj, + struct kobj_attribute *a, + char *buf) +{ + struct btrfs_fs_info *fs_info = discard_to_fs_info(kobj); + + return snprintf(buf, PAGE_SIZE, "%lld\n", + atomic64_read(&fs_info->discard_ctl.discardable_bytes)); +} +BTRFS_ATTR(discard, discardable_bytes, btrfs_discardable_bytes_show); + +static ssize_t btrfs_discardable_extents_show(struct kobject *kobj, + struct kobj_attribute *a, + char *buf) +{ + struct btrfs_fs_info *fs_info = discard_to_fs_info(kobj); + + return snprintf(buf, PAGE_SIZE, "%d\n", + atomic_read(&fs_info->discard_ctl.discardable_extents)); +} +BTRFS_ATTR(discard, discardable_extents, btrfs_discardable_extents_show); + +static ssize_t btrfs_discard_bitmap_bytes_show(struct kobject *kobj, + struct kobj_attribute *a, + char *buf) +{ + struct btrfs_fs_info *fs_info = discard_to_fs_info(kobj); + + return snprintf(buf, PAGE_SIZE, "%lld\n", + fs_info->discard_ctl.discard_bitmap_bytes); +} +BTRFS_ATTR(discard, discard_bitmap_bytes, btrfs_discard_bitmap_bytes_show); + +static ssize_t btrfs_discard_bytes_saved_show(struct kobject *kobj, + struct kobj_attribute *a, + char *buf) +{ + struct btrfs_fs_info *fs_info = discard_to_fs_info(kobj); + + return snprintf(buf, PAGE_SIZE, "%lld\n", + atomic64_read(&fs_info->discard_ctl.discard_bytes_saved)); +} +BTRFS_ATTR(discard, discard_bytes_saved, btrfs_discard_bytes_saved_show); + +static ssize_t btrfs_discard_extent_bytes_show(struct kobject *kobj, + struct kobj_attribute *a, + char *buf) +{ + struct btrfs_fs_info *fs_info = discard_to_fs_info(kobj); + + return snprintf(buf, PAGE_SIZE, "%lld\n", + fs_info->discard_ctl.discard_extent_bytes); +} +BTRFS_ATTR(discard, discard_extent_bytes, btrfs_discard_extent_bytes_show); + +static ssize_t btrfs_discard_iops_limit_show(struct kobject *kobj, + struct kobj_attribute *a, + char *buf) +{ + struct btrfs_fs_info *fs_info = discard_to_fs_info(kobj); + + return snprintf(buf, PAGE_SIZE, "%u\n", + READ_ONCE(fs_info->discard_ctl.iops_limit)); +} + +static ssize_t btrfs_discard_iops_limit_store(struct kobject *kobj, + struct kobj_attribute *a, + const char *buf, size_t len) +{ + struct btrfs_fs_info *fs_info = discard_to_fs_info(kobj); + struct btrfs_discard_ctl *discard_ctl = &fs_info->discard_ctl; + u32 iops_limit; + int ret; + + ret = kstrtou32(buf, 10, &iops_limit); + if (ret) + return -EINVAL; + + WRITE_ONCE(discard_ctl->iops_limit, iops_limit); + + return len; +} +BTRFS_ATTR_RW(discard, iops_limit, btrfs_discard_iops_limit_show, + btrfs_discard_iops_limit_store); + +static ssize_t btrfs_discard_kbps_limit_show(struct kobject *kobj, + struct kobj_attribute *a, + char *buf) +{ + struct btrfs_fs_info *fs_info = discard_to_fs_info(kobj); + + return snprintf(buf, PAGE_SIZE, "%u\n", + READ_ONCE(fs_info->discard_ctl.kbps_limit)); +} + +static ssize_t btrfs_discard_kbps_limit_store(struct kobject *kobj, + struct kobj_attribute *a, + const char *buf, size_t len) +{ + struct btrfs_fs_info *fs_info = discard_to_fs_info(kobj); + struct btrfs_discard_ctl *discard_ctl = &fs_info->discard_ctl; + u32 kbps_limit; + int ret; + + ret = kstrtou32(buf, 10, &kbps_limit); + if (ret) + return -EINVAL; + + WRITE_ONCE(discard_ctl->kbps_limit, kbps_limit); + + return len; +} +BTRFS_ATTR_RW(discard, kbps_limit, btrfs_discard_kbps_limit_show, + btrfs_discard_kbps_limit_store); + +static ssize_t btrfs_discard_max_discard_size_show(struct kobject *kobj, + struct kobj_attribute *a, + char *buf) +{ + struct btrfs_fs_info *fs_info = discard_to_fs_info(kobj); + + return snprintf(buf, PAGE_SIZE, "%llu\n", + READ_ONCE(fs_info->discard_ctl.max_discard_size)); +} + +static ssize_t btrfs_discard_max_discard_size_store(struct kobject *kobj, + struct kobj_attribute *a, + const char *buf, size_t len) +{ + struct btrfs_fs_info *fs_info = discard_to_fs_info(kobj); + struct btrfs_discard_ctl *discard_ctl = &fs_info->discard_ctl; + u64 max_discard_size; + int ret; + + ret = kstrtou64(buf, 10, &max_discard_size); + if (ret) + return -EINVAL; + + WRITE_ONCE(discard_ctl->max_discard_size, max_discard_size); + + return len; +} +BTRFS_ATTR_RW(discard, max_discard_size, btrfs_discard_max_discard_size_show, + btrfs_discard_max_discard_size_store); + +static const struct attribute *discard_debug_attrs[] = { + BTRFS_ATTR_PTR(discard, discardable_bytes), + BTRFS_ATTR_PTR(discard, discardable_extents), + BTRFS_ATTR_PTR(discard, discard_bitmap_bytes), + BTRFS_ATTR_PTR(discard, discard_bytes_saved), + BTRFS_ATTR_PTR(discard, discard_extent_bytes), + BTRFS_ATTR_PTR(discard, iops_limit), + BTRFS_ATTR_PTR(discard, kbps_limit), + BTRFS_ATTR_PTR(discard, max_discard_size), + NULL, +}; + +/* * Runtime debugging exported via sysfs * * /sys/fs/btrfs/debug - applies to module or all filesystems * /sys/fs/btrfs/UUID - applies only to the given filesystem */ +static const struct attribute *btrfs_debug_mount_attrs[] = { + NULL, +}; + static struct attribute *btrfs_debug_feature_attrs[] = { NULL }; @@ -734,10 +901,10 @@ static int addrm_unknown_feature_attrs(struct btrfs_fs_info *fs_info, bool add) static void __btrfs_sysfs_remove_fsid(struct btrfs_fs_devices *fs_devs) { - if (fs_devs->device_dir_kobj) { - kobject_del(fs_devs->device_dir_kobj); - kobject_put(fs_devs->device_dir_kobj); - fs_devs->device_dir_kobj = NULL; + if (fs_devs->devices_kobj) { + kobject_del(fs_devs->devices_kobj); + kobject_put(fs_devs->devices_kobj); + fs_devs->devices_kobj = NULL; } if (fs_devs->fsid_kobj.state_initialized) { @@ -771,6 +938,19 @@ void btrfs_sysfs_remove_mounted(struct btrfs_fs_info *fs_info) kobject_del(fs_info->space_info_kobj); kobject_put(fs_info->space_info_kobj); } +#ifdef CONFIG_BTRFS_DEBUG + if (fs_info->discard_debug_kobj) { + sysfs_remove_files(fs_info->discard_debug_kobj, + discard_debug_attrs); + kobject_del(fs_info->discard_debug_kobj); + kobject_put(fs_info->discard_debug_kobj); + } + if (fs_info->debug_kobj) { + sysfs_remove_files(fs_info->debug_kobj, btrfs_debug_mount_attrs); + kobject_del(fs_info->debug_kobj); + kobject_put(fs_info->debug_kobj); + } +#endif addrm_unknown_feature_attrs(fs_info, false); sysfs_remove_group(&fs_info->fs_devices->fsid_kobj, &btrfs_feature_attr_group); sysfs_remove_files(&fs_info->fs_devices->fsid_kobj, btrfs_attrs); @@ -969,46 +1149,120 @@ int btrfs_sysfs_rm_device_link(struct btrfs_fs_devices *fs_devices, struct hd_struct *disk; struct kobject *disk_kobj; - if (!fs_devices->device_dir_kobj) + if (!fs_devices->devices_kobj) return -EINVAL; - if (one_device && one_device->bdev) { - disk = one_device->bdev->bd_part; - disk_kobj = &part_to_dev(disk)->kobj; + if (one_device) { + if (one_device->bdev) { + disk = one_device->bdev->bd_part; + disk_kobj = &part_to_dev(disk)->kobj; + sysfs_remove_link(fs_devices->devices_kobj, + disk_kobj->name); + } - sysfs_remove_link(fs_devices->device_dir_kobj, - disk_kobj->name); - } + kobject_del(&one_device->devid_kobj); + kobject_put(&one_device->devid_kobj); + + wait_for_completion(&one_device->kobj_unregister); - if (one_device) return 0; + } - list_for_each_entry(one_device, - &fs_devices->devices, dev_list) { - if (!one_device->bdev) - continue; - disk = one_device->bdev->bd_part; - disk_kobj = &part_to_dev(disk)->kobj; + list_for_each_entry(one_device, &fs_devices->devices, dev_list) { + + if (one_device->bdev) { + disk = one_device->bdev->bd_part; + disk_kobj = &part_to_dev(disk)->kobj; + sysfs_remove_link(fs_devices->devices_kobj, + disk_kobj->name); + } + kobject_del(&one_device->devid_kobj); + kobject_put(&one_device->devid_kobj); - sysfs_remove_link(fs_devices->device_dir_kobj, - disk_kobj->name); + wait_for_completion(&one_device->kobj_unregister); } return 0; } -int btrfs_sysfs_add_device(struct btrfs_fs_devices *fs_devs) +static ssize_t btrfs_devinfo_in_fs_metadata_show(struct kobject *kobj, + struct kobj_attribute *a, + char *buf) { - if (!fs_devs->device_dir_kobj) - fs_devs->device_dir_kobj = kobject_create_and_add("devices", - &fs_devs->fsid_kobj); + int val; + struct btrfs_device *device = container_of(kobj, struct btrfs_device, + devid_kobj); - if (!fs_devs->device_dir_kobj) - return -ENOMEM; + val = !!test_bit(BTRFS_DEV_STATE_IN_FS_METADATA, &device->dev_state); - return 0; + return snprintf(buf, PAGE_SIZE, "%d\n", val); +} +BTRFS_ATTR(devid, in_fs_metadata, btrfs_devinfo_in_fs_metadata_show); + +static ssize_t btrfs_sysfs_missing_show(struct kobject *kobj, + struct kobj_attribute *a, char *buf) +{ + int val; + struct btrfs_device *device = container_of(kobj, struct btrfs_device, + devid_kobj); + + val = !!test_bit(BTRFS_DEV_STATE_MISSING, &device->dev_state); + + return snprintf(buf, PAGE_SIZE, "%d\n", val); +} +BTRFS_ATTR(devid, missing, btrfs_sysfs_missing_show); + +static ssize_t btrfs_devinfo_replace_target_show(struct kobject *kobj, + struct kobj_attribute *a, + char *buf) +{ + int val; + struct btrfs_device *device = container_of(kobj, struct btrfs_device, + devid_kobj); + + val = !!test_bit(BTRFS_DEV_STATE_REPLACE_TGT, &device->dev_state); + + return snprintf(buf, PAGE_SIZE, "%d\n", val); +} +BTRFS_ATTR(devid, replace_target, btrfs_devinfo_replace_target_show); + +static ssize_t btrfs_devinfo_writeable_show(struct kobject *kobj, + struct kobj_attribute *a, char *buf) +{ + int val; + struct btrfs_device *device = container_of(kobj, struct btrfs_device, + devid_kobj); + + val = !!test_bit(BTRFS_DEV_STATE_WRITEABLE, &device->dev_state); + + return snprintf(buf, PAGE_SIZE, "%d\n", val); +} +BTRFS_ATTR(devid, writeable, btrfs_devinfo_writeable_show); + +static struct attribute *devid_attrs[] = { + BTRFS_ATTR_PTR(devid, in_fs_metadata), + BTRFS_ATTR_PTR(devid, missing), + BTRFS_ATTR_PTR(devid, replace_target), + BTRFS_ATTR_PTR(devid, writeable), + NULL +}; +ATTRIBUTE_GROUPS(devid); + +static void btrfs_release_devid_kobj(struct kobject *kobj) +{ + struct btrfs_device *device = container_of(kobj, struct btrfs_device, + devid_kobj); + + memset(&device->devid_kobj, 0, sizeof(struct kobject)); + complete(&device->kobj_unregister); } +static struct kobj_type devid_ktype = { + .sysfs_ops = &kobj_sysfs_ops, + .default_groups = devid_groups, + .release = btrfs_release_devid_kobj, +}; + int btrfs_sysfs_add_device_link(struct btrfs_fs_devices *fs_devices, struct btrfs_device *one_device) { @@ -1016,22 +1270,31 @@ int btrfs_sysfs_add_device_link(struct btrfs_fs_devices *fs_devices, struct btrfs_device *dev; list_for_each_entry(dev, &fs_devices->devices, dev_list) { - struct hd_struct *disk; - struct kobject *disk_kobj; - - if (!dev->bdev) - continue; if (one_device && one_device != dev) continue; - disk = dev->bdev->bd_part; - disk_kobj = &part_to_dev(disk)->kobj; + if (dev->bdev) { + struct hd_struct *disk; + struct kobject *disk_kobj; + + disk = dev->bdev->bd_part; + disk_kobj = &part_to_dev(disk)->kobj; - error = sysfs_create_link(fs_devices->device_dir_kobj, - disk_kobj, disk_kobj->name); - if (error) + error = sysfs_create_link(fs_devices->devices_kobj, + disk_kobj, disk_kobj->name); + if (error) + break; + } + + init_completion(&dev->kobj_unregister); + error = kobject_init_and_add(&dev->devid_kobj, &devid_ktype, + fs_devices->devices_kobj, "%llu", + dev->devid); + if (error) { + kobject_put(&dev->devid_kobj); break; + } } return error; @@ -1063,27 +1326,49 @@ void btrfs_sysfs_update_sprout_fsid(struct btrfs_fs_devices *fs_devices, "sysfs: failed to create fsid for sprout"); } +void btrfs_sysfs_update_devid(struct btrfs_device *device) +{ + char tmp[24]; + + snprintf(tmp, sizeof(tmp), "%llu", device->devid); + + if (kobject_rename(&device->devid_kobj, tmp)) + btrfs_warn(device->fs_devices->fs_info, + "sysfs: failed to update devid for %llu", + device->devid); +} + /* /sys/fs/btrfs/ entry */ static struct kset *btrfs_kset; /* + * Creates: + * /sys/fs/btrfs/UUID + * * Can be called by the device discovery thread. - * And parent can be specified for seed device */ -int btrfs_sysfs_add_fsid(struct btrfs_fs_devices *fs_devs, - struct kobject *parent) +int btrfs_sysfs_add_fsid(struct btrfs_fs_devices *fs_devs) { int error; init_completion(&fs_devs->kobj_unregister); fs_devs->fsid_kobj.kset = btrfs_kset; - error = kobject_init_and_add(&fs_devs->fsid_kobj, - &btrfs_ktype, parent, "%pU", fs_devs->fsid); + error = kobject_init_and_add(&fs_devs->fsid_kobj, &btrfs_ktype, NULL, + "%pU", fs_devs->fsid); if (error) { kobject_put(&fs_devs->fsid_kobj); return error; } + fs_devs->devices_kobj = kobject_create_and_add("devices", + &fs_devs->fsid_kobj); + if (!fs_devs->devices_kobj) { + btrfs_err(fs_devs->fs_info, + "failed to init sysfs device interface"); + kobject_put(&fs_devs->fsid_kobj); + return -ENOMEM; + } + return 0; } @@ -1111,8 +1396,26 @@ int btrfs_sysfs_add_mounted(struct btrfs_fs_info *fs_info) goto failure; #ifdef CONFIG_BTRFS_DEBUG - error = sysfs_create_group(fsid_kobj, - &btrfs_debug_feature_attr_group); + fs_info->debug_kobj = kobject_create_and_add("debug", fsid_kobj); + if (!fs_info->debug_kobj) { + error = -ENOMEM; + goto failure; + } + + error = sysfs_create_files(fs_info->debug_kobj, btrfs_debug_mount_attrs); + if (error) + goto failure; + + /* Discard directory */ + fs_info->discard_debug_kobj = kobject_create_and_add("discard", + fs_info->debug_kobj); + if (!fs_info->discard_debug_kobj) { + error = -ENOMEM; + goto failure; + } + + error = sysfs_create_files(fs_info->discard_debug_kobj, + discard_debug_attrs); if (error) goto failure; #endif @@ -1209,6 +1512,9 @@ void __cold btrfs_exit_sysfs(void) sysfs_unmerge_group(&btrfs_kset->kobj, &btrfs_static_feature_attr_group); sysfs_remove_group(&btrfs_kset->kobj, &btrfs_feature_attr_group); +#ifdef CONFIG_BTRFS_DEBUG + sysfs_remove_group(&btrfs_kset->kobj, &btrfs_debug_feature_attr_group); +#endif kset_unregister(btrfs_kset); } diff --git a/fs/btrfs/sysfs.h b/fs/btrfs/sysfs.h index e10c3adfc30f..c68582add92e 100644 --- a/fs/btrfs/sysfs.h +++ b/fs/btrfs/sysfs.h @@ -18,9 +18,7 @@ int btrfs_sysfs_add_device_link(struct btrfs_fs_devices *fs_devices, struct btrfs_device *one_device); int btrfs_sysfs_rm_device_link(struct btrfs_fs_devices *fs_devices, struct btrfs_device *one_device); -int btrfs_sysfs_add_fsid(struct btrfs_fs_devices *fs_devs, - struct kobject *parent); -int btrfs_sysfs_add_device(struct btrfs_fs_devices *fs_devs); +int btrfs_sysfs_add_fsid(struct btrfs_fs_devices *fs_devs); void btrfs_sysfs_remove_fsid(struct btrfs_fs_devices *fs_devs); void btrfs_sysfs_update_sprout_fsid(struct btrfs_fs_devices *fs_devices, const u8 *fsid); @@ -36,5 +34,6 @@ void btrfs_sysfs_add_block_group_type(struct btrfs_block_group *cache); int btrfs_sysfs_add_space_info_type(struct btrfs_fs_info *fs_info, struct btrfs_space_info *space_info); void btrfs_sysfs_remove_space_info(struct btrfs_space_info *space_info); +void btrfs_sysfs_update_devid(struct btrfs_device *device); #endif diff --git a/fs/btrfs/tests/btrfs-tests.c b/fs/btrfs/tests/btrfs-tests.c index a7aca4141788..c12b91ff5f56 100644 --- a/fs/btrfs/tests/btrfs-tests.c +++ b/fs/btrfs/tests/btrfs-tests.c @@ -86,6 +86,27 @@ static void btrfs_destroy_test_fs(void) unregister_filesystem(&test_type); } +struct btrfs_device *btrfs_alloc_dummy_device(struct btrfs_fs_info *fs_info) +{ + struct btrfs_device *dev; + + dev = kzalloc(sizeof(*dev), GFP_KERNEL); + if (!dev) + return ERR_PTR(-ENOMEM); + + extent_io_tree_init(NULL, &dev->alloc_state, 0, NULL); + INIT_LIST_HEAD(&dev->dev_list); + list_add(&dev->dev_list, &fs_info->fs_devices->devices); + + return dev; +} + +static void btrfs_free_dummy_device(struct btrfs_device *dev) +{ + extent_io_tree_release(&dev->alloc_state); + kfree(dev); +} + struct btrfs_fs_info *btrfs_alloc_dummy_fs_info(u32 nodesize, u32 sectorsize) { struct btrfs_fs_info *fs_info = kzalloc(sizeof(struct btrfs_fs_info), @@ -132,12 +153,14 @@ struct btrfs_fs_info *btrfs_alloc_dummy_fs_info(u32 nodesize, u32 sectorsize) INIT_LIST_HEAD(&fs_info->dirty_qgroups); INIT_LIST_HEAD(&fs_info->dead_roots); INIT_LIST_HEAD(&fs_info->tree_mod_seq_list); + INIT_LIST_HEAD(&fs_info->fs_devices->devices); INIT_RADIX_TREE(&fs_info->buffer_radix, GFP_ATOMIC); INIT_RADIX_TREE(&fs_info->fs_roots_radix, GFP_ATOMIC); extent_io_tree_init(fs_info, &fs_info->freed_extents[0], IO_TREE_FS_INFO_FREED_EXTENTS0, NULL); extent_io_tree_init(fs_info, &fs_info->freed_extents[1], IO_TREE_FS_INFO_FREED_EXTENTS1, NULL); + extent_map_tree_init(&fs_info->mapping_tree); fs_info->pinned_extents = &fs_info->freed_extents[0]; set_bit(BTRFS_FS_STATE_DUMMY_FS_INFO, &fs_info->fs_state); @@ -150,6 +173,7 @@ void btrfs_free_dummy_fs_info(struct btrfs_fs_info *fs_info) { struct radix_tree_iter iter; void **slot; + struct btrfs_device *dev, *tmp; if (!fs_info) return; @@ -180,6 +204,11 @@ void btrfs_free_dummy_fs_info(struct btrfs_fs_info *fs_info) } spin_unlock(&fs_info->buffer_lock); + btrfs_mapping_tree_free(&fs_info->mapping_tree); + list_for_each_entry_safe(dev, tmp, &fs_info->fs_devices->devices, + dev_list) { + btrfs_free_dummy_device(dev); + } btrfs_free_qgroup_config(fs_info); btrfs_free_fs_roots(fs_info); cleanup_srcu_struct(&fs_info->subvol_srcu); diff --git a/fs/btrfs/tests/btrfs-tests.h b/fs/btrfs/tests/btrfs-tests.h index 9e52527357d8..7a2d7ffbe30e 100644 --- a/fs/btrfs/tests/btrfs-tests.h +++ b/fs/btrfs/tests/btrfs-tests.h @@ -46,6 +46,7 @@ btrfs_alloc_dummy_block_group(struct btrfs_fs_info *fs_info, unsigned long lengt void btrfs_free_dummy_block_group(struct btrfs_block_group *cache); void btrfs_init_dummy_trans(struct btrfs_trans_handle *trans, struct btrfs_fs_info *fs_info); +struct btrfs_device *btrfs_alloc_dummy_device(struct btrfs_fs_info *fs_info); #else static inline int btrfs_run_sanity_tests(void) { diff --git a/fs/btrfs/tests/extent-map-tests.c b/fs/btrfs/tests/extent-map-tests.c index 4a7f796c9900..57379e96ccc9 100644 --- a/fs/btrfs/tests/extent-map-tests.c +++ b/fs/btrfs/tests/extent-map-tests.c @@ -6,6 +6,9 @@ #include <linux/types.h> #include "btrfs-tests.h" #include "../ctree.h" +#include "../volumes.h" +#include "../disk-io.h" +#include "../block-group.h" static void free_extent_map_tree(struct extent_map_tree *em_tree) { @@ -437,11 +440,153 @@ static int test_case_4(struct btrfs_fs_info *fs_info, return ret; } +struct rmap_test_vector { + u64 raid_type; + u64 physical_start; + u64 data_stripe_size; + u64 num_data_stripes; + u64 num_stripes; + /* Assume we won't have more than 5 physical stripes */ + u64 data_stripe_phys_start[5]; + bool expected_mapped_addr; + /* Physical to logical addresses */ + u64 mapped_logical[5]; +}; + +static int test_rmap_block(struct btrfs_fs_info *fs_info, + struct rmap_test_vector *test) +{ + struct extent_map *em; + struct map_lookup *map = NULL; + u64 *logical = NULL; + int i, out_ndaddrs, out_stripe_len; + int ret; + + em = alloc_extent_map(); + if (!em) { + test_std_err(TEST_ALLOC_EXTENT_MAP); + return -ENOMEM; + } + + map = kmalloc(map_lookup_size(test->num_stripes), GFP_KERNEL); + if (!map) { + kfree(em); + test_std_err(TEST_ALLOC_EXTENT_MAP); + return -ENOMEM; + } + + set_bit(EXTENT_FLAG_FS_MAPPING, &em->flags); + /* Start at 4GiB logical address */ + em->start = SZ_4G; + em->len = test->data_stripe_size * test->num_data_stripes; + em->block_len = em->len; + em->orig_block_len = test->data_stripe_size; + em->map_lookup = map; + + map->num_stripes = test->num_stripes; + map->stripe_len = BTRFS_STRIPE_LEN; + map->type = test->raid_type; + + for (i = 0; i < map->num_stripes; i++) { + struct btrfs_device *dev = btrfs_alloc_dummy_device(fs_info); + + if (IS_ERR(dev)) { + test_err("cannot allocate device"); + ret = PTR_ERR(dev); + goto out; + } + map->stripes[i].dev = dev; + map->stripes[i].physical = test->data_stripe_phys_start[i]; + } + + write_lock(&fs_info->mapping_tree.lock); + ret = add_extent_mapping(&fs_info->mapping_tree, em, 0); + write_unlock(&fs_info->mapping_tree.lock); + if (ret) { + test_err("error adding block group mapping to mapping tree"); + goto out_free; + } + + ret = btrfs_rmap_block(fs_info, em->start, btrfs_sb_offset(1), + &logical, &out_ndaddrs, &out_stripe_len); + if (ret || (out_ndaddrs == 0 && test->expected_mapped_addr)) { + test_err("didn't rmap anything but expected %d", + test->expected_mapped_addr); + goto out; + } + + if (out_stripe_len != BTRFS_STRIPE_LEN) { + test_err("calculated stripe length doesn't match"); + goto out; + } + + if (out_ndaddrs != test->expected_mapped_addr) { + for (i = 0; i < out_ndaddrs; i++) + test_msg("mapped %llu", logical[i]); + test_err("unexpected number of mapped addresses: %d", out_ndaddrs); + goto out; + } + + for (i = 0; i < out_ndaddrs; i++) { + if (logical[i] != test->mapped_logical[i]) { + test_err("unexpected logical address mapped"); + goto out; + } + } + + ret = 0; +out: + write_lock(&fs_info->mapping_tree.lock); + remove_extent_mapping(&fs_info->mapping_tree, em); + write_unlock(&fs_info->mapping_tree.lock); + /* For us */ + free_extent_map(em); +out_free: + /* For the tree */ + free_extent_map(em); + kfree(logical); + return ret; +} + int btrfs_test_extent_map(void) { struct btrfs_fs_info *fs_info = NULL; struct extent_map_tree *em_tree; - int ret = 0; + int ret = 0, i; + struct rmap_test_vector rmap_tests[] = { + { + /* + * Test a chunk with 2 data stripes one of which + * interesects the physical address of the super block + * is correctly recognised. + */ + .raid_type = BTRFS_BLOCK_GROUP_RAID1, + .physical_start = SZ_64M - SZ_4M, + .data_stripe_size = SZ_256M, + .num_data_stripes = 2, + .num_stripes = 2, + .data_stripe_phys_start = + {SZ_64M - SZ_4M, SZ_64M - SZ_4M + SZ_256M}, + .expected_mapped_addr = true, + .mapped_logical= {SZ_4G + SZ_4M} + }, + { + /* + * Test that out-of-range physical addresses are + * ignored + */ + + /* SINGLE chunk type */ + .raid_type = 0, + .physical_start = SZ_4G, + .data_stripe_size = SZ_256M, + .num_data_stripes = 1, + .num_stripes = 1, + .data_stripe_phys_start = {SZ_256M}, + .expected_mapped_addr = false, + .mapped_logical = {0} + } + }; test_msg("running extent_map tests"); @@ -474,6 +619,13 @@ int btrfs_test_extent_map(void) goto out; ret = test_case_4(fs_info, em_tree); + test_msg("running rmap tests"); + for (i = 0; i < ARRAY_SIZE(rmap_tests); i++) { + ret = test_rmap_block(fs_info, &rmap_tests[i]); + if (ret) + goto out; + } + out: kfree(em_tree); btrfs_free_dummy_fs_info(fs_info); diff --git a/fs/btrfs/tests/free-space-tree-tests.c b/fs/btrfs/tests/free-space-tree-tests.c index 1a846bf6e197..914eea5ba6a7 100644 --- a/fs/btrfs/tests/free-space-tree-tests.c +++ b/fs/btrfs/tests/free-space-tree-tests.c @@ -452,9 +452,9 @@ static int run_test(test_func_t test_func, int bitmaps, u32 sectorsize, root->fs_info->tree_root = root; root->node = alloc_test_extent_buffer(root->fs_info, nodesize); - if (!root->node) { + if (IS_ERR(root->node)) { test_std_err(TEST_ALLOC_EXTENT_BUFFER); - ret = -ENOMEM; + ret = PTR_ERR(root->node); goto out; } btrfs_set_header_level(root->node, 0); diff --git a/fs/btrfs/tests/inode-tests.c b/fs/btrfs/tests/inode-tests.c index 09ecf7dc7b08..24a8c714f56c 100644 --- a/fs/btrfs/tests/inode-tests.c +++ b/fs/btrfs/tests/inode-tests.c @@ -263,7 +263,7 @@ static noinline int test_btrfs_get_extent(u32 sectorsize, u32 nodesize) /* First with no extents */ BTRFS_I(inode)->root = root; - em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, 0, sectorsize, 0); + em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, 0, sectorsize); if (IS_ERR(em)) { em = NULL; test_err("got an error when we shouldn't have"); @@ -283,7 +283,7 @@ static noinline int test_btrfs_get_extent(u32 sectorsize, u32 nodesize) */ setup_file_extents(root, sectorsize); - em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, 0, (u64)-1, 0); + em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, 0, (u64)-1); if (IS_ERR(em)) { test_err("got an error when we shouldn't have"); goto out; @@ -305,7 +305,7 @@ static noinline int test_btrfs_get_extent(u32 sectorsize, u32 nodesize) offset = em->start + em->len; free_extent_map(em); - em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize, 0); + em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize); if (IS_ERR(em)) { test_err("got an error when we shouldn't have"); goto out; @@ -333,7 +333,7 @@ static noinline int test_btrfs_get_extent(u32 sectorsize, u32 nodesize) offset = em->start + em->len; free_extent_map(em); - em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize, 0); + em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize); if (IS_ERR(em)) { test_err("got an error when we shouldn't have"); goto out; @@ -356,7 +356,7 @@ static noinline int test_btrfs_get_extent(u32 sectorsize, u32 nodesize) free_extent_map(em); /* Regular extent */ - em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize, 0); + em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize); if (IS_ERR(em)) { test_err("got an error when we shouldn't have"); goto out; @@ -384,7 +384,7 @@ static noinline int test_btrfs_get_extent(u32 sectorsize, u32 nodesize) free_extent_map(em); /* The next 3 are split extents */ - em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize, 0); + em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize); if (IS_ERR(em)) { test_err("got an error when we shouldn't have"); goto out; @@ -413,7 +413,7 @@ static noinline int test_btrfs_get_extent(u32 sectorsize, u32 nodesize) offset = em->start + em->len; free_extent_map(em); - em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize, 0); + em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize); if (IS_ERR(em)) { test_err("got an error when we shouldn't have"); goto out; @@ -435,7 +435,7 @@ static noinline int test_btrfs_get_extent(u32 sectorsize, u32 nodesize) offset = em->start + em->len; free_extent_map(em); - em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize, 0); + em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize); if (IS_ERR(em)) { test_err("got an error when we shouldn't have"); goto out; @@ -469,7 +469,7 @@ static noinline int test_btrfs_get_extent(u32 sectorsize, u32 nodesize) free_extent_map(em); /* Prealloc extent */ - em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize, 0); + em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize); if (IS_ERR(em)) { test_err("got an error when we shouldn't have"); goto out; @@ -498,7 +498,7 @@ static noinline int test_btrfs_get_extent(u32 sectorsize, u32 nodesize) free_extent_map(em); /* The next 3 are a half written prealloc extent */ - em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize, 0); + em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize); if (IS_ERR(em)) { test_err("got an error when we shouldn't have"); goto out; @@ -528,7 +528,7 @@ static noinline int test_btrfs_get_extent(u32 sectorsize, u32 nodesize) offset = em->start + em->len; free_extent_map(em); - em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize, 0); + em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize); if (IS_ERR(em)) { test_err("got an error when we shouldn't have"); goto out; @@ -561,7 +561,7 @@ static noinline int test_btrfs_get_extent(u32 sectorsize, u32 nodesize) offset = em->start + em->len; free_extent_map(em); - em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize, 0); + em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize); if (IS_ERR(em)) { test_err("got an error when we shouldn't have"); goto out; @@ -596,7 +596,7 @@ static noinline int test_btrfs_get_extent(u32 sectorsize, u32 nodesize) free_extent_map(em); /* Now for the compressed extent */ - em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize, 0); + em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize); if (IS_ERR(em)) { test_err("got an error when we shouldn't have"); goto out; @@ -630,7 +630,7 @@ static noinline int test_btrfs_get_extent(u32 sectorsize, u32 nodesize) free_extent_map(em); /* Split compressed extent */ - em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize, 0); + em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize); if (IS_ERR(em)) { test_err("got an error when we shouldn't have"); goto out; @@ -665,7 +665,7 @@ static noinline int test_btrfs_get_extent(u32 sectorsize, u32 nodesize) offset = em->start + em->len; free_extent_map(em); - em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize, 0); + em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize); if (IS_ERR(em)) { test_err("got an error when we shouldn't have"); goto out; @@ -692,7 +692,7 @@ static noinline int test_btrfs_get_extent(u32 sectorsize, u32 nodesize) offset = em->start + em->len; free_extent_map(em); - em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize, 0); + em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize); if (IS_ERR(em)) { test_err("got an error when we shouldn't have"); goto out; @@ -727,8 +727,7 @@ static noinline int test_btrfs_get_extent(u32 sectorsize, u32 nodesize) free_extent_map(em); /* A hole between regular extents but no hole extent */ - em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset + 6, - sectorsize, 0); + em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset + 6, sectorsize); if (IS_ERR(em)) { test_err("got an error when we shouldn't have"); goto out; @@ -755,7 +754,7 @@ static noinline int test_btrfs_get_extent(u32 sectorsize, u32 nodesize) offset = em->start + em->len; free_extent_map(em); - em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, SZ_4M, 0); + em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, SZ_4M); if (IS_ERR(em)) { test_err("got an error when we shouldn't have"); goto out; @@ -788,7 +787,7 @@ static noinline int test_btrfs_get_extent(u32 sectorsize, u32 nodesize) offset = em->start + em->len; free_extent_map(em); - em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize, 0); + em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, offset, sectorsize); if (IS_ERR(em)) { test_err("got an error when we shouldn't have"); goto out; @@ -872,7 +871,7 @@ static int test_hole_first(u32 sectorsize, u32 nodesize) insert_inode_item_key(root); insert_extent(root, sectorsize, sectorsize, sectorsize, 0, sectorsize, sectorsize, BTRFS_FILE_EXTENT_REG, 0, 1); - em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, 0, 2 * sectorsize, 0); + em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, 0, 2 * sectorsize); if (IS_ERR(em)) { test_err("got an error when we shouldn't have"); goto out; @@ -894,8 +893,7 @@ static int test_hole_first(u32 sectorsize, u32 nodesize) } free_extent_map(em); - em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, sectorsize, - 2 * sectorsize, 0); + em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, sectorsize, 2 * sectorsize); if (IS_ERR(em)) { test_err("got an error when we shouldn't have"); goto out; diff --git a/fs/btrfs/tests/qgroup-tests.c b/fs/btrfs/tests/qgroup-tests.c index 09aaca1efd62..ac035a6fa003 100644 --- a/fs/btrfs/tests/qgroup-tests.c +++ b/fs/btrfs/tests/qgroup-tests.c @@ -484,9 +484,9 @@ int btrfs_test_qgroups(u32 sectorsize, u32 nodesize) * *cough*backref walking code*cough* */ root->node = alloc_test_extent_buffer(root->fs_info, nodesize); - if (!root->node) { + if (IS_ERR(root->node)) { test_err("couldn't allocate dummy buffer"); - ret = -ENOMEM; + ret = PTR_ERR(root->node); goto out; } btrfs_set_header_level(root->node, 0); diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index cfc08ef9b876..33dcc88b428a 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -147,13 +147,14 @@ void btrfs_put_transaction(struct btrfs_transaction *transaction) } } -static noinline void switch_commit_roots(struct btrfs_transaction *trans) +static noinline void switch_commit_roots(struct btrfs_trans_handle *trans) { + struct btrfs_transaction *cur_trans = trans->transaction; struct btrfs_fs_info *fs_info = trans->fs_info; struct btrfs_root *root, *tmp; down_write(&fs_info->commit_root_sem); - list_for_each_entry_safe(root, tmp, &trans->switch_commits, + list_for_each_entry_safe(root, tmp, &cur_trans->switch_commits, dirty_list) { list_del_init(&root->dirty_list); free_extent_buffer(root->commit_root); @@ -165,16 +166,17 @@ static noinline void switch_commit_roots(struct btrfs_transaction *trans) } /* We can free old roots now. */ - spin_lock(&trans->dropped_roots_lock); - while (!list_empty(&trans->dropped_roots)) { - root = list_first_entry(&trans->dropped_roots, + spin_lock(&cur_trans->dropped_roots_lock); + while (!list_empty(&cur_trans->dropped_roots)) { + root = list_first_entry(&cur_trans->dropped_roots, struct btrfs_root, root_list); list_del_init(&root->root_list); - spin_unlock(&trans->dropped_roots_lock); + spin_unlock(&cur_trans->dropped_roots_lock); + btrfs_free_log(trans, root); btrfs_drop_and_free_fs_root(fs_info, root); - spin_lock(&trans->dropped_roots_lock); + spin_lock(&cur_trans->dropped_roots_lock); } - spin_unlock(&trans->dropped_roots_lock); + spin_unlock(&cur_trans->dropped_roots_lock); up_write(&fs_info->commit_root_sem); } @@ -1421,7 +1423,7 @@ static int qgroup_account_snapshot(struct btrfs_trans_handle *trans, ret = commit_cowonly_roots(trans); if (ret) goto out; - switch_commit_roots(trans->transaction); + switch_commit_roots(trans); ret = btrfs_write_and_wait_transaction(trans); if (ret) btrfs_handle_fs_error(fs_info, ret, @@ -2013,6 +2015,14 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans) ASSERT(refcount_read(&trans->use_count) == 1); + /* + * Some places just start a transaction to commit it. We need to make + * sure that if this commit fails that the abort code actually marks the + * transaction as failed, so set trans->dirty to make the abort code do + * the right thing. + */ + trans->dirty = true; + /* Stop the commit early if ->aborted is set */ if (unlikely(READ_ONCE(cur_trans->aborted))) { ret = cur_trans->aborted; @@ -2301,7 +2311,7 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans) list_add_tail(&fs_info->chunk_root->dirty_list, &cur_trans->switch_commits); - switch_commit_roots(cur_trans); + switch_commit_roots(trans); ASSERT(list_empty(&cur_trans->dirty_bgs)); ASSERT(list_empty(&cur_trans->io_bgs)); diff --git a/fs/btrfs/tree-checker.c b/fs/btrfs/tree-checker.c index 493d4d9e0f79..a92f8a6dd192 100644 --- a/fs/btrfs/tree-checker.c +++ b/fs/btrfs/tree-checker.c @@ -227,7 +227,7 @@ static int check_extent_data_item(struct extent_buffer *leaf, */ if (item_size < BTRFS_FILE_EXTENT_INLINE_DATA_START) { file_extent_err(leaf, slot, - "invalid item size, have %u expect [%lu, %u)", + "invalid item size, have %u expect [%zu, %u)", item_size, BTRFS_FILE_EXTENT_INLINE_DATA_START, SZ_4K); return -EUCLEAN; @@ -332,7 +332,7 @@ static int check_extent_data_item(struct extent_buffer *leaf, } static int check_csum_item(struct extent_buffer *leaf, struct btrfs_key *key, - int slot) + int slot, struct btrfs_key *prev_key) { struct btrfs_fs_info *fs_info = leaf->fs_info; u32 sectorsize = fs_info->sectorsize; @@ -356,6 +356,118 @@ static int check_csum_item(struct extent_buffer *leaf, struct btrfs_key *key, btrfs_item_size_nr(leaf, slot), csumsize); return -EUCLEAN; } + if (slot > 0 && prev_key->type == BTRFS_EXTENT_CSUM_KEY) { + u64 prev_csum_end; + u32 prev_item_size; + + prev_item_size = btrfs_item_size_nr(leaf, slot - 1); + prev_csum_end = (prev_item_size / csumsize) * sectorsize; + prev_csum_end += prev_key->offset; + if (prev_csum_end > key->offset) { + generic_err(leaf, slot - 1, +"csum end range (%llu) goes beyond the start range (%llu) of the next csum item", + prev_csum_end, key->offset); + return -EUCLEAN; + } + } + return 0; +} + +/* Inode item error output has the same format as dir_item_err() */ +#define inode_item_err(eb, slot, fmt, ...) \ + dir_item_err(eb, slot, fmt, __VA_ARGS__) + +static int check_inode_key(struct extent_buffer *leaf, struct btrfs_key *key, + int slot) +{ + struct btrfs_key item_key; + bool is_inode_item; + + btrfs_item_key_to_cpu(leaf, &item_key, slot); + is_inode_item = (item_key.type == BTRFS_INODE_ITEM_KEY); + + /* For XATTR_ITEM, location key should be all 0 */ + if (item_key.type == BTRFS_XATTR_ITEM_KEY) { + if (key->type != 0 || key->objectid != 0 || key->offset != 0) + return -EUCLEAN; + return 0; + } + + if ((key->objectid < BTRFS_FIRST_FREE_OBJECTID || + key->objectid > BTRFS_LAST_FREE_OBJECTID) && + key->objectid != BTRFS_ROOT_TREE_DIR_OBJECTID && + key->objectid != BTRFS_FREE_INO_OBJECTID) { + if (is_inode_item) { + generic_err(leaf, slot, + "invalid key objectid: has %llu expect %llu or [%llu, %llu] or %llu", + key->objectid, BTRFS_ROOT_TREE_DIR_OBJECTID, + BTRFS_FIRST_FREE_OBJECTID, + BTRFS_LAST_FREE_OBJECTID, + BTRFS_FREE_INO_OBJECTID); + } else { + dir_item_err(leaf, slot, +"invalid location key objectid: has %llu expect %llu or [%llu, %llu] or %llu", + key->objectid, BTRFS_ROOT_TREE_DIR_OBJECTID, + BTRFS_FIRST_FREE_OBJECTID, + BTRFS_LAST_FREE_OBJECTID, + BTRFS_FREE_INO_OBJECTID); + } + return -EUCLEAN; + } + if (key->offset != 0) { + if (is_inode_item) + inode_item_err(leaf, slot, + "invalid key offset: has %llu expect 0", + key->offset); + else + dir_item_err(leaf, slot, + "invalid location key offset:has %llu expect 0", + key->offset); + return -EUCLEAN; + } + return 0; +} + +static int check_root_key(struct extent_buffer *leaf, struct btrfs_key *key, + int slot) +{ + struct btrfs_key item_key; + bool is_root_item; + + btrfs_item_key_to_cpu(leaf, &item_key, slot); + is_root_item = (item_key.type == BTRFS_ROOT_ITEM_KEY); + + /* No such tree id */ + if (key->objectid == 0) { + if (is_root_item) + generic_err(leaf, slot, "invalid root id 0"); + else + dir_item_err(leaf, slot, + "invalid location key root id 0"); + return -EUCLEAN; + } + + /* DIR_ITEM/INDEX/INODE_REF is not allowed to point to non-fs trees */ + if (!is_fstree(key->objectid) && !is_root_item) { + dir_item_err(leaf, slot, + "invalid location key objectid, have %llu expect [%llu, %llu]", + key->objectid, BTRFS_FIRST_FREE_OBJECTID, + BTRFS_LAST_FREE_OBJECTID); + return -EUCLEAN; + } + + /* + * ROOT_ITEM with non-zero offset means this is a snapshot, created at + * @offset transid. + * Furthermore, for location key in DIR_ITEM, its offset is always -1. + * + * So here we only check offset for reloc tree whose key->offset must + * be a valid tree. + */ + if (key->objectid == BTRFS_TREE_RELOC_OBJECTID && key->offset == 0) { + generic_err(leaf, slot, "invalid root id 0 for reloc tree"); + return -EUCLEAN; + } return 0; } @@ -372,12 +484,14 @@ static int check_dir_item(struct extent_buffer *leaf, return -EUCLEAN; di = btrfs_item_ptr(leaf, slot, struct btrfs_dir_item); while (cur < item_size) { + struct btrfs_key location_key; u32 name_len; u32 data_len; u32 max_name_len; u32 total_size; u32 name_hash; u8 dir_type; + int ret; /* header itself should not cross item boundary */ if (cur + sizeof(*di) > item_size) { @@ -387,6 +501,25 @@ static int check_dir_item(struct extent_buffer *leaf, return -EUCLEAN; } + /* Location key check */ + btrfs_dir_item_key_to_cpu(leaf, di, &location_key); + if (location_key.type == BTRFS_ROOT_ITEM_KEY) { + ret = check_root_key(leaf, &location_key, slot); + if (ret < 0) + return ret; + } else if (location_key.type == BTRFS_INODE_ITEM_KEY || + location_key.type == 0) { + ret = check_inode_key(leaf, &location_key, slot); + if (ret < 0) + return ret; + } else { + dir_item_err(leaf, slot, + "invalid location key type, have %u, expect %u or %u", + location_key.type, BTRFS_ROOT_ITEM_KEY, + BTRFS_INODE_ITEM_KEY); + return -EUCLEAN; + } + /* dir type check */ dir_type = btrfs_dir_type(leaf, di); if (dir_type >= BTRFS_FT_MAX) { @@ -724,6 +857,44 @@ int btrfs_check_chunk_valid(struct extent_buffer *leaf, return 0; } +/* + * Enhanced version of chunk item checker. + * + * The common btrfs_check_chunk_valid() doesn't check item size since it needs + * to work on super block sys_chunk_array which doesn't have full item ptr. + */ +static int check_leaf_chunk_item(struct extent_buffer *leaf, + struct btrfs_chunk *chunk, + struct btrfs_key *key, int slot) +{ + int num_stripes; + + if (btrfs_item_size_nr(leaf, slot) < sizeof(struct btrfs_chunk)) { + chunk_err(leaf, chunk, key->offset, + "invalid chunk item size: have %u expect [%zu, %u)", + btrfs_item_size_nr(leaf, slot), + sizeof(struct btrfs_chunk), + BTRFS_LEAF_DATA_SIZE(leaf->fs_info)); + return -EUCLEAN; + } + + num_stripes = btrfs_chunk_num_stripes(leaf, chunk); + /* Let btrfs_check_chunk_valid() handle this error type */ + if (num_stripes == 0) + goto out; + + if (btrfs_chunk_item_size(num_stripes) != + btrfs_item_size_nr(leaf, slot)) { + chunk_err(leaf, chunk, key->offset, + "invalid chunk item size: have %u expect %lu", + btrfs_item_size_nr(leaf, slot), + btrfs_chunk_item_size(num_stripes)); + return -EUCLEAN; + } +out: + return btrfs_check_chunk_valid(leaf, chunk, key->offset); +} + __printf(3, 4) __cold static void dev_item_err(const struct extent_buffer *eb, int slot, @@ -787,7 +958,7 @@ static int check_dev_item(struct extent_buffer *leaf, } /* Inode item error output has the same format as dir_item_err() */ -#define inode_item_err(fs_info, eb, slot, fmt, ...) \ +#define inode_item_err(eb, slot, fmt, ...) \ dir_item_err(eb, slot, fmt, __VA_ARGS__) static int check_inode_item(struct extent_buffer *leaf, @@ -798,30 +969,17 @@ static int check_inode_item(struct extent_buffer *leaf, u64 super_gen = btrfs_super_generation(fs_info->super_copy); u32 valid_mask = (S_IFMT | S_ISUID | S_ISGID | S_ISVTX | 0777); u32 mode; + int ret; + + ret = check_inode_key(leaf, key, slot); + if (ret < 0) + return ret; - if ((key->objectid < BTRFS_FIRST_FREE_OBJECTID || - key->objectid > BTRFS_LAST_FREE_OBJECTID) && - key->objectid != BTRFS_ROOT_TREE_DIR_OBJECTID && - key->objectid != BTRFS_FREE_INO_OBJECTID) { - generic_err(leaf, slot, - "invalid key objectid: has %llu expect %llu or [%llu, %llu] or %llu", - key->objectid, BTRFS_ROOT_TREE_DIR_OBJECTID, - BTRFS_FIRST_FREE_OBJECTID, - BTRFS_LAST_FREE_OBJECTID, - BTRFS_FREE_INO_OBJECTID); - return -EUCLEAN; - } - if (key->offset != 0) { - inode_item_err(fs_info, leaf, slot, - "invalid key offset: has %llu expect 0", - key->offset); - return -EUCLEAN; - } iitem = btrfs_item_ptr(leaf, slot, struct btrfs_inode_item); /* Here we use super block generation + 1 to handle log tree */ if (btrfs_inode_generation(leaf, iitem) > super_gen + 1) { - inode_item_err(fs_info, leaf, slot, + inode_item_err(leaf, slot, "invalid inode generation: has %llu expect (0, %llu]", btrfs_inode_generation(leaf, iitem), super_gen + 1); @@ -829,7 +987,7 @@ static int check_inode_item(struct extent_buffer *leaf, } /* Note for ROOT_TREE_DIR_ITEM, mkfs could set its transid 0 */ if (btrfs_inode_transid(leaf, iitem) > super_gen + 1) { - inode_item_err(fs_info, leaf, slot, + inode_item_err(leaf, slot, "invalid inode generation: has %llu expect [0, %llu]", btrfs_inode_transid(leaf, iitem), super_gen + 1); return -EUCLEAN; @@ -842,7 +1000,7 @@ static int check_inode_item(struct extent_buffer *leaf, */ mode = btrfs_inode_mode(leaf, iitem); if (mode & ~valid_mask) { - inode_item_err(fs_info, leaf, slot, + inode_item_err(leaf, slot, "unknown mode bit detected: 0x%x", mode & ~valid_mask); return -EUCLEAN; @@ -855,20 +1013,20 @@ static int check_inode_item(struct extent_buffer *leaf, */ if (!has_single_bit_set(mode & S_IFMT)) { if (!S_ISLNK(mode) && !S_ISBLK(mode) && !S_ISSOCK(mode)) { - inode_item_err(fs_info, leaf, slot, + inode_item_err(leaf, slot, "invalid mode: has 0%o expect valid S_IF* bit(s)", mode & S_IFMT); return -EUCLEAN; } } if (S_ISDIR(mode) && btrfs_inode_nlink(leaf, iitem) > 1) { - inode_item_err(fs_info, leaf, slot, + inode_item_err(leaf, slot, "invalid nlink: has %u expect no more than 1 for dir", btrfs_inode_nlink(leaf, iitem)); return -EUCLEAN; } if (btrfs_inode_flags(leaf, iitem) & ~BTRFS_INODE_FLAG_MASK) { - inode_item_err(fs_info, leaf, slot, + inode_item_err(leaf, slot, "unknown flags detected: 0x%llx", btrfs_inode_flags(leaf, iitem) & ~BTRFS_INODE_FLAG_MASK); @@ -884,22 +1042,11 @@ static int check_root_item(struct extent_buffer *leaf, struct btrfs_key *key, struct btrfs_root_item ri; const u64 valid_root_flags = BTRFS_ROOT_SUBVOL_RDONLY | BTRFS_ROOT_SUBVOL_DEAD; + int ret; - /* No such tree id */ - if (key->objectid == 0) { - generic_err(leaf, slot, "invalid root id 0"); - return -EUCLEAN; - } - - /* - * Some older kernel may create ROOT_ITEM with non-zero offset, so here - * we only check offset for reloc tree whose key->offset must be a - * valid tree. - */ - if (key->objectid == BTRFS_TREE_RELOC_OBJECTID && key->offset == 0) { - generic_err(leaf, slot, "invalid root id 0 for reloc tree"); - return -EUCLEAN; - } + ret = check_root_key(leaf, key, slot); + if (ret < 0) + return ret; if (btrfs_item_size_nr(leaf, slot) != sizeof(ri)) { generic_err(leaf, slot, @@ -1288,8 +1435,8 @@ static int check_extent_data_ref(struct extent_buffer *leaf, return 0; } -#define inode_ref_err(fs_info, eb, slot, fmt, args...) \ - inode_item_err(fs_info, eb, slot, fmt, ##args) +#define inode_ref_err(eb, slot, fmt, args...) \ + inode_item_err(eb, slot, fmt, ##args) static int check_inode_ref(struct extent_buffer *leaf, struct btrfs_key *key, struct btrfs_key *prev_key, int slot) @@ -1302,7 +1449,7 @@ static int check_inode_ref(struct extent_buffer *leaf, return -EUCLEAN; /* namelen can't be 0, so item_size == sizeof() is also invalid */ if (btrfs_item_size_nr(leaf, slot) <= sizeof(*iref)) { - inode_ref_err(fs_info, leaf, slot, + inode_ref_err(leaf, slot, "invalid item size, have %u expect (%zu, %u)", btrfs_item_size_nr(leaf, slot), sizeof(*iref), BTRFS_LEAF_DATA_SIZE(leaf->fs_info)); @@ -1315,7 +1462,7 @@ static int check_inode_ref(struct extent_buffer *leaf, u16 namelen; if (ptr + sizeof(iref) > end) { - inode_ref_err(fs_info, leaf, slot, + inode_ref_err(leaf, slot, "inode ref overflow, ptr %lu end %lu inode_ref_size %zu", ptr, end, sizeof(iref)); return -EUCLEAN; @@ -1324,7 +1471,7 @@ static int check_inode_ref(struct extent_buffer *leaf, iref = (struct btrfs_inode_ref *)ptr; namelen = btrfs_inode_ref_name_len(leaf, iref); if (ptr + sizeof(*iref) + namelen > end) { - inode_ref_err(fs_info, leaf, slot, + inode_ref_err(leaf, slot, "inode ref overflow, ptr %lu end %lu namelen %u", ptr, end, namelen); return -EUCLEAN; @@ -1355,7 +1502,7 @@ static int check_leaf_item(struct extent_buffer *leaf, ret = check_extent_data_item(leaf, key, slot, prev_key); break; case BTRFS_EXTENT_CSUM_KEY: - ret = check_csum_item(leaf, key, slot); + ret = check_csum_item(leaf, key, slot, prev_key); break; case BTRFS_DIR_ITEM_KEY: case BTRFS_DIR_INDEX_KEY: @@ -1370,7 +1517,7 @@ static int check_leaf_item(struct extent_buffer *leaf, break; case BTRFS_CHUNK_ITEM_KEY: chunk = btrfs_item_ptr(leaf, slot, struct btrfs_chunk); - ret = btrfs_check_chunk_valid(leaf, chunk, key->offset); + ret = check_leaf_chunk_item(leaf, chunk, key, slot); break; case BTRFS_DEV_ITEM_KEY: ret = check_dev_item(leaf, key, slot); diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index 6f757361db53..7dd7552f53a4 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -808,7 +808,8 @@ static noinline int replay_one_extent(struct btrfs_trans_handle *trans, struct btrfs_ordered_sum, list); if (!ret) - ret = btrfs_del_csums(trans, fs_info, + ret = btrfs_del_csums(trans, + fs_info->csum_root, sums->bytenr, sums->len); if (!ret) @@ -2673,14 +2674,9 @@ static noinline int walk_down_log_tree(struct btrfs_trans_handle *trans, u32 blocksize; int ret = 0; - WARN_ON(*level < 0); - WARN_ON(*level >= BTRFS_MAX_LEVEL); - while (*level > 0) { struct btrfs_key first_key; - WARN_ON(*level < 0); - WARN_ON(*level >= BTRFS_MAX_LEVEL); cur = path->nodes[*level]; WARN_ON(btrfs_header_level(cur) != *level); @@ -2731,9 +2727,8 @@ static noinline int walk_down_log_tree(struct btrfs_trans_handle *trans, WARN_ON(root_owner != BTRFS_TREE_LOG_OBJECTID); - ret = btrfs_free_and_pin_reserved_extent( - fs_info, bytenr, - blocksize); + ret = btrfs_pin_reserved_extent(fs_info, + bytenr, blocksize); if (ret) { free_extent_buffer(next); return ret; @@ -2748,7 +2743,6 @@ static noinline int walk_down_log_tree(struct btrfs_trans_handle *trans, return ret; } - WARN_ON(*level <= 0); if (path->nodes[*level-1]) free_extent_buffer(path->nodes[*level-1]); path->nodes[*level-1] = next; @@ -2756,9 +2750,6 @@ static noinline int walk_down_log_tree(struct btrfs_trans_handle *trans, path->slots[*level] = 0; cond_resched(); } - WARN_ON(*level < 0); - WARN_ON(*level >= BTRFS_MAX_LEVEL); - path->slots[*level] = btrfs_header_nritems(path->nodes[*level]); cond_resched(); @@ -2814,8 +2805,7 @@ static noinline int walk_up_log_tree(struct btrfs_trans_handle *trans, } WARN_ON(root_owner != BTRFS_TREE_LOG_OBJECTID); - ret = btrfs_free_and_pin_reserved_extent( - fs_info, + ret = btrfs_pin_reserved_extent(fs_info, path->nodes[*level]->start, path->nodes[*level]->len); if (ret) @@ -2895,10 +2885,8 @@ static int walk_log_tree(struct btrfs_trans_handle *trans, clear_extent_buffer_dirty(next); } - WARN_ON(log->root_key.objectid != - BTRFS_TREE_LOG_OBJECTID); - ret = btrfs_free_and_pin_reserved_extent(fs_info, - next->start, next->len); + ret = btrfs_pin_reserved_extent(fs_info, next->start, + next->len); if (ret) goto out; } @@ -3909,10 +3897,32 @@ static int log_inode_item(struct btrfs_trans_handle *trans, return 0; } +static int log_csums(struct btrfs_trans_handle *trans, + struct btrfs_root *log_root, + struct btrfs_ordered_sum *sums) +{ + int ret; + + /* + * Due to extent cloning, we might have logged a csum item that covers a + * subrange of a cloned extent, and later we can end up logging a csum + * item for a larger subrange of the same extent or the entire range. + * This would leave csum items in the log tree that cover the same range + * and break the searches for checksums in the log tree, resulting in + * some checksums missing in the fs/subvolume tree. So just delete (or + * trim and adjust) any existing csum items in the log for this range. + */ + ret = btrfs_del_csums(trans, log_root, sums->bytenr, sums->len); + if (ret) + return ret; + + return btrfs_csum_file_blocks(trans, log_root, sums); +} + static noinline int copy_items(struct btrfs_trans_handle *trans, struct btrfs_inode *inode, struct btrfs_path *dst_path, - struct btrfs_path *src_path, u64 *last_extent, + struct btrfs_path *src_path, int start_slot, int nr, int inode_only, u64 logged_isize) { @@ -3923,7 +3933,6 @@ static noinline int copy_items(struct btrfs_trans_handle *trans, struct btrfs_file_extent_item *extent; struct btrfs_inode_item *inode_item; struct extent_buffer *src = src_path->nodes[0]; - struct btrfs_key first_key, last_key, key; int ret; struct btrfs_key *ins_keys; u32 *ins_sizes; @@ -3931,9 +3940,6 @@ static noinline int copy_items(struct btrfs_trans_handle *trans, int i; struct list_head ordered_sums; int skip_csum = inode->flags & BTRFS_INODE_NODATASUM; - bool has_extents = false; - bool need_find_last_extent = true; - bool done = false; INIT_LIST_HEAD(&ordered_sums); @@ -3942,8 +3948,6 @@ static noinline int copy_items(struct btrfs_trans_handle *trans, if (!ins_data) return -ENOMEM; - first_key.objectid = (u64)-1; - ins_sizes = (u32 *)ins_data; ins_keys = (struct btrfs_key *)(ins_data + nr * sizeof(u32)); @@ -3964,9 +3968,6 @@ static noinline int copy_items(struct btrfs_trans_handle *trans, src_offset = btrfs_item_ptr_offset(src, start_slot + i); - if (i == nr - 1) - last_key = ins_keys[i]; - if (ins_keys[i].type == BTRFS_INODE_ITEM_KEY) { inode_item = btrfs_item_ptr(dst_path->nodes[0], dst_path->slots[0], @@ -3980,20 +3981,6 @@ static noinline int copy_items(struct btrfs_trans_handle *trans, src_offset, ins_sizes[i]); } - /* - * We set need_find_last_extent here in case we know we were - * processing other items and then walk into the first extent in - * the inode. If we don't hit an extent then nothing changes, - * we'll do the last search the next time around. - */ - if (ins_keys[i].type == BTRFS_EXTENT_DATA_KEY) { - has_extents = true; - if (first_key.objectid == (u64)-1) - first_key = ins_keys[i]; - } else { - need_find_last_extent = false; - } - /* take a reference on file data extents so that truncates * or deletes of this inode don't have to relog the inode * again @@ -4054,172 +4041,11 @@ static noinline int copy_items(struct btrfs_trans_handle *trans, struct btrfs_ordered_sum, list); if (!ret) - ret = btrfs_csum_file_blocks(trans, log, sums); + ret = log_csums(trans, log, sums); list_del(&sums->list); kfree(sums); } - if (!has_extents) - return ret; - - if (need_find_last_extent && *last_extent == first_key.offset) { - /* - * We don't have any leafs between our current one and the one - * we processed before that can have file extent items for our - * inode (and have a generation number smaller than our current - * transaction id). - */ - need_find_last_extent = false; - } - - /* - * Because we use btrfs_search_forward we could skip leaves that were - * not modified and then assume *last_extent is valid when it really - * isn't. So back up to the previous leaf and read the end of the last - * extent before we go and fill in holes. - */ - if (need_find_last_extent) { - u64 len; - - ret = btrfs_prev_leaf(inode->root, src_path); - if (ret < 0) - return ret; - if (ret) - goto fill_holes; - if (src_path->slots[0]) - src_path->slots[0]--; - src = src_path->nodes[0]; - btrfs_item_key_to_cpu(src, &key, src_path->slots[0]); - if (key.objectid != btrfs_ino(inode) || - key.type != BTRFS_EXTENT_DATA_KEY) - goto fill_holes; - extent = btrfs_item_ptr(src, src_path->slots[0], - struct btrfs_file_extent_item); - if (btrfs_file_extent_type(src, extent) == - BTRFS_FILE_EXTENT_INLINE) { - len = btrfs_file_extent_ram_bytes(src, extent); - *last_extent = ALIGN(key.offset + len, - fs_info->sectorsize); - } else { - len = btrfs_file_extent_num_bytes(src, extent); - *last_extent = key.offset + len; - } - } -fill_holes: - /* So we did prev_leaf, now we need to move to the next leaf, but a few - * things could have happened - * - * 1) A merge could have happened, so we could currently be on a leaf - * that holds what we were copying in the first place. - * 2) A split could have happened, and now not all of the items we want - * are on the same leaf. - * - * So we need to adjust how we search for holes, we need to drop the - * path and re-search for the first extent key we found, and then walk - * forward until we hit the last one we copied. - */ - if (need_find_last_extent) { - /* btrfs_prev_leaf could return 1 without releasing the path */ - btrfs_release_path(src_path); - ret = btrfs_search_slot(NULL, inode->root, &first_key, - src_path, 0, 0); - if (ret < 0) - return ret; - ASSERT(ret == 0); - src = src_path->nodes[0]; - i = src_path->slots[0]; - } else { - i = start_slot; - } - - /* - * Ok so here we need to go through and fill in any holes we may have - * to make sure that holes are punched for those areas in case they had - * extents previously. - */ - while (!done) { - u64 offset, len; - u64 extent_end; - - if (i >= btrfs_header_nritems(src_path->nodes[0])) { - ret = btrfs_next_leaf(inode->root, src_path); - if (ret < 0) - return ret; - ASSERT(ret == 0); - src = src_path->nodes[0]; - i = 0; - need_find_last_extent = true; - } - - btrfs_item_key_to_cpu(src, &key, i); - if (!btrfs_comp_cpu_keys(&key, &last_key)) - done = true; - if (key.objectid != btrfs_ino(inode) || - key.type != BTRFS_EXTENT_DATA_KEY) { - i++; - continue; - } - extent = btrfs_item_ptr(src, i, struct btrfs_file_extent_item); - if (btrfs_file_extent_type(src, extent) == - BTRFS_FILE_EXTENT_INLINE) { - len = btrfs_file_extent_ram_bytes(src, extent); - extent_end = ALIGN(key.offset + len, - fs_info->sectorsize); - } else { - len = btrfs_file_extent_num_bytes(src, extent); - extent_end = key.offset + len; - } - i++; - - if (*last_extent == key.offset) { - *last_extent = extent_end; - continue; - } - offset = *last_extent; - len = key.offset - *last_extent; - ret = btrfs_insert_file_extent(trans, log, btrfs_ino(inode), - offset, 0, 0, len, 0, len, 0, 0, 0); - if (ret) - break; - *last_extent = extent_end; - } - - /* - * Check if there is a hole between the last extent found in our leaf - * and the first extent in the next leaf. If there is one, we need to - * log an explicit hole so that at replay time we can punch the hole. - */ - if (ret == 0 && - key.objectid == btrfs_ino(inode) && - key.type == BTRFS_EXTENT_DATA_KEY && - i == btrfs_header_nritems(src_path->nodes[0])) { - ret = btrfs_next_leaf(inode->root, src_path); - need_find_last_extent = true; - if (ret > 0) { - ret = 0; - } else if (ret == 0) { - btrfs_item_key_to_cpu(src_path->nodes[0], &key, - src_path->slots[0]); - if (key.objectid == btrfs_ino(inode) && - key.type == BTRFS_EXTENT_DATA_KEY && - *last_extent < key.offset) { - const u64 len = key.offset - *last_extent; - - ret = btrfs_insert_file_extent(trans, log, - btrfs_ino(inode), - *last_extent, 0, - 0, len, 0, len, - 0, 0, 0); - *last_extent += len; - } - } - } - /* - * Need to let the callers know we dropped the path so they should - * re-search. - */ - if (!ret && need_find_last_extent) - ret = 1; return ret; } @@ -4274,7 +4100,7 @@ static int log_extent_csums(struct btrfs_trans_handle *trans, struct btrfs_ordered_sum, list); if (!ret) - ret = btrfs_csum_file_blocks(trans, log_root, sums); + ret = log_csums(trans, log_root, sums); list_del(&sums->list); kfree(sums); } @@ -4384,7 +4210,7 @@ static int btrfs_log_prealloc_extents(struct btrfs_trans_handle *trans, const u64 i_size = i_size_read(&inode->vfs_inode); const u64 ino = btrfs_ino(inode); struct btrfs_path *dst_path = NULL; - u64 last_extent = (u64)-1; + bool dropped_extents = false; int ins_nr = 0; int start_slot; int ret; @@ -4406,8 +4232,7 @@ static int btrfs_log_prealloc_extents(struct btrfs_trans_handle *trans, if (slot >= btrfs_header_nritems(leaf)) { if (ins_nr > 0) { ret = copy_items(trans, inode, dst_path, path, - &last_extent, start_slot, - ins_nr, 1, 0); + start_slot, ins_nr, 1, 0); if (ret < 0) goto out; ins_nr = 0; @@ -4431,8 +4256,7 @@ static int btrfs_log_prealloc_extents(struct btrfs_trans_handle *trans, path->slots[0]++; continue; } - if (last_extent == (u64)-1) { - last_extent = key.offset; + if (!dropped_extents) { /* * Avoid logging extent items logged in past fsync calls * and leading to duplicate keys in the log tree. @@ -4446,6 +4270,7 @@ static int btrfs_log_prealloc_extents(struct btrfs_trans_handle *trans, } while (ret == -EAGAIN); if (ret) goto out; + dropped_extents = true; } if (ins_nr == 0) start_slot = slot; @@ -4460,7 +4285,7 @@ static int btrfs_log_prealloc_extents(struct btrfs_trans_handle *trans, } } if (ins_nr > 0) { - ret = copy_items(trans, inode, dst_path, path, &last_extent, + ret = copy_items(trans, inode, dst_path, path, start_slot, ins_nr, 1, 0); if (ret > 0) ret = 0; @@ -4647,13 +4472,8 @@ static int btrfs_log_all_xattrs(struct btrfs_trans_handle *trans, if (slot >= nritems) { if (ins_nr > 0) { - u64 last_extent = 0; - ret = copy_items(trans, inode, dst_path, path, - &last_extent, start_slot, - ins_nr, 1, 0); - /* can't be 1, extent items aren't processed */ - ASSERT(ret <= 0); + start_slot, ins_nr, 1, 0); if (ret < 0) return ret; ins_nr = 0; @@ -4677,13 +4497,8 @@ static int btrfs_log_all_xattrs(struct btrfs_trans_handle *trans, cond_resched(); } if (ins_nr > 0) { - u64 last_extent = 0; - ret = copy_items(trans, inode, dst_path, path, - &last_extent, start_slot, - ins_nr, 1, 0); - /* can't be 1, extent items aren't processed */ - ASSERT(ret <= 0); + start_slot, ins_nr, 1, 0); if (ret < 0) return ret; } @@ -4692,100 +4507,119 @@ static int btrfs_log_all_xattrs(struct btrfs_trans_handle *trans, } /* - * If the no holes feature is enabled we need to make sure any hole between the - * last extent and the i_size of our inode is explicitly marked in the log. This - * is to make sure that doing something like: - * - * 1) create file with 128Kb of data - * 2) truncate file to 64Kb - * 3) truncate file to 256Kb - * 4) fsync file - * 5) <crash/power failure> - * 6) mount fs and trigger log replay - * - * Will give us a file with a size of 256Kb, the first 64Kb of data match what - * the file had in its first 64Kb of data at step 1 and the last 192Kb of the - * file correspond to a hole. The presence of explicit holes in a log tree is - * what guarantees that log replay will remove/adjust file extent items in the - * fs/subvol tree. - * - * Here we do not need to care about holes between extents, that is already done - * by copy_items(). We also only need to do this in the full sync path, where we - * lookup for extents from the fs/subvol tree only. In the fast path case, we - * lookup the list of modified extent maps and if any represents a hole, we - * insert a corresponding extent representing a hole in the log tree. + * When using the NO_HOLES feature if we punched a hole that causes the + * deletion of entire leafs or all the extent items of the first leaf (the one + * that contains the inode item and references) we may end up not processing + * any extents, because there are no leafs with a generation matching the + * current transaction that have extent items for our inode. So we need to find + * if any holes exist and then log them. We also need to log holes after any + * truncate operation that changes the inode's size. */ -static int btrfs_log_trailing_hole(struct btrfs_trans_handle *trans, - struct btrfs_root *root, - struct btrfs_inode *inode, - struct btrfs_path *path) +static int btrfs_log_holes(struct btrfs_trans_handle *trans, + struct btrfs_root *root, + struct btrfs_inode *inode, + struct btrfs_path *path) { struct btrfs_fs_info *fs_info = root->fs_info; - int ret; struct btrfs_key key; - u64 hole_start; - u64 hole_size; - struct extent_buffer *leaf; - struct btrfs_root *log = root->log_root; const u64 ino = btrfs_ino(inode); const u64 i_size = i_size_read(&inode->vfs_inode); + u64 prev_extent_end = 0; + int ret; - if (!btrfs_fs_incompat(fs_info, NO_HOLES)) + if (!btrfs_fs_incompat(fs_info, NO_HOLES) || i_size == 0) return 0; key.objectid = ino; key.type = BTRFS_EXTENT_DATA_KEY; - key.offset = (u64)-1; + key.offset = 0; ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); - ASSERT(ret != 0); if (ret < 0) return ret; - ASSERT(path->slots[0] > 0); - path->slots[0]--; - leaf = path->nodes[0]; - btrfs_item_key_to_cpu(leaf, &key, path->slots[0]); - - if (key.objectid != ino || key.type != BTRFS_EXTENT_DATA_KEY) { - /* inode does not have any extents */ - hole_start = 0; - hole_size = i_size; - } else { + while (true) { struct btrfs_file_extent_item *extent; + struct extent_buffer *leaf = path->nodes[0]; u64 len; - /* - * If there's an extent beyond i_size, an explicit hole was - * already inserted by copy_items(). - */ - if (key.offset >= i_size) - return 0; + if (path->slots[0] >= btrfs_header_nritems(path->nodes[0])) { + ret = btrfs_next_leaf(root, path); + if (ret < 0) + return ret; + if (ret > 0) { + ret = 0; + break; + } + leaf = path->nodes[0]; + } + + btrfs_item_key_to_cpu(leaf, &key, path->slots[0]); + if (key.objectid != ino || key.type != BTRFS_EXTENT_DATA_KEY) + break; + + /* We have a hole, log it. */ + if (prev_extent_end < key.offset) { + const u64 hole_len = key.offset - prev_extent_end; + + /* + * Release the path to avoid deadlocks with other code + * paths that search the root while holding locks on + * leafs from the log root. + */ + btrfs_release_path(path); + ret = btrfs_insert_file_extent(trans, root->log_root, + ino, prev_extent_end, 0, + 0, hole_len, 0, hole_len, + 0, 0, 0); + if (ret < 0) + return ret; + + /* + * Search for the same key again in the root. Since it's + * an extent item and we are holding the inode lock, the + * key must still exist. If it doesn't just emit warning + * and return an error to fall back to a transaction + * commit. + */ + ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); + if (ret < 0) + return ret; + if (WARN_ON(ret > 0)) + return -ENOENT; + leaf = path->nodes[0]; + } extent = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_file_extent_item); - if (btrfs_file_extent_type(leaf, extent) == - BTRFS_FILE_EXTENT_INLINE) - return 0; + BTRFS_FILE_EXTENT_INLINE) { + len = btrfs_file_extent_ram_bytes(leaf, extent); + prev_extent_end = ALIGN(key.offset + len, + fs_info->sectorsize); + } else { + len = btrfs_file_extent_num_bytes(leaf, extent); + prev_extent_end = key.offset + len; + } - len = btrfs_file_extent_num_bytes(leaf, extent); - /* Last extent goes beyond i_size, no need to log a hole. */ - if (key.offset + len > i_size) - return 0; - hole_start = key.offset + len; - hole_size = i_size - hole_start; + path->slots[0]++; + cond_resched(); } - btrfs_release_path(path); - /* Last extent ends at i_size. */ - if (hole_size == 0) - return 0; + if (prev_extent_end < i_size) { + u64 hole_len; - hole_size = ALIGN(hole_size, fs_info->sectorsize); - ret = btrfs_insert_file_extent(trans, log, ino, hole_start, 0, 0, - hole_size, 0, hole_size, 0, 0, 0); - return ret; + btrfs_release_path(path); + hole_len = ALIGN(i_size - prev_extent_end, fs_info->sectorsize); + ret = btrfs_insert_file_extent(trans, root->log_root, + ino, prev_extent_end, 0, 0, + hole_len, 0, hole_len, + 0, 0, 0); + if (ret < 0) + return ret; + } + + return 0; } /* @@ -4989,6 +4823,50 @@ static int log_conflicting_inodes(struct btrfs_trans_handle *trans, continue; } /* + * If the inode was already logged skip it - otherwise we can + * hit an infinite loop. Example: + * + * From the commit root (previous transaction) we have the + * following inodes: + * + * inode 257 a directory + * inode 258 with references "zz" and "zz_link" on inode 257 + * inode 259 with reference "a" on inode 257 + * + * And in the current (uncommitted) transaction we have: + * + * inode 257 a directory, unchanged + * inode 258 with references "a" and "a2" on inode 257 + * inode 259 with reference "zz_link" on inode 257 + * inode 261 with reference "zz" on inode 257 + * + * When logging inode 261 the following infinite loop could + * happen if we don't skip already logged inodes: + * + * - we detect inode 258 as a conflicting inode, with inode 261 + * on reference "zz", and log it; + * + * - we detect inode 259 as a conflicting inode, with inode 258 + * on reference "a", and log it; + * + * - we detect inode 258 as a conflicting inode, with inode 259 + * on reference "zz_link", and log it - again! After this we + * repeat the above steps forever. + */ + spin_lock(&BTRFS_I(inode)->lock); + /* + * Check the inode's logged_trans only instead of + * btrfs_inode_in_log(). This is because the last_log_commit of + * the inode is not updated when we only log that it exists and + * and it has the full sync bit set (see btrfs_log_inode()). + */ + if (BTRFS_I(inode)->logged_trans == trans->transid) { + spin_unlock(&BTRFS_I(inode)->lock); + btrfs_add_delayed_iput(inode); + continue; + } + spin_unlock(&BTRFS_I(inode)->lock); + /* * We are safe logging the other inode without acquiring its * lock as long as we log with the LOG_INODE_EXISTS mode. We * are safe against concurrent renames of the other inode as @@ -5087,7 +4965,6 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans, struct btrfs_key min_key; struct btrfs_key max_key; struct btrfs_root *log = root->log_root; - u64 last_extent = 0; int err = 0; int ret; int nritems; @@ -5265,7 +5142,7 @@ again: ins_start_slot = path->slots[0]; } ret = copy_items(trans, inode, dst_path, path, - &last_extent, ins_start_slot, + ins_start_slot, ins_nr, inode_only, logged_isize); if (ret < 0) { @@ -5288,17 +5165,13 @@ again: if (ins_nr == 0) goto next_slot; ret = copy_items(trans, inode, dst_path, path, - &last_extent, ins_start_slot, + ins_start_slot, ins_nr, inode_only, logged_isize); if (ret < 0) { err = ret; goto out_unlock; } ins_nr = 0; - if (ret) { - btrfs_release_path(path); - continue; - } goto next_slot; } @@ -5311,18 +5184,13 @@ again: goto next_slot; } - ret = copy_items(trans, inode, dst_path, path, &last_extent, + ret = copy_items(trans, inode, dst_path, path, ins_start_slot, ins_nr, inode_only, logged_isize); if (ret < 0) { err = ret; goto out_unlock; } - if (ret) { - ins_nr = 0; - btrfs_release_path(path); - continue; - } ins_nr = 1; ins_start_slot = path->slots[0]; next_slot: @@ -5336,13 +5204,12 @@ next_slot: } if (ins_nr) { ret = copy_items(trans, inode, dst_path, path, - &last_extent, ins_start_slot, + ins_start_slot, ins_nr, inode_only, logged_isize); if (ret < 0) { err = ret; goto out_unlock; } - ret = 0; ins_nr = 0; } btrfs_release_path(path); @@ -5357,14 +5224,13 @@ next_key: } } if (ins_nr) { - ret = copy_items(trans, inode, dst_path, path, &last_extent, + ret = copy_items(trans, inode, dst_path, path, ins_start_slot, ins_nr, inode_only, logged_isize); if (ret < 0) { err = ret; goto out_unlock; } - ret = 0; ins_nr = 0; } @@ -5377,7 +5243,7 @@ next_key: if (max_key.type >= BTRFS_EXTENT_DATA_KEY && !fast_search) { btrfs_release_path(path); btrfs_release_path(dst_path); - err = btrfs_log_trailing_hole(trans, root, inode, path); + err = btrfs_log_holes(trans, root, inode, path); if (err) goto out_unlock; } @@ -6294,9 +6160,28 @@ again: wc.replay_dest = btrfs_read_fs_root_no_name(fs_info, &tmp_key); if (IS_ERR(wc.replay_dest)) { ret = PTR_ERR(wc.replay_dest); + + /* + * We didn't find the subvol, likely because it was + * deleted. This is ok, simply skip this log and go to + * the next one. + * + * We need to exclude the root because we can't have + * other log replays overwriting this log as we'll read + * it back in a few more times. This will keep our + * block from being modified, and we'll just bail for + * each subsequent pass. + */ + if (ret == -ENOENT) + ret = btrfs_pin_extent_for_log_replay(fs_info, + log->node->start, + log->node->len); free_extent_buffer(log->node); free_extent_buffer(log->commit_root); kfree(log); + + if (!ret) + goto next; btrfs_handle_fs_error(fs_info, ret, "Couldn't read target root for tree log recovery."); goto error; @@ -6328,7 +6213,6 @@ again: &root->highest_objectid); } - key.offset = found_key.offset - 1; wc.replay_dest->log_root = NULL; free_extent_buffer(log->node); free_extent_buffer(log->commit_root); @@ -6336,9 +6220,10 @@ again: if (ret) goto error; - +next: if (found_key.offset == 0) break; + key.offset = found_key.offset - 1; } btrfs_release_path(path); diff --git a/fs/btrfs/uuid-tree.c b/fs/btrfs/uuid-tree.c index 91caab63bdf5..76b84f2397b1 100644 --- a/fs/btrfs/uuid-tree.c +++ b/fs/btrfs/uuid-tree.c @@ -324,6 +324,8 @@ again_search_slot: } if (ret < 0 && ret != -ENOENT) goto out; + key.offset++; + goto again_search_slot; } item_size -= sizeof(subid_le); offset += sizeof(subid_le); diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index d8e5560db285..9cfc668f91f4 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -30,6 +30,7 @@ #include "tree-checker.h" #include "space-info.h" #include "block-group.h" +#include "discard.h" const struct btrfs_raid_attr btrfs_raid_array[BTRFS_NR_RAID_TYPES] = { [BTRFS_RAID_RAID10] = { @@ -61,11 +62,12 @@ const struct btrfs_raid_attr btrfs_raid_array[BTRFS_NR_RAID_TYPES] = { [BTRFS_RAID_RAID1C3] = { .sub_stripes = 1, .dev_stripes = 1, - .devs_max = 0, + .devs_max = 3, .devs_min = 3, .tolerated_failures = 2, .devs_increment = 3, .ncopies = 3, + .nparity = 0, .raid_name = "raid1c3", .bg_flag = BTRFS_BLOCK_GROUP_RAID1C3, .mindev_error = BTRFS_ERROR_DEV_RAID1C3_MIN_NOT_MET, @@ -73,11 +75,12 @@ const struct btrfs_raid_attr btrfs_raid_array[BTRFS_NR_RAID_TYPES] = { [BTRFS_RAID_RAID1C4] = { .sub_stripes = 1, .dev_stripes = 1, - .devs_max = 0, + .devs_max = 4, .devs_min = 4, .tolerated_failures = 3, .devs_increment = 4, .ncopies = 4, + .nparity = 0, .raid_name = "raid1c4", .bg_flag = BTRFS_BLOCK_GROUP_RAID1C4, .mindev_error = BTRFS_ERROR_DEV_RAID1C4_MIN_NOT_MET, @@ -438,39 +441,6 @@ static noinline struct btrfs_fs_devices *find_fsid( ASSERT(fsid); - if (metadata_fsid) { - /* - * Handle scanned device having completed its fsid change but - * belonging to a fs_devices that was created by first scanning - * a device which didn't have its fsid/metadata_uuid changed - * at all and the CHANGING_FSID_V2 flag set. - */ - list_for_each_entry(fs_devices, &fs_uuids, fs_list) { - if (fs_devices->fsid_change && - memcmp(metadata_fsid, fs_devices->fsid, - BTRFS_FSID_SIZE) == 0 && - memcmp(fs_devices->fsid, fs_devices->metadata_uuid, - BTRFS_FSID_SIZE) == 0) { - return fs_devices; - } - } - /* - * Handle scanned device having completed its fsid change but - * belonging to a fs_devices that was created by a device that - * has an outdated pair of fsid/metadata_uuid and - * CHANGING_FSID_V2 flag set. - */ - list_for_each_entry(fs_devices, &fs_uuids, fs_list) { - if (fs_devices->fsid_change && - memcmp(fs_devices->metadata_uuid, - fs_devices->fsid, BTRFS_FSID_SIZE) != 0 && - memcmp(metadata_fsid, fs_devices->metadata_uuid, - BTRFS_FSID_SIZE) == 0) { - return fs_devices; - } - } - } - /* Handle non-split brain cases */ list_for_each_entry(fs_devices, &fs_uuids, fs_list) { if (metadata_fsid) { @@ -486,6 +456,47 @@ static noinline struct btrfs_fs_devices *find_fsid( return NULL; } +static struct btrfs_fs_devices *find_fsid_with_metadata_uuid( + struct btrfs_super_block *disk_super) +{ + + struct btrfs_fs_devices *fs_devices; + + /* + * Handle scanned device having completed its fsid change but + * belonging to a fs_devices that was created by first scanning + * a device which didn't have its fsid/metadata_uuid changed + * at all and the CHANGING_FSID_V2 flag set. + */ + list_for_each_entry(fs_devices, &fs_uuids, fs_list) { + if (fs_devices->fsid_change && + memcmp(disk_super->metadata_uuid, fs_devices->fsid, + BTRFS_FSID_SIZE) == 0 && + memcmp(fs_devices->fsid, fs_devices->metadata_uuid, + BTRFS_FSID_SIZE) == 0) { + return fs_devices; + } + } + /* + * Handle scanned device having completed its fsid change but + * belonging to a fs_devices that was created by a device that + * has an outdated pair of fsid/metadata_uuid and + * CHANGING_FSID_V2 flag set. + */ + list_for_each_entry(fs_devices, &fs_uuids, fs_list) { + if (fs_devices->fsid_change && + memcmp(fs_devices->metadata_uuid, + fs_devices->fsid, BTRFS_FSID_SIZE) != 0 && + memcmp(disk_super->metadata_uuid, fs_devices->metadata_uuid, + BTRFS_FSID_SIZE) == 0) { + return fs_devices; + } + } + + return find_fsid(disk_super->fsid, disk_super->metadata_uuid); +} + + static int btrfs_get_bdev_and_sb(const char *device_path, fmode_t flags, void *holder, int flush, struct block_device **bdev, @@ -669,7 +680,9 @@ error_brelse: /* * Handle scanned device having its CHANGING_FSID_V2 flag set and the fs_devices - * being created with a disk that has already completed its fsid change. + * being created with a disk that has already completed its fsid change. Such + * disk can belong to an fs which has its FSID changed or to one which doesn't. + * Handle both cases here. */ static struct btrfs_fs_devices *find_fsid_inprogress( struct btrfs_super_block *disk_super) @@ -685,7 +698,7 @@ static struct btrfs_fs_devices *find_fsid_inprogress( } } - return NULL; + return find_fsid(disk_super->fsid, NULL); } @@ -697,17 +710,54 @@ static struct btrfs_fs_devices *find_fsid_changed( /* * Handles the case where scanned device is part of an fs that had * multiple successful changes of FSID but curently device didn't - * observe it. Meaning our fsid will be different than theirs. + * observe it. Meaning our fsid will be different than theirs. We need + * to handle two subcases : + * 1 - The fs still continues to have different METADATA/FSID uuids. + * 2 - The fs is switched back to its original FSID (METADATA/FSID + * are equal). */ list_for_each_entry(fs_devices, &fs_uuids, fs_list) { + /* Changed UUIDs */ if (memcmp(fs_devices->metadata_uuid, fs_devices->fsid, BTRFS_FSID_SIZE) != 0 && memcmp(fs_devices->metadata_uuid, disk_super->metadata_uuid, BTRFS_FSID_SIZE) == 0 && memcmp(fs_devices->fsid, disk_super->fsid, - BTRFS_FSID_SIZE) != 0) { + BTRFS_FSID_SIZE) != 0) + return fs_devices; + + /* Unchanged UUIDs */ + if (memcmp(fs_devices->metadata_uuid, fs_devices->fsid, + BTRFS_FSID_SIZE) == 0 && + memcmp(fs_devices->fsid, disk_super->metadata_uuid, + BTRFS_FSID_SIZE) == 0) + return fs_devices; + } + + return NULL; +} + +static struct btrfs_fs_devices *find_fsid_reverted_metadata( + struct btrfs_super_block *disk_super) +{ + struct btrfs_fs_devices *fs_devices; + + /* + * Handle the case where the scanned device is part of an fs whose last + * metadata UUID change reverted it to the original FSID. At the same + * time * fs_devices was first created by another constitutent device + * which didn't fully observe the operation. This results in an + * btrfs_fs_devices created with metadata/fsid different AND + * btrfs_fs_devices::fsid_change set AND the metadata_uuid of the + * fs_devices equal to the FSID of the disk. + */ + list_for_each_entry(fs_devices, &fs_uuids, fs_list) { + if (memcmp(fs_devices->fsid, fs_devices->metadata_uuid, + BTRFS_FSID_SIZE) != 0 && + memcmp(fs_devices->metadata_uuid, disk_super->fsid, + BTRFS_FSID_SIZE) == 0 && + fs_devices->fsid_change) return fs_devices; - } } return NULL; @@ -734,24 +784,16 @@ static noinline struct btrfs_device *device_list_add(const char *path, BTRFS_SUPER_FLAG_CHANGING_FSID_V2); if (fsid_change_in_progress) { - if (!has_metadata_uuid) { - /* - * When we have an image which has CHANGING_FSID_V2 set - * it might belong to either a filesystem which has - * disks with completed fsid change or it might belong - * to fs with no UUID changes in effect, handle both. - */ + if (!has_metadata_uuid) fs_devices = find_fsid_inprogress(disk_super); - if (!fs_devices) - fs_devices = find_fsid(disk_super->fsid, NULL); - } else { + else fs_devices = find_fsid_changed(disk_super); - } } else if (has_metadata_uuid) { - fs_devices = find_fsid(disk_super->fsid, - disk_super->metadata_uuid); + fs_devices = find_fsid_with_metadata_uuid(disk_super); } else { - fs_devices = find_fsid(disk_super->fsid, NULL); + fs_devices = find_fsid_reverted_metadata(disk_super); + if (!fs_devices) + fs_devices = find_fsid(disk_super->fsid, NULL); } @@ -781,12 +823,18 @@ static noinline struct btrfs_device *device_list_add(const char *path, * a device which had the CHANGING_FSID_V2 flag then replace the * metadata_uuid/fsid values of the fs_devices. */ - if (has_metadata_uuid && fs_devices->fsid_change && + if (fs_devices->fsid_change && found_transid > fs_devices->latest_generation) { memcpy(fs_devices->fsid, disk_super->fsid, BTRFS_FSID_SIZE); - memcpy(fs_devices->metadata_uuid, - disk_super->metadata_uuid, BTRFS_FSID_SIZE); + + if (has_metadata_uuid) + memcpy(fs_devices->metadata_uuid, + disk_super->metadata_uuid, + BTRFS_FSID_SIZE); + else + memcpy(fs_devices->metadata_uuid, + disk_super->fsid, BTRFS_FSID_SIZE); fs_devices->fsid_change = false; } @@ -1064,11 +1112,6 @@ static void btrfs_close_bdev(struct btrfs_device *device) static void btrfs_close_one_device(struct btrfs_device *device) { struct btrfs_fs_devices *fs_devices = device->fs_devices; - struct btrfs_device *new_device; - struct rcu_string *name; - - if (device->bdev) - fs_devices->open_devices--; if (test_bit(BTRFS_DEV_STATE_WRITEABLE, &device->dev_state) && device->devid != BTRFS_DEV_REPLACE_DEVID) { @@ -1080,23 +1123,22 @@ static void btrfs_close_one_device(struct btrfs_device *device) fs_devices->missing_devices--; btrfs_close_bdev(device); - - new_device = btrfs_alloc_device(NULL, &device->devid, - device->uuid); - BUG_ON(IS_ERR(new_device)); /* -ENOMEM */ - - /* Safe because we are under uuid_mutex */ - if (device->name) { - name = rcu_string_strdup(device->name->str, GFP_NOFS); - BUG_ON(!name); /* -ENOMEM */ - rcu_assign_pointer(new_device->name, name); + if (device->bdev) { + fs_devices->open_devices--; + device->bdev = NULL; } + clear_bit(BTRFS_DEV_STATE_WRITEABLE, &device->dev_state); - list_replace_rcu(&device->dev_list, &new_device->dev_list); - new_device->fs_devices = device->fs_devices; + device->fs_info = NULL; + atomic_set(&device->dev_stats_ccnt, 0); + extent_io_tree_release(&device->alloc_state); - synchronize_rcu(); - btrfs_free_device(device); + /* Verify the device is back in a pristine state */ + ASSERT(!test_bit(BTRFS_DEV_STATE_FLUSH_SENT, &device->dev_state)); + ASSERT(!test_bit(BTRFS_DEV_STATE_REPLACE_TGT, &device->dev_state)); + ASSERT(list_empty(&device->dev_alloc_list)); + ASSERT(list_empty(&device->post_commit_list)); + ASSERT(atomic_read(&device->reada_in_flight) == 0); } static int close_fs_devices(struct btrfs_fs_devices *fs_devices) @@ -2130,7 +2172,6 @@ void btrfs_destroy_dev_replace_tgtdev(struct btrfs_device *tgtdev) { struct btrfs_fs_devices *fs_devices = tgtdev->fs_info->fs_devices; - WARN_ON(!tgtdev); mutex_lock(&fs_devices->device_list_mutex); btrfs_sysfs_rm_device_link(fs_devices, tgtdev); @@ -2875,6 +2916,7 @@ static int btrfs_relocate_chunk(struct btrfs_fs_info *fs_info, u64 chunk_offset) { struct btrfs_root *root = fs_info->chunk_root; struct btrfs_trans_handle *trans; + struct btrfs_block_group *block_group; int ret; /* @@ -2898,6 +2940,12 @@ static int btrfs_relocate_chunk(struct btrfs_fs_info *fs_info, u64 chunk_offset) if (ret) return ret; + block_group = btrfs_lookup_block_group(fs_info, chunk_offset); + if (!block_group) + return -ENOENT; + btrfs_discard_cancel_work(&fs_info->discard_ctl, block_group); + btrfs_put_block_group(block_group); + trans = btrfs_start_trans_remove_block_group(root->fs_info, chunk_offset); if (IS_ERR(trans)) { @@ -3881,7 +3929,11 @@ int btrfs_balance(struct btrfs_fs_info *fs_info, } } - num_devices = btrfs_num_devices(fs_info); + /* + * rw_devices will not change at the moment, device add/delete/replace + * are excluded by EXCL_OP + */ + num_devices = fs_info->fs_devices->rw_devices; /* * SINGLE profile on-disk has no profile bit, but in-memory we have a @@ -6107,75 +6159,6 @@ int btrfs_map_sblock(struct btrfs_fs_info *fs_info, enum btrfs_map_op op, return __btrfs_map_block(fs_info, op, logical, length, bbio_ret, 0, 1); } -int btrfs_rmap_block(struct btrfs_fs_info *fs_info, u64 chunk_start, - u64 physical, u64 **logical, int *naddrs, int *stripe_len) -{ - struct extent_map *em; - struct map_lookup *map; - u64 *buf; - u64 bytenr; - u64 length; - u64 stripe_nr; - u64 rmap_len; - int i, j, nr = 0; - - em = btrfs_get_chunk_map(fs_info, chunk_start, 1); - if (IS_ERR(em)) - return -EIO; - - map = em->map_lookup; - length = em->len; - rmap_len = map->stripe_len; - - if (map->type & BTRFS_BLOCK_GROUP_RAID10) - length = div_u64(length, map->num_stripes / map->sub_stripes); - else if (map->type & BTRFS_BLOCK_GROUP_RAID0) - length = div_u64(length, map->num_stripes); - else if (map->type & BTRFS_BLOCK_GROUP_RAID56_MASK) { - length = div_u64(length, nr_data_stripes(map)); - rmap_len = map->stripe_len * nr_data_stripes(map); - } - - buf = kcalloc(map->num_stripes, sizeof(u64), GFP_NOFS); - BUG_ON(!buf); /* -ENOMEM */ - - for (i = 0; i < map->num_stripes; i++) { - if (map->stripes[i].physical > physical || - map->stripes[i].physical + length <= physical) - continue; - - stripe_nr = physical - map->stripes[i].physical; - stripe_nr = div64_u64(stripe_nr, map->stripe_len); - - if (map->type & BTRFS_BLOCK_GROUP_RAID10) { - stripe_nr = stripe_nr * map->num_stripes + i; - stripe_nr = div_u64(stripe_nr, map->sub_stripes); - } else if (map->type & BTRFS_BLOCK_GROUP_RAID0) { - stripe_nr = stripe_nr * map->num_stripes + i; - } /* else if RAID[56], multiply by nr_data_stripes(). - * Alternatively, just use rmap_len below instead of - * map->stripe_len */ - - bytenr = chunk_start + stripe_nr * rmap_len; - WARN_ON(nr >= map->num_stripes); - for (j = 0; j < nr; j++) { - if (buf[j] == bytenr) - break; - } - if (j == nr) { - WARN_ON(nr >= map->num_stripes); - buf[nr++] = bytenr; - } - } - - *logical = buf; - *naddrs = nr; - *stripe_len = rmap_len; - - free_extent_map(em); - return 0; -} - static inline void btrfs_end_bbio(struct btrfs_bio *bbio, struct bio *bio) { bio->bi_private = bbio->private; @@ -6476,19 +6459,14 @@ static u64 calc_stripe_length(u64 type, u64 chunk_len, int num_stripes) { int index = btrfs_bg_flags_to_raid_index(type); int ncopies = btrfs_raid_array[index].ncopies; + const int nparity = btrfs_raid_array[index].nparity; int data_stripes; - switch (type & BTRFS_BLOCK_GROUP_PROFILE_MASK) { - case BTRFS_BLOCK_GROUP_RAID5: - data_stripes = num_stripes - 1; - break; - case BTRFS_BLOCK_GROUP_RAID6: - data_stripes = num_stripes - 2; - break; - default: + if (nparity) + data_stripes = num_stripes - nparity; + else data_stripes = num_stripes / ncopies; - break; - } + return div_u64(chunk_len, data_stripes); } @@ -7327,6 +7305,8 @@ int btrfs_get_dev_stats(struct btrfs_fs_info *fs_info, else btrfs_dev_stat_set(dev, i, 0); } + btrfs_info(fs_info, "device stats zeroed by %s (%d)", + current->comm, task_pid_nr(current)); } else { for (i = 0; i < BTRFS_DEV_STAT_VALUES_MAX; i++) if (stats->nr_items > i) diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h index fc1b564b9cfe..409f4816fb89 100644 --- a/fs/btrfs/volumes.h +++ b/fs/btrfs/volumes.h @@ -120,8 +120,6 @@ struct btrfs_device { /* per-device scrub information */ struct scrub_ctx *scrub_ctx; - struct btrfs_work work; - /* readahead state */ atomic_t reada_in_flight; u64 reada_next; @@ -138,6 +136,10 @@ struct btrfs_device { atomic_t dev_stat_values[BTRFS_DEV_STAT_VALUES_MAX]; struct extent_io_tree alloc_state; + + struct completion kobj_unregister; + /* For sysfs/FSID/devinfo/devid/ */ + struct kobject devid_kobj; }; /* @@ -168,7 +170,7 @@ btrfs_device_set_##name(struct btrfs_device *dev, u64 size) \ write_seqcount_end(&dev->data_seqcount); \ preempt_enable(); \ } -#elif BITS_PER_LONG==32 && defined(CONFIG_PREEMPT) +#elif BITS_PER_LONG==32 && defined(CONFIG_PREEMPTION) #define BTRFS_DEVICE_GETSET_FUNCS(name) \ static inline u64 \ btrfs_device_get_##name(const struct btrfs_device *dev) \ @@ -255,7 +257,7 @@ struct btrfs_fs_devices { struct btrfs_fs_info *fs_info; /* sysfs kobjects */ struct kobject fsid_kobj; - struct kobject *device_dir_kobj; + struct kobject *devices_kobj; struct completion kobj_unregister; }; @@ -417,8 +419,6 @@ int btrfs_map_sblock(struct btrfs_fs_info *fs_info, enum btrfs_map_op op, struct btrfs_bio **bbio_ret); int btrfs_get_io_geometry(struct btrfs_fs_info *fs_info, enum btrfs_map_op op, u64 logical, u64 len, struct btrfs_io_geometry *io_geom); -int btrfs_rmap_block(struct btrfs_fs_info *fs_info, u64 chunk_start, - u64 physical, u64 **logical, int *naddrs, int *stripe_len); int btrfs_read_sys_array(struct btrfs_fs_info *fs_info); int btrfs_read_chunk_tree(struct btrfs_fs_info *fs_info); int btrfs_alloc_chunk(struct btrfs_trans_handle *trans, u64 type); diff --git a/fs/buffer.c b/fs/buffer.c index d8c7242426bb..b8d28370cfd7 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -1433,7 +1433,7 @@ static bool has_bh_in_lru(int cpu, void *dummy) void invalidate_bh_lrus(void) { - on_each_cpu_cond(has_bh_in_lru, invalidate_bh_lru, NULL, 1, GFP_KERNEL); + on_each_cpu_cond(has_bh_in_lru, invalidate_bh_lru, NULL, 1); } EXPORT_SYMBOL_GPL(invalidate_bh_lrus); @@ -3031,11 +3031,9 @@ static void end_bio_bh_io_sync(struct bio *bio) * errors, this only handles the "we need to be able to * do IO at the final sector" case. */ -void guard_bio_eod(int op, struct bio *bio) +void guard_bio_eod(struct bio *bio) { sector_t maxsector; - struct bio_vec *bvec = bio_last_bvec_all(bio); - unsigned truncated_bytes; struct hd_struct *part; rcu_read_lock(); @@ -3061,28 +3059,7 @@ void guard_bio_eod(int op, struct bio *bio) if (likely((bio->bi_iter.bi_size >> 9) <= maxsector)) return; - /* Uhhuh. We've got a bio that straddles the device size! */ - truncated_bytes = bio->bi_iter.bi_size - (maxsector << 9); - - /* - * The bio contains more than one segment which spans EOD, just return - * and let IO layer turn it into an EIO - */ - if (truncated_bytes > bvec->bv_len) - return; - - /* Truncate the bio.. */ - bio->bi_iter.bi_size -= truncated_bytes; - bvec->bv_len -= truncated_bytes; - - /* ..and clear the end of the buffer for reads */ - if (op == REQ_OP_READ) { - struct bio_vec bv; - - mp_bvec_last_segment(bvec, &bv); - zero_user(bv.bv_page, bv.bv_offset + bv.bv_len, - truncated_bytes); - } + bio_truncate(bio, maxsector << 9); } static int submit_bh_wbc(int op, int op_flags, struct buffer_head *bh, @@ -3118,15 +3095,15 @@ static int submit_bh_wbc(int op, int op_flags, struct buffer_head *bh, bio->bi_end_io = end_bio_bh_io_sync; bio->bi_private = bh; - /* Take care of bh's that straddle the end of the device */ - guard_bio_eod(op, bio); - if (buffer_meta(bh)) op_flags |= REQ_META; if (buffer_prio(bh)) op_flags |= REQ_PRIO; bio_set_op_attrs(bio, op, op_flags); + /* Take care of bh's that straddle the end of the device */ + guard_bio_eod(bio); + if (wbc) { wbc_init_bio(wbc, bio); wbc_account_cgroup_owner(wbc, bh->b_page, bh->b_size); diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c index f5a38910a82b..9d09bb53c1ab 100644 --- a/fs/ceph/caps.c +++ b/fs/ceph/caps.c @@ -1011,18 +1011,13 @@ static int __ceph_is_single_caps(struct ceph_inode_info *ci) return rb_first(&ci->i_caps) == rb_last(&ci->i_caps); } -static int __ceph_is_any_caps(struct ceph_inode_info *ci) -{ - return !RB_EMPTY_ROOT(&ci->i_caps); -} - int ceph_is_any_caps(struct inode *inode) { struct ceph_inode_info *ci = ceph_inode(inode); int ret; spin_lock(&ci->i_ceph_lock); - ret = __ceph_is_any_caps(ci); + ret = __ceph_is_any_real_caps(ci); spin_unlock(&ci->i_ceph_lock); return ret; @@ -1099,15 +1094,16 @@ void __ceph_remove_cap(struct ceph_cap *cap, bool queue_release) if (removed) ceph_put_cap(mdsc, cap); - /* when reconnect denied, we remove session caps forcibly, - * i_wr_ref can be non-zero. If there are ongoing write, - * keep i_snap_realm. - */ - if (!__ceph_is_any_caps(ci) && ci->i_wr_ref == 0 && ci->i_snap_realm) - drop_inode_snap_realm(ci); + if (!__ceph_is_any_real_caps(ci)) { + /* when reconnect denied, we remove session caps forcibly, + * i_wr_ref can be non-zero. If there are ongoing write, + * keep i_snap_realm. + */ + if (ci->i_wr_ref == 0 && ci->i_snap_realm) + drop_inode_snap_realm(ci); - if (!__ceph_is_any_real_caps(ci)) __cap_delay_cancel(mdsc, ci); + } } struct cap_msg_args { @@ -2764,7 +2760,19 @@ int ceph_get_caps(struct file *filp, int need, int want, if (ret == -EAGAIN) continue; if (!ret) { + struct ceph_mds_client *mdsc = fsc->mdsc; + struct cap_wait cw; DEFINE_WAIT_FUNC(wait, woken_wake_function); + + cw.ino = inode->i_ino; + cw.tgid = current->tgid; + cw.need = need; + cw.want = want; + + spin_lock(&mdsc->caps_list_lock); + list_add(&cw.list, &mdsc->cap_wait_list); + spin_unlock(&mdsc->caps_list_lock); + add_wait_queue(&ci->i_cap_wq, &wait); flags |= NON_BLOCKING; @@ -2778,6 +2786,11 @@ int ceph_get_caps(struct file *filp, int need, int want, } remove_wait_queue(&ci->i_cap_wq, &wait); + + spin_lock(&mdsc->caps_list_lock); + list_del(&cw.list); + spin_unlock(&mdsc->caps_list_lock); + if (ret == -EAGAIN) continue; } @@ -2928,7 +2941,7 @@ void ceph_put_cap_refs(struct ceph_inode_info *ci, int had) ci->i_head_snapc = NULL; } /* see comment in __ceph_remove_cap() */ - if (!__ceph_is_any_caps(ci) && ci->i_snap_realm) + if (!__ceph_is_any_real_caps(ci) && ci->i_snap_realm) drop_inode_snap_realm(ci); } spin_unlock(&ci->i_ceph_lock); diff --git a/fs/ceph/debugfs.c b/fs/ceph/debugfs.c index facb387c2735..c281f32b54f7 100644 --- a/fs/ceph/debugfs.c +++ b/fs/ceph/debugfs.c @@ -139,6 +139,7 @@ static int caps_show(struct seq_file *s, void *p) struct ceph_fs_client *fsc = s->private; struct ceph_mds_client *mdsc = fsc->mdsc; int total, avail, used, reserved, min, i; + struct cap_wait *cw; ceph_reservation_status(fsc, &total, &avail, &used, &reserved, &min); seq_printf(s, "total\t\t%d\n" @@ -166,6 +167,18 @@ static int caps_show(struct seq_file *s, void *p) } mutex_unlock(&mdsc->mutex); + seq_printf(s, "\n\nWaiters:\n--------\n"); + seq_printf(s, "tgid ino need want\n"); + seq_printf(s, "-----------------------------------------------------\n"); + + spin_lock(&mdsc->caps_list_lock); + list_for_each_entry(cw, &mdsc->cap_wait_list, list) { + seq_printf(s, "%-13d0x%-17lx%-17s%-17s\n", cw->tgid, cw->ino, + ceph_cap_string(cw->need), + ceph_cap_string(cw->want)); + } + spin_unlock(&mdsc->caps_list_lock); + return 0; } diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c index 068b029cf073..145d46ba25ae 100644 --- a/fs/ceph/mds_client.c +++ b/fs/ceph/mds_client.c @@ -708,8 +708,10 @@ void ceph_mdsc_release_request(struct kref *kref) /* avoid calling iput_final() in mds dispatch threads */ ceph_async_iput(req->r_inode); } - if (req->r_parent) + if (req->r_parent) { ceph_put_cap_refs(ceph_inode(req->r_parent), CEPH_CAP_PIN); + ceph_async_iput(req->r_parent); + } ceph_async_iput(req->r_target_inode); if (req->r_dentry) dput(req->r_dentry); @@ -2015,7 +2017,7 @@ void ceph_reclaim_caps_nr(struct ceph_mds_client *mdsc, int nr) if (!nr) return; val = atomic_add_return(nr, &mdsc->cap_reclaim_pending); - if (!(val % CEPH_CAPS_PER_RELEASE)) { + if ((val % CEPH_CAPS_PER_RELEASE) < nr) { atomic_set(&mdsc->cap_reclaim_pending, 0); ceph_queue_cap_reclaim_work(mdsc); } @@ -2032,12 +2034,13 @@ int ceph_alloc_readdir_reply_buffer(struct ceph_mds_request *req, struct ceph_mds_reply_info_parsed *rinfo = &req->r_reply_info; struct ceph_mount_options *opt = req->r_mdsc->fsc->mount_options; size_t size = sizeof(struct ceph_mds_reply_dir_entry); - int order, num_entries; + unsigned int num_entries; + int order; spin_lock(&ci->i_ceph_lock); num_entries = ci->i_files + ci->i_subdirs; spin_unlock(&ci->i_ceph_lock); - num_entries = max(num_entries, 1); + num_entries = max(num_entries, 1U); num_entries = min(num_entries, opt->max_readdir); order = get_order(size * num_entries); @@ -2675,8 +2678,10 @@ int ceph_mdsc_submit_request(struct ceph_mds_client *mdsc, struct inode *dir, /* take CAP_PIN refs for r_inode, r_parent, r_old_dentry */ if (req->r_inode) ceph_get_cap_refs(ceph_inode(req->r_inode), CEPH_CAP_PIN); - if (req->r_parent) + if (req->r_parent) { ceph_get_cap_refs(ceph_inode(req->r_parent), CEPH_CAP_PIN); + ihold(req->r_parent); + } if (req->r_old_dentry_dir) ceph_get_cap_refs(ceph_inode(req->r_old_dentry_dir), CEPH_CAP_PIN); @@ -4168,6 +4173,7 @@ int ceph_mdsc_init(struct ceph_fs_client *fsc) INIT_DELAYED_WORK(&mdsc->delayed_work, delayed_work); mdsc->last_renew_caps = jiffies; INIT_LIST_HEAD(&mdsc->cap_delay_list); + INIT_LIST_HEAD(&mdsc->cap_wait_list); spin_lock_init(&mdsc->cap_delay_lock); INIT_LIST_HEAD(&mdsc->snap_flush_list); spin_lock_init(&mdsc->snap_flush_lock); diff --git a/fs/ceph/mds_client.h b/fs/ceph/mds_client.h index 5cd131b41d84..14c7e8c49970 100644 --- a/fs/ceph/mds_client.h +++ b/fs/ceph/mds_client.h @@ -340,6 +340,14 @@ struct ceph_quotarealm_inode { struct inode *inode; }; +struct cap_wait { + struct list_head list; + unsigned long ino; + pid_t tgid; + int need; + int want; +}; + /* * mds client state */ @@ -416,6 +424,7 @@ struct ceph_mds_client { spinlock_t caps_list_lock; struct list_head caps_list; /* unused (reserved or unreserved) */ + struct list_head cap_wait_list; int caps_total_count; /* total caps allocated */ int caps_use_count; /* in use */ int caps_use_max; /* max used caps */ diff --git a/fs/ceph/mdsmap.c b/fs/ceph/mdsmap.c index aeec1d6e3769..471bac335fae 100644 --- a/fs/ceph/mdsmap.c +++ b/fs/ceph/mdsmap.c @@ -158,6 +158,7 @@ struct ceph_mdsmap *ceph_mdsmap_decode(void **p, void *end) void *pexport_targets = NULL; struct ceph_timespec laggy_since; struct ceph_mds_info *info; + bool laggy; ceph_decode_need(p, end, sizeof(u64) + 1, bad); global_id = ceph_decode_64(p); @@ -190,6 +191,7 @@ struct ceph_mdsmap *ceph_mdsmap_decode(void **p, void *end) if (err) goto corrupt; ceph_decode_copy(p, &laggy_since, sizeof(laggy_since)); + laggy = laggy_since.tv_sec != 0 || laggy_since.tv_nsec != 0; *p += sizeof(u32); ceph_decode_32_safe(p, end, namelen, bad); *p += namelen; @@ -207,10 +209,11 @@ struct ceph_mdsmap *ceph_mdsmap_decode(void **p, void *end) *p = info_end; } - dout("mdsmap_decode %d/%d %lld mds%d.%d %s %s\n", + dout("mdsmap_decode %d/%d %lld mds%d.%d %s %s%s\n", i+1, n, global_id, mds, inc, ceph_pr_addr(&addr), - ceph_mds_state_name(state)); + ceph_mds_state_name(state), + laggy ? "(laggy)" : ""); if (mds < 0 || state <= 0) continue; @@ -230,8 +233,7 @@ struct ceph_mdsmap *ceph_mdsmap_decode(void **p, void *end) info->global_id = global_id; info->state = state; info->addr = addr; - info->laggy = (laggy_since.tv_sec != 0 || - laggy_since.tv_nsec != 0); + info->laggy = laggy; info->num_export_targets = num_export_targets; if (num_export_targets) { info->export_targets = kcalloc(num_export_targets, @@ -355,6 +357,8 @@ struct ceph_mdsmap *ceph_mdsmap_decode(void **p, void *end) m->m_damaged = false; } bad_ext: + dout("mdsmap_decode m_enabled: %d, m_damaged: %d, m_num_laggy: %d\n", + !!m->m_enabled, !!m->m_damaged, m->m_num_laggy); *p = end; dout("mdsmap_decode success epoch %u\n", m->m_epoch); return m; diff --git a/fs/ceph/super.c b/fs/ceph/super.c index 9c9a7c68eea3..29a795f975df 100644 --- a/fs/ceph/super.c +++ b/fs/ceph/super.c @@ -172,10 +172,10 @@ static const struct fs_parameter_enum ceph_mount_param_enums[] = { static const struct fs_parameter_spec ceph_mount_param_specs[] = { fsparam_flag_no ("acl", Opt_acl), fsparam_flag_no ("asyncreaddir", Opt_asyncreaddir), - fsparam_u32 ("caps_max", Opt_caps_max), + fsparam_s32 ("caps_max", Opt_caps_max), fsparam_u32 ("caps_wanted_delay_max", Opt_caps_wanted_delay_max), fsparam_u32 ("caps_wanted_delay_min", Opt_caps_wanted_delay_min), - fsparam_s32 ("write_congestion_kb", Opt_congestion_kb), + fsparam_u32 ("write_congestion_kb", Opt_congestion_kb), fsparam_flag_no ("copyfrom", Opt_copyfrom), fsparam_flag_no ("dcache", Opt_dcache), fsparam_flag_no ("dirstat", Opt_dirstat), @@ -187,8 +187,8 @@ static const struct fs_parameter_spec ceph_mount_param_specs[] = { fsparam_flag_no ("quotadf", Opt_quotadf), fsparam_u32 ("rasize", Opt_rasize), fsparam_flag_no ("rbytes", Opt_rbytes), - fsparam_s32 ("readdir_max_bytes", Opt_readdir_max_bytes), - fsparam_s32 ("readdir_max_entries", Opt_readdir_max_entries), + fsparam_u32 ("readdir_max_bytes", Opt_readdir_max_bytes), + fsparam_u32 ("readdir_max_entries", Opt_readdir_max_entries), fsparam_enum ("recover_session", Opt_recover_session), fsparam_flag_no ("require_active_mds", Opt_require_active_mds), fsparam_u32 ("rsize", Opt_rsize), @@ -328,7 +328,9 @@ static int ceph_parse_mount_param(struct fs_context *fc, fsopt->caps_wanted_delay_max = result.uint_32; break; case Opt_caps_max: - fsopt->caps_max = result.uint_32; + if (result.int_32 < 0) + goto out_of_range; + fsopt->caps_max = result.int_32; break; case Opt_readdir_max_entries: if (result.uint_32 < 1) @@ -547,25 +549,25 @@ static int ceph_show_options(struct seq_file *m, struct dentry *root) seq_show_option(m, "recover_session", "clean"); if (fsopt->wsize != CEPH_MAX_WRITE_SIZE) - seq_printf(m, ",wsize=%d", fsopt->wsize); + seq_printf(m, ",wsize=%u", fsopt->wsize); if (fsopt->rsize != CEPH_MAX_READ_SIZE) - seq_printf(m, ",rsize=%d", fsopt->rsize); + seq_printf(m, ",rsize=%u", fsopt->rsize); if (fsopt->rasize != CEPH_RASIZE_DEFAULT) - seq_printf(m, ",rasize=%d", fsopt->rasize); + seq_printf(m, ",rasize=%u", fsopt->rasize); if (fsopt->congestion_kb != default_congestion_kb()) - seq_printf(m, ",write_congestion_kb=%d", fsopt->congestion_kb); + seq_printf(m, ",write_congestion_kb=%u", fsopt->congestion_kb); if (fsopt->caps_max) seq_printf(m, ",caps_max=%d", fsopt->caps_max); if (fsopt->caps_wanted_delay_min != CEPH_CAPS_WANTED_DELAY_MIN_DEFAULT) - seq_printf(m, ",caps_wanted_delay_min=%d", + seq_printf(m, ",caps_wanted_delay_min=%u", fsopt->caps_wanted_delay_min); if (fsopt->caps_wanted_delay_max != CEPH_CAPS_WANTED_DELAY_MAX_DEFAULT) - seq_printf(m, ",caps_wanted_delay_max=%d", + seq_printf(m, ",caps_wanted_delay_max=%u", fsopt->caps_wanted_delay_max); if (fsopt->max_readdir != CEPH_MAX_READDIR_DEFAULT) - seq_printf(m, ",readdir_max_entries=%d", fsopt->max_readdir); + seq_printf(m, ",readdir_max_entries=%u", fsopt->max_readdir); if (fsopt->max_readdir_bytes != CEPH_MAX_READDIR_BYTES_DEFAULT) - seq_printf(m, ",readdir_max_bytes=%d", fsopt->max_readdir_bytes); + seq_printf(m, ",readdir_max_bytes=%u", fsopt->max_readdir_bytes); if (strcmp(fsopt->snapdir_name, CEPH_SNAPDIRNAME_DEFAULT)) seq_show_option(m, "snapdirname", fsopt->snapdir_name); diff --git a/fs/ceph/super.h b/fs/ceph/super.h index f0f9cb7447ac..3bf1a01cd536 100644 --- a/fs/ceph/super.h +++ b/fs/ceph/super.h @@ -73,16 +73,16 @@ #define CEPH_CAPS_WANTED_DELAY_MAX_DEFAULT 60 /* cap release delay */ struct ceph_mount_options { - int flags; + unsigned int flags; - int wsize; /* max write size */ - int rsize; /* max read size */ - int rasize; /* max readahead */ - int congestion_kb; /* max writeback in flight */ - int caps_wanted_delay_min, caps_wanted_delay_max; + unsigned int wsize; /* max write size */ + unsigned int rsize; /* max read size */ + unsigned int rasize; /* max readahead */ + unsigned int congestion_kb; /* max writeback in flight */ + unsigned int caps_wanted_delay_min, caps_wanted_delay_max; int caps_max; - int max_readdir; /* max readdir result (entires) */ - int max_readdir_bytes; /* max readdir result (bytes) */ + unsigned int max_readdir; /* max readdir result (entries) */ + unsigned int max_readdir_bytes; /* max readdir result (bytes) */ /* * everything above this point can be memcmp'd; everything below diff --git a/fs/char_dev.c b/fs/char_dev.c index 00dfe17871ac..c5e6eff5a381 100644 --- a/fs/char_dev.c +++ b/fs/char_dev.c @@ -352,7 +352,7 @@ static struct kobject *cdev_get(struct cdev *p) if (owner && !try_module_get(owner)) return NULL; - kobj = kobject_get(&p->kobj); + kobj = kobject_get_unless_zero(&p->kobj); if (!kobj) module_put(owner); return kobj; diff --git a/fs/cifs/cifs_dfs_ref.c b/fs/cifs/cifs_dfs_ref.c index 41957b82d796..606f26d862dc 100644 --- a/fs/cifs/cifs_dfs_ref.c +++ b/fs/cifs/cifs_dfs_ref.c @@ -120,17 +120,17 @@ cifs_build_devname(char *nodename, const char *prepath) /** - * cifs_compose_mount_options - creates mount options for refferral + * cifs_compose_mount_options - creates mount options for referral * @sb_mountdata: parent/root DFS mount options (template) * @fullpath: full path in UNC format - * @ref: server's referral + * @ref: optional server's referral * @devname: optional pointer for saving device name * * creates mount options for submount based on template options sb_mountdata * and replacing unc,ip,prefixpath options with ones we've got form ref_unc. * * Returns: pointer to new mount options or ERR_PTR. - * Caller is responcible for freeing retunrned value if it is not error. + * Caller is responsible for freeing returned value if it is not error. */ char *cifs_compose_mount_options(const char *sb_mountdata, const char *fullpath, @@ -150,18 +150,27 @@ char *cifs_compose_mount_options(const char *sb_mountdata, if (sb_mountdata == NULL) return ERR_PTR(-EINVAL); - if (strlen(fullpath) - ref->path_consumed) { - prepath = fullpath + ref->path_consumed; - /* skip initial delimiter */ - if (*prepath == '/' || *prepath == '\\') - prepath++; - } + if (ref) { + if (strlen(fullpath) - ref->path_consumed) { + prepath = fullpath + ref->path_consumed; + /* skip initial delimiter */ + if (*prepath == '/' || *prepath == '\\') + prepath++; + } - name = cifs_build_devname(ref->node_name, prepath); - if (IS_ERR(name)) { - rc = PTR_ERR(name); - name = NULL; - goto compose_mount_options_err; + name = cifs_build_devname(ref->node_name, prepath); + if (IS_ERR(name)) { + rc = PTR_ERR(name); + name = NULL; + goto compose_mount_options_err; + } + } else { + name = cifs_build_devname((char *)fullpath, NULL); + if (IS_ERR(name)) { + rc = PTR_ERR(name); + name = NULL; + goto compose_mount_options_err; + } } rc = dns_resolve_server_name_to_ip(name, &srvIP); @@ -225,6 +234,8 @@ char *cifs_compose_mount_options(const char *sb_mountdata, if (devname) *devname = name; + else + kfree(name); /*cifs_dbg(FYI, "%s: parent mountdata: %s\n", __func__, sb_mountdata);*/ /*cifs_dbg(FYI, "%s: submount mountdata: %s\n", __func__, mountdata );*/ @@ -241,23 +252,23 @@ compose_mount_options_err: } /** - * cifs_dfs_do_refmount - mounts specified path using provided refferal + * cifs_dfs_do_mount - mounts specified path using DFS full path + * + * Always pass down @fullpath to smb3_do_mount() so we can use the root server + * to perform failover in case we failed to connect to the first target in the + * referral. + * * @cifs_sb: parent/root superblock * @fullpath: full path in UNC format - * @ref: server's referral */ -static struct vfsmount *cifs_dfs_do_refmount(struct dentry *mntpt, - struct cifs_sb_info *cifs_sb, - const char *fullpath, const struct dfs_info3_param *ref) +static struct vfsmount *cifs_dfs_do_mount(struct dentry *mntpt, + struct cifs_sb_info *cifs_sb, + const char *fullpath) { struct vfsmount *mnt; char *mountdata; char *devname; - /* - * Always pass down the DFS full path to smb3_do_mount() so we - * can use it later for failover. - */ devname = kstrndup(fullpath, strlen(fullpath), GFP_KERNEL); if (!devname) return ERR_PTR(-ENOMEM); @@ -266,7 +277,7 @@ static struct vfsmount *cifs_dfs_do_refmount(struct dentry *mntpt, /* strip first '\' from fullpath */ mountdata = cifs_compose_mount_options(cifs_sb->mountdata, - fullpath + 1, ref, NULL); + fullpath + 1, NULL, NULL); if (IS_ERR(mountdata)) { kfree(devname); return (struct vfsmount *)mountdata; @@ -278,28 +289,16 @@ static struct vfsmount *cifs_dfs_do_refmount(struct dentry *mntpt, return mnt; } -static void dump_referral(const struct dfs_info3_param *ref) -{ - cifs_dbg(FYI, "DFS: ref path: %s\n", ref->path_name); - cifs_dbg(FYI, "DFS: node path: %s\n", ref->node_name); - cifs_dbg(FYI, "DFS: fl: %d, srv_type: %d\n", - ref->flags, ref->server_type); - cifs_dbg(FYI, "DFS: ref_flags: %d, path_consumed: %d\n", - ref->ref_flag, ref->path_consumed); -} - /* * Create a vfsmount that we can automount */ static struct vfsmount *cifs_dfs_do_automount(struct dentry *mntpt) { - struct dfs_info3_param referral = {0}; struct cifs_sb_info *cifs_sb; struct cifs_ses *ses; struct cifs_tcon *tcon; char *full_path, *root_path; unsigned int xid; - int len; int rc; struct vfsmount *mnt; @@ -357,7 +356,7 @@ static struct vfsmount *cifs_dfs_do_automount(struct dentry *mntpt) if (!rc) { rc = dfs_cache_find(xid, ses, cifs_sb->local_nls, cifs_remap(cifs_sb), full_path + 1, - &referral, NULL); + NULL, NULL); } free_xid(xid); @@ -366,26 +365,16 @@ static struct vfsmount *cifs_dfs_do_automount(struct dentry *mntpt) mnt = ERR_PTR(rc); goto free_root_path; } - - dump_referral(&referral); - - len = strlen(referral.node_name); - if (len < 2) { - cifs_dbg(VFS, "%s: Net Address path too short: %s\n", - __func__, referral.node_name); - mnt = ERR_PTR(-EINVAL); - goto free_dfs_ref; - } /* - * cifs_mount() will retry every available node server in case - * of failures. + * OK - we were able to get and cache a referral for @full_path. + * + * Now, pass it down to cifs_mount() and it will retry every available + * node server in case of failures - no need to do it here. */ - mnt = cifs_dfs_do_refmount(mntpt, cifs_sb, full_path, &referral); - cifs_dbg(FYI, "%s: cifs_dfs_do_refmount:%s , mnt:%p\n", __func__, - referral.node_name, mnt); + mnt = cifs_dfs_do_mount(mntpt, cifs_sb, full_path); + cifs_dbg(FYI, "%s: cifs_dfs_do_mount:%s , mnt:%p\n", __func__, + full_path + 1, mnt); -free_dfs_ref: - free_dfs_info_param(&referral); free_root_path: kfree(root_path); free_full_path: diff --git a/fs/cifs/cifsacl.c b/fs/cifs/cifsacl.c index 96ae72b556ac..fb41e51dd574 100644 --- a/fs/cifs/cifsacl.c +++ b/fs/cifs/cifsacl.c @@ -802,6 +802,26 @@ static void parse_dacl(struct cifs_acl *pdacl, char *end_of_acl, return; } +unsigned int setup_authusers_ACE(struct cifs_ace *pntace) +{ + int i; + unsigned int ace_size = 20; + + pntace->type = ACCESS_ALLOWED_ACE_TYPE; + pntace->flags = 0x0; + pntace->access_req = cpu_to_le32(GENERIC_ALL); + pntace->sid.num_subauth = 1; + pntace->sid.revision = 1; + for (i = 0; i < NUM_AUTHS; i++) + pntace->sid.authority[i] = sid_authusers.authority[i]; + + pntace->sid.sub_auth[0] = sid_authusers.sub_auth[0]; + + /* size = 1 + 1 + 2 + 4 + 1 + 1 + 6 + (psid->num_subauth*4) */ + pntace->size = cpu_to_le16(ace_size); + return ace_size; +} + /* * Fill in the special SID based on the mode. See * http://technet.microsoft.com/en-us/library/hh509017(v=ws.10).aspx diff --git a/fs/cifs/cifsfs.h b/fs/cifs/cifsfs.h index b59dc7478130..096a4c18fbd0 100644 --- a/fs/cifs/cifsfs.h +++ b/fs/cifs/cifsfs.h @@ -149,6 +149,9 @@ extern ssize_t cifs_file_copychunk_range(unsigned int xid, size_t len, unsigned int flags); extern long cifs_ioctl(struct file *filep, unsigned int cmd, unsigned long arg); +extern void cifs_setsize(struct inode *inode, loff_t offset); +extern int cifs_truncate_page(struct address_space *mapping, loff_t from); + #ifdef CONFIG_CIFS_NFSD_EXPORT extern const struct export_operations cifs_export_ops; #endif /* CONFIG_CIFS_NFSD_EXPORT */ diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index fd0262ce5ad5..239338d57086 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -1061,7 +1061,7 @@ cap_unix(struct cifs_ses *ses) struct cached_fid { bool is_valid:1; /* Do we have a useable root fid */ bool file_all_info_is_valid:1; - + bool has_lease:1; struct kref refcount; struct cifs_fid *fid; struct mutex fid_mutex; @@ -1588,6 +1588,7 @@ struct mid_q_entry { mid_callback_t *callback; /* call completion callback */ mid_handle_t *handle; /* call handle mid callback */ void *callback_data; /* general purpose pointer for callback */ + struct task_struct *creator; void *resp_buf; /* pointer to received SMB header */ unsigned int resp_buf_size; int mid_state; /* wish this were enum but can not pass to wait_event */ @@ -1693,6 +1694,7 @@ struct cifs_fattr { struct timespec64 cf_atime; struct timespec64 cf_mtime; struct timespec64 cf_ctime; + u32 cf_cifstag; }; static inline void free_dfs_info_param(struct dfs_info3_param *param) diff --git a/fs/cifs/cifsproto.h b/fs/cifs/cifsproto.h index 9c229408a251..948bf3474db1 100644 --- a/fs/cifs/cifsproto.h +++ b/fs/cifs/cifsproto.h @@ -213,6 +213,7 @@ extern struct cifs_ntsd *get_cifs_acl_by_fid(struct cifs_sb_info *, const struct cifs_fid *, u32 *); extern int set_cifs_acl(struct cifs_ntsd *, __u32, struct inode *, const char *, int); +extern unsigned int setup_authusers_ACE(struct cifs_ace *pace); extern unsigned int setup_special_mode_ACE(struct cifs_ace *pace, __u64 nmode); extern void dequeue_mid(struct mid_q_entry *mid, bool malformed); @@ -596,6 +597,9 @@ bool is_ses_using_iface(struct cifs_ses *ses, struct cifs_server_iface *iface); void extract_unc_hostname(const char *unc, const char **h, size_t *len); int copy_path_name(char *dst, const char *src); +int smb2_parse_query_directory(struct cifs_tcon *tcon, struct kvec *rsp_iov, + int resp_buftype, + struct cifs_search_info *srch_inf); #ifdef CONFIG_CIFS_DFS_UPCALL static inline int get_dfs_path(const unsigned int xid, struct cifs_ses *ses, diff --git a/fs/cifs/cifssmb.c b/fs/cifs/cifssmb.c index 4f554f019a98..a481296f417f 100644 --- a/fs/cifs/cifssmb.c +++ b/fs/cifs/cifssmb.c @@ -42,6 +42,7 @@ #include "cifsproto.h" #include "cifs_unicode.h" #include "cifs_debug.h" +#include "smb2proto.h" #include "fscache.h" #include "smbdirect.h" #ifdef CONFIG_CIFS_DFS_UPCALL @@ -112,6 +113,8 @@ cifs_mark_open_files_invalid(struct cifs_tcon *tcon) mutex_lock(&tcon->crfid.fid_mutex); tcon->crfid.is_valid = false; + /* cached handle is not valid, so SMB2_CLOSE won't be sent below */ + close_shroot_lease_locked(&tcon->crfid); memset(tcon->crfid.fid, 0, sizeof(struct cifs_fid)); mutex_unlock(&tcon->crfid.fid_mutex); @@ -4616,7 +4619,7 @@ findFirstRetry: psrch_inf->unicode = false; psrch_inf->ntwrk_buf_start = (char *)pSMBr; - psrch_inf->smallBuf = 0; + psrch_inf->smallBuf = false; psrch_inf->srch_entries_start = (char *) &pSMBr->hdr.Protocol + le16_to_cpu(pSMBr->t2.DataOffset); @@ -4750,7 +4753,7 @@ int CIFSFindNext(const unsigned int xid, struct cifs_tcon *tcon, cifs_buf_release(psrch_inf->ntwrk_buf_start); psrch_inf->srch_entries_start = response_data; psrch_inf->ntwrk_buf_start = (char *)pSMB; - psrch_inf->smallBuf = 0; + psrch_inf->smallBuf = false; if (parms->EndofSearch) psrch_inf->endOfSearch = true; else diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 05ea0e2b7e0e..0aa3623ae0e1 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -3709,8 +3709,10 @@ match_prepath(struct super_block *sb, struct cifs_mnt_data *mnt_data) { struct cifs_sb_info *old = CIFS_SB(sb); struct cifs_sb_info *new = mnt_data->cifs_sb; - bool old_set = old->mnt_cifs_flags & CIFS_MOUNT_USE_PREFIX_PATH; - bool new_set = new->mnt_cifs_flags & CIFS_MOUNT_USE_PREFIX_PATH; + bool old_set = (old->mnt_cifs_flags & CIFS_MOUNT_USE_PREFIX_PATH) && + old->prepath; + bool new_set = (new->mnt_cifs_flags & CIFS_MOUNT_USE_PREFIX_PATH) && + new->prepath; if (old_set && new_set && !strcmp(new->prepath, old->prepath)) return 1; diff --git a/fs/cifs/dfs_cache.c b/fs/cifs/dfs_cache.c index 2faa05860a48..9a384d1e27b4 100644 --- a/fs/cifs/dfs_cache.c +++ b/fs/cifs/dfs_cache.c @@ -5,8 +5,6 @@ * Copyright (c) 2018-2019 Paulo Alcantara <palcantara@suse.de> */ -#include <linux/rcupdate.h> -#include <linux/rculist.h> #include <linux/jhash.h> #include <linux/ktime.h> #include <linux/slab.h> @@ -22,67 +20,68 @@ #include "dfs_cache.h" -#define DFS_CACHE_HTABLE_SIZE 32 -#define DFS_CACHE_MAX_ENTRIES 64 +#define CACHE_HTABLE_SIZE 32 +#define CACHE_MAX_ENTRIES 64 #define IS_INTERLINK_SET(v) ((v) & (DFSREF_REFERRAL_SERVER | \ DFSREF_STORAGE_SERVER)) -struct dfs_cache_tgt { - char *t_name; - struct list_head t_list; +struct cache_dfs_tgt { + char *name; + struct list_head list; }; -struct dfs_cache_entry { - struct hlist_node ce_hlist; - const char *ce_path; - int ce_ttl; - int ce_srvtype; - int ce_flags; - struct timespec64 ce_etime; - int ce_path_consumed; - int ce_numtgts; - struct list_head ce_tlist; - struct dfs_cache_tgt *ce_tgthint; - struct rcu_head ce_rcu; +struct cache_entry { + struct hlist_node hlist; + const char *path; + int ttl; + int srvtype; + int flags; + struct timespec64 etime; + int path_consumed; + int numtgts; + struct list_head tlist; + struct cache_dfs_tgt *tgthint; }; -static struct kmem_cache *dfs_cache_slab __read_mostly; - -struct dfs_cache_vol_info { - char *vi_fullpath; - struct smb_vol vi_vol; - char *vi_mntdata; - struct list_head vi_list; +struct vol_info { + char *fullpath; + spinlock_t smb_vol_lock; + struct smb_vol smb_vol; + char *mntdata; + struct list_head list; + struct list_head rlist; + struct kref refcnt; }; -struct dfs_cache { - struct mutex dc_lock; - struct nls_table *dc_nlsc; - struct list_head dc_vol_list; - int dc_ttl; - struct delayed_work dc_refresh; -}; +static struct kmem_cache *cache_slab __read_mostly; +static struct workqueue_struct *dfscache_wq __read_mostly; -static struct dfs_cache dfs_cache; +static int cache_ttl; +static DEFINE_SPINLOCK(cache_ttl_lock); + +static struct nls_table *cache_nlsc; /* * Number of entries in the cache */ -static size_t dfs_cache_count; +static atomic_t cache_count; + +static struct hlist_head cache_htable[CACHE_HTABLE_SIZE]; +static DECLARE_RWSEM(htable_rw_lock); -static DEFINE_MUTEX(dfs_cache_list_lock); -static struct hlist_head dfs_cache_htable[DFS_CACHE_HTABLE_SIZE]; +static LIST_HEAD(vol_list); +static DEFINE_SPINLOCK(vol_list_lock); static void refresh_cache_worker(struct work_struct *work); -static inline bool is_path_valid(const char *path) -{ - return path && (strchr(path + 1, '\\') || strchr(path + 1, '/')); -} +static DECLARE_DELAYED_WORK(refresh_task, refresh_cache_worker); -static inline int get_normalized_path(const char *path, char **npath) +static int get_normalized_path(const char *path, char **npath) { + if (!path || strlen(path) < 3 || (*path != '\\' && *path != '/')) + return -EINVAL; + if (*path == '\\') { *npath = (char *)path; } else { @@ -100,57 +99,48 @@ static inline void free_normalized_path(const char *path, char *npath) kfree(npath); } -static inline bool cache_entry_expired(const struct dfs_cache_entry *ce) +static inline bool cache_entry_expired(const struct cache_entry *ce) { struct timespec64 ts; ktime_get_coarse_real_ts64(&ts); - return timespec64_compare(&ts, &ce->ce_etime) >= 0; + return timespec64_compare(&ts, &ce->etime) >= 0; } -static inline void free_tgts(struct dfs_cache_entry *ce) +static inline void free_tgts(struct cache_entry *ce) { - struct dfs_cache_tgt *t, *n; + struct cache_dfs_tgt *t, *n; - list_for_each_entry_safe(t, n, &ce->ce_tlist, t_list) { - list_del(&t->t_list); - kfree(t->t_name); + list_for_each_entry_safe(t, n, &ce->tlist, list) { + list_del(&t->list); + kfree(t->name); kfree(t); } } -static void free_cache_entry(struct rcu_head *rcu) +static inline void flush_cache_ent(struct cache_entry *ce) { - struct dfs_cache_entry *ce = container_of(rcu, struct dfs_cache_entry, - ce_rcu); - kmem_cache_free(dfs_cache_slab, ce); -} - -static inline void flush_cache_ent(struct dfs_cache_entry *ce) -{ - if (hlist_unhashed(&ce->ce_hlist)) - return; - - hlist_del_init_rcu(&ce->ce_hlist); - kfree_const(ce->ce_path); + hlist_del_init(&ce->hlist); + kfree(ce->path); free_tgts(ce); - dfs_cache_count--; - call_rcu(&ce->ce_rcu, free_cache_entry); + atomic_dec(&cache_count); + kmem_cache_free(cache_slab, ce); } static void flush_cache_ents(void) { int i; - rcu_read_lock(); - for (i = 0; i < DFS_CACHE_HTABLE_SIZE; i++) { - struct hlist_head *l = &dfs_cache_htable[i]; - struct dfs_cache_entry *ce; + for (i = 0; i < CACHE_HTABLE_SIZE; i++) { + struct hlist_head *l = &cache_htable[i]; + struct hlist_node *n; + struct cache_entry *ce; - hlist_for_each_entry_rcu(ce, l, ce_hlist) - flush_cache_ent(ce); + hlist_for_each_entry_safe(ce, n, l, hlist) { + if (!hlist_unhashed(&ce->hlist)) + flush_cache_ent(ce); + } } - rcu_read_unlock(); } /* @@ -158,36 +148,39 @@ static void flush_cache_ents(void) */ static int dfscache_proc_show(struct seq_file *m, void *v) { - int bucket; - struct dfs_cache_entry *ce; - struct dfs_cache_tgt *t; + int i; + struct cache_entry *ce; + struct cache_dfs_tgt *t; seq_puts(m, "DFS cache\n---------\n"); - mutex_lock(&dfs_cache_list_lock); - - rcu_read_lock(); - hash_for_each_rcu(dfs_cache_htable, bucket, ce, ce_hlist) { - seq_printf(m, - "cache entry: path=%s,type=%s,ttl=%d,etime=%ld," - "interlink=%s,path_consumed=%d,expired=%s\n", - ce->ce_path, - ce->ce_srvtype == DFS_TYPE_ROOT ? "root" : "link", - ce->ce_ttl, ce->ce_etime.tv_nsec, - IS_INTERLINK_SET(ce->ce_flags) ? "yes" : "no", - ce->ce_path_consumed, - cache_entry_expired(ce) ? "yes" : "no"); - - list_for_each_entry(t, &ce->ce_tlist, t_list) { - seq_printf(m, " %s%s\n", - t->t_name, - ce->ce_tgthint == t ? " (target hint)" : ""); + down_read(&htable_rw_lock); + for (i = 0; i < CACHE_HTABLE_SIZE; i++) { + struct hlist_head *l = &cache_htable[i]; + + hlist_for_each_entry(ce, l, hlist) { + if (hlist_unhashed(&ce->hlist)) + continue; + + seq_printf(m, + "cache entry: path=%s,type=%s,ttl=%d,etime=%ld," + "interlink=%s,path_consumed=%d,expired=%s\n", + ce->path, + ce->srvtype == DFS_TYPE_ROOT ? "root" : "link", + ce->ttl, ce->etime.tv_nsec, + IS_INTERLINK_SET(ce->flags) ? "yes" : "no", + ce->path_consumed, + cache_entry_expired(ce) ? "yes" : "no"); + + list_for_each_entry(t, &ce->tlist, list) { + seq_printf(m, " %s%s\n", + t->name, + ce->tgthint == t ? " (target hint)" : ""); + } } - } - rcu_read_unlock(); + up_read(&htable_rw_lock); - mutex_unlock(&dfs_cache_list_lock); return 0; } @@ -205,9 +198,10 @@ static ssize_t dfscache_proc_write(struct file *file, const char __user *buffer, return -EINVAL; cifs_dbg(FYI, "clearing dfs cache"); - mutex_lock(&dfs_cache_list_lock); + + down_write(&htable_rw_lock); flush_cache_ents(); - mutex_unlock(&dfs_cache_list_lock); + up_write(&htable_rw_lock); return count; } @@ -226,25 +220,25 @@ const struct file_operations dfscache_proc_fops = { }; #ifdef CONFIG_CIFS_DEBUG2 -static inline void dump_tgts(const struct dfs_cache_entry *ce) +static inline void dump_tgts(const struct cache_entry *ce) { - struct dfs_cache_tgt *t; + struct cache_dfs_tgt *t; cifs_dbg(FYI, "target list:\n"); - list_for_each_entry(t, &ce->ce_tlist, t_list) { - cifs_dbg(FYI, " %s%s\n", t->t_name, - ce->ce_tgthint == t ? " (target hint)" : ""); + list_for_each_entry(t, &ce->tlist, list) { + cifs_dbg(FYI, " %s%s\n", t->name, + ce->tgthint == t ? " (target hint)" : ""); } } -static inline void dump_ce(const struct dfs_cache_entry *ce) +static inline void dump_ce(const struct cache_entry *ce) { cifs_dbg(FYI, "cache entry: path=%s,type=%s,ttl=%d,etime=%ld," - "interlink=%s,path_consumed=%d,expired=%s\n", ce->ce_path, - ce->ce_srvtype == DFS_TYPE_ROOT ? "root" : "link", ce->ce_ttl, - ce->ce_etime.tv_nsec, - IS_INTERLINK_SET(ce->ce_flags) ? "yes" : "no", - ce->ce_path_consumed, + "interlink=%s,path_consumed=%d,expired=%s\n", ce->path, + ce->srvtype == DFS_TYPE_ROOT ? "root" : "link", ce->ttl, + ce->etime.tv_nsec, + IS_INTERLINK_SET(ce->flags) ? "yes" : "no", + ce->path_consumed, cache_entry_expired(ce) ? "yes" : "no"); dump_tgts(ce); } @@ -284,25 +278,34 @@ static inline void dump_refs(const struct dfs_info3_param *refs, int numrefs) */ int dfs_cache_init(void) { + int rc; int i; - dfs_cache_slab = kmem_cache_create("cifs_dfs_cache", - sizeof(struct dfs_cache_entry), 0, - SLAB_HWCACHE_ALIGN, NULL); - if (!dfs_cache_slab) + dfscache_wq = alloc_workqueue("cifs-dfscache", + WQ_FREEZABLE | WQ_MEM_RECLAIM, 1); + if (!dfscache_wq) return -ENOMEM; - for (i = 0; i < DFS_CACHE_HTABLE_SIZE; i++) - INIT_HLIST_HEAD(&dfs_cache_htable[i]); + cache_slab = kmem_cache_create("cifs_dfs_cache", + sizeof(struct cache_entry), 0, + SLAB_HWCACHE_ALIGN, NULL); + if (!cache_slab) { + rc = -ENOMEM; + goto out_destroy_wq; + } + + for (i = 0; i < CACHE_HTABLE_SIZE; i++) + INIT_HLIST_HEAD(&cache_htable[i]); - INIT_LIST_HEAD(&dfs_cache.dc_vol_list); - mutex_init(&dfs_cache.dc_lock); - INIT_DELAYED_WORK(&dfs_cache.dc_refresh, refresh_cache_worker); - dfs_cache.dc_ttl = -1; - dfs_cache.dc_nlsc = load_nls_default(); + atomic_set(&cache_count, 0); + cache_nlsc = load_nls_default(); cifs_dbg(FYI, "%s: initialized DFS referral cache\n", __func__); return 0; + +out_destroy_wq: + destroy_workqueue(dfscache_wq); + return rc; } static inline unsigned int cache_entry_hash(const void *data, int size) @@ -310,7 +313,7 @@ static inline unsigned int cache_entry_hash(const void *data, int size) unsigned int h; h = jhash(data, size, 0); - return h & (DFS_CACHE_HTABLE_SIZE - 1); + return h & (CACHE_HTABLE_SIZE - 1); } /* Check whether second path component of @path is SYSVOL or NETLOGON */ @@ -325,11 +328,11 @@ static inline bool is_sysvol_or_netlogon(const char *path) } /* Return target hint of a DFS cache entry */ -static inline char *get_tgt_name(const struct dfs_cache_entry *ce) +static inline char *get_tgt_name(const struct cache_entry *ce) { - struct dfs_cache_tgt *t = ce->ce_tgthint; + struct cache_dfs_tgt *t = ce->tgthint; - return t ? t->t_name : ERR_PTR(-ENOENT); + return t ? t->name : ERR_PTR(-ENOENT); } /* Return expire time out of a new entry's TTL */ @@ -346,19 +349,19 @@ static inline struct timespec64 get_expire_time(int ttl) } /* Allocate a new DFS target */ -static inline struct dfs_cache_tgt *alloc_tgt(const char *name) +static struct cache_dfs_tgt *alloc_target(const char *name) { - struct dfs_cache_tgt *t; + struct cache_dfs_tgt *t; - t = kmalloc(sizeof(*t), GFP_KERNEL); + t = kmalloc(sizeof(*t), GFP_ATOMIC); if (!t) return ERR_PTR(-ENOMEM); - t->t_name = kstrndup(name, strlen(name), GFP_KERNEL); - if (!t->t_name) { + t->name = kstrndup(name, strlen(name), GFP_ATOMIC); + if (!t->name) { kfree(t); return ERR_PTR(-ENOMEM); } - INIT_LIST_HEAD(&t->t_list); + INIT_LIST_HEAD(&t->list); return t; } @@ -367,180 +370,184 @@ static inline struct dfs_cache_tgt *alloc_tgt(const char *name) * target hint. */ static int copy_ref_data(const struct dfs_info3_param *refs, int numrefs, - struct dfs_cache_entry *ce, const char *tgthint) + struct cache_entry *ce, const char *tgthint) { int i; - ce->ce_ttl = refs[0].ttl; - ce->ce_etime = get_expire_time(ce->ce_ttl); - ce->ce_srvtype = refs[0].server_type; - ce->ce_flags = refs[0].ref_flag; - ce->ce_path_consumed = refs[0].path_consumed; + ce->ttl = refs[0].ttl; + ce->etime = get_expire_time(ce->ttl); + ce->srvtype = refs[0].server_type; + ce->flags = refs[0].ref_flag; + ce->path_consumed = refs[0].path_consumed; for (i = 0; i < numrefs; i++) { - struct dfs_cache_tgt *t; + struct cache_dfs_tgt *t; - t = alloc_tgt(refs[i].node_name); + t = alloc_target(refs[i].node_name); if (IS_ERR(t)) { free_tgts(ce); return PTR_ERR(t); } - if (tgthint && !strcasecmp(t->t_name, tgthint)) { - list_add(&t->t_list, &ce->ce_tlist); + if (tgthint && !strcasecmp(t->name, tgthint)) { + list_add(&t->list, &ce->tlist); tgthint = NULL; } else { - list_add_tail(&t->t_list, &ce->ce_tlist); + list_add_tail(&t->list, &ce->tlist); } - ce->ce_numtgts++; + ce->numtgts++; } - ce->ce_tgthint = list_first_entry_or_null(&ce->ce_tlist, - struct dfs_cache_tgt, t_list); + ce->tgthint = list_first_entry_or_null(&ce->tlist, + struct cache_dfs_tgt, list); return 0; } /* Allocate a new cache entry */ -static struct dfs_cache_entry * -alloc_cache_entry(const char *path, const struct dfs_info3_param *refs, - int numrefs) +static struct cache_entry *alloc_cache_entry(const char *path, + const struct dfs_info3_param *refs, + int numrefs) { - struct dfs_cache_entry *ce; + struct cache_entry *ce; int rc; - ce = kmem_cache_zalloc(dfs_cache_slab, GFP_KERNEL); + ce = kmem_cache_zalloc(cache_slab, GFP_KERNEL); if (!ce) return ERR_PTR(-ENOMEM); - ce->ce_path = kstrdup_const(path, GFP_KERNEL); - if (!ce->ce_path) { - kmem_cache_free(dfs_cache_slab, ce); + ce->path = kstrndup(path, strlen(path), GFP_KERNEL); + if (!ce->path) { + kmem_cache_free(cache_slab, ce); return ERR_PTR(-ENOMEM); } - INIT_HLIST_NODE(&ce->ce_hlist); - INIT_LIST_HEAD(&ce->ce_tlist); + INIT_HLIST_NODE(&ce->hlist); + INIT_LIST_HEAD(&ce->tlist); rc = copy_ref_data(refs, numrefs, ce, NULL); if (rc) { - kfree_const(ce->ce_path); - kmem_cache_free(dfs_cache_slab, ce); + kfree(ce->path); + kmem_cache_free(cache_slab, ce); ce = ERR_PTR(rc); } return ce; } +/* Must be called with htable_rw_lock held */ static void remove_oldest_entry(void) { - int bucket; - struct dfs_cache_entry *ce; - struct dfs_cache_entry *to_del = NULL; - - rcu_read_lock(); - hash_for_each_rcu(dfs_cache_htable, bucket, ce, ce_hlist) { - if (!to_del || timespec64_compare(&ce->ce_etime, - &to_del->ce_etime) < 0) - to_del = ce; + int i; + struct cache_entry *ce; + struct cache_entry *to_del = NULL; + + for (i = 0; i < CACHE_HTABLE_SIZE; i++) { + struct hlist_head *l = &cache_htable[i]; + + hlist_for_each_entry(ce, l, hlist) { + if (hlist_unhashed(&ce->hlist)) + continue; + if (!to_del || timespec64_compare(&ce->etime, + &to_del->etime) < 0) + to_del = ce; + } } + if (!to_del) { cifs_dbg(FYI, "%s: no entry to remove", __func__); - goto out; + return; } + cifs_dbg(FYI, "%s: removing entry", __func__); dump_ce(to_del); flush_cache_ent(to_del); -out: - rcu_read_unlock(); } /* Add a new DFS cache entry */ -static inline struct dfs_cache_entry * -add_cache_entry(unsigned int hash, const char *path, - const struct dfs_info3_param *refs, int numrefs) +static int add_cache_entry(const char *path, unsigned int hash, + struct dfs_info3_param *refs, int numrefs) { - struct dfs_cache_entry *ce; + struct cache_entry *ce; ce = alloc_cache_entry(path, refs, numrefs); if (IS_ERR(ce)) - return ce; + return PTR_ERR(ce); - hlist_add_head_rcu(&ce->ce_hlist, &dfs_cache_htable[hash]); - - mutex_lock(&dfs_cache.dc_lock); - if (dfs_cache.dc_ttl < 0) { - dfs_cache.dc_ttl = ce->ce_ttl; - queue_delayed_work(cifsiod_wq, &dfs_cache.dc_refresh, - dfs_cache.dc_ttl * HZ); + spin_lock(&cache_ttl_lock); + if (!cache_ttl) { + cache_ttl = ce->ttl; + queue_delayed_work(dfscache_wq, &refresh_task, cache_ttl * HZ); } else { - dfs_cache.dc_ttl = min_t(int, dfs_cache.dc_ttl, ce->ce_ttl); - mod_delayed_work(cifsiod_wq, &dfs_cache.dc_refresh, - dfs_cache.dc_ttl * HZ); + cache_ttl = min_t(int, cache_ttl, ce->ttl); + mod_delayed_work(dfscache_wq, &refresh_task, cache_ttl * HZ); } - mutex_unlock(&dfs_cache.dc_lock); + spin_unlock(&cache_ttl_lock); - return ce; + down_write(&htable_rw_lock); + hlist_add_head(&ce->hlist, &cache_htable[hash]); + dump_ce(ce); + up_write(&htable_rw_lock); + + return 0; } -static struct dfs_cache_entry *__find_cache_entry(unsigned int hash, - const char *path) +/* + * Find a DFS cache entry in hash table and optionally check prefix path against + * @path. + * Use whole path components in the match. + * Must be called with htable_rw_lock held. + * + * Return ERR_PTR(-ENOENT) if the entry is not found. + */ +static struct cache_entry *lookup_cache_entry(const char *path, + unsigned int *hash) { - struct dfs_cache_entry *ce; + struct cache_entry *ce; + unsigned int h; bool found = false; - rcu_read_lock(); - hlist_for_each_entry_rcu(ce, &dfs_cache_htable[hash], ce_hlist) { - if (!strcasecmp(path, ce->ce_path)) { -#ifdef CONFIG_CIFS_DEBUG2 - char *name = get_tgt_name(ce); + h = cache_entry_hash(path, strlen(path)); - if (IS_ERR(name)) { - rcu_read_unlock(); - return ERR_CAST(name); - } - cifs_dbg(FYI, "%s: cache hit\n", __func__); - cifs_dbg(FYI, "%s: target hint: %s\n", __func__, name); -#endif + hlist_for_each_entry(ce, &cache_htable[h], hlist) { + if (!strcasecmp(path, ce->path)) { found = true; + dump_ce(ce); break; } } - rcu_read_unlock(); - return found ? ce : ERR_PTR(-ENOENT); -} -/* - * Find a DFS cache entry in hash table and optionally check prefix path against - * @path. - * Use whole path components in the match. - * Return ERR_PTR(-ENOENT) if the entry is not found. - */ -static inline struct dfs_cache_entry *find_cache_entry(const char *path, - unsigned int *hash) -{ - *hash = cache_entry_hash(path, strlen(path)); - return __find_cache_entry(*hash, path); + if (!found) + ce = ERR_PTR(-ENOENT); + if (hash) + *hash = h; + + return ce; } -static inline void destroy_slab_cache(void) +static void __vol_release(struct vol_info *vi) { - rcu_barrier(); - kmem_cache_destroy(dfs_cache_slab); + kfree(vi->fullpath); + kfree(vi->mntdata); + cifs_cleanup_volume_info_contents(&vi->smb_vol); + kfree(vi); } -static inline void free_vol(struct dfs_cache_vol_info *vi) +static void vol_release(struct kref *kref) { - list_del(&vi->vi_list); - kfree(vi->vi_fullpath); - kfree(vi->vi_mntdata); - cifs_cleanup_volume_info_contents(&vi->vi_vol); - kfree(vi); + struct vol_info *vi = container_of(kref, struct vol_info, refcnt); + + spin_lock(&vol_list_lock); + list_del(&vi->list); + spin_unlock(&vol_list_lock); + __vol_release(vi); } static inline void free_vol_list(void) { - struct dfs_cache_vol_info *vi, *nvi; + struct vol_info *vi, *nvi; - list_for_each_entry_safe(vi, nvi, &dfs_cache.dc_vol_list, vi_list) - free_vol(vi); + list_for_each_entry_safe(vi, nvi, &vol_list, list) { + list_del_init(&vi->list); + __vol_release(vi); + } } /** @@ -548,83 +555,78 @@ static inline void free_vol_list(void) */ void dfs_cache_destroy(void) { - cancel_delayed_work_sync(&dfs_cache.dc_refresh); - unload_nls(dfs_cache.dc_nlsc); + cancel_delayed_work_sync(&refresh_task); + unload_nls(cache_nlsc); free_vol_list(); - mutex_destroy(&dfs_cache.dc_lock); - flush_cache_ents(); - destroy_slab_cache(); - mutex_destroy(&dfs_cache_list_lock); + kmem_cache_destroy(cache_slab); + destroy_workqueue(dfscache_wq); cifs_dbg(FYI, "%s: destroyed DFS referral cache\n", __func__); } -static inline struct dfs_cache_entry * -__update_cache_entry(const char *path, const struct dfs_info3_param *refs, - int numrefs) +/* Must be called with htable_rw_lock held */ +static int __update_cache_entry(const char *path, + const struct dfs_info3_param *refs, + int numrefs) { int rc; - unsigned int h; - struct dfs_cache_entry *ce; + struct cache_entry *ce; char *s, *th = NULL; - ce = find_cache_entry(path, &h); + ce = lookup_cache_entry(path, NULL); if (IS_ERR(ce)) - return ce; + return PTR_ERR(ce); - if (ce->ce_tgthint) { - s = ce->ce_tgthint->t_name; - th = kstrndup(s, strlen(s), GFP_KERNEL); + if (ce->tgthint) { + s = ce->tgthint->name; + th = kstrndup(s, strlen(s), GFP_ATOMIC); if (!th) - return ERR_PTR(-ENOMEM); + return -ENOMEM; } free_tgts(ce); - ce->ce_numtgts = 0; + ce->numtgts = 0; rc = copy_ref_data(refs, numrefs, ce, th); - kfree(th); - if (rc) - ce = ERR_PTR(rc); + kfree(th); - return ce; + return rc; } -/* Update an expired cache entry by getting a new DFS referral from server */ -static struct dfs_cache_entry * -update_cache_entry(const unsigned int xid, struct cifs_ses *ses, - const struct nls_table *nls_codepage, int remap, - const char *path, struct dfs_cache_entry *ce) +static int get_dfs_referral(const unsigned int xid, struct cifs_ses *ses, + const struct nls_table *nls_codepage, int remap, + const char *path, struct dfs_info3_param **refs, + int *numrefs) { - int rc; - struct dfs_info3_param *refs = NULL; - int numrefs = 0; + cifs_dbg(FYI, "%s: get an DFS referral for %s\n", __func__, path); - cifs_dbg(FYI, "%s: update expired cache entry\n", __func__); - /* - * Check if caller provided enough parameters to update an expired - * entry. - */ if (!ses || !ses->server || !ses->server->ops->get_dfs_refer) - return ERR_PTR(-ETIME); + return -EOPNOTSUPP; if (unlikely(!nls_codepage)) - return ERR_PTR(-ETIME); + return -EINVAL; - cifs_dbg(FYI, "%s: DFS referral request for %s\n", __func__, path); + *refs = NULL; + *numrefs = 0; - rc = ses->server->ops->get_dfs_refer(xid, ses, path, &refs, &numrefs, - nls_codepage, remap); - if (rc) - ce = ERR_PTR(rc); - else - ce = __update_cache_entry(path, refs, numrefs); + return ses->server->ops->get_dfs_refer(xid, ses, path, refs, numrefs, + nls_codepage, remap); +} - dump_refs(refs, numrefs); - free_dfs_info_array(refs, numrefs); +/* Update an expired cache entry by getting a new DFS referral from server */ +static int update_cache_entry(const char *path, + const struct dfs_info3_param *refs, + int numrefs) +{ - return ce; + int rc; + + down_write(&htable_rw_lock); + rc = __update_cache_entry(path, refs, numrefs); + up_write(&htable_rw_lock); + + return rc; } /* @@ -636,95 +638,86 @@ update_cache_entry(const unsigned int xid, struct cifs_ses *ses, * For interlinks, __cifs_dfs_mount() and expand_dfs_referral() are supposed to * handle them properly. */ -static struct dfs_cache_entry * -do_dfs_cache_find(const unsigned int xid, struct cifs_ses *ses, - const struct nls_table *nls_codepage, int remap, - const char *path, bool noreq) +static int __dfs_cache_find(const unsigned int xid, struct cifs_ses *ses, + const struct nls_table *nls_codepage, int remap, + const char *path, bool noreq) { int rc; - unsigned int h; - struct dfs_cache_entry *ce; - struct dfs_info3_param *nrefs; - int numnrefs; + unsigned int hash; + struct cache_entry *ce; + struct dfs_info3_param *refs = NULL; + int numrefs = 0; + bool newent = false; cifs_dbg(FYI, "%s: search path: %s\n", __func__, path); - ce = find_cache_entry(path, &h); - if (IS_ERR(ce)) { - cifs_dbg(FYI, "%s: cache miss\n", __func__); - /* - * If @noreq is set, no requests will be sent to the server for - * either updating or getting a new DFS referral. - */ - if (noreq) - return ce; - /* - * No cache entry was found, so check for valid parameters that - * will be required to get a new DFS referral and then create a - * new cache entry. - */ - if (!ses || !ses->server || !ses->server->ops->get_dfs_refer) { - ce = ERR_PTR(-EOPNOTSUPP); - return ce; - } - if (unlikely(!nls_codepage)) { - ce = ERR_PTR(-EINVAL); - return ce; - } + down_read(&htable_rw_lock); - nrefs = NULL; - numnrefs = 0; + ce = lookup_cache_entry(path, &hash); - cifs_dbg(FYI, "%s: DFS referral request for %s\n", __func__, - path); + /* + * If @noreq is set, no requests will be sent to the server. Just return + * the cache entry. + */ + if (noreq) { + up_read(&htable_rw_lock); + return PTR_ERR_OR_ZERO(ce); + } - rc = ses->server->ops->get_dfs_refer(xid, ses, path, &nrefs, - &numnrefs, nls_codepage, - remap); - if (rc) { - ce = ERR_PTR(rc); - return ce; + if (!IS_ERR(ce)) { + if (!cache_entry_expired(ce)) { + dump_ce(ce); + up_read(&htable_rw_lock); + return 0; } + } else { + newent = true; + } - dump_refs(nrefs, numnrefs); + up_read(&htable_rw_lock); - cifs_dbg(FYI, "%s: new cache entry\n", __func__); + /* + * No entry was found. + * + * Request a new DFS referral in order to create a new cache entry, or + * updating an existing one. + */ + rc = get_dfs_referral(xid, ses, nls_codepage, remap, path, + &refs, &numrefs); + if (rc) + return rc; - if (dfs_cache_count >= DFS_CACHE_MAX_ENTRIES) { - cifs_dbg(FYI, "%s: reached max cache size (%d)", - __func__, DFS_CACHE_MAX_ENTRIES); - remove_oldest_entry(); - } - ce = add_cache_entry(h, path, nrefs, numnrefs); - free_dfs_info_array(nrefs, numnrefs); + dump_refs(refs, numrefs); - if (IS_ERR(ce)) - return ce; + if (!newent) { + rc = update_cache_entry(path, refs, numrefs); + goto out_free_refs; + } - dfs_cache_count++; + if (atomic_read(&cache_count) >= CACHE_MAX_ENTRIES) { + cifs_dbg(FYI, "%s: reached max cache size (%d)", __func__, + CACHE_MAX_ENTRIES); + down_write(&htable_rw_lock); + remove_oldest_entry(); + up_write(&htable_rw_lock); } - dump_ce(ce); + rc = add_cache_entry(path, hash, refs, numrefs); + if (!rc) + atomic_inc(&cache_count); - /* Just return the found cache entry in case @noreq is set */ - if (noreq) - return ce; - - if (cache_entry_expired(ce)) { - cifs_dbg(FYI, "%s: expired cache entry\n", __func__); - ce = update_cache_entry(xid, ses, nls_codepage, remap, path, - ce); - if (IS_ERR(ce)) { - cifs_dbg(FYI, "%s: failed to update expired entry\n", - __func__); - } - } - return ce; +out_free_refs: + free_dfs_info_array(refs, numrefs); + return rc; } -/* Set up a new DFS referral from a given cache entry */ -static int setup_ref(const char *path, const struct dfs_cache_entry *ce, - struct dfs_info3_param *ref, const char *tgt) +/* + * Set up a DFS referral from a given cache entry. + * + * Must be called with htable_rw_lock held. + */ +static int setup_referral(const char *path, struct cache_entry *ce, + struct dfs_info3_param *ref, const char *target) { int rc; @@ -732,21 +725,20 @@ static int setup_ref(const char *path, const struct dfs_cache_entry *ce, memset(ref, 0, sizeof(*ref)); - ref->path_name = kstrndup(path, strlen(path), GFP_KERNEL); + ref->path_name = kstrndup(path, strlen(path), GFP_ATOMIC); if (!ref->path_name) return -ENOMEM; - ref->path_consumed = ce->ce_path_consumed; - - ref->node_name = kstrndup(tgt, strlen(tgt), GFP_KERNEL); + ref->node_name = kstrndup(target, strlen(target), GFP_ATOMIC); if (!ref->node_name) { rc = -ENOMEM; goto err_free_path; } - ref->ttl = ce->ce_ttl; - ref->server_type = ce->ce_srvtype; - ref->ref_flag = ce->ce_flags; + ref->path_consumed = ce->path_consumed; + ref->ttl = ce->ttl; + ref->server_type = ce->srvtype; + ref->ref_flag = ce->flags; return 0; @@ -757,38 +749,37 @@ err_free_path: } /* Return target list of a DFS cache entry */ -static int get_tgt_list(const struct dfs_cache_entry *ce, - struct dfs_cache_tgt_list *tl) +static int get_targets(struct cache_entry *ce, struct dfs_cache_tgt_list *tl) { int rc; struct list_head *head = &tl->tl_list; - struct dfs_cache_tgt *t; + struct cache_dfs_tgt *t; struct dfs_cache_tgt_iterator *it, *nit; memset(tl, 0, sizeof(*tl)); INIT_LIST_HEAD(head); - list_for_each_entry(t, &ce->ce_tlist, t_list) { - it = kzalloc(sizeof(*it), GFP_KERNEL); + list_for_each_entry(t, &ce->tlist, list) { + it = kzalloc(sizeof(*it), GFP_ATOMIC); if (!it) { rc = -ENOMEM; goto err_free_it; } - it->it_name = kstrndup(t->t_name, strlen(t->t_name), - GFP_KERNEL); + it->it_name = kstrndup(t->name, strlen(t->name), GFP_ATOMIC); if (!it->it_name) { kfree(it); rc = -ENOMEM; goto err_free_it; } - if (ce->ce_tgthint == t) + if (ce->tgthint == t) list_add(&it->it_list, head); else list_add_tail(&it->it_list, head); } - tl->tl_numtgts = ce->ce_numtgts; + + tl->tl_numtgts = ce->numtgts; return 0; @@ -829,28 +820,35 @@ int dfs_cache_find(const unsigned int xid, struct cifs_ses *ses, { int rc; char *npath; - struct dfs_cache_entry *ce; - - if (unlikely(!is_path_valid(path))) - return -EINVAL; + struct cache_entry *ce; rc = get_normalized_path(path, &npath); if (rc) return rc; - mutex_lock(&dfs_cache_list_lock); - ce = do_dfs_cache_find(xid, ses, nls_codepage, remap, npath, false); - if (!IS_ERR(ce)) { - if (ref) - rc = setup_ref(path, ce, ref, get_tgt_name(ce)); - else - rc = 0; - if (!rc && tgt_list) - rc = get_tgt_list(ce, tgt_list); - } else { + rc = __dfs_cache_find(xid, ses, nls_codepage, remap, npath, false); + if (rc) + goto out_free_path; + + down_read(&htable_rw_lock); + + ce = lookup_cache_entry(npath, NULL); + if (IS_ERR(ce)) { + up_read(&htable_rw_lock); rc = PTR_ERR(ce); + goto out_free_path; } - mutex_unlock(&dfs_cache_list_lock); + + if (ref) + rc = setup_referral(path, ce, ref, get_tgt_name(ce)); + else + rc = 0; + if (!rc && tgt_list) + rc = get_targets(ce, tgt_list); + + up_read(&htable_rw_lock); + +out_free_path: free_normalized_path(path, npath); return rc; } @@ -876,31 +874,33 @@ int dfs_cache_noreq_find(const char *path, struct dfs_info3_param *ref, { int rc; char *npath; - struct dfs_cache_entry *ce; - - if (unlikely(!is_path_valid(path))) - return -EINVAL; + struct cache_entry *ce; rc = get_normalized_path(path, &npath); if (rc) return rc; - mutex_lock(&dfs_cache_list_lock); - ce = do_dfs_cache_find(0, NULL, NULL, 0, npath, true); + cifs_dbg(FYI, "%s: path: %s\n", __func__, npath); + + down_read(&htable_rw_lock); + + ce = lookup_cache_entry(npath, NULL); if (IS_ERR(ce)) { rc = PTR_ERR(ce); - goto out; + goto out_unlock; } if (ref) - rc = setup_ref(path, ce, ref, get_tgt_name(ce)); + rc = setup_referral(path, ce, ref, get_tgt_name(ce)); else rc = 0; if (!rc && tgt_list) - rc = get_tgt_list(ce, tgt_list); -out: - mutex_unlock(&dfs_cache_list_lock); + rc = get_targets(ce, tgt_list); + +out_unlock: + up_read(&htable_rw_lock); free_normalized_path(path, npath); + return rc; } @@ -929,44 +929,46 @@ int dfs_cache_update_tgthint(const unsigned int xid, struct cifs_ses *ses, { int rc; char *npath; - struct dfs_cache_entry *ce; - struct dfs_cache_tgt *t; - - if (unlikely(!is_path_valid(path))) - return -EINVAL; + struct cache_entry *ce; + struct cache_dfs_tgt *t; rc = get_normalized_path(path, &npath); if (rc) return rc; - cifs_dbg(FYI, "%s: path: %s\n", __func__, npath); + cifs_dbg(FYI, "%s: update target hint - path: %s\n", __func__, npath); - mutex_lock(&dfs_cache_list_lock); - ce = do_dfs_cache_find(xid, ses, nls_codepage, remap, npath, false); + rc = __dfs_cache_find(xid, ses, nls_codepage, remap, npath, false); + if (rc) + goto out_free_path; + + down_write(&htable_rw_lock); + + ce = lookup_cache_entry(npath, NULL); if (IS_ERR(ce)) { rc = PTR_ERR(ce); - goto out; + goto out_unlock; } - rc = 0; - - t = ce->ce_tgthint; + t = ce->tgthint; - if (likely(!strcasecmp(it->it_name, t->t_name))) - goto out; + if (likely(!strcasecmp(it->it_name, t->name))) + goto out_unlock; - list_for_each_entry(t, &ce->ce_tlist, t_list) { - if (!strcasecmp(t->t_name, it->it_name)) { - ce->ce_tgthint = t; + list_for_each_entry(t, &ce->tlist, list) { + if (!strcasecmp(t->name, it->it_name)) { + ce->tgthint = t; cifs_dbg(FYI, "%s: new target hint: %s\n", __func__, it->it_name); break; } } -out: - mutex_unlock(&dfs_cache_list_lock); +out_unlock: + up_write(&htable_rw_lock); +out_free_path: free_normalized_path(path, npath); + return rc; } @@ -989,10 +991,10 @@ int dfs_cache_noreq_update_tgthint(const char *path, { int rc; char *npath; - struct dfs_cache_entry *ce; - struct dfs_cache_tgt *t; + struct cache_entry *ce; + struct cache_dfs_tgt *t; - if (unlikely(!is_path_valid(path)) || !it) + if (!it) return -EINVAL; rc = get_normalized_path(path, &npath); @@ -1001,33 +1003,33 @@ int dfs_cache_noreq_update_tgthint(const char *path, cifs_dbg(FYI, "%s: path: %s\n", __func__, npath); - mutex_lock(&dfs_cache_list_lock); + down_write(&htable_rw_lock); - ce = do_dfs_cache_find(0, NULL, NULL, 0, npath, true); + ce = lookup_cache_entry(npath, NULL); if (IS_ERR(ce)) { rc = PTR_ERR(ce); - goto out; + goto out_unlock; } rc = 0; + t = ce->tgthint; - t = ce->ce_tgthint; + if (unlikely(!strcasecmp(it->it_name, t->name))) + goto out_unlock; - if (unlikely(!strcasecmp(it->it_name, t->t_name))) - goto out; - - list_for_each_entry(t, &ce->ce_tlist, t_list) { - if (!strcasecmp(t->t_name, it->it_name)) { - ce->ce_tgthint = t; + list_for_each_entry(t, &ce->tlist, list) { + if (!strcasecmp(t->name, it->it_name)) { + ce->tgthint = t; cifs_dbg(FYI, "%s: new target hint: %s\n", __func__, it->it_name); break; } } -out: - mutex_unlock(&dfs_cache_list_lock); +out_unlock: + up_write(&htable_rw_lock); free_normalized_path(path, npath); + return rc; } @@ -1047,13 +1049,10 @@ int dfs_cache_get_tgt_referral(const char *path, { int rc; char *npath; - struct dfs_cache_entry *ce; - unsigned int h; + struct cache_entry *ce; if (!it || !ref) return -EINVAL; - if (unlikely(!is_path_valid(path))) - return -EINVAL; rc = get_normalized_path(path, &npath); if (rc) @@ -1061,21 +1060,22 @@ int dfs_cache_get_tgt_referral(const char *path, cifs_dbg(FYI, "%s: path: %s\n", __func__, npath); - mutex_lock(&dfs_cache_list_lock); + down_read(&htable_rw_lock); - ce = find_cache_entry(npath, &h); + ce = lookup_cache_entry(npath, NULL); if (IS_ERR(ce)) { rc = PTR_ERR(ce); - goto out; + goto out_unlock; } cifs_dbg(FYI, "%s: target name: %s\n", __func__, it->it_name); - rc = setup_ref(path, ce, ref, it->it_name); + rc = setup_referral(path, ce, ref, it->it_name); -out: - mutex_unlock(&dfs_cache_list_lock); +out_unlock: + up_read(&htable_rw_lock); free_normalized_path(path, npath); + return rc; } @@ -1085,7 +1085,7 @@ static int dup_vol(struct smb_vol *vol, struct smb_vol *new) if (vol->username) { new->username = kstrndup(vol->username, strlen(vol->username), - GFP_KERNEL); + GFP_KERNEL); if (!new->username) return -ENOMEM; } @@ -1103,7 +1103,7 @@ static int dup_vol(struct smb_vol *vol, struct smb_vol *new) } if (vol->domainname) { new->domainname = kstrndup(vol->domainname, - strlen(vol->domainname), GFP_KERNEL); + strlen(vol->domainname), GFP_KERNEL); if (!new->domainname) goto err_free_unc; } @@ -1150,7 +1150,7 @@ err_free_username: int dfs_cache_add_vol(char *mntdata, struct smb_vol *vol, const char *fullpath) { int rc; - struct dfs_cache_vol_info *vi; + struct vol_info *vi; if (!vol || !fullpath || !mntdata) return -EINVAL; @@ -1161,38 +1161,41 @@ int dfs_cache_add_vol(char *mntdata, struct smb_vol *vol, const char *fullpath) if (!vi) return -ENOMEM; - vi->vi_fullpath = kstrndup(fullpath, strlen(fullpath), GFP_KERNEL); - if (!vi->vi_fullpath) { + vi->fullpath = kstrndup(fullpath, strlen(fullpath), GFP_KERNEL); + if (!vi->fullpath) { rc = -ENOMEM; goto err_free_vi; } - rc = dup_vol(vol, &vi->vi_vol); + rc = dup_vol(vol, &vi->smb_vol); if (rc) goto err_free_fullpath; - vi->vi_mntdata = mntdata; + vi->mntdata = mntdata; + spin_lock_init(&vi->smb_vol_lock); + kref_init(&vi->refcnt); + + spin_lock(&vol_list_lock); + list_add_tail(&vi->list, &vol_list); + spin_unlock(&vol_list_lock); - mutex_lock(&dfs_cache.dc_lock); - list_add_tail(&vi->vi_list, &dfs_cache.dc_vol_list); - mutex_unlock(&dfs_cache.dc_lock); return 0; err_free_fullpath: - kfree(vi->vi_fullpath); + kfree(vi->fullpath); err_free_vi: kfree(vi); return rc; } -static inline struct dfs_cache_vol_info *find_vol(const char *fullpath) +/* Must be called with vol_list_lock held */ +static struct vol_info *find_vol(const char *fullpath) { - struct dfs_cache_vol_info *vi; + struct vol_info *vi; - list_for_each_entry(vi, &dfs_cache.dc_vol_list, vi_list) { - cifs_dbg(FYI, "%s: vi->vi_fullpath: %s\n", __func__, - vi->vi_fullpath); - if (!strcasecmp(vi->vi_fullpath, fullpath)) + list_for_each_entry(vi, &vol_list, list) { + cifs_dbg(FYI, "%s: vi->fullpath: %s\n", __func__, vi->fullpath); + if (!strcasecmp(vi->fullpath, fullpath)) return vi; } return ERR_PTR(-ENOENT); @@ -1208,30 +1211,31 @@ static inline struct dfs_cache_vol_info *find_vol(const char *fullpath) */ int dfs_cache_update_vol(const char *fullpath, struct TCP_Server_Info *server) { - int rc; - struct dfs_cache_vol_info *vi; + struct vol_info *vi; if (!fullpath || !server) return -EINVAL; cifs_dbg(FYI, "%s: fullpath: %s\n", __func__, fullpath); - mutex_lock(&dfs_cache.dc_lock); - + spin_lock(&vol_list_lock); vi = find_vol(fullpath); if (IS_ERR(vi)) { - rc = PTR_ERR(vi); - goto out; + spin_unlock(&vol_list_lock); + return PTR_ERR(vi); } + kref_get(&vi->refcnt); + spin_unlock(&vol_list_lock); cifs_dbg(FYI, "%s: updating volume info\n", __func__); - memcpy(&vi->vi_vol.dstaddr, &server->dstaddr, - sizeof(vi->vi_vol.dstaddr)); - rc = 0; + spin_lock(&vi->smb_vol_lock); + memcpy(&vi->smb_vol.dstaddr, &server->dstaddr, + sizeof(vi->smb_vol.dstaddr)); + spin_unlock(&vi->smb_vol_lock); -out: - mutex_unlock(&dfs_cache.dc_lock); - return rc; + kref_put(&vi->refcnt, vol_release); + + return 0; } /** @@ -1241,18 +1245,18 @@ out: */ void dfs_cache_del_vol(const char *fullpath) { - struct dfs_cache_vol_info *vi; + struct vol_info *vi; if (!fullpath || !*fullpath) return; cifs_dbg(FYI, "%s: fullpath: %s\n", __func__, fullpath); - mutex_lock(&dfs_cache.dc_lock); + spin_lock(&vol_list_lock); vi = find_vol(fullpath); - if (!IS_ERR(vi)) - free_vol(vi); - mutex_unlock(&dfs_cache.dc_lock); + spin_unlock(&vol_list_lock); + + kref_put(&vi->refcnt, vol_release); } /* Get all tcons that are within a DFS namespace and can be refreshed */ @@ -1280,7 +1284,7 @@ static void get_tcons(struct TCP_Server_Info *server, struct list_head *head) spin_unlock(&cifs_tcp_ses_lock); } -static inline bool is_dfs_link(const char *path) +static bool is_dfs_link(const char *path) { char *s; @@ -1290,7 +1294,7 @@ static inline bool is_dfs_link(const char *path) return !!strchr(s + 1, '\\'); } -static inline char *get_dfs_root(const char *path) +static char *get_dfs_root(const char *path) { char *s, *npath; @@ -1309,31 +1313,67 @@ static inline char *get_dfs_root(const char *path) return npath; } +static inline void put_tcp_server(struct TCP_Server_Info *server) +{ + cifs_put_tcp_session(server, 0); +} + +static struct TCP_Server_Info *get_tcp_server(struct smb_vol *vol) +{ + struct TCP_Server_Info *server; + + server = cifs_find_tcp_session(vol); + if (IS_ERR_OR_NULL(server)) + return NULL; + + spin_lock(&GlobalMid_Lock); + if (server->tcpStatus != CifsGood) { + spin_unlock(&GlobalMid_Lock); + put_tcp_server(server); + return NULL; + } + spin_unlock(&GlobalMid_Lock); + + return server; +} + /* Find root SMB session out of a DFS link path */ -static struct cifs_ses *find_root_ses(struct dfs_cache_vol_info *vi, - struct cifs_tcon *tcon, const char *path) +static struct cifs_ses *find_root_ses(struct vol_info *vi, + struct cifs_tcon *tcon, + const char *path) { char *rpath; int rc; + struct cache_entry *ce; struct dfs_info3_param ref = {0}; char *mdata = NULL, *devname = NULL; struct TCP_Server_Info *server; struct cifs_ses *ses; - struct smb_vol vol; + struct smb_vol vol = {NULL}; rpath = get_dfs_root(path); if (IS_ERR(rpath)) return ERR_CAST(rpath); - memset(&vol, 0, sizeof(vol)); + down_read(&htable_rw_lock); + + ce = lookup_cache_entry(rpath, NULL); + if (IS_ERR(ce)) { + up_read(&htable_rw_lock); + ses = ERR_CAST(ce); + goto out; + } - rc = dfs_cache_noreq_find(rpath, &ref, NULL); + rc = setup_referral(path, ce, &ref, get_tgt_name(ce)); if (rc) { + up_read(&htable_rw_lock); ses = ERR_PTR(rc); goto out; } - mdata = cifs_compose_mount_options(vi->vi_mntdata, rpath, &ref, + up_read(&htable_rw_lock); + + mdata = cifs_compose_mount_options(vi->mntdata, rpath, &ref, &devname); free_dfs_info_param(&ref); @@ -1351,13 +1391,8 @@ static struct cifs_ses *find_root_ses(struct dfs_cache_vol_info *vi, goto out; } - server = cifs_find_tcp_session(&vol); - if (IS_ERR_OR_NULL(server)) { - ses = ERR_PTR(-EHOSTDOWN); - goto out; - } - if (server->tcpStatus != CifsGood) { - cifs_put_tcp_session(server, 0); + server = get_tcp_server(&vol); + if (!server) { ses = ERR_PTR(-EHOSTDOWN); goto out; } @@ -1373,17 +1408,15 @@ out: } /* Refresh DFS cache entry from a given tcon */ -static void do_refresh_tcon(struct dfs_cache *dc, struct dfs_cache_vol_info *vi, - struct cifs_tcon *tcon) +static int refresh_tcon(struct vol_info *vi, struct cifs_tcon *tcon) { int rc = 0; unsigned int xid; char *path, *npath; - unsigned int h; - struct dfs_cache_entry *ce; + struct cache_entry *ce; + struct cifs_ses *root_ses = NULL, *ses; struct dfs_info3_param *refs = NULL; int numrefs = 0; - struct cifs_ses *root_ses = NULL, *ses; xid = get_xid(); @@ -1391,19 +1424,23 @@ static void do_refresh_tcon(struct dfs_cache *dc, struct dfs_cache_vol_info *vi, rc = get_normalized_path(path, &npath); if (rc) - goto out; + goto out_free_xid; - mutex_lock(&dfs_cache_list_lock); - ce = find_cache_entry(npath, &h); - mutex_unlock(&dfs_cache_list_lock); + down_read(&htable_rw_lock); + ce = lookup_cache_entry(npath, NULL); if (IS_ERR(ce)) { rc = PTR_ERR(ce); - goto out; + up_read(&htable_rw_lock); + goto out_free_path; } - if (!cache_entry_expired(ce)) - goto out; + if (!cache_entry_expired(ce)) { + up_read(&htable_rw_lock); + goto out_free_path; + } + + up_read(&htable_rw_lock); /* If it's a DFS Link, then use root SMB session for refreshing it */ if (is_dfs_link(npath)) { @@ -1411,35 +1448,29 @@ static void do_refresh_tcon(struct dfs_cache *dc, struct dfs_cache_vol_info *vi, if (IS_ERR(ses)) { rc = PTR_ERR(ses); root_ses = NULL; - goto out; + goto out_free_path; } } else { ses = tcon->ses; } - if (unlikely(!ses->server->ops->get_dfs_refer)) { - rc = -EOPNOTSUPP; - } else { - rc = ses->server->ops->get_dfs_refer(xid, ses, path, &refs, - &numrefs, dc->dc_nlsc, - tcon->remap); - if (!rc) { - mutex_lock(&dfs_cache_list_lock); - ce = __update_cache_entry(npath, refs, numrefs); - mutex_unlock(&dfs_cache_list_lock); - dump_refs(refs, numrefs); - free_dfs_info_array(refs, numrefs); - if (IS_ERR(ce)) - rc = PTR_ERR(ce); - } + rc = get_dfs_referral(xid, ses, cache_nlsc, tcon->remap, npath, &refs, + &numrefs); + if (!rc) { + dump_refs(refs, numrefs); + rc = update_cache_entry(npath, refs, numrefs); + free_dfs_info_array(refs, numrefs); } -out: if (root_ses) cifs_put_smb_ses(root_ses); - free_xid(xid); +out_free_path: free_normalized_path(path, npath); + +out_free_xid: + free_xid(xid); + return rc; } /* @@ -1448,30 +1479,61 @@ out: */ static void refresh_cache_worker(struct work_struct *work) { - struct dfs_cache *dc = container_of(work, struct dfs_cache, - dc_refresh.work); - struct dfs_cache_vol_info *vi; + struct vol_info *vi, *nvi; struct TCP_Server_Info *server; - LIST_HEAD(list); + LIST_HEAD(vols); + LIST_HEAD(tcons); struct cifs_tcon *tcon, *ntcon; + int rc; - mutex_lock(&dc->dc_lock); - - list_for_each_entry(vi, &dc->dc_vol_list, vi_list) { - server = cifs_find_tcp_session(&vi->vi_vol); - if (IS_ERR_OR_NULL(server)) + /* + * Find SMB volumes that are eligible (server->tcpStatus == CifsGood) + * for refreshing. + */ + spin_lock(&vol_list_lock); + list_for_each_entry(vi, &vol_list, list) { + server = get_tcp_server(&vi->smb_vol); + if (!server) continue; - if (server->tcpStatus != CifsGood) - goto next; - get_tcons(server, &list); - list_for_each_entry_safe(tcon, ntcon, &list, ulist) { - do_refresh_tcon(dc, vi, tcon); + + kref_get(&vi->refcnt); + list_add_tail(&vi->rlist, &vols); + put_tcp_server(server); + } + spin_unlock(&vol_list_lock); + + /* Walk through all TCONs and refresh any expired cache entry */ + list_for_each_entry_safe(vi, nvi, &vols, rlist) { + spin_lock(&vi->smb_vol_lock); + server = get_tcp_server(&vi->smb_vol); + spin_unlock(&vi->smb_vol_lock); + + if (!server) + goto next_vol; + + get_tcons(server, &tcons); + rc = 0; + + list_for_each_entry_safe(tcon, ntcon, &tcons, ulist) { + /* + * Skip tcp server if any of its tcons failed to refresh + * (possibily due to reconnects). + */ + if (!rc) + rc = refresh_tcon(vi, tcon); + list_del_init(&tcon->ulist); cifs_put_tcon(tcon); } -next: - cifs_put_tcp_session(server, 0); + + put_tcp_server(server); + +next_vol: + list_del_init(&vi->rlist); + kref_put(&vi->refcnt, vol_release); } - queue_delayed_work(cifsiod_wq, &dc->dc_refresh, dc->dc_ttl * HZ); - mutex_unlock(&dc->dc_lock); + + spin_lock(&cache_ttl_lock); + queue_delayed_work(dfscache_wq, &refresh_task, cache_ttl * HZ); + spin_unlock(&cache_ttl_lock); } diff --git a/fs/cifs/file.c b/fs/cifs/file.c index 043288b5c728..a4e8f7d445ac 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -2921,7 +2921,7 @@ cifs_write_from_iter(loff_t offset, size_t len, struct iov_iter *from, "direct_writev couldn't get user pages " "(rc=%zd) iter type %d iov_offset %zd " "count %zd\n", - result, from->type, + result, iov_iter_type(from), from->iov_offset, from->count); dump_stack(); @@ -3132,7 +3132,7 @@ static ssize_t __cifs_writev( * In this case, fall back to non-direct write function. * this could be improved by getting pages directly in ITER_KVEC */ - if (direct && from->type & ITER_KVEC) { + if (direct && iov_iter_is_kvec(from)) { cifs_dbg(FYI, "use non-direct cifs_writev for kvec I/O\n"); direct = false; } @@ -3652,7 +3652,7 @@ cifs_send_async_read(loff_t offset, size_t len, struct cifsFileInfo *open_file, "couldn't get user pages (rc=%zd)" " iter type %d" " iov_offset %zd count %zd\n", - result, direct_iov.type, + result, iov_iter_type(&direct_iov), direct_iov.iov_offset, direct_iov.count); dump_stack(); @@ -3863,7 +3863,7 @@ static ssize_t __cifs_readv( * fall back to data copy read path * this could be improved by getting pages directly in ITER_KVEC */ - if (direct && to->type & ITER_KVEC) { + if (direct && iov_iter_is_kvec(to)) { cifs_dbg(FYI, "use non-direct cifs_user_readv for kvec I/O\n"); direct = false; } diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index ca76a9287456..9b547f7f5f5d 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -2228,7 +2228,7 @@ int cifs_fiemap(struct inode *inode, struct fiemap_extent_info *fei, u64 start, return -ENOTSUPP; } -static int cifs_truncate_page(struct address_space *mapping, loff_t from) +int cifs_truncate_page(struct address_space *mapping, loff_t from) { pgoff_t index = from >> PAGE_SHIFT; unsigned offset = from & (PAGE_SIZE - 1); @@ -2245,7 +2245,7 @@ static int cifs_truncate_page(struct address_space *mapping, loff_t from) return rc; } -static void cifs_setsize(struct inode *inode, loff_t offset) +void cifs_setsize(struct inode *inode, loff_t offset) { struct cifsInodeInfo *cifs_i = CIFS_I(inode); diff --git a/fs/cifs/readdir.c b/fs/cifs/readdir.c index 3925a7bfc74d..d17587c2c4ab 100644 --- a/fs/cifs/readdir.c +++ b/fs/cifs/readdir.c @@ -139,6 +139,28 @@ retry: dput(dentry); } +static bool reparse_file_needs_reval(const struct cifs_fattr *fattr) +{ + if (!(fattr->cf_cifsattrs & ATTR_REPARSE)) + return false; + /* + * The DFS tags should be only intepreted by server side as per + * MS-FSCC 2.1.2.1, but let's include them anyway. + * + * Besides, if cf_cifstag is unset (0), then we still need it to be + * revalidated to know exactly what reparse point it is. + */ + switch (fattr->cf_cifstag) { + case IO_REPARSE_TAG_DFS: + case IO_REPARSE_TAG_DFSR: + case IO_REPARSE_TAG_SYMLINK: + case IO_REPARSE_TAG_NFS: + case 0: + return true; + } + return false; +} + static void cifs_fill_common_info(struct cifs_fattr *fattr, struct cifs_sb_info *cifs_sb) { @@ -158,7 +180,7 @@ cifs_fill_common_info(struct cifs_fattr *fattr, struct cifs_sb_info *cifs_sb) * is a symbolic link, DFS referral or a reparse point with a direct * access like junctions, deduplicated files, NFS symlinks. */ - if (fattr->cf_cifsattrs & ATTR_REPARSE) + if (reparse_file_needs_reval(fattr)) fattr->cf_flags |= CIFS_FATTR_NEED_REVAL; /* non-unix readdir doesn't provide nlink */ @@ -194,19 +216,37 @@ cifs_fill_common_info(struct cifs_fattr *fattr, struct cifs_sb_info *cifs_sb) } } +static void __dir_info_to_fattr(struct cifs_fattr *fattr, const void *info) +{ + const FILE_DIRECTORY_INFO *fi = info; + + memset(fattr, 0, sizeof(*fattr)); + fattr->cf_cifsattrs = le32_to_cpu(fi->ExtFileAttributes); + fattr->cf_eof = le64_to_cpu(fi->EndOfFile); + fattr->cf_bytes = le64_to_cpu(fi->AllocationSize); + fattr->cf_createtime = le64_to_cpu(fi->CreationTime); + fattr->cf_atime = cifs_NTtimeToUnix(fi->LastAccessTime); + fattr->cf_ctime = cifs_NTtimeToUnix(fi->ChangeTime); + fattr->cf_mtime = cifs_NTtimeToUnix(fi->LastWriteTime); +} + void cifs_dir_info_to_fattr(struct cifs_fattr *fattr, FILE_DIRECTORY_INFO *info, struct cifs_sb_info *cifs_sb) { - memset(fattr, 0, sizeof(*fattr)); - fattr->cf_cifsattrs = le32_to_cpu(info->ExtFileAttributes); - fattr->cf_eof = le64_to_cpu(info->EndOfFile); - fattr->cf_bytes = le64_to_cpu(info->AllocationSize); - fattr->cf_createtime = le64_to_cpu(info->CreationTime); - fattr->cf_atime = cifs_NTtimeToUnix(info->LastAccessTime); - fattr->cf_ctime = cifs_NTtimeToUnix(info->ChangeTime); - fattr->cf_mtime = cifs_NTtimeToUnix(info->LastWriteTime); + __dir_info_to_fattr(fattr, info); + cifs_fill_common_info(fattr, cifs_sb); +} +static void cifs_fulldir_info_to_fattr(struct cifs_fattr *fattr, + SEARCH_ID_FULL_DIR_INFO *info, + struct cifs_sb_info *cifs_sb) +{ + __dir_info_to_fattr(fattr, info); + + /* See MS-FSCC 2.4.18 FileIdFullDirectoryInformation */ + if (fattr->cf_cifsattrs & ATTR_REPARSE) + fattr->cf_cifstag = le32_to_cpu(info->EaSize); cifs_fill_common_info(fattr, cifs_sb); } @@ -755,6 +795,11 @@ static int cifs_filldir(char *find_entry, struct file *file, (FIND_FILE_STANDARD_INFO *)find_entry, cifs_sb); break; + case SMB_FIND_FILE_ID_FULL_DIR_INFO: + cifs_fulldir_info_to_fattr(&fattr, + (SEARCH_ID_FULL_DIR_INFO *)find_entry, + cifs_sb); + break; default: cifs_dir_info_to_fattr(&fattr, (FILE_DIRECTORY_INFO *)find_entry, diff --git a/fs/cifs/smb2file.c b/fs/cifs/smb2file.c index 8b0b512c5792..afe1f03aabe3 100644 --- a/fs/cifs/smb2file.c +++ b/fs/cifs/smb2file.c @@ -67,7 +67,7 @@ smb2_open_file(const unsigned int xid, struct cifs_open_parms *oparms, goto out; - if (oparms->tcon->use_resilient) { + if (oparms->tcon->use_resilient) { /* default timeout is 0, servers pick default (120 seconds) */ nr_ioctl_req.Timeout = cpu_to_le32(oparms->tcon->handle_timeout); diff --git a/fs/cifs/smb2inode.c b/fs/cifs/smb2inode.c index 18c7a33adceb..5ef5e97a6d13 100644 --- a/fs/cifs/smb2inode.c +++ b/fs/cifs/smb2inode.c @@ -95,6 +95,7 @@ smb2_compound_op(const unsigned int xid, struct cifs_tcon *tcon, goto finished; } + memset(&oparms, 0, sizeof(struct cifs_open_parms)); oparms.tcon = tcon; oparms.desired_access = desired_access; oparms.disposition = create_disposition; diff --git a/fs/cifs/smb2misc.c b/fs/cifs/smb2misc.c index 0516fc482d43..0511aaf451d4 100644 --- a/fs/cifs/smb2misc.c +++ b/fs/cifs/smb2misc.c @@ -743,7 +743,7 @@ __smb2_handle_cancelled_cmd(struct cifs_tcon *tcon, __u16 cmd, __u64 mid, { struct close_cancelled_open *cancelled; - cancelled = kzalloc(sizeof(*cancelled), GFP_KERNEL); + cancelled = kzalloc(sizeof(*cancelled), GFP_ATOMIC); if (!cancelled) return -ENOMEM; diff --git a/fs/cifs/smb2ops.c b/fs/cifs/smb2ops.c index a5c96bc522cb..6787fce26f20 100644 --- a/fs/cifs/smb2ops.c +++ b/fs/cifs/smb2ops.c @@ -12,6 +12,7 @@ #include <linux/uuid.h> #include <linux/sort.h> #include <crypto/aead.h> +#include "cifsfs.h" #include "cifsglob.h" #include "smb2pdu.h" #include "smb2proto.h" @@ -616,6 +617,7 @@ smb2_close_cached_fid(struct kref *ref) cfid->fid->volatile_fid); cfid->is_valid = false; cfid->file_all_info_is_valid = false; + cfid->has_lease = false; } } @@ -626,13 +628,28 @@ void close_shroot(struct cached_fid *cfid) mutex_unlock(&cfid->fid_mutex); } +void close_shroot_lease_locked(struct cached_fid *cfid) +{ + if (cfid->has_lease) { + cfid->has_lease = false; + kref_put(&cfid->refcount, smb2_close_cached_fid); + } +} + +void close_shroot_lease(struct cached_fid *cfid) +{ + mutex_lock(&cfid->fid_mutex); + close_shroot_lease_locked(cfid); + mutex_unlock(&cfid->fid_mutex); +} + void smb2_cached_lease_break(struct work_struct *work) { struct cached_fid *cfid = container_of(work, struct cached_fid, lease_break); - close_shroot(cfid); + close_shroot_lease(cfid); } /* @@ -773,6 +790,7 @@ int open_shroot(unsigned int xid, struct cifs_tcon *tcon, struct cifs_fid *pfid) /* BB TBD check to see if oplock level check can be removed below */ if (o_rsp->OplockLevel == SMB2_OPLOCK_LEVEL_LEASE) { kref_get(&tcon->crfid.refcount); + tcon->crfid.has_lease = true; smb2_parse_contexts(server, o_rsp, &oparms.fid->epoch, oparms.fid->lease_key, &oplock, NULL); @@ -787,7 +805,7 @@ int open_shroot(unsigned int xid, struct cifs_tcon *tcon, struct cifs_fid *pfid) sizeof(struct smb2_file_all_info), &rsp_iov[1], sizeof(struct smb2_file_all_info), (char *)&tcon->crfid.file_all_info)) - tcon->crfid.file_all_info_is_valid = 1; + tcon->crfid.file_all_info_is_valid = true; oshr_exit: mutex_unlock(&tcon->crfid.fid_mutex); @@ -1506,7 +1524,9 @@ smb2_ioctl_query_info(const unsigned int xid, COMPOUND_FID, COMPOUND_FID, qi.info_type, true, buffer, qi.output_buffer_length, - CIFSMaxBufSize); + CIFSMaxBufSize - + MAX_SMB2_CREATE_RESPONSE_SIZE - + MAX_SMB2_CLOSE_RESPONSE_SIZE); } } else if (qi.flags == PASSTHRU_SET_INFO) { /* Can eventually relax perm check since server enforces too */ @@ -2036,14 +2056,33 @@ smb2_query_dir_first(const unsigned int xid, struct cifs_tcon *tcon, struct cifs_search_info *srch_inf) { __le16 *utf16_path; - int rc; - __u8 oplock = SMB2_OPLOCK_LEVEL_NONE; + struct smb_rqst rqst[2]; + struct kvec rsp_iov[2]; + int resp_buftype[2]; + struct kvec open_iov[SMB2_CREATE_IOV_SIZE]; + struct kvec qd_iov[SMB2_QUERY_DIRECTORY_IOV_SIZE]; + int rc, flags = 0; + u8 oplock = SMB2_OPLOCK_LEVEL_NONE; struct cifs_open_parms oparms; + struct smb2_query_directory_rsp *qd_rsp = NULL; + struct smb2_create_rsp *op_rsp = NULL; utf16_path = cifs_convert_path_to_utf16(path, cifs_sb); if (!utf16_path) return -ENOMEM; + if (smb3_encryption_required(tcon)) + flags |= CIFS_TRANSFORM_REQ; + + memset(rqst, 0, sizeof(rqst)); + resp_buftype[0] = resp_buftype[1] = CIFS_NO_BUFFER; + memset(rsp_iov, 0, sizeof(rsp_iov)); + + /* Open */ + memset(&open_iov, 0, sizeof(open_iov)); + rqst[0].rq_iov = open_iov; + rqst[0].rq_nvec = SMB2_CREATE_IOV_SIZE; + oparms.tcon = tcon; oparms.desired_access = FILE_READ_ATTRIBUTES | FILE_READ_DATA; oparms.disposition = FILE_OPEN; @@ -2054,22 +2093,75 @@ smb2_query_dir_first(const unsigned int xid, struct cifs_tcon *tcon, oparms.fid = fid; oparms.reconnect = false; - rc = SMB2_open(xid, &oparms, utf16_path, &oplock, NULL, NULL, NULL); - kfree(utf16_path); - if (rc) { - cifs_dbg(FYI, "open dir failed rc=%d\n", rc); - return rc; - } + rc = SMB2_open_init(tcon, &rqst[0], &oplock, &oparms, utf16_path); + if (rc) + goto qdf_free; + smb2_set_next_command(tcon, &rqst[0]); + /* Query directory */ srch_inf->entries_in_buffer = 0; srch_inf->index_of_last_entry = 2; - rc = SMB2_query_directory(xid, tcon, fid->persistent_fid, - fid->volatile_fid, 0, srch_inf); - if (rc) { - cifs_dbg(FYI, "query directory failed rc=%d\n", rc); + memset(&qd_iov, 0, sizeof(qd_iov)); + rqst[1].rq_iov = qd_iov; + rqst[1].rq_nvec = SMB2_QUERY_DIRECTORY_IOV_SIZE; + + rc = SMB2_query_directory_init(xid, tcon, &rqst[1], + COMPOUND_FID, COMPOUND_FID, + 0, srch_inf->info_level); + if (rc) + goto qdf_free; + + smb2_set_related(&rqst[1]); + + rc = compound_send_recv(xid, tcon->ses, flags, 2, rqst, + resp_buftype, rsp_iov); + + /* If the open failed there is nothing to do */ + op_rsp = (struct smb2_create_rsp *)rsp_iov[0].iov_base; + if (op_rsp == NULL || op_rsp->sync_hdr.Status != STATUS_SUCCESS) { + cifs_dbg(FYI, "query_dir_first: open failed rc=%d\n", rc); + goto qdf_free; + } + fid->persistent_fid = op_rsp->PersistentFileId; + fid->volatile_fid = op_rsp->VolatileFileId; + + /* Anything else than ENODATA means a genuine error */ + if (rc && rc != -ENODATA) { SMB2_close(xid, tcon, fid->persistent_fid, fid->volatile_fid); + cifs_dbg(FYI, "query_dir_first: query directory failed rc=%d\n", rc); + trace_smb3_query_dir_err(xid, fid->persistent_fid, + tcon->tid, tcon->ses->Suid, 0, 0, rc); + goto qdf_free; } + + qd_rsp = (struct smb2_query_directory_rsp *)rsp_iov[1].iov_base; + if (qd_rsp->sync_hdr.Status == STATUS_NO_MORE_FILES) { + trace_smb3_query_dir_done(xid, fid->persistent_fid, + tcon->tid, tcon->ses->Suid, 0, 0); + srch_inf->endOfSearch = true; + rc = 0; + goto qdf_free; + } + + rc = smb2_parse_query_directory(tcon, &rsp_iov[1], resp_buftype[1], + srch_inf); + if (rc) { + trace_smb3_query_dir_err(xid, fid->persistent_fid, tcon->tid, + tcon->ses->Suid, 0, 0, rc); + goto qdf_free; + } + resp_buftype[1] = CIFS_NO_BUFFER; + + trace_smb3_query_dir_done(xid, fid->persistent_fid, tcon->tid, + tcon->ses->Suid, 0, srch_inf->entries_in_buffer); + + qdf_free: + kfree(utf16_path); + SMB2_open_free(&rqst[0]); + SMB2_query_directory_free(&rqst[1]); + free_rsp_buf(resp_buftype[0], rsp_iov[0].iov_base); + free_rsp_buf(resp_buftype[1], rsp_iov[1].iov_base); return rc; } @@ -2680,7 +2772,10 @@ smb2_query_symlink(const unsigned int xid, struct cifs_tcon *tcon, rc = SMB2_ioctl_init(tcon, &rqst[1], fid.persistent_fid, fid.volatile_fid, FSCTL_GET_REPARSE_POINT, - true /* is_fctl */, NULL, 0, CIFSMaxBufSize); + true /* is_fctl */, NULL, 0, + CIFSMaxBufSize - + MAX_SMB2_CREATE_RESPONSE_SIZE - + MAX_SMB2_CLOSE_RESPONSE_SIZE); if (rc) goto querty_exit; @@ -3078,28 +3173,32 @@ static long smb3_simple_falloc(struct file *file, struct cifs_tcon *tcon, } /* + * Extending the file + */ + if ((keep_size == false) && i_size_read(inode) < off + len) { + if ((cifsi->cifsAttrs & FILE_ATTRIBUTE_SPARSE_FILE) == 0) + smb2_set_sparse(xid, tcon, cfile, inode, false); + + eof = cpu_to_le64(off + len); + rc = SMB2_set_eof(xid, tcon, cfile->fid.persistent_fid, + cfile->fid.volatile_fid, cfile->pid, &eof); + if (rc == 0) { + cifsi->server_eof = off + len; + cifs_setsize(inode, off + len); + cifs_truncate_page(inode->i_mapping, inode->i_size); + truncate_setsize(inode, off + len); + } + goto out; + } + + /* * Files are non-sparse by default so falloc may be a no-op - * Must check if file sparse. If not sparse, and not extending - * then no need to do anything since file already allocated + * Must check if file sparse. If not sparse, and since we are not + * extending then no need to do anything since file already allocated */ if ((cifsi->cifsAttrs & FILE_ATTRIBUTE_SPARSE_FILE) == 0) { - if (keep_size == true) - rc = 0; - /* check if extending file */ - else if (i_size_read(inode) >= off + len) - /* not extending file and already not sparse */ - rc = 0; - /* BB: in future add else clause to extend file */ - else - rc = -EOPNOTSUPP; - if (rc) - trace_smb3_falloc_err(xid, cfile->fid.persistent_fid, - tcon->tid, tcon->ses->Suid, off, len, rc); - else - trace_smb3_falloc_done(xid, cfile->fid.persistent_fid, - tcon->tid, tcon->ses->Suid, off, len); - free_xid(xid); - return rc; + rc = 0; + goto out; } if ((keep_size == true) || (i_size_read(inode) >= off + len)) { @@ -3113,25 +3212,14 @@ static long smb3_simple_falloc(struct file *file, struct cifs_tcon *tcon, */ if ((off > 8192) || (off + len + 8192 < i_size_read(inode))) { rc = -EOPNOTSUPP; - trace_smb3_falloc_err(xid, cfile->fid.persistent_fid, - tcon->tid, tcon->ses->Suid, off, len, rc); - free_xid(xid); - return rc; - } - - smb2_set_sparse(xid, tcon, cfile, inode, false); - rc = 0; - } else { - smb2_set_sparse(xid, tcon, cfile, inode, false); - rc = 0; - if (i_size_read(inode) < off + len) { - eof = cpu_to_le64(off + len); - rc = SMB2_set_eof(xid, tcon, cfile->fid.persistent_fid, - cfile->fid.volatile_fid, cfile->pid, - &eof); + goto out; } } + smb2_set_sparse(xid, tcon, cfile, inode, false); + rc = 0; + +out: if (rc) trace_smb3_falloc_err(xid, cfile->fid.persistent_fid, tcon->tid, tcon->ses->Suid, off, len, rc); diff --git a/fs/cifs/smb2pdu.c b/fs/cifs/smb2pdu.c index 0ab6b1200288..7edba3e6d5e6 100644 --- a/fs/cifs/smb2pdu.c +++ b/fs/cifs/smb2pdu.c @@ -1847,7 +1847,7 @@ SMB2_tdis(const unsigned int xid, struct cifs_tcon *tcon) if ((tcon->need_reconnect) || (tcon->ses->need_reconnect)) return 0; - close_shroot(&tcon->crfid); + close_shroot_lease(&tcon->crfid); rc = smb2_plain_req_init(SMB2_TREE_DISCONNECT, tcon, (void **) &req, &total_len); @@ -2199,13 +2199,14 @@ create_sd_buf(umode_t mode, unsigned int *len) struct cifs_ace *pace; unsigned int sdlen, acelen; - *len = roundup(sizeof(struct crt_sd_ctxt) + sizeof(struct cifs_ace), 8); + *len = roundup(sizeof(struct crt_sd_ctxt) + sizeof(struct cifs_ace) * 2, + 8); buf = kzalloc(*len, GFP_KERNEL); if (buf == NULL) return buf; sdlen = sizeof(struct smb3_sd) + sizeof(struct smb3_acl) + - sizeof(struct cifs_ace); + 2 * sizeof(struct cifs_ace); buf->ccontext.DataOffset = cpu_to_le16(offsetof (struct crt_sd_ctxt, sd)); @@ -2232,8 +2233,12 @@ create_sd_buf(umode_t mode, unsigned int *len) /* create one ACE to hold the mode embedded in reserved special SID */ pace = (struct cifs_ace *)(sizeof(struct crt_sd_ctxt) + (char *)buf); acelen = setup_special_mode_ACE(pace, (__u64)mode); + /* and one more ACE to allow access for authenticated users */ + pace = (struct cifs_ace *)(acelen + (sizeof(struct crt_sd_ctxt) + + (char *)buf)); + acelen += setup_authusers_ACE(pace); buf->acl.AclSize = cpu_to_le16(sizeof(struct cifs_acl) + acelen); - buf->acl.AceCount = cpu_to_le16(1); + buf->acl.AceCount = cpu_to_le16(2); return buf; } @@ -4296,56 +4301,38 @@ num_entries(char *bufstart, char *end_of_buf, char **lastentry, size_t size) /* * Readdir/FindFirst */ -int -SMB2_query_directory(const unsigned int xid, struct cifs_tcon *tcon, - u64 persistent_fid, u64 volatile_fid, int index, - struct cifs_search_info *srch_inf) +int SMB2_query_directory_init(const unsigned int xid, + struct cifs_tcon *tcon, struct smb_rqst *rqst, + u64 persistent_fid, u64 volatile_fid, + int index, int info_level) { - struct smb_rqst rqst; + struct TCP_Server_Info *server = tcon->ses->server; struct smb2_query_directory_req *req; - struct smb2_query_directory_rsp *rsp = NULL; - struct kvec iov[2]; - struct kvec rsp_iov; - int rc = 0; - int len; - int resp_buftype = CIFS_NO_BUFFER; unsigned char *bufptr; - struct TCP_Server_Info *server; - struct cifs_ses *ses = tcon->ses; __le16 asteriks = cpu_to_le16('*'); - char *end_of_smb; - unsigned int output_size = CIFSMaxBufSize; - size_t info_buf_size; - int flags = 0; + unsigned int output_size = CIFSMaxBufSize - + MAX_SMB2_CREATE_RESPONSE_SIZE - + MAX_SMB2_CLOSE_RESPONSE_SIZE; unsigned int total_len; - - if (ses && (ses->server)) - server = ses->server; - else - return -EIO; + struct kvec *iov = rqst->rq_iov; + int len, rc; rc = smb2_plain_req_init(SMB2_QUERY_DIRECTORY, tcon, (void **) &req, &total_len); if (rc) return rc; - if (smb3_encryption_required(tcon)) - flags |= CIFS_TRANSFORM_REQ; - - switch (srch_inf->info_level) { + switch (info_level) { case SMB_FIND_FILE_DIRECTORY_INFO: req->FileInformationClass = FILE_DIRECTORY_INFORMATION; - info_buf_size = sizeof(FILE_DIRECTORY_INFO) - 1; break; case SMB_FIND_FILE_ID_FULL_DIR_INFO: req->FileInformationClass = FILEID_FULL_DIRECTORY_INFORMATION; - info_buf_size = sizeof(SEARCH_ID_FULL_DIR_INFO) - 1; break; default: cifs_tcon_dbg(VFS, "info level %u isn't supported\n", - srch_inf->info_level); - rc = -EINVAL; - goto qdir_exit; + info_level); + return -EINVAL; } req->FileIndex = cpu_to_le32(index); @@ -4374,40 +4361,50 @@ SMB2_query_directory(const unsigned int xid, struct cifs_tcon *tcon, iov[1].iov_base = (char *)(req->Buffer); iov[1].iov_len = len; - memset(&rqst, 0, sizeof(struct smb_rqst)); - rqst.rq_iov = iov; - rqst.rq_nvec = 2; - trace_smb3_query_dir_enter(xid, persistent_fid, tcon->tid, tcon->ses->Suid, index, output_size); - rc = cifs_send_recv(xid, ses, &rqst, &resp_buftype, flags, &rsp_iov); - cifs_small_buf_release(req); - rsp = (struct smb2_query_directory_rsp *)rsp_iov.iov_base; + return 0; +} - if (rc) { - if (rc == -ENODATA && - rsp->sync_hdr.Status == STATUS_NO_MORE_FILES) { - trace_smb3_query_dir_done(xid, persistent_fid, - tcon->tid, tcon->ses->Suid, index, 0); - srch_inf->endOfSearch = true; - rc = 0; - } else { - trace_smb3_query_dir_err(xid, persistent_fid, tcon->tid, - tcon->ses->Suid, index, 0, rc); - cifs_stats_fail_inc(tcon, SMB2_QUERY_DIRECTORY_HE); - } - goto qdir_exit; +void SMB2_query_directory_free(struct smb_rqst *rqst) +{ + if (rqst && rqst->rq_iov) { + cifs_small_buf_release(rqst->rq_iov[0].iov_base); /* request */ + } +} + +int +smb2_parse_query_directory(struct cifs_tcon *tcon, + struct kvec *rsp_iov, + int resp_buftype, + struct cifs_search_info *srch_inf) +{ + struct smb2_query_directory_rsp *rsp; + size_t info_buf_size; + char *end_of_smb; + int rc; + + rsp = (struct smb2_query_directory_rsp *)rsp_iov->iov_base; + + switch (srch_inf->info_level) { + case SMB_FIND_FILE_DIRECTORY_INFO: + info_buf_size = sizeof(FILE_DIRECTORY_INFO) - 1; + break; + case SMB_FIND_FILE_ID_FULL_DIR_INFO: + info_buf_size = sizeof(SEARCH_ID_FULL_DIR_INFO) - 1; + break; + default: + cifs_tcon_dbg(VFS, "info level %u isn't supported\n", + srch_inf->info_level); + return -EINVAL; } rc = smb2_validate_iov(le16_to_cpu(rsp->OutputBufferOffset), - le32_to_cpu(rsp->OutputBufferLength), &rsp_iov, + le32_to_cpu(rsp->OutputBufferLength), rsp_iov, info_buf_size); - if (rc) { - trace_smb3_query_dir_err(xid, persistent_fid, tcon->tid, - tcon->ses->Suid, index, 0, rc); - goto qdir_exit; - } + if (rc) + return rc; srch_inf->unicode = true; @@ -4420,7 +4417,7 @@ SMB2_query_directory(const unsigned int xid, struct cifs_tcon *tcon, srch_inf->ntwrk_buf_start = (char *)rsp; srch_inf->srch_entries_start = srch_inf->last_entry = (char *)rsp + le16_to_cpu(rsp->OutputBufferOffset); - end_of_smb = rsp_iov.iov_len + (char *)rsp; + end_of_smb = rsp_iov->iov_len + (char *)rsp; srch_inf->entries_in_buffer = num_entries(srch_inf->srch_entries_start, end_of_smb, &srch_inf->last_entry, info_buf_size); @@ -4435,11 +4432,72 @@ SMB2_query_directory(const unsigned int xid, struct cifs_tcon *tcon, else cifs_tcon_dbg(VFS, "illegal search buffer type\n"); + return 0; +} + +int +SMB2_query_directory(const unsigned int xid, struct cifs_tcon *tcon, + u64 persistent_fid, u64 volatile_fid, int index, + struct cifs_search_info *srch_inf) +{ + struct smb_rqst rqst; + struct kvec iov[SMB2_QUERY_DIRECTORY_IOV_SIZE]; + struct smb2_query_directory_rsp *rsp = NULL; + int resp_buftype = CIFS_NO_BUFFER; + struct kvec rsp_iov; + int rc = 0; + struct cifs_ses *ses = tcon->ses; + int flags = 0; + + if (!ses || !(ses->server)) + return -EIO; + + if (smb3_encryption_required(tcon)) + flags |= CIFS_TRANSFORM_REQ; + + memset(&rqst, 0, sizeof(struct smb_rqst)); + memset(&iov, 0, sizeof(iov)); + rqst.rq_iov = iov; + rqst.rq_nvec = SMB2_QUERY_DIRECTORY_IOV_SIZE; + + rc = SMB2_query_directory_init(xid, tcon, &rqst, persistent_fid, + volatile_fid, index, + srch_inf->info_level); + if (rc) + goto qdir_exit; + + rc = cifs_send_recv(xid, ses, &rqst, &resp_buftype, flags, &rsp_iov); + rsp = (struct smb2_query_directory_rsp *)rsp_iov.iov_base; + + if (rc) { + if (rc == -ENODATA && + rsp->sync_hdr.Status == STATUS_NO_MORE_FILES) { + trace_smb3_query_dir_done(xid, persistent_fid, + tcon->tid, tcon->ses->Suid, index, 0); + srch_inf->endOfSearch = true; + rc = 0; + } else { + trace_smb3_query_dir_err(xid, persistent_fid, tcon->tid, + tcon->ses->Suid, index, 0, rc); + cifs_stats_fail_inc(tcon, SMB2_QUERY_DIRECTORY_HE); + } + goto qdir_exit; + } + + rc = smb2_parse_query_directory(tcon, &rsp_iov, resp_buftype, + srch_inf); + if (rc) { + trace_smb3_query_dir_err(xid, persistent_fid, tcon->tid, + tcon->ses->Suid, index, 0, rc); + goto qdir_exit; + } + resp_buftype = CIFS_NO_BUFFER; + trace_smb3_query_dir_done(xid, persistent_fid, tcon->tid, tcon->ses->Suid, index, srch_inf->entries_in_buffer); - return rc; qdir_exit: + SMB2_query_directory_free(&rqst); free_rsp_buf(resp_buftype, rsp); return rc; } diff --git a/fs/cifs/smb2pdu.h b/fs/cifs/smb2pdu.h index 7b1c379fdf7a..4c43dbd1e089 100644 --- a/fs/cifs/smb2pdu.h +++ b/fs/cifs/smb2pdu.h @@ -1282,6 +1282,8 @@ struct smb2_echo_rsp { #define SMB2_INDEX_SPECIFIED 0x04 #define SMB2_REOPEN 0x10 +#define SMB2_QUERY_DIRECTORY_IOV_SIZE 2 + struct smb2_query_directory_req { struct smb2_sync_hdr sync_hdr; __le16 StructureSize; /* Must be 33 */ diff --git a/fs/cifs/smb2proto.h b/fs/cifs/smb2proto.h index a18272c987fe..6c678e00046f 100644 --- a/fs/cifs/smb2proto.h +++ b/fs/cifs/smb2proto.h @@ -70,6 +70,8 @@ extern int smb3_handle_read_data(struct TCP_Server_Info *server, extern int open_shroot(unsigned int xid, struct cifs_tcon *tcon, struct cifs_fid *pfid); extern void close_shroot(struct cached_fid *cfid); +extern void close_shroot_lease(struct cached_fid *cfid); +extern void close_shroot_lease_locked(struct cached_fid *cfid); extern void move_smb2_info_to_cifs(FILE_ALL_INFO *dst, struct smb2_file_all_info *src); extern int smb2_query_path_info(const unsigned int xid, struct cifs_tcon *tcon, @@ -195,6 +197,11 @@ extern int SMB2_echo(struct TCP_Server_Info *server); extern int SMB2_query_directory(const unsigned int xid, struct cifs_tcon *tcon, u64 persistent_fid, u64 volatile_fid, int index, struct cifs_search_info *srch_inf); +extern int SMB2_query_directory_init(unsigned int xid, struct cifs_tcon *tcon, + struct smb_rqst *rqst, + u64 persistent_fid, u64 volatile_fid, + int index, int info_level); +extern void SMB2_query_directory_free(struct smb_rqst *rqst); extern int SMB2_set_eof(const unsigned int xid, struct cifs_tcon *tcon, u64 persistent_fid, u64 volatile_fid, u32 pid, __le64 *eof); diff --git a/fs/cifs/smb2transport.c b/fs/cifs/smb2transport.c index 387c88704c52..fe6acfce3390 100644 --- a/fs/cifs/smb2transport.c +++ b/fs/cifs/smb2transport.c @@ -685,6 +685,8 @@ smb2_mid_entry_alloc(const struct smb2_sync_hdr *shdr, * The default is for the mid to be synchronous, so the * default callback just wakes up the current task. */ + get_task_struct(current); + temp->creator = current; temp->callback = cifs_wake_up_task; temp->callback_data = current; diff --git a/fs/cifs/transport.c b/fs/cifs/transport.c index 3d2e11f85cba..cb3ee916f527 100644 --- a/fs/cifs/transport.c +++ b/fs/cifs/transport.c @@ -76,6 +76,8 @@ AllocMidQEntry(const struct smb_hdr *smb_buffer, struct TCP_Server_Info *server) * The default is for the mid to be synchronous, so the * default callback just wakes up the current task. */ + get_task_struct(current); + temp->creator = current; temp->callback = cifs_wake_up_task; temp->callback_data = current; @@ -158,6 +160,7 @@ static void _cifs_mid_q_entry_release(struct kref *refcount) } } #endif + put_task_struct(midEntry->creator); mempool_free(midEntry, cifs_mid_poolp); } diff --git a/fs/cifs/xattr.c b/fs/cifs/xattr.c index db4ba8f6077e..b8299173ea7e 100644 --- a/fs/cifs/xattr.c +++ b/fs/cifs/xattr.c @@ -32,7 +32,8 @@ #include "cifs_unicode.h" #define MAX_EA_VALUE_SIZE CIFSMaxBufSize -#define CIFS_XATTR_CIFS_ACL "system.cifs_acl" +#define CIFS_XATTR_CIFS_ACL "system.cifs_acl" /* DACL only */ +#define CIFS_XATTR_CIFS_NTSD "system.cifs_ntsd" /* owner plus DACL */ #define CIFS_XATTR_ATTRIB "cifs.dosattrib" /* full name: user.cifs.dosattrib */ #define CIFS_XATTR_CREATETIME "cifs.creationtime" /* user.cifs.creationtime */ /* @@ -40,12 +41,62 @@ * confusing users and using the 20+ year old term 'cifs' when it is no longer * secure, replaced by SMB2 (then even more highly secure SMB3) many years ago */ -#define SMB3_XATTR_CIFS_ACL "system.smb3_acl" +#define SMB3_XATTR_CIFS_ACL "system.smb3_acl" /* DACL only */ +#define SMB3_XATTR_CIFS_NTSD "system.smb3_ntsd" /* owner plus DACL */ #define SMB3_XATTR_ATTRIB "smb3.dosattrib" /* full name: user.smb3.dosattrib */ #define SMB3_XATTR_CREATETIME "smb3.creationtime" /* user.smb3.creationtime */ /* BB need to add server (Samba e.g) support for security and trusted prefix */ -enum { XATTR_USER, XATTR_CIFS_ACL, XATTR_ACL_ACCESS, XATTR_ACL_DEFAULT }; +enum { XATTR_USER, XATTR_CIFS_ACL, XATTR_ACL_ACCESS, XATTR_ACL_DEFAULT, + XATTR_CIFS_NTSD }; + +static int cifs_attrib_set(unsigned int xid, struct cifs_tcon *pTcon, + struct inode *inode, char *full_path, + const void *value, size_t size) +{ + ssize_t rc = -EOPNOTSUPP; + __u32 *pattrib = (__u32 *)value; + __u32 attrib; + FILE_BASIC_INFO info_buf; + + if ((value == NULL) || (size != sizeof(__u32))) + return -ERANGE; + + memset(&info_buf, 0, sizeof(info_buf)); + attrib = *pattrib; + info_buf.Attributes = cpu_to_le32(attrib); + if (pTcon->ses->server->ops->set_file_info) + rc = pTcon->ses->server->ops->set_file_info(inode, full_path, + &info_buf, xid); + if (rc == 0) + CIFS_I(inode)->cifsAttrs = attrib; + + return rc; +} + +static int cifs_creation_time_set(unsigned int xid, struct cifs_tcon *pTcon, + struct inode *inode, char *full_path, + const void *value, size_t size) +{ + ssize_t rc = -EOPNOTSUPP; + __u64 *pcreation_time = (__u64 *)value; + __u64 creation_time; + FILE_BASIC_INFO info_buf; + + if ((value == NULL) || (size != sizeof(__u64))) + return -ERANGE; + + memset(&info_buf, 0, sizeof(info_buf)); + creation_time = *pcreation_time; + info_buf.CreationTime = cpu_to_le64(creation_time); + if (pTcon->ses->server->ops->set_file_info) + rc = pTcon->ses->server->ops->set_file_info(inode, full_path, + &info_buf, xid); + if (rc == 0) + CIFS_I(inode)->createtime = creation_time; + + return rc; +} static int cifs_xattr_set(const struct xattr_handler *handler, struct dentry *dentry, struct inode *inode, @@ -86,6 +137,23 @@ static int cifs_xattr_set(const struct xattr_handler *handler, switch (handler->flags) { case XATTR_USER: + cifs_dbg(FYI, "%s:setting user xattr %s\n", __func__, name); + if ((strcmp(name, CIFS_XATTR_ATTRIB) == 0) || + (strcmp(name, SMB3_XATTR_ATTRIB) == 0)) { + rc = cifs_attrib_set(xid, pTcon, inode, full_path, + value, size); + if (rc == 0) /* force revalidate of the inode */ + CIFS_I(inode)->time = 0; + break; + } else if ((strcmp(name, CIFS_XATTR_CREATETIME) == 0) || + (strcmp(name, SMB3_XATTR_CREATETIME) == 0)) { + rc = cifs_creation_time_set(xid, pTcon, inode, + full_path, value, size); + if (rc == 0) /* force revalidate of the inode */ + CIFS_I(inode)->time = 0; + break; + } + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_XATTR) goto out; @@ -95,7 +163,8 @@ static int cifs_xattr_set(const struct xattr_handler *handler, cifs_sb->local_nls, cifs_sb); break; - case XATTR_CIFS_ACL: { + case XATTR_CIFS_ACL: + case XATTR_CIFS_NTSD: { struct cifs_ntsd *pacl; if (!value) @@ -106,12 +175,25 @@ static int cifs_xattr_set(const struct xattr_handler *handler, } else { memcpy(pacl, value, size); if (value && - pTcon->ses->server->ops->set_acl) - rc = pTcon->ses->server->ops->set_acl(pacl, - size, inode, - full_path, CIFS_ACL_DACL); - else + pTcon->ses->server->ops->set_acl) { + rc = 0; + if (handler->flags == XATTR_CIFS_NTSD) { + /* set owner and DACL */ + rc = pTcon->ses->server->ops->set_acl( + pacl, size, inode, + full_path, + CIFS_ACL_OWNER); + } + if (rc == 0) { + /* set DACL */ + rc = pTcon->ses->server->ops->set_acl( + pacl, size, inode, + full_path, + CIFS_ACL_DACL); + } + } else { rc = -EOPNOTSUPP; + } if (rc == 0) /* force revalidate of the inode */ CIFS_I(inode)->time = 0; kfree(pacl); @@ -179,7 +261,7 @@ static int cifs_creation_time_get(struct dentry *dentry, struct inode *inode, void *value, size_t size) { ssize_t rc; - __u64 * pcreatetime; + __u64 *pcreatetime; rc = cifs_revalidate_dentry_attr(dentry); if (rc) @@ -244,7 +326,9 @@ static int cifs_xattr_get(const struct xattr_handler *handler, full_path, name, value, size, cifs_sb); break; - case XATTR_CIFS_ACL: { + case XATTR_CIFS_ACL: + case XATTR_CIFS_NTSD: { + /* the whole ntsd is fetched regardless */ u32 acllen; struct cifs_ntsd *pacl; @@ -382,6 +466,26 @@ static const struct xattr_handler smb3_acl_xattr_handler = { .set = cifs_xattr_set, }; +static const struct xattr_handler cifs_cifs_ntsd_xattr_handler = { + .name = CIFS_XATTR_CIFS_NTSD, + .flags = XATTR_CIFS_NTSD, + .get = cifs_xattr_get, + .set = cifs_xattr_set, +}; + +/* + * Although this is just an alias for the above, need to move away from + * confusing users and using the 20 year old term 'cifs' when it is no + * longer secure and was replaced by SMB2/SMB3 a long time ago, and + * SMB3 and later are highly secure. + */ +static const struct xattr_handler smb3_ntsd_xattr_handler = { + .name = SMB3_XATTR_CIFS_NTSD, + .flags = XATTR_CIFS_NTSD, + .get = cifs_xattr_get, + .set = cifs_xattr_set, +}; + static const struct xattr_handler cifs_posix_acl_access_xattr_handler = { .name = XATTR_NAME_POSIX_ACL_ACCESS, .flags = XATTR_ACL_ACCESS, @@ -401,6 +505,8 @@ const struct xattr_handler *cifs_xattr_handlers[] = { &cifs_os2_xattr_handler, &cifs_cifs_acl_xattr_handler, &smb3_acl_xattr_handler, /* alias for above since avoiding "cifs" */ + &cifs_cifs_ntsd_xattr_handler, + &smb3_ntsd_xattr_handler, /* alias for above since avoiding "cifs" */ &cifs_posix_acl_access_xattr_handler, &cifs_posix_acl_default_xattr_handler, NULL diff --git a/fs/crypto/Kconfig b/fs/crypto/Kconfig index ff5a1746cbae..8046d7c7a3e9 100644 --- a/fs/crypto/Kconfig +++ b/fs/crypto/Kconfig @@ -2,13 +2,8 @@ config FS_ENCRYPTION bool "FS Encryption (Per-file encryption)" select CRYPTO - select CRYPTO_AES - select CRYPTO_CBC - select CRYPTO_ECB - select CRYPTO_XTS - select CRYPTO_CTS - select CRYPTO_SHA512 - select CRYPTO_HMAC + select CRYPTO_HASH + select CRYPTO_SKCIPHER select KEYS help Enable encryption of files and directories. This @@ -16,3 +11,16 @@ config FS_ENCRYPTION efficient since it avoids caching the encrypted and decrypted pages in the page cache. Currently Ext4, F2FS and UBIFS make use of this feature. + +# Filesystems supporting encryption must select this if FS_ENCRYPTION. This +# allows the algorithms to be built as modules when all the filesystems are. +config FS_ENCRYPTION_ALGS + tristate + select CRYPTO_AES + select CRYPTO_CBC + select CRYPTO_CTS + select CRYPTO_ECB + select CRYPTO_HMAC + select CRYPTO_SHA256 + select CRYPTO_SHA512 + select CRYPTO_XTS diff --git a/fs/crypto/bio.c b/fs/crypto/bio.c index 1f4b8a277060..4fa18fff9c4e 100644 --- a/fs/crypto/bio.c +++ b/fs/crypto/bio.c @@ -41,53 +41,101 @@ void fscrypt_decrypt_bio(struct bio *bio) } EXPORT_SYMBOL(fscrypt_decrypt_bio); +/** + * fscrypt_zeroout_range() - zero out a range of blocks in an encrypted file + * @inode: the file's inode + * @lblk: the first file logical block to zero out + * @pblk: the first filesystem physical block to zero out + * @len: number of blocks to zero out + * + * Zero out filesystem blocks in an encrypted regular file on-disk, i.e. write + * ciphertext blocks which decrypt to the all-zeroes block. The blocks must be + * both logically and physically contiguous. It's also assumed that the + * filesystem only uses a single block device, ->s_bdev. + * + * Note that since each block uses a different IV, this involves writing a + * different ciphertext to each block; we can't simply reuse the same one. + * + * Return: 0 on success; -errno on failure. + */ int fscrypt_zeroout_range(const struct inode *inode, pgoff_t lblk, - sector_t pblk, unsigned int len) + sector_t pblk, unsigned int len) { const unsigned int blockbits = inode->i_blkbits; const unsigned int blocksize = 1 << blockbits; - struct page *ciphertext_page; + const unsigned int blocks_per_page_bits = PAGE_SHIFT - blockbits; + const unsigned int blocks_per_page = 1 << blocks_per_page_bits; + struct page *pages[16]; /* write up to 16 pages at a time */ + unsigned int nr_pages; + unsigned int i; + unsigned int offset; struct bio *bio; - int ret, err = 0; + int ret, err; - ciphertext_page = fscrypt_alloc_bounce_page(GFP_NOWAIT); - if (!ciphertext_page) - return -ENOMEM; + if (len == 0) + return 0; - while (len--) { - err = fscrypt_crypt_block(inode, FS_ENCRYPT, lblk, - ZERO_PAGE(0), ciphertext_page, - blocksize, 0, GFP_NOFS); - if (err) - goto errout; + BUILD_BUG_ON(ARRAY_SIZE(pages) > BIO_MAX_PAGES); + nr_pages = min_t(unsigned int, ARRAY_SIZE(pages), + (len + blocks_per_page - 1) >> blocks_per_page_bits); - bio = bio_alloc(GFP_NOWAIT, 1); - if (!bio) { - err = -ENOMEM; - goto errout; - } + /* + * We need at least one page for ciphertext. Allocate the first one + * from a mempool, with __GFP_DIRECT_RECLAIM set so that it can't fail. + * + * Any additional page allocations are allowed to fail, as they only + * help performance, and waiting on the mempool for them could deadlock. + */ + for (i = 0; i < nr_pages; i++) { + pages[i] = fscrypt_alloc_bounce_page(i == 0 ? GFP_NOFS : + GFP_NOWAIT | __GFP_NOWARN); + if (!pages[i]) + break; + } + nr_pages = i; + if (WARN_ON(nr_pages <= 0)) + return -EINVAL; + + /* This always succeeds since __GFP_DIRECT_RECLAIM is set. */ + bio = bio_alloc(GFP_NOFS, nr_pages); + + do { bio_set_dev(bio, inode->i_sb->s_bdev); bio->bi_iter.bi_sector = pblk << (blockbits - 9); bio_set_op_attrs(bio, REQ_OP_WRITE, 0); - ret = bio_add_page(bio, ciphertext_page, blocksize, 0); - if (WARN_ON(ret != blocksize)) { - /* should never happen! */ - bio_put(bio); - err = -EIO; - goto errout; - } + + i = 0; + offset = 0; + do { + err = fscrypt_crypt_block(inode, FS_ENCRYPT, lblk, + ZERO_PAGE(0), pages[i], + blocksize, offset, GFP_NOFS); + if (err) + goto out; + lblk++; + pblk++; + len--; + offset += blocksize; + if (offset == PAGE_SIZE || len == 0) { + ret = bio_add_page(bio, pages[i++], offset, 0); + if (WARN_ON(ret != offset)) { + err = -EIO; + goto out; + } + offset = 0; + } + } while (i != nr_pages && len != 0); + err = submit_bio_wait(bio); - if (err == 0 && bio->bi_status) - err = -EIO; - bio_put(bio); if (err) - goto errout; - lblk++; - pblk++; - } + goto out; + bio_reset(bio); + } while (len != 0); err = 0; -errout: - fscrypt_free_bounce_page(ciphertext_page); +out: + bio_put(bio); + for (i = 0; i < nr_pages; i++) + fscrypt_free_bounce_page(pages[i]); return err; } EXPORT_SYMBOL(fscrypt_zeroout_range); diff --git a/fs/crypto/crypto.c b/fs/crypto/crypto.c index 3719efa546c6..1ecaac7ee3cb 100644 --- a/fs/crypto/crypto.c +++ b/fs/crypto/crypto.c @@ -25,8 +25,6 @@ #include <linux/module.h> #include <linux/scatterlist.h> #include <linux/ratelimit.h> -#include <linux/dcache.h> -#include <linux/namei.h> #include <crypto/skcipher.h> #include "fscrypt_private.h" @@ -140,7 +138,7 @@ int fscrypt_crypt_block(const struct inode *inode, fscrypt_direction_t rw, * multiple of the filesystem's block size. * @offs: Byte offset within @page of the first block to encrypt. Must be * a multiple of the filesystem's block size. - * @gfp_flags: Memory allocation flags + * @gfp_flags: Memory allocation flags. See details below. * * A new bounce page is allocated, and the specified block(s) are encrypted into * it. In the bounce page, the ciphertext block(s) will be located at the same @@ -150,6 +148,11 @@ int fscrypt_crypt_block(const struct inode *inode, fscrypt_direction_t rw, * * This is for use by the filesystem's ->writepages() method. * + * The bounce page allocation is mempool-backed, so it will always succeed when + * @gfp_flags includes __GFP_DIRECT_RECLAIM, e.g. when it's GFP_NOFS. However, + * only the first page of each bio can be allocated this way. To prevent + * deadlocks, for any additional pages a mask like GFP_NOWAIT must be used. + * * Return: the new encrypted bounce page on success; an ERR_PTR() on failure */ struct page *fscrypt_encrypt_pagecache_blocks(struct page *page, @@ -286,54 +289,6 @@ int fscrypt_decrypt_block_inplace(const struct inode *inode, struct page *page, } EXPORT_SYMBOL(fscrypt_decrypt_block_inplace); -/* - * Validate dentries in encrypted directories to make sure we aren't potentially - * caching stale dentries after a key has been added. - */ -static int fscrypt_d_revalidate(struct dentry *dentry, unsigned int flags) -{ - struct dentry *dir; - int err; - int valid; - - /* - * Plaintext names are always valid, since fscrypt doesn't support - * reverting to ciphertext names without evicting the directory's inode - * -- which implies eviction of the dentries in the directory. - */ - if (!(dentry->d_flags & DCACHE_ENCRYPTED_NAME)) - return 1; - - /* - * Ciphertext name; valid if the directory's key is still unavailable. - * - * Although fscrypt forbids rename() on ciphertext names, we still must - * use dget_parent() here rather than use ->d_parent directly. That's - * because a corrupted fs image may contain directory hard links, which - * the VFS handles by moving the directory's dentry tree in the dcache - * each time ->lookup() finds the directory and it already has a dentry - * elsewhere. Thus ->d_parent can be changing, and we must safely grab - * a reference to some ->d_parent to prevent it from being freed. - */ - - if (flags & LOOKUP_RCU) - return -ECHILD; - - dir = dget_parent(dentry); - err = fscrypt_get_encryption_info(d_inode(dir)); - valid = !fscrypt_has_encryption_key(d_inode(dir)); - dput(dir); - - if (err < 0) - return err; - - return valid; -} - -const struct dentry_operations fscrypt_d_ops = { - .d_revalidate = fscrypt_d_revalidate, -}; - /** * fscrypt_initialize() - allocate major buffers for fs encryption. * @cop_flags: fscrypt operations flags diff --git a/fs/crypto/fname.c b/fs/crypto/fname.c index 3da3707c10e3..4c212442a8f7 100644 --- a/fs/crypto/fname.c +++ b/fs/crypto/fname.c @@ -11,10 +11,87 @@ * This has not yet undergone a rigorous security audit. */ +#include <linux/namei.h> #include <linux/scatterlist.h> +#include <crypto/hash.h> +#include <crypto/sha.h> #include <crypto/skcipher.h> #include "fscrypt_private.h" +/** + * struct fscrypt_nokey_name - identifier for directory entry when key is absent + * + * When userspace lists an encrypted directory without access to the key, the + * filesystem must present a unique "no-key name" for each filename that allows + * it to find the directory entry again if requested. Naively, that would just + * mean using the ciphertext filenames. However, since the ciphertext filenames + * can contain illegal characters ('\0' and '/'), they must be encoded in some + * way. We use base64. But that can cause names to exceed NAME_MAX (255 + * bytes), so we also need to use a strong hash to abbreviate long names. + * + * The filesystem may also need another kind of hash, the "dirhash", to quickly + * find the directory entry. Since filesystems normally compute the dirhash + * over the on-disk filename (i.e. the ciphertext), it's not computable from + * no-key names that abbreviate the ciphertext using the strong hash to fit in + * NAME_MAX. It's also not computable if it's a keyed hash taken over the + * plaintext (but it may still be available in the on-disk directory entry); + * casefolded directories use this type of dirhash. At least in these cases, + * each no-key name must include the name's dirhash too. + * + * To meet all these requirements, we base64-encode the following + * variable-length structure. It contains the dirhash, or 0's if the filesystem + * didn't provide one; up to 149 bytes of the ciphertext name; and for + * ciphertexts longer than 149 bytes, also the SHA-256 of the remaining bytes. + * + * This ensures that each no-key name contains everything needed to find the + * directory entry again, contains only legal characters, doesn't exceed + * NAME_MAX, is unambiguous unless there's a SHA-256 collision, and that we only + * take the performance hit of SHA-256 on very long filenames (which are rare). + */ +struct fscrypt_nokey_name { + u32 dirhash[2]; + u8 bytes[149]; + u8 sha256[SHA256_DIGEST_SIZE]; +}; /* 189 bytes => 252 bytes base64-encoded, which is <= NAME_MAX (255) */ + +/* + * Decoded size of max-size nokey name, i.e. a name that was abbreviated using + * the strong hash and thus includes the 'sha256' field. This isn't simply + * sizeof(struct fscrypt_nokey_name), as the padding at the end isn't included. + */ +#define FSCRYPT_NOKEY_NAME_MAX offsetofend(struct fscrypt_nokey_name, sha256) + +static struct crypto_shash *sha256_hash_tfm; + +static int fscrypt_do_sha256(const u8 *data, unsigned int data_len, u8 *result) +{ + struct crypto_shash *tfm = READ_ONCE(sha256_hash_tfm); + + if (unlikely(!tfm)) { + struct crypto_shash *prev_tfm; + + tfm = crypto_alloc_shash("sha256", 0, 0); + if (IS_ERR(tfm)) { + fscrypt_err(NULL, + "Error allocating SHA-256 transform: %ld", + PTR_ERR(tfm)); + return PTR_ERR(tfm); + } + prev_tfm = cmpxchg(&sha256_hash_tfm, NULL, tfm); + if (prev_tfm) { + crypto_free_shash(tfm); + tfm = prev_tfm; + } + } + { + SHASH_DESC_ON_STACK(desc, tfm); + + desc->tfm = tfm; + + return crypto_shash_digest(desc, data, data_len, result); + } +} + static inline bool fscrypt_is_dot_dotdot(const struct qstr *str) { if (str->len == 1 && str->name[0] == '.') @@ -27,19 +104,19 @@ static inline bool fscrypt_is_dot_dotdot(const struct qstr *str) } /** - * fname_encrypt() - encrypt a filename + * fscrypt_fname_encrypt() - encrypt a filename * * The output buffer must be at least as large as the input buffer. * Any extra space is filled with NUL padding before encryption. * * Return: 0 on success, -errno on failure */ -int fname_encrypt(struct inode *inode, const struct qstr *iname, - u8 *out, unsigned int olen) +int fscrypt_fname_encrypt(const struct inode *inode, const struct qstr *iname, + u8 *out, unsigned int olen) { struct skcipher_request *req = NULL; DECLARE_CRYPTO_WAIT(wait); - struct fscrypt_info *ci = inode->i_crypt_info; + const struct fscrypt_info *ci = inode->i_crypt_info; struct crypto_skcipher *tfm = ci->ci_ctfm; union fscrypt_iv iv; struct scatterlist sg; @@ -85,14 +162,14 @@ int fname_encrypt(struct inode *inode, const struct qstr *iname, * * Return: 0 on success, -errno on failure */ -static int fname_decrypt(struct inode *inode, - const struct fscrypt_str *iname, - struct fscrypt_str *oname) +static int fname_decrypt(const struct inode *inode, + const struct fscrypt_str *iname, + struct fscrypt_str *oname) { struct skcipher_request *req = NULL; DECLARE_CRYPTO_WAIT(wait); struct scatterlist src_sg, dst_sg; - struct fscrypt_info *ci = inode->i_crypt_info; + const struct fscrypt_info *ci = inode->i_crypt_info; struct crypto_skcipher *tfm = ci->ci_ctfm; union fscrypt_iv iv; int res; @@ -206,9 +283,7 @@ int fscrypt_fname_alloc_buffer(const struct inode *inode, u32 max_encrypted_len, struct fscrypt_str *crypto_str) { - const u32 max_encoded_len = - max_t(u32, BASE64_CHARS(FSCRYPT_FNAME_MAX_UNDIGESTED_SIZE), - 1 + BASE64_CHARS(sizeof(struct fscrypt_digested_name))); + const u32 max_encoded_len = BASE64_CHARS(FSCRYPT_NOKEY_NAME_MAX); u32 max_presented_len; max_presented_len = max(max_encoded_len, max_encrypted_len); @@ -241,19 +316,21 @@ EXPORT_SYMBOL(fscrypt_fname_free_buffer); * * The caller must have allocated sufficient memory for the @oname string. * - * If the key is available, we'll decrypt the disk name; otherwise, we'll encode - * it for presentation. Short names are directly base64-encoded, while long - * names are encoded in fscrypt_digested_name format. + * If the key is available, we'll decrypt the disk name. Otherwise, we'll + * encode it for presentation in fscrypt_nokey_name format. + * See struct fscrypt_nokey_name for details. * * Return: 0 on success, -errno on failure */ -int fscrypt_fname_disk_to_usr(struct inode *inode, - u32 hash, u32 minor_hash, - const struct fscrypt_str *iname, - struct fscrypt_str *oname) +int fscrypt_fname_disk_to_usr(const struct inode *inode, + u32 hash, u32 minor_hash, + const struct fscrypt_str *iname, + struct fscrypt_str *oname) { const struct qstr qname = FSTR_TO_QSTR(iname); - struct fscrypt_digested_name digested_name; + struct fscrypt_nokey_name nokey_name; + u32 size; /* size of the unencoded no-key name */ + int err; if (fscrypt_is_dot_dotdot(&qname)) { oname->name[0] = '.'; @@ -268,24 +345,37 @@ int fscrypt_fname_disk_to_usr(struct inode *inode, if (fscrypt_has_encryption_key(inode)) return fname_decrypt(inode, iname, oname); - if (iname->len <= FSCRYPT_FNAME_MAX_UNDIGESTED_SIZE) { - oname->len = base64_encode(iname->name, iname->len, - oname->name); - return 0; - } + /* + * Sanity check that struct fscrypt_nokey_name doesn't have padding + * between fields and that its encoded size never exceeds NAME_MAX. + */ + BUILD_BUG_ON(offsetofend(struct fscrypt_nokey_name, dirhash) != + offsetof(struct fscrypt_nokey_name, bytes)); + BUILD_BUG_ON(offsetofend(struct fscrypt_nokey_name, bytes) != + offsetof(struct fscrypt_nokey_name, sha256)); + BUILD_BUG_ON(BASE64_CHARS(FSCRYPT_NOKEY_NAME_MAX) > NAME_MAX); + if (hash) { - digested_name.hash = hash; - digested_name.minor_hash = minor_hash; + nokey_name.dirhash[0] = hash; + nokey_name.dirhash[1] = minor_hash; + } else { + nokey_name.dirhash[0] = 0; + nokey_name.dirhash[1] = 0; + } + if (iname->len <= sizeof(nokey_name.bytes)) { + memcpy(nokey_name.bytes, iname->name, iname->len); + size = offsetof(struct fscrypt_nokey_name, bytes[iname->len]); } else { - digested_name.hash = 0; - digested_name.minor_hash = 0; + memcpy(nokey_name.bytes, iname->name, sizeof(nokey_name.bytes)); + /* Compute strong hash of remaining part of name. */ + err = fscrypt_do_sha256(&iname->name[sizeof(nokey_name.bytes)], + iname->len - sizeof(nokey_name.bytes), + nokey_name.sha256); + if (err) + return err; + size = FSCRYPT_NOKEY_NAME_MAX; } - memcpy(digested_name.digest, - FSCRYPT_FNAME_DIGEST(iname->name, iname->len), - FSCRYPT_FNAME_DIGEST_SIZE); - oname->name[0] = '_'; - oname->len = 1 + base64_encode((const u8 *)&digested_name, - sizeof(digested_name), oname->name + 1); + oname->len = base64_encode((const u8 *)&nokey_name, size, oname->name); return 0; } EXPORT_SYMBOL(fscrypt_fname_disk_to_usr); @@ -306,8 +396,7 @@ EXPORT_SYMBOL(fscrypt_fname_disk_to_usr); * get the disk_name. * * Else, for keyless @lookup operations, @iname is the presented ciphertext, so - * we decode it to get either the ciphertext disk_name (for short names) or the - * fscrypt_digested_name (for long names). Non-@lookup operations will be + * we decode it to get the fscrypt_nokey_name. Non-@lookup operations will be * impossible in this case, so we fail them with ENOKEY. * * If successful, fscrypt_free_filename() must be called later to clean up. @@ -317,8 +406,8 @@ EXPORT_SYMBOL(fscrypt_fname_disk_to_usr); int fscrypt_setup_filename(struct inode *dir, const struct qstr *iname, int lookup, struct fscrypt_name *fname) { + struct fscrypt_nokey_name *nokey_name; int ret; - int digested; memset(fname, 0, sizeof(struct fscrypt_name)); fname->usr_fname = iname; @@ -342,8 +431,8 @@ int fscrypt_setup_filename(struct inode *dir, const struct qstr *iname, if (!fname->crypto_buf.name) return -ENOMEM; - ret = fname_encrypt(dir, iname, fname->crypto_buf.name, - fname->crypto_buf.len); + ret = fscrypt_fname_encrypt(dir, iname, fname->crypto_buf.name, + fname->crypto_buf.len); if (ret) goto errout; fname->disk_name.name = fname->crypto_buf.name; @@ -358,40 +447,31 @@ int fscrypt_setup_filename(struct inode *dir, const struct qstr *iname, * We don't have the key and we are doing a lookup; decode the * user-supplied name */ - if (iname->name[0] == '_') { - if (iname->len != - 1 + BASE64_CHARS(sizeof(struct fscrypt_digested_name))) - return -ENOENT; - digested = 1; - } else { - if (iname->len > - BASE64_CHARS(FSCRYPT_FNAME_MAX_UNDIGESTED_SIZE)) - return -ENOENT; - digested = 0; - } - fname->crypto_buf.name = - kmalloc(max_t(size_t, FSCRYPT_FNAME_MAX_UNDIGESTED_SIZE, - sizeof(struct fscrypt_digested_name)), - GFP_KERNEL); + if (iname->len > BASE64_CHARS(FSCRYPT_NOKEY_NAME_MAX)) + return -ENOENT; + + fname->crypto_buf.name = kmalloc(FSCRYPT_NOKEY_NAME_MAX, GFP_KERNEL); if (fname->crypto_buf.name == NULL) return -ENOMEM; - ret = base64_decode(iname->name + digested, iname->len - digested, - fname->crypto_buf.name); - if (ret < 0) { + ret = base64_decode(iname->name, iname->len, fname->crypto_buf.name); + if (ret < (int)offsetof(struct fscrypt_nokey_name, bytes[1]) || + (ret > offsetof(struct fscrypt_nokey_name, sha256) && + ret != FSCRYPT_NOKEY_NAME_MAX)) { ret = -ENOENT; goto errout; } fname->crypto_buf.len = ret; - if (digested) { - const struct fscrypt_digested_name *n = - (const void *)fname->crypto_buf.name; - fname->hash = n->hash; - fname->minor_hash = n->minor_hash; - } else { - fname->disk_name.name = fname->crypto_buf.name; - fname->disk_name.len = fname->crypto_buf.len; + + nokey_name = (void *)fname->crypto_buf.name; + fname->hash = nokey_name->dirhash[0]; + fname->minor_hash = nokey_name->dirhash[1]; + if (ret != FSCRYPT_NOKEY_NAME_MAX) { + /* The full ciphertext filename is available. */ + fname->disk_name.name = nokey_name->bytes; + fname->disk_name.len = + ret - offsetof(struct fscrypt_nokey_name, bytes); } return 0; @@ -400,3 +480,109 @@ errout: return ret; } EXPORT_SYMBOL(fscrypt_setup_filename); + +/** + * fscrypt_match_name() - test whether the given name matches a directory entry + * @fname: the name being searched for + * @de_name: the name from the directory entry + * @de_name_len: the length of @de_name in bytes + * + * Normally @fname->disk_name will be set, and in that case we simply compare + * that to the name stored in the directory entry. The only exception is that + * if we don't have the key for an encrypted directory and the name we're + * looking for is very long, then we won't have the full disk_name and instead + * we'll need to match against a fscrypt_nokey_name that includes a strong hash. + * + * Return: %true if the name matches, otherwise %false. + */ +bool fscrypt_match_name(const struct fscrypt_name *fname, + const u8 *de_name, u32 de_name_len) +{ + const struct fscrypt_nokey_name *nokey_name = + (const void *)fname->crypto_buf.name; + u8 sha256[SHA256_DIGEST_SIZE]; + + if (likely(fname->disk_name.name)) { + if (de_name_len != fname->disk_name.len) + return false; + return !memcmp(de_name, fname->disk_name.name, de_name_len); + } + if (de_name_len <= sizeof(nokey_name->bytes)) + return false; + if (memcmp(de_name, nokey_name->bytes, sizeof(nokey_name->bytes))) + return false; + if (fscrypt_do_sha256(&de_name[sizeof(nokey_name->bytes)], + de_name_len - sizeof(nokey_name->bytes), sha256)) + return false; + return !memcmp(sha256, nokey_name->sha256, sizeof(sha256)); +} +EXPORT_SYMBOL_GPL(fscrypt_match_name); + +/** + * fscrypt_fname_siphash() - calculate the SipHash of a filename + * @dir: the parent directory + * @name: the filename to calculate the SipHash of + * + * Given a plaintext filename @name and a directory @dir which uses SipHash as + * its dirhash method and has had its fscrypt key set up, this function + * calculates the SipHash of that name using the directory's secret dirhash key. + * + * Return: the SipHash of @name using the hash key of @dir + */ +u64 fscrypt_fname_siphash(const struct inode *dir, const struct qstr *name) +{ + const struct fscrypt_info *ci = dir->i_crypt_info; + + WARN_ON(!ci->ci_dirhash_key_initialized); + + return siphash(name->name, name->len, &ci->ci_dirhash_key); +} +EXPORT_SYMBOL_GPL(fscrypt_fname_siphash); + +/* + * Validate dentries in encrypted directories to make sure we aren't potentially + * caching stale dentries after a key has been added. + */ +static int fscrypt_d_revalidate(struct dentry *dentry, unsigned int flags) +{ + struct dentry *dir; + int err; + int valid; + + /* + * Plaintext names are always valid, since fscrypt doesn't support + * reverting to ciphertext names without evicting the directory's inode + * -- which implies eviction of the dentries in the directory. + */ + if (!(dentry->d_flags & DCACHE_ENCRYPTED_NAME)) + return 1; + + /* + * Ciphertext name; valid if the directory's key is still unavailable. + * + * Although fscrypt forbids rename() on ciphertext names, we still must + * use dget_parent() here rather than use ->d_parent directly. That's + * because a corrupted fs image may contain directory hard links, which + * the VFS handles by moving the directory's dentry tree in the dcache + * each time ->lookup() finds the directory and it already has a dentry + * elsewhere. Thus ->d_parent can be changing, and we must safely grab + * a reference to some ->d_parent to prevent it from being freed. + */ + + if (flags & LOOKUP_RCU) + return -ECHILD; + + dir = dget_parent(dentry); + err = fscrypt_get_encryption_info(d_inode(dir)); + valid = !fscrypt_has_encryption_key(d_inode(dir)); + dput(dir); + + if (err < 0) + return err; + + return valid; +} + +const struct dentry_operations fscrypt_d_ops = { + .d_revalidate = fscrypt_d_revalidate, +}; diff --git a/fs/crypto/fscrypt_private.h b/fs/crypto/fscrypt_private.h index 130b50e5a011..9aae851409e5 100644 --- a/fs/crypto/fscrypt_private.h +++ b/fs/crypto/fscrypt_private.h @@ -12,6 +12,7 @@ #define _FSCRYPT_PRIVATE_H #include <linux/fscrypt.h> +#include <linux/siphash.h> #include <crypto/hash.h> #define CONST_STRLEN(str) (sizeof(str) - 1) @@ -136,12 +137,6 @@ fscrypt_policy_flags(const union fscrypt_policy *policy) BUG(); } -static inline bool -fscrypt_is_direct_key_policy(const union fscrypt_policy *policy) -{ - return fscrypt_policy_flags(policy) & FSCRYPT_POLICY_FLAG_DIRECT_KEY; -} - /** * For encrypted symlinks, the ciphertext length is stored at the beginning * of the string in little-endian format. @@ -194,6 +189,14 @@ struct fscrypt_info { */ struct fscrypt_direct_key *ci_direct_key; + /* + * This inode's hash key for filenames. This is a 128-bit SipHash-2-4 + * key. This is only set for directories that use a keyed dirhash over + * the plaintext filenames -- currently just casefolded directories. + */ + siphash_key_t ci_dirhash_key; + bool ci_dirhash_key_initialized; + /* The encryption policy used by this inode */ union fscrypt_policy ci_policy; @@ -206,24 +209,6 @@ typedef enum { FS_ENCRYPT, } fscrypt_direction_t; -static inline bool fscrypt_valid_enc_modes(u32 contents_mode, - u32 filenames_mode) -{ - if (contents_mode == FSCRYPT_MODE_AES_128_CBC && - filenames_mode == FSCRYPT_MODE_AES_128_CTS) - return true; - - if (contents_mode == FSCRYPT_MODE_AES_256_XTS && - filenames_mode == FSCRYPT_MODE_AES_256_CTS) - return true; - - if (contents_mode == FSCRYPT_MODE_ADIANTUM && - filenames_mode == FSCRYPT_MODE_ADIANTUM) - return true; - - return false; -} - /* crypto.c */ extern struct kmem_cache *fscrypt_info_cachep; extern int fscrypt_initialize(unsigned int cop_flags); @@ -233,7 +218,6 @@ extern int fscrypt_crypt_block(const struct inode *inode, unsigned int len, unsigned int offs, gfp_t gfp_flags); extern struct page *fscrypt_alloc_bounce_page(gfp_t gfp_flags); -extern const struct dentry_operations fscrypt_d_ops; extern void __printf(3, 4) __cold fscrypt_msg(const struct inode *inode, const char *level, const char *fmt, ...); @@ -260,11 +244,13 @@ void fscrypt_generate_iv(union fscrypt_iv *iv, u64 lblk_num, const struct fscrypt_info *ci); /* fname.c */ -extern int fname_encrypt(struct inode *inode, const struct qstr *iname, - u8 *out, unsigned int olen); +extern int fscrypt_fname_encrypt(const struct inode *inode, + const struct qstr *iname, + u8 *out, unsigned int olen); extern bool fscrypt_fname_encrypted_size(const struct inode *inode, u32 orig_len, u32 max_len, u32 *encrypted_len_ret); +extern const struct dentry_operations fscrypt_d_ops; /* hkdf.c */ @@ -283,11 +269,12 @@ extern int fscrypt_init_hkdf(struct fscrypt_hkdf *hkdf, const u8 *master_key, * output doesn't reveal another. */ #define HKDF_CONTEXT_KEY_IDENTIFIER 1 -#define HKDF_CONTEXT_PER_FILE_KEY 2 +#define HKDF_CONTEXT_PER_FILE_ENC_KEY 2 #define HKDF_CONTEXT_DIRECT_KEY 3 #define HKDF_CONTEXT_IV_INO_LBLK_64_KEY 4 +#define HKDF_CONTEXT_DIRHASH_KEY 5 -extern int fscrypt_hkdf_expand(struct fscrypt_hkdf *hkdf, u8 context, +extern int fscrypt_hkdf_expand(const struct fscrypt_hkdf *hkdf, u8 context, const u8 *info, unsigned int infolen, u8 *okm, unsigned int okmlen); @@ -448,18 +435,17 @@ struct fscrypt_mode { int logged_impl_name; }; -static inline bool -fscrypt_mode_supports_direct_key(const struct fscrypt_mode *mode) -{ - return mode->ivsize >= offsetofend(union fscrypt_iv, nonce); -} +extern struct fscrypt_mode fscrypt_modes[]; extern struct crypto_skcipher * fscrypt_allocate_skcipher(struct fscrypt_mode *mode, const u8 *raw_key, const struct inode *inode); -extern int fscrypt_set_derived_key(struct fscrypt_info *ci, - const u8 *derived_key); +extern int fscrypt_set_per_file_enc_key(struct fscrypt_info *ci, + const u8 *raw_key); + +extern int fscrypt_derive_dirhash_key(struct fscrypt_info *ci, + const struct fscrypt_master_key *mk); /* keysetup_v1.c */ diff --git a/fs/crypto/hkdf.c b/fs/crypto/hkdf.c index f21873e1b467..efb95bd19a89 100644 --- a/fs/crypto/hkdf.c +++ b/fs/crypto/hkdf.c @@ -112,7 +112,7 @@ out: * adds to its application-specific info strings to guarantee that it doesn't * accidentally repeat an info string when using HKDF for different purposes.) */ -int fscrypt_hkdf_expand(struct fscrypt_hkdf *hkdf, u8 context, +int fscrypt_hkdf_expand(const struct fscrypt_hkdf *hkdf, u8 context, const u8 *info, unsigned int infolen, u8 *okm, unsigned int okmlen) { diff --git a/fs/crypto/hooks.c b/fs/crypto/hooks.c index bb3b7fcfdd48..5ef861742921 100644 --- a/fs/crypto/hooks.c +++ b/fs/crypto/hooks.c @@ -5,6 +5,8 @@ * Encryption hooks for higher-level filesystem operations. */ +#include <linux/key.h> + #include "fscrypt_private.h" /** @@ -122,6 +124,48 @@ int __fscrypt_prepare_lookup(struct inode *dir, struct dentry *dentry, } EXPORT_SYMBOL_GPL(__fscrypt_prepare_lookup); +/** + * fscrypt_prepare_setflags() - prepare to change flags with FS_IOC_SETFLAGS + * @inode: the inode on which flags are being changed + * @oldflags: the old flags + * @flags: the new flags + * + * The caller should be holding i_rwsem for write. + * + * Return: 0 on success; -errno if the flags change isn't allowed or if + * another error occurs. + */ +int fscrypt_prepare_setflags(struct inode *inode, + unsigned int oldflags, unsigned int flags) +{ + struct fscrypt_info *ci; + struct fscrypt_master_key *mk; + int err; + + /* + * When the CASEFOLD flag is set on an encrypted directory, we must + * derive the secret key needed for the dirhash. This is only possible + * if the directory uses a v2 encryption policy. + */ + if (IS_ENCRYPTED(inode) && (flags & ~oldflags & FS_CASEFOLD_FL)) { + err = fscrypt_require_key(inode); + if (err) + return err; + ci = inode->i_crypt_info; + if (ci->ci_policy.version != FSCRYPT_POLICY_V2) + return -EINVAL; + mk = ci->ci_master_key->payload.data[0]; + down_read(&mk->mk_secret_sem); + if (is_master_key_secret_present(&mk->mk_secret)) + err = fscrypt_derive_dirhash_key(ci, mk); + else + err = -ENOKEY; + up_read(&mk->mk_secret_sem); + return err; + } + return 0; +} + int __fscrypt_prepare_symlink(struct inode *dir, unsigned int len, unsigned int max_len, struct fscrypt_str *disk_link) @@ -188,7 +232,8 @@ int __fscrypt_encrypt_symlink(struct inode *inode, const char *target, ciphertext_len = disk_link->len - sizeof(*sd); sd->len = cpu_to_le16(ciphertext_len); - err = fname_encrypt(inode, &iname, sd->encrypted_path, ciphertext_len); + err = fscrypt_fname_encrypt(inode, &iname, sd->encrypted_path, + ciphertext_len); if (err) goto err_free_sd; diff --git a/fs/crypto/keyring.c b/fs/crypto/keyring.c index 040df1f5e1c8..ab41b25d4fa1 100644 --- a/fs/crypto/keyring.c +++ b/fs/crypto/keyring.c @@ -151,7 +151,7 @@ static struct key *search_fscrypt_keyring(struct key *keyring, } #define FSCRYPT_FS_KEYRING_DESCRIPTION_SIZE \ - (CONST_STRLEN("fscrypt-") + FIELD_SIZEOF(struct super_block, s_id)) + (CONST_STRLEN("fscrypt-") + sizeof_field(struct super_block, s_id)) #define FSCRYPT_MK_DESCRIPTION_SIZE (2 * FSCRYPT_KEY_IDENTIFIER_SIZE + 1) @@ -465,6 +465,109 @@ out_unlock: return err; } +static int fscrypt_provisioning_key_preparse(struct key_preparsed_payload *prep) +{ + const struct fscrypt_provisioning_key_payload *payload = prep->data; + + if (prep->datalen < sizeof(*payload) + FSCRYPT_MIN_KEY_SIZE || + prep->datalen > sizeof(*payload) + FSCRYPT_MAX_KEY_SIZE) + return -EINVAL; + + if (payload->type != FSCRYPT_KEY_SPEC_TYPE_DESCRIPTOR && + payload->type != FSCRYPT_KEY_SPEC_TYPE_IDENTIFIER) + return -EINVAL; + + if (payload->__reserved) + return -EINVAL; + + prep->payload.data[0] = kmemdup(payload, prep->datalen, GFP_KERNEL); + if (!prep->payload.data[0]) + return -ENOMEM; + + prep->quotalen = prep->datalen; + return 0; +} + +static void fscrypt_provisioning_key_free_preparse( + struct key_preparsed_payload *prep) +{ + kzfree(prep->payload.data[0]); +} + +static void fscrypt_provisioning_key_describe(const struct key *key, + struct seq_file *m) +{ + seq_puts(m, key->description); + if (key_is_positive(key)) { + const struct fscrypt_provisioning_key_payload *payload = + key->payload.data[0]; + + seq_printf(m, ": %u [%u]", key->datalen, payload->type); + } +} + +static void fscrypt_provisioning_key_destroy(struct key *key) +{ + kzfree(key->payload.data[0]); +} + +static struct key_type key_type_fscrypt_provisioning = { + .name = "fscrypt-provisioning", + .preparse = fscrypt_provisioning_key_preparse, + .free_preparse = fscrypt_provisioning_key_free_preparse, + .instantiate = generic_key_instantiate, + .describe = fscrypt_provisioning_key_describe, + .destroy = fscrypt_provisioning_key_destroy, +}; + +/* + * Retrieve the raw key from the Linux keyring key specified by 'key_id', and + * store it into 'secret'. + * + * The key must be of type "fscrypt-provisioning" and must have the field + * fscrypt_provisioning_key_payload::type set to 'type', indicating that it's + * only usable with fscrypt with the particular KDF version identified by + * 'type'. We don't use the "logon" key type because there's no way to + * completely restrict the use of such keys; they can be used by any kernel API + * that accepts "logon" keys and doesn't require a specific service prefix. + * + * The ability to specify the key via Linux keyring key is intended for cases + * where userspace needs to re-add keys after the filesystem is unmounted and + * re-mounted. Most users should just provide the raw key directly instead. + */ +static int get_keyring_key(u32 key_id, u32 type, + struct fscrypt_master_key_secret *secret) +{ + key_ref_t ref; + struct key *key; + const struct fscrypt_provisioning_key_payload *payload; + int err; + + ref = lookup_user_key(key_id, 0, KEY_NEED_SEARCH); + if (IS_ERR(ref)) + return PTR_ERR(ref); + key = key_ref_to_ptr(ref); + + if (key->type != &key_type_fscrypt_provisioning) + goto bad_key; + payload = key->payload.data[0]; + + /* Don't allow fscrypt v1 keys to be used as v2 keys and vice versa. */ + if (payload->type != type) + goto bad_key; + + secret->size = key->datalen - sizeof(*payload); + memcpy(secret->raw, payload->raw, secret->size); + err = 0; + goto out_put; + +bad_key: + err = -EKEYREJECTED; +out_put: + key_ref_put(ref); + return err; +} + /* * Add a master encryption key to the filesystem, causing all files which were * encrypted with it to appear "unlocked" (decrypted) when accessed. @@ -503,18 +606,25 @@ int fscrypt_ioctl_add_key(struct file *filp, void __user *_uarg) if (!valid_key_spec(&arg.key_spec)) return -EINVAL; - if (arg.raw_size < FSCRYPT_MIN_KEY_SIZE || - arg.raw_size > FSCRYPT_MAX_KEY_SIZE) - return -EINVAL; - if (memchr_inv(arg.__reserved, 0, sizeof(arg.__reserved))) return -EINVAL; memset(&secret, 0, sizeof(secret)); - secret.size = arg.raw_size; - err = -EFAULT; - if (copy_from_user(secret.raw, uarg->raw, secret.size)) - goto out_wipe_secret; + if (arg.key_id) { + if (arg.raw_size != 0) + return -EINVAL; + err = get_keyring_key(arg.key_id, arg.key_spec.type, &secret); + if (err) + goto out_wipe_secret; + } else { + if (arg.raw_size < FSCRYPT_MIN_KEY_SIZE || + arg.raw_size > FSCRYPT_MAX_KEY_SIZE) + return -EINVAL; + secret.size = arg.raw_size; + err = -EFAULT; + if (copy_from_user(secret.raw, uarg->raw, secret.size)) + goto out_wipe_secret; + } switch (arg.key_spec.type) { case FSCRYPT_KEY_SPEC_TYPE_DESCRIPTOR: @@ -666,9 +776,6 @@ static int check_for_busy_inodes(struct super_block *sb, struct list_head *pos; size_t busy_count = 0; unsigned long ino; - struct dentry *dentry; - char _path[256]; - char *path = NULL; spin_lock(&mk->mk_decrypted_inodes_lock); @@ -687,22 +794,14 @@ static int check_for_busy_inodes(struct super_block *sb, struct fscrypt_info, ci_master_key_link)->ci_inode; ino = inode->i_ino; - dentry = d_find_alias(inode); } spin_unlock(&mk->mk_decrypted_inodes_lock); - if (dentry) { - path = dentry_path(dentry, _path, sizeof(_path)); - dput(dentry); - } - if (IS_ERR_OR_NULL(path)) - path = "(unknown)"; - fscrypt_warn(NULL, - "%s: %zu inode(s) still busy after removing key with %s %*phN, including ino %lu (%s)", + "%s: %zu inode(s) still busy after removing key with %s %*phN, including ino %lu", sb->s_id, busy_count, master_key_spec_type(&mk->mk_spec), master_key_spec_len(&mk->mk_spec), (u8 *)&mk->mk_spec.u, - ino, path); + ino); return -EBUSY; } @@ -978,8 +1077,14 @@ int __init fscrypt_init_keyring(void) if (err) goto err_unregister_fscrypt; + err = register_key_type(&key_type_fscrypt_provisioning); + if (err) + goto err_unregister_fscrypt_user; + return 0; +err_unregister_fscrypt_user: + unregister_key_type(&key_type_fscrypt_user); err_unregister_fscrypt: unregister_key_type(&key_type_fscrypt); return err; diff --git a/fs/crypto/keysetup.c b/fs/crypto/keysetup.c index f577bb6613f9..65cb09fa6ead 100644 --- a/fs/crypto/keysetup.c +++ b/fs/crypto/keysetup.c @@ -13,7 +13,7 @@ #include "fscrypt_private.h" -static struct fscrypt_mode available_modes[] = { +struct fscrypt_mode fscrypt_modes[] = { [FSCRYPT_MODE_AES_256_XTS] = { .friendly_name = "AES-256-XTS", .cipher_str = "xts(aes)", @@ -51,10 +51,10 @@ select_encryption_mode(const union fscrypt_policy *policy, const struct inode *inode) { if (S_ISREG(inode->i_mode)) - return &available_modes[fscrypt_policy_contents_mode(policy)]; + return &fscrypt_modes[fscrypt_policy_contents_mode(policy)]; if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode)) - return &available_modes[fscrypt_policy_fnames_mode(policy)]; + return &fscrypt_modes[fscrypt_policy_fnames_mode(policy)]; WARN_ONCE(1, "fscrypt: filesystem tried to load encryption info for inode %lu, which is not encryptable (file type %d)\n", inode->i_ino, (inode->i_mode & S_IFMT)); @@ -89,8 +89,11 @@ struct crypto_skcipher *fscrypt_allocate_skcipher(struct fscrypt_mode *mode, * first time a mode is used. */ pr_info("fscrypt: %s using implementation \"%s\"\n", - mode->friendly_name, - crypto_skcipher_alg(tfm)->base.cra_driver_name); + mode->friendly_name, crypto_skcipher_driver_name(tfm)); + } + if (WARN_ON(crypto_skcipher_ivsize(tfm) != mode->ivsize)) { + err = -EINVAL; + goto err_free_tfm; } crypto_skcipher_set_flags(tfm, CRYPTO_TFM_REQ_FORBID_WEAK_KEYS); err = crypto_skcipher_setkey(tfm, raw_key, mode->keysize); @@ -104,12 +107,12 @@ err_free_tfm: return ERR_PTR(err); } -/* Given the per-file key, set up the file's crypto transform object */ -int fscrypt_set_derived_key(struct fscrypt_info *ci, const u8 *derived_key) +/* Given a per-file encryption key, set up the file's crypto transform object */ +int fscrypt_set_per_file_enc_key(struct fscrypt_info *ci, const u8 *raw_key) { struct crypto_skcipher *tfm; - tfm = fscrypt_allocate_skcipher(ci->ci_mode, derived_key, ci->ci_inode); + tfm = fscrypt_allocate_skcipher(ci->ci_mode, raw_key, ci->ci_inode); if (IS_ERR(tfm)) return PTR_ERR(tfm); @@ -118,15 +121,15 @@ int fscrypt_set_derived_key(struct fscrypt_info *ci, const u8 *derived_key) return 0; } -static int setup_per_mode_key(struct fscrypt_info *ci, - struct fscrypt_master_key *mk, - struct crypto_skcipher **tfms, - u8 hkdf_context, bool include_fs_uuid) +static int setup_per_mode_enc_key(struct fscrypt_info *ci, + struct fscrypt_master_key *mk, + struct crypto_skcipher **tfms, + u8 hkdf_context, bool include_fs_uuid) { const struct inode *inode = ci->ci_inode; const struct super_block *sb = inode->i_sb; struct fscrypt_mode *mode = ci->ci_mode; - u8 mode_num = mode - available_modes; + const u8 mode_num = mode - fscrypt_modes; struct crypto_skcipher *tfm, *prev_tfm; u8 mode_key[FSCRYPT_MAX_KEY_SIZE]; u8 hkdf_info[sizeof(mode_num) + sizeof(sb->s_uuid)]; @@ -171,29 +174,37 @@ done: return 0; } +int fscrypt_derive_dirhash_key(struct fscrypt_info *ci, + const struct fscrypt_master_key *mk) +{ + int err; + + err = fscrypt_hkdf_expand(&mk->mk_secret.hkdf, HKDF_CONTEXT_DIRHASH_KEY, + ci->ci_nonce, FS_KEY_DERIVATION_NONCE_SIZE, + (u8 *)&ci->ci_dirhash_key, + sizeof(ci->ci_dirhash_key)); + if (err) + return err; + ci->ci_dirhash_key_initialized = true; + return 0; +} + static int fscrypt_setup_v2_file_key(struct fscrypt_info *ci, struct fscrypt_master_key *mk) { - u8 derived_key[FSCRYPT_MAX_KEY_SIZE]; int err; if (ci->ci_policy.v2.flags & FSCRYPT_POLICY_FLAG_DIRECT_KEY) { /* - * DIRECT_KEY: instead of deriving per-file keys, the per-file - * nonce will be included in all the IVs. But unlike v1 - * policies, for v2 policies in this case we don't encrypt with - * the master key directly but rather derive a per-mode key. - * This ensures that the master key is consistently used only - * for HKDF, avoiding key reuse issues. + * DIRECT_KEY: instead of deriving per-file encryption keys, the + * per-file nonce will be included in all the IVs. But unlike + * v1 policies, for v2 policies in this case we don't encrypt + * with the master key directly but rather derive a per-mode + * encryption key. This ensures that the master key is + * consistently used only for HKDF, avoiding key reuse issues. */ - if (!fscrypt_mode_supports_direct_key(ci->ci_mode)) { - fscrypt_warn(ci->ci_inode, - "Direct key flag not allowed with %s", - ci->ci_mode->friendly_name); - return -EINVAL; - } - return setup_per_mode_key(ci, mk, mk->mk_direct_tfms, - HKDF_CONTEXT_DIRECT_KEY, false); + err = setup_per_mode_enc_key(ci, mk, mk->mk_direct_tfms, + HKDF_CONTEXT_DIRECT_KEY, false); } else if (ci->ci_policy.v2.flags & FSCRYPT_POLICY_FLAG_IV_INO_LBLK_64) { /* @@ -202,21 +213,34 @@ static int fscrypt_setup_v2_file_key(struct fscrypt_info *ci, * the IVs. This format is optimized for use with inline * encryption hardware compliant with the UFS or eMMC standards. */ - return setup_per_mode_key(ci, mk, mk->mk_iv_ino_lblk_64_tfms, - HKDF_CONTEXT_IV_INO_LBLK_64_KEY, - true); + err = setup_per_mode_enc_key(ci, mk, mk->mk_iv_ino_lblk_64_tfms, + HKDF_CONTEXT_IV_INO_LBLK_64_KEY, + true); + } else { + u8 derived_key[FSCRYPT_MAX_KEY_SIZE]; + + err = fscrypt_hkdf_expand(&mk->mk_secret.hkdf, + HKDF_CONTEXT_PER_FILE_ENC_KEY, + ci->ci_nonce, + FS_KEY_DERIVATION_NONCE_SIZE, + derived_key, ci->ci_mode->keysize); + if (err) + return err; + + err = fscrypt_set_per_file_enc_key(ci, derived_key); + memzero_explicit(derived_key, ci->ci_mode->keysize); } - - err = fscrypt_hkdf_expand(&mk->mk_secret.hkdf, - HKDF_CONTEXT_PER_FILE_KEY, - ci->ci_nonce, FS_KEY_DERIVATION_NONCE_SIZE, - derived_key, ci->ci_mode->keysize); if (err) return err; - err = fscrypt_set_derived_key(ci, derived_key); - memzero_explicit(derived_key, ci->ci_mode->keysize); - return err; + /* Derive a secret dirhash key for directories that need it. */ + if (S_ISDIR(ci->ci_inode->i_mode) && IS_CASEFOLDED(ci->ci_inode)) { + err = fscrypt_derive_dirhash_key(ci, mk); + if (err) + return err; + } + + return 0; } /* diff --git a/fs/crypto/keysetup_v1.c b/fs/crypto/keysetup_v1.c index 5298ef22aa85..801b48c0cd7f 100644 --- a/fs/crypto/keysetup_v1.c +++ b/fs/crypto/keysetup_v1.c @@ -9,7 +9,7 @@ * This file implements compatibility functions for the original encryption * policy version ("v1"), including: * - * - Deriving per-file keys using the AES-128-ECB based KDF + * - Deriving per-file encryption keys using the AES-128-ECB based KDF * (rather than the new method of using HKDF-SHA512) * * - Retrieving fscrypt master keys from process-subscribed keyrings @@ -253,23 +253,8 @@ err_free_dk: static int setup_v1_file_key_direct(struct fscrypt_info *ci, const u8 *raw_master_key) { - const struct fscrypt_mode *mode = ci->ci_mode; struct fscrypt_direct_key *dk; - if (!fscrypt_mode_supports_direct_key(mode)) { - fscrypt_warn(ci->ci_inode, - "Direct key mode not allowed with %s", - mode->friendly_name); - return -EINVAL; - } - - if (ci->ci_policy.v1.contents_encryption_mode != - ci->ci_policy.v1.filenames_encryption_mode) { - fscrypt_warn(ci->ci_inode, - "Direct key mode not allowed with different contents and filenames modes"); - return -EINVAL; - } - dk = fscrypt_get_direct_key(ci, raw_master_key); if (IS_ERR(dk)) return PTR_ERR(dk); @@ -298,7 +283,7 @@ static int setup_v1_file_key_derived(struct fscrypt_info *ci, if (err) goto out; - err = fscrypt_set_derived_key(ci, derived_key); + err = fscrypt_set_per_file_enc_key(ci, derived_key); out: kzfree(derived_key); return err; diff --git a/fs/crypto/policy.c b/fs/crypto/policy.c index 96f528071bed..cf2a9d26ef7d 100644 --- a/fs/crypto/policy.c +++ b/fs/crypto/policy.c @@ -29,6 +29,43 @@ bool fscrypt_policies_equal(const union fscrypt_policy *policy1, return !memcmp(policy1, policy2, fscrypt_policy_size(policy1)); } +static bool fscrypt_valid_enc_modes(u32 contents_mode, u32 filenames_mode) +{ + if (contents_mode == FSCRYPT_MODE_AES_256_XTS && + filenames_mode == FSCRYPT_MODE_AES_256_CTS) + return true; + + if (contents_mode == FSCRYPT_MODE_AES_128_CBC && + filenames_mode == FSCRYPT_MODE_AES_128_CTS) + return true; + + if (contents_mode == FSCRYPT_MODE_ADIANTUM && + filenames_mode == FSCRYPT_MODE_ADIANTUM) + return true; + + return false; +} + +static bool supported_direct_key_modes(const struct inode *inode, + u32 contents_mode, u32 filenames_mode) +{ + const struct fscrypt_mode *mode; + + if (contents_mode != filenames_mode) { + fscrypt_warn(inode, + "Direct key flag not allowed with different contents and filenames modes"); + return false; + } + mode = &fscrypt_modes[contents_mode]; + + if (mode->ivsize < offsetofend(union fscrypt_iv, nonce)) { + fscrypt_warn(inode, "Direct key flag not allowed with %s", + mode->friendly_name); + return false; + } + return true; +} + static bool supported_iv_ino_lblk_64_policy( const struct fscrypt_policy_v2 *policy, const struct inode *inode) @@ -63,13 +100,82 @@ static bool supported_iv_ino_lblk_64_policy( return true; } +static bool fscrypt_supported_v1_policy(const struct fscrypt_policy_v1 *policy, + const struct inode *inode) +{ + if (!fscrypt_valid_enc_modes(policy->contents_encryption_mode, + policy->filenames_encryption_mode)) { + fscrypt_warn(inode, + "Unsupported encryption modes (contents %d, filenames %d)", + policy->contents_encryption_mode, + policy->filenames_encryption_mode); + return false; + } + + if (policy->flags & ~(FSCRYPT_POLICY_FLAGS_PAD_MASK | + FSCRYPT_POLICY_FLAG_DIRECT_KEY)) { + fscrypt_warn(inode, "Unsupported encryption flags (0x%02x)", + policy->flags); + return false; + } + + if ((policy->flags & FSCRYPT_POLICY_FLAG_DIRECT_KEY) && + !supported_direct_key_modes(inode, policy->contents_encryption_mode, + policy->filenames_encryption_mode)) + return false; + + if (IS_CASEFOLDED(inode)) { + /* With v1, there's no way to derive dirhash keys. */ + fscrypt_warn(inode, + "v1 policies can't be used on casefolded directories"); + return false; + } + + return true; +} + +static bool fscrypt_supported_v2_policy(const struct fscrypt_policy_v2 *policy, + const struct inode *inode) +{ + if (!fscrypt_valid_enc_modes(policy->contents_encryption_mode, + policy->filenames_encryption_mode)) { + fscrypt_warn(inode, + "Unsupported encryption modes (contents %d, filenames %d)", + policy->contents_encryption_mode, + policy->filenames_encryption_mode); + return false; + } + + if (policy->flags & ~FSCRYPT_POLICY_FLAGS_VALID) { + fscrypt_warn(inode, "Unsupported encryption flags (0x%02x)", + policy->flags); + return false; + } + + if ((policy->flags & FSCRYPT_POLICY_FLAG_DIRECT_KEY) && + !supported_direct_key_modes(inode, policy->contents_encryption_mode, + policy->filenames_encryption_mode)) + return false; + + if ((policy->flags & FSCRYPT_POLICY_FLAG_IV_INO_LBLK_64) && + !supported_iv_ino_lblk_64_policy(policy, inode)) + return false; + + if (memchr_inv(policy->__reserved, 0, sizeof(policy->__reserved))) { + fscrypt_warn(inode, "Reserved bits set in encryption policy"); + return false; + } + + return true; +} + /** * fscrypt_supported_policy - check whether an encryption policy is supported * * Given an encryption policy, check whether all its encryption modes and other - * settings are supported by this kernel. (But we don't currently don't check - * for crypto API support here, so attempting to use an algorithm not configured - * into the crypto API will still fail later.) + * settings are supported by this kernel on the given inode. (But we don't + * currently don't check for crypto API support here, so attempting to use an + * algorithm not configured into the crypto API will still fail later.) * * Return: %true if supported, else %false */ @@ -77,60 +183,10 @@ bool fscrypt_supported_policy(const union fscrypt_policy *policy_u, const struct inode *inode) { switch (policy_u->version) { - case FSCRYPT_POLICY_V1: { - const struct fscrypt_policy_v1 *policy = &policy_u->v1; - - if (!fscrypt_valid_enc_modes(policy->contents_encryption_mode, - policy->filenames_encryption_mode)) { - fscrypt_warn(inode, - "Unsupported encryption modes (contents %d, filenames %d)", - policy->contents_encryption_mode, - policy->filenames_encryption_mode); - return false; - } - - if (policy->flags & ~(FSCRYPT_POLICY_FLAGS_PAD_MASK | - FSCRYPT_POLICY_FLAG_DIRECT_KEY)) { - fscrypt_warn(inode, - "Unsupported encryption flags (0x%02x)", - policy->flags); - return false; - } - - return true; - } - case FSCRYPT_POLICY_V2: { - const struct fscrypt_policy_v2 *policy = &policy_u->v2; - - if (!fscrypt_valid_enc_modes(policy->contents_encryption_mode, - policy->filenames_encryption_mode)) { - fscrypt_warn(inode, - "Unsupported encryption modes (contents %d, filenames %d)", - policy->contents_encryption_mode, - policy->filenames_encryption_mode); - return false; - } - - if (policy->flags & ~FSCRYPT_POLICY_FLAGS_VALID) { - fscrypt_warn(inode, - "Unsupported encryption flags (0x%02x)", - policy->flags); - return false; - } - - if ((policy->flags & FSCRYPT_POLICY_FLAG_IV_INO_LBLK_64) && - !supported_iv_ino_lblk_64_policy(policy, inode)) - return false; - - if (memchr_inv(policy->__reserved, 0, - sizeof(policy->__reserved))) { - fscrypt_warn(inode, - "Reserved bits set in encryption policy"); - return false; - } - - return true; - } + case FSCRYPT_POLICY_V1: + return fscrypt_supported_v1_policy(&policy_u->v1, inode); + case FSCRYPT_POLICY_V2: + return fscrypt_supported_v2_policy(&policy_u->v2, inode); } return false; } diff --git a/fs/direct-io.c b/fs/direct-io.c index 0ec4f270139f..00b4d15bb811 100644 --- a/fs/direct-io.c +++ b/fs/direct-io.c @@ -39,6 +39,8 @@ #include <linux/atomic.h> #include <linux/prefetch.h> +#include "internal.h" + /* * How many user pages to map in one call to get_user_pages(). This determines * the size of a structure in the slab cache diff --git a/fs/drop_caches.c b/fs/drop_caches.c index d31b6c72b476..dc1a1d5d825b 100644 --- a/fs/drop_caches.c +++ b/fs/drop_caches.c @@ -35,11 +35,11 @@ static void drop_pagecache_sb(struct super_block *sb, void *unused) spin_unlock(&inode->i_lock); spin_unlock(&sb->s_inode_list_lock); - cond_resched(); invalidate_mapping_pages(inode->i_mapping, 0, -1); iput(toput_inode); toput_inode = inode; + cond_resched(); spin_lock(&sb->s_inode_list_lock); } spin_unlock(&sb->s_inode_list_lock); diff --git a/fs/erofs/xattr.c b/fs/erofs/xattr.c index a13a78725c57..b766c3ee5fa8 100644 --- a/fs/erofs/xattr.c +++ b/fs/erofs/xattr.c @@ -649,6 +649,8 @@ ssize_t erofs_listxattr(struct dentry *dentry, struct listxattr_iter it; ret = init_inode_xattrs(d_inode(dentry)); + if (ret == -ENOATTR) + return 0; if (ret) return ret; diff --git a/fs/ext4/Kconfig b/fs/ext4/Kconfig index ef42ab040905..db9bfa08d3e0 100644 --- a/fs/ext4/Kconfig +++ b/fs/ext4/Kconfig @@ -39,6 +39,7 @@ config EXT4_FS select CRYPTO select CRYPTO_CRC32C select FS_IOMAP + select FS_ENCRYPTION_ALGS if FS_ENCRYPTION help This is the next generation of the ext3 filesystem. diff --git a/fs/ext4/block_validity.c b/fs/ext4/block_validity.c index d4d4fdfac1a6..1ee04e76bbe0 100644 --- a/fs/ext4/block_validity.c +++ b/fs/ext4/block_validity.c @@ -133,10 +133,13 @@ static void debug_print_tree(struct ext4_sb_info *sbi) { struct rb_node *node; struct ext4_system_zone *entry; + struct ext4_system_blocks *system_blks; int first = 1; printk(KERN_INFO "System zones: "); - node = rb_first(&sbi->system_blks->root); + rcu_read_lock(); + system_blks = rcu_dereference(sbi->system_blks); + node = rb_first(&system_blks->root); while (node) { entry = rb_entry(node, struct ext4_system_zone, node); printk(KERN_CONT "%s%llu-%llu", first ? "" : ", ", @@ -144,6 +147,7 @@ static void debug_print_tree(struct ext4_sb_info *sbi) first = 0; node = rb_next(node); } + rcu_read_unlock(); printk(KERN_CONT "\n"); } diff --git a/fs/ext4/dir.c b/fs/ext4/dir.c index 9fdd2b269d61..4e093277c8bf 100644 --- a/fs/ext4/dir.c +++ b/fs/ext4/dir.c @@ -72,6 +72,7 @@ int __ext4_check_dir_entry(const char *function, unsigned int line, const char *error_msg = NULL; const int rlen = ext4_rec_len_from_disk(de->rec_len, dir->i_sb->s_blocksize); + const int next_offset = ((char *) de - buf) + rlen; if (unlikely(rlen < EXT4_DIR_REC_LEN(1))) error_msg = "rec_len is smaller than minimal"; @@ -79,8 +80,11 @@ int __ext4_check_dir_entry(const char *function, unsigned int line, error_msg = "rec_len % 4 != 0"; else if (unlikely(rlen < EXT4_DIR_REC_LEN(de->name_len))) error_msg = "rec_len is too small for name_len"; - else if (unlikely(((char *) de - buf) + rlen > size)) + else if (unlikely(next_offset > size)) error_msg = "directory entry overrun"; + else if (unlikely(next_offset > size - EXT4_DIR_REC_LEN(1) && + next_offset != size)) + error_msg = "directory entry too close to block end"; else if (unlikely(le32_to_cpu(de->inode) > le32_to_cpu(EXT4_SB(dir->i_sb)->s_es->s_inodes_count))) error_msg = "inode out of bounds"; @@ -116,7 +120,7 @@ static int ext4_readdir(struct file *file, struct dir_context *ctx) if (IS_ENCRYPTED(inode)) { err = fscrypt_get_encryption_info(inode); - if (err && err != -ENOKEY) + if (err) return err; } diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c index dc333e8e51e8..8ca4a23129aa 100644 --- a/fs/ext4/ialloc.c +++ b/fs/ext4/ialloc.c @@ -921,8 +921,8 @@ repeat_in_this_group: if (!handle) { BUG_ON(nblocks <= 0); handle = __ext4_journal_start_sb(dir->i_sb, line_no, - handle_type, nblocks, - 0, 0); + handle_type, nblocks, 0, + ext4_trans_default_revoke_credits(sb)); if (IS_ERR(handle)) { err = PTR_ERR(handle); ext4_std_error(sb, err); diff --git a/fs/ext4/inode-test.c b/fs/ext4/inode-test.c index 92a9da1774aa..bbce1c328d85 100644 --- a/fs/ext4/inode-test.c +++ b/fs/ext4/inode-test.c @@ -25,7 +25,7 @@ * For constructing the negative timestamp lower bound value. * binary: 10000000 00000000 00000000 00000000 */ -#define LOWER_MSB_1 (-0x80000000L) +#define LOWER_MSB_1 (-(UPPER_MSB_0) - 1L) /* avoid overflow */ /* * For constructing the negative timestamp upper bound value. * binary: 11111111 11111111 11111111 11111111 diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 28f28de0c1b6..629a25d999f0 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -5692,7 +5692,7 @@ int ext4_expand_extra_isize(struct inode *inode, error = ext4_journal_get_write_access(handle, iloc->bh); if (error) { brelse(iloc->bh); - goto out_stop; + goto out_unlock; } error = __ext4_expand_extra_isize(inode, new_extra_isize, iloc, @@ -5702,8 +5702,8 @@ int ext4_expand_extra_isize(struct inode *inode, if (!error) error = rc; +out_unlock: ext4_write_unlock_xattr(inode, &no_expand); -out_stop: ext4_journal_stop(handle); return error; } diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index a856997d87b5..1cb42d940784 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -2164,7 +2164,9 @@ static int ext4_add_entry(handle_t *handle, struct dentry *dentry, struct buffer_head *bh = NULL; struct ext4_dir_entry_2 *de; struct super_block *sb; +#ifdef CONFIG_UNICODE struct ext4_sb_info *sbi; +#endif struct ext4_filename fname; int retval; int dx_fallback=0; @@ -2176,12 +2178,12 @@ static int ext4_add_entry(handle_t *handle, struct dentry *dentry, csum_size = sizeof(struct ext4_dir_entry_tail); sb = dir->i_sb; - sbi = EXT4_SB(sb); blocksize = sb->s_blocksize; if (!dentry->d_name.len) return -EINVAL; #ifdef CONFIG_UNICODE + sbi = EXT4_SB(sb); if (ext4_has_strict_mode(sbi) && IS_CASEFOLDED(dir) && sbi->s_encoding && utf8_validate(sbi->s_encoding, &dentry->d_name)) return -EINVAL; @@ -2822,7 +2824,7 @@ bool ext4_empty_dir(struct inode *inode) { unsigned int offset; struct buffer_head *bh; - struct ext4_dir_entry_2 *de, *de1; + struct ext4_dir_entry_2 *de; struct super_block *sb; if (ext4_has_inline_data(inode)) { @@ -2847,19 +2849,25 @@ bool ext4_empty_dir(struct inode *inode) return true; de = (struct ext4_dir_entry_2 *) bh->b_data; - de1 = ext4_next_entry(de, sb->s_blocksize); - if (le32_to_cpu(de->inode) != inode->i_ino || - le32_to_cpu(de1->inode) == 0 || - strcmp(".", de->name) || strcmp("..", de1->name)) { - ext4_warning_inode(inode, "directory missing '.' and/or '..'"); + if (ext4_check_dir_entry(inode, NULL, de, bh, bh->b_data, bh->b_size, + 0) || + le32_to_cpu(de->inode) != inode->i_ino || strcmp(".", de->name)) { + ext4_warning_inode(inode, "directory missing '.'"); + brelse(bh); + return true; + } + offset = ext4_rec_len_from_disk(de->rec_len, sb->s_blocksize); + de = ext4_next_entry(de, sb->s_blocksize); + if (ext4_check_dir_entry(inode, NULL, de, bh, bh->b_data, bh->b_size, + offset) || + le32_to_cpu(de->inode) == 0 || strcmp("..", de->name)) { + ext4_warning_inode(inode, "directory missing '..'"); brelse(bh); return true; } - offset = ext4_rec_len_from_disk(de->rec_len, sb->s_blocksize) + - ext4_rec_len_from_disk(de1->rec_len, sb->s_blocksize); - de = ext4_next_entry(de1, sb->s_blocksize); + offset += ext4_rec_len_from_disk(de->rec_len, sb->s_blocksize); while (offset < inode->i_size) { - if ((void *) de >= (void *) (bh->b_data+sb->s_blocksize)) { + if (!(offset & (sb->s_blocksize - 1))) { unsigned int lblock; brelse(bh); lblock = offset >> EXT4_BLOCK_SIZE_BITS(sb); @@ -2870,12 +2878,11 @@ bool ext4_empty_dir(struct inode *inode) } if (IS_ERR(bh)) return true; - de = (struct ext4_dir_entry_2 *) bh->b_data; } + de = (struct ext4_dir_entry_2 *) (bh->b_data + + (offset & (sb->s_blocksize - 1))); if (ext4_check_dir_entry(inode, NULL, de, bh, bh->b_data, bh->b_size, offset)) { - de = (struct ext4_dir_entry_2 *)(bh->b_data + - sb->s_blocksize); offset = (offset | (sb->s_blocksize - 1)) + 1; continue; } @@ -2884,7 +2891,6 @@ bool ext4_empty_dir(struct inode *inode) return false; } offset += ext4_rec_len_from_disk(de->rec_len, sb->s_blocksize); - de = ext4_next_entry(de, sb->s_blocksize); } brelse(bh); return true; diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 1d82b56d9b11..2937a8873fe1 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -1900,6 +1900,13 @@ static int handle_mount_opt(struct super_block *sb, char *opt, int token, } sbi->s_commit_interval = HZ * arg; } else if (token == Opt_debug_want_extra_isize) { + if ((arg & 1) || + (arg < 4) || + (arg > (sbi->s_inode_size - EXT4_GOOD_OLD_INODE_SIZE))) { + ext4_msg(sb, KERN_ERR, + "Invalid want_extra_isize %d", arg); + return -1; + } sbi->s_want_extra_isize = arg; } else if (token == Opt_max_batch_time) { sbi->s_max_batch_time = arg; @@ -3554,40 +3561,6 @@ int ext4_calculate_overhead(struct super_block *sb) return 0; } -static void ext4_clamp_want_extra_isize(struct super_block *sb) -{ - struct ext4_sb_info *sbi = EXT4_SB(sb); - struct ext4_super_block *es = sbi->s_es; - unsigned def_extra_isize = sizeof(struct ext4_inode) - - EXT4_GOOD_OLD_INODE_SIZE; - - if (sbi->s_inode_size == EXT4_GOOD_OLD_INODE_SIZE) { - sbi->s_want_extra_isize = 0; - return; - } - if (sbi->s_want_extra_isize < 4) { - sbi->s_want_extra_isize = def_extra_isize; - if (ext4_has_feature_extra_isize(sb)) { - if (sbi->s_want_extra_isize < - le16_to_cpu(es->s_want_extra_isize)) - sbi->s_want_extra_isize = - le16_to_cpu(es->s_want_extra_isize); - if (sbi->s_want_extra_isize < - le16_to_cpu(es->s_min_extra_isize)) - sbi->s_want_extra_isize = - le16_to_cpu(es->s_min_extra_isize); - } - } - /* Check if enough inode space is available */ - if ((sbi->s_want_extra_isize > sbi->s_inode_size) || - (EXT4_GOOD_OLD_INODE_SIZE + sbi->s_want_extra_isize > - sbi->s_inode_size)) { - sbi->s_want_extra_isize = def_extra_isize; - ext4_msg(sb, KERN_INFO, - "required extra inode space not available"); - } -} - static void ext4_set_resv_clusters(struct super_block *sb) { ext4_fsblk_t resv_clusters; @@ -3795,6 +3768,68 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) */ sbi->s_li_wait_mult = EXT4_DEF_LI_WAIT_MULT; + if (le32_to_cpu(es->s_rev_level) == EXT4_GOOD_OLD_REV) { + sbi->s_inode_size = EXT4_GOOD_OLD_INODE_SIZE; + sbi->s_first_ino = EXT4_GOOD_OLD_FIRST_INO; + } else { + sbi->s_inode_size = le16_to_cpu(es->s_inode_size); + sbi->s_first_ino = le32_to_cpu(es->s_first_ino); + if (sbi->s_first_ino < EXT4_GOOD_OLD_FIRST_INO) { + ext4_msg(sb, KERN_ERR, "invalid first ino: %u", + sbi->s_first_ino); + goto failed_mount; + } + if ((sbi->s_inode_size < EXT4_GOOD_OLD_INODE_SIZE) || + (!is_power_of_2(sbi->s_inode_size)) || + (sbi->s_inode_size > blocksize)) { + ext4_msg(sb, KERN_ERR, + "unsupported inode size: %d", + sbi->s_inode_size); + goto failed_mount; + } + /* + * i_atime_extra is the last extra field available for + * [acm]times in struct ext4_inode. Checking for that + * field should suffice to ensure we have extra space + * for all three. + */ + if (sbi->s_inode_size >= offsetof(struct ext4_inode, i_atime_extra) + + sizeof(((struct ext4_inode *)0)->i_atime_extra)) { + sb->s_time_gran = 1; + sb->s_time_max = EXT4_EXTRA_TIMESTAMP_MAX; + } else { + sb->s_time_gran = NSEC_PER_SEC; + sb->s_time_max = EXT4_NON_EXTRA_TIMESTAMP_MAX; + } + sb->s_time_min = EXT4_TIMESTAMP_MIN; + } + if (sbi->s_inode_size > EXT4_GOOD_OLD_INODE_SIZE) { + sbi->s_want_extra_isize = sizeof(struct ext4_inode) - + EXT4_GOOD_OLD_INODE_SIZE; + if (ext4_has_feature_extra_isize(sb)) { + unsigned v, max = (sbi->s_inode_size - + EXT4_GOOD_OLD_INODE_SIZE); + + v = le16_to_cpu(es->s_want_extra_isize); + if (v > max) { + ext4_msg(sb, KERN_ERR, + "bad s_want_extra_isize: %d", v); + goto failed_mount; + } + if (sbi->s_want_extra_isize < v) + sbi->s_want_extra_isize = v; + + v = le16_to_cpu(es->s_min_extra_isize); + if (v > max) { + ext4_msg(sb, KERN_ERR, + "bad s_min_extra_isize: %d", v); + goto failed_mount; + } + if (sbi->s_want_extra_isize < v) + sbi->s_want_extra_isize = v; + } + } + if (sbi->s_es->s_mount_opts[0]) { char *s_mount_opts = kstrndup(sbi->s_es->s_mount_opts, sizeof(sbi->s_es->s_mount_opts), @@ -4033,42 +4068,6 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) has_huge_files); sb->s_maxbytes = ext4_max_size(sb->s_blocksize_bits, has_huge_files); - if (le32_to_cpu(es->s_rev_level) == EXT4_GOOD_OLD_REV) { - sbi->s_inode_size = EXT4_GOOD_OLD_INODE_SIZE; - sbi->s_first_ino = EXT4_GOOD_OLD_FIRST_INO; - } else { - sbi->s_inode_size = le16_to_cpu(es->s_inode_size); - sbi->s_first_ino = le32_to_cpu(es->s_first_ino); - if (sbi->s_first_ino < EXT4_GOOD_OLD_FIRST_INO) { - ext4_msg(sb, KERN_ERR, "invalid first ino: %u", - sbi->s_first_ino); - goto failed_mount; - } - if ((sbi->s_inode_size < EXT4_GOOD_OLD_INODE_SIZE) || - (!is_power_of_2(sbi->s_inode_size)) || - (sbi->s_inode_size > blocksize)) { - ext4_msg(sb, KERN_ERR, - "unsupported inode size: %d", - sbi->s_inode_size); - goto failed_mount; - } - /* - * i_atime_extra is the last extra field available for [acm]times in - * struct ext4_inode. Checking for that field should suffice to ensure - * we have extra space for all three. - */ - if (sbi->s_inode_size >= offsetof(struct ext4_inode, i_atime_extra) + - sizeof(((struct ext4_inode *)0)->i_atime_extra)) { - sb->s_time_gran = 1; - sb->s_time_max = EXT4_EXTRA_TIMESTAMP_MAX; - } else { - sb->s_time_gran = NSEC_PER_SEC; - sb->s_time_max = EXT4_NON_EXTRA_TIMESTAMP_MAX; - } - - sb->s_time_min = EXT4_TIMESTAMP_MIN; - } - sbi->s_desc_size = le16_to_cpu(es->s_desc_size); if (ext4_has_feature_64bit(sb)) { if (sbi->s_desc_size < EXT4_MIN_DESC_SIZE_64BIT || @@ -4517,8 +4516,6 @@ no_journal: } else if (ret) goto failed_mount4a; - ext4_clamp_want_extra_isize(sb); - ext4_set_resv_clusters(sb); err = ext4_setup_system_zone(sb); @@ -5306,8 +5303,6 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) goto restore_opts; } - ext4_clamp_want_extra_isize(sb); - if ((old_opts.s_mount_opt & EXT4_MOUNT_JOURNAL_CHECKSUM) ^ test_opt(sb, JOURNAL_CHECKSUM)) { ext4_msg(sb, KERN_ERR, "changing journal_checksum " diff --git a/fs/ext4/verity.c b/fs/ext4/verity.c index d0d8a9795dd6..dc5ec724d889 100644 --- a/fs/ext4/verity.c +++ b/fs/ext4/verity.c @@ -342,12 +342,55 @@ static int ext4_get_verity_descriptor(struct inode *inode, void *buf, return desc_size; } +/* + * Prefetch some pages from the file's Merkle tree. + * + * This is basically a stripped-down version of __do_page_cache_readahead() + * which works on pages past i_size. + */ +static void ext4_merkle_tree_readahead(struct address_space *mapping, + pgoff_t start_index, unsigned long count) +{ + LIST_HEAD(pages); + unsigned int nr_pages = 0; + struct page *page; + pgoff_t index; + struct blk_plug plug; + + for (index = start_index; index < start_index + count; index++) { + page = xa_load(&mapping->i_pages, index); + if (!page || xa_is_value(page)) { + page = __page_cache_alloc(readahead_gfp_mask(mapping)); + if (!page) + break; + page->index = index; + list_add(&page->lru, &pages); + nr_pages++; + } + } + blk_start_plug(&plug); + ext4_mpage_readpages(mapping, &pages, NULL, nr_pages, true); + blk_finish_plug(&plug); +} + static struct page *ext4_read_merkle_tree_page(struct inode *inode, - pgoff_t index) + pgoff_t index, + unsigned long num_ra_pages) { + struct page *page; + index += ext4_verity_metadata_pos(inode) >> PAGE_SHIFT; - return read_mapping_page(inode->i_mapping, index, NULL); + page = find_get_page_flags(inode->i_mapping, index, FGP_ACCESSED); + if (!page || !PageUptodate(page)) { + if (page) + put_page(page); + else if (num_ra_pages > 1) + ext4_merkle_tree_readahead(inode->i_mapping, index, + num_ra_pages); + page = read_mapping_page(inode->i_mapping, index, NULL); + } + return page; } static int ext4_write_merkle_tree_block(struct inode *inode, const void *buf, diff --git a/fs/f2fs/Kconfig b/fs/f2fs/Kconfig index 652fd2e2b23d..599fb9194c6a 100644 --- a/fs/f2fs/Kconfig +++ b/fs/f2fs/Kconfig @@ -6,6 +6,7 @@ config F2FS_FS select CRYPTO select CRYPTO_CRC32 select F2FS_FS_XATTR if FS_ENCRYPTION + select FS_ENCRYPTION_ALGS if FS_ENCRYPTION help F2FS is based on Log-structured File System (LFS), which supports versatile "flash-friendly" features. The design has been focused on diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index a034cd0ce021..0fa356e94ef5 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -1881,7 +1881,7 @@ out: * use ->readpage() or do the necessary surgery to decouple ->readpages() * from read-ahead. */ -static int f2fs_mpage_readpages(struct address_space *mapping, +int f2fs_mpage_readpages(struct address_space *mapping, struct list_head *pages, struct page *page, unsigned nr_pages, bool is_readahead) { diff --git a/fs/f2fs/dir.c b/fs/f2fs/dir.c index c967cacf979e..d9ad842945df 100644 --- a/fs/f2fs/dir.c +++ b/fs/f2fs/dir.c @@ -987,7 +987,7 @@ static int f2fs_readdir(struct file *file, struct dir_context *ctx) if (IS_ENCRYPTED(inode)) { err = fscrypt_get_encryption_info(inode); - if (err && err != -ENOKEY) + if (err) goto out; err = fscrypt_fname_alloc_buffer(inode, F2FS_NAME_LEN, &fstr); diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 5a888a063c7f..059ade83bfb1 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -3229,6 +3229,9 @@ int f2fs_reserve_new_block(struct dnode_of_data *dn); int f2fs_get_block(struct dnode_of_data *dn, pgoff_t index); int f2fs_preallocate_blocks(struct kiocb *iocb, struct iov_iter *from); int f2fs_reserve_block(struct dnode_of_data *dn, pgoff_t index); +int f2fs_mpage_readpages(struct address_space *mapping, + struct list_head *pages, struct page *page, + unsigned nr_pages, bool is_readahead); struct page *f2fs_get_read_data_page(struct inode *inode, pgoff_t index, int op_flags, bool for_write); struct page *f2fs_find_data_page(struct inode *inode, pgoff_t index); diff --git a/fs/f2fs/verity.c b/fs/f2fs/verity.c index a401ef72bc82..d7d430a6f130 100644 --- a/fs/f2fs/verity.c +++ b/fs/f2fs/verity.c @@ -222,12 +222,55 @@ static int f2fs_get_verity_descriptor(struct inode *inode, void *buf, return size; } +/* + * Prefetch some pages from the file's Merkle tree. + * + * This is basically a stripped-down version of __do_page_cache_readahead() + * which works on pages past i_size. + */ +static void f2fs_merkle_tree_readahead(struct address_space *mapping, + pgoff_t start_index, unsigned long count) +{ + LIST_HEAD(pages); + unsigned int nr_pages = 0; + struct page *page; + pgoff_t index; + struct blk_plug plug; + + for (index = start_index; index < start_index + count; index++) { + page = xa_load(&mapping->i_pages, index); + if (!page || xa_is_value(page)) { + page = __page_cache_alloc(readahead_gfp_mask(mapping)); + if (!page) + break; + page->index = index; + list_add(&page->lru, &pages); + nr_pages++; + } + } + blk_start_plug(&plug); + f2fs_mpage_readpages(mapping, &pages, NULL, nr_pages, true); + blk_finish_plug(&plug); +} + static struct page *f2fs_read_merkle_tree_page(struct inode *inode, - pgoff_t index) + pgoff_t index, + unsigned long num_ra_pages) { + struct page *page; + index += f2fs_verity_metadata_pos(inode) >> PAGE_SHIFT; - return read_mapping_page(inode->i_mapping, index, NULL); + page = find_get_page_flags(inode->i_mapping, index, FGP_ACCESSED); + if (!page || !PageUptodate(page)) { + if (page) + put_page(page); + else if (num_ra_pages > 1) + f2fs_merkle_tree_readahead(inode->i_mapping, index, + num_ra_pages); + page = read_mapping_page(inode->i_mapping, index, NULL); + } + return page; } static int f2fs_write_merkle_tree_block(struct inode *inode, const void *buf, diff --git a/fs/fuse/file.c b/fs/fuse/file.c index a63d779eac10..ce715380143c 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -882,6 +882,7 @@ static void fuse_send_readpages(struct fuse_io_args *ia, struct file *file) struct fuse_args_pages *ap = &ia->ap; loff_t pos = page_offset(ap->pages[0]); size_t count = ap->num_pages << PAGE_SHIFT; + ssize_t res; int err; ap->args.out_pages = true; @@ -896,7 +897,8 @@ static void fuse_send_readpages(struct fuse_io_args *ia, struct file *file) if (!err) return; } else { - err = fuse_simple_request(fc, &ap->args); + res = fuse_simple_request(fc, &ap->args); + err = res < 0 ? res : 0; } fuse_readpages_end(fc, &ap->args, err); } diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c index d5c2a3158610..a66e425884d1 100644 --- a/fs/hugetlbfs/inode.c +++ b/fs/hugetlbfs/inode.c @@ -1498,8 +1498,10 @@ static int __init init_hugetlbfs_fs(void) /* other hstates are optional */ i = 0; for_each_hstate(h) { - if (i == default_hstate_idx) + if (i == default_hstate_idx) { + i++; continue; + } mnt = mount_one_hugetlbfs(h); if (IS_ERR(mnt)) diff --git a/fs/inode.c b/fs/inode.c index fef457a42882..ea15c6d9f274 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -12,6 +12,7 @@ #include <linux/security.h> #include <linux/cdev.h> #include <linux/memblock.h> +#include <linux/fscrypt.h> #include <linux/fsnotify.h> #include <linux/mount.h> #include <linux/posix_acl.h> @@ -676,6 +677,7 @@ int invalidate_inodes(struct super_block *sb, bool kill_dirty) struct inode *inode, *next; LIST_HEAD(dispose); +again: spin_lock(&sb->s_inode_list_lock); list_for_each_entry_safe(inode, next, &sb->s_inodes, i_sb_list) { spin_lock(&inode->i_lock); @@ -698,6 +700,12 @@ int invalidate_inodes(struct super_block *sb, bool kill_dirty) inode_lru_list_del(inode); spin_unlock(&inode->i_lock); list_add(&inode->i_lru, &dispose); + if (need_resched()) { + spin_unlock(&sb->s_inode_list_lock); + cond_resched(); + dispose_list(&dispose); + goto again; + } } spin_unlock(&sb->s_inode_list_lock); @@ -2245,7 +2253,7 @@ int vfs_ioc_setflags_prepare(struct inode *inode, unsigned int oldflags, !capable(CAP_LINUX_IMMUTABLE)) return -EPERM; - return 0; + return fscrypt_prepare_setflags(inode, oldflags, flags); } EXPORT_SYMBOL(vfs_ioc_setflags_prepare); diff --git a/fs/internal.h b/fs/internal.h index 4a7da1df573d..e3fa69544b66 100644 --- a/fs/internal.h +++ b/fs/internal.h @@ -38,7 +38,7 @@ static inline int __sync_blockdev(struct block_device *bdev, int wait) /* * buffer.c */ -extern void guard_bio_eod(int rw, struct bio *bio); +extern void guard_bio_eod(struct bio *bio); extern int __block_write_begin_int(struct page *page, loff_t pos, unsigned len, get_block_t *get_block, struct iomap *iomap); diff --git a/fs/io-wq.c b/fs/io-wq.c index 74b40506c5d9..5147d2213b01 100644 --- a/fs/io-wq.c +++ b/fs/io-wq.c @@ -49,7 +49,6 @@ struct io_worker { struct hlist_nulls_node nulls_node; struct list_head all_list; struct task_struct *task; - wait_queue_head_t wait; struct io_wqe *wqe; struct io_wq_work *cur_work; @@ -93,7 +92,6 @@ struct io_wqe { struct io_wqe_acct acct[2]; struct hlist_nulls_head free_list; - struct hlist_nulls_head busy_list; struct list_head all_list; struct io_wq *wq; @@ -258,7 +256,7 @@ static bool io_wqe_activate_free_worker(struct io_wqe *wqe) worker = hlist_nulls_entry(n, struct io_worker, nulls_node); if (io_worker_get(worker)) { - wake_up(&worker->wait); + wake_up_process(worker->task); io_worker_release(worker); return true; } @@ -328,7 +326,6 @@ static void __io_worker_busy(struct io_wqe *wqe, struct io_worker *worker, if (worker->flags & IO_WORKER_F_FREE) { worker->flags &= ~IO_WORKER_F_FREE; hlist_nulls_del_init_rcu(&worker->nulls_node); - hlist_nulls_add_head_rcu(&worker->nulls_node, &wqe->busy_list); } /* @@ -366,7 +363,6 @@ static bool __io_worker_idle(struct io_wqe *wqe, struct io_worker *worker) { if (!(worker->flags & IO_WORKER_F_FREE)) { worker->flags |= IO_WORKER_F_FREE; - hlist_nulls_del_init_rcu(&worker->nulls_node); hlist_nulls_add_head_rcu(&worker->nulls_node, &wqe->free_list); } @@ -433,6 +429,8 @@ next: if (signal_pending(current)) flush_signals(current); + cond_resched(); + spin_lock_irq(&worker->lock); worker->cur_work = work; spin_unlock_irq(&worker->lock); @@ -447,10 +445,14 @@ next: task_unlock(current); } if ((work->flags & IO_WQ_WORK_NEEDS_USER) && !worker->mm && - wq->mm && mmget_not_zero(wq->mm)) { - use_mm(wq->mm); - set_fs(USER_DS); - worker->mm = wq->mm; + wq->mm) { + if (mmget_not_zero(wq->mm)) { + use_mm(wq->mm); + set_fs(USER_DS); + worker->mm = wq->mm; + } else { + work->flags |= IO_WQ_WORK_CANCEL; + } } if (!worker->creds) worker->creds = override_creds(wq->creds); @@ -492,28 +494,46 @@ next: } while (1); } +static inline void io_worker_spin_for_work(struct io_wqe *wqe) +{ + int i = 0; + + while (++i < 1000) { + if (io_wqe_run_queue(wqe)) + break; + if (need_resched()) + break; + cpu_relax(); + } +} + static int io_wqe_worker(void *data) { struct io_worker *worker = data; struct io_wqe *wqe = worker->wqe; struct io_wq *wq = wqe->wq; - DEFINE_WAIT(wait); + bool did_work; io_worker_start(wqe, worker); + did_work = false; while (!test_bit(IO_WQ_BIT_EXIT, &wq->state)) { - prepare_to_wait(&worker->wait, &wait, TASK_INTERRUPTIBLE); - + set_current_state(TASK_INTERRUPTIBLE); +loop: + if (did_work) + io_worker_spin_for_work(wqe); spin_lock_irq(&wqe->lock); if (io_wqe_run_queue(wqe)) { __set_current_state(TASK_RUNNING); io_worker_handle_work(worker); - continue; + did_work = true; + goto loop; } + did_work = false; /* drops the lock on success, retry */ if (__io_worker_idle(wqe, worker)) { __release(&wqe->lock); - continue; + goto loop; } spin_unlock_irq(&wqe->lock); if (signal_pending(current)) @@ -526,8 +546,6 @@ static int io_wqe_worker(void *data) break; } - finish_wait(&worker->wait, &wait); - if (test_bit(IO_WQ_BIT_EXIT, &wq->state)) { spin_lock_irq(&wqe->lock); if (!wq_list_empty(&wqe->work_list)) @@ -589,7 +607,6 @@ static bool create_io_worker(struct io_wq *wq, struct io_wqe *wqe, int index) refcount_set(&worker->ref, 1); worker->nulls_node.pprev = NULL; - init_waitqueue_head(&worker->wait); worker->wqe = wqe; spin_lock_init(&worker->lock); @@ -784,10 +801,6 @@ void io_wq_cancel_all(struct io_wq *wq) set_bit(IO_WQ_BIT_CANCEL, &wq->state); - /* - * Browse both lists, as there's a gap between handing work off - * to a worker and the worker putting itself on the busy_list - */ rcu_read_lock(); for_each_node(node) { struct io_wqe *wqe = wq->wqes[node]; @@ -934,7 +947,7 @@ static enum io_wq_cancel io_wqe_cancel_work(struct io_wqe *wqe, /* * Now check if a free (going busy) or busy worker has the work * currently running. If we find it there, we'll return CANCEL_RUNNING - * as an indication that we attempte to signal cancellation. The + * as an indication that we attempt to signal cancellation. The * completion will run normally in this case. */ rcu_read_lock(); @@ -1035,7 +1048,6 @@ struct io_wq *io_wq_create(unsigned bounded, struct io_wq_data *data) spin_lock_init(&wqe->lock); INIT_WQ_LIST(&wqe->work_list); INIT_HLIST_NULLS_HEAD(&wqe->free_list, 0); - INIT_HLIST_NULLS_HEAD(&wqe->busy_list, 1); INIT_LIST_HEAD(&wqe->all_list); } diff --git a/fs/io-wq.h b/fs/io-wq.h index 7c333a28e2a7..3f5e356de980 100644 --- a/fs/io-wq.h +++ b/fs/io-wq.h @@ -35,7 +35,8 @@ static inline void wq_list_add_tail(struct io_wq_work_node *node, struct io_wq_work_list *list) { if (!list->first) { - list->first = list->last = node; + list->last = node; + WRITE_ONCE(list->first, node); } else { list->last->next = node; list->last = node; @@ -47,7 +48,7 @@ static inline void wq_node_del(struct io_wq_work_list *list, struct io_wq_work_node *prev) { if (node == list->first) - list->first = node->next; + WRITE_ONCE(list->first, node->next); if (node == list->last) list->last = prev; if (prev) @@ -58,7 +59,7 @@ static inline void wq_node_del(struct io_wq_work_list *list, #define wq_list_for_each(pos, prv, head) \ for (pos = (head)->first, prv = NULL; pos; prv = pos, pos = (pos)->next) -#define wq_list_empty(list) ((list)->first == NULL) +#define wq_list_empty(list) (READ_ONCE((list)->first) == NULL) #define INIT_WQ_LIST(list) do { \ (list)->first = NULL; \ (list)->last = NULL; \ @@ -119,6 +120,10 @@ static inline void io_wq_worker_sleeping(struct task_struct *tsk) static inline void io_wq_worker_running(struct task_struct *tsk) { } -#endif /* CONFIG_IO_WQ */ +#endif -#endif /* INTERNAL_IO_WQ_H */ +static inline bool io_wq_current_is_worker(void) +{ + return in_task() && (current->flags & PF_IO_WORKER); +} +#endif diff --git a/fs/io_uring.c b/fs/io_uring.c index 405be10da73d..e54556b0fcc6 100644 --- a/fs/io_uring.c +++ b/fs/io_uring.c @@ -289,11 +289,14 @@ struct io_ring_ctx { */ struct io_poll_iocb { struct file *file; - struct wait_queue_head *head; + union { + struct wait_queue_head *head; + u64 addr; + }; __poll_t events; bool done; bool canceled; - struct wait_queue_entry *wait; + struct wait_queue_entry wait; }; struct io_timeout_data { @@ -304,6 +307,51 @@ struct io_timeout_data { u32 seq_offset; }; +struct io_accept { + struct file *file; + struct sockaddr __user *addr; + int __user *addr_len; + int flags; +}; + +struct io_sync { + struct file *file; + loff_t len; + loff_t off; + int flags; +}; + +struct io_cancel { + struct file *file; + u64 addr; +}; + +struct io_timeout { + struct file *file; + u64 addr; + int flags; + unsigned count; +}; + +struct io_rw { + /* NOTE: kiocb has the file as the first member, so don't do it here */ + struct kiocb kiocb; + u64 addr; + u64 len; +}; + +struct io_connect { + struct file *file; + struct sockaddr __user *addr; + int addr_len; +}; + +struct io_sr_msg { + struct file *file; + struct user_msghdr __user *msg; + int msg_flags; +}; + struct io_async_connect { struct sockaddr_storage address; }; @@ -323,7 +371,6 @@ struct io_async_rw { }; struct io_async_ctx { - struct io_uring_sqe sqe; union { struct io_async_rw rw; struct io_async_msghdr msg; @@ -341,17 +388,23 @@ struct io_async_ctx { struct io_kiocb { union { struct file *file; - struct kiocb rw; + struct io_rw rw; struct io_poll_iocb poll; + struct io_accept accept; + struct io_sync sync; + struct io_cancel cancel; + struct io_timeout timeout; + struct io_connect connect; + struct io_sr_msg sr_msg; }; - const struct io_uring_sqe *sqe; struct io_async_ctx *io; struct file *ring_file; int ring_fd; bool has_user; bool in_async; bool needs_fixed_file; + u8 opcode; struct io_ring_ctx *ctx; union { @@ -377,6 +430,7 @@ struct io_kiocb { #define REQ_F_TIMEOUT_NOSEQ 8192 /* no timeout sequence */ #define REQ_F_INFLIGHT 16384 /* on inflight list */ #define REQ_F_COMP_LOCKED 32768 /* completion under lock */ +#define REQ_F_HARDLINK 65536 /* doesn't sever on completion < 0 */ u64 user_data; u32 result; u32 sequence; @@ -563,12 +617,10 @@ static void __io_commit_cqring(struct io_ring_ctx *ctx) } } -static inline bool io_sqe_needs_user(const struct io_uring_sqe *sqe) +static inline bool io_req_needs_user(struct io_kiocb *req) { - u8 opcode = READ_ONCE(sqe->opcode); - - return !(opcode == IORING_OP_READ_FIXED || - opcode == IORING_OP_WRITE_FIXED); + return !(req->opcode == IORING_OP_READ_FIXED || + req->opcode == IORING_OP_WRITE_FIXED); } static inline bool io_prep_async_work(struct io_kiocb *req, @@ -576,31 +628,31 @@ static inline bool io_prep_async_work(struct io_kiocb *req, { bool do_hashed = false; - if (req->sqe) { - switch (req->sqe->opcode) { - case IORING_OP_WRITEV: - case IORING_OP_WRITE_FIXED: + switch (req->opcode) { + case IORING_OP_WRITEV: + case IORING_OP_WRITE_FIXED: + /* only regular files should be hashed for writes */ + if (req->flags & REQ_F_ISREG) do_hashed = true; - /* fall-through */ - case IORING_OP_READV: - case IORING_OP_READ_FIXED: - case IORING_OP_SENDMSG: - case IORING_OP_RECVMSG: - case IORING_OP_ACCEPT: - case IORING_OP_POLL_ADD: - case IORING_OP_CONNECT: - /* - * We know REQ_F_ISREG is not set on some of these - * opcodes, but this enables us to keep the check in - * just one place. - */ - if (!(req->flags & REQ_F_ISREG)) - req->work.flags |= IO_WQ_WORK_UNBOUND; - break; - } - if (io_sqe_needs_user(req->sqe)) - req->work.flags |= IO_WQ_WORK_NEEDS_USER; + /* fall-through */ + case IORING_OP_READV: + case IORING_OP_READ_FIXED: + case IORING_OP_SENDMSG: + case IORING_OP_RECVMSG: + case IORING_OP_ACCEPT: + case IORING_OP_POLL_ADD: + case IORING_OP_CONNECT: + /* + * We know REQ_F_ISREG is not set on some of these + * opcodes, but this enables us to keep the check in + * just one place. + */ + if (!(req->flags & REQ_F_ISREG)) + req->work.flags |= IO_WQ_WORK_UNBOUND; + break; } + if (io_req_needs_user(req)) + req->work.flags |= IO_WQ_WORK_NEEDS_USER; *link = io_prep_linked_timeout(req); return do_hashed; @@ -969,7 +1021,7 @@ static void io_fail_links(struct io_kiocb *req) trace_io_uring_fail_link(req, link); if ((req->flags & REQ_F_LINK_TIMEOUT) && - link->sqe->opcode == IORING_OP_LINK_TIMEOUT) { + link->opcode == IORING_OP_LINK_TIMEOUT) { io_link_cancel_timeout(link); } else { io_cqring_fill_event(link, -ECANCELED); @@ -1145,7 +1197,7 @@ static int io_do_iopoll(struct io_ring_ctx *ctx, unsigned int *nr_events, ret = 0; list_for_each_entry_safe(req, tmp, &ctx->poll_list, list) { - struct kiocb *kiocb = &req->rw; + struct kiocb *kiocb = &req->rw.kiocb; /* * Move completed entries to our local list. If we find a @@ -1175,7 +1227,7 @@ static int io_do_iopoll(struct io_ring_ctx *ctx, unsigned int *nr_events, } /* - * Poll for a mininum of 'min' events. Note that if min == 0 we consider that a + * Poll for a minimum of 'min' events. Note that if min == 0 we consider that a * non-spinning poll check - we'll still enter the driver poll loop, but only * as a non-spinning completion check. */ @@ -1292,21 +1344,27 @@ static void kiocb_end_write(struct io_kiocb *req) file_end_write(req->file); } +static inline void req_set_fail_links(struct io_kiocb *req) +{ + if ((req->flags & (REQ_F_LINK | REQ_F_HARDLINK)) == REQ_F_LINK) + req->flags |= REQ_F_FAIL_LINK; +} + static void io_complete_rw_common(struct kiocb *kiocb, long res) { - struct io_kiocb *req = container_of(kiocb, struct io_kiocb, rw); + struct io_kiocb *req = container_of(kiocb, struct io_kiocb, rw.kiocb); if (kiocb->ki_flags & IOCB_WRITE) kiocb_end_write(req); - if ((req->flags & REQ_F_LINK) && res != req->result) - req->flags |= REQ_F_FAIL_LINK; + if (res != req->result) + req_set_fail_links(req); io_cqring_add_event(req, res); } static void io_complete_rw(struct kiocb *kiocb, long res, long res2) { - struct io_kiocb *req = container_of(kiocb, struct io_kiocb, rw); + struct io_kiocb *req = container_of(kiocb, struct io_kiocb, rw.kiocb); io_complete_rw_common(kiocb, res); io_put_req(req); @@ -1314,7 +1372,7 @@ static void io_complete_rw(struct kiocb *kiocb, long res, long res2) static struct io_kiocb *__io_complete_rw(struct kiocb *kiocb, long res) { - struct io_kiocb *req = container_of(kiocb, struct io_kiocb, rw); + struct io_kiocb *req = container_of(kiocb, struct io_kiocb, rw.kiocb); struct io_kiocb *nxt = NULL; io_complete_rw_common(kiocb, res); @@ -1325,13 +1383,13 @@ static struct io_kiocb *__io_complete_rw(struct kiocb *kiocb, long res) static void io_complete_rw_iopoll(struct kiocb *kiocb, long res, long res2) { - struct io_kiocb *req = container_of(kiocb, struct io_kiocb, rw); + struct io_kiocb *req = container_of(kiocb, struct io_kiocb, rw.kiocb); if (kiocb->ki_flags & IOCB_WRITE) kiocb_end_write(req); - if ((req->flags & REQ_F_LINK) && res != req->result) - req->flags |= REQ_F_FAIL_LINK; + if (res != req->result) + req_set_fail_links(req); req->result = res; if (res != -EAGAIN) req->flags |= REQ_F_IOPOLL_COMPLETED; @@ -1359,7 +1417,7 @@ static void io_iopoll_req_issued(struct io_kiocb *req) list_req = list_first_entry(&ctx->poll_list, struct io_kiocb, list); - if (list_req->rw.ki_filp != req->rw.ki_filp) + if (list_req->file != req->file) ctx->poll_multi_file = true; } @@ -1422,7 +1480,7 @@ static bool io_file_supports_async(struct file *file) { umode_t mode = file_inode(file)->i_mode; - if (S_ISBLK(mode) || S_ISCHR(mode)) + if (S_ISBLK(mode) || S_ISCHR(mode) || S_ISSOCK(mode)) return true; if (S_ISREG(mode) && file->f_op != &io_uring_fops) return true; @@ -1430,11 +1488,11 @@ static bool io_file_supports_async(struct file *file) return false; } -static int io_prep_rw(struct io_kiocb *req, bool force_nonblock) +static int io_prep_rw(struct io_kiocb *req, const struct io_uring_sqe *sqe, + bool force_nonblock) { - const struct io_uring_sqe *sqe = req->sqe; struct io_ring_ctx *ctx = req->ctx; - struct kiocb *kiocb = &req->rw; + struct kiocb *kiocb = &req->rw.kiocb; unsigned ioprio; int ret; @@ -1483,6 +1541,12 @@ static int io_prep_rw(struct io_kiocb *req, bool force_nonblock) return -EINVAL; kiocb->ki_complete = io_complete_rw; } + + req->rw.addr = READ_ONCE(sqe->addr); + req->rw.len = READ_ONCE(sqe->len); + /* we own ->private, reuse it for the buffer index */ + req->rw.kiocb.private = (void *) (unsigned long) + READ_ONCE(sqe->buf_index); return 0; } @@ -1516,11 +1580,11 @@ static void kiocb_done(struct kiocb *kiocb, ssize_t ret, struct io_kiocb **nxt, io_rw_done(kiocb, ret); } -static ssize_t io_import_fixed(struct io_ring_ctx *ctx, int rw, - const struct io_uring_sqe *sqe, +static ssize_t io_import_fixed(struct io_kiocb *req, int rw, struct iov_iter *iter) { - size_t len = READ_ONCE(sqe->len); + struct io_ring_ctx *ctx = req->ctx; + size_t len = req->rw.len; struct io_mapped_ubuf *imu; unsigned index, buf_index; size_t offset; @@ -1530,13 +1594,13 @@ static ssize_t io_import_fixed(struct io_ring_ctx *ctx, int rw, if (unlikely(!ctx->user_bufs)) return -EFAULT; - buf_index = READ_ONCE(sqe->buf_index); + buf_index = (unsigned long) req->rw.kiocb.private; if (unlikely(buf_index >= ctx->nr_user_bufs)) return -EFAULT; index = array_index_nospec(buf_index, ctx->nr_user_bufs); imu = &ctx->user_bufs[index]; - buf_addr = READ_ONCE(sqe->addr); + buf_addr = req->rw.addr; /* overflow */ if (buf_addr + len < buf_addr) @@ -1593,25 +1657,20 @@ static ssize_t io_import_fixed(struct io_ring_ctx *ctx, int rw, static ssize_t io_import_iovec(int rw, struct io_kiocb *req, struct iovec **iovec, struct iov_iter *iter) { - const struct io_uring_sqe *sqe = req->sqe; - void __user *buf = u64_to_user_ptr(READ_ONCE(sqe->addr)); - size_t sqe_len = READ_ONCE(sqe->len); + void __user *buf = u64_to_user_ptr(req->rw.addr); + size_t sqe_len = req->rw.len; u8 opcode; - /* - * We're reading ->opcode for the second time, but the first read - * doesn't care whether it's _FIXED or not, so it doesn't matter - * whether ->opcode changes concurrently. The first read does care - * about whether it is a READ or a WRITE, so we don't trust this read - * for that purpose and instead let the caller pass in the read/write - * flag. - */ - opcode = READ_ONCE(sqe->opcode); + opcode = req->opcode; if (opcode == IORING_OP_READ_FIXED || opcode == IORING_OP_WRITE_FIXED) { *iovec = NULL; - return io_import_fixed(req->ctx, rw, sqe, iter); + return io_import_fixed(req, rw, iter); } + /* buffer index only valid with fixed read/write */ + if (req->rw.kiocb.private) + return -EINVAL; + if (req->io) { struct io_async_rw *iorw = &req->io->rw; @@ -1692,7 +1751,7 @@ static ssize_t loop_rw_iter(int rw, struct file *file, struct kiocb *kiocb, return ret; } -static void io_req_map_io(struct io_kiocb *req, ssize_t io_size, +static void io_req_map_rw(struct io_kiocb *req, ssize_t io_size, struct iovec *iovec, struct iovec *fast_iov, struct iov_iter *iter) { @@ -1706,57 +1765,85 @@ static void io_req_map_io(struct io_kiocb *req, ssize_t io_size, } } -static int io_setup_async_io(struct io_kiocb *req, ssize_t io_size, +static int io_alloc_async_ctx(struct io_kiocb *req) +{ + req->io = kmalloc(sizeof(*req->io), GFP_KERNEL); + return req->io == NULL; +} + +static void io_rw_async(struct io_wq_work **workptr) +{ + struct io_kiocb *req = container_of(*workptr, struct io_kiocb, work); + struct iovec *iov = NULL; + + if (req->io->rw.iov != req->io->rw.fast_iov) + iov = req->io->rw.iov; + io_wq_submit_work(workptr); + kfree(iov); +} + +static int io_setup_async_rw(struct io_kiocb *req, ssize_t io_size, struct iovec *iovec, struct iovec *fast_iov, struct iov_iter *iter) { - req->io = kmalloc(sizeof(*req->io), GFP_KERNEL); - if (req->io) { - io_req_map_io(req, io_size, iovec, fast_iov, iter); - memcpy(&req->io->sqe, req->sqe, sizeof(req->io->sqe)); - req->sqe = &req->io->sqe; + if (req->opcode == IORING_OP_READ_FIXED || + req->opcode == IORING_OP_WRITE_FIXED) return 0; - } + if (!req->io && io_alloc_async_ctx(req)) + return -ENOMEM; - return -ENOMEM; + io_req_map_rw(req, io_size, iovec, fast_iov, iter); + req->work.func = io_rw_async; + return 0; } -static int io_read_prep(struct io_kiocb *req, struct iovec **iovec, - struct iov_iter *iter, bool force_nonblock) +static int io_read_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe, + bool force_nonblock) { + struct io_async_ctx *io; + struct iov_iter iter; ssize_t ret; - ret = io_prep_rw(req, force_nonblock); + ret = io_prep_rw(req, sqe, force_nonblock); if (ret) return ret; if (unlikely(!(req->file->f_mode & FMODE_READ))) return -EBADF; - return io_import_iovec(READ, req, iovec, iter); + if (!req->io) + return 0; + + io = req->io; + io->rw.iov = io->rw.fast_iov; + req->io = NULL; + ret = io_import_iovec(READ, req, &io->rw.iov, &iter); + req->io = io; + if (ret < 0) + return ret; + + io_req_map_rw(req, ret, io->rw.iov, io->rw.fast_iov, &iter); + return 0; } static int io_read(struct io_kiocb *req, struct io_kiocb **nxt, bool force_nonblock) { struct iovec inline_vecs[UIO_FASTIOV], *iovec = inline_vecs; - struct kiocb *kiocb = &req->rw; + struct kiocb *kiocb = &req->rw.kiocb; struct iov_iter iter; - struct file *file; size_t iov_count; ssize_t io_size, ret; - if (!req->io) { - ret = io_read_prep(req, &iovec, &iter, force_nonblock); - if (ret < 0) - return ret; - } else { - ret = io_import_iovec(READ, req, &iovec, &iter); - if (ret < 0) - return ret; - } + ret = io_import_iovec(READ, req, &iovec, &iter); + if (ret < 0) + return ret; + + /* Ensure we clear previously set non-block flag */ + if (!force_nonblock) + req->rw.kiocb.ki_flags &= ~IOCB_NOWAIT; - file = req->file; + req->result = 0; io_size = ret; if (req->flags & REQ_F_LINK) req->result = io_size; @@ -1765,39 +1852,27 @@ static int io_read(struct io_kiocb *req, struct io_kiocb **nxt, * If the file doesn't support async, mark it as REQ_F_MUST_PUNT so * we know to async punt it even if it was opened O_NONBLOCK */ - if (force_nonblock && !io_file_supports_async(file)) { + if (force_nonblock && !io_file_supports_async(req->file)) { req->flags |= REQ_F_MUST_PUNT; goto copy_iov; } iov_count = iov_iter_count(&iter); - ret = rw_verify_area(READ, file, &kiocb->ki_pos, iov_count); + ret = rw_verify_area(READ, req->file, &kiocb->ki_pos, iov_count); if (!ret) { ssize_t ret2; - if (file->f_op->read_iter) - ret2 = call_read_iter(file, kiocb, &iter); + if (req->file->f_op->read_iter) + ret2 = call_read_iter(req->file, kiocb, &iter); else - ret2 = loop_rw_iter(READ, file, kiocb, &iter); + ret2 = loop_rw_iter(READ, req->file, kiocb, &iter); - /* - * In case of a short read, punt to async. This can happen - * if we have data partially cached. Alternatively we can - * return the short read, in which case the application will - * need to issue another SQE and wait for it. That SQE will - * need async punt anyway, so it's more efficient to do it - * here. - */ - if (force_nonblock && !(req->flags & REQ_F_NOWAIT) && - (req->flags & REQ_F_ISREG) && - ret2 > 0 && ret2 < io_size) - ret2 = -EAGAIN; /* Catch -EAGAIN return for forced non-blocking submission */ if (!force_nonblock || ret2 != -EAGAIN) { kiocb_done(kiocb, ret2, nxt, req->in_async); } else { copy_iov: - ret = io_setup_async_io(req, io_size, iovec, + ret = io_setup_async_rw(req, io_size, iovec, inline_vecs, &iter); if (ret) goto out_free; @@ -1805,46 +1880,58 @@ copy_iov: } } out_free: - kfree(iovec); + if (!io_wq_current_is_worker()) + kfree(iovec); return ret; } -static int io_write_prep(struct io_kiocb *req, struct iovec **iovec, - struct iov_iter *iter, bool force_nonblock) +static int io_write_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe, + bool force_nonblock) { + struct io_async_ctx *io; + struct iov_iter iter; ssize_t ret; - ret = io_prep_rw(req, force_nonblock); + ret = io_prep_rw(req, sqe, force_nonblock); if (ret) return ret; if (unlikely(!(req->file->f_mode & FMODE_WRITE))) return -EBADF; - return io_import_iovec(WRITE, req, iovec, iter); + if (!req->io) + return 0; + + io = req->io; + io->rw.iov = io->rw.fast_iov; + req->io = NULL; + ret = io_import_iovec(WRITE, req, &io->rw.iov, &iter); + req->io = io; + if (ret < 0) + return ret; + + io_req_map_rw(req, ret, io->rw.iov, io->rw.fast_iov, &iter); + return 0; } static int io_write(struct io_kiocb *req, struct io_kiocb **nxt, bool force_nonblock) { struct iovec inline_vecs[UIO_FASTIOV], *iovec = inline_vecs; - struct kiocb *kiocb = &req->rw; + struct kiocb *kiocb = &req->rw.kiocb; struct iov_iter iter; - struct file *file; size_t iov_count; ssize_t ret, io_size; - if (!req->io) { - ret = io_write_prep(req, &iovec, &iter, force_nonblock); - if (ret < 0) - return ret; - } else { - ret = io_import_iovec(WRITE, req, &iovec, &iter); - if (ret < 0) - return ret; - } + ret = io_import_iovec(WRITE, req, &iovec, &iter); + if (ret < 0) + return ret; - file = kiocb->ki_filp; + /* Ensure we clear previously set non-block flag */ + if (!force_nonblock) + req->rw.kiocb.ki_flags &= ~IOCB_NOWAIT; + + req->result = 0; io_size = ret; if (req->flags & REQ_F_LINK) req->result = io_size; @@ -1858,11 +1945,13 @@ static int io_write(struct io_kiocb *req, struct io_kiocb **nxt, goto copy_iov; } - if (force_nonblock && !(kiocb->ki_flags & IOCB_DIRECT)) + /* file path doesn't support NOWAIT for non-direct_IO */ + if (force_nonblock && !(kiocb->ki_flags & IOCB_DIRECT) && + (req->flags & REQ_F_ISREG)) goto copy_iov; iov_count = iov_iter_count(&iter); - ret = rw_verify_area(WRITE, file, &kiocb->ki_pos, iov_count); + ret = rw_verify_area(WRITE, req->file, &kiocb->ki_pos, iov_count); if (!ret) { ssize_t ret2; @@ -1874,22 +1963,22 @@ static int io_write(struct io_kiocb *req, struct io_kiocb **nxt, * we return to userspace. */ if (req->flags & REQ_F_ISREG) { - __sb_start_write(file_inode(file)->i_sb, + __sb_start_write(file_inode(req->file)->i_sb, SB_FREEZE_WRITE, true); - __sb_writers_release(file_inode(file)->i_sb, + __sb_writers_release(file_inode(req->file)->i_sb, SB_FREEZE_WRITE); } kiocb->ki_flags |= IOCB_WRITE; - if (file->f_op->write_iter) - ret2 = call_write_iter(file, kiocb, &iter); + if (req->file->f_op->write_iter) + ret2 = call_write_iter(req->file, kiocb, &iter); else - ret2 = loop_rw_iter(WRITE, file, kiocb, &iter); + ret2 = loop_rw_iter(WRITE, req->file, kiocb, &iter); if (!force_nonblock || ret2 != -EAGAIN) { kiocb_done(kiocb, ret2, nxt, req->in_async); } else { copy_iov: - ret = io_setup_async_io(req, io_size, iovec, + ret = io_setup_async_rw(req, io_size, iovec, inline_vecs, &iter); if (ret) goto out_free; @@ -1897,7 +1986,8 @@ copy_iov: } } out_free: - kfree(iovec); + if (!io_wq_current_is_worker()) + kfree(iovec); return ret; } @@ -1928,45 +2018,92 @@ static int io_prep_fsync(struct io_kiocb *req, const struct io_uring_sqe *sqe) if (unlikely(sqe->addr || sqe->ioprio || sqe->buf_index)) return -EINVAL; + req->sync.flags = READ_ONCE(sqe->fsync_flags); + if (unlikely(req->sync.flags & ~IORING_FSYNC_DATASYNC)) + return -EINVAL; + + req->sync.off = READ_ONCE(sqe->off); + req->sync.len = READ_ONCE(sqe->len); return 0; } -static int io_fsync(struct io_kiocb *req, const struct io_uring_sqe *sqe, - struct io_kiocb **nxt, bool force_nonblock) +static bool io_req_cancelled(struct io_kiocb *req) +{ + if (req->work.flags & IO_WQ_WORK_CANCEL) { + req_set_fail_links(req); + io_cqring_add_event(req, -ECANCELED); + io_put_req(req); + return true; + } + + return false; +} + +static void io_link_work_cb(struct io_wq_work **workptr) +{ + struct io_wq_work *work = *workptr; + struct io_kiocb *link = work->data; + + io_queue_linked_timeout(link); + work->func = io_wq_submit_work; +} + +static void io_wq_assign_next(struct io_wq_work **workptr, struct io_kiocb *nxt) +{ + struct io_kiocb *link; + + io_prep_async_work(nxt, &link); + *workptr = &nxt->work; + if (link) { + nxt->work.flags |= IO_WQ_WORK_CB; + nxt->work.func = io_link_work_cb; + nxt->work.data = link; + } +} + +static void io_fsync_finish(struct io_wq_work **workptr) { - loff_t sqe_off = READ_ONCE(sqe->off); - loff_t sqe_len = READ_ONCE(sqe->len); - loff_t end = sqe_off + sqe_len; - unsigned fsync_flags; + struct io_kiocb *req = container_of(*workptr, struct io_kiocb, work); + loff_t end = req->sync.off + req->sync.len; + struct io_kiocb *nxt = NULL; int ret; - fsync_flags = READ_ONCE(sqe->fsync_flags); - if (unlikely(fsync_flags & ~IORING_FSYNC_DATASYNC)) - return -EINVAL; + if (io_req_cancelled(req)) + return; - ret = io_prep_fsync(req, sqe); - if (ret) - return ret; + ret = vfs_fsync_range(req->file, req->sync.off, + end > 0 ? end : LLONG_MAX, + req->sync.flags & IORING_FSYNC_DATASYNC); + if (ret < 0) + req_set_fail_links(req); + io_cqring_add_event(req, ret); + io_put_req_find_next(req, &nxt); + if (nxt) + io_wq_assign_next(workptr, nxt); +} + +static int io_fsync(struct io_kiocb *req, struct io_kiocb **nxt, + bool force_nonblock) +{ + struct io_wq_work *work, *old_work; /* fsync always requires a blocking context */ - if (force_nonblock) + if (force_nonblock) { + io_put_req(req); + req->work.func = io_fsync_finish; return -EAGAIN; + } - ret = vfs_fsync_range(req->rw.ki_filp, sqe_off, - end > 0 ? end : LLONG_MAX, - fsync_flags & IORING_FSYNC_DATASYNC); - - if (ret < 0 && (req->flags & REQ_F_LINK)) - req->flags |= REQ_F_FAIL_LINK; - io_cqring_add_event(req, ret); - io_put_req_find_next(req, nxt); + work = old_work = &req->work; + io_fsync_finish(&work); + if (work && work != old_work) + *nxt = container_of(work, struct io_kiocb, work); return 0; } static int io_prep_sfr(struct io_kiocb *req, const struct io_uring_sqe *sqe) { struct io_ring_ctx *ctx = req->ctx; - int ret = 0; if (!req->file) return -EBADF; @@ -1976,59 +2113,88 @@ static int io_prep_sfr(struct io_kiocb *req, const struct io_uring_sqe *sqe) if (unlikely(sqe->addr || sqe->ioprio || sqe->buf_index)) return -EINVAL; - return ret; + req->sync.off = READ_ONCE(sqe->off); + req->sync.len = READ_ONCE(sqe->len); + req->sync.flags = READ_ONCE(sqe->sync_range_flags); + return 0; } -static int io_sync_file_range(struct io_kiocb *req, - const struct io_uring_sqe *sqe, - struct io_kiocb **nxt, - bool force_nonblock) +static void io_sync_file_range_finish(struct io_wq_work **workptr) { - loff_t sqe_off; - loff_t sqe_len; - unsigned flags; + struct io_kiocb *req = container_of(*workptr, struct io_kiocb, work); + struct io_kiocb *nxt = NULL; int ret; - ret = io_prep_sfr(req, sqe); - if (ret) - return ret; + if (io_req_cancelled(req)) + return; + + ret = sync_file_range(req->file, req->sync.off, req->sync.len, + req->sync.flags); + if (ret < 0) + req_set_fail_links(req); + io_cqring_add_event(req, ret); + io_put_req_find_next(req, &nxt); + if (nxt) + io_wq_assign_next(workptr, nxt); +} + +static int io_sync_file_range(struct io_kiocb *req, struct io_kiocb **nxt, + bool force_nonblock) +{ + struct io_wq_work *work, *old_work; /* sync_file_range always requires a blocking context */ - if (force_nonblock) + if (force_nonblock) { + io_put_req(req); + req->work.func = io_sync_file_range_finish; return -EAGAIN; + } - sqe_off = READ_ONCE(sqe->off); - sqe_len = READ_ONCE(sqe->len); - flags = READ_ONCE(sqe->sync_range_flags); + work = old_work = &req->work; + io_sync_file_range_finish(&work); + if (work && work != old_work) + *nxt = container_of(work, struct io_kiocb, work); + return 0; +} - ret = sync_file_range(req->rw.ki_filp, sqe_off, sqe_len, flags); +#if defined(CONFIG_NET) +static void io_sendrecv_async(struct io_wq_work **workptr) +{ + struct io_kiocb *req = container_of(*workptr, struct io_kiocb, work); + struct iovec *iov = NULL; - if (ret < 0 && (req->flags & REQ_F_LINK)) - req->flags |= REQ_F_FAIL_LINK; - io_cqring_add_event(req, ret); - io_put_req_find_next(req, nxt); - return 0; + if (req->io->rw.iov != req->io->rw.fast_iov) + iov = req->io->msg.iov; + io_wq_submit_work(workptr); + kfree(iov); } +#endif -static int io_sendmsg_prep(struct io_kiocb *req, struct io_async_ctx *io) +static int io_sendmsg_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe) { #if defined(CONFIG_NET) - const struct io_uring_sqe *sqe = req->sqe; - struct user_msghdr __user *msg; - unsigned flags; + struct io_sr_msg *sr = &req->sr_msg; + struct io_async_ctx *io = req->io; + + sr->msg_flags = READ_ONCE(sqe->msg_flags); + sr->msg = u64_to_user_ptr(READ_ONCE(sqe->addr)); + + if (!io) + return 0; - flags = READ_ONCE(sqe->msg_flags); - msg = (struct user_msghdr __user *)(unsigned long) READ_ONCE(sqe->addr); - return sendmsg_copy_msghdr(&io->msg.msg, msg, flags, &io->msg.iov); + io->msg.iov = io->msg.fast_iov; + return sendmsg_copy_msghdr(&io->msg.msg, sr->msg, sr->msg_flags, + &io->msg.iov); #else - return 0; + return -EOPNOTSUPP; #endif } -static int io_sendmsg(struct io_kiocb *req, const struct io_uring_sqe *sqe, - struct io_kiocb **nxt, bool force_nonblock) +static int io_sendmsg(struct io_kiocb *req, struct io_kiocb **nxt, + bool force_nonblock) { #if defined(CONFIG_NET) + struct io_async_msghdr *kmsg = NULL; struct socket *sock; int ret; @@ -2037,50 +2203,55 @@ static int io_sendmsg(struct io_kiocb *req, const struct io_uring_sqe *sqe, sock = sock_from_file(req->file, &ret); if (sock) { - struct io_async_ctx io, *copy; + struct io_async_ctx io; struct sockaddr_storage addr; - struct msghdr *kmsg; unsigned flags; - flags = READ_ONCE(sqe->msg_flags); - if (flags & MSG_DONTWAIT) - req->flags |= REQ_F_NOWAIT; - else if (force_nonblock) - flags |= MSG_DONTWAIT; - if (req->io) { - kmsg = &req->io->msg.msg; - kmsg->msg_name = &addr; + kmsg = &req->io->msg; + kmsg->msg.msg_name = &addr; + /* if iov is set, it's allocated already */ + if (!kmsg->iov) + kmsg->iov = kmsg->fast_iov; + kmsg->msg.msg_iter.iov = kmsg->iov; } else { - kmsg = &io.msg.msg; - kmsg->msg_name = &addr; + struct io_sr_msg *sr = &req->sr_msg; + + kmsg = &io.msg; + kmsg->msg.msg_name = &addr; + io.msg.iov = io.msg.fast_iov; - ret = io_sendmsg_prep(req, &io); + ret = sendmsg_copy_msghdr(&io.msg.msg, sr->msg, + sr->msg_flags, &io.msg.iov); if (ret) - goto out; + return ret; } - ret = __sys_sendmsg_sock(sock, kmsg, flags); + flags = req->sr_msg.msg_flags; + if (flags & MSG_DONTWAIT) + req->flags |= REQ_F_NOWAIT; + else if (force_nonblock) + flags |= MSG_DONTWAIT; + + ret = __sys_sendmsg_sock(sock, &kmsg->msg, flags); if (force_nonblock && ret == -EAGAIN) { - copy = kmalloc(sizeof(*copy), GFP_KERNEL); - if (!copy) { - ret = -ENOMEM; - goto out; - } - memcpy(©->msg, &io.msg, sizeof(copy->msg)); - req->io = copy; - memcpy(&req->io->sqe, req->sqe, sizeof(*req->sqe)); - req->sqe = &req->io->sqe; - return ret; + if (req->io) + return -EAGAIN; + if (io_alloc_async_ctx(req)) + return -ENOMEM; + memcpy(&req->io->msg, &io.msg, sizeof(io.msg)); + req->work.func = io_sendrecv_async; + return -EAGAIN; } if (ret == -ERESTARTSYS) ret = -EINTR; } -out: + if (!io_wq_current_is_worker() && kmsg && kmsg->iov != kmsg->fast_iov) + kfree(kmsg->iov); io_cqring_add_event(req, ret); - if (ret < 0 && (req->flags & REQ_F_LINK)) - req->flags |= REQ_F_FAIL_LINK; + if (ret < 0) + req_set_fail_links(req); io_put_req_find_next(req, nxt); return 0; #else @@ -2088,26 +2259,32 @@ out: #endif } -static int io_recvmsg_prep(struct io_kiocb *req, struct io_async_ctx *io) +static int io_recvmsg_prep(struct io_kiocb *req, + const struct io_uring_sqe *sqe) { #if defined(CONFIG_NET) - const struct io_uring_sqe *sqe = req->sqe; - struct user_msghdr __user *msg; - unsigned flags; + struct io_sr_msg *sr = &req->sr_msg; + struct io_async_ctx *io = req->io; - flags = READ_ONCE(sqe->msg_flags); - msg = (struct user_msghdr __user *)(unsigned long) READ_ONCE(sqe->addr); - return recvmsg_copy_msghdr(&io->msg.msg, msg, flags, &io->msg.uaddr, - &io->msg.iov); + sr->msg_flags = READ_ONCE(sqe->msg_flags); + sr->msg = u64_to_user_ptr(READ_ONCE(sqe->addr)); + + if (!io) + return 0; + + io->msg.iov = io->msg.fast_iov; + return recvmsg_copy_msghdr(&io->msg.msg, sr->msg, sr->msg_flags, + &io->msg.uaddr, &io->msg.iov); #else - return 0; + return -EOPNOTSUPP; #endif } -static int io_recvmsg(struct io_kiocb *req, const struct io_uring_sqe *sqe, - struct io_kiocb **nxt, bool force_nonblock) +static int io_recvmsg(struct io_kiocb *req, struct io_kiocb **nxt, + bool force_nonblock) { #if defined(CONFIG_NET) + struct io_async_msghdr *kmsg = NULL; struct socket *sock; int ret; @@ -2116,53 +2293,57 @@ static int io_recvmsg(struct io_kiocb *req, const struct io_uring_sqe *sqe, sock = sock_from_file(req->file, &ret); if (sock) { - struct user_msghdr __user *msg; - struct io_async_ctx io, *copy; + struct io_async_ctx io; struct sockaddr_storage addr; - struct msghdr *kmsg; unsigned flags; - flags = READ_ONCE(sqe->msg_flags); - if (flags & MSG_DONTWAIT) - req->flags |= REQ_F_NOWAIT; - else if (force_nonblock) - flags |= MSG_DONTWAIT; - - msg = (struct user_msghdr __user *) (unsigned long) - READ_ONCE(sqe->addr); if (req->io) { - kmsg = &req->io->msg.msg; - kmsg->msg_name = &addr; + kmsg = &req->io->msg; + kmsg->msg.msg_name = &addr; + /* if iov is set, it's allocated already */ + if (!kmsg->iov) + kmsg->iov = kmsg->fast_iov; + kmsg->msg.msg_iter.iov = kmsg->iov; } else { - kmsg = &io.msg.msg; - kmsg->msg_name = &addr; + struct io_sr_msg *sr = &req->sr_msg; + + kmsg = &io.msg; + kmsg->msg.msg_name = &addr; + io.msg.iov = io.msg.fast_iov; - ret = io_recvmsg_prep(req, &io); + ret = recvmsg_copy_msghdr(&io.msg.msg, sr->msg, + sr->msg_flags, &io.msg.uaddr, + &io.msg.iov); if (ret) - goto out; + return ret; } - ret = __sys_recvmsg_sock(sock, kmsg, msg, io.msg.uaddr, flags); + flags = req->sr_msg.msg_flags; + if (flags & MSG_DONTWAIT) + req->flags |= REQ_F_NOWAIT; + else if (force_nonblock) + flags |= MSG_DONTWAIT; + + ret = __sys_recvmsg_sock(sock, &kmsg->msg, req->sr_msg.msg, + kmsg->uaddr, flags); if (force_nonblock && ret == -EAGAIN) { - copy = kmalloc(sizeof(*copy), GFP_KERNEL); - if (!copy) { - ret = -ENOMEM; - goto out; - } - memcpy(copy, &io, sizeof(*copy)); - req->io = copy; - memcpy(&req->io->sqe, req->sqe, sizeof(*req->sqe)); - req->sqe = &req->io->sqe; - return ret; + if (req->io) + return -EAGAIN; + if (io_alloc_async_ctx(req)) + return -ENOMEM; + memcpy(&req->io->msg, &io.msg, sizeof(io.msg)); + req->work.func = io_sendrecv_async; + return -EAGAIN; } if (ret == -ERESTARTSYS) ret = -EINTR; } -out: + if (!io_wq_current_is_worker() && kmsg && kmsg->iov != kmsg->fast_iov) + kfree(kmsg->iov); io_cqring_add_event(req, ret); - if (ret < 0 && (req->flags & REQ_F_LINK)) - req->flags |= REQ_F_FAIL_LINK; + if (ret < 0) + req_set_fail_links(req); io_put_req_find_next(req, nxt); return 0; #else @@ -2170,101 +2351,141 @@ out: #endif } -static int io_accept(struct io_kiocb *req, const struct io_uring_sqe *sqe, - struct io_kiocb **nxt, bool force_nonblock) +static int io_accept_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe) { #if defined(CONFIG_NET) - struct sockaddr __user *addr; - int __user *addr_len; - unsigned file_flags; - int flags, ret; + struct io_accept *accept = &req->accept; if (unlikely(req->ctx->flags & (IORING_SETUP_IOPOLL|IORING_SETUP_SQPOLL))) return -EINVAL; if (sqe->ioprio || sqe->len || sqe->buf_index) return -EINVAL; - addr = (struct sockaddr __user *) (unsigned long) READ_ONCE(sqe->addr); - addr_len = (int __user *) (unsigned long) READ_ONCE(sqe->addr2); - flags = READ_ONCE(sqe->accept_flags); - file_flags = force_nonblock ? O_NONBLOCK : 0; + accept->addr = u64_to_user_ptr(READ_ONCE(sqe->addr)); + accept->addr_len = u64_to_user_ptr(READ_ONCE(sqe->addr2)); + accept->flags = READ_ONCE(sqe->accept_flags); + return 0; +#else + return -EOPNOTSUPP; +#endif +} - ret = __sys_accept4_file(req->file, file_flags, addr, addr_len, flags); - if (ret == -EAGAIN && force_nonblock) { - req->work.flags |= IO_WQ_WORK_NEEDS_FILES; +#if defined(CONFIG_NET) +static int __io_accept(struct io_kiocb *req, struct io_kiocb **nxt, + bool force_nonblock) +{ + struct io_accept *accept = &req->accept; + unsigned file_flags; + int ret; + + file_flags = force_nonblock ? O_NONBLOCK : 0; + ret = __sys_accept4_file(req->file, file_flags, accept->addr, + accept->addr_len, accept->flags); + if (ret == -EAGAIN && force_nonblock) return -EAGAIN; - } if (ret == -ERESTARTSYS) ret = -EINTR; - if (ret < 0 && (req->flags & REQ_F_LINK)) - req->flags |= REQ_F_FAIL_LINK; + if (ret < 0) + req_set_fail_links(req); io_cqring_add_event(req, ret); io_put_req_find_next(req, nxt); return 0; -#else - return -EOPNOTSUPP; -#endif } -static int io_connect_prep(struct io_kiocb *req, struct io_async_ctx *io) +static void io_accept_finish(struct io_wq_work **workptr) +{ + struct io_kiocb *req = container_of(*workptr, struct io_kiocb, work); + struct io_kiocb *nxt = NULL; + + if (io_req_cancelled(req)) + return; + __io_accept(req, &nxt, false); + if (nxt) + io_wq_assign_next(workptr, nxt); +} +#endif + +static int io_accept(struct io_kiocb *req, struct io_kiocb **nxt, + bool force_nonblock) { #if defined(CONFIG_NET) - const struct io_uring_sqe *sqe = req->sqe; - struct sockaddr __user *addr; - int addr_len; + int ret; - addr = (struct sockaddr __user *) (unsigned long) READ_ONCE(sqe->addr); - addr_len = READ_ONCE(sqe->addr2); - return move_addr_to_kernel(addr, addr_len, &io->connect.address); -#else + ret = __io_accept(req, nxt, force_nonblock); + if (ret == -EAGAIN && force_nonblock) { + req->work.func = io_accept_finish; + req->work.flags |= IO_WQ_WORK_NEEDS_FILES; + io_put_req(req); + return -EAGAIN; + } return 0; +#else + return -EOPNOTSUPP; #endif } -static int io_connect(struct io_kiocb *req, const struct io_uring_sqe *sqe, - struct io_kiocb **nxt, bool force_nonblock) +static int io_connect_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe) { #if defined(CONFIG_NET) - struct io_async_ctx __io, *io; - unsigned file_flags; - int addr_len, ret; + struct io_connect *conn = &req->connect; + struct io_async_ctx *io = req->io; if (unlikely(req->ctx->flags & (IORING_SETUP_IOPOLL|IORING_SETUP_SQPOLL))) return -EINVAL; if (sqe->ioprio || sqe->len || sqe->buf_index || sqe->rw_flags) return -EINVAL; - addr_len = READ_ONCE(sqe->addr2); - file_flags = force_nonblock ? O_NONBLOCK : 0; + conn->addr = u64_to_user_ptr(READ_ONCE(sqe->addr)); + conn->addr_len = READ_ONCE(sqe->addr2); + + if (!io) + return 0; + + return move_addr_to_kernel(conn->addr, conn->addr_len, + &io->connect.address); +#else + return -EOPNOTSUPP; +#endif +} + +static int io_connect(struct io_kiocb *req, struct io_kiocb **nxt, + bool force_nonblock) +{ +#if defined(CONFIG_NET) + struct io_async_ctx __io, *io; + unsigned file_flags; + int ret; if (req->io) { io = req->io; } else { - ret = io_connect_prep(req, &__io); + ret = move_addr_to_kernel(req->connect.addr, + req->connect.addr_len, + &__io.connect.address); if (ret) goto out; io = &__io; } - ret = __sys_connect_file(req->file, &io->connect.address, addr_len, - file_flags); + file_flags = force_nonblock ? O_NONBLOCK : 0; + + ret = __sys_connect_file(req->file, &io->connect.address, + req->connect.addr_len, file_flags); if ((ret == -EAGAIN || ret == -EINPROGRESS) && force_nonblock) { - io = kmalloc(sizeof(*io), GFP_KERNEL); - if (!io) { + if (req->io) + return -EAGAIN; + if (io_alloc_async_ctx(req)) { ret = -ENOMEM; goto out; } - memcpy(&io->connect, &__io.connect, sizeof(io->connect)); - req->io = io; - memcpy(&io->sqe, req->sqe, sizeof(*req->sqe)); - req->sqe = &io->sqe; + memcpy(&req->io->connect, &__io.connect, sizeof(__io.connect)); return -EAGAIN; } if (ret == -ERESTARTSYS) ret = -EINTR; out: - if (ret < 0 && (req->flags & REQ_F_LINK)) - req->flags |= REQ_F_FAIL_LINK; + if (ret < 0) + req_set_fail_links(req); io_cqring_add_event(req, ret); io_put_req_find_next(req, nxt); return 0; @@ -2279,8 +2500,8 @@ static void io_poll_remove_one(struct io_kiocb *req) spin_lock(&poll->head->lock); WRITE_ONCE(poll->canceled, true); - if (!list_empty(&poll->wait->entry)) { - list_del_init(&poll->wait->entry); + if (!list_empty(&poll->wait.entry)) { + list_del_init(&poll->wait.entry); io_queue_async_work(req); } spin_unlock(&poll->head->lock); @@ -2320,28 +2541,37 @@ static int io_poll_cancel(struct io_ring_ctx *ctx, __u64 sqe_addr) return -ENOENT; } +static int io_poll_remove_prep(struct io_kiocb *req, + const struct io_uring_sqe *sqe) +{ + if (unlikely(req->ctx->flags & IORING_SETUP_IOPOLL)) + return -EINVAL; + if (sqe->ioprio || sqe->off || sqe->len || sqe->buf_index || + sqe->poll_events) + return -EINVAL; + + req->poll.addr = READ_ONCE(sqe->addr); + return 0; +} + /* * Find a running poll command that matches one specified in sqe->addr, * and remove it if found. */ -static int io_poll_remove(struct io_kiocb *req, const struct io_uring_sqe *sqe) +static int io_poll_remove(struct io_kiocb *req) { struct io_ring_ctx *ctx = req->ctx; + u64 addr; int ret; - if (unlikely(req->ctx->flags & IORING_SETUP_IOPOLL)) - return -EINVAL; - if (sqe->ioprio || sqe->off || sqe->len || sqe->buf_index || - sqe->poll_events) - return -EINVAL; - + addr = req->poll.addr; spin_lock_irq(&ctx->completion_lock); - ret = io_poll_cancel(ctx, READ_ONCE(sqe->addr)); + ret = io_poll_cancel(ctx, addr); spin_unlock_irq(&ctx->completion_lock); io_cqring_add_event(req, ret); - if (ret < 0 && (req->flags & REQ_F_LINK)) - req->flags |= REQ_F_FAIL_LINK; + if (ret < 0) + req_set_fail_links(req); io_put_req(req); return 0; } @@ -2351,7 +2581,6 @@ static void io_poll_complete(struct io_kiocb *req, __poll_t mask, int error) struct io_ring_ctx *ctx = req->ctx; req->poll.done = true; - kfree(req->poll.wait); if (error) io_cqring_fill_event(req, error); else @@ -2389,7 +2618,7 @@ static void io_poll_complete_work(struct io_wq_work **workptr) */ spin_lock_irq(&ctx->completion_lock); if (!mask && ret != -ECANCELED) { - add_wait_queue(poll->head, poll->wait); + add_wait_queue(poll->head, &poll->wait); spin_unlock_irq(&ctx->completion_lock); return; } @@ -2399,11 +2628,11 @@ static void io_poll_complete_work(struct io_wq_work **workptr) io_cqring_ev_posted(ctx); - if (ret < 0 && req->flags & REQ_F_LINK) - req->flags |= REQ_F_FAIL_LINK; + if (ret < 0) + req_set_fail_links(req); io_put_req_find_next(req, &nxt); if (nxt) - *workptr = &nxt->work; + io_wq_assign_next(workptr, nxt); } static int io_poll_wake(struct wait_queue_entry *wait, unsigned mode, int sync, @@ -2419,7 +2648,7 @@ static int io_poll_wake(struct wait_queue_entry *wait, unsigned mode, int sync, if (mask && !(mask & poll->events)) return 0; - list_del_init(&poll->wait->entry); + list_del_init(&poll->wait.entry); /* * Run completion inline if we can. We're using trylock here because @@ -2460,7 +2689,7 @@ static void io_poll_queue_proc(struct file *file, struct wait_queue_head *head, pt->error = 0; pt->req->poll.head = head; - add_wait_queue(head, pt->req->poll.wait); + add_wait_queue(head, &pt->req->poll.wait); } static void io_poll_req_insert(struct io_kiocb *req) @@ -2472,14 +2701,9 @@ static void io_poll_req_insert(struct io_kiocb *req) hlist_add_head(&req->hash_node, list); } -static int io_poll_add(struct io_kiocb *req, const struct io_uring_sqe *sqe, - struct io_kiocb **nxt) +static int io_poll_add_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe) { struct io_poll_iocb *poll = &req->poll; - struct io_ring_ctx *ctx = req->ctx; - struct io_poll_table ipt; - bool cancel = false; - __poll_t mask; u16 events; if (unlikely(req->ctx->flags & IORING_SETUP_IOPOLL)) @@ -2489,14 +2713,20 @@ static int io_poll_add(struct io_kiocb *req, const struct io_uring_sqe *sqe, if (!poll->file) return -EBADF; - poll->wait = kmalloc(sizeof(*poll->wait), GFP_KERNEL); - if (!poll->wait) - return -ENOMEM; - - req->io = NULL; - INIT_IO_WORK(&req->work, io_poll_complete_work); events = READ_ONCE(sqe->poll_events); poll->events = demangle_poll(events) | EPOLLERR | EPOLLHUP; + return 0; +} + +static int io_poll_add(struct io_kiocb *req, struct io_kiocb **nxt) +{ + struct io_poll_iocb *poll = &req->poll; + struct io_ring_ctx *ctx = req->ctx; + struct io_poll_table ipt; + bool cancel = false; + __poll_t mask; + + INIT_IO_WORK(&req->work, io_poll_complete_work); INIT_HLIST_NODE(&req->hash_node); poll->head = NULL; @@ -2509,9 +2739,9 @@ static int io_poll_add(struct io_kiocb *req, const struct io_uring_sqe *sqe, ipt.error = -EINVAL; /* same as no support for IOCB_CMD_POLL */ /* initialized the list so that we can do list_empty checks */ - INIT_LIST_HEAD(&poll->wait->entry); - init_waitqueue_func_entry(poll->wait, io_poll_wake); - poll->wait->private = poll; + INIT_LIST_HEAD(&poll->wait.entry); + init_waitqueue_func_entry(&poll->wait, io_poll_wake); + poll->wait.private = poll; INIT_LIST_HEAD(&req->list); @@ -2520,14 +2750,14 @@ static int io_poll_add(struct io_kiocb *req, const struct io_uring_sqe *sqe, spin_lock_irq(&ctx->completion_lock); if (likely(poll->head)) { spin_lock(&poll->head->lock); - if (unlikely(list_empty(&poll->wait->entry))) { + if (unlikely(list_empty(&poll->wait.entry))) { if (ipt.error) cancel = true; ipt.error = 0; mask = 0; } if (mask || ipt.error) - list_del_init(&poll->wait->entry); + list_del_init(&poll->wait.entry); else if (cancel) WRITE_ONCE(poll->canceled, true); else if (!poll->done) /* actually waiting for an event */ @@ -2567,7 +2797,7 @@ static enum hrtimer_restart io_timeout_fn(struct hrtimer *timer) /* * Adjust the reqs sequence before the current one because it - * will consume a slot in the cq_ring and the the cq_tail + * will consume a slot in the cq_ring and the cq_tail * pointer will be increased, otherwise other timeout reqs may * return in advance without waiting for enough wait_nr. */ @@ -2582,8 +2812,7 @@ static enum hrtimer_restart io_timeout_fn(struct hrtimer *timer) spin_unlock_irqrestore(&ctx->completion_lock, flags); io_cqring_ev_posted(ctx); - if (req->flags & REQ_F_LINK) - req->flags |= REQ_F_FAIL_LINK; + req_set_fail_links(req); io_put_req(req); return HRTIMER_NORESTART; } @@ -2608,48 +2837,52 @@ static int io_timeout_cancel(struct io_ring_ctx *ctx, __u64 user_data) if (ret == -1) return -EALREADY; - if (req->flags & REQ_F_LINK) - req->flags |= REQ_F_FAIL_LINK; + req_set_fail_links(req); io_cqring_fill_event(req, -ECANCELED); io_put_req(req); return 0; } +static int io_timeout_remove_prep(struct io_kiocb *req, + const struct io_uring_sqe *sqe) +{ + if (unlikely(req->ctx->flags & IORING_SETUP_IOPOLL)) + return -EINVAL; + if (sqe->flags || sqe->ioprio || sqe->buf_index || sqe->len) + return -EINVAL; + + req->timeout.addr = READ_ONCE(sqe->addr); + req->timeout.flags = READ_ONCE(sqe->timeout_flags); + if (req->timeout.flags) + return -EINVAL; + + return 0; +} + /* * Remove or update an existing timeout command */ -static int io_timeout_remove(struct io_kiocb *req, - const struct io_uring_sqe *sqe) +static int io_timeout_remove(struct io_kiocb *req) { struct io_ring_ctx *ctx = req->ctx; - unsigned flags; int ret; - if (unlikely(ctx->flags & IORING_SETUP_IOPOLL)) - return -EINVAL; - if (sqe->flags || sqe->ioprio || sqe->buf_index || sqe->len) - return -EINVAL; - flags = READ_ONCE(sqe->timeout_flags); - if (flags) - return -EINVAL; - spin_lock_irq(&ctx->completion_lock); - ret = io_timeout_cancel(ctx, READ_ONCE(sqe->addr)); + ret = io_timeout_cancel(ctx, req->timeout.addr); io_cqring_fill_event(req, ret); io_commit_cqring(ctx); spin_unlock_irq(&ctx->completion_lock); io_cqring_ev_posted(ctx); - if (ret < 0 && req->flags & REQ_F_LINK) - req->flags |= REQ_F_FAIL_LINK; + if (ret < 0) + req_set_fail_links(req); io_put_req(req); return 0; } -static int io_timeout_prep(struct io_kiocb *req, struct io_async_ctx *io, +static int io_timeout_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe, bool is_timeout_link) { - const struct io_uring_sqe *sqe = req->sqe; struct io_timeout_data *data; unsigned flags; @@ -2663,7 +2896,12 @@ static int io_timeout_prep(struct io_kiocb *req, struct io_async_ctx *io, if (flags & ~IORING_TIMEOUT_ABS) return -EINVAL; - data = &io->timeout; + req->timeout.count = READ_ONCE(sqe->off); + + if (!req->io && io_alloc_async_ctx(req)) + return -ENOMEM; + + data = &req->io->timeout; data->req = req; req->flags |= REQ_F_TIMEOUT; @@ -2676,32 +2914,17 @@ static int io_timeout_prep(struct io_kiocb *req, struct io_async_ctx *io, data->mode = HRTIMER_MODE_REL; hrtimer_init(&data->timer, CLOCK_MONOTONIC, data->mode); - req->io = io; return 0; } -static int io_timeout(struct io_kiocb *req, const struct io_uring_sqe *sqe) +static int io_timeout(struct io_kiocb *req) { unsigned count; struct io_ring_ctx *ctx = req->ctx; struct io_timeout_data *data; - struct io_async_ctx *io; struct list_head *entry; unsigned span = 0; - io = req->io; - if (!io) { - int ret; - - io = kmalloc(sizeof(*io), GFP_KERNEL); - if (!io) - return -ENOMEM; - ret = io_timeout_prep(req, io, false); - if (ret) { - kfree(io); - return ret; - } - } data = &req->io->timeout; /* @@ -2709,7 +2932,7 @@ static int io_timeout(struct io_kiocb *req, const struct io_uring_sqe *sqe) * timeout event to be satisfied. If it isn't set, then this is * a pure timeout request, sequence isn't used. */ - count = READ_ONCE(sqe->off); + count = req->timeout.count; if (!count) { req->flags |= REQ_F_TIMEOUT_NOSEQ; spin_lock_irq(&ctx->completion_lock); @@ -2822,89 +3045,109 @@ done: spin_unlock_irqrestore(&ctx->completion_lock, flags); io_cqring_ev_posted(ctx); - if (ret < 0 && (req->flags & REQ_F_LINK)) - req->flags |= REQ_F_FAIL_LINK; + if (ret < 0) + req_set_fail_links(req); io_put_req_find_next(req, nxt); } -static int io_async_cancel(struct io_kiocb *req, const struct io_uring_sqe *sqe, - struct io_kiocb **nxt) +static int io_async_cancel_prep(struct io_kiocb *req, + const struct io_uring_sqe *sqe) { - struct io_ring_ctx *ctx = req->ctx; - - if (unlikely(ctx->flags & IORING_SETUP_IOPOLL)) + if (unlikely(req->ctx->flags & IORING_SETUP_IOPOLL)) return -EINVAL; if (sqe->flags || sqe->ioprio || sqe->off || sqe->len || sqe->cancel_flags) return -EINVAL; - io_async_find_and_cancel(ctx, req, READ_ONCE(sqe->addr), nxt, 0); + req->cancel.addr = READ_ONCE(sqe->addr); return 0; } -static int io_req_defer_prep(struct io_kiocb *req, struct io_async_ctx *io) +static int io_async_cancel(struct io_kiocb *req, struct io_kiocb **nxt) { - struct iovec inline_vecs[UIO_FASTIOV], *iovec = inline_vecs; - struct iov_iter iter; - ssize_t ret; + struct io_ring_ctx *ctx = req->ctx; - memcpy(&io->sqe, req->sqe, sizeof(io->sqe)); - req->sqe = &io->sqe; + io_async_find_and_cancel(ctx, req, req->cancel.addr, nxt, 0); + return 0; +} - switch (io->sqe.opcode) { +static int io_req_defer_prep(struct io_kiocb *req, + const struct io_uring_sqe *sqe) +{ + ssize_t ret = 0; + + switch (req->opcode) { + case IORING_OP_NOP: + break; case IORING_OP_READV: case IORING_OP_READ_FIXED: - ret = io_read_prep(req, &iovec, &iter, true); + ret = io_read_prep(req, sqe, true); break; case IORING_OP_WRITEV: case IORING_OP_WRITE_FIXED: - ret = io_write_prep(req, &iovec, &iter, true); + ret = io_write_prep(req, sqe, true); + break; + case IORING_OP_POLL_ADD: + ret = io_poll_add_prep(req, sqe); + break; + case IORING_OP_POLL_REMOVE: + ret = io_poll_remove_prep(req, sqe); + break; + case IORING_OP_FSYNC: + ret = io_prep_fsync(req, sqe); + break; + case IORING_OP_SYNC_FILE_RANGE: + ret = io_prep_sfr(req, sqe); break; case IORING_OP_SENDMSG: - ret = io_sendmsg_prep(req, io); + ret = io_sendmsg_prep(req, sqe); break; case IORING_OP_RECVMSG: - ret = io_recvmsg_prep(req, io); + ret = io_recvmsg_prep(req, sqe); break; case IORING_OP_CONNECT: - ret = io_connect_prep(req, io); + ret = io_connect_prep(req, sqe); break; case IORING_OP_TIMEOUT: - return io_timeout_prep(req, io, false); + ret = io_timeout_prep(req, sqe, false); + break; + case IORING_OP_TIMEOUT_REMOVE: + ret = io_timeout_remove_prep(req, sqe); + break; + case IORING_OP_ASYNC_CANCEL: + ret = io_async_cancel_prep(req, sqe); + break; case IORING_OP_LINK_TIMEOUT: - return io_timeout_prep(req, io, true); + ret = io_timeout_prep(req, sqe, true); + break; + case IORING_OP_ACCEPT: + ret = io_accept_prep(req, sqe); + break; default: - req->io = io; - return 0; + printk_once(KERN_WARNING "io_uring: unhandled opcode %d\n", + req->opcode); + ret = -EINVAL; + break; } - if (ret < 0) - return ret; - - req->io = io; - io_req_map_io(req, ret, iovec, inline_vecs, &iter); - return 0; + return ret; } -static int io_req_defer(struct io_kiocb *req) +static int io_req_defer(struct io_kiocb *req, const struct io_uring_sqe *sqe) { struct io_ring_ctx *ctx = req->ctx; - struct io_async_ctx *io; int ret; /* Still need defer if there is pending req in defer list. */ if (!req_need_defer(req) && list_empty(&ctx->defer_list)) return 0; - io = kmalloc(sizeof(*io), GFP_KERNEL); - if (!io) + if (!req->io && io_alloc_async_ctx(req)) return -EAGAIN; - ret = io_req_defer_prep(req, io); - if (ret < 0) { - kfree(io); + ret = io_req_defer_prep(req, sqe); + if (ret < 0) return ret; - } spin_lock_irq(&ctx->completion_lock); if (!req_need_defer(req) && list_empty(&ctx->defer_list)) { @@ -2918,66 +3161,121 @@ static int io_req_defer(struct io_kiocb *req) return -EIOCBQUEUED; } -__attribute__((nonnull)) -static int io_issue_sqe(struct io_kiocb *req, struct io_kiocb **nxt, - bool force_nonblock) +static int io_issue_sqe(struct io_kiocb *req, const struct io_uring_sqe *sqe, + struct io_kiocb **nxt, bool force_nonblock) { - int ret, opcode; struct io_ring_ctx *ctx = req->ctx; + int ret; - opcode = READ_ONCE(req->sqe->opcode); - switch (opcode) { + switch (req->opcode) { case IORING_OP_NOP: ret = io_nop(req); break; case IORING_OP_READV: - if (unlikely(req->sqe->buf_index)) - return -EINVAL; - ret = io_read(req, nxt, force_nonblock); - break; - case IORING_OP_WRITEV: - if (unlikely(req->sqe->buf_index)) - return -EINVAL; - ret = io_write(req, nxt, force_nonblock); - break; case IORING_OP_READ_FIXED: + if (sqe) { + ret = io_read_prep(req, sqe, force_nonblock); + if (ret < 0) + break; + } ret = io_read(req, nxt, force_nonblock); break; + case IORING_OP_WRITEV: case IORING_OP_WRITE_FIXED: + if (sqe) { + ret = io_write_prep(req, sqe, force_nonblock); + if (ret < 0) + break; + } ret = io_write(req, nxt, force_nonblock); break; case IORING_OP_FSYNC: - ret = io_fsync(req, req->sqe, nxt, force_nonblock); + if (sqe) { + ret = io_prep_fsync(req, sqe); + if (ret < 0) + break; + } + ret = io_fsync(req, nxt, force_nonblock); break; case IORING_OP_POLL_ADD: - ret = io_poll_add(req, req->sqe, nxt); + if (sqe) { + ret = io_poll_add_prep(req, sqe); + if (ret) + break; + } + ret = io_poll_add(req, nxt); break; case IORING_OP_POLL_REMOVE: - ret = io_poll_remove(req, req->sqe); + if (sqe) { + ret = io_poll_remove_prep(req, sqe); + if (ret < 0) + break; + } + ret = io_poll_remove(req); break; case IORING_OP_SYNC_FILE_RANGE: - ret = io_sync_file_range(req, req->sqe, nxt, force_nonblock); + if (sqe) { + ret = io_prep_sfr(req, sqe); + if (ret < 0) + break; + } + ret = io_sync_file_range(req, nxt, force_nonblock); break; case IORING_OP_SENDMSG: - ret = io_sendmsg(req, req->sqe, nxt, force_nonblock); + if (sqe) { + ret = io_sendmsg_prep(req, sqe); + if (ret < 0) + break; + } + ret = io_sendmsg(req, nxt, force_nonblock); break; case IORING_OP_RECVMSG: - ret = io_recvmsg(req, req->sqe, nxt, force_nonblock); + if (sqe) { + ret = io_recvmsg_prep(req, sqe); + if (ret) + break; + } + ret = io_recvmsg(req, nxt, force_nonblock); break; case IORING_OP_TIMEOUT: - ret = io_timeout(req, req->sqe); + if (sqe) { + ret = io_timeout_prep(req, sqe, false); + if (ret) + break; + } + ret = io_timeout(req); break; case IORING_OP_TIMEOUT_REMOVE: - ret = io_timeout_remove(req, req->sqe); + if (sqe) { + ret = io_timeout_remove_prep(req, sqe); + if (ret) + break; + } + ret = io_timeout_remove(req); break; case IORING_OP_ACCEPT: - ret = io_accept(req, req->sqe, nxt, force_nonblock); + if (sqe) { + ret = io_accept_prep(req, sqe); + if (ret) + break; + } + ret = io_accept(req, nxt, force_nonblock); break; case IORING_OP_CONNECT: - ret = io_connect(req, req->sqe, nxt, force_nonblock); + if (sqe) { + ret = io_connect_prep(req, sqe); + if (ret) + break; + } + ret = io_connect(req, nxt, force_nonblock); break; case IORING_OP_ASYNC_CANCEL: - ret = io_async_cancel(req, req->sqe, nxt); + if (sqe) { + ret = io_async_cancel_prep(req, sqe); + if (ret) + break; + } + ret = io_async_cancel(req, nxt); break; default: ret = -EINVAL; @@ -2988,29 +3286,24 @@ static int io_issue_sqe(struct io_kiocb *req, struct io_kiocb **nxt, return ret; if (ctx->flags & IORING_SETUP_IOPOLL) { + const bool in_async = io_wq_current_is_worker(); + if (req->result == -EAGAIN) return -EAGAIN; /* workqueue context doesn't hold uring_lock, grab it now */ - if (req->in_async) + if (in_async) mutex_lock(&ctx->uring_lock); + io_iopoll_req_issued(req); - if (req->in_async) + + if (in_async) mutex_unlock(&ctx->uring_lock); } return 0; } -static void io_link_work_cb(struct io_wq_work **workptr) -{ - struct io_wq_work *work = *workptr; - struct io_kiocb *link = work->data; - - io_queue_linked_timeout(link); - work->func = io_wq_submit_work; -} - static void io_wq_submit_work(struct io_wq_work **workptr) { struct io_wq_work *work = *workptr; @@ -3018,9 +3311,6 @@ static void io_wq_submit_work(struct io_wq_work **workptr) struct io_kiocb *nxt = NULL; int ret = 0; - /* Ensure we clear previously set non-block flag */ - req->rw.ki_flags &= ~IOCB_NOWAIT; - if (work->flags & IO_WQ_WORK_CANCEL) ret = -ECANCELED; @@ -3028,7 +3318,7 @@ static void io_wq_submit_work(struct io_wq_work **workptr) req->has_user = (work->flags & IO_WQ_WORK_HAS_MM) != 0; req->in_async = true; do { - ret = io_issue_sqe(req, &nxt, false); + ret = io_issue_sqe(req, NULL, &nxt, false); /* * We can get EAGAIN for polled IO even though we're * forcing a sync submission from here, since we can't @@ -3044,40 +3334,35 @@ static void io_wq_submit_work(struct io_wq_work **workptr) io_put_req(req); if (ret) { - if (req->flags & REQ_F_LINK) - req->flags |= REQ_F_FAIL_LINK; + req_set_fail_links(req); io_cqring_add_event(req, ret); io_put_req(req); } /* if a dependent link is ready, pass it back */ - if (!ret && nxt) { - struct io_kiocb *link; - - io_prep_async_work(nxt, &link); - *workptr = &nxt->work; - if (link) { - nxt->work.flags |= IO_WQ_WORK_CB; - nxt->work.func = io_link_work_cb; - nxt->work.data = link; - } - } + if (!ret && nxt) + io_wq_assign_next(workptr, nxt); } -static bool io_op_needs_file(const struct io_uring_sqe *sqe) +static bool io_req_op_valid(int op) { - int op = READ_ONCE(sqe->opcode); + return op >= IORING_OP_NOP && op < IORING_OP_LAST; +} - switch (op) { +static int io_req_needs_file(struct io_kiocb *req) +{ + switch (req->opcode) { case IORING_OP_NOP: case IORING_OP_POLL_REMOVE: case IORING_OP_TIMEOUT: case IORING_OP_TIMEOUT_REMOVE: case IORING_OP_ASYNC_CANCEL: case IORING_OP_LINK_TIMEOUT: - return false; + return 0; default: - return true; + if (io_req_op_valid(req->opcode)) + return 1; + return -EINVAL; } } @@ -3090,20 +3375,22 @@ static inline struct file *io_file_from_index(struct io_ring_ctx *ctx, return table->files[index & IORING_FILE_TABLE_MASK]; } -static int io_req_set_file(struct io_submit_state *state, struct io_kiocb *req) +static int io_req_set_file(struct io_submit_state *state, struct io_kiocb *req, + const struct io_uring_sqe *sqe) { struct io_ring_ctx *ctx = req->ctx; unsigned flags; - int fd; + int fd, ret; - flags = READ_ONCE(req->sqe->flags); - fd = READ_ONCE(req->sqe->fd); + flags = READ_ONCE(sqe->flags); + fd = READ_ONCE(sqe->fd); if (flags & IOSQE_IO_DRAIN) req->flags |= REQ_F_IO_DRAIN; - if (!io_op_needs_file(req->sqe)) - return 0; + ret = io_req_needs_file(req); + if (ret <= 0) + return ret; if (flags & IOSQE_FIXED_FILE) { if (unlikely(!ctx->file_table || @@ -3179,8 +3466,7 @@ static enum hrtimer_restart io_link_timeout_fn(struct hrtimer *timer) spin_unlock_irqrestore(&ctx->completion_lock, flags); if (prev) { - if (prev->flags & REQ_F_LINK) - prev->flags |= REQ_F_FAIL_LINK; + req_set_fail_links(prev); io_async_find_and_cancel(ctx, req, prev->user_data, NULL, -ETIME); io_put_req(prev); @@ -3222,22 +3508,23 @@ static struct io_kiocb *io_prep_linked_timeout(struct io_kiocb *req) nxt = list_first_entry_or_null(&req->link_list, struct io_kiocb, link_list); - if (!nxt || nxt->sqe->opcode != IORING_OP_LINK_TIMEOUT) + if (!nxt || nxt->opcode != IORING_OP_LINK_TIMEOUT) return NULL; req->flags |= REQ_F_LINK_TIMEOUT; return nxt; } -static void __io_queue_sqe(struct io_kiocb *req) +static void __io_queue_sqe(struct io_kiocb *req, const struct io_uring_sqe *sqe) { - struct io_kiocb *linked_timeout = io_prep_linked_timeout(req); + struct io_kiocb *linked_timeout; struct io_kiocb *nxt = NULL; int ret; - ret = io_issue_sqe(req, &nxt, true); - if (nxt) - io_queue_async_work(nxt); +again: + linked_timeout = io_prep_linked_timeout(req); + + ret = io_issue_sqe(req, sqe, &nxt, true); /* * We async punt it if the file wasn't marked NOWAIT, or if the file @@ -3256,7 +3543,7 @@ static void __io_queue_sqe(struct io_kiocb *req) * submit reference when the iocb is actually submitted. */ io_queue_async_work(req); - return; + goto done_req; } err: @@ -3273,13 +3560,18 @@ err: /* and drop final reference, if we failed */ if (ret) { io_cqring_add_event(req, ret); - if (req->flags & REQ_F_LINK) - req->flags |= REQ_F_FAIL_LINK; + req_set_fail_links(req); io_put_req(req); } +done_req: + if (nxt) { + req = nxt; + nxt = NULL; + goto again; + } } -static void io_queue_sqe(struct io_kiocb *req) +static void io_queue_sqe(struct io_kiocb *req, const struct io_uring_sqe *sqe) { int ret; @@ -3289,16 +3581,15 @@ static void io_queue_sqe(struct io_kiocb *req) } req->ctx->drain_next = (req->flags & REQ_F_DRAIN_LINK); - ret = io_req_defer(req); + ret = io_req_defer(req, sqe); if (ret) { if (ret != -EIOCBQUEUED) { io_cqring_add_event(req, ret); - if (req->flags & REQ_F_LINK) - req->flags |= REQ_F_FAIL_LINK; + req_set_fail_links(req); io_double_put_req(req); } } else - __io_queue_sqe(req); + __io_queue_sqe(req, sqe); } static inline void io_queue_link_head(struct io_kiocb *req) @@ -3307,27 +3598,25 @@ static inline void io_queue_link_head(struct io_kiocb *req) io_cqring_add_event(req, -ECANCELED); io_double_put_req(req); } else - io_queue_sqe(req); + io_queue_sqe(req, NULL); } +#define SQE_VALID_FLAGS (IOSQE_FIXED_FILE|IOSQE_IO_DRAIN|IOSQE_IO_LINK| \ + IOSQE_IO_HARDLINK) -#define SQE_VALID_FLAGS (IOSQE_FIXED_FILE|IOSQE_IO_DRAIN|IOSQE_IO_LINK) - -static bool io_submit_sqe(struct io_kiocb *req, struct io_submit_state *state, - struct io_kiocb **link) +static bool io_submit_sqe(struct io_kiocb *req, const struct io_uring_sqe *sqe, + struct io_submit_state *state, struct io_kiocb **link) { struct io_ring_ctx *ctx = req->ctx; int ret; - req->user_data = req->sqe->user_data; - /* enforce forwards compatibility on users */ - if (unlikely(req->sqe->flags & ~SQE_VALID_FLAGS)) { + if (unlikely(sqe->flags & ~SQE_VALID_FLAGS)) { ret = -EINVAL; goto err_req; } - ret = io_req_set_file(state, req); + ret = io_req_set_file(state, req, sqe); if (unlikely(ret)) { err_req: io_cqring_add_event(req, ret); @@ -3344,32 +3633,38 @@ err_req: */ if (*link) { struct io_kiocb *prev = *link; - struct io_async_ctx *io; - if (req->sqe->flags & IOSQE_IO_DRAIN) + if (sqe->flags & IOSQE_IO_DRAIN) (*link)->flags |= REQ_F_DRAIN_LINK | REQ_F_IO_DRAIN; - io = kmalloc(sizeof(*io), GFP_KERNEL); - if (!io) { + if (sqe->flags & IOSQE_IO_HARDLINK) + req->flags |= REQ_F_HARDLINK; + + if (io_alloc_async_ctx(req)) { ret = -EAGAIN; goto err_req; } - ret = io_req_defer_prep(req, io); + ret = io_req_defer_prep(req, sqe); if (ret) { - kfree(io); + /* fail even hard links since we don't submit */ prev->flags |= REQ_F_FAIL_LINK; goto err_req; } trace_io_uring_link(ctx, req, prev); list_add_tail(&req->link_list, &prev->link_list); - } else if (req->sqe->flags & IOSQE_IO_LINK) { + } else if (sqe->flags & (IOSQE_IO_LINK|IOSQE_IO_HARDLINK)) { req->flags |= REQ_F_LINK; + if (sqe->flags & IOSQE_IO_HARDLINK) + req->flags |= REQ_F_HARDLINK; INIT_LIST_HEAD(&req->link_list); + ret = io_req_defer_prep(req, sqe); + if (ret) + req->flags |= REQ_F_FAIL_LINK; *link = req; } else { - io_queue_sqe(req); + io_queue_sqe(req, sqe); } return true; @@ -3414,14 +3709,15 @@ static void io_commit_sqring(struct io_ring_ctx *ctx) } /* - * Fetch an sqe, if one is available. Note that s->sqe will point to memory + * Fetch an sqe, if one is available. Note that sqe_ptr will point to memory * that is mapped by userspace. This means that care needs to be taken to * ensure that reads are stable, as we cannot rely on userspace always * being a good citizen. If members of the sqe are validated and then later * used, it's important that those reads are done through READ_ONCE() to * prevent a re-load down the line. */ -static bool io_get_sqring(struct io_ring_ctx *ctx, struct io_kiocb *req) +static bool io_get_sqring(struct io_ring_ctx *ctx, struct io_kiocb *req, + const struct io_uring_sqe **sqe_ptr) { struct io_rings *rings = ctx->rings; u32 *sq_array = ctx->sq_array; @@ -3448,7 +3744,9 @@ static bool io_get_sqring(struct io_ring_ctx *ctx, struct io_kiocb *req) * link list. */ req->sequence = ctx->cached_sq_head; - req->sqe = &ctx->sq_sqes[head]; + *sqe_ptr = &ctx->sq_sqes[head]; + req->opcode = READ_ONCE((*sqe_ptr)->opcode); + req->user_data = READ_ONCE((*sqe_ptr)->user_data); ctx->cached_sq_head++; return true; } @@ -3480,6 +3778,7 @@ static int io_submit_sqes(struct io_ring_ctx *ctx, unsigned int nr, } for (i = 0; i < nr; i++) { + const struct io_uring_sqe *sqe; struct io_kiocb *req; unsigned int sqe_flags; @@ -3489,12 +3788,12 @@ static int io_submit_sqes(struct io_ring_ctx *ctx, unsigned int nr, submitted = -EAGAIN; break; } - if (!io_get_sqring(ctx, req)) { + if (!io_get_sqring(ctx, req, &sqe)) { __io_free_req(req); break; } - if (io_sqe_needs_user(req->sqe) && !*mm) { + if (io_req_needs_user(req) && !*mm) { mm_fault = mm_fault || !mmget_not_zero(ctx->sqo_mm); if (!mm_fault) { use_mm(ctx->sqo_mm); @@ -3503,22 +3802,21 @@ static int io_submit_sqes(struct io_ring_ctx *ctx, unsigned int nr, } submitted++; - sqe_flags = req->sqe->flags; + sqe_flags = sqe->flags; req->ring_file = ring_file; req->ring_fd = ring_fd; req->has_user = *mm != NULL; req->in_async = async; req->needs_fixed_file = async; - trace_io_uring_submit_sqe(ctx, req->sqe->user_data, - true, async); - if (!io_submit_sqe(req, statep, &link)) + trace_io_uring_submit_sqe(ctx, req->user_data, true, async); + if (!io_submit_sqe(req, sqe, statep, &link)) break; /* * If previous wasn't linked and we have a linked command, * that's the end of the chain. Submit the previous link. */ - if (!(sqe_flags & IOSQE_IO_LINK) && link) { + if (!(sqe_flags & (IOSQE_IO_LINK|IOSQE_IO_HARDLINK)) && link) { io_queue_link_head(link); link = NULL; } @@ -3647,7 +3945,9 @@ static int io_sq_thread(void *data) } to_submit = min(to_submit, ctx->sq_entries); + mutex_lock(&ctx->uring_lock); ret = io_submit_sqes(ctx, to_submit, NULL, -1, &cur_mm, true); + mutex_unlock(&ctx->uring_lock); if (ret > 0) inflight += ret; } @@ -3676,7 +3976,7 @@ static inline bool io_should_wake(struct io_wait_queue *iowq, bool noflush) struct io_ring_ctx *ctx = iowq->ctx; /* - * Wake up if we have enough events, or if a timeout occured since we + * Wake up if we have enough events, or if a timeout occurred since we * started waiting. For timeouts, we always want to return to userspace, * regardless of event count. */ @@ -4163,13 +4463,15 @@ static int io_sqe_files_update(struct io_ring_ctx *ctx, void __user *arg, return -EINVAL; if (copy_from_user(&up, arg, sizeof(up))) return -EFAULT; + if (up.resv) + return -EINVAL; if (check_add_overflow(up.offset, nr_args, &done)) return -EOVERFLOW; if (done > ctx->nr_user_files) return -EINVAL; done = 0; - fds = (__s32 __user *) up.fds; + fds = u64_to_user_ptr(up.fds); while (nr_args) { struct fixed_file_table *table; unsigned index; @@ -4428,7 +4730,7 @@ static int io_copy_iov(struct io_ring_ctx *ctx, struct iovec *dst, if (copy_from_user(&ciov, &ciovs[index], sizeof(ciov))) return -EFAULT; - dst->iov_base = (void __user *) (unsigned long) ciov.iov_base; + dst->iov_base = u64_to_user_ptr((u64)ciov.iov_base); dst->iov_len = ciov.iov_len; return 0; } @@ -4742,10 +5044,6 @@ static int io_uring_flush(struct file *file, void *data) struct io_ring_ctx *ctx = file->private_data; io_uring_cancel_files(ctx, data); - if (fatal_signal_pending(current) || (current->flags & PF_EXITING)) { - io_cqring_overflow_flush(ctx, true); - io_wq_cancel_all(ctx->io_wq); - } return 0; } @@ -4866,6 +5164,9 @@ SYSCALL_DEFINE6(io_uring_enter, unsigned int, fd, u32, to_submit, submitted = io_submit_sqes(ctx, to_submit, f.file, fd, &cur_mm, false); mutex_unlock(&ctx->uring_lock); + + if (submitted != to_submit) + goto out; } if (flags & IORING_ENTER_GETEVENTS) { unsigned nr_events = 0; @@ -4879,6 +5180,7 @@ SYSCALL_DEFINE6(io_uring_enter, unsigned int, fd, u32, to_submit, } } +out: percpu_ref_put(&ctx->refs); out_fput: fdput(f); diff --git a/fs/locks.c b/fs/locks.c index 6970f55daf54..44b6da032842 100644 --- a/fs/locks.c +++ b/fs/locks.c @@ -2853,7 +2853,7 @@ static void lock_get_status(struct seq_file *f, struct file_lock *fl, } if (inode) { /* userspace relies on this representation of dev_t */ - seq_printf(f, "%d %02x:%02x:%ld ", fl_pid, + seq_printf(f, "%d %02x:%02x:%lu ", fl_pid, MAJOR(inode->i_sb->s_dev), MINOR(inode->i_sb->s_dev), inode->i_ino); } else { diff --git a/fs/mpage.c b/fs/mpage.c index a63620cdb73a..ccba3c4c4479 100644 --- a/fs/mpage.c +++ b/fs/mpage.c @@ -62,7 +62,7 @@ static struct bio *mpage_bio_submit(int op, int op_flags, struct bio *bio) { bio->bi_end_io = mpage_end_io; bio_set_op_attrs(bio, op, op_flags); - guard_bio_eod(op, bio); + guard_bio_eod(bio); submit_bio(bio); return NULL; } diff --git a/fs/namei.c b/fs/namei.c index d6c91d1e88cb..4fb61e0754ed 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -1001,7 +1001,8 @@ static int may_linkat(struct path *link) * may_create_in_sticky - Check whether an O_CREAT open in a sticky directory * should be allowed, or not, on files that already * exist. - * @dir: the sticky parent directory + * @dir_mode: mode bits of directory + * @dir_uid: owner of directory * @inode: the inode of the file to open * * Block an O_CREAT open of a FIFO (or a regular file) when: @@ -1017,18 +1018,18 @@ static int may_linkat(struct path *link) * * Returns 0 if the open is allowed, -ve on error. */ -static int may_create_in_sticky(struct dentry * const dir, +static int may_create_in_sticky(umode_t dir_mode, kuid_t dir_uid, struct inode * const inode) { if ((!sysctl_protected_fifos && S_ISFIFO(inode->i_mode)) || (!sysctl_protected_regular && S_ISREG(inode->i_mode)) || - likely(!(dir->d_inode->i_mode & S_ISVTX)) || - uid_eq(inode->i_uid, dir->d_inode->i_uid) || + likely(!(dir_mode & S_ISVTX)) || + uid_eq(inode->i_uid, dir_uid) || uid_eq(current_fsuid(), inode->i_uid)) return 0; - if (likely(dir->d_inode->i_mode & 0002) || - (dir->d_inode->i_mode & 0020 && + if (likely(dir_mode & 0002) || + (dir_mode & 0020 && ((sysctl_protected_fifos >= 2 && S_ISFIFO(inode->i_mode)) || (sysctl_protected_regular >= 2 && S_ISREG(inode->i_mode))))) { const char *operation = S_ISFIFO(inode->i_mode) ? @@ -1232,6 +1233,7 @@ static int follow_managed(struct path *path, struct nameidata *nd) BUG_ON(!path->dentry->d_op); BUG_ON(!path->dentry->d_op->d_manage); ret = path->dentry->d_op->d_manage(path, false); + flags = smp_load_acquire(&path->dentry->d_flags); if (ret < 0) break; } @@ -1649,17 +1651,15 @@ again: if (IS_ERR(dentry)) return dentry; if (unlikely(!d_in_lookup(dentry))) { - if (!(flags & LOOKUP_NO_REVAL)) { - int error = d_revalidate(dentry, flags); - if (unlikely(error <= 0)) { - if (!error) { - d_invalidate(dentry); - dput(dentry); - goto again; - } + int error = d_revalidate(dentry, flags); + if (unlikely(error <= 0)) { + if (!error) { + d_invalidate(dentry); dput(dentry); - dentry = ERR_PTR(error); + goto again; } + dput(dentry); + dentry = ERR_PTR(error); } } else { old = inode->i_op->lookup(inode, dentry, flags); @@ -2618,72 +2618,6 @@ int user_path_at_empty(int dfd, const char __user *name, unsigned flags, EXPORT_SYMBOL(user_path_at_empty); /** - * mountpoint_last - look up last component for umount - * @nd: pathwalk nameidata - currently pointing at parent directory of "last" - * - * This is a special lookup_last function just for umount. In this case, we - * need to resolve the path without doing any revalidation. - * - * The nameidata should be the result of doing a LOOKUP_PARENT pathwalk. Since - * mountpoints are always pinned in the dcache, their ancestors are too. Thus, - * in almost all cases, this lookup will be served out of the dcache. The only - * cases where it won't are if nd->last refers to a symlink or the path is - * bogus and it doesn't exist. - * - * Returns: - * -error: if there was an error during lookup. This includes -ENOENT if the - * lookup found a negative dentry. - * - * 0: if we successfully resolved nd->last and found it to not to be a - * symlink that needs to be followed. - * - * 1: if we successfully resolved nd->last and found it to be a symlink - * that needs to be followed. - */ -static int -mountpoint_last(struct nameidata *nd) -{ - int error = 0; - struct dentry *dir = nd->path.dentry; - struct path path; - - /* If we're in rcuwalk, drop out of it to handle last component */ - if (nd->flags & LOOKUP_RCU) { - if (unlazy_walk(nd)) - return -ECHILD; - } - - nd->flags &= ~LOOKUP_PARENT; - - if (unlikely(nd->last_type != LAST_NORM)) { - error = handle_dots(nd, nd->last_type); - if (error) - return error; - path.dentry = dget(nd->path.dentry); - } else { - path.dentry = d_lookup(dir, &nd->last); - if (!path.dentry) { - /* - * No cached dentry. Mounted dentries are pinned in the - * cache, so that means that this dentry is probably - * a symlink or the path doesn't actually point - * to a mounted dentry. - */ - path.dentry = lookup_slow(&nd->last, dir, - nd->flags | LOOKUP_NO_REVAL); - if (IS_ERR(path.dentry)) - return PTR_ERR(path.dentry); - } - } - if (d_flags_negative(smp_load_acquire(&path.dentry->d_flags))) { - dput(path.dentry); - return -ENOENT; - } - path.mnt = nd->path.mnt; - return step_into(nd, &path, 0, d_backing_inode(path.dentry), 0); -} - -/** * path_mountpoint - look up a path to be umounted * @nd: lookup context * @flags: lookup flags @@ -2699,14 +2633,17 @@ path_mountpoint(struct nameidata *nd, unsigned flags, struct path *path) int err; while (!(err = link_path_walk(s, nd)) && - (err = mountpoint_last(nd)) > 0) { + (err = lookup_last(nd)) > 0) { s = trailing_symlink(nd); } + if (!err && (nd->flags & LOOKUP_RCU)) + err = unlazy_walk(nd); + if (!err) + err = handle_lookup_down(nd); if (!err) { *path = nd->path; nd->path.mnt = NULL; nd->path.dentry = NULL; - follow_mount(path); } terminate_walk(nd); return err; @@ -3265,6 +3202,8 @@ static int do_last(struct nameidata *nd, struct file *file, const struct open_flags *op) { struct dentry *dir = nd->path.dentry; + kuid_t dir_uid = dir->d_inode->i_uid; + umode_t dir_mode = dir->d_inode->i_mode; int open_flag = op->open_flag; bool will_truncate = (open_flag & O_TRUNC) != 0; bool got_write = false; @@ -3395,7 +3334,7 @@ finish_open: error = -EISDIR; if (d_is_dir(nd->path.dentry)) goto out; - error = may_create_in_sticky(dir, + error = may_create_in_sticky(dir_mode, dir_uid, d_backing_inode(nd->path.dentry)); if (unlikely(error)) goto out; diff --git a/fs/namespace.c b/fs/namespace.c index 2fd0c8bcb8c1..5e1bf611a9eb 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -1728,7 +1728,7 @@ static bool is_mnt_ns_file(struct dentry *dentry) dentry->d_fsdata == &mntns_operations; } -struct mnt_namespace *to_mnt_ns(struct ns_common *ns) +static struct mnt_namespace *to_mnt_ns(struct ns_common *ns) { return container_of(ns, struct mnt_namespace, ns); } @@ -3325,8 +3325,8 @@ struct dentry *mount_subtree(struct vfsmount *m, const char *name) } EXPORT_SYMBOL(mount_subtree); -int ksys_mount(const char __user *dev_name, const char __user *dir_name, - const char __user *type, unsigned long flags, void __user *data) +SYSCALL_DEFINE5(mount, char __user *, dev_name, char __user *, dir_name, + char __user *, type, unsigned long, flags, void __user *, data) { int ret; char *kernel_type; @@ -3359,12 +3359,6 @@ out_type: return ret; } -SYSCALL_DEFINE5(mount, char __user *, dev_name, char __user *, dir_name, - char __user *, type, unsigned long, flags, void __user *, data) -{ - return ksys_mount(dev_name, dir_name, type, flags, data); -} - /* * Create a kernel mount representation for a new, prepared superblock * (specified by fs_fd) and attach to an open_tree-like file descriptor. diff --git a/fs/nfs/nfstrace.h b/fs/nfs/nfstrace.h index f64a33d2a1d1..2a82dcce5fc1 100644 --- a/fs/nfs/nfstrace.h +++ b/fs/nfs/nfstrace.h @@ -206,7 +206,6 @@ TRACE_DEFINE_ENUM(LOOKUP_AUTOMOUNT); TRACE_DEFINE_ENUM(LOOKUP_PARENT); TRACE_DEFINE_ENUM(LOOKUP_REVAL); TRACE_DEFINE_ENUM(LOOKUP_RCU); -TRACE_DEFINE_ENUM(LOOKUP_NO_REVAL); TRACE_DEFINE_ENUM(LOOKUP_OPEN); TRACE_DEFINE_ENUM(LOOKUP_CREATE); TRACE_DEFINE_ENUM(LOOKUP_EXCL); @@ -224,7 +223,6 @@ TRACE_DEFINE_ENUM(LOOKUP_DOWN); { LOOKUP_PARENT, "PARENT" }, \ { LOOKUP_REVAL, "REVAL" }, \ { LOOKUP_RCU, "RCU" }, \ - { LOOKUP_NO_REVAL, "NO_REVAL" }, \ { LOOKUP_OPEN, "OPEN" }, \ { LOOKUP_CREATE, "CREATE" }, \ { LOOKUP_EXCL, "EXCL" }, \ diff --git a/fs/notify/fsnotify.c b/fs/notify/fsnotify.c index 3e77b728a22b..46f225580009 100644 --- a/fs/notify/fsnotify.c +++ b/fs/notify/fsnotify.c @@ -57,6 +57,9 @@ static void fsnotify_unmount_inodes(struct super_block *sb) * doing an __iget/iput with SB_ACTIVE clear would actually * evict all inodes with zero i_count from icache which is * unnecessarily violent and may in fact be illegal to do. + * However, we should have been called /after/ evict_inodes + * removed all zero refcount inodes, in any case. Test to + * be sure. */ if (!atomic_read(&inode->i_count)) { spin_unlock(&inode->i_lock); @@ -77,6 +80,7 @@ static void fsnotify_unmount_inodes(struct super_block *sb) iput_inode = inode; + cond_resched(); spin_lock(&sb->s_inode_list_lock); } spin_unlock(&sb->s_inode_list_lock); diff --git a/fs/nsfs.c b/fs/nsfs.c index a0431642c6b5..f75767bd623a 100644 --- a/fs/nsfs.c +++ b/fs/nsfs.c @@ -3,6 +3,7 @@ #include <linux/pseudo_fs.h> #include <linux/file.h> #include <linux/fs.h> +#include <linux/proc_fs.h> #include <linux/proc_ns.h> #include <linux/magic.h> #include <linux/ktime.h> @@ -11,6 +12,8 @@ #include <linux/nsfs.h> #include <linux/uaccess.h> +#include "internal.h" + static struct vfsmount *nsfs_mnt; static long ns_ioctl(struct file *filp, unsigned int ioctl, diff --git a/fs/ocfs2/dlmglue.c b/fs/ocfs2/dlmglue.c index 1c4c51f3df60..cda1027d0819 100644 --- a/fs/ocfs2/dlmglue.c +++ b/fs/ocfs2/dlmglue.c @@ -3282,6 +3282,7 @@ static void ocfs2_dlm_init_debug(struct ocfs2_super *osb) debugfs_create_u32("locking_filter", 0600, osb->osb_debug_root, &dlm_debug->d_filter_secs); + ocfs2_get_dlm_debug(dlm_debug); } static void ocfs2_dlm_shutdown_debug(struct ocfs2_super *osb) diff --git a/fs/ocfs2/journal.c b/fs/ocfs2/journal.c index 1afe57f425a0..68ba354cf361 100644 --- a/fs/ocfs2/journal.c +++ b/fs/ocfs2/journal.c @@ -1066,6 +1066,14 @@ int ocfs2_journal_load(struct ocfs2_journal *journal, int local, int replayed) ocfs2_clear_journal_error(osb->sb, journal->j_journal, osb->slot_num); + if (replayed) { + jbd2_journal_lock_updates(journal->j_journal); + status = jbd2_journal_flush(journal->j_journal); + jbd2_journal_unlock_updates(journal->j_journal); + if (status < 0) + mlog_errno(status); + } + status = ocfs2_journal_toggle_dirty(osb, 1, replayed); if (status < 0) { mlog_errno(status); diff --git a/fs/overlayfs/copy_up.c b/fs/overlayfs/copy_up.c index b801c6353100..6220642fe113 100644 --- a/fs/overlayfs/copy_up.c +++ b/fs/overlayfs/copy_up.c @@ -227,13 +227,17 @@ int ovl_set_attr(struct dentry *upperdentry, struct kstat *stat) struct ovl_fh *ovl_encode_real_fh(struct dentry *real, bool is_upper) { struct ovl_fh *fh; - int fh_type, fh_len, dwords; - void *buf; + int fh_type, dwords; int buflen = MAX_HANDLE_SZ; uuid_t *uuid = &real->d_sb->s_uuid; + int err; - buf = kmalloc(buflen, GFP_KERNEL); - if (!buf) + /* Make sure the real fid stays 32bit aligned */ + BUILD_BUG_ON(OVL_FH_FID_OFFSET % 4); + BUILD_BUG_ON(MAX_HANDLE_SZ + OVL_FH_FID_OFFSET > 255); + + fh = kzalloc(buflen + OVL_FH_FID_OFFSET, GFP_KERNEL); + if (!fh) return ERR_PTR(-ENOMEM); /* @@ -242,27 +246,19 @@ struct ovl_fh *ovl_encode_real_fh(struct dentry *real, bool is_upper) * the price or reconnecting the dentry. */ dwords = buflen >> 2; - fh_type = exportfs_encode_fh(real, buf, &dwords, 0); + fh_type = exportfs_encode_fh(real, (void *)fh->fb.fid, &dwords, 0); buflen = (dwords << 2); - fh = ERR_PTR(-EIO); + err = -EIO; if (WARN_ON(fh_type < 0) || WARN_ON(buflen > MAX_HANDLE_SZ) || WARN_ON(fh_type == FILEID_INVALID)) - goto out; + goto out_err; - BUILD_BUG_ON(MAX_HANDLE_SZ + offsetof(struct ovl_fh, fid) > 255); - fh_len = offsetof(struct ovl_fh, fid) + buflen; - fh = kmalloc(fh_len, GFP_KERNEL); - if (!fh) { - fh = ERR_PTR(-ENOMEM); - goto out; - } - - fh->version = OVL_FH_VERSION; - fh->magic = OVL_FH_MAGIC; - fh->type = fh_type; - fh->flags = OVL_FH_FLAG_CPU_ENDIAN; + fh->fb.version = OVL_FH_VERSION; + fh->fb.magic = OVL_FH_MAGIC; + fh->fb.type = fh_type; + fh->fb.flags = OVL_FH_FLAG_CPU_ENDIAN; /* * When we will want to decode an overlay dentry from this handle * and all layers are on the same fs, if we get a disconncted real @@ -270,14 +266,15 @@ struct ovl_fh *ovl_encode_real_fh(struct dentry *real, bool is_upper) * it to upperdentry or to lowerstack is by checking this flag. */ if (is_upper) - fh->flags |= OVL_FH_FLAG_PATH_UPPER; - fh->len = fh_len; - fh->uuid = *uuid; - memcpy(fh->fid, buf, buflen); + fh->fb.flags |= OVL_FH_FLAG_PATH_UPPER; + fh->fb.len = sizeof(fh->fb) + buflen; + fh->fb.uuid = *uuid; -out: - kfree(buf); return fh; + +out_err: + kfree(fh); + return ERR_PTR(err); } int ovl_set_origin(struct dentry *dentry, struct dentry *lower, @@ -300,8 +297,8 @@ int ovl_set_origin(struct dentry *dentry, struct dentry *lower, /* * Do not fail when upper doesn't support xattrs. */ - err = ovl_check_setxattr(dentry, upper, OVL_XATTR_ORIGIN, fh, - fh ? fh->len : 0, 0); + err = ovl_check_setxattr(dentry, upper, OVL_XATTR_ORIGIN, fh->buf, + fh ? fh->fb.len : 0, 0); kfree(fh); return err; @@ -317,7 +314,7 @@ static int ovl_set_upper_fh(struct dentry *upper, struct dentry *index) if (IS_ERR(fh)) return PTR_ERR(fh); - err = ovl_do_setxattr(index, OVL_XATTR_UPPER, fh, fh->len, 0); + err = ovl_do_setxattr(index, OVL_XATTR_UPPER, fh->buf, fh->fb.len, 0); kfree(fh); return err; diff --git a/fs/overlayfs/dir.c b/fs/overlayfs/dir.c index 702aa63f6774..29abdb1d3b5c 100644 --- a/fs/overlayfs/dir.c +++ b/fs/overlayfs/dir.c @@ -1170,7 +1170,7 @@ static int ovl_rename(struct inode *olddir, struct dentry *old, if (newdentry == trap) goto out_dput; - if (WARN_ON(olddentry->d_inode == newdentry->d_inode)) + if (olddentry->d_inode == newdentry->d_inode) goto out_dput; err = 0; diff --git a/fs/overlayfs/export.c b/fs/overlayfs/export.c index 73c9775215b3..70e55588aedc 100644 --- a/fs/overlayfs/export.c +++ b/fs/overlayfs/export.c @@ -211,10 +211,11 @@ static int ovl_check_encode_origin(struct dentry *dentry) return 1; } -static int ovl_d_to_fh(struct dentry *dentry, char *buf, int buflen) +static int ovl_dentry_to_fid(struct dentry *dentry, u32 *fid, int buflen) { struct ovl_fh *fh = NULL; int err, enc_lower; + int len; /* * Check if we should encode a lower or upper file handle and maybe @@ -231,11 +232,12 @@ static int ovl_d_to_fh(struct dentry *dentry, char *buf, int buflen) return PTR_ERR(fh); err = -EOVERFLOW; - if (fh->len > buflen) + len = OVL_FH_LEN(fh); + if (len > buflen) goto fail; - memcpy(buf, (char *)fh, fh->len); - err = fh->len; + memcpy(fid, fh, len); + err = len; out: kfree(fh); @@ -243,31 +245,16 @@ out: fail: pr_warn_ratelimited("overlayfs: failed to encode file handle (%pd2, err=%i, buflen=%d, len=%d, type=%d)\n", - dentry, err, buflen, fh ? (int)fh->len : 0, - fh ? fh->type : 0); + dentry, err, buflen, fh ? (int)fh->fb.len : 0, + fh ? fh->fb.type : 0); goto out; } -static int ovl_dentry_to_fh(struct dentry *dentry, u32 *fid, int *max_len) -{ - int res, len = *max_len << 2; - - res = ovl_d_to_fh(dentry, (char *)fid, len); - if (res <= 0) - return FILEID_INVALID; - - len = res; - - /* Round up to dwords */ - *max_len = (len + 3) >> 2; - return OVL_FILEID; -} - static int ovl_encode_fh(struct inode *inode, u32 *fid, int *max_len, struct inode *parent) { struct dentry *dentry; - int type; + int bytes = *max_len << 2; /* TODO: encode connectable file handles */ if (parent) @@ -277,10 +264,14 @@ static int ovl_encode_fh(struct inode *inode, u32 *fid, int *max_len, if (WARN_ON(!dentry)) return FILEID_INVALID; - type = ovl_dentry_to_fh(dentry, fid, max_len); - + bytes = ovl_dentry_to_fid(dentry, fid, bytes); dput(dentry); - return type; + if (bytes <= 0) + return FILEID_INVALID; + + *max_len = bytes >> 2; + + return OVL_FILEID_V1; } /* @@ -777,24 +768,45 @@ out_err: goto out; } +static struct ovl_fh *ovl_fid_to_fh(struct fid *fid, int buflen, int fh_type) +{ + struct ovl_fh *fh; + + /* If on-wire inner fid is aligned - nothing to do */ + if (fh_type == OVL_FILEID_V1) + return (struct ovl_fh *)fid; + + if (fh_type != OVL_FILEID_V0) + return ERR_PTR(-EINVAL); + + fh = kzalloc(buflen, GFP_KERNEL); + if (!fh) + return ERR_PTR(-ENOMEM); + + /* Copy unaligned inner fh into aligned buffer */ + memcpy(&fh->fb, fid, buflen - OVL_FH_WIRE_OFFSET); + return fh; +} + static struct dentry *ovl_fh_to_dentry(struct super_block *sb, struct fid *fid, int fh_len, int fh_type) { struct dentry *dentry = NULL; - struct ovl_fh *fh = (struct ovl_fh *) fid; + struct ovl_fh *fh = NULL; int len = fh_len << 2; unsigned int flags = 0; int err; - err = -EINVAL; - if (fh_type != OVL_FILEID) + fh = ovl_fid_to_fh(fid, len, fh_type); + err = PTR_ERR(fh); + if (IS_ERR(fh)) goto out_err; err = ovl_check_fh_len(fh, len); if (err) goto out_err; - flags = fh->flags; + flags = fh->fb.flags; dentry = (flags & OVL_FH_FLAG_PATH_UPPER) ? ovl_upper_fh_to_d(sb, fh) : ovl_lower_fh_to_d(sb, fh); @@ -802,12 +814,18 @@ static struct dentry *ovl_fh_to_dentry(struct super_block *sb, struct fid *fid, if (IS_ERR(dentry) && err != -ESTALE) goto out_err; +out: + /* We may have needed to re-align OVL_FILEID_V0 */ + if (!IS_ERR_OR_NULL(fh) && fh != (void *)fid) + kfree(fh); + return dentry; out_err: pr_warn_ratelimited("overlayfs: failed to decode file handle (len=%d, type=%d, flags=%x, err=%i)\n", - len, fh_type, flags, err); - return ERR_PTR(err); + fh_len, fh_type, flags, err); + dentry = ERR_PTR(err); + goto out; } static struct dentry *ovl_fh_to_parent(struct super_block *sb, struct fid *fid, diff --git a/fs/overlayfs/inode.c b/fs/overlayfs/inode.c index bc14781886bf..b045cf1826fc 100644 --- a/fs/overlayfs/inode.c +++ b/fs/overlayfs/inode.c @@ -200,8 +200,14 @@ int ovl_getattr(const struct path *path, struct kstat *stat, if (ovl_test_flag(OVL_INDEX, d_inode(dentry)) || (!ovl_verify_lower(dentry->d_sb) && (is_dir || lowerstat.nlink == 1))) { - stat->ino = lowerstat.ino; lower_layer = ovl_layer_lower(dentry); + /* + * Cannot use origin st_dev;st_ino because + * origin inode content may differ from overlay + * inode content. + */ + if (samefs || lower_layer->fsid) + stat->ino = lowerstat.ino; } /* diff --git a/fs/overlayfs/namei.c b/fs/overlayfs/namei.c index c269d6033525..76ff66339173 100644 --- a/fs/overlayfs/namei.c +++ b/fs/overlayfs/namei.c @@ -84,21 +84,21 @@ static int ovl_acceptable(void *ctx, struct dentry *dentry) * Return -ENODATA for "origin unknown". * Return <0 for an invalid file handle. */ -int ovl_check_fh_len(struct ovl_fh *fh, int fh_len) +int ovl_check_fb_len(struct ovl_fb *fb, int fb_len) { - if (fh_len < sizeof(struct ovl_fh) || fh_len < fh->len) + if (fb_len < sizeof(struct ovl_fb) || fb_len < fb->len) return -EINVAL; - if (fh->magic != OVL_FH_MAGIC) + if (fb->magic != OVL_FH_MAGIC) return -EINVAL; /* Treat larger version and unknown flags as "origin unknown" */ - if (fh->version > OVL_FH_VERSION || fh->flags & ~OVL_FH_FLAG_ALL) + if (fb->version > OVL_FH_VERSION || fb->flags & ~OVL_FH_FLAG_ALL) return -ENODATA; /* Treat endianness mismatch as "origin unknown" */ - if (!(fh->flags & OVL_FH_FLAG_ANY_ENDIAN) && - (fh->flags & OVL_FH_FLAG_BIG_ENDIAN) != OVL_FH_FLAG_CPU_ENDIAN) + if (!(fb->flags & OVL_FH_FLAG_ANY_ENDIAN) && + (fb->flags & OVL_FH_FLAG_BIG_ENDIAN) != OVL_FH_FLAG_CPU_ENDIAN) return -ENODATA; return 0; @@ -119,15 +119,15 @@ static struct ovl_fh *ovl_get_fh(struct dentry *dentry, const char *name) if (res == 0) return NULL; - fh = kzalloc(res, GFP_KERNEL); + fh = kzalloc(res + OVL_FH_WIRE_OFFSET, GFP_KERNEL); if (!fh) return ERR_PTR(-ENOMEM); - res = vfs_getxattr(dentry, name, fh, res); + res = vfs_getxattr(dentry, name, fh->buf, res); if (res < 0) goto fail; - err = ovl_check_fh_len(fh, res); + err = ovl_check_fb_len(&fh->fb, res); if (err < 0) { if (err == -ENODATA) goto out; @@ -158,12 +158,12 @@ struct dentry *ovl_decode_real_fh(struct ovl_fh *fh, struct vfsmount *mnt, * Make sure that the stored uuid matches the uuid of the lower * layer where file handle will be decoded. */ - if (!uuid_equal(&fh->uuid, &mnt->mnt_sb->s_uuid)) + if (!uuid_equal(&fh->fb.uuid, &mnt->mnt_sb->s_uuid)) return NULL; - bytes = (fh->len - offsetof(struct ovl_fh, fid)); - real = exportfs_decode_fh(mnt, (struct fid *)fh->fid, - bytes >> 2, (int)fh->type, + bytes = (fh->fb.len - offsetof(struct ovl_fb, fid)); + real = exportfs_decode_fh(mnt, (struct fid *)fh->fb.fid, + bytes >> 2, (int)fh->fb.type, connected ? ovl_acceptable : NULL, mnt); if (IS_ERR(real)) { /* @@ -173,7 +173,7 @@ struct dentry *ovl_decode_real_fh(struct ovl_fh *fh, struct vfsmount *mnt, * index entries correctly. */ if (real == ERR_PTR(-ESTALE) && - !(fh->flags & OVL_FH_FLAG_PATH_UPPER)) + !(fh->fb.flags & OVL_FH_FLAG_PATH_UPPER)) real = NULL; return real; } @@ -323,6 +323,14 @@ int ovl_check_origin_fh(struct ovl_fs *ofs, struct ovl_fh *fh, bool connected, int i; for (i = 0; i < ofs->numlower; i++) { + /* + * If lower fs uuid is not unique among lower fs we cannot match + * fh->uuid to layer. + */ + if (ofs->lower_layers[i].fsid && + ofs->lower_layers[i].fs->bad_uuid) + continue; + origin = ovl_decode_real_fh(fh, ofs->lower_layers[i].mnt, connected); if (origin) @@ -400,7 +408,7 @@ static int ovl_verify_fh(struct dentry *dentry, const char *name, if (IS_ERR(ofh)) return PTR_ERR(ofh); - if (fh->len != ofh->len || memcmp(fh, ofh, fh->len)) + if (fh->fb.len != ofh->fb.len || memcmp(&fh->fb, &ofh->fb, fh->fb.len)) err = -ESTALE; kfree(ofh); @@ -431,7 +439,7 @@ int ovl_verify_set_fh(struct dentry *dentry, const char *name, err = ovl_verify_fh(dentry, name, fh); if (set && err == -ENODATA) - err = ovl_do_setxattr(dentry, name, fh, fh->len, 0); + err = ovl_do_setxattr(dentry, name, fh->buf, fh->fb.len, 0); if (err) goto fail; @@ -505,20 +513,20 @@ int ovl_verify_index(struct ovl_fs *ofs, struct dentry *index) goto fail; err = -EINVAL; - if (index->d_name.len < sizeof(struct ovl_fh)*2) + if (index->d_name.len < sizeof(struct ovl_fb)*2) goto fail; err = -ENOMEM; len = index->d_name.len / 2; - fh = kzalloc(len, GFP_KERNEL); + fh = kzalloc(len + OVL_FH_WIRE_OFFSET, GFP_KERNEL); if (!fh) goto fail; err = -EINVAL; - if (hex2bin((u8 *)fh, index->d_name.name, len)) + if (hex2bin(fh->buf, index->d_name.name, len)) goto fail; - err = ovl_check_fh_len(fh, len); + err = ovl_check_fb_len(&fh->fb, len); if (err) goto fail; @@ -597,11 +605,11 @@ static int ovl_get_index_name_fh(struct ovl_fh *fh, struct qstr *name) { char *n, *s; - n = kcalloc(fh->len, 2, GFP_KERNEL); + n = kcalloc(fh->fb.len, 2, GFP_KERNEL); if (!n) return -ENOMEM; - s = bin2hex(n, fh, fh->len); + s = bin2hex(n, fh->buf, fh->fb.len); *name = (struct qstr) QSTR_INIT(n, s - n); return 0; diff --git a/fs/overlayfs/overlayfs.h b/fs/overlayfs/overlayfs.h index 6934bcf030f0..f283b1d69a9e 100644 --- a/fs/overlayfs/overlayfs.h +++ b/fs/overlayfs/overlayfs.h @@ -71,20 +71,36 @@ enum ovl_entry_flag { #error Endianness not defined #endif -/* The type returned by overlay exportfs ops when encoding an ovl_fh handle */ -#define OVL_FILEID 0xfb +/* The type used to be returned by overlay exportfs for misaligned fid */ +#define OVL_FILEID_V0 0xfb +/* The type returned by overlay exportfs for 32bit aligned fid */ +#define OVL_FILEID_V1 0xf8 -/* On-disk and in-memeory format for redirect by file handle */ -struct ovl_fh { +/* On-disk format for "origin" file handle */ +struct ovl_fb { u8 version; /* 0 */ u8 magic; /* 0xfb */ u8 len; /* size of this header + size of fid */ u8 flags; /* OVL_FH_FLAG_* */ u8 type; /* fid_type of fid */ uuid_t uuid; /* uuid of filesystem */ - u8 fid[0]; /* file identifier */ + u32 fid[0]; /* file identifier should be 32bit aligned in-memory */ } __packed; +/* In-memory and on-wire format for overlay file handle */ +struct ovl_fh { + u8 padding[3]; /* make sure fb.fid is 32bit aligned */ + union { + struct ovl_fb fb; + u8 buf[0]; + }; +} __packed; + +#define OVL_FH_WIRE_OFFSET offsetof(struct ovl_fh, fb) +#define OVL_FH_LEN(fh) (OVL_FH_WIRE_OFFSET + (fh)->fb.len) +#define OVL_FH_FID_OFFSET (OVL_FH_WIRE_OFFSET + \ + offsetof(struct ovl_fb, fid)) + static inline int ovl_do_rmdir(struct inode *dir, struct dentry *dentry) { int err = vfs_rmdir(dir, dentry); @@ -302,7 +318,13 @@ static inline void ovl_inode_unlock(struct inode *inode) /* namei.c */ -int ovl_check_fh_len(struct ovl_fh *fh, int fh_len); +int ovl_check_fb_len(struct ovl_fb *fb, int fb_len); + +static inline int ovl_check_fh_len(struct ovl_fh *fh, int fh_len) +{ + return ovl_check_fb_len(&fh->fb, fh_len - OVL_FH_WIRE_OFFSET); +} + struct dentry *ovl_decode_real_fh(struct ovl_fh *fh, struct vfsmount *mnt, bool connected); int ovl_check_origin_fh(struct ovl_fs *ofs, struct ovl_fh *fh, bool connected, diff --git a/fs/overlayfs/ovl_entry.h b/fs/overlayfs/ovl_entry.h index a8279280e88d..28348c44ea5b 100644 --- a/fs/overlayfs/ovl_entry.h +++ b/fs/overlayfs/ovl_entry.h @@ -22,6 +22,8 @@ struct ovl_config { struct ovl_sb { struct super_block *sb; dev_t pseudo_dev; + /* Unusable (conflicting) uuid */ + bool bad_uuid; }; struct ovl_layer { diff --git a/fs/overlayfs/super.c b/fs/overlayfs/super.c index afbcb116a7f1..7621ff176d15 100644 --- a/fs/overlayfs/super.c +++ b/fs/overlayfs/super.c @@ -1255,7 +1255,7 @@ static bool ovl_lower_uuid_ok(struct ovl_fs *ofs, const uuid_t *uuid) { unsigned int i; - if (!ofs->config.nfs_export && !(ofs->config.index && ofs->upper_mnt)) + if (!ofs->config.nfs_export && !ofs->upper_mnt) return true; for (i = 0; i < ofs->numlowerfs; i++) { @@ -1263,9 +1263,13 @@ static bool ovl_lower_uuid_ok(struct ovl_fs *ofs, const uuid_t *uuid) * We use uuid to associate an overlay lower file handle with a * lower layer, so we can accept lower fs with null uuid as long * as all lower layers with null uuid are on the same fs. + * if we detect multiple lower fs with the same uuid, we + * disable lower file handle decoding on all of them. */ - if (uuid_equal(&ofs->lower_fs[i].sb->s_uuid, uuid)) + if (uuid_equal(&ofs->lower_fs[i].sb->s_uuid, uuid)) { + ofs->lower_fs[i].bad_uuid = true; return false; + } } return true; } @@ -1277,6 +1281,7 @@ static int ovl_get_fsid(struct ovl_fs *ofs, const struct path *path) unsigned int i; dev_t dev; int err; + bool bad_uuid = false; /* fsid 0 is reserved for upper fs even with non upper overlay */ if (ofs->upper_mnt && ofs->upper_mnt->mnt_sb == sb) @@ -1288,11 +1293,15 @@ static int ovl_get_fsid(struct ovl_fs *ofs, const struct path *path) } if (!ovl_lower_uuid_ok(ofs, &sb->s_uuid)) { - ofs->config.index = false; - ofs->config.nfs_export = false; - pr_warn("overlayfs: %s uuid detected in lower fs '%pd2', falling back to index=off,nfs_export=off.\n", - uuid_is_null(&sb->s_uuid) ? "null" : "conflicting", - path->dentry); + bad_uuid = true; + if (ofs->config.index || ofs->config.nfs_export) { + ofs->config.index = false; + ofs->config.nfs_export = false; + pr_warn("overlayfs: %s uuid detected in lower fs '%pd2', falling back to index=off,nfs_export=off.\n", + uuid_is_null(&sb->s_uuid) ? "null" : + "conflicting", + path->dentry); + } } err = get_anon_bdev(&dev); @@ -1303,6 +1312,7 @@ static int ovl_get_fsid(struct ovl_fs *ofs, const struct path *path) ofs->lower_fs[ofs->numlowerfs].sb = sb; ofs->lower_fs[ofs->numlowerfs].pseudo_dev = dev; + ofs->lower_fs[ofs->numlowerfs].bad_uuid = bad_uuid; ofs->numlowerfs++; return ofs->numlowerfs; diff --git a/fs/pipe.c b/fs/pipe.c index 87109e761fa5..57502c3c0fba 100644 --- a/fs/pipe.c +++ b/fs/pipe.c @@ -364,17 +364,39 @@ pipe_read(struct kiocb *iocb, struct iov_iter *to) ret = -EAGAIN; break; } - if (signal_pending(current)) { - if (!ret) - ret = -ERESTARTSYS; - break; - } __pipe_unlock(pipe); - if (was_full) { + + /* + * We only get here if we didn't actually read anything. + * + * However, we could have seen (and removed) a zero-sized + * pipe buffer, and might have made space in the buffers + * that way. + * + * You can't make zero-sized pipe buffers by doing an empty + * write (not even in packet mode), but they can happen if + * the writer gets an EFAULT when trying to fill a buffer + * that already got allocated and inserted in the buffer + * array. + * + * So we still need to wake up any pending writers in the + * _very_ unlikely case that the pipe was full, but we got + * no data. + */ + if (unlikely(was_full)) { wake_up_interruptible_sync_poll(&pipe->wait, EPOLLOUT | EPOLLWRNORM); kill_fasync(&pipe->fasync_writers, SIGIO, POLL_OUT); } - wait_event_interruptible(pipe->wait, pipe_readable(pipe)); + + /* + * But because we didn't read anything, at this point we can + * just return directly with -ERESTARTSYS if we're interrupted, + * since we've done any required wakeups and there's no need + * to mark anything accessed. And we've dropped the lock. + */ + if (wait_event_interruptible(pipe->wait, pipe_readable(pipe)) < 0) + return -ERESTARTSYS; + __pipe_lock(pipe); was_full = pipe_full(pipe->head, pipe->tail, pipe->max_usage); } @@ -559,7 +581,7 @@ pipe_write(struct kiocb *iocb, struct iov_iter *from) } wait_event_interruptible(pipe->wait, pipe_writable(pipe)); __pipe_lock(pipe); - was_empty = pipe_empty(head, pipe->tail); + was_empty = pipe_empty(pipe->head, pipe->tail); } out: __pipe_unlock(pipe); diff --git a/fs/posix_acl.c b/fs/posix_acl.c index 84ad1c90d535..249672bf54fe 100644 --- a/fs/posix_acl.c +++ b/fs/posix_acl.c @@ -631,12 +631,15 @@ EXPORT_SYMBOL_GPL(posix_acl_create); /** * posix_acl_update_mode - update mode in set_acl + * @inode: target inode + * @mode_p: mode (pointer) for update + * @acl: acl pointer * * Update the file mode when setting an ACL: compute the new file permission * bits based on the ACL. In addition, if the ACL is equivalent to the new - * file mode, set *acl to NULL to indicate that no ACL should be set. + * file mode, set *@acl to NULL to indicate that no ACL should be set. * - * As with chmod, clear the setgit bit if the caller is not in the owning group + * As with chmod, clear the setgid bit if the caller is not in the owning group * or capable of CAP_FSETID (see inode_change_ok). * * Called from set_acl inode operations. diff --git a/fs/proc/Kconfig b/fs/proc/Kconfig index 733881a6387b..27ef84d99f59 100644 --- a/fs/proc/Kconfig +++ b/fs/proc/Kconfig @@ -103,3 +103,7 @@ config PROC_CHILDREN config PROC_PID_ARCH_STATUS def_bool n depends on PROC_FS + +config PROC_CPU_RESCTRL + def_bool n + depends on PROC_FS diff --git a/fs/proc/base.c b/fs/proc/base.c index ebea9501afb8..915686772f0e 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -94,6 +94,8 @@ #include <linux/sched/debug.h> #include <linux/sched/stat.h> #include <linux/posix-timers.h> +#include <linux/time_namespace.h> +#include <linux/resctrl.h> #include <trace/events/oom.h> #include "internal.h" #include "fd.h" @@ -1533,6 +1535,96 @@ static const struct file_operations proc_pid_sched_autogroup_operations = { #endif /* CONFIG_SCHED_AUTOGROUP */ +#ifdef CONFIG_TIME_NS +static int timens_offsets_show(struct seq_file *m, void *v) +{ + struct task_struct *p; + + p = get_proc_task(file_inode(m->file)); + if (!p) + return -ESRCH; + proc_timens_show_offsets(p, m); + + put_task_struct(p); + + return 0; +} + +static ssize_t timens_offsets_write(struct file *file, const char __user *buf, + size_t count, loff_t *ppos) +{ + struct inode *inode = file_inode(file); + struct proc_timens_offset offsets[2]; + char *kbuf = NULL, *pos, *next_line; + struct task_struct *p; + int ret, noffsets; + + /* Only allow < page size writes at the beginning of the file */ + if ((*ppos != 0) || (count >= PAGE_SIZE)) + return -EINVAL; + + /* Slurp in the user data */ + kbuf = memdup_user_nul(buf, count); + if (IS_ERR(kbuf)) + return PTR_ERR(kbuf); + + /* Parse the user data */ + ret = -EINVAL; + noffsets = 0; + for (pos = kbuf; pos; pos = next_line) { + struct proc_timens_offset *off = &offsets[noffsets]; + int err; + + /* Find the end of line and ensure we don't look past it */ + next_line = strchr(pos, '\n'); + if (next_line) { + *next_line = '\0'; + next_line++; + if (*next_line == '\0') + next_line = NULL; + } + + err = sscanf(pos, "%u %lld %lu", &off->clockid, + &off->val.tv_sec, &off->val.tv_nsec); + if (err != 3 || off->val.tv_nsec >= NSEC_PER_SEC) + goto out; + noffsets++; + if (noffsets == ARRAY_SIZE(offsets)) { + if (next_line) + count = next_line - kbuf; + break; + } + } + + ret = -ESRCH; + p = get_proc_task(inode); + if (!p) + goto out; + ret = proc_timens_set_offset(file, p, offsets, noffsets); + put_task_struct(p); + if (ret) + goto out; + + ret = count; +out: + kfree(kbuf); + return ret; +} + +static int timens_offsets_open(struct inode *inode, struct file *filp) +{ + return single_open(filp, timens_offsets_show, inode); +} + +static const struct file_operations proc_timens_offsets_operations = { + .open = timens_offsets_open, + .read = seq_read, + .write = timens_offsets_write, + .llseek = seq_lseek, + .release = single_release, +}; +#endif /* CONFIG_TIME_NS */ + static ssize_t comm_write(struct file *file, const char __user *buf, size_t count, loff_t *offset) { @@ -3016,6 +3108,9 @@ static const struct pid_entry tgid_base_stuff[] = { #ifdef CONFIG_SCHED_AUTOGROUP REG("autogroup", S_IRUGO|S_IWUSR, proc_pid_sched_autogroup_operations), #endif +#ifdef CONFIG_TIME_NS + REG("timens_offsets", S_IRUGO|S_IWUSR, proc_timens_offsets_operations), +#endif REG("comm", S_IRUGO|S_IWUSR, proc_pid_set_comm_operations), #ifdef CONFIG_HAVE_ARCH_TRACEHOOK ONE("syscall", S_IRUSR, proc_pid_syscall), @@ -3061,6 +3156,9 @@ static const struct pid_entry tgid_base_stuff[] = { #ifdef CONFIG_CGROUPS ONE("cgroup", S_IRUGO, proc_cgroup_show), #endif +#ifdef CONFIG_PROC_CPU_RESCTRL + ONE("cpu_resctrl_groups", S_IRUGO, proc_resctrl_show), +#endif ONE("oom_score", S_IRUGO, proc_oom_score), REG("oom_adj", S_IRUGO|S_IWUSR, proc_oom_adj_operations), REG("oom_score_adj", S_IRUGO|S_IWUSR, proc_oom_score_adj_operations), @@ -3461,6 +3559,9 @@ static const struct pid_entry tid_base_stuff[] = { #ifdef CONFIG_CGROUPS ONE("cgroup", S_IRUGO, proc_cgroup_show), #endif +#ifdef CONFIG_PROC_CPU_RESCTRL + ONE("cpu_resctrl_groups", S_IRUGO, proc_resctrl_show), +#endif ONE("oom_score", S_IRUGO, proc_oom_score), REG("oom_adj", S_IRUGO|S_IWUSR, proc_oom_adj_operations), REG("oom_score_adj", S_IRUGO|S_IWUSR, proc_oom_score_adj_operations), diff --git a/fs/proc/namespaces.c b/fs/proc/namespaces.c index dd2b35f78b09..8b5c720fe5d7 100644 --- a/fs/proc/namespaces.c +++ b/fs/proc/namespaces.c @@ -33,6 +33,10 @@ static const struct proc_ns_operations *ns_entries[] = { #ifdef CONFIG_CGROUPS &cgroupns_operations, #endif +#ifdef CONFIG_TIME_NS + &timens_operations, + &timens_for_children_operations, +#endif }; static const char *proc_ns_get_link(struct dentry *dentry, diff --git a/fs/proc/stat.c b/fs/proc/stat.c index 37bdbec5b402..fd931d3e77be 100644 --- a/fs/proc/stat.c +++ b/fs/proc/stat.c @@ -134,7 +134,7 @@ static int show_stat(struct seq_file *p, void *v) softirq += cpustat[CPUTIME_SOFTIRQ]; steal += cpustat[CPUTIME_STEAL]; guest += cpustat[CPUTIME_GUEST]; - guest_nice += cpustat[CPUTIME_USER]; + guest_nice += cpustat[CPUTIME_GUEST_NICE]; sum += kstat_cpu_irqs_sum(i); sum += arch_irq_stat_cpu(i); @@ -175,7 +175,7 @@ static int show_stat(struct seq_file *p, void *v) softirq = cpustat[CPUTIME_SOFTIRQ]; steal = cpustat[CPUTIME_STEAL]; guest = cpustat[CPUTIME_GUEST]; - guest_nice = cpustat[CPUTIME_USER]; + guest_nice = cpustat[CPUTIME_GUEST_NICE]; seq_printf(p, "cpu%d", i); seq_put_decimal_ull(p, " ", nsec_to_clock_t(user)); seq_put_decimal_ull(p, " ", nsec_to_clock_t(nice)); diff --git a/fs/proc/uptime.c b/fs/proc/uptime.c index a4c2791ab70b..5a1b228964fb 100644 --- a/fs/proc/uptime.c +++ b/fs/proc/uptime.c @@ -5,6 +5,7 @@ #include <linux/sched.h> #include <linux/seq_file.h> #include <linux/time.h> +#include <linux/time_namespace.h> #include <linux/kernel_stat.h> static int uptime_proc_show(struct seq_file *m, void *v) @@ -20,6 +21,8 @@ static int uptime_proc_show(struct seq_file *m, void *v) nsec += (__force u64) kcpustat_cpu(i).cpustat[CPUTIME_IDLE]; ktime_get_boottime_ts64(&uptime); + timens_add_boottime(&uptime); + idle.tv_sec = div_u64_rem(nsec, NSEC_PER_SEC, &rem); idle.tv_nsec = rem; seq_printf(m, "%lu.%02lu %lu.%02lu\n", diff --git a/fs/pstore/ram.c b/fs/pstore/ram.c index 8caff834f002..013486b5125e 100644 --- a/fs/pstore/ram.c +++ b/fs/pstore/ram.c @@ -407,6 +407,17 @@ static int notrace ramoops_pstore_write(struct pstore_record *record) prz = cxt->dprzs[cxt->dump_write_cnt]; + /* + * Since this is a new crash dump, we need to reset the buffer in + * case it still has an old dump present. Without this, the new dump + * will get appended, which would seriously confuse anything trying + * to check dump file contents. Specifically, ramoops_read_kmsg_hdr() + * expects to find a dump header in the beginning of buffer data, so + * we must to reset the buffer values, in order to ensure that the + * header will be written to the beginning of the buffer. + */ + persistent_ram_zap(prz); + /* Build header and append record contents. */ hlen = ramoops_write_kmsg_hdr(prz, record); if (!hlen) @@ -572,6 +583,7 @@ static int ramoops_init_przs(const char *name, prz_ar[i] = persistent_ram_new(*paddr, zone_sz, sig, &cxt->ecc_info, cxt->memtype, flags, label); + kfree(label); if (IS_ERR(prz_ar[i])) { err = PTR_ERR(prz_ar[i]); dev_err(dev, "failed to request %s mem region (0x%zx@0x%llx): %d\n", @@ -617,6 +629,7 @@ static int ramoops_init_prz(const char *name, label = kasprintf(GFP_KERNEL, "ramoops:%s", name); *prz = persistent_ram_new(*paddr, sz, sig, &cxt->ecc_info, cxt->memtype, PRZ_FLAG_ZAP_OLD, label); + kfree(label); if (IS_ERR(*prz)) { int err = PTR_ERR(*prz); diff --git a/fs/pstore/ram_core.c b/fs/pstore/ram_core.c index 8823f65888f0..1f4d8c06f9be 100644 --- a/fs/pstore/ram_core.c +++ b/fs/pstore/ram_core.c @@ -574,7 +574,7 @@ struct persistent_ram_zone *persistent_ram_new(phys_addr_t start, size_t size, /* Initialize general buffer state. */ raw_spin_lock_init(&prz->buffer_lock); prz->flags = flags; - prz->label = label; + prz->label = kstrdup(label, GFP_KERNEL); ret = persistent_ram_buffer_map(start, size, prz, memtype); if (ret) diff --git a/fs/quota/dquot.c b/fs/quota/dquot.c index b0688c02dc90..b6a4f692d345 100644 --- a/fs/quota/dquot.c +++ b/fs/quota/dquot.c @@ -984,6 +984,7 @@ static int add_dquot_ref(struct super_block *sb, int type) * later. */ old_inode = inode; + cond_resched(); spin_lock(&sb->s_inode_list_lock); } spin_unlock(&sb->s_inode_list_lock); diff --git a/fs/read_write.c b/fs/read_write.c index 5bbf587f5bc1..7458fccc59e1 100644 --- a/fs/read_write.c +++ b/fs/read_write.c @@ -1777,10 +1777,9 @@ static int remap_verify_area(struct file *file, loff_t pos, loff_t len, * else. Assume that the offsets have already been checked for block * alignment. * - * For deduplication we always scale down to the previous block because we - * can't meaningfully compare post-EOF contents. - * - * For clone we only link a partial EOF block above the destination file's EOF. + * For clone we only link a partial EOF block above or at the destination file's + * EOF. For deduplication we accept a partial EOF block only if it ends at the + * destination file's EOF (can not link it into the middle of a file). * * Shorten the request if possible. */ @@ -1796,8 +1795,7 @@ static int generic_remap_check_len(struct inode *inode_in, if ((*len & blkmask) == 0) return 0; - if ((remap_flags & REMAP_FILE_DEDUP) || - pos_out + *len < i_size_read(inode_out)) + if (pos_out + *len < i_size_read(inode_out)) new_len &= ~blkmask; if (new_len == *len) diff --git a/fs/readdir.c b/fs/readdir.c index d26d5ea4de7b..de2eceffdee8 100644 --- a/fs/readdir.c +++ b/fs/readdir.c @@ -102,10 +102,14 @@ EXPORT_SYMBOL(iterate_dir); * filename length, and the above "soft error" worry means * that it's probably better left alone until we have that * issue clarified. + * + * Note the PATH_MAX check - it's arbitrary but the real + * kernel limit on a possible path component, not NAME_MAX, + * which is the technical standard limit. */ static int verify_dirent_name(const char *name, int len) { - if (!len) + if (len <= 0 || len >= PATH_MAX) return -EIO; if (memchr(name, '/', len)) return -EIO; @@ -206,7 +210,7 @@ struct linux_dirent { struct getdents_callback { struct dir_context ctx; struct linux_dirent __user * current_dir; - struct linux_dirent __user * previous; + int prev_reclen; int count; int error; }; @@ -214,12 +218,13 @@ struct getdents_callback { static int filldir(struct dir_context *ctx, const char *name, int namlen, loff_t offset, u64 ino, unsigned int d_type) { - struct linux_dirent __user * dirent; + struct linux_dirent __user *dirent, *prev; struct getdents_callback *buf = container_of(ctx, struct getdents_callback, ctx); unsigned long d_ino; int reclen = ALIGN(offsetof(struct linux_dirent, d_name) + namlen + 2, sizeof(long)); + int prev_reclen; buf->error = verify_dirent_name(name, namlen); if (unlikely(buf->error)) @@ -232,28 +237,24 @@ static int filldir(struct dir_context *ctx, const char *name, int namlen, buf->error = -EOVERFLOW; return -EOVERFLOW; } - dirent = buf->previous; - if (dirent && signal_pending(current)) + prev_reclen = buf->prev_reclen; + if (prev_reclen && signal_pending(current)) return -EINTR; - - /* - * Note! This range-checks 'previous' (which may be NULL). - * The real range was checked in getdents - */ - if (!user_access_begin(dirent, sizeof(*dirent))) - goto efault; - if (dirent) - unsafe_put_user(offset, &dirent->d_off, efault_end); dirent = buf->current_dir; + prev = (void __user *) dirent - prev_reclen; + if (!user_access_begin(prev, reclen + prev_reclen)) + goto efault; + + /* This might be 'dirent->d_off', but if so it will get overwritten */ + unsafe_put_user(offset, &prev->d_off, efault_end); unsafe_put_user(d_ino, &dirent->d_ino, efault_end); unsafe_put_user(reclen, &dirent->d_reclen, efault_end); unsafe_put_user(d_type, (char __user *) dirent + reclen - 1, efault_end); unsafe_copy_dirent_name(dirent->d_name, name, namlen, efault_end); user_access_end(); - buf->previous = dirent; - dirent = (void __user *)dirent + reclen; - buf->current_dir = dirent; + buf->current_dir = (void __user *)dirent + reclen; + buf->prev_reclen = reclen; buf->count -= reclen; return 0; efault_end: @@ -267,7 +268,6 @@ SYSCALL_DEFINE3(getdents, unsigned int, fd, struct linux_dirent __user *, dirent, unsigned int, count) { struct fd f; - struct linux_dirent __user * lastdirent; struct getdents_callback buf = { .ctx.actor = filldir, .count = count, @@ -285,8 +285,10 @@ SYSCALL_DEFINE3(getdents, unsigned int, fd, error = iterate_dir(f.file, &buf.ctx); if (error >= 0) error = buf.error; - lastdirent = buf.previous; - if (lastdirent) { + if (buf.prev_reclen) { + struct linux_dirent __user * lastdirent; + lastdirent = (void __user *)buf.current_dir - buf.prev_reclen; + if (put_user(buf.ctx.pos, &lastdirent->d_off)) error = -EFAULT; else @@ -299,7 +301,7 @@ SYSCALL_DEFINE3(getdents, unsigned int, fd, struct getdents_callback64 { struct dir_context ctx; struct linux_dirent64 __user * current_dir; - struct linux_dirent64 __user * previous; + int prev_reclen; int count; int error; }; @@ -307,11 +309,12 @@ struct getdents_callback64 { static int filldir64(struct dir_context *ctx, const char *name, int namlen, loff_t offset, u64 ino, unsigned int d_type) { - struct linux_dirent64 __user *dirent; + struct linux_dirent64 __user *dirent, *prev; struct getdents_callback64 *buf = container_of(ctx, struct getdents_callback64, ctx); int reclen = ALIGN(offsetof(struct linux_dirent64, d_name) + namlen + 1, sizeof(u64)); + int prev_reclen; buf->error = verify_dirent_name(name, namlen); if (unlikely(buf->error)) @@ -319,30 +322,27 @@ static int filldir64(struct dir_context *ctx, const char *name, int namlen, buf->error = -EINVAL; /* only used if we fail.. */ if (reclen > buf->count) return -EINVAL; - dirent = buf->previous; - if (dirent && signal_pending(current)) + prev_reclen = buf->prev_reclen; + if (prev_reclen && signal_pending(current)) return -EINTR; - - /* - * Note! This range-checks 'previous' (which may be NULL). - * The real range was checked in getdents - */ - if (!user_access_begin(dirent, sizeof(*dirent))) - goto efault; - if (dirent) - unsafe_put_user(offset, &dirent->d_off, efault_end); dirent = buf->current_dir; + prev = (void __user *)dirent - prev_reclen; + if (!user_access_begin(prev, reclen + prev_reclen)) + goto efault; + + /* This might be 'dirent->d_off', but if so it will get overwritten */ + unsafe_put_user(offset, &prev->d_off, efault_end); unsafe_put_user(ino, &dirent->d_ino, efault_end); unsafe_put_user(reclen, &dirent->d_reclen, efault_end); unsafe_put_user(d_type, &dirent->d_type, efault_end); unsafe_copy_dirent_name(dirent->d_name, name, namlen, efault_end); user_access_end(); - buf->previous = dirent; - dirent = (void __user *)dirent + reclen; - buf->current_dir = dirent; + buf->prev_reclen = reclen; + buf->current_dir = (void __user *)dirent + reclen; buf->count -= reclen; return 0; + efault_end: user_access_end(); efault: @@ -354,7 +354,6 @@ int ksys_getdents64(unsigned int fd, struct linux_dirent64 __user *dirent, unsigned int count) { struct fd f; - struct linux_dirent64 __user * lastdirent; struct getdents_callback64 buf = { .ctx.actor = filldir64, .count = count, @@ -372,9 +371,11 @@ int ksys_getdents64(unsigned int fd, struct linux_dirent64 __user *dirent, error = iterate_dir(f.file, &buf.ctx); if (error >= 0) error = buf.error; - lastdirent = buf.previous; - if (lastdirent) { + if (buf.prev_reclen) { + struct linux_dirent64 __user * lastdirent; typeof(lastdirent->d_off) d_off = buf.ctx.pos; + + lastdirent = (void __user *) buf.current_dir - buf.prev_reclen; if (__put_user(d_off, &lastdirent->d_off)) error = -EFAULT; else diff --git a/fs/reiserfs/xattr.c b/fs/reiserfs/xattr.c index 62b40df36c98..28b241cd6987 100644 --- a/fs/reiserfs/xattr.c +++ b/fs/reiserfs/xattr.c @@ -319,8 +319,12 @@ static int reiserfs_for_each_xattr(struct inode *inode, out_dir: dput(dir); out: - /* -ENODATA isn't an error */ - if (err == -ENODATA) + /* + * -ENODATA: this object doesn't have any xattrs + * -EOPNOTSUPP: this file system doesn't have xattrs enabled on disk. + * Neither are errors + */ + if (err == -ENODATA || err == -EOPNOTSUPP) err = 0; return err; } diff --git a/fs/stack.c b/fs/stack.c index 4ef2c056269d..c9830924eb12 100644 --- a/fs/stack.c +++ b/fs/stack.c @@ -23,7 +23,7 @@ void fsstack_copy_inode_size(struct inode *dst, struct inode *src) /* * But on 32-bit, we ought to make an effort to keep the two halves of - * i_blocks in sync despite SMP or PREEMPT - though stat's + * i_blocks in sync despite SMP or PREEMPTION - though stat's * generic_fillattr() doesn't bother, and we won't be applying quotas * (where i_blocks does become important) at the upper level. * @@ -38,14 +38,14 @@ void fsstack_copy_inode_size(struct inode *dst, struct inode *src) spin_unlock(&src->i_lock); /* - * If CONFIG_SMP or CONFIG_PREEMPT on 32-bit, it's vital for + * If CONFIG_SMP or CONFIG_PREEMPTION on 32-bit, it's vital for * fsstack_copy_inode_size() to hold some lock around * i_size_write(), otherwise i_size_read() may spin forever (see * include/linux/fs.h). We don't necessarily hold i_mutex when this * is called, so take i_lock for that case. * * And if on 32-bit, continue our effort to keep the two halves of - * i_blocks in sync despite SMP or PREEMPT: use i_lock for that case + * i_blocks in sync despite SMP or PREEMPTION: use i_lock for that case * too, and do both at once by combining the tests. * * There is none of this locking overhead in the 64-bit case. diff --git a/fs/super.c b/fs/super.c index cfadab2cbf35..cd352530eca9 100644 --- a/fs/super.c +++ b/fs/super.c @@ -448,10 +448,12 @@ void generic_shutdown_super(struct super_block *sb) sync_filesystem(sb); sb->s_flags &= ~SB_ACTIVE; - fsnotify_sb_delete(sb); cgroup_writeback_umount(); + /* evict all inodes with zero refcount */ evict_inodes(sb); + /* only nonzero refcount inodes can have marks */ + fsnotify_sb_delete(sb); if (sb->s_dio_done_wq) { destroy_workqueue(sb->s_dio_done_wq); diff --git a/fs/timerfd.c b/fs/timerfd.c index ac7f59a58f94..c5509d2448e3 100644 --- a/fs/timerfd.c +++ b/fs/timerfd.c @@ -26,6 +26,7 @@ #include <linux/syscalls.h> #include <linux/compat.h> #include <linux/rcupdate.h> +#include <linux/time_namespace.h> struct timerfd_ctx { union { @@ -196,6 +197,8 @@ static int timerfd_setup(struct timerfd_ctx *ctx, int flags, } if (texp != 0) { + if (flags & TFD_TIMER_ABSTIME) + texp = timens_ktime_to_host(clockid, texp); if (isalarm(ctx)) { if (flags & TFD_TIMER_ABSTIME) alarm_start(&ctx->t.alarm, texp); diff --git a/fs/ubifs/Kconfig b/fs/ubifs/Kconfig index 69932bcfa920..45d3d207fb99 100644 --- a/fs/ubifs/Kconfig +++ b/fs/ubifs/Kconfig @@ -12,6 +12,7 @@ config UBIFS_FS select CRYPTO_ZSTD if UBIFS_FS_ZSTD select CRYPTO_HASH_INFO select UBIFS_FS_XATTR if FS_ENCRYPTION + select FS_ENCRYPTION_ALGS if FS_ENCRYPTION depends on MTD_UBI help UBIFS is a file system for flash devices which works on top of UBI. diff --git a/fs/ubifs/dir.c b/fs/ubifs/dir.c index 0b98e3c8b461..ef85ec167a84 100644 --- a/fs/ubifs/dir.c +++ b/fs/ubifs/dir.c @@ -81,7 +81,7 @@ struct inode *ubifs_new_inode(struct ubifs_info *c, struct inode *dir, struct ubifs_inode *ui; bool encrypted = false; - if (ubifs_crypt_is_encrypted(dir)) { + if (IS_ENCRYPTED(dir)) { err = fscrypt_get_encryption_info(dir); if (err) { ubifs_err(c, "fscrypt_get_encryption_info failed: %i", err); @@ -225,9 +225,9 @@ static struct dentry *ubifs_lookup(struct inode *dir, struct dentry *dentry, goto done; } - if (nm.hash) { - ubifs_assert(c, fname_len(&nm) == 0); - ubifs_assert(c, fname_name(&nm) == NULL); + if (fname_name(&nm) == NULL) { + if (nm.hash & ~UBIFS_S_KEY_HASH_MASK) + goto done; /* ENOENT */ dent_key_init_hash(c, &key, dir->i_ino, nm.hash); err = ubifs_tnc_lookup_dh(c, &key, dent, nm.minor_hash); } else { @@ -261,7 +261,7 @@ static struct dentry *ubifs_lookup(struct inode *dir, struct dentry *dentry, goto done; } - if (ubifs_crypt_is_encrypted(dir) && + if (IS_ENCRYPTED(dir) && (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode)) && !fscrypt_has_permitted_context(dir, inode)) { ubifs_warn(c, "Inconsistent encryption contexts: %lu/%lu", @@ -499,7 +499,7 @@ static int ubifs_readdir(struct file *file, struct dir_context *ctx) struct ubifs_dent_node *dent; struct inode *dir = file_inode(file); struct ubifs_info *c = dir->i_sb->s_fs_info; - bool encrypted = ubifs_crypt_is_encrypted(dir); + bool encrypted = IS_ENCRYPTED(dir); dbg_gen("dir ino %lu, f_pos %#llx", dir->i_ino, ctx->pos); @@ -512,7 +512,7 @@ static int ubifs_readdir(struct file *file, struct dir_context *ctx) if (encrypted) { err = fscrypt_get_encryption_info(dir); - if (err && err != -ENOKEY) + if (err) return err; err = fscrypt_fname_alloc_buffer(dir, UBIFS_MAX_NLEN, &fstr); @@ -1618,7 +1618,7 @@ int ubifs_getattr(const struct path *path, struct kstat *stat, static int ubifs_dir_open(struct inode *dir, struct file *file) { - if (ubifs_crypt_is_encrypted(dir)) + if (IS_ENCRYPTED(dir)) return fscrypt_get_encryption_info(dir) ? -EACCES : 0; return 0; diff --git a/fs/ubifs/file.c b/fs/ubifs/file.c index cd52585c8f4f..c8e8f50c6054 100644 --- a/fs/ubifs/file.c +++ b/fs/ubifs/file.c @@ -67,7 +67,7 @@ static int read_block(struct inode *inode, void *addr, unsigned int block, dlen = le32_to_cpu(dn->ch.len) - UBIFS_DATA_NODE_SZ; - if (ubifs_crypt_is_encrypted(inode)) { + if (IS_ENCRYPTED(inode)) { err = ubifs_decrypt(inode, dn, &dlen, block); if (err) goto dump; @@ -647,7 +647,7 @@ static int populate_page(struct ubifs_info *c, struct page *page, dlen = le32_to_cpu(dn->ch.len) - UBIFS_DATA_NODE_SZ; out_len = UBIFS_BLOCK_SIZE; - if (ubifs_crypt_is_encrypted(inode)) { + if (IS_ENCRYPTED(inode)) { err = ubifs_decrypt(inode, dn, &dlen, page_block); if (err) goto out_err; diff --git a/fs/ubifs/journal.c b/fs/ubifs/journal.c index 388fe8f5dc51..3bf8b1fda9d7 100644 --- a/fs/ubifs/journal.c +++ b/fs/ubifs/journal.c @@ -588,7 +588,7 @@ int ubifs_jnl_update(struct ubifs_info *c, const struct inode *dir, if (!xent) { dent->ch.node_type = UBIFS_DENT_NODE; - if (nm->hash) + if (fname_name(nm) == NULL) dent_key_init_hash(c, &dent_key, dir->i_ino, nm->hash); else dent_key_init(c, &dent_key, dir->i_ino, nm); @@ -646,7 +646,7 @@ int ubifs_jnl_update(struct ubifs_info *c, const struct inode *dir, ubifs_add_auth_dirt(c, lnum); if (deletion) { - if (nm->hash) + if (fname_name(nm) == NULL) err = ubifs_tnc_remove_dh(c, &dent_key, nm->minor_hash); else err = ubifs_tnc_remove_nm(c, &dent_key, nm); @@ -727,7 +727,7 @@ int ubifs_jnl_write_data(struct ubifs_info *c, const struct inode *inode, int dlen = COMPRESSED_DATA_NODE_BUF_SZ, allocated = 1; int write_len; struct ubifs_inode *ui = ubifs_inode(inode); - bool encrypted = ubifs_crypt_is_encrypted(inode); + bool encrypted = IS_ENCRYPTED(inode); u8 hash[UBIFS_HASH_ARR_SZ]; dbg_jnlk(key, "ino %lu, blk %u, len %d, key ", @@ -1449,7 +1449,7 @@ static int truncate_data_node(const struct ubifs_info *c, const struct inode *in dlen = old_dlen = le32_to_cpu(dn->ch.len) - UBIFS_DATA_NODE_SZ; compr_type = le16_to_cpu(dn->compr_type); - if (ubifs_crypt_is_encrypted(inode)) { + if (IS_ENCRYPTED(inode)) { err = ubifs_decrypt(inode, dn, &dlen, block); if (err) goto out; @@ -1465,7 +1465,7 @@ static int truncate_data_node(const struct ubifs_info *c, const struct inode *in ubifs_compress(c, buf, *new_len, &dn->data, &out_len, &compr_type); } - if (ubifs_crypt_is_encrypted(inode)) { + if (IS_ENCRYPTED(inode)) { err = ubifs_encrypt(inode, dn, out_len, &old_dlen, block); if (err) goto out; diff --git a/fs/ubifs/key.h b/fs/ubifs/key.h index afa704ff5ca0..8142d9d6fe5d 100644 --- a/fs/ubifs/key.h +++ b/fs/ubifs/key.h @@ -150,7 +150,6 @@ static inline void dent_key_init(const struct ubifs_info *c, uint32_t hash = c->key_hash(fname_name(nm), fname_len(nm)); ubifs_assert(c, !(hash & ~UBIFS_S_KEY_HASH_MASK)); - ubifs_assert(c, !nm->hash && !nm->minor_hash); key->u32[0] = inum; key->u32[1] = hash | (UBIFS_DENT_KEY << UBIFS_S_KEY_HASH_BITS); } diff --git a/fs/ubifs/ubifs.h b/fs/ubifs/ubifs.h index c55f212dcb75..bff682309fbe 100644 --- a/fs/ubifs/ubifs.h +++ b/fs/ubifs/ubifs.h @@ -2095,13 +2095,6 @@ int ubifs_decrypt(const struct inode *inode, struct ubifs_data_node *dn, extern const struct fscrypt_operations ubifs_crypt_operations; -static inline bool ubifs_crypt_is_encrypted(const struct inode *inode) -{ - const struct ubifs_inode *ui = ubifs_inode(inode); - - return ui->flags & UBIFS_CRYPT_FL; -} - /* Normal UBIFS messages */ __printf(2, 3) void ubifs_msg(const struct ubifs_info *c, const char *fmt, ...); diff --git a/fs/verity/enable.c b/fs/verity/enable.c index eabc6ac19906..d98bea308fd7 100644 --- a/fs/verity/enable.c +++ b/fs/verity/enable.c @@ -8,18 +8,48 @@ #include "fsverity_private.h" #include <crypto/hash.h> +#include <linux/backing-dev.h> #include <linux/mount.h> #include <linux/pagemap.h> #include <linux/sched/signal.h> #include <linux/uaccess.h> -static int build_merkle_tree_level(struct inode *inode, unsigned int level, +/* + * Read a file data page for Merkle tree construction. Do aggressive readahead, + * since we're sequentially reading the entire file. + */ +static struct page *read_file_data_page(struct file *filp, pgoff_t index, + struct file_ra_state *ra, + unsigned long remaining_pages) +{ + struct page *page; + + page = find_get_page_flags(filp->f_mapping, index, FGP_ACCESSED); + if (!page || !PageUptodate(page)) { + if (page) + put_page(page); + else + page_cache_sync_readahead(filp->f_mapping, ra, filp, + index, remaining_pages); + page = read_mapping_page(filp->f_mapping, index, NULL); + if (IS_ERR(page)) + return page; + } + if (PageReadahead(page)) + page_cache_async_readahead(filp->f_mapping, ra, filp, page, + index, remaining_pages); + return page; +} + +static int build_merkle_tree_level(struct file *filp, unsigned int level, u64 num_blocks_to_hash, const struct merkle_tree_params *params, u8 *pending_hashes, struct ahash_request *req) { + struct inode *inode = file_inode(filp); const struct fsverity_operations *vops = inode->i_sb->s_vop; + struct file_ra_state ra = { 0 }; unsigned int pending_size = 0; u64 dst_block_num; u64 i; @@ -36,6 +66,8 @@ static int build_merkle_tree_level(struct inode *inode, unsigned int level, dst_block_num = 0; /* unused */ } + file_ra_state_init(&ra, filp->f_mapping); + for (i = 0; i < num_blocks_to_hash; i++) { struct page *src_page; @@ -45,7 +77,8 @@ static int build_merkle_tree_level(struct inode *inode, unsigned int level, if (level == 0) { /* Leaf: hashing a data block */ - src_page = read_mapping_page(inode->i_mapping, i, NULL); + src_page = read_file_data_page(filp, i, &ra, + num_blocks_to_hash - i); if (IS_ERR(src_page)) { err = PTR_ERR(src_page); fsverity_err(inode, @@ -54,9 +87,14 @@ static int build_merkle_tree_level(struct inode *inode, unsigned int level, return err; } } else { + unsigned long num_ra_pages = + min_t(unsigned long, num_blocks_to_hash - i, + inode->i_sb->s_bdi->io_pages); + /* Non-leaf: hashing hash block from level below */ src_page = vops->read_merkle_tree_page(inode, - params->level_start[level - 1] + i); + params->level_start[level - 1] + i, + num_ra_pages); if (IS_ERR(src_page)) { err = PTR_ERR(src_page); fsverity_err(inode, @@ -103,17 +141,18 @@ static int build_merkle_tree_level(struct inode *inode, unsigned int level, } /* - * Build the Merkle tree for the given inode using the given parameters, and + * Build the Merkle tree for the given file using the given parameters, and * return the root hash in @root_hash. * * The tree is written to a filesystem-specific location as determined by the * ->write_merkle_tree_block() method. However, the blocks that comprise the * tree are the same for all filesystems. */ -static int build_merkle_tree(struct inode *inode, +static int build_merkle_tree(struct file *filp, const struct merkle_tree_params *params, u8 *root_hash) { + struct inode *inode = file_inode(filp); u8 *pending_hashes; struct ahash_request *req; u64 blocks; @@ -126,9 +165,11 @@ static int build_merkle_tree(struct inode *inode, return 0; } + /* This allocation never fails, since it's mempool-backed. */ + req = fsverity_alloc_hash_request(params->hash_alg, GFP_KERNEL); + pending_hashes = kmalloc(params->block_size, GFP_KERNEL); - req = ahash_request_alloc(params->hash_alg->tfm, GFP_KERNEL); - if (!pending_hashes || !req) + if (!pending_hashes) goto out; /* @@ -139,7 +180,7 @@ static int build_merkle_tree(struct inode *inode, blocks = (inode->i_size + params->block_size - 1) >> params->log_blocksize; for (level = 0; level <= params->num_levels; level++) { - err = build_merkle_tree_level(inode, level, blocks, params, + err = build_merkle_tree_level(filp, level, blocks, params, pending_hashes, req); if (err) goto out; @@ -150,7 +191,7 @@ static int build_merkle_tree(struct inode *inode, err = 0; out: kfree(pending_hashes); - ahash_request_free(req); + fsverity_free_hash_request(params->hash_alg, req); return err; } @@ -175,8 +216,7 @@ static int enable_verity(struct file *filp, /* Get the salt if the user provided one */ if (arg->salt_size && - copy_from_user(desc->salt, - (const u8 __user *)(uintptr_t)arg->salt_ptr, + copy_from_user(desc->salt, u64_to_user_ptr(arg->salt_ptr), arg->salt_size)) { err = -EFAULT; goto out; @@ -185,8 +225,7 @@ static int enable_verity(struct file *filp, /* Get the signature if the user provided one */ if (arg->sig_size && - copy_from_user(desc->signature, - (const u8 __user *)(uintptr_t)arg->sig_ptr, + copy_from_user(desc->signature, u64_to_user_ptr(arg->sig_ptr), arg->sig_size)) { err = -EFAULT; goto out; @@ -227,7 +266,7 @@ static int enable_verity(struct file *filp, */ pr_debug("Building Merkle tree...\n"); BUILD_BUG_ON(sizeof(desc->root_hash) < FS_VERITY_MAX_DIGEST_SIZE); - err = build_merkle_tree(inode, ¶ms, desc->root_hash); + err = build_merkle_tree(filp, ¶ms, desc->root_hash); if (err) { fsverity_err(inode, "Error %d building Merkle tree", err); goto rollback; @@ -315,7 +354,7 @@ int fsverity_ioctl_enable(struct file *filp, const void __user *uarg) if (arg.block_size != PAGE_SIZE) return -EINVAL; - if (arg.salt_size > FIELD_SIZEOF(struct fsverity_descriptor, salt)) + if (arg.salt_size > sizeof_field(struct fsverity_descriptor, salt)) return -EMSGSIZE; if (arg.sig_size > FS_VERITY_MAX_SIGNATURE_SIZE) diff --git a/fs/verity/fsverity_private.h b/fs/verity/fsverity_private.h index e74c79b64d88..74768cf539da 100644 --- a/fs/verity/fsverity_private.h +++ b/fs/verity/fsverity_private.h @@ -16,6 +16,7 @@ #include <crypto/sha.h> #include <linux/fsverity.h> +#include <linux/mempool.h> struct ahash_request; @@ -37,11 +38,12 @@ struct fsverity_hash_alg { const char *name; /* crypto API name, e.g. sha256 */ unsigned int digest_size; /* digest size in bytes, e.g. 32 for SHA-256 */ unsigned int block_size; /* block size in bytes, e.g. 64 for SHA-256 */ + mempool_t req_pool; /* mempool with a preallocated hash request */ }; /* Merkle tree parameters: hash algorithm, initial hash state, and topology */ struct merkle_tree_params { - const struct fsverity_hash_alg *hash_alg; /* the hash algorithm */ + struct fsverity_hash_alg *hash_alg; /* the hash algorithm */ const u8 *hashstate; /* initial hash state or NULL */ unsigned int digest_size; /* same as hash_alg->digest_size */ unsigned int block_size; /* size of data and tree blocks */ @@ -50,6 +52,7 @@ struct merkle_tree_params { unsigned int log_arity; /* log2(hashes_per_block) */ unsigned int num_levels; /* number of levels in Merkle tree */ u64 tree_size; /* Merkle tree size in bytes */ + unsigned long level0_blocks; /* number of blocks in tree level 0 */ /* * Starting block index for each tree level, ordered from leaf level (0) @@ -114,14 +117,18 @@ struct fsverity_signed_digest { extern struct fsverity_hash_alg fsverity_hash_algs[]; -const struct fsverity_hash_alg *fsverity_get_hash_alg(const struct inode *inode, - unsigned int num); -const u8 *fsverity_prepare_hash_state(const struct fsverity_hash_alg *alg, +struct fsverity_hash_alg *fsverity_get_hash_alg(const struct inode *inode, + unsigned int num); +struct ahash_request *fsverity_alloc_hash_request(struct fsverity_hash_alg *alg, + gfp_t gfp_flags); +void fsverity_free_hash_request(struct fsverity_hash_alg *alg, + struct ahash_request *req); +const u8 *fsverity_prepare_hash_state(struct fsverity_hash_alg *alg, const u8 *salt, size_t salt_size); int fsverity_hash_page(const struct merkle_tree_params *params, const struct inode *inode, struct ahash_request *req, struct page *page, u8 *out); -int fsverity_hash_buffer(const struct fsverity_hash_alg *alg, +int fsverity_hash_buffer(struct fsverity_hash_alg *alg, const void *data, size_t size, u8 *out); void __init fsverity_check_hash_algs(void); diff --git a/fs/verity/hash_algs.c b/fs/verity/hash_algs.c index 31e6d7d2389a..c37e186ebeb6 100644 --- a/fs/verity/hash_algs.c +++ b/fs/verity/hash_algs.c @@ -24,6 +24,8 @@ struct fsverity_hash_alg fsverity_hash_algs[] = { }, }; +static DEFINE_MUTEX(fsverity_hash_alg_init_mutex); + /** * fsverity_get_hash_alg() - validate and prepare a hash algorithm * @inode: optional inode for logging purposes @@ -36,8 +38,8 @@ struct fsverity_hash_alg fsverity_hash_algs[] = { * * Return: pointer to the hash alg on success, else an ERR_PTR() */ -const struct fsverity_hash_alg *fsverity_get_hash_alg(const struct inode *inode, - unsigned int num) +struct fsverity_hash_alg *fsverity_get_hash_alg(const struct inode *inode, + unsigned int num) { struct fsverity_hash_alg *alg; struct crypto_ahash *tfm; @@ -50,10 +52,15 @@ const struct fsverity_hash_alg *fsverity_get_hash_alg(const struct inode *inode, } alg = &fsverity_hash_algs[num]; - /* pairs with cmpxchg() below */ - tfm = READ_ONCE(alg->tfm); - if (likely(tfm != NULL)) + /* pairs with smp_store_release() below */ + if (likely(smp_load_acquire(&alg->tfm) != NULL)) return alg; + + mutex_lock(&fsverity_hash_alg_init_mutex); + + if (alg->tfm != NULL) + goto out_unlock; + /* * Using the shash API would make things a bit simpler, but the ahash * API is preferable as it allows the use of crypto accelerators. @@ -64,12 +71,14 @@ const struct fsverity_hash_alg *fsverity_get_hash_alg(const struct inode *inode, fsverity_warn(inode, "Missing crypto API support for hash algorithm \"%s\"", alg->name); - return ERR_PTR(-ENOPKG); + alg = ERR_PTR(-ENOPKG); + goto out_unlock; } fsverity_err(inode, "Error allocating hash algorithm \"%s\": %ld", alg->name, PTR_ERR(tfm)); - return ERR_CAST(tfm); + alg = ERR_CAST(tfm); + goto out_unlock; } err = -EINVAL; @@ -78,18 +87,61 @@ const struct fsverity_hash_alg *fsverity_get_hash_alg(const struct inode *inode, if (WARN_ON(alg->block_size != crypto_ahash_blocksize(tfm))) goto err_free_tfm; + err = mempool_init_kmalloc_pool(&alg->req_pool, 1, + sizeof(struct ahash_request) + + crypto_ahash_reqsize(tfm)); + if (err) + goto err_free_tfm; + pr_info("%s using implementation \"%s\"\n", alg->name, crypto_ahash_driver_name(tfm)); - /* pairs with READ_ONCE() above */ - if (cmpxchg(&alg->tfm, NULL, tfm) != NULL) - crypto_free_ahash(tfm); - - return alg; + /* pairs with smp_load_acquire() above */ + smp_store_release(&alg->tfm, tfm); + goto out_unlock; err_free_tfm: crypto_free_ahash(tfm); - return ERR_PTR(err); + alg = ERR_PTR(err); +out_unlock: + mutex_unlock(&fsverity_hash_alg_init_mutex); + return alg; +} + +/** + * fsverity_alloc_hash_request() - allocate a hash request object + * @alg: the hash algorithm for which to allocate the request + * @gfp_flags: memory allocation flags + * + * This is mempool-backed, so this never fails if __GFP_DIRECT_RECLAIM is set in + * @gfp_flags. However, in that case this might need to wait for all + * previously-allocated requests to be freed. So to avoid deadlocks, callers + * must never need multiple requests at a time to make forward progress. + * + * Return: the request object on success; NULL on failure (but see above) + */ +struct ahash_request *fsverity_alloc_hash_request(struct fsverity_hash_alg *alg, + gfp_t gfp_flags) +{ + struct ahash_request *req = mempool_alloc(&alg->req_pool, gfp_flags); + + if (req) + ahash_request_set_tfm(req, alg->tfm); + return req; +} + +/** + * fsverity_free_hash_request() - free a hash request object + * @alg: the hash algorithm + * @req: the hash request object to free + */ +void fsverity_free_hash_request(struct fsverity_hash_alg *alg, + struct ahash_request *req) +{ + if (req) { + ahash_request_zero(req); + mempool_free(req, &alg->req_pool); + } } /** @@ -101,7 +153,7 @@ err_free_tfm: * Return: NULL if the salt is empty, otherwise the kmalloc()'ed precomputed * initial hash state on success or an ERR_PTR() on failure. */ -const u8 *fsverity_prepare_hash_state(const struct fsverity_hash_alg *alg, +const u8 *fsverity_prepare_hash_state(struct fsverity_hash_alg *alg, const u8 *salt, size_t salt_size) { u8 *hashstate = NULL; @@ -119,11 +171,8 @@ const u8 *fsverity_prepare_hash_state(const struct fsverity_hash_alg *alg, if (!hashstate) return ERR_PTR(-ENOMEM); - req = ahash_request_alloc(alg->tfm, GFP_KERNEL); - if (!req) { - err = -ENOMEM; - goto err_free; - } + /* This allocation never fails, since it's mempool-backed. */ + req = fsverity_alloc_hash_request(alg, GFP_KERNEL); /* * Zero-pad the salt to the next multiple of the input size of the hash @@ -158,7 +207,7 @@ const u8 *fsverity_prepare_hash_state(const struct fsverity_hash_alg *alg, if (err) goto err_free; out: - ahash_request_free(req); + fsverity_free_hash_request(alg, req); kfree(padded_salt); return hashstate; @@ -229,7 +278,7 @@ int fsverity_hash_page(const struct merkle_tree_params *params, * * Return: 0 on success, -errno on failure */ -int fsverity_hash_buffer(const struct fsverity_hash_alg *alg, +int fsverity_hash_buffer(struct fsverity_hash_alg *alg, const void *data, size_t size, u8 *out) { struct ahash_request *req; @@ -237,9 +286,8 @@ int fsverity_hash_buffer(const struct fsverity_hash_alg *alg, DECLARE_CRYPTO_WAIT(wait); int err; - req = ahash_request_alloc(alg->tfm, GFP_KERNEL); - if (!req) - return -ENOMEM; + /* This allocation never fails, since it's mempool-backed. */ + req = fsverity_alloc_hash_request(alg, GFP_KERNEL); sg_init_one(&sg, data, size); ahash_request_set_callback(req, CRYPTO_TFM_REQ_MAY_SLEEP | @@ -249,7 +297,7 @@ int fsverity_hash_buffer(const struct fsverity_hash_alg *alg, err = crypto_wait_req(crypto_ahash_digest(req), &wait); - ahash_request_free(req); + fsverity_free_hash_request(alg, req); return err; } diff --git a/fs/verity/open.c b/fs/verity/open.c index 63d1004b688c..c5fe6948e262 100644 --- a/fs/verity/open.c +++ b/fs/verity/open.c @@ -31,7 +31,7 @@ int fsverity_init_merkle_tree_params(struct merkle_tree_params *params, unsigned int log_blocksize, const u8 *salt, size_t salt_size) { - const struct fsverity_hash_alg *hash_alg; + struct fsverity_hash_alg *hash_alg; int err; u64 blocks; u64 offset; @@ -102,6 +102,7 @@ int fsverity_init_merkle_tree_params(struct merkle_tree_params *params, /* temporarily using level_start[] to store blocks in level */ params->level_start[params->num_levels++] = blocks; } + params->level0_blocks = params->level_start[0]; /* Compute the starting block of each level */ offset = 0; @@ -126,7 +127,7 @@ out_err: * Compute the file measurement by hashing the fsverity_descriptor excluding the * signature and with the sig_size field set to 0. */ -static int compute_file_measurement(const struct fsverity_hash_alg *hash_alg, +static int compute_file_measurement(struct fsverity_hash_alg *hash_alg, struct fsverity_descriptor *desc, u8 *measurement) { diff --git a/fs/verity/verify.c b/fs/verity/verify.c index 3e8f2de44667..e0cb62da3864 100644 --- a/fs/verity/verify.c +++ b/fs/verity/verify.c @@ -84,7 +84,8 @@ static inline int cmp_hashes(const struct fsverity_info *vi, * Return: true if the page is valid, else false. */ static bool verify_page(struct inode *inode, const struct fsverity_info *vi, - struct ahash_request *req, struct page *data_page) + struct ahash_request *req, struct page *data_page, + unsigned long level0_ra_pages) { const struct merkle_tree_params *params = &vi->tree_params; const unsigned int hsize = params->digest_size; @@ -117,8 +118,8 @@ static bool verify_page(struct inode *inode, const struct fsverity_info *vi, pr_debug_ratelimited("Level %d: hindex=%lu, hoffset=%u\n", level, hindex, hoffset); - hpage = inode->i_sb->s_vop->read_merkle_tree_page(inode, - hindex); + hpage = inode->i_sb->s_vop->read_merkle_tree_page(inode, hindex, + level == 0 ? level0_ra_pages : 0); if (IS_ERR(hpage)) { err = PTR_ERR(hpage); fsverity_err(inode, @@ -191,13 +192,12 @@ bool fsverity_verify_page(struct page *page) struct ahash_request *req; bool valid; - req = ahash_request_alloc(vi->tree_params.hash_alg->tfm, GFP_NOFS); - if (unlikely(!req)) - return false; + /* This allocation never fails, since it's mempool-backed. */ + req = fsverity_alloc_hash_request(vi->tree_params.hash_alg, GFP_NOFS); - valid = verify_page(inode, vi, req, page); + valid = verify_page(inode, vi, req, page, 0); - ahash_request_free(req); + fsverity_free_hash_request(vi->tree_params.hash_alg, req); return valid; } @@ -222,25 +222,42 @@ void fsverity_verify_bio(struct bio *bio) { struct inode *inode = bio_first_page_all(bio)->mapping->host; const struct fsverity_info *vi = inode->i_verity_info; + const struct merkle_tree_params *params = &vi->tree_params; struct ahash_request *req; struct bio_vec *bv; struct bvec_iter_all iter_all; - - req = ahash_request_alloc(vi->tree_params.hash_alg->tfm, GFP_NOFS); - if (unlikely(!req)) { + unsigned long max_ra_pages = 0; + + /* This allocation never fails, since it's mempool-backed. */ + req = fsverity_alloc_hash_request(params->hash_alg, GFP_NOFS); + + if (bio->bi_opf & REQ_RAHEAD) { + /* + * If this bio is for data readahead, then we also do readahead + * of the first (largest) level of the Merkle tree. Namely, + * when a Merkle tree page is read, we also try to piggy-back on + * some additional pages -- up to 1/4 the number of data pages. + * + * This improves sequential read performance, as it greatly + * reduces the number of I/O requests made to the Merkle tree. + */ bio_for_each_segment_all(bv, bio, iter_all) - SetPageError(bv->bv_page); - return; + max_ra_pages++; + max_ra_pages /= 4; } bio_for_each_segment_all(bv, bio, iter_all) { struct page *page = bv->bv_page; + unsigned long level0_index = page->index >> params->log_arity; + unsigned long level0_ra_pages = + min(max_ra_pages, params->level0_blocks - level0_index); - if (!PageError(page) && !verify_page(inode, vi, req, page)) + if (!PageError(page) && + !verify_page(inode, vi, req, page, level0_ra_pages)) SetPageError(page); } - ahash_request_free(req); + fsverity_free_hash_request(params->hash_alg, req); } EXPORT_SYMBOL_GPL(fsverity_verify_bio); #endif /* CONFIG_BLOCK */ diff --git a/fs/xfs/libxfs/xfs_alloc.c b/fs/xfs/libxfs/xfs_alloc.c index c284e10af491..fc93fd88ec89 100644 --- a/fs/xfs/libxfs/xfs_alloc.c +++ b/fs/xfs/libxfs/xfs_alloc.c @@ -2248,24 +2248,32 @@ xfs_alloc_longest_free_extent( return pag->pagf_flcount > 0 || pag->pagf_longest > 0; } +/* + * Compute the minimum length of the AGFL in the given AG. If @pag is NULL, + * return the largest possible minimum length. + */ unsigned int xfs_alloc_min_freelist( struct xfs_mount *mp, struct xfs_perag *pag) { + /* AG btrees have at least 1 level. */ + static const uint8_t fake_levels[XFS_BTNUM_AGF] = {1, 1, 1}; + const uint8_t *levels = pag ? pag->pagf_levels : fake_levels; unsigned int min_free; + ASSERT(mp->m_ag_maxlevels > 0); + /* space needed by-bno freespace btree */ - min_free = min_t(unsigned int, pag->pagf_levels[XFS_BTNUM_BNOi] + 1, + min_free = min_t(unsigned int, levels[XFS_BTNUM_BNOi] + 1, mp->m_ag_maxlevels); /* space needed by-size freespace btree */ - min_free += min_t(unsigned int, pag->pagf_levels[XFS_BTNUM_CNTi] + 1, + min_free += min_t(unsigned int, levels[XFS_BTNUM_CNTi] + 1, mp->m_ag_maxlevels); /* space needed reverse mapping used space btree */ if (xfs_sb_version_hasrmapbt(&mp->m_sb)) - min_free += min_t(unsigned int, - pag->pagf_levels[XFS_BTNUM_RMAPi] + 1, - mp->m_rmap_maxlevels); + min_free += min_t(unsigned int, levels[XFS_BTNUM_RMAPi] + 1, + mp->m_rmap_maxlevels); return min_free; } diff --git a/fs/xfs/libxfs/xfs_bmap.c b/fs/xfs/libxfs/xfs_bmap.c index a9ad1f991ba3..4c2e046fbfad 100644 --- a/fs/xfs/libxfs/xfs_bmap.c +++ b/fs/xfs/libxfs/xfs_bmap.c @@ -4561,7 +4561,7 @@ xfs_bmapi_convert_delalloc( struct xfs_mount *mp = ip->i_mount; xfs_fileoff_t offset_fsb = XFS_B_TO_FSBT(mp, offset); struct xfs_bmalloca bma = { NULL }; - u16 flags = 0; + uint16_t flags = 0; struct xfs_trans *tp; int error; @@ -5972,8 +5972,7 @@ xfs_bmap_insert_extents( goto del_cursor; } - if (XFS_IS_CORRUPT(mp, - stop_fsb >= got.br_startoff + got.br_blockcount)) { + if (XFS_IS_CORRUPT(mp, stop_fsb > got.br_startoff)) { error = -EFSCORRUPTED; goto del_cursor; } diff --git a/fs/xfs/libxfs/xfs_dir2.c b/fs/xfs/libxfs/xfs_dir2.c index 0aa87cbde49e..dd6fcaaea318 100644 --- a/fs/xfs/libxfs/xfs_dir2.c +++ b/fs/xfs/libxfs/xfs_dir2.c @@ -724,3 +724,24 @@ xfs_dir2_namecheck( /* There shouldn't be any slashes or nulls here */ return !memchr(name, '/', length) && !memchr(name, 0, length); } + +xfs_dahash_t +xfs_dir2_hashname( + struct xfs_mount *mp, + struct xfs_name *name) +{ + if (unlikely(xfs_sb_version_hasasciici(&mp->m_sb))) + return xfs_ascii_ci_hashname(name); + return xfs_da_hashname(name->name, name->len); +} + +enum xfs_dacmp +xfs_dir2_compname( + struct xfs_da_args *args, + const unsigned char *name, + int len) +{ + if (unlikely(xfs_sb_version_hasasciici(&args->dp->i_mount->m_sb))) + return xfs_ascii_ci_compname(args, name, len); + return xfs_da_compname(args, name, len); +} diff --git a/fs/xfs/libxfs/xfs_dir2_priv.h b/fs/xfs/libxfs/xfs_dir2_priv.h index c031c53d0f0d..01ee0b926572 100644 --- a/fs/xfs/libxfs/xfs_dir2_priv.h +++ b/fs/xfs/libxfs/xfs_dir2_priv.h @@ -175,6 +175,12 @@ extern int xfs_dir2_sf_lookup(struct xfs_da_args *args); extern int xfs_dir2_sf_removename(struct xfs_da_args *args); extern int xfs_dir2_sf_replace(struct xfs_da_args *args); extern xfs_failaddr_t xfs_dir2_sf_verify(struct xfs_inode *ip); +int xfs_dir2_sf_entsize(struct xfs_mount *mp, + struct xfs_dir2_sf_hdr *hdr, int len); +void xfs_dir2_sf_put_ino(struct xfs_mount *mp, struct xfs_dir2_sf_hdr *hdr, + struct xfs_dir2_sf_entry *sfep, xfs_ino_t ino); +void xfs_dir2_sf_put_ftype(struct xfs_mount *mp, + struct xfs_dir2_sf_entry *sfep, uint8_t ftype); /* xfs_dir2_readdir.c */ extern int xfs_readdir(struct xfs_trans *tp, struct xfs_inode *dp, @@ -194,25 +200,8 @@ xfs_dir2_data_entsize( return round_up(len, XFS_DIR2_DATA_ALIGN); } -static inline xfs_dahash_t -xfs_dir2_hashname( - struct xfs_mount *mp, - struct xfs_name *name) -{ - if (unlikely(xfs_sb_version_hasasciici(&mp->m_sb))) - return xfs_ascii_ci_hashname(name); - return xfs_da_hashname(name->name, name->len); -} - -static inline enum xfs_dacmp -xfs_dir2_compname( - struct xfs_da_args *args, - const unsigned char *name, - int len) -{ - if (unlikely(xfs_sb_version_hasasciici(&args->dp->i_mount->m_sb))) - return xfs_ascii_ci_compname(args, name, len); - return xfs_da_compname(args, name, len); -} +xfs_dahash_t xfs_dir2_hashname(struct xfs_mount *mp, struct xfs_name *name); +enum xfs_dacmp xfs_dir2_compname(struct xfs_da_args *args, + const unsigned char *name, int len); #endif /* __XFS_DIR2_PRIV_H__ */ diff --git a/fs/xfs/libxfs/xfs_dir2_sf.c b/fs/xfs/libxfs/xfs_dir2_sf.c index 8b94d33d232f..7b7f6fb2ea3b 100644 --- a/fs/xfs/libxfs/xfs_dir2_sf.c +++ b/fs/xfs/libxfs/xfs_dir2_sf.c @@ -37,7 +37,7 @@ static void xfs_dir2_sf_check(xfs_da_args_t *args); static void xfs_dir2_sf_toino4(xfs_da_args_t *args); static void xfs_dir2_sf_toino8(xfs_da_args_t *args); -static int +int xfs_dir2_sf_entsize( struct xfs_mount *mp, struct xfs_dir2_sf_hdr *hdr, @@ -84,7 +84,7 @@ xfs_dir2_sf_get_ino( return get_unaligned_be64(from) & XFS_MAXINUMBER; } -static void +void xfs_dir2_sf_put_ino( struct xfs_mount *mp, struct xfs_dir2_sf_hdr *hdr, @@ -145,7 +145,7 @@ xfs_dir2_sf_get_ftype( return XFS_DIR3_FT_UNKNOWN; } -static void +void xfs_dir2_sf_put_ftype( struct xfs_mount *mp, struct xfs_dir2_sf_entry *sfep, diff --git a/fs/xfs/libxfs/xfs_ialloc.c b/fs/xfs/libxfs/xfs_ialloc.c index 988cde7744e6..5b759af4d165 100644 --- a/fs/xfs/libxfs/xfs_ialloc.c +++ b/fs/xfs/libxfs/xfs_ialloc.c @@ -2909,3 +2909,67 @@ xfs_ialloc_setup_geometry( else igeo->ialloc_align = 0; } + +/* Compute the location of the root directory inode that is laid out by mkfs. */ +xfs_ino_t +xfs_ialloc_calc_rootino( + struct xfs_mount *mp, + int sunit) +{ + struct xfs_ino_geometry *igeo = M_IGEO(mp); + xfs_agblock_t first_bno; + + /* + * Pre-calculate the geometry of AG 0. We know what it looks like + * because libxfs knows how to create allocation groups now. + * + * first_bno is the first block in which mkfs could possibly have + * allocated the root directory inode, once we factor in the metadata + * that mkfs formats before it. Namely, the four AG headers... + */ + first_bno = howmany(4 * mp->m_sb.sb_sectsize, mp->m_sb.sb_blocksize); + + /* ...the two free space btree roots... */ + first_bno += 2; + + /* ...the inode btree root... */ + first_bno += 1; + + /* ...the initial AGFL... */ + first_bno += xfs_alloc_min_freelist(mp, NULL); + + /* ...the free inode btree root... */ + if (xfs_sb_version_hasfinobt(&mp->m_sb)) + first_bno++; + + /* ...the reverse mapping btree root... */ + if (xfs_sb_version_hasrmapbt(&mp->m_sb)) + first_bno++; + + /* ...the reference count btree... */ + if (xfs_sb_version_hasreflink(&mp->m_sb)) + first_bno++; + + /* + * ...and the log, if it is allocated in the first allocation group. + * + * This can happen with filesystems that only have a single + * allocation group, or very odd geometries created by old mkfs + * versions on very small filesystems. + */ + if (mp->m_sb.sb_logstart && + XFS_FSB_TO_AGNO(mp, mp->m_sb.sb_logstart) == 0) + first_bno += mp->m_sb.sb_logblocks; + + /* + * Now round first_bno up to whatever allocation alignment is given + * by the filesystem or was passed in. + */ + if (xfs_sb_version_hasdalign(&mp->m_sb) && igeo->ialloc_align > 0) + first_bno = roundup(first_bno, sunit); + else if (xfs_sb_version_hasalign(&mp->m_sb) && + mp->m_sb.sb_inoalignmt > 1) + first_bno = roundup(first_bno, mp->m_sb.sb_inoalignmt); + + return XFS_AGINO_TO_INO(mp, 0, XFS_AGB_TO_AGINO(mp, first_bno)); +} diff --git a/fs/xfs/libxfs/xfs_ialloc.h b/fs/xfs/libxfs/xfs_ialloc.h index 323592d563d5..72b3468b97b1 100644 --- a/fs/xfs/libxfs/xfs_ialloc.h +++ b/fs/xfs/libxfs/xfs_ialloc.h @@ -152,5 +152,6 @@ int xfs_inobt_insert_rec(struct xfs_btree_cur *cur, uint16_t holemask, int xfs_ialloc_cluster_alignment(struct xfs_mount *mp); void xfs_ialloc_setup_geometry(struct xfs_mount *mp); +xfs_ino_t xfs_ialloc_calc_rootino(struct xfs_mount *mp, int sunit); #endif /* __XFS_IALLOC_H__ */ diff --git a/fs/xfs/libxfs/xfs_trans_resv.c b/fs/xfs/libxfs/xfs_trans_resv.c index c55cd9a3dec9..7a9c04920505 100644 --- a/fs/xfs/libxfs/xfs_trans_resv.c +++ b/fs/xfs/libxfs/xfs_trans_resv.c @@ -197,6 +197,24 @@ xfs_calc_inode_chunk_res( } /* + * Per-extent log reservation for the btree changes involved in freeing or + * allocating a realtime extent. We have to be able to log as many rtbitmap + * blocks as needed to mark inuse MAXEXTLEN blocks' worth of realtime extents, + * as well as the realtime summary block. + */ +static unsigned int +xfs_rtalloc_log_count( + struct xfs_mount *mp, + unsigned int num_ops) +{ + unsigned int blksz = XFS_FSB_TO_B(mp, 1); + unsigned int rtbmp_bytes; + + rtbmp_bytes = (MAXEXTLEN / mp->m_sb.sb_rextsize) / NBBY; + return (howmany(rtbmp_bytes, blksz) + 1) * num_ops; +} + +/* * Various log reservation values. * * These are based on the size of the file system block because that is what @@ -218,13 +236,21 @@ xfs_calc_inode_chunk_res( /* * In a write transaction we can allocate a maximum of 2 - * extents. This gives: + * extents. This gives (t1): * the inode getting the new extents: inode size * the inode's bmap btree: max depth * block size * the agfs of the ags from which the extents are allocated: 2 * sector * the superblock free block counter: sector size * the allocation btrees: 2 exts * 2 trees * (2 * max depth - 1) * block size - * And the bmap_finish transaction can free bmap blocks in a join: + * Or, if we're writing to a realtime file (t2): + * the inode getting the new extents: inode size + * the inode's bmap btree: max depth * block size + * the agfs of the ags from which the extents are allocated: 2 * sector + * the superblock free block counter: sector size + * the realtime bitmap: ((MAXEXTLEN / rtextsize) / NBBY) bytes + * the realtime summary: 1 block + * the allocation btrees: 2 trees * (2 * max depth - 1) * block size + * And the bmap_finish transaction can free bmap blocks in a join (t3): * the agfs of the ags containing the blocks: 2 * sector size * the agfls of the ags containing the blocks: 2 * sector size * the super block free block counter: sector size @@ -234,40 +260,72 @@ STATIC uint xfs_calc_write_reservation( struct xfs_mount *mp) { - return XFS_DQUOT_LOGRES(mp) + - max((xfs_calc_inode_res(mp, 1) + + unsigned int t1, t2, t3; + unsigned int blksz = XFS_FSB_TO_B(mp, 1); + + t1 = xfs_calc_inode_res(mp, 1) + + xfs_calc_buf_res(XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK), blksz) + + xfs_calc_buf_res(3, mp->m_sb.sb_sectsize) + + xfs_calc_buf_res(xfs_allocfree_log_count(mp, 2), blksz); + + if (xfs_sb_version_hasrealtime(&mp->m_sb)) { + t2 = xfs_calc_inode_res(mp, 1) + xfs_calc_buf_res(XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK), - XFS_FSB_TO_B(mp, 1)) + + blksz) + xfs_calc_buf_res(3, mp->m_sb.sb_sectsize) + - xfs_calc_buf_res(xfs_allocfree_log_count(mp, 2), - XFS_FSB_TO_B(mp, 1))), - (xfs_calc_buf_res(5, mp->m_sb.sb_sectsize) + - xfs_calc_buf_res(xfs_allocfree_log_count(mp, 2), - XFS_FSB_TO_B(mp, 1)))); + xfs_calc_buf_res(xfs_rtalloc_log_count(mp, 1), blksz) + + xfs_calc_buf_res(xfs_allocfree_log_count(mp, 1), blksz); + } else { + t2 = 0; + } + + t3 = xfs_calc_buf_res(5, mp->m_sb.sb_sectsize) + + xfs_calc_buf_res(xfs_allocfree_log_count(mp, 2), blksz); + + return XFS_DQUOT_LOGRES(mp) + max3(t1, t2, t3); } /* - * In truncating a file we free up to two extents at once. We can modify: + * In truncating a file we free up to two extents at once. We can modify (t1): * the inode being truncated: inode size * the inode's bmap btree: (max depth + 1) * block size - * And the bmap_finish transaction can free the blocks and bmap blocks: + * And the bmap_finish transaction can free the blocks and bmap blocks (t2): * the agf for each of the ags: 4 * sector size * the agfl for each of the ags: 4 * sector size * the super block to reflect the freed blocks: sector size * worst case split in allocation btrees per extent assuming 4 extents: * 4 exts * 2 trees * (2 * max depth - 1) * block size + * Or, if it's a realtime file (t3): + * the agf for each of the ags: 2 * sector size + * the agfl for each of the ags: 2 * sector size + * the super block to reflect the freed blocks: sector size + * the realtime bitmap: 2 exts * ((MAXEXTLEN / rtextsize) / NBBY) bytes + * the realtime summary: 2 exts * 1 block + * worst case split in allocation btrees per extent assuming 2 extents: + * 2 exts * 2 trees * (2 * max depth - 1) * block size */ STATIC uint xfs_calc_itruncate_reservation( struct xfs_mount *mp) { - return XFS_DQUOT_LOGRES(mp) + - max((xfs_calc_inode_res(mp, 1) + - xfs_calc_buf_res(XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK) + 1, - XFS_FSB_TO_B(mp, 1))), - (xfs_calc_buf_res(9, mp->m_sb.sb_sectsize) + - xfs_calc_buf_res(xfs_allocfree_log_count(mp, 4), - XFS_FSB_TO_B(mp, 1)))); + unsigned int t1, t2, t3; + unsigned int blksz = XFS_FSB_TO_B(mp, 1); + + t1 = xfs_calc_inode_res(mp, 1) + + xfs_calc_buf_res(XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK) + 1, blksz); + + t2 = xfs_calc_buf_res(9, mp->m_sb.sb_sectsize) + + xfs_calc_buf_res(xfs_allocfree_log_count(mp, 4), blksz); + + if (xfs_sb_version_hasrealtime(&mp->m_sb)) { + t3 = xfs_calc_buf_res(5, mp->m_sb.sb_sectsize) + + xfs_calc_buf_res(xfs_rtalloc_log_count(mp, 2), blksz) + + xfs_calc_buf_res(xfs_allocfree_log_count(mp, 2), blksz); + } else { + t3 = 0; + } + + return XFS_DQUOT_LOGRES(mp) + max3(t1, t2, t3); } /* diff --git a/fs/xfs/scrub/trace.h b/fs/xfs/scrub/trace.h index 3362bae28b46..096203119934 100644 --- a/fs/xfs/scrub/trace.h +++ b/fs/xfs/scrub/trace.h @@ -329,7 +329,7 @@ TRACE_EVENT(xchk_btree_op_error, __field(int, level) __field(xfs_agnumber_t, agno) __field(xfs_agblock_t, bno) - __field(int, ptr); + __field(int, ptr) __field(int, error) __field(void *, ret_ip) ), @@ -414,7 +414,7 @@ TRACE_EVENT(xchk_btree_error, __field(int, level) __field(xfs_agnumber_t, agno) __field(xfs_agblock_t, bno) - __field(int, ptr); + __field(int, ptr) __field(void *, ret_ip) ), TP_fast_assign( @@ -452,7 +452,7 @@ TRACE_EVENT(xchk_ifork_btree_error, __field(int, level) __field(xfs_agnumber_t, agno) __field(xfs_agblock_t, bno) - __field(int, ptr); + __field(int, ptr) __field(void *, ret_ip) ), TP_fast_assign( diff --git a/fs/xfs/xfs_bmap_util.c b/fs/xfs/xfs_bmap_util.c index 2efd78a9719e..e62fb5216341 100644 --- a/fs/xfs/xfs_bmap_util.c +++ b/fs/xfs/xfs_bmap_util.c @@ -992,6 +992,7 @@ xfs_prepare_shift( struct xfs_inode *ip, loff_t offset) { + struct xfs_mount *mp = ip->i_mount; int error; /* @@ -1005,6 +1006,17 @@ xfs_prepare_shift( } /* + * Shift operations must stabilize the start block offset boundary along + * with the full range of the operation. If we don't, a COW writeback + * completion could race with an insert, front merge with the start + * extent (after split) during the shift and corrupt the file. Start + * with the block just prior to the start to stabilize the boundary. + */ + offset = round_down(offset, 1 << mp->m_sb.sb_blocklog); + if (offset) + offset -= (1 << mp->m_sb.sb_blocklog); + + /* * Writeback and invalidate cache for the remainder of the file as we're * about to shift down every extent from offset to EOF. */ diff --git a/fs/xfs/xfs_buf_item.c b/fs/xfs/xfs_buf_item.c index 3458a1264a3f..3984779e5911 100644 --- a/fs/xfs/xfs_buf_item.c +++ b/fs/xfs/xfs_buf_item.c @@ -956,7 +956,7 @@ xfs_buf_item_relse( struct xfs_buf_log_item *bip = bp->b_log_item; trace_xfs_buf_item_relse(bp, _RET_IP_); - ASSERT(!(bip->bli_item.li_flags & XFS_LI_IN_AIL)); + ASSERT(!test_bit(XFS_LI_IN_AIL, &bip->bli_item.li_flags)); bp->b_log_item = NULL; if (list_empty(&bp->b_li_list)) diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c index fca65109cf24..56efe140c923 100644 --- a/fs/xfs/xfs_mount.c +++ b/fs/xfs/xfs_mount.c @@ -31,7 +31,7 @@ #include "xfs_reflink.h" #include "xfs_extent_busy.h" #include "xfs_health.h" - +#include "xfs_trace.h" static DEFINE_MUTEX(xfs_uuid_table_mutex); static int xfs_uuid_table_size; @@ -360,66 +360,119 @@ release_buf: } /* - * Update alignment values based on mount options and sb values + * If the sunit/swidth change would move the precomputed root inode value, we + * must reject the ondisk change because repair will stumble over that. + * However, we allow the mount to proceed because we never rejected this + * combination before. Returns true to update the sb, false otherwise. + */ +static inline int +xfs_check_new_dalign( + struct xfs_mount *mp, + int new_dalign, + bool *update_sb) +{ + struct xfs_sb *sbp = &mp->m_sb; + xfs_ino_t calc_ino; + + calc_ino = xfs_ialloc_calc_rootino(mp, new_dalign); + trace_xfs_check_new_dalign(mp, new_dalign, calc_ino); + + if (sbp->sb_rootino == calc_ino) { + *update_sb = true; + return 0; + } + + xfs_warn(mp, +"Cannot change stripe alignment; would require moving root inode."); + + /* + * XXX: Next time we add a new incompat feature, this should start + * returning -EINVAL to fail the mount. Until then, spit out a warning + * that we're ignoring the administrator's instructions. + */ + xfs_warn(mp, "Skipping superblock stripe alignment update."); + *update_sb = false; + return 0; +} + +/* + * If we were provided with new sunit/swidth values as mount options, make sure + * that they pass basic alignment and superblock feature checks, and convert + * them into the same units (FSB) that everything else expects. This step + * /must/ be done before computing the inode geometry. */ STATIC int -xfs_update_alignment(xfs_mount_t *mp) +xfs_validate_new_dalign( + struct xfs_mount *mp) { - xfs_sb_t *sbp = &(mp->m_sb); + if (mp->m_dalign == 0) + return 0; - if (mp->m_dalign) { + /* + * If stripe unit and stripe width are not multiples + * of the fs blocksize turn off alignment. + */ + if ((BBTOB(mp->m_dalign) & mp->m_blockmask) || + (BBTOB(mp->m_swidth) & mp->m_blockmask)) { + xfs_warn(mp, + "alignment check failed: sunit/swidth vs. blocksize(%d)", + mp->m_sb.sb_blocksize); + return -EINVAL; + } else { /* - * If stripe unit and stripe width are not multiples - * of the fs blocksize turn off alignment. + * Convert the stripe unit and width to FSBs. */ - if ((BBTOB(mp->m_dalign) & mp->m_blockmask) || - (BBTOB(mp->m_swidth) & mp->m_blockmask)) { + mp->m_dalign = XFS_BB_TO_FSBT(mp, mp->m_dalign); + if (mp->m_dalign && (mp->m_sb.sb_agblocks % mp->m_dalign)) { xfs_warn(mp, - "alignment check failed: sunit/swidth vs. blocksize(%d)", - sbp->sb_blocksize); + "alignment check failed: sunit/swidth vs. agsize(%d)", + mp->m_sb.sb_agblocks); return -EINVAL; - } else { - /* - * Convert the stripe unit and width to FSBs. - */ - mp->m_dalign = XFS_BB_TO_FSBT(mp, mp->m_dalign); - if (mp->m_dalign && (sbp->sb_agblocks % mp->m_dalign)) { - xfs_warn(mp, - "alignment check failed: sunit/swidth vs. agsize(%d)", - sbp->sb_agblocks); - return -EINVAL; - } else if (mp->m_dalign) { - mp->m_swidth = XFS_BB_TO_FSBT(mp, mp->m_swidth); - } else { - xfs_warn(mp, - "alignment check failed: sunit(%d) less than bsize(%d)", - mp->m_dalign, sbp->sb_blocksize); - return -EINVAL; - } - } - - /* - * Update superblock with new values - * and log changes - */ - if (xfs_sb_version_hasdalign(sbp)) { - if (sbp->sb_unit != mp->m_dalign) { - sbp->sb_unit = mp->m_dalign; - mp->m_update_sb = true; - } - if (sbp->sb_width != mp->m_swidth) { - sbp->sb_width = mp->m_swidth; - mp->m_update_sb = true; - } + } else if (mp->m_dalign) { + mp->m_swidth = XFS_BB_TO_FSBT(mp, mp->m_swidth); } else { xfs_warn(mp, - "cannot change alignment: superblock does not support data alignment"); + "alignment check failed: sunit(%d) less than bsize(%d)", + mp->m_dalign, mp->m_sb.sb_blocksize); return -EINVAL; } + } + + if (!xfs_sb_version_hasdalign(&mp->m_sb)) { + xfs_warn(mp, +"cannot change alignment: superblock does not support data alignment"); + return -EINVAL; + } + + return 0; +} + +/* Update alignment values based on mount options and sb values. */ +STATIC int +xfs_update_alignment( + struct xfs_mount *mp) +{ + struct xfs_sb *sbp = &mp->m_sb; + + if (mp->m_dalign) { + bool update_sb; + int error; + + if (sbp->sb_unit == mp->m_dalign && + sbp->sb_width == mp->m_swidth) + return 0; + + error = xfs_check_new_dalign(mp, mp->m_dalign, &update_sb); + if (error || !update_sb) + return error; + + sbp->sb_unit = mp->m_dalign; + sbp->sb_width = mp->m_swidth; + mp->m_update_sb = true; } else if ((mp->m_flags & XFS_MOUNT_NOALIGN) != XFS_MOUNT_NOALIGN && xfs_sb_version_hasdalign(&mp->m_sb)) { - mp->m_dalign = sbp->sb_unit; - mp->m_swidth = sbp->sb_width; + mp->m_dalign = sbp->sb_unit; + mp->m_swidth = sbp->sb_width; } return 0; @@ -648,12 +701,12 @@ xfs_mountfs( } /* - * Check if sb_agblocks is aligned at stripe boundary - * If sb_agblocks is NOT aligned turn off m_dalign since - * allocator alignment is within an ag, therefore ag has - * to be aligned at stripe boundary. + * If we were given new sunit/swidth options, do some basic validation + * checks and convert the incore dalign and swidth values to the + * same units (FSB) that everything else uses. This /must/ happen + * before computing the inode geometry. */ - error = xfs_update_alignment(mp); + error = xfs_validate_new_dalign(mp); if (error) goto out; @@ -664,6 +717,17 @@ xfs_mountfs( xfs_rmapbt_compute_maxlevels(mp); xfs_refcountbt_compute_maxlevels(mp); + /* + * Check if sb_agblocks is aligned at stripe boundary. If sb_agblocks + * is NOT aligned turn off m_dalign since allocator alignment is within + * an ag, therefore ag has to be aligned at stripe boundary. Note that + * we must compute the free space and rmap btree geometry before doing + * this. + */ + error = xfs_update_alignment(mp); + if (error) + goto out; + /* enable fail_at_unmount as default */ mp->m_fail_unmount = true; diff --git a/fs/xfs/xfs_trace.h b/fs/xfs/xfs_trace.h index c13bb3655e48..e242988f57fb 100644 --- a/fs/xfs/xfs_trace.h +++ b/fs/xfs/xfs_trace.h @@ -218,8 +218,8 @@ DECLARE_EVENT_CLASS(xfs_bmap_class, TP_STRUCT__entry( __field(dev_t, dev) __field(xfs_ino_t, ino) - __field(void *, leaf); - __field(int, pos); + __field(void *, leaf) + __field(int, pos) __field(xfs_fileoff_t, startoff) __field(xfs_fsblock_t, startblock) __field(xfs_filblks_t, blockcount) @@ -3573,6 +3573,27 @@ DEFINE_KMEM_EVENT(kmem_alloc_large); DEFINE_KMEM_EVENT(kmem_realloc); DEFINE_KMEM_EVENT(kmem_zone_alloc); +TRACE_EVENT(xfs_check_new_dalign, + TP_PROTO(struct xfs_mount *mp, int new_dalign, xfs_ino_t calc_rootino), + TP_ARGS(mp, new_dalign, calc_rootino), + TP_STRUCT__entry( + __field(dev_t, dev) + __field(int, new_dalign) + __field(xfs_ino_t, sb_rootino) + __field(xfs_ino_t, calc_rootino) + ), + TP_fast_assign( + __entry->dev = mp->m_super->s_dev; + __entry->new_dalign = new_dalign; + __entry->sb_rootino = mp->m_sb.sb_rootino; + __entry->calc_rootino = calc_rootino; + ), + TP_printk("dev %d:%d new_dalign %d sb_rootino %llu calc_rootino %llu", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->new_dalign, __entry->sb_rootino, + __entry->calc_rootino) +) + #endif /* _TRACE_XFS_H */ #undef TRACE_INCLUDE_PATH |