From c15271f4e74cd6dbdf461335d6d1450949c4b956 Mon Sep 17 00:00:00 2001 From: Jan Blunck Date: Wed, 14 Apr 2010 14:38:38 +0200 Subject: ext2: Add ext2_sb_info s_lock spinlock Add a spinlock that protects against concurrent modifications of s_mount_state, s_blocks_last, s_overhead_last and the content of the superblock's buffer pointed to by sbi->s_es. The spinlock is now used in ext2_xattr_update_super_block() which was setting the EXT2_FEATURE_COMPAT_EXT_ATTR flag on the superblock without protection before. Likewise the spinlock is used in ext2_show_options() to have a consistent view of the mount options. This is a preparation patch for removing the BKL from ext2 in the next patch. Signed-off-by: Jan Blunck Cc: Andi Kleen Cc: Jan Kara Cc: OGAWA Hirofumi Signed-off-by: Jan Kara --- include/linux/ext2_fs_sb.h | 9 +++++++++ 1 file changed, 9 insertions(+) (limited to 'include') diff --git a/include/linux/ext2_fs_sb.h b/include/linux/ext2_fs_sb.h index 1cdb66367c98..db4d9f586bb6 100644 --- a/include/linux/ext2_fs_sb.h +++ b/include/linux/ext2_fs_sb.h @@ -106,6 +106,15 @@ struct ext2_sb_info { spinlock_t s_rsv_window_lock; struct rb_root s_rsv_window_root; struct ext2_reserve_window_node s_rsv_window_head; + /* + * s_lock protects against concurrent modifications of s_mount_state, + * s_blocks_last, s_overhead_last and the content of superblock's + * buffer pointed to by sbi->s_es. + * + * Note: It is used in ext2_show_options() to provide a consistent view + * of the mount options. + */ + spinlock_t s_lock; }; static inline spinlock_t * -- cgit v1.2.3 From 03f4d804a1b4748885dc4613a4afe10089a731c8 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Thu, 15 Apr 2010 22:16:24 +0200 Subject: jbd: Provide function to check whether transaction will issue data barrier Provide a function which returns whether a transaction with given tid will send a barrier to the filesystem device. The function will be used by ext3 to detect whether fsync needs to send a separate barrier or not. Signed-off-by: Jan Kara --- fs/jbd/commit.c | 8 +++++++- fs/jbd/journal.c | 33 +++++++++++++++++++++++++++++++++ include/linux/jbd.h | 3 ++- 3 files changed, 42 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/fs/jbd/commit.c b/fs/jbd/commit.c index ecb44c94ba8d..28a9ddaa0c49 100644 --- a/fs/jbd/commit.c +++ b/fs/jbd/commit.c @@ -786,6 +786,12 @@ wait_for_iobuf: jbd_debug(3, "JBD: commit phase 6\n"); + /* All metadata is written, now write commit record and do cleanup */ + spin_lock(&journal->j_state_lock); + J_ASSERT(commit_transaction->t_state == T_COMMIT); + commit_transaction->t_state = T_COMMIT_RECORD; + spin_unlock(&journal->j_state_lock); + if (journal_write_commit_record(journal, commit_transaction)) err = -EIO; @@ -923,7 +929,7 @@ restart_loop: jbd_debug(3, "JBD: commit phase 8\n"); - J_ASSERT(commit_transaction->t_state == T_COMMIT); + J_ASSERT(commit_transaction->t_state == T_COMMIT_RECORD); commit_transaction->t_state = T_FINISHED; J_ASSERT(commit_transaction == journal->j_committing_transaction); diff --git a/fs/jbd/journal.c b/fs/jbd/journal.c index bd224eec9b07..99c71940155a 100644 --- a/fs/jbd/journal.c +++ b/fs/jbd/journal.c @@ -564,6 +564,38 @@ int log_wait_commit(journal_t *journal, tid_t tid) return err; } +/* + * Return 1 if a given transaction has not yet sent barrier request + * connected with a transaction commit. If 0 is returned, transaction + * may or may not have sent the barrier. Used to avoid sending barrier + * twice in common cases. + */ +int journal_trans_will_send_data_barrier(journal_t *journal, tid_t tid) +{ + int ret = 0; + transaction_t *commit_trans; + + if (!(journal->j_flags & JFS_BARRIER)) + return 0; + spin_lock(&journal->j_state_lock); + /* Transaction already committed? */ + if (tid_geq(journal->j_commit_sequence, tid)) + goto out; + /* + * Transaction is being committed and we already proceeded to + * writing commit record? + */ + commit_trans = journal->j_committing_transaction; + if (commit_trans && commit_trans->t_tid == tid && + commit_trans->t_state >= T_COMMIT_RECORD) + goto out; + ret = 1; +out: + spin_unlock(&journal->j_state_lock); + return ret; +} +EXPORT_SYMBOL(journal_commit_will_send_barrier); + /* * Log buffer allocation routines: */ @@ -1157,6 +1189,7 @@ int journal_destroy(journal_t *journal) { int err = 0; + /* Wait for the commit thread to wake up and die. */ journal_kill_thread(journal); diff --git a/include/linux/jbd.h b/include/linux/jbd.h index 516a2a27e87a..e06965081ba5 100644 --- a/include/linux/jbd.h +++ b/include/linux/jbd.h @@ -427,9 +427,9 @@ struct transaction_s enum { T_RUNNING, T_LOCKED, - T_RUNDOWN, T_FLUSH, T_COMMIT, + T_COMMIT_RECORD, T_FINISHED } t_state; @@ -991,6 +991,7 @@ int journal_start_commit(journal_t *journal, tid_t *tid); int journal_force_commit_nested(journal_t *journal); int log_wait_commit(journal_t *journal, tid_t tid); int log_do_checkpoint(journal_t *journal); +int journal_trans_will_send_data_barrier(journal_t *journal, tid_t tid); void __log_wait_for_space(journal_t *journal); extern void __journal_drop_transaction(journal_t *, transaction_t *); -- cgit v1.2.3 From dde9588853b1bde542eab247f8838c472806688f Mon Sep 17 00:00:00 2001 From: Dmitry Monakhov Date: Mon, 26 Apr 2010 20:03:33 +0400 Subject: quota: Make quota stat accounting lockless. Quota stats is mostly writable data structure. Let's alloc percpu bucket for each value. NOTE: dqstats_read() function is racy against dqstats_{inc,dec} and may return inconsistent value. But this is ok since absolute accuracy is not required. Signed-off-by: Dmitry Monakhov Signed-off-by: Jan Kara --- fs/quota/dquot.c | 102 +++++++++++++++++++++++++++++++++----------------- fs/quota/quota_tree.c | 4 +- fs/quota/quota_v1.c | 4 +- include/linux/quota.h | 42 +++++++++++++++++---- 4 files changed, 106 insertions(+), 46 deletions(-) (limited to 'include') diff --git a/fs/quota/dquot.c b/fs/quota/dquot.c index ae766056350d..01347e81d0ca 100644 --- a/fs/quota/dquot.c +++ b/fs/quota/dquot.c @@ -82,7 +82,7 @@ /* * There are three quota SMP locks. dq_list_lock protects all lists with quotas - * and quota formats, dqstats structure containing statistics about the lists + * and quota formats. * dq_data_lock protects data from dq_dqb and also mem_dqinfo structures and * also guards consistency of dquot->dq_dqb with inode->i_blocks, i_bytes. * i_blocks and i_bytes updates itself are guarded by i_lock acquired directly @@ -228,6 +228,10 @@ static struct hlist_head *dquot_hash; struct dqstats dqstats; EXPORT_SYMBOL(dqstats); +#ifdef CONFIG_SMP +struct dqstats *dqstats_pcpu; +EXPORT_SYMBOL(dqstats_pcpu); +#endif static qsize_t inode_get_rsv_space(struct inode *inode); static void __dquot_initialize(struct inode *inode, int type); @@ -275,7 +279,7 @@ static struct dquot *find_dquot(unsigned int hashent, struct super_block *sb, static inline void put_dquot_last(struct dquot *dquot) { list_add_tail(&dquot->dq_free, &free_dquots); - dqstats.free_dquots++; + dqstats_inc(DQST_FREE_DQUOTS); } static inline void remove_free_dquot(struct dquot *dquot) @@ -283,7 +287,7 @@ static inline void remove_free_dquot(struct dquot *dquot) if (list_empty(&dquot->dq_free)) return; list_del_init(&dquot->dq_free); - dqstats.free_dquots--; + dqstats_dec(DQST_FREE_DQUOTS); } static inline void put_inuse(struct dquot *dquot) @@ -291,12 +295,12 @@ static inline void put_inuse(struct dquot *dquot) /* We add to the back of inuse list so we don't have to restart * when traversing this list and we block */ list_add_tail(&dquot->dq_inuse, &inuse_list); - dqstats.allocated_dquots++; + dqstats_inc(DQST_ALLOC_DQUOTS); } static inline void remove_inuse(struct dquot *dquot) { - dqstats.allocated_dquots--; + dqstats_dec(DQST_ALLOC_DQUOTS); list_del(&dquot->dq_inuse); } /* @@ -561,8 +565,8 @@ int dquot_scan_active(struct super_block *sb, continue; /* Now we have active dquot so we can just increase use count */ atomic_inc(&dquot->dq_count); - dqstats.lookups++; spin_unlock(&dq_list_lock); + dqstats_inc(DQST_LOOKUPS); dqput(old_dquot); old_dquot = dquot; ret = fn(dquot, priv); @@ -607,8 +611,8 @@ int vfs_quota_sync(struct super_block *sb, int type, int wait) * holding reference so we can safely just increase * use count */ atomic_inc(&dquot->dq_count); - dqstats.lookups++; spin_unlock(&dq_list_lock); + dqstats_inc(DQST_LOOKUPS); sb->dq_op->write_dquot(dquot); dqput(dquot); spin_lock(&dq_list_lock); @@ -620,9 +624,7 @@ int vfs_quota_sync(struct super_block *sb, int type, int wait) if ((cnt == type || type == -1) && sb_has_quota_active(sb, cnt) && info_dirty(&dqopt->info[cnt])) sb->dq_op->write_info(sb, cnt); - spin_lock(&dq_list_lock); - dqstats.syncs++; - spin_unlock(&dq_list_lock); + dqstats_inc(DQST_SYNCS); mutex_unlock(&dqopt->dqonoff_mutex); if (!wait || (sb_dqopt(sb)->flags & DQUOT_QUOTA_SYS_FILE)) @@ -674,6 +676,22 @@ static void prune_dqcache(int count) } } +static int dqstats_read(unsigned int type) +{ + int count = 0; +#ifdef CONFIG_SMP + int cpu; + for_each_possible_cpu(cpu) + count += per_cpu_ptr(dqstats_pcpu, cpu)->stat[type]; + /* Statistics reading is racy, but absolute accuracy isn't required */ + if (count < 0) + count = 0; +#else + count = dqstats.stat[type]; +#endif + return count; +} + /* * This is called from kswapd when we think we need some * more memory @@ -686,7 +704,7 @@ static int shrink_dqcache_memory(int nr, gfp_t gfp_mask) prune_dqcache(nr); spin_unlock(&dq_list_lock); } - return (dqstats.free_dquots / 100) * sysctl_vfs_cache_pressure; + return (dqstats_read(DQST_FREE_DQUOTS)/100) * sysctl_vfs_cache_pressure; } static struct shrinker dqcache_shrinker = { @@ -714,10 +732,7 @@ void dqput(struct dquot *dquot) BUG(); } #endif - - spin_lock(&dq_list_lock); - dqstats.drops++; - spin_unlock(&dq_list_lock); + dqstats_inc(DQST_DROPS); we_slept: spin_lock(&dq_list_lock); if (atomic_read(&dquot->dq_count) > 1) { @@ -834,15 +849,15 @@ we_slept: put_inuse(dquot); /* hash it first so it can be found */ insert_dquot_hash(dquot); - dqstats.lookups++; spin_unlock(&dq_list_lock); + dqstats_inc(DQST_LOOKUPS); } else { if (!atomic_read(&dquot->dq_count)) remove_free_dquot(dquot); atomic_inc(&dquot->dq_count); - dqstats.cache_hits++; - dqstats.lookups++; spin_unlock(&dq_list_lock); + dqstats_inc(DQST_CACHE_HITS); + dqstats_inc(DQST_LOOKUPS); } /* Wait for dq_lock - after this we know that either dquot_release() is * already finished or it will be canceled due to dq_count > 1 test */ @@ -2476,62 +2491,74 @@ const struct quotactl_ops vfs_quotactl_ops = { .set_dqblk = vfs_set_dqblk }; + +static int do_proc_dqstats(struct ctl_table *table, int write, + void __user *buffer, size_t *lenp, loff_t *ppos) +{ +#ifdef CONFIG_SMP + /* Update global table */ + unsigned int type = (int *)table->data - dqstats.stat; + dqstats.stat[type] = dqstats_read(type); +#endif + return proc_dointvec(table, write, buffer, lenp, ppos); +} + static ctl_table fs_dqstats_table[] = { { .procname = "lookups", - .data = &dqstats.lookups, + .data = &dqstats.stat[DQST_LOOKUPS], .maxlen = sizeof(int), .mode = 0444, - .proc_handler = proc_dointvec, + .proc_handler = do_proc_dqstats, }, { .procname = "drops", - .data = &dqstats.drops, + .data = &dqstats.stat[DQST_DROPS], .maxlen = sizeof(int), .mode = 0444, - .proc_handler = proc_dointvec, + .proc_handler = do_proc_dqstats, }, { .procname = "reads", - .data = &dqstats.reads, + .data = &dqstats.stat[DQST_READS], .maxlen = sizeof(int), .mode = 0444, - .proc_handler = proc_dointvec, + .proc_handler = do_proc_dqstats, }, { .procname = "writes", - .data = &dqstats.writes, + .data = &dqstats.stat[DQST_WRITES], .maxlen = sizeof(int), .mode = 0444, - .proc_handler = proc_dointvec, + .proc_handler = do_proc_dqstats, }, { .procname = "cache_hits", - .data = &dqstats.cache_hits, + .data = &dqstats.stat[DQST_CACHE_HITS], .maxlen = sizeof(int), .mode = 0444, - .proc_handler = proc_dointvec, + .proc_handler = do_proc_dqstats, }, { .procname = "allocated_dquots", - .data = &dqstats.allocated_dquots, + .data = &dqstats.stat[DQST_ALLOC_DQUOTS], .maxlen = sizeof(int), .mode = 0444, - .proc_handler = proc_dointvec, + .proc_handler = do_proc_dqstats, }, { .procname = "free_dquots", - .data = &dqstats.free_dquots, + .data = &dqstats.stat[DQST_FREE_DQUOTS], .maxlen = sizeof(int), .mode = 0444, - .proc_handler = proc_dointvec, + .proc_handler = do_proc_dqstats, }, { .procname = "syncs", - .data = &dqstats.syncs, + .data = &dqstats.stat[DQST_SYNCS], .maxlen = sizeof(int), .mode = 0444, - .proc_handler = proc_dointvec, + .proc_handler = do_proc_dqstats, }, #ifdef CONFIG_PRINT_QUOTA_WARNING { @@ -2583,6 +2610,13 @@ static int __init dquot_init(void) if (!dquot_hash) panic("Cannot create dquot hash table"); +#ifdef CONFIG_SMP + dqstats_pcpu = alloc_percpu(struct dqstats); + if (!dqstats_pcpu) + panic("Cannot create dquot stats table"); +#endif + memset(&dqstats, 0, sizeof(struct dqstats)); + /* Find power-of-two hlist_heads which can fit into allocation */ nr_hash = (1UL << order) * PAGE_SIZE / sizeof(struct hlist_head); dq_hash_bits = 0; diff --git a/fs/quota/quota_tree.c b/fs/quota/quota_tree.c index f81f4bcfb178..5b7f7416ec7a 100644 --- a/fs/quota/quota_tree.c +++ b/fs/quota/quota_tree.c @@ -384,7 +384,7 @@ int qtree_write_dquot(struct qtree_mem_dqinfo *info, struct dquot *dquot) } else { ret = 0; } - dqstats.writes++; + dqstats_inc(DQST_WRITES); kfree(ddquot); return ret; @@ -634,7 +634,7 @@ int qtree_read_dquot(struct qtree_mem_dqinfo *info, struct dquot *dquot) spin_unlock(&dq_data_lock); kfree(ddquot); out: - dqstats.reads++; + dqstats_inc(DQST_READS); return ret; } EXPORT_SYMBOL(qtree_read_dquot); diff --git a/fs/quota/quota_v1.c b/fs/quota/quota_v1.c index 2ae757e9c008..4af344c5852a 100644 --- a/fs/quota/quota_v1.c +++ b/fs/quota/quota_v1.c @@ -71,7 +71,7 @@ static int v1_read_dqblk(struct dquot *dquot) dquot->dq_dqb.dqb_ihardlimit == 0 && dquot->dq_dqb.dqb_isoftlimit == 0) set_bit(DQ_FAKE_B, &dquot->dq_flags); - dqstats.reads++; + dqstats_inc(DQST_READS); return 0; } @@ -104,7 +104,7 @@ static int v1_commit_dqblk(struct dquot *dquot) ret = 0; out: - dqstats.writes++; + dqstats_inc(DQST_WRITES); return ret; } diff --git a/include/linux/quota.h b/include/linux/quota.h index b462916b2a0a..cdfde10481b7 100644 --- a/include/linux/quota.h +++ b/include/linux/quota.h @@ -174,6 +174,8 @@ enum { #include #include #include +#include +#include #include #include @@ -238,19 +240,43 @@ static inline int info_dirty(struct mem_dqinfo *info) return test_bit(DQF_INFO_DIRTY_B, &info->dqi_flags); } +enum { + DQST_LOOKUPS, + DQST_DROPS, + DQST_READS, + DQST_WRITES, + DQST_CACHE_HITS, + DQST_ALLOC_DQUOTS, + DQST_FREE_DQUOTS, + DQST_SYNCS, + _DQST_DQSTAT_LAST +}; + struct dqstats { - int lookups; - int drops; - int reads; - int writes; - int cache_hits; - int allocated_dquots; - int free_dquots; - int syncs; + int stat[_DQST_DQSTAT_LAST]; }; +extern struct dqstats *dqstats_pcpu; extern struct dqstats dqstats; +static inline void dqstats_inc(unsigned int type) +{ +#ifdef CONFIG_SMP + per_cpu_ptr(dqstats_pcpu, smp_processor_id())->stat[type]++; +#else + dqstats.stat[type]++; +#endif +} + +static inline void dqstats_dec(unsigned int type) +{ +#ifdef CONFIG_SMP + per_cpu_ptr(dqstats_pcpu, smp_processor_id())->stat[type]--; +#else + dqstats.stat[type]--; +#endif +} + #define DQ_MOD_B 0 /* dquot modified since read */ #define DQ_BLKS_B 1 /* uid/gid has been warned about blk limit */ #define DQ_INODES_B 2 /* uid/gid has been warned about inode limit */ -- cgit v1.2.3 From b9b2dd36c1bc64430f8e13990ab135cbecc10076 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Thu, 6 May 2010 17:04:58 -0400 Subject: quota: unify ->get_dqblk Pass the larger struct fs_disk_quota to the ->get_dqblk operation so that the Q_GETQUOTA and Q_XGETQUOTA operations can be implemented with a single filesystem operation and we can retire the ->get_xquota operation. The additional information (RT-subvolume accounting and warn counts) are left zero for the VFS quota implementation. Signed-off-by: Christoph Hellwig Signed-off-by: Jan Kara --- fs/gfs2/quota.c | 6 +++--- fs/quota/dquot.c | 27 ++++++++++++++++----------- fs/quota/quota.c | 23 ++++++++++++++++++----- fs/xfs/linux-2.6/xfs_quotaops.c | 4 ++-- include/linux/quota.h | 3 +-- include/linux/quotaops.h | 3 ++- 6 files changed, 42 insertions(+), 24 deletions(-) (limited to 'include') diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c index d5f4661287f9..dec93577a783 100644 --- a/fs/gfs2/quota.c +++ b/fs/gfs2/quota.c @@ -1476,8 +1476,8 @@ static int gfs2_quota_get_xstate(struct super_block *sb, return 0; } -static int gfs2_xquota_get(struct super_block *sb, int type, qid_t id, - struct fs_disk_quota *fdq) +static int gfs2_get_dqblk(struct super_block *sb, int type, qid_t id, + struct fs_disk_quota *fdq) { struct gfs2_sbd *sdp = sb->s_fs_info; struct gfs2_quota_lvb *qlvb; @@ -1629,7 +1629,7 @@ out_put: const struct quotactl_ops gfs2_quotactl_ops = { .quota_sync = gfs2_quota_sync, .get_xstate = gfs2_quota_get_xstate, - .get_xquota = gfs2_xquota_get, + .get_dqblk = gfs2_get_dqblk, .set_xquota = gfs2_xquota_set, }; diff --git a/fs/quota/dquot.c b/fs/quota/dquot.c index 01347e81d0ca..6aed77fc99c7 100644 --- a/fs/quota/dquot.c +++ b/fs/quota/dquot.c @@ -2301,25 +2301,30 @@ static inline qsize_t stoqb(qsize_t space) } /* Generic routine for getting common part of quota structure */ -static void do_get_dqblk(struct dquot *dquot, struct if_dqblk *di) +static void do_get_dqblk(struct dquot *dquot, struct fs_disk_quota *di) { struct mem_dqblk *dm = &dquot->dq_dqb; + memset(di, 0, sizeof(*di)); + di->d_version = FS_DQUOT_VERSION; + di->d_flags = dquot->dq_type == USRQUOTA ? + XFS_USER_QUOTA : XFS_GROUP_QUOTA; + di->d_id = dquot->dq_id; + spin_lock(&dq_data_lock); - di->dqb_bhardlimit = stoqb(dm->dqb_bhardlimit); - di->dqb_bsoftlimit = stoqb(dm->dqb_bsoftlimit); - di->dqb_curspace = dm->dqb_curspace + dm->dqb_rsvspace; - di->dqb_ihardlimit = dm->dqb_ihardlimit; - di->dqb_isoftlimit = dm->dqb_isoftlimit; - di->dqb_curinodes = dm->dqb_curinodes; - di->dqb_btime = dm->dqb_btime; - di->dqb_itime = dm->dqb_itime; - di->dqb_valid = QIF_ALL; + di->d_blk_hardlimit = stoqb(dm->dqb_bhardlimit); + di->d_blk_softlimit = stoqb(dm->dqb_bsoftlimit); + di->d_ino_hardlimit = dm->dqb_ihardlimit; + di->d_ino_softlimit = dm->dqb_isoftlimit; + di->d_bcount = dm->dqb_curspace + dm->dqb_rsvspace; + di->d_icount = dm->dqb_curinodes; + di->d_btimer = dm->dqb_btime; + di->d_itimer = dm->dqb_itime; spin_unlock(&dq_data_lock); } int vfs_get_dqblk(struct super_block *sb, int type, qid_t id, - struct if_dqblk *di) + struct fs_disk_quota *di) { struct dquot *dquot; diff --git a/fs/quota/quota.c b/fs/quota/quota.c index 95388f9b7356..8680e257c2bd 100644 --- a/fs/quota/quota.c +++ b/fs/quota/quota.c @@ -136,19 +136,32 @@ static int quota_setinfo(struct super_block *sb, int type, void __user *addr) return sb->s_qcop->set_info(sb, type, &info); } +static void copy_to_if_dqblk(struct if_dqblk *dst, struct fs_disk_quota *src) +{ + dst->dqb_bhardlimit = src->d_blk_hardlimit; + dst->dqb_bsoftlimit = src->d_blk_softlimit; + dst->dqb_curspace = src->d_bcount; + dst->dqb_ihardlimit = src->d_ino_hardlimit; + dst->dqb_isoftlimit = src->d_ino_softlimit; + dst->dqb_curinodes = src->d_icount; + dst->dqb_btime = src->d_btimer; + dst->dqb_itime = src->d_itimer; + dst->dqb_valid = QIF_ALL; +} + static int quota_getquota(struct super_block *sb, int type, qid_t id, void __user *addr) { + struct fs_disk_quota fdq; struct if_dqblk idq; int ret; - if (!sb_has_quota_active(sb, type)) - return -ESRCH; if (!sb->s_qcop->get_dqblk) return -ENOSYS; - ret = sb->s_qcop->get_dqblk(sb, type, id, &idq); + ret = sb->s_qcop->get_dqblk(sb, type, id, &fdq); if (ret) return ret; + copy_to_if_dqblk(&idq, &fdq); if (copy_to_user(addr, &idq, sizeof(idq))) return -EFAULT; return 0; @@ -210,9 +223,9 @@ static int quota_getxquota(struct super_block *sb, int type, qid_t id, struct fs_disk_quota fdq; int ret; - if (!sb->s_qcop->get_xquota) + if (!sb->s_qcop->get_dqblk) return -ENOSYS; - ret = sb->s_qcop->get_xquota(sb, type, id, &fdq); + ret = sb->s_qcop->get_dqblk(sb, type, id, &fdq); if (!ret && copy_to_user(addr, &fdq, sizeof(fdq))) return -EFAULT; return ret; diff --git a/fs/xfs/linux-2.6/xfs_quotaops.c b/fs/xfs/linux-2.6/xfs_quotaops.c index 1947514ce1ad..3d473f43c9a9 100644 --- a/fs/xfs/linux-2.6/xfs_quotaops.c +++ b/fs/xfs/linux-2.6/xfs_quotaops.c @@ -97,7 +97,7 @@ xfs_fs_set_xstate( } STATIC int -xfs_fs_get_xquota( +xfs_fs_get_dqblk( struct super_block *sb, int type, qid_t id, @@ -135,6 +135,6 @@ xfs_fs_set_xquota( const struct quotactl_ops xfs_quotactl_operations = { .get_xstate = xfs_fs_get_xstate, .set_xstate = xfs_fs_set_xstate, - .get_xquota = xfs_fs_get_xquota, + .get_dqblk = xfs_fs_get_dqblk, .set_xquota = xfs_fs_set_xquota, }; diff --git a/include/linux/quota.h b/include/linux/quota.h index cdfde10481b7..42364219dc9b 100644 --- a/include/linux/quota.h +++ b/include/linux/quota.h @@ -337,11 +337,10 @@ struct quotactl_ops { int (*quota_sync)(struct super_block *, int, int); int (*get_info)(struct super_block *, int, struct if_dqinfo *); int (*set_info)(struct super_block *, int, struct if_dqinfo *); - int (*get_dqblk)(struct super_block *, int, qid_t, struct if_dqblk *); + int (*get_dqblk)(struct super_block *, int, qid_t, struct fs_disk_quota *); int (*set_dqblk)(struct super_block *, int, qid_t, struct if_dqblk *); int (*get_xstate)(struct super_block *, struct fs_quota_stat *); int (*set_xstate)(struct super_block *, unsigned int, int); - int (*get_xquota)(struct super_block *, int, qid_t, struct fs_disk_quota *); int (*set_xquota)(struct super_block *, int, qid_t, struct fs_disk_quota *); }; diff --git a/include/linux/quotaops.h b/include/linux/quotaops.h index e6fa7acce290..d32a48631b0d 100644 --- a/include/linux/quotaops.h +++ b/include/linux/quotaops.h @@ -63,7 +63,8 @@ int vfs_quota_disable(struct super_block *sb, int type, unsigned int flags); int vfs_quota_sync(struct super_block *sb, int type, int wait); int vfs_get_dqinfo(struct super_block *sb, int type, struct if_dqinfo *ii); int vfs_set_dqinfo(struct super_block *sb, int type, struct if_dqinfo *ii); -int vfs_get_dqblk(struct super_block *sb, int type, qid_t id, struct if_dqblk *di); +int vfs_get_dqblk(struct super_block *sb, int type, qid_t id, + struct fs_disk_quota *di); int vfs_set_dqblk(struct super_block *sb, int type, qid_t id, struct if_dqblk *di); int dquot_transfer(struct inode *inode, struct iattr *iattr); -- cgit v1.2.3 From c472b43275976512e4c1c32da5ced03f339cb380 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Thu, 6 May 2010 17:05:17 -0400 Subject: quota: unify ->set_dqblk Pass the larger struct fs_disk_quota to the ->set_dqblk operation so that the Q_SETQUOTA and Q_XSETQUOTA operations can be implemented with a single filesystem operation and we can retire the ->set_xquota operation. The additional information (RT-subvolume accounting and warn counts) are left zero for the VFS quota implementation. Add new fieldmask values for setting the numer of blocks and inodes values which is required for the VFS quota, but wasn't for XFS. Signed-off-by: Christoph Hellwig Signed-off-by: Jan Kara --- fs/gfs2/quota.c | 6 ++-- fs/quota/dquot.c | 67 ++++++++++++++++++++++++++--------------- fs/quota/quota.c | 36 +++++++++++++++++++--- fs/xfs/linux-2.6/xfs_quotaops.c | 4 +-- fs/xfs/quota/xfs_qm_syscalls.c | 10 ++++-- include/linux/dqblk_xfs.h | 9 ++++++ include/linux/quota.h | 3 +- include/linux/quotaops.h | 3 +- 8 files changed, 98 insertions(+), 40 deletions(-) (limited to 'include') diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c index dec93577a783..49667d68769e 100644 --- a/fs/gfs2/quota.c +++ b/fs/gfs2/quota.c @@ -1521,8 +1521,8 @@ out: /* GFS2 only supports a subset of the XFS fields */ #define GFS2_FIELDMASK (FS_DQ_BSOFT|FS_DQ_BHARD) -static int gfs2_xquota_set(struct super_block *sb, int type, qid_t id, - struct fs_disk_quota *fdq) +static int gfs2_set_dqblk(struct super_block *sb, int type, qid_t id, + struct fs_disk_quota *fdq) { struct gfs2_sbd *sdp = sb->s_fs_info; struct gfs2_inode *ip = GFS2_I(sdp->sd_quota_inode); @@ -1630,6 +1630,6 @@ const struct quotactl_ops gfs2_quotactl_ops = { .quota_sync = gfs2_quota_sync, .get_xstate = gfs2_quota_get_xstate, .get_dqblk = gfs2_get_dqblk, - .set_xquota = gfs2_xquota_set, + .set_dqblk = gfs2_set_dqblk, }; diff --git a/fs/quota/dquot.c b/fs/quota/dquot.c index 6aed77fc99c7..b1a5036560a9 100644 --- a/fs/quota/dquot.c +++ b/fs/quota/dquot.c @@ -2338,51 +2338,70 @@ int vfs_get_dqblk(struct super_block *sb, int type, qid_t id, } EXPORT_SYMBOL(vfs_get_dqblk); +#define VFS_FS_DQ_MASK \ + (FS_DQ_BCOUNT | FS_DQ_BSOFT | FS_DQ_BHARD | \ + FS_DQ_ICOUNT | FS_DQ_ISOFT | FS_DQ_IHARD | \ + FS_DQ_BTIMER | FS_DQ_ITIMER) + /* Generic routine for setting common part of quota structure */ -static int do_set_dqblk(struct dquot *dquot, struct if_dqblk *di) +static int do_set_dqblk(struct dquot *dquot, struct fs_disk_quota *di) { struct mem_dqblk *dm = &dquot->dq_dqb; int check_blim = 0, check_ilim = 0; struct mem_dqinfo *dqi = &sb_dqopt(dquot->dq_sb)->info[dquot->dq_type]; - if ((di->dqb_valid & QIF_BLIMITS && - (di->dqb_bhardlimit > dqi->dqi_maxblimit || - di->dqb_bsoftlimit > dqi->dqi_maxblimit)) || - (di->dqb_valid & QIF_ILIMITS && - (di->dqb_ihardlimit > dqi->dqi_maxilimit || - di->dqb_isoftlimit > dqi->dqi_maxilimit))) + if (di->d_fieldmask & ~VFS_FS_DQ_MASK) + return -EINVAL; + + if (((di->d_fieldmask & FS_DQ_BSOFT) && + (di->d_blk_softlimit > dqi->dqi_maxblimit)) || + ((di->d_fieldmask & FS_DQ_BHARD) && + (di->d_blk_hardlimit > dqi->dqi_maxblimit)) || + ((di->d_fieldmask & FS_DQ_ISOFT) && + (di->d_ino_softlimit > dqi->dqi_maxilimit)) || + ((di->d_fieldmask & FS_DQ_IHARD) && + (di->d_ino_hardlimit > dqi->dqi_maxilimit))) return -ERANGE; spin_lock(&dq_data_lock); - if (di->dqb_valid & QIF_SPACE) { - dm->dqb_curspace = di->dqb_curspace - dm->dqb_rsvspace; + if (di->d_fieldmask & FS_DQ_BCOUNT) { + dm->dqb_curspace = di->d_bcount - dm->dqb_rsvspace; check_blim = 1; set_bit(DQ_LASTSET_B + QIF_SPACE_B, &dquot->dq_flags); } - if (di->dqb_valid & QIF_BLIMITS) { - dm->dqb_bsoftlimit = qbtos(di->dqb_bsoftlimit); - dm->dqb_bhardlimit = qbtos(di->dqb_bhardlimit); + + if (di->d_fieldmask & FS_DQ_BSOFT) + dm->dqb_bsoftlimit = qbtos(di->d_blk_softlimit); + if (di->d_fieldmask & FS_DQ_BHARD) + dm->dqb_bhardlimit = qbtos(di->d_blk_hardlimit); + if (di->d_fieldmask & (FS_DQ_BSOFT | FS_DQ_BHARD)) { check_blim = 1; set_bit(DQ_LASTSET_B + QIF_BLIMITS_B, &dquot->dq_flags); } - if (di->dqb_valid & QIF_INODES) { - dm->dqb_curinodes = di->dqb_curinodes; + + if (di->d_fieldmask & FS_DQ_ICOUNT) { + dm->dqb_curinodes = di->d_icount; check_ilim = 1; set_bit(DQ_LASTSET_B + QIF_INODES_B, &dquot->dq_flags); } - if (di->dqb_valid & QIF_ILIMITS) { - dm->dqb_isoftlimit = di->dqb_isoftlimit; - dm->dqb_ihardlimit = di->dqb_ihardlimit; + + if (di->d_fieldmask & FS_DQ_ISOFT) + dm->dqb_isoftlimit = di->d_ino_softlimit; + if (di->d_fieldmask & FS_DQ_IHARD) + dm->dqb_ihardlimit = di->d_ino_hardlimit; + if (di->d_fieldmask & (FS_DQ_ISOFT | FS_DQ_IHARD)) { check_ilim = 1; set_bit(DQ_LASTSET_B + QIF_ILIMITS_B, &dquot->dq_flags); } - if (di->dqb_valid & QIF_BTIME) { - dm->dqb_btime = di->dqb_btime; + + if (di->d_fieldmask & FS_DQ_BTIMER) { + dm->dqb_btime = di->d_btimer; check_blim = 1; set_bit(DQ_LASTSET_B + QIF_BTIME_B, &dquot->dq_flags); } - if (di->dqb_valid & QIF_ITIME) { - dm->dqb_itime = di->dqb_itime; + + if (di->d_fieldmask & FS_DQ_ITIMER) { + dm->dqb_itime = di->d_itimer; check_ilim = 1; set_bit(DQ_LASTSET_B + QIF_ITIME_B, &dquot->dq_flags); } @@ -2392,7 +2411,7 @@ static int do_set_dqblk(struct dquot *dquot, struct if_dqblk *di) dm->dqb_curspace < dm->dqb_bsoftlimit) { dm->dqb_btime = 0; clear_bit(DQ_BLKS_B, &dquot->dq_flags); - } else if (!(di->dqb_valid & QIF_BTIME)) + } else if (!(di->d_fieldmask & FS_DQ_BTIMER)) /* Set grace only if user hasn't provided his own... */ dm->dqb_btime = get_seconds() + dqi->dqi_bgrace; } @@ -2401,7 +2420,7 @@ static int do_set_dqblk(struct dquot *dquot, struct if_dqblk *di) dm->dqb_curinodes < dm->dqb_isoftlimit) { dm->dqb_itime = 0; clear_bit(DQ_INODES_B, &dquot->dq_flags); - } else if (!(di->dqb_valid & QIF_ITIME)) + } else if (!(di->d_fieldmask & FS_DQ_ITIMER)) /* Set grace only if user hasn't provided his own... */ dm->dqb_itime = get_seconds() + dqi->dqi_igrace; } @@ -2417,7 +2436,7 @@ static int do_set_dqblk(struct dquot *dquot, struct if_dqblk *di) } int vfs_set_dqblk(struct super_block *sb, int type, qid_t id, - struct if_dqblk *di) + struct fs_disk_quota *di) { struct dquot *dquot; int rc; diff --git a/fs/quota/quota.c b/fs/quota/quota.c index 8680e257c2bd..d6ee49dda4fd 100644 --- a/fs/quota/quota.c +++ b/fs/quota/quota.c @@ -167,18 +167,44 @@ static int quota_getquota(struct super_block *sb, int type, qid_t id, return 0; } +static void copy_from_if_dqblk(struct fs_disk_quota *dst, struct if_dqblk *src) +{ + dst->d_blk_hardlimit = src->dqb_bhardlimit; + dst->d_blk_softlimit = src->dqb_bsoftlimit; + dst->d_bcount = src->dqb_curspace; + dst->d_ino_hardlimit = src->dqb_ihardlimit; + dst->d_ino_softlimit = src->dqb_isoftlimit; + dst->d_icount = src->dqb_curinodes; + dst->d_btimer = src->dqb_btime; + dst->d_itimer = src->dqb_itime; + + dst->d_fieldmask = 0; + if (src->dqb_valid & QIF_BLIMITS) + dst->d_fieldmask |= FS_DQ_BSOFT | FS_DQ_BHARD; + if (src->dqb_valid & QIF_SPACE) + dst->d_fieldmask |= FS_DQ_BCOUNT; + if (src->dqb_valid & QIF_ILIMITS) + dst->d_fieldmask |= FS_DQ_ISOFT | FS_DQ_IHARD; + if (src->dqb_valid & QIF_INODES) + dst->d_fieldmask |= FS_DQ_ICOUNT; + if (src->dqb_valid & QIF_BTIME) + dst->d_fieldmask |= FS_DQ_BTIMER; + if (src->dqb_valid & QIF_ITIME) + dst->d_fieldmask |= FS_DQ_ITIMER; +} + static int quota_setquota(struct super_block *sb, int type, qid_t id, void __user *addr) { + struct fs_disk_quota fdq; struct if_dqblk idq; if (copy_from_user(&idq, addr, sizeof(idq))) return -EFAULT; - if (!sb_has_quota_active(sb, type)) - return -ESRCH; if (!sb->s_qcop->set_dqblk) return -ENOSYS; - return sb->s_qcop->set_dqblk(sb, type, id, &idq); + copy_from_if_dqblk(&fdq, &idq); + return sb->s_qcop->set_dqblk(sb, type, id, &fdq); } static int quota_setxstate(struct super_block *sb, int cmd, void __user *addr) @@ -212,9 +238,9 @@ static int quota_setxquota(struct super_block *sb, int type, qid_t id, if (copy_from_user(&fdq, addr, sizeof(fdq))) return -EFAULT; - if (!sb->s_qcop->set_xquota) + if (!sb->s_qcop->set_dqblk) return -ENOSYS; - return sb->s_qcop->set_xquota(sb, type, id, &fdq); + return sb->s_qcop->set_dqblk(sb, type, id, &fdq); } static int quota_getxquota(struct super_block *sb, int type, qid_t id, diff --git a/fs/xfs/linux-2.6/xfs_quotaops.c b/fs/xfs/linux-2.6/xfs_quotaops.c index 3d473f43c9a9..e31bf21fe5d3 100644 --- a/fs/xfs/linux-2.6/xfs_quotaops.c +++ b/fs/xfs/linux-2.6/xfs_quotaops.c @@ -114,7 +114,7 @@ xfs_fs_get_dqblk( } STATIC int -xfs_fs_set_xquota( +xfs_fs_set_dqblk( struct super_block *sb, int type, qid_t id, @@ -136,5 +136,5 @@ const struct quotactl_ops xfs_quotactl_operations = { .get_xstate = xfs_fs_get_xstate, .set_xstate = xfs_fs_set_xstate, .get_dqblk = xfs_fs_get_dqblk, - .set_xquota = xfs_fs_set_xquota, + .set_dqblk = xfs_fs_set_dqblk, }; diff --git a/fs/xfs/quota/xfs_qm_syscalls.c b/fs/xfs/quota/xfs_qm_syscalls.c index 26fa43140f2e..92b002f1805f 100644 --- a/fs/xfs/quota/xfs_qm_syscalls.c +++ b/fs/xfs/quota/xfs_qm_syscalls.c @@ -448,6 +448,9 @@ xfs_qm_scall_getqstat( return 0; } +#define XFS_DQ_MASK \ + (FS_DQ_LIMIT_MASK | FS_DQ_TIMER_MASK | FS_DQ_WARNS_MASK) + /* * Adjust quota limits, and start/stop timers accordingly. */ @@ -465,9 +468,10 @@ xfs_qm_scall_setqlim( int error; xfs_qcnt_t hard, soft; - if ((newlim->d_fieldmask & - (FS_DQ_LIMIT_MASK|FS_DQ_TIMER_MASK|FS_DQ_WARNS_MASK)) == 0) - return (0); + if (newlim->d_fieldmask & ~XFS_DQ_MASK) + return EINVAL; + if ((newlim->d_fieldmask & XFS_DQ_MASK) == 0) + return 0; tp = xfs_trans_alloc(mp, XFS_TRANS_QM_SETQLIM); if ((error = xfs_trans_reserve(tp, 0, sizeof(xfs_disk_dquot_t) + 128, diff --git a/include/linux/dqblk_xfs.h b/include/linux/dqblk_xfs.h index 527504c11c5e..4389ae72024e 100644 --- a/include/linux/dqblk_xfs.h +++ b/include/linux/dqblk_xfs.h @@ -109,6 +109,15 @@ typedef struct fs_disk_quota { #define FS_DQ_RTBWARNS (1<<11) #define FS_DQ_WARNS_MASK (FS_DQ_BWARNS | FS_DQ_IWARNS | FS_DQ_RTBWARNS) +/* + * Accounting values. These can only be set for filesystem with + * non-transactional quotas that require quotacheck(8) in userspace. + */ +#define FS_DQ_BCOUNT (1<<12) +#define FS_DQ_ICOUNT (1<<13) +#define FS_DQ_RTBCOUNT (1<<14) +#define FS_DQ_ACCT_MASK (FS_DQ_BCOUNT | FS_DQ_ICOUNT | FS_DQ_RTBCOUNT) + /* * Various flags related to quotactl(2). Only relevant to XFS filesystems. */ diff --git a/include/linux/quota.h b/include/linux/quota.h index 42364219dc9b..7126a15467f1 100644 --- a/include/linux/quota.h +++ b/include/linux/quota.h @@ -338,10 +338,9 @@ struct quotactl_ops { int (*get_info)(struct super_block *, int, struct if_dqinfo *); int (*set_info)(struct super_block *, int, struct if_dqinfo *); int (*get_dqblk)(struct super_block *, int, qid_t, struct fs_disk_quota *); - int (*set_dqblk)(struct super_block *, int, qid_t, struct if_dqblk *); + int (*set_dqblk)(struct super_block *, int, qid_t, struct fs_disk_quota *); int (*get_xstate)(struct super_block *, struct fs_quota_stat *); int (*set_xstate)(struct super_block *, unsigned int, int); - int (*set_xquota)(struct super_block *, int, qid_t, struct fs_disk_quota *); }; struct quota_format_type { diff --git a/include/linux/quotaops.h b/include/linux/quotaops.h index d32a48631b0d..82c70c42d035 100644 --- a/include/linux/quotaops.h +++ b/include/linux/quotaops.h @@ -65,7 +65,8 @@ int vfs_get_dqinfo(struct super_block *sb, int type, struct if_dqinfo *ii); int vfs_set_dqinfo(struct super_block *sb, int type, struct if_dqinfo *ii); int vfs_get_dqblk(struct super_block *sb, int type, qid_t id, struct fs_disk_quota *di); -int vfs_set_dqblk(struct super_block *sb, int type, qid_t id, struct if_dqblk *di); +int vfs_set_dqblk(struct super_block *sb, int type, qid_t id, + struct fs_disk_quota *di); int dquot_transfer(struct inode *inode, struct iattr *iattr); int vfs_dq_quota_on_remount(struct super_block *sb); -- cgit v1.2.3 From 12755627bdcddcdb30a1bfb9a09395a52b1d6838 Mon Sep 17 00:00:00 2001 From: Dmitry Monakhov Date: Thu, 8 Apr 2010 22:04:20 +0400 Subject: quota: unify quota init condition in setattr Quota must being initialized if size or uid/git changes requested. But initialization performed in two different places: in case of i_size file system is responsible for dquot init , but in case of uid/gid init will be called internally in dquot_transfer(). This ambiguity makes code harder to understand. Let's move this logic to one common helper function. Signed-off-by: Dmitry Monakhov Signed-off-by: Jan Kara --- fs/ext2/inode.c | 2 +- fs/ext3/inode.c | 2 +- fs/ext4/inode.c | 2 +- fs/jfs/file.c | 2 +- fs/ocfs2/file.c | 4 ++-- fs/quota/dquot.c | 5 ++--- fs/reiserfs/inode.c | 3 ++- fs/udf/file.c | 2 +- fs/ufs/truncate.c | 8 ++++---- include/linux/quotaops.h | 8 ++++++++ 10 files changed, 23 insertions(+), 15 deletions(-) (limited to 'include') diff --git a/fs/ext2/inode.c b/fs/ext2/inode.c index b90c3bf6e9ba..527c46d9bc1f 100644 --- a/fs/ext2/inode.c +++ b/fs/ext2/inode.c @@ -1466,7 +1466,7 @@ int ext2_setattr(struct dentry *dentry, struct iattr *iattr) if (error) return error; - if (iattr->ia_valid & ATTR_SIZE) + if (is_quota_modification(inode, iattr)) dquot_initialize(inode); if ((iattr->ia_valid & ATTR_UID && iattr->ia_uid != inode->i_uid) || (iattr->ia_valid & ATTR_GID && iattr->ia_gid != inode->i_gid)) { diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c index ea33bdf0a300..735f0190ec2a 100644 --- a/fs/ext3/inode.c +++ b/fs/ext3/inode.c @@ -3151,7 +3151,7 @@ int ext3_setattr(struct dentry *dentry, struct iattr *attr) if (error) return error; - if (ia_valid & ATTR_SIZE) + if (is_quota_modification(inode, attr)) dquot_initialize(inode); if ((ia_valid & ATTR_UID && attr->ia_uid != inode->i_uid) || (ia_valid & ATTR_GID && attr->ia_gid != inode->i_gid)) { diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 81d605412844..3e0f6af9d08d 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -5425,7 +5425,7 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr) if (error) return error; - if (ia_valid & ATTR_SIZE) + if (is_quota_modification(inode, attr)) dquot_initialize(inode); if ((ia_valid & ATTR_UID && attr->ia_uid != inode->i_uid) || (ia_valid & ATTR_GID && attr->ia_gid != inode->i_gid)) { diff --git a/fs/jfs/file.c b/fs/jfs/file.c index 14ba982b3f24..85d9ec659225 100644 --- a/fs/jfs/file.c +++ b/fs/jfs/file.c @@ -98,7 +98,7 @@ int jfs_setattr(struct dentry *dentry, struct iattr *iattr) if (rc) return rc; - if (iattr->ia_valid & ATTR_SIZE) + if (is_quota_modification(inode, iattr)) dquot_initialize(inode); if ((iattr->ia_valid & ATTR_UID && iattr->ia_uid != inode->i_uid) || (iattr->ia_valid & ATTR_GID && iattr->ia_gid != inode->i_gid)) { diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c index f74f1400eccd..e127c53ec2e7 100644 --- a/fs/ocfs2/file.c +++ b/fs/ocfs2/file.c @@ -966,10 +966,10 @@ int ocfs2_setattr(struct dentry *dentry, struct iattr *attr) if (status) return status; + if (is_quota_modification(inode, attr)) + dquot_initialize(inode); size_change = S_ISREG(inode->i_mode) && attr->ia_valid & ATTR_SIZE; if (size_change) { - dquot_initialize(inode); - status = ocfs2_rw_lock(inode, 1); if (status < 0) { mlog_errno(status); diff --git a/fs/quota/dquot.c b/fs/quota/dquot.c index b1a5036560a9..1056a21f0300 100644 --- a/fs/quota/dquot.c +++ b/fs/quota/dquot.c @@ -1822,10 +1822,9 @@ int dquot_transfer(struct inode *inode, struct iattr *iattr) mask |= 1 << GRPQUOTA; chid[GRPQUOTA] = iattr->ia_gid; } - if (sb_any_quota_active(inode->i_sb) && !IS_NOQUOTA(inode)) { - dquot_initialize(inode); + if (sb_any_quota_active(inode->i_sb) && !IS_NOQUOTA(inode)) return __dquot_transfer(inode, chid, mask); - } + return 0; } EXPORT_SYMBOL(dquot_transfer); diff --git a/fs/reiserfs/inode.c b/fs/reiserfs/inode.c index dc2c65e04853..0f22fdaf54ac 100644 --- a/fs/reiserfs/inode.c +++ b/fs/reiserfs/inode.c @@ -3076,9 +3076,10 @@ int reiserfs_setattr(struct dentry *dentry, struct iattr *attr) ia_valid = attr->ia_valid &= ~(ATTR_KILL_SUID|ATTR_KILL_SGID); depth = reiserfs_write_lock_once(inode->i_sb); - if (attr->ia_valid & ATTR_SIZE) { + if (is_quota_modification(inode, attr)) dquot_initialize(inode); + if (attr->ia_valid & ATTR_SIZE) { /* version 2 items will be caught by the s_maxbytes check ** done for us in vmtruncate */ diff --git a/fs/udf/file.c b/fs/udf/file.c index 4b6a46ccbf46..6ebc043f3a2a 100644 --- a/fs/udf/file.c +++ b/fs/udf/file.c @@ -227,7 +227,7 @@ int udf_setattr(struct dentry *dentry, struct iattr *iattr) if (error) return error; - if (iattr->ia_valid & ATTR_SIZE) + if (is_quota_modification(inode, iattr)) dquot_initialize(inode); if ((iattr->ia_valid & ATTR_UID && iattr->ia_uid != inode->i_uid) || diff --git a/fs/ufs/truncate.c b/fs/ufs/truncate.c index ee8db3e77bfe..f294c44577dc 100644 --- a/fs/ufs/truncate.c +++ b/fs/ufs/truncate.c @@ -518,18 +518,18 @@ int ufs_setattr(struct dentry *dentry, struct iattr *attr) if (error) return error; + if (is_quota_modification(inode, attr)) + dquot_initialize(inode); + if ((ia_valid & ATTR_UID && attr->ia_uid != inode->i_uid) || (ia_valid & ATTR_GID && attr->ia_gid != inode->i_gid)) { error = dquot_transfer(inode, attr); if (error) return error; } - if (ia_valid & ATTR_SIZE && - attr->ia_size != i_size_read(inode)) { + if (ia_valid & ATTR_SIZE && attr->ia_size != inode->i_size) { loff_t old_i_size = inode->i_size; - dquot_initialize(inode); - error = vmtruncate(inode, attr->ia_size); if (error) return error; diff --git a/include/linux/quotaops.h b/include/linux/quotaops.h index 82c70c42d035..8a7818764a67 100644 --- a/include/linux/quotaops.h +++ b/include/linux/quotaops.h @@ -14,6 +14,14 @@ static inline struct quota_info *sb_dqopt(struct super_block *sb) return &sb->s_dquot; } +/* i_mutex must being held */ +static inline bool is_quota_modification(struct inode *inode, struct iattr *ia) +{ + return (ia->ia_valid & ATTR_SIZE && ia->ia_size != inode->i_size) || + (ia->ia_valid & ATTR_UID && ia->ia_uid != inode->i_uid) || + (ia->ia_valid & ATTR_GID && ia->ia_gid != inode->i_gid); +} + #if defined(CONFIG_QUOTA) /* -- cgit v1.2.3 From bc8e5f07392f05c47c8bdeff4f7098db440d065c Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Thu, 13 May 2010 19:58:50 +0200 Subject: quota: Refactor dquot_transfer code so that OCFS2 can pass in its references Currently, __dquot_transfer() acquires its own references of dquot structures that will be put into inode. But for OCFS2, this creates a lock inversion between dq_lock (waited on in dqget) and transaction start (started in ocfs2_setattr). Currently, deadlock is impossible because dq_lock is acquired only during dquot_acquire and dquot_release and we already hold a reference to dquot structures in ocfs2_setattr so neither of these functions can be called while we call dquot_transfer. But this is rather subtle and it is hard to teach lockdep about it. So provide __dquot_transfer function that can be passed dquot references directly. OCFS2 can then pass acquired dquot references directly to __dquot_transfer with proper locking. Signed-off-by: Jan Kara --- fs/quota/dquot.c | 61 +++++++++++++++++++++--------------------------- include/linux/quotaops.h | 1 + 2 files changed, 27 insertions(+), 35 deletions(-) (limited to 'include') diff --git a/fs/quota/dquot.c b/fs/quota/dquot.c index 1056a21f0300..655a4c52b8c3 100644 --- a/fs/quota/dquot.c +++ b/fs/quota/dquot.c @@ -1703,16 +1703,19 @@ EXPORT_SYMBOL(dquot_free_inode); /* * Transfer the number of inode and blocks from one diskquota to an other. + * On success, dquot references in transfer_to are consumed and references + * to original dquots that need to be released are placed there. On failure, + * references are kept untouched. * * This operation can block, but only after everything is updated * A transaction must be started when entering this function. + * */ -static int __dquot_transfer(struct inode *inode, qid_t *chid, unsigned long mask) +int __dquot_transfer(struct inode *inode, struct dquot **transfer_to) { qsize_t space, cur_space; qsize_t rsv_space = 0; - struct dquot *transfer_from[MAXQUOTAS]; - struct dquot *transfer_to[MAXQUOTAS]; + struct dquot *transfer_from[MAXQUOTAS] = {}; int cnt, ret = 0; char warntype_to[MAXQUOTAS]; char warntype_from_inodes[MAXQUOTAS], warntype_from_space[MAXQUOTAS]; @@ -1722,19 +1725,12 @@ static int __dquot_transfer(struct inode *inode, qid_t *chid, unsigned long mask if (IS_NOQUOTA(inode)) return 0; /* Initialize the arrays */ - for (cnt = 0; cnt < MAXQUOTAS; cnt++) { - transfer_from[cnt] = NULL; - transfer_to[cnt] = NULL; + for (cnt = 0; cnt < MAXQUOTAS; cnt++) warntype_to[cnt] = QUOTA_NL_NOWARN; - } - for (cnt = 0; cnt < MAXQUOTAS; cnt++) { - if (mask & (1 << cnt)) - transfer_to[cnt] = dqget(inode->i_sb, chid[cnt], cnt); - } down_write(&sb_dqopt(inode->i_sb)->dqptr_sem); if (IS_NOQUOTA(inode)) { /* File without quota accounting? */ up_write(&sb_dqopt(inode->i_sb)->dqptr_sem); - goto put_all; + return 0; } spin_lock(&dq_data_lock); cur_space = inode_get_bytes(inode); @@ -1786,46 +1782,41 @@ static int __dquot_transfer(struct inode *inode, qid_t *chid, unsigned long mask mark_all_dquot_dirty(transfer_from); mark_all_dquot_dirty(transfer_to); - /* The reference we got is transferred to the inode */ + /* Pass back references to put */ for (cnt = 0; cnt < MAXQUOTAS; cnt++) - transfer_to[cnt] = NULL; -warn_put_all: + transfer_to[cnt] = transfer_from[cnt]; +warn: flush_warnings(transfer_to, warntype_to); flush_warnings(transfer_from, warntype_from_inodes); flush_warnings(transfer_from, warntype_from_space); -put_all: - dqput_all(transfer_from); - dqput_all(transfer_to); return ret; over_quota: spin_unlock(&dq_data_lock); up_write(&sb_dqopt(inode->i_sb)->dqptr_sem); - /* Clear dquot pointers we don't want to dqput() */ - for (cnt = 0; cnt < MAXQUOTAS; cnt++) - transfer_from[cnt] = NULL; - goto warn_put_all; + goto warn; } +EXPORT_SYMBOL(__dquot_transfer); /* Wrapper for transferring ownership of an inode for uid/gid only * Called from FSXXX_setattr() */ int dquot_transfer(struct inode *inode, struct iattr *iattr) { - qid_t chid[MAXQUOTAS]; - unsigned long mask = 0; + struct dquot *transfer_to[MAXQUOTAS] = {}; + struct super_block *sb = inode->i_sb; + int ret; - if (iattr->ia_valid & ATTR_UID && iattr->ia_uid != inode->i_uid) { - mask |= 1 << USRQUOTA; - chid[USRQUOTA] = iattr->ia_uid; - } - if (iattr->ia_valid & ATTR_GID && iattr->ia_gid != inode->i_gid) { - mask |= 1 << GRPQUOTA; - chid[GRPQUOTA] = iattr->ia_gid; - } - if (sb_any_quota_active(inode->i_sb) && !IS_NOQUOTA(inode)) - return __dquot_transfer(inode, chid, mask); + if (!sb_any_quota_active(sb) || IS_NOQUOTA(inode)) + return 0; - return 0; + if (iattr->ia_valid & ATTR_UID && iattr->ia_uid != inode->i_uid) + transfer_to[USRQUOTA] = dqget(sb, iattr->ia_uid, USRQUOTA); + if (iattr->ia_valid & ATTR_GID && iattr->ia_gid != inode->i_gid) + transfer_to[GRPQUOTA] = dqget(sb, iattr->ia_uid, GRPQUOTA); + + ret = __dquot_transfer(inode, transfer_to); + dqput_all(transfer_to); + return ret; } EXPORT_SYMBOL(dquot_transfer); diff --git a/include/linux/quotaops.h b/include/linux/quotaops.h index 8a7818764a67..370abb1e99cb 100644 --- a/include/linux/quotaops.h +++ b/include/linux/quotaops.h @@ -76,6 +76,7 @@ int vfs_get_dqblk(struct super_block *sb, int type, qid_t id, int vfs_set_dqblk(struct super_block *sb, int type, qid_t id, struct fs_disk_quota *di); +int __dquot_transfer(struct inode *inode, struct dquot **transfer_to); int dquot_transfer(struct inode *inode, struct iattr *iattr); int vfs_dq_quota_on_remount(struct super_block *sb); -- cgit v1.2.3