diff options
Diffstat (limited to 'fs/quota')
| -rw-r--r-- | fs/quota/Kconfig | 59 | ||||
| -rw-r--r-- | fs/quota/Makefile | 14 | ||||
| -rw-r--r-- | fs/quota/dquot.c | 2617 | ||||
| -rw-r--r-- | fs/quota/quota.c | 524 | ||||
| -rw-r--r-- | fs/quota/quota_tree.c | 651 | ||||
| -rw-r--r-- | fs/quota/quota_tree.h | 25 | ||||
| -rw-r--r-- | fs/quota/quota_v1.c | 234 | ||||
| -rw-r--r-- | fs/quota/quota_v2.c | 237 | ||||
| -rw-r--r-- | fs/quota/quotaio_v1.h | 33 | ||||
| -rw-r--r-- | fs/quota/quotaio_v2.h | 60 |
10 files changed, 4454 insertions, 0 deletions
diff --git a/fs/quota/Kconfig b/fs/quota/Kconfig new file mode 100644 index 000000000000..8047e01ef46b --- /dev/null +++ b/fs/quota/Kconfig @@ -0,0 +1,59 @@ +# +# Quota configuration +# + +config QUOTA + bool "Quota support" + help + If you say Y here, you will be able to set per user limits for disk + usage (also called disk quotas). Currently, it works for the + ext2, ext3, and reiserfs file system. ext3 also supports journalled + quotas for which you don't need to run quotacheck(8) after an unclean + shutdown. + For further details, read the Quota mini-HOWTO, available from + <http://www.tldp.org/docs.html#howto>, or the documentation provided + with the quota tools. Probably the quota support is only useful for + multi user systems. If unsure, say N. + +config QUOTA_NETLINK_INTERFACE + bool "Report quota messages through netlink interface" + depends on QUOTA && NET + help + If you say Y here, quota warnings (about exceeding softlimit, reaching + hardlimit, etc.) will be reported through netlink interface. If unsure, + say Y. + +config PRINT_QUOTA_WARNING + bool "Print quota warnings to console (OBSOLETE)" + depends on QUOTA + default y + help + If you say Y here, quota warnings (about exceeding softlimit, reaching + hardlimit, etc.) will be printed to the process' controlling terminal. + Note that this behavior is currently deprecated and may go away in + future. Please use notification via netlink socket instead. + +# Generic support for tree structured quota files. Selected when needed. +config QUOTA_TREE + tristate + +config QFMT_V1 + tristate "Old quota format support" + depends on QUOTA + help + This quota format was (is) used by kernels earlier than 2.4.22. If + you have quota working and you don't want to convert to new quota + format say Y here. + +config QFMT_V2 + tristate "Quota format v2 support" + depends on QUOTA + select QUOTA_TREE + help + This quota format allows using quotas with 32-bit UIDs/GIDs. If you + need this functionality say Y here. + +config QUOTACTL + bool + depends on XFS_QUOTA || QUOTA + default y diff --git a/fs/quota/Makefile b/fs/quota/Makefile new file mode 100644 index 000000000000..385a0831cc99 --- /dev/null +++ b/fs/quota/Makefile @@ -0,0 +1,14 @@ +# +# Makefile for the Linux filesystems. +# +# 14 Sep 2000, Christoph Hellwig <hch@infradead.org> +# Rewritten to use lists instead of if-statements. +# + +obj-y := + +obj-$(CONFIG_QUOTA) += dquot.o +obj-$(CONFIG_QFMT_V1) += quota_v1.o +obj-$(CONFIG_QFMT_V2) += quota_v2.o +obj-$(CONFIG_QUOTA_TREE) += quota_tree.o +obj-$(CONFIG_QUOTACTL) += quota.o diff --git a/fs/quota/dquot.c b/fs/quota/dquot.c new file mode 100644 index 000000000000..2ca967a5ef77 --- /dev/null +++ b/fs/quota/dquot.c @@ -0,0 +1,2617 @@ +/* + * Implementation of the diskquota system for the LINUX operating system. QUOTA + * is implemented using the BSD system call interface as the means of + * communication with the user level. This file contains the generic routines + * called by the different filesystems on allocation of an inode or block. + * These routines take care of the administration needed to have a consistent + * diskquota tracking system. The ideas of both user and group quotas are based + * on the Melbourne quota system as used on BSD derived systems. The internal + * implementation is based on one of the several variants of the LINUX + * inode-subsystem with added complexity of the diskquota system. + * + * Author: Marco van Wieringen <mvw@planets.elm.net> + * + * Fixes: Dmitry Gorodchanin <pgmdsg@ibi.com>, 11 Feb 96 + * + * Revised list management to avoid races + * -- Bill Hawes, <whawes@star.net>, 9/98 + * + * Fixed races in dquot_transfer(), dqget() and dquot_alloc_...(). + * As the consequence the locking was moved from dquot_decr_...(), + * dquot_incr_...() to calling functions. + * invalidate_dquots() now writes modified dquots. + * Serialized quota_off() and quota_on() for mount point. + * Fixed a few bugs in grow_dquots(). + * Fixed deadlock in write_dquot() - we no longer account quotas on + * quota files + * remove_dquot_ref() moved to inode.c - it now traverses through inodes + * add_dquot_ref() restarts after blocking + * Added check for bogus uid and fixed check for group in quotactl. + * Jan Kara, <jack@suse.cz>, sponsored by SuSE CR, 10-11/99 + * + * Used struct list_head instead of own list struct + * Invalidation of referenced dquots is no longer possible + * Improved free_dquots list management + * Quota and i_blocks are now updated in one place to avoid races + * Warnings are now delayed so we won't block in critical section + * Write updated not to require dquot lock + * Jan Kara, <jack@suse.cz>, 9/2000 + * + * Added dynamic quota structure allocation + * Jan Kara <jack@suse.cz> 12/2000 + * + * Rewritten quota interface. Implemented new quota format and + * formats registering. + * Jan Kara, <jack@suse.cz>, 2001,2002 + * + * New SMP locking. + * Jan Kara, <jack@suse.cz>, 10/2002 + * + * Added journalled quota support, fix lock inversion problems + * Jan Kara, <jack@suse.cz>, 2003,2004 + * + * (C) Copyright 1994 - 1997 Marco van Wieringen + */ + +#include <linux/errno.h> +#include <linux/kernel.h> +#include <linux/fs.h> +#include <linux/mount.h> +#include <linux/mm.h> +#include <linux/time.h> +#include <linux/types.h> +#include <linux/string.h> +#include <linux/fcntl.h> +#include <linux/stat.h> +#include <linux/tty.h> +#include <linux/file.h> +#include <linux/slab.h> +#include <linux/sysctl.h> +#include <linux/init.h> +#include <linux/module.h> +#include <linux/proc_fs.h> +#include <linux/security.h> +#include <linux/kmod.h> +#include <linux/namei.h> +#include <linux/buffer_head.h> +#include <linux/capability.h> +#include <linux/quotaops.h> +#include <linux/writeback.h> /* for inode_lock, oddly enough.. */ +#ifdef CONFIG_QUOTA_NETLINK_INTERFACE +#include <net/netlink.h> +#include <net/genetlink.h> +#endif + +#include <asm/uaccess.h> + +#define __DQUOT_PARANOIA + +/* + * There are three quota SMP locks. dq_list_lock protects all lists with quotas + * and quota formats, dqstats structure containing statistics about the lists + * 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 + * in inode_add_bytes() and inode_sub_bytes(). dq_state_lock protects + * modifications of quota state (on quotaon and quotaoff) and readers who care + * about latest values take it as well. + * + * The spinlock ordering is hence: dq_data_lock > dq_list_lock > i_lock, + * dq_list_lock > dq_state_lock + * + * Note that some things (eg. sb pointer, type, id) doesn't change during + * the life of the dquot structure and so needn't to be protected by a lock + * + * Any operation working on dquots via inode pointers must hold dqptr_sem. If + * operation is just reading pointers from inode (or not using them at all) the + * read lock is enough. If pointers are altered function must hold write lock + * (these locking rules also apply for S_NOQUOTA flag in the inode - note that + * for altering the flag i_mutex is also needed). + * + * Each dquot has its dq_lock mutex. Locked dquots might not be referenced + * from inodes (dquot_alloc_space() and such don't check the dq_lock). + * Currently dquot is locked only when it is being read to memory (or space for + * it is being allocated) on the first dqget() and when it is being released on + * the last dqput(). The allocation and release oparations are serialized by + * the dq_lock and by checking the use count in dquot_release(). Write + * operations on dquots don't hold dq_lock as they copy data under dq_data_lock + * spinlock to internal buffers before writing. + * + * Lock ordering (including related VFS locks) is the following: + * i_mutex > dqonoff_sem > journal_lock > dqptr_sem > dquot->dq_lock > + * dqio_mutex + * The lock ordering of dqptr_sem imposed by quota code is only dqonoff_sem > + * dqptr_sem. But filesystem has to count with the fact that functions such as + * dquot_alloc_space() acquire dqptr_sem and they usually have to be called + * from inside a transaction to keep filesystem consistency after a crash. Also + * filesystems usually want to do some IO on dquot from ->mark_dirty which is + * called with dqptr_sem held. + * i_mutex on quota files is special (it's below dqio_mutex) + */ + +static __cacheline_aligned_in_smp DEFINE_SPINLOCK(dq_list_lock); +static __cacheline_aligned_in_smp DEFINE_SPINLOCK(dq_state_lock); +__cacheline_aligned_in_smp DEFINE_SPINLOCK(dq_data_lock); +EXPORT_SYMBOL(dq_data_lock); + +static char *quotatypes[] = INITQFNAMES; +static struct quota_format_type *quota_formats; /* List of registered formats */ +static struct quota_module_name module_names[] = INIT_QUOTA_MODULE_NAMES; + +/* SLAB cache for dquot structures */ +static struct kmem_cache *dquot_cachep; + +int register_quota_format(struct quota_format_type *fmt) +{ + spin_lock(&dq_list_lock); + fmt->qf_next = quota_formats; + quota_formats = fmt; + spin_unlock(&dq_list_lock); + return 0; +} +EXPORT_SYMBOL(register_quota_format); + +void unregister_quota_format(struct quota_format_type *fmt) +{ + struct quota_format_type **actqf; + + spin_lock(&dq_list_lock); + for (actqf = "a_formats; *actqf && *actqf != fmt; + actqf = &(*actqf)->qf_next) + ; + if (*actqf) + *actqf = (*actqf)->qf_next; + spin_unlock(&dq_list_lock); +} +EXPORT_SYMBOL(unregister_quota_format); + +static struct quota_format_type *find_quota_format(int id) +{ + struct quota_format_type *actqf; + + spin_lock(&dq_list_lock); + for (actqf = quota_formats; actqf && actqf->qf_fmt_id != id; + actqf = actqf->qf_next) + ; + if (!actqf || !try_module_get(actqf->qf_owner)) { + int qm; + + spin_unlock(&dq_list_lock); + + for (qm = 0; module_names[qm].qm_fmt_id && + module_names[qm].qm_fmt_id != id; qm++) + ; + if (!module_names[qm].qm_fmt_id || + request_module(module_names[qm].qm_mod_name)) + return NULL; + + spin_lock(&dq_list_lock); + for (actqf = quota_formats; actqf && actqf->qf_fmt_id != id; + actqf = actqf->qf_next) + ; + if (actqf && !try_module_get(actqf->qf_owner)) + actqf = NULL; + } + spin_unlock(&dq_list_lock); + return actqf; +} + +static void put_quota_format(struct quota_format_type *fmt) +{ + module_put(fmt->qf_owner); +} + +/* + * Dquot List Management: + * The quota code uses three lists for dquot management: the inuse_list, + * free_dquots, and dquot_hash[] array. A single dquot structure may be + * on all three lists, depending on its current state. + * + * All dquots are placed to the end of inuse_list when first created, and this + * list is used for invalidate operation, which must look at every dquot. + * + * Unused dquots (dq_count == 0) are added to the free_dquots list when freed, + * and this list is searched whenever we need an available dquot. Dquots are + * removed from the list as soon as they are used again, and + * dqstats.free_dquots gives the number of dquots on the list. When + * dquot is invalidated it's completely released from memory. + * + * Dquots with a specific identity (device, type and id) are placed on + * one of the dquot_hash[] hash chains. The provides an efficient search + * mechanism to locate a specific dquot. + */ + +static LIST_HEAD(inuse_list); +static LIST_HEAD(free_dquots); +static unsigned int dq_hash_bits, dq_hash_mask; +static struct hlist_head *dquot_hash; + +struct dqstats dqstats; +EXPORT_SYMBOL(dqstats); + +static inline unsigned int +hashfn(const struct super_block *sb, unsigned int id, int type) +{ + unsigned long tmp; + + tmp = (((unsigned long)sb>>L1_CACHE_SHIFT) ^ id) * (MAXQUOTAS - type); + return (tmp + (tmp >> dq_hash_bits)) & dq_hash_mask; +} + +/* + * Following list functions expect dq_list_lock to be held + */ +static inline void insert_dquot_hash(struct dquot *dquot) +{ + struct hlist_head *head; + head = dquot_hash + hashfn(dquot->dq_sb, dquot->dq_id, dquot->dq_type); + hlist_add_head(&dquot->dq_hash, head); +} + +static inline void remove_dquot_hash(struct dquot *dquot) +{ + hlist_del_init(&dquot->dq_hash); +} + +static struct dquot *find_dquot(unsigned int hashent, struct super_block *sb, + unsigned int id, int type) +{ + struct hlist_node *node; + struct dquot *dquot; + + hlist_for_each (node, dquot_hash+hashent) { + dquot = hlist_entry(node, struct dquot, dq_hash); + if (dquot->dq_sb == sb && dquot->dq_id == id && + dquot->dq_type == type) + return dquot; + } + return NULL; +} + +/* Add a dquot to the tail of the free list */ +static inline void put_dquot_last(struct dquot *dquot) +{ + list_add_tail(&dquot->dq_free, &free_dquots); + dqstats.free_dquots++; +} + +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--; +} + +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++; +} + +static inline void remove_inuse(struct dquot *dquot) +{ + dqstats.allocated_dquots--; + list_del(&dquot->dq_inuse); +} +/* + * End of list functions needing dq_list_lock + */ + +static void wait_on_dquot(struct dquot *dquot) +{ + mutex_lock(&dquot->dq_lock); + mutex_unlock(&dquot->dq_lock); +} + +static inline int dquot_dirty(struct dquot *dquot) +{ + return test_bit(DQ_MOD_B, &dquot->dq_flags); +} + +static inline int mark_dquot_dirty(struct dquot *dquot) +{ + return dquot->dq_sb->dq_op->mark_dirty(dquot); +} + +int dquot_mark_dquot_dirty(struct dquot *dquot) +{ + spin_lock(&dq_list_lock); + if (!test_and_set_bit(DQ_MOD_B, &dquot->dq_flags)) + list_add(&dquot->dq_dirty, &sb_dqopt(dquot->dq_sb)-> + info[dquot->dq_type].dqi_dirty_list); + spin_unlock(&dq_list_lock); + return 0; +} +EXPORT_SYMBOL(dquot_mark_dquot_dirty); + +/* This function needs dq_list_lock */ +static inline int clear_dquot_dirty(struct dquot *dquot) +{ + if (!test_and_clear_bit(DQ_MOD_B, &dquot->dq_flags)) + return 0; + list_del_init(&dquot->dq_dirty); + return 1; +} + +void mark_info_dirty(struct super_block *sb, int type) +{ + set_bit(DQF_INFO_DIRTY_B, &sb_dqopt(sb)->info[type].dqi_flags); +} +EXPORT_SYMBOL(mark_info_dirty); + +/* + * Read dquot from disk and alloc space for it + */ + +int dquot_acquire(struct dquot *dquot) +{ + int ret = 0, ret2 = 0; + struct quota_info *dqopt = sb_dqopt(dquot->dq_sb); + + mutex_lock(&dquot->dq_lock); + mutex_lock(&dqopt->dqio_mutex); + if (!test_bit(DQ_READ_B, &dquot->dq_flags)) + ret = dqopt->ops[dquot->dq_type]->read_dqblk(dquot); + if (ret < 0) + goto out_iolock; + set_bit(DQ_READ_B, &dquot->dq_flags); + /* Instantiate dquot if needed */ + if (!test_bit(DQ_ACTIVE_B, &dquot->dq_flags) && !dquot->dq_off) { + ret = dqopt->ops[dquot->dq_type]->commit_dqblk(dquot); + /* Write the info if needed */ + if (info_dirty(&dqopt->info[dquot->dq_type])) { + ret2 = dqopt->ops[dquot->dq_type]->write_file_info( + dquot->dq_sb, dquot->dq_type); + } + if (ret < 0) + goto out_iolock; + if (ret2 < 0) { + ret = ret2; + goto out_iolock; + } + } + set_bit(DQ_ACTIVE_B, &dquot->dq_flags); +out_iolock: + mutex_unlock(&dqopt->dqio_mutex); + mutex_unlock(&dquot->dq_lock); + return ret; +} +EXPORT_SYMBOL(dquot_acquire); + +/* + * Write dquot to disk + */ +int dquot_commit(struct dquot *dquot) +{ + int ret = 0, ret2 = 0; + struct quota_info *dqopt = sb_dqopt(dquot->dq_sb); + + mutex_lock(&dqopt->dqio_mutex); + spin_lock(&dq_list_lock); + if (!clear_dquot_dirty(dquot)) { + spin_unlock(&dq_list_lock); + goto out_sem; + } + spin_unlock(&dq_list_lock); + /* Inactive dquot can be only if there was error during read/init + * => we have better not writing it */ + if (test_bit(DQ_ACTIVE_B, &dquot->dq_flags)) { + ret = dqopt->ops[dquot->dq_type]->commit_dqblk(dquot); + if (info_dirty(&dqopt->info[dquot->dq_type])) { + ret2 = dqopt->ops[dquot->dq_type]->write_file_info( + dquot->dq_sb, dquot->dq_type); + } + if (ret >= 0) + ret = ret2; + } +out_sem: + mutex_unlock(&dqopt->dqio_mutex); + return ret; +} +EXPORT_SYMBOL(dquot_commit); + +/* + * Release dquot + */ +int dquot_release(struct dquot *dquot) +{ + int ret = 0, ret2 = 0; + struct quota_info *dqopt = sb_dqopt(dquot->dq_sb); + + mutex_lock(&dquot->dq_lock); + /* Check whether we are not racing with some other dqget() */ + if (atomic_read(&dquot->dq_count) > 1) + goto out_dqlock; + mutex_lock(&dqopt->dqio_mutex); + if (dqopt->ops[dquot->dq_type]->release_dqblk) { + ret = dqopt->ops[dquot->dq_type]->release_dqblk(dquot); + /* Write the info */ + if (info_dirty(&dqopt->info[dquot->dq_type])) { + ret2 = dqopt->ops[dquot->dq_type]->write_file_info( + dquot->dq_sb, dquot->dq_type); + } + if (ret >= 0) + ret = ret2; + } + clear_bit(DQ_ACTIVE_B, &dquot->dq_flags); + mutex_unlock(&dqopt->dqio_mutex); +out_dqlock: + mutex_unlock(&dquot->dq_lock); + return ret; +} +EXPORT_SYMBOL(dquot_release); + +void dquot_destroy(struct dquot *dquot) +{ + kmem_cache_free(dquot_cachep, dquot); +} +EXPORT_SYMBOL(dquot_destroy); + +static inline void do_destroy_dquot(struct dquot *dquot) +{ + dquot->dq_sb->dq_op->destroy_dquot(dquot); +} + +/* Invalidate all dquots on the list. Note that this function is called after + * quota is disabled and pointers from inodes removed so there cannot be new + * quota users. There can still be some users of quotas due to inodes being + * just deleted or pruned by prune_icache() (those are not attached to any + * list) or parallel quotactl call. We have to wait for such users. + */ +static void invalidate_dquots(struct super_block *sb, int type) +{ + struct dquot *dquot, *tmp; + +restart: + spin_lock(&dq_list_lock); + list_for_each_entry_safe(dquot, tmp, &inuse_list, dq_inuse) { + if (dquot->dq_sb != sb) + continue; + if (dquot->dq_type != type) + continue; + /* Wait for dquot users */ + if (atomic_read(&dquot->dq_count)) { + DEFINE_WAIT(wait); + + atomic_inc(&dquot->dq_count); + prepare_to_wait(&dquot->dq_wait_unused, &wait, + TASK_UNINTERRUPTIBLE); + spin_unlock(&dq_list_lock); + /* Once dqput() wakes us up, we know it's time to free + * the dquot. + * IMPORTANT: we rely on the fact that there is always + * at most one process waiting for dquot to free. + * Otherwise dq_count would be > 1 and we would never + * wake up. + */ + if (atomic_read(&dquot->dq_count) > 1) + schedule(); + finish_wait(&dquot->dq_wait_unused, &wait); + dqput(dquot); + /* At this moment dquot() need not exist (it could be + * reclaimed by prune_dqcache(). Hence we must + * restart. */ + goto restart; + } + /* + * Quota now has no users and it has been written on last + * dqput() + */ + remove_dquot_hash(dquot); + remove_free_dquot(dquot); + remove_inuse(dquot); + do_destroy_dquot(dquot); + } + spin_unlock(&dq_list_lock); +} + +/* Call callback for every active dquot on given filesystem */ +int dquot_scan_active(struct super_block *sb, + int (*fn)(struct dquot *dquot, unsigned long priv), + unsigned long priv) +{ + struct dquot *dquot, *old_dquot = NULL; + int ret = 0; + + mutex_lock(&sb_dqopt(sb)->dqonoff_mutex); + spin_lock(&dq_list_lock); + list_for_each_entry(dquot, &inuse_list, dq_inuse) { + if (!test_bit(DQ_ACTIVE_B, &dquot->dq_flags)) + continue; + if (dquot->dq_sb != 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); + dqput(old_dquot); + old_dquot = dquot; + ret = fn(dquot, priv); + if (ret < 0) + goto out; + spin_lock(&dq_list_lock); + /* We are safe to continue now because our dquot could not + * be moved out of the inuse list while we hold the reference */ + } + spin_unlock(&dq_list_lock); +out: + dqput(old_dquot); + mutex_unlock(&sb_dqopt(sb)->dqonoff_mutex); + return ret; +} +EXPORT_SYMBOL(dquot_scan_active); + +int vfs_quota_sync(struct super_block *sb, int type) +{ + struct list_head *dirty; + struct dquot *dquot; + struct quota_info *dqopt = sb_dqopt(sb); + int cnt; + + mutex_lock(&dqopt->dqonoff_mutex); + for (cnt = 0; cnt < MAXQUOTAS; cnt++) { + if (type != -1 && cnt != type) + continue; + if (!sb_has_quota_active(sb, cnt)) + continue; + spin_lock(&dq_list_lock); + dirty = &dqopt->info[cnt].dqi_dirty_list; + while (!list_empty(dirty)) { + dquot = list_first_entry(dirty, struct dquot, + dq_dirty); + /* Dirty and inactive can be only bad dquot... */ + if (!test_bit(DQ_ACTIVE_B, &dquot->dq_flags)) { + clear_dquot_dirty(dquot); + continue; + } + /* Now we have active dquot from which someone is + * holding reference so we can safely just increase + * use count */ + atomic_inc(&dquot->dq_count); + dqstats.lookups++; + spin_unlock(&dq_list_lock); + sb->dq_op->write_dquot(dquot); + dqput(dquot); + spin_lock(&dq_list_lock); + } + spin_unlock(&dq_list_lock); + } + + for (cnt = 0; cnt < MAXQUOTAS; cnt++) + 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); + mutex_unlock(&dqopt->dqonoff_mutex); + + return 0; +} +EXPORT_SYMBOL(vfs_quota_sync); + +/* Free unused dquots from cache */ +static void prune_dqcache(int count) +{ + struct list_head *head; + struct dquot *dquot; + + head = free_dquots.prev; + while (head != &free_dquots && count) { + dquot = list_entry(head, struct dquot, dq_free); + remove_dquot_hash(dquot); + remove_free_dquot(dquot); + remove_inuse(dquot); + do_destroy_dquot(dquot); + count--; + head = free_dquots.prev; + } +} + +/* + * This is called from kswapd when we think we need some + * more memory + */ + +static int shrink_dqcache_memory(int nr, gfp_t gfp_mask) +{ + if (nr) { + spin_lock(&dq_list_lock); + prune_dqcache(nr); + spin_unlock(&dq_list_lock); + } + return (dqstats.free_dquots / 100) * sysctl_vfs_cache_pressure; +} + +static struct shrinker dqcache_shrinker = { + .shrink = shrink_dqcache_memory, + .seeks = DEFAULT_SEEKS, +}; + +/* + * Put reference to dquot + * NOTE: If you change this function please check whether dqput_blocks() works right... + */ +void dqput(struct dquot *dquot) +{ + int ret; + + if (!dquot) + return; +#ifdef __DQUOT_PARANOIA + if (!atomic_read(&dquot->dq_count)) { + printk("VFS: dqput: trying to free free dquot\n"); + printk("VFS: device %s, dquot of %s %d\n", + dquot->dq_sb->s_id, + quotatypes[dquot->dq_type], + dquot->dq_id); + BUG(); + } +#endif + + spin_lock(&dq_list_lock); + dqstats.drops++; + spin_unlock(&dq_list_lock); +we_slept: + spin_lock(&dq_list_lock); + if (atomic_read(&dquot->dq_count) > 1) { + /* We have more than one user... nothing to do */ + atomic_dec(&dquot->dq_count); + /* Releasing dquot during quotaoff phase? */ + if (!sb_has_quota_active(dquot->dq_sb, dquot->dq_type) && + atomic_read(&dquot->dq_count) == 1) + wake_up(&dquot->dq_wait_unused); + spin_unlock(&dq_list_lock); + return; + } + /* Need to release dquot? */ + if (test_bit(DQ_ACTIVE_B, &dquot->dq_flags) && dquot_dirty(dquot)) { + spin_unlock(&dq_list_lock); + /* Commit dquot before releasing */ + ret = dquot->dq_sb->dq_op->write_dquot(dquot); + if (ret < 0) { + printk(KERN_ERR "VFS: cannot write quota structure on " + "device %s (error %d). Quota may get out of " + "sync!\n", dquot->dq_sb->s_id, ret); + /* + * We clear dirty bit anyway, so that we avoid + * infinite loop here + */ + spin_lock(&dq_list_lock); + clear_dquot_dirty(dquot); + spin_unlock(&dq_list_lock); + } + goto we_slept; + } + /* Clear flag in case dquot was inactive (something bad happened) */ + clear_dquot_dirty(dquot); + if (test_bit(DQ_ACTIVE_B, &dquot->dq_flags)) { + spin_unlock(&dq_list_lock); + dquot->dq_sb->dq_op->release_dquot(dquot); + goto we_slept; + } + atomic_dec(&dquot->dq_count); +#ifdef __DQUOT_PARANOIA + /* sanity check */ + BUG_ON(!list_empty(&dquot->dq_free)); +#endif + put_dquot_last(dquot); + spin_unlock(&dq_list_lock); +} +EXPORT_SYMBOL(dqput); + +struct dquot *dquot_alloc(struct super_block *sb, int type) +{ + return kmem_cache_zalloc(dquot_cachep, GFP_NOFS); +} +EXPORT_SYMBOL(dquot_alloc); + +static struct dquot *get_empty_dquot(struct super_block *sb, int type) +{ + struct dquot *dquot; + + dquot = sb->dq_op->alloc_dquot(sb, type); + if(!dquot) + return NULL; + + mutex_init(&dquot->dq_lock); + INIT_LIST_HEAD(&dquot->dq_free); + INIT_LIST_HEAD(&dquot->dq_inuse); + INIT_HLIST_NODE(&dquot->dq_hash); + INIT_LIST_HEAD(&dquot->dq_dirty); + init_waitqueue_head(&dquot->dq_wait_unused); + dquot->dq_sb = sb; + dquot->dq_type = type; + atomic_set(&dquot->dq_count, 1); + + return dquot; +} + +/* + * Get reference to dquot + * + * Locking is slightly tricky here. We are guarded from parallel quotaoff() + * destroying our dquot by: + * a) checking for quota flags under dq_list_lock and + * b) getting a reference to dquot before we release dq_list_lock + */ +struct dquot *dqget(struct super_block *sb, unsigned int id, int type) +{ + unsigned int hashent = hashfn(sb, id, type); + struct dquot *dquot = NULL, *empty = NULL; + + if (!sb_has_quota_active(sb, type)) + return NULL; +we_slept: + spin_lock(&dq_list_lock); + spin_lock(&dq_state_lock); + if (!sb_has_quota_active(sb, type)) { + spin_unlock(&dq_state_lock); + spin_unlock(&dq_list_lock); + goto out; + } + spin_unlock(&dq_state_lock); + + dquot = find_dquot(hashent, sb, id, type); + if (!dquot) { + if (!empty) { + spin_unlock(&dq_list_lock); + empty = get_empty_dquot(sb, type); + if (!empty) + schedule(); /* Try to wait for a moment... */ + goto we_slept; + } + dquot = empty; + empty = NULL; + dquot->dq_id = id; + /* all dquots go on the inuse_list */ + put_inuse(dquot); + /* hash it first so it can be found */ + insert_dquot_hash(dquot); + dqstats.lookups++; + spin_unlock(&dq_list_lock); + } 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); + } + /* 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 */ + wait_on_dquot(dquot); + /* Read the dquot / allocate space in quota file */ + if (!test_bit(DQ_ACTIVE_B, &dquot->dq_flags) && + sb->dq_op->acquire_dquot(dquot) < 0) { + dqput(dquot); + dquot = NULL; + goto out; + } +#ifdef __DQUOT_PARANOIA + BUG_ON(!dquot->dq_sb); /* Has somebody invalidated entry under us? */ +#endif +out: + if (empty) + do_destroy_dquot(empty); + + return dquot; +} +EXPORT_SYMBOL(dqget); + +static int dqinit_needed(struct inode *inode, int type) +{ + int cnt; + + if (IS_NOQUOTA(inode)) + return 0; + if (type != -1) + return !inode->i_dquot[type]; + for (cnt = 0; cnt < MAXQUOTAS; cnt++) + if (!inode->i_dquot[cnt]) + return 1; + return 0; +} + +/* This routine is guarded by dqonoff_mutex mutex */ +static void add_dquot_ref(struct super_block *sb, int type) +{ + struct inode *inode, *old_inode = NULL; + + spin_lock(&inode_lock); + list_for_each_entry(inode, &sb->s_inodes, i_sb_list) { + if (inode->i_state & (I_FREEING|I_WILL_FREE|I_NEW)) + continue; + if (!atomic_read(&inode->i_writecount)) + continue; + if (!dqinit_needed(inode, type)) + continue; + + __iget(inode); + spin_unlock(&inode_lock); + + iput(old_inode); + sb->dq_op->initialize(inode, type); + /* We hold a reference to 'inode' so it couldn't have been + * removed from s_inodes list while we dropped the inode_lock. + * We cannot iput the inode now as we can be holding the last + * reference and we cannot iput it under inode_lock. So we + * keep the reference and iput it later. */ + old_inode = inode; + spin_lock(&inode_lock); + } + spin_unlock(&inode_lock); + iput(old_inode); +} + +/* + * Return 0 if dqput() won't block. + * (note that 1 doesn't necessarily mean blocking) + */ +static inline int dqput_blocks(struct dquot *dquot) +{ + if (atomic_read(&dquot->dq_count) <= 1) + return 1; + return 0; +} + +/* + * Remove references to dquots from inode and add dquot to list for freeing + * if we have the last referece to dquot + * We can't race with anybody because we hold dqptr_sem for writing... + */ +static int remove_inode_dquot_ref(struct inode *inode, int type, + struct list_head *tofree_head) +{ + struct dquot *dquot = inode->i_dquot[type]; + + inode->i_dquot[type] = NULL; + if (dquot) { + if (dqput_blocks(dquot)) { +#ifdef __DQUOT_PARANOIA + if (atomic_read(&dquot->dq_count) != 1) + printk(KERN_WARNING "VFS: Adding dquot with dq_count %d to dispose list.\n", atomic_read(&dquot->dq_count)); +#endif + spin_lock(&dq_list_lock); + /* As dquot must have currently users it can't be on + * the free list... */ + list_add(&dquot->dq_free, tofree_head); + spin_unlock(&dq_list_lock); + return 1; + } + else + dqput(dquot); /* We have guaranteed we won't block */ + } + return 0; +} + +/* + * Free list of dquots + * Dquots are removed from inodes and no new references can be got so we are + * the only ones holding reference + */ +static void put_dquot_list(struct list_head *tofree_head) +{ + struct list_head *act_head; + struct dquot *dquot; + + act_head = tofree_head->next; + while (act_head != tofree_head) { + dquot = list_entry(act_head, struct dquot, dq_free); + act_head = act_head->next; + /* Remove dquot from the list so we won't have problems... */ + list_del_init(&dquot->dq_free); + dqput(dquot); + } +} + +static void remove_dquot_ref(struct super_block *sb, int type, + struct list_head *tofree_head) +{ + struct inode *inode; + + spin_lock(&inode_lock); + list_for_each_entry(inode, &sb->s_inodes, i_sb_list) { + /* + * We have to scan also I_NEW inodes because they can already + * have quota pointer initialized. Luckily, we need to touch + * only quota pointers and these have separate locking + * (dqptr_sem). + */ + if (!IS_NOQUOTA(inode)) + remove_inode_dquot_ref(inode, type, tofree_head); + } + spin_unlock(&inode_lock); +} + +/* Gather all references from inodes and drop them */ +static void drop_dquot_ref(struct super_block *sb, int type) +{ + LIST_HEAD(tofree_head); + + if (sb->dq_op) { + down_write(&sb_dqopt(sb)->dqptr_sem); + remove_dquot_ref(sb, type, &tofree_head); + up_write(&sb_dqopt(sb)->dqptr_sem); + put_dquot_list(&tofree_head); + } +} + +static inline void dquot_incr_inodes(struct dquot *dquot, qsize_t number) +{ + dquot->dq_dqb.dqb_curinodes += number; +} + +static inline void dquot_incr_space(struct dquot *dquot, qsize_t number) +{ + dquot->dq_dqb.dqb_curspace += number; +} + +static inline void dquot_resv_space(struct dquot *dquot, qsize_t number) +{ + dquot->dq_dqb.dqb_rsvspace += number; +} + +/* + * Claim reserved quota space + */ |