kernel-fxtec-pro1x/fs/quota.c
Kirill Korotaev 618f06362a [PATCH] O(1) sb list traversing on syncs
This patch removes O(n^2) super block loops in sync_inodes(),
sync_filesystems() etc.  in favour of using __put_super_and_need_restart()
which I introduced earlier.  We faced a noticably long freezes on sb
syncing when there are thousands of super blocks in the system.

Signed-Off-By: Kirill Korotaev <dev@sw.ru>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-23 09:45:27 -07:00

370 lines
8.6 KiB
C

/*
* Quota code necessary even when VFS quota support is not compiled
* into the kernel. The interesting stuff is over in dquot.c, here
* we have symbols for initial quotactl(2) handling, the sysctl(2)
* variables, etc - things needed even when quota support disabled.
*/
#include <linux/fs.h>
#include <linux/namei.h>
#include <linux/slab.h>
#include <asm/current.h>
#include <asm/uaccess.h>
#include <linux/kernel.h>
#include <linux/smp_lock.h>
#include <linux/security.h>
#include <linux/syscalls.h>
#include <linux/buffer_head.h>
/* Check validity of generic quotactl commands */
static int generic_quotactl_valid(struct super_block *sb, int type, int cmd, qid_t id)
{
if (type >= MAXQUOTAS)
return -EINVAL;
if (!sb && cmd != Q_SYNC)
return -ENODEV;
/* Is operation supported? */
if (sb && !sb->s_qcop)
return -ENOSYS;
switch (cmd) {
case Q_GETFMT:
break;
case Q_QUOTAON:
if (!sb->s_qcop->quota_on)
return -ENOSYS;
break;
case Q_QUOTAOFF:
if (!sb->s_qcop->quota_off)
return -ENOSYS;
break;
case Q_SETINFO:
if (!sb->s_qcop->set_info)
return -ENOSYS;
break;
case Q_GETINFO:
if (!sb->s_qcop->get_info)
return -ENOSYS;
break;
case Q_SETQUOTA:
if (!sb->s_qcop->set_dqblk)
return -ENOSYS;
break;
case Q_GETQUOTA:
if (!sb->s_qcop->get_dqblk)
return -ENOSYS;
break;
case Q_SYNC:
if (sb && !sb->s_qcop->quota_sync)
return -ENOSYS;
break;
default:
return -EINVAL;
}
/* Is quota turned on for commands which need it? */
switch (cmd) {
case Q_GETFMT:
case Q_GETINFO:
case Q_QUOTAOFF:
case Q_SETINFO:
case Q_SETQUOTA:
case Q_GETQUOTA:
/* This is just informative test so we are satisfied without a lock */
if (!sb_has_quota_enabled(sb, type))
return -ESRCH;
}
/* Check privileges */
if (cmd == Q_GETQUOTA) {
if (((type == USRQUOTA && current->euid != id) ||
(type == GRPQUOTA && !in_egroup_p(id))) &&
!capable(CAP_SYS_ADMIN))
return -EPERM;
}
else if (cmd != Q_GETFMT && cmd != Q_SYNC && cmd != Q_GETINFO)
if (!capable(CAP_SYS_ADMIN))
return -EPERM;
return 0;
}
/* Check validity of XFS Quota Manager commands */
static int xqm_quotactl_valid(struct super_block *sb, int type, int cmd, qid_t id)
{
if (type >= XQM_MAXQUOTAS)
return -EINVAL;
if (!sb)
return -ENODEV;
if (!sb->s_qcop)
return -ENOSYS;
switch (cmd) {
case Q_XQUOTAON:
case Q_XQUOTAOFF:
case Q_XQUOTARM:
if (!sb->s_qcop->set_xstate)
return -ENOSYS;
break;
case Q_XGETQSTAT:
if (!sb->s_qcop->get_xstate)
return -ENOSYS;
break;
case Q_XSETQLIM:
if (!sb->s_qcop->set_xquota)
return -ENOSYS;
break;
case Q_XGETQUOTA:
if (!sb->s_qcop->get_xquota)
return -ENOSYS;
break;
default:
return -EINVAL;
}
/* Check privileges */
if (cmd == Q_XGETQUOTA) {
if (((type == XQM_USRQUOTA && current->euid != id) ||
(type == XQM_GRPQUOTA && !in_egroup_p(id))) &&
!capable(CAP_SYS_ADMIN))
return -EPERM;
} else if (cmd != Q_XGETQSTAT) {
if (!capable(CAP_SYS_ADMIN))
return -EPERM;
}
return 0;
}
static int check_quotactl_valid(struct super_block *sb, int type, int cmd, qid_t id)
{
int error;
if (XQM_COMMAND(cmd))
error = xqm_quotactl_valid(sb, type, cmd, id);
else
error = generic_quotactl_valid(sb, type, cmd, id);
if (!error)
error = security_quotactl(cmd, type, id, sb);
return error;
}
static void quota_sync_sb(struct super_block *sb, int type)
{
int cnt;
struct inode *discard[MAXQUOTAS];
sb->s_qcop->quota_sync(sb, type);
/* This is not very clever (and fast) but currently I don't know about
* any other simple way of getting quota data to disk and we must get
* them there for userspace to be visible... */
if (sb->s_op->sync_fs)
sb->s_op->sync_fs(sb, 1);
sync_blockdev(sb->s_bdev);
/* Now when everything is written we can discard the pagecache so
* that userspace sees the changes. We need i_sem and so we could
* not do it inside dqonoff_sem. Moreover we need to be carefull
* about races with quotaoff() (that is the reason why we have own
* reference to inode). */
down(&sb_dqopt(sb)->dqonoff_sem);
for (cnt = 0; cnt < MAXQUOTAS; cnt++) {
discard[cnt] = NULL;
if (type != -1 && cnt != type)
continue;
if (!sb_has_quota_enabled(sb, cnt))
continue;
discard[cnt] = igrab(sb_dqopt(sb)->files[cnt]);
}
up(&sb_dqopt(sb)->dqonoff_sem);
for (cnt = 0; cnt < MAXQUOTAS; cnt++) {
if (discard[cnt]) {
down(&discard[cnt]->i_sem);
truncate_inode_pages(&discard[cnt]->i_data, 0);
up(&discard[cnt]->i_sem);
iput(discard[cnt]);
}
}
}
void sync_dquots(struct super_block *sb, int type)
{
int cnt, dirty;
if (sb) {
if (sb->s_qcop->quota_sync)
quota_sync_sb(sb, type);
return;
}
spin_lock(&sb_lock);
restart:
list_for_each_entry(sb, &super_blocks, s_list) {
/* This test just improves performance so it needn't be reliable... */
for (cnt = 0, dirty = 0; cnt < MAXQUOTAS; cnt++)
if ((type == cnt || type == -1) && sb_has_quota_enabled(sb, cnt)
&& info_any_dirty(&sb_dqopt(sb)->info[cnt]))
dirty = 1;
if (!dirty)
continue;
sb->s_count++;
spin_unlock(&sb_lock);
down_read(&sb->s_umount);
if (sb->s_root && sb->s_qcop->quota_sync)
quota_sync_sb(sb, type);
up_read(&sb->s_umount);
spin_lock(&sb_lock);
if (__put_super_and_need_restart(sb))
goto restart;
}
spin_unlock(&sb_lock);
}
/* Copy parameters and call proper function */
static int do_quotactl(struct super_block *sb, int type, int cmd, qid_t id, void __user *addr)
{
int ret;
switch (cmd) {
case Q_QUOTAON: {
char *pathname;
if (IS_ERR(pathname = getname(addr)))
return PTR_ERR(pathname);
ret = sb->s_qcop->quota_on(sb, type, id, pathname);
putname(pathname);
return ret;
}
case Q_QUOTAOFF:
return sb->s_qcop->quota_off(sb, type);
case Q_GETFMT: {
__u32 fmt;
down_read(&sb_dqopt(sb)->dqptr_sem);
if (!sb_has_quota_enabled(sb, type)) {
up_read(&sb_dqopt(sb)->dqptr_sem);
return -ESRCH;
}
fmt = sb_dqopt(sb)->info[type].dqi_format->qf_fmt_id;
up_read(&sb_dqopt(sb)->dqptr_sem);
if (copy_to_user(addr, &fmt, sizeof(fmt)))
return -EFAULT;
return 0;
}
case Q_GETINFO: {
struct if_dqinfo info;
if ((ret = sb->s_qcop->get_info(sb, type, &info)))
return ret;
if (copy_to_user(addr, &info, sizeof(info)))
return -EFAULT;
return 0;
}
case Q_SETINFO: {
struct if_dqinfo info;
if (copy_from_user(&info, addr, sizeof(info)))
return -EFAULT;
return sb->s_qcop->set_info(sb, type, &info);
}
case Q_GETQUOTA: {
struct if_dqblk idq;
if ((ret = sb->s_qcop->get_dqblk(sb, type, id, &idq)))
return ret;
if (copy_to_user(addr, &idq, sizeof(idq)))
return -EFAULT;
return 0;
}
case Q_SETQUOTA: {
struct if_dqblk idq;
if (copy_from_user(&idq, addr, sizeof(idq)))
return -EFAULT;
return sb->s_qcop->set_dqblk(sb, type, id, &idq);
}
case Q_SYNC:
sync_dquots(sb, type);
return 0;
case Q_XQUOTAON:
case Q_XQUOTAOFF:
case Q_XQUOTARM: {
__u32 flags;
if (copy_from_user(&flags, addr, sizeof(flags)))
return -EFAULT;
return sb->s_qcop->set_xstate(sb, flags, cmd);
}
case Q_XGETQSTAT: {
struct fs_quota_stat fqs;
if ((ret = sb->s_qcop->get_xstate(sb, &fqs)))
return ret;
if (copy_to_user(addr, &fqs, sizeof(fqs)))
return -EFAULT;
return 0;
}
case Q_XSETQLIM: {
struct fs_disk_quota fdq;
if (copy_from_user(&fdq, addr, sizeof(fdq)))
return -EFAULT;
return sb->s_qcop->set_xquota(sb, type, id, &fdq);
}
case Q_XGETQUOTA: {
struct fs_disk_quota fdq;
if ((ret = sb->s_qcop->get_xquota(sb, type, id, &fdq)))
return ret;
if (copy_to_user(addr, &fdq, sizeof(fdq)))
return -EFAULT;
return 0;
}
/* We never reach here unless validity check is broken */
default:
BUG();
}
return 0;
}
/*
* This is the system call interface. This communicates with
* the user-level programs. Currently this only supports diskquota
* calls. Maybe we need to add the process quotas etc. in the future,
* but we probably should use rlimits for that.
*/
asmlinkage long sys_quotactl(unsigned int cmd, const char __user *special, qid_t id, void __user *addr)
{
uint cmds, type;
struct super_block *sb = NULL;
struct block_device *bdev;
char *tmp;
int ret;
cmds = cmd >> SUBCMDSHIFT;
type = cmd & SUBCMDMASK;
if (cmds != Q_SYNC || special) {
tmp = getname(special);
if (IS_ERR(tmp))
return PTR_ERR(tmp);
bdev = lookup_bdev(tmp);
putname(tmp);
if (IS_ERR(bdev))
return PTR_ERR(bdev);
sb = get_super(bdev);
bdput(bdev);
if (!sb)
return -ENODEV;
}
ret = check_quotactl_valid(sb, type, cmds, id);
if (ret >= 0)
ret = do_quotactl(sb, type, cmds, id, addr);
if (sb)
drop_super(sb);
return ret;
}