Global File System
------------------
-http://sources.redhat.com/cluster/
+http://sources.redhat.com/cluster/wiki/
GFS is a cluster file system. It allows a cluster of computers to
simultaneously use a block device that is shared between them (with FC,
is pretty close.
The following man pages can be found at the URL above:
- fsck.gfs2 to repair a filesystem
- gfs2_grow to expand a filesystem online
- gfs2_jadd to add journals to a filesystem online
- gfs2_tool to manipulate, examine and tune a filesystem
+ fsck.gfs2 to repair a filesystem
+ gfs2_grow to expand a filesystem online
+ gfs2_jadd to add journals to a filesystem online
+ gfs2_tool to manipulate, examine and tune a filesystem
gfs2_quota to examine and change quota values in a filesystem
gfs2_convert to convert a gfs filesystem to gfs2 in-place
mount.gfs2 to help mount(8) mount a filesystem
- mkfs.gfs2 to make a filesystem
+ mkfs.gfs2 to make a filesystem
static int stuffed_readpage(struct gfs2_inode *ip, struct page *page)
{
struct buffer_head *dibh;
+ u64 dsize = i_size_read(&ip->i_inode);
void *kaddr;
int error;
return error;
kaddr = kmap_atomic(page, KM_USER0);
- memcpy(kaddr, dibh->b_data + sizeof(struct gfs2_dinode),
- ip->i_disksize);
- memset(kaddr + ip->i_disksize, 0, PAGE_CACHE_SIZE - ip->i_disksize);
+ if (dsize > (dibh->b_size - sizeof(struct gfs2_dinode)))
+ dsize = (dibh->b_size - sizeof(struct gfs2_dinode));
+ memcpy(kaddr, dibh->b_data + sizeof(struct gfs2_dinode), dsize);
+ memset(kaddr + dsize, 0, PAGE_CACHE_SIZE - dsize);
kunmap_atomic(kaddr, KM_USER0);
flush_dcache_page(page);
brelse(dibh);
if (!PageUptodate(page)) {
void *kaddr = kmap(page);
+ u64 dsize = i_size_read(inode);
+
+ if (dsize > (dibh->b_size - sizeof(struct gfs2_dinode)))
+ dsize = dibh->b_size - sizeof(struct gfs2_dinode);
- memcpy(kaddr, dibh->b_data + sizeof(struct gfs2_dinode),
- ip->i_disksize);
- memset(kaddr + ip->i_disksize, 0,
- PAGE_CACHE_SIZE - ip->i_disksize);
+ memcpy(kaddr, dibh->b_data + sizeof(struct gfs2_dinode), dsize);
+ memset(kaddr + dsize, 0, PAGE_CACHE_SIZE - dsize);
kunmap(page);
SetPageUptodate(page);
goto out;
if (gfs2_is_stuffed(ip)) {
- ip->i_disksize = size;
+ u64 dsize = size + sizeof(struct gfs2_inode);
ip->i_inode.i_mtime = ip->i_inode.i_ctime = CURRENT_TIME;
gfs2_trans_add_bh(ip->i_gl, dibh, 1);
gfs2_dinode_out(ip, dibh->b_data);
- gfs2_buffer_clear_tail(dibh, sizeof(struct gfs2_dinode) + size);
+ if (dsize > dibh->b_size)
+ dsize = dibh->b_size;
+ gfs2_buffer_clear_tail(dibh, dsize);
error = 1;
-
} else {
if (size & (u64)(sdp->sd_sb.sb_bsize - 1))
error = gfs2_block_truncate_page(ip->i_inode.i_mapping);
inode = gfs2_inode_lookup(dir->i_sb,
be16_to_cpu(dent->de_type),
be64_to_cpu(dent->de_inum.no_addr),
- be64_to_cpu(dent->de_inum.no_formal_ino), 0);
+ be64_to_cpu(dent->de_inum.no_formal_ino));
brelse(bh);
return inode;
}
if (error)
goto fail;
- inode = gfs2_inode_lookup(sb, DT_UNKNOWN, inum->no_addr, 0, 0);
+ inode = gfs2_inode_lookup(sb, DT_UNKNOWN, inum->no_addr, 0);
if (IS_ERR(inode)) {
error = PTR_ERR(inode);
goto fail;
gh->gh_flags = flags;
gh->gh_iflags = 0;
gh->gh_ip = (unsigned long)__builtin_return_address(0);
+ if (gh->gh_owner_pid)
+ put_pid(gh->gh_owner_pid);
+ gh->gh_owner_pid = get_pid(task_pid(current));
}
/**
struct gfs2_tune {
spinlock_t gt_spin;
- unsigned int gt_incore_log_blocks;
- unsigned int gt_log_flush_secs;
-
unsigned int gt_logd_secs;
unsigned int gt_quota_simul_sync; /* Max quotavals to sync at once */
SDF_SHUTDOWN = 2,
SDF_NOBARRIERS = 3,
SDF_NORECOVERY = 4,
+ SDF_DEMOTE = 5,
};
#define GFS2_FSNAME_LEN 256
unsigned int sd_log_commited_databuf;
int sd_log_commited_revoke;
+ atomic_t sd_log_pinned;
unsigned int sd_log_num_buf;
unsigned int sd_log_num_revoke;
unsigned int sd_log_num_rg;
struct list_head sd_log_le_databuf;
struct list_head sd_log_le_ordered;
+ atomic_t sd_log_thresh1;
+ atomic_t sd_log_thresh2;
atomic_t sd_log_blks_free;
- struct mutex sd_log_reserve_mutex;
+ wait_queue_head_t sd_log_waitq;
+ wait_queue_head_t sd_logd_waitq;
u64 sd_log_sequence;
unsigned int sd_log_head;
unsigned int sd_log_tail;
int sd_log_idle;
- unsigned long sd_log_flush_time;
struct rw_semaphore sd_log_flush_lock;
atomic_t sd_log_in_flight;
wait_queue_head_t sd_log_flush_wait;
* @sb: The super block
* @no_addr: The inode number
* @type: The type of the inode
- * @skip_freeing: set this not return an inode if it is currently being freed.
*
* Returns: A VFS inode, or an error
*/
struct inode *gfs2_inode_lookup(struct super_block *sb,
unsigned int type,
u64 no_addr,
- u64 no_formal_ino, int skip_freeing)
+ u64 no_formal_ino)
{
struct inode *inode;
struct gfs2_inode *ip;
struct gfs2_glock *io_gl;
int error;
- if (skip_freeing)
- inode = gfs2_iget_skip(sb, no_addr);
- else
- inode = gfs2_iget(sb, no_addr);
+ inode = gfs2_iget(sb, no_addr);
ip = GFS2_I(inode);
if (!inode)
fail_iopen:
gfs2_glock_put(io_gl);
fail_put:
- ip->i_gl->gl_object = NULL;
+ if (inode->i_state & I_NEW)
+ ip->i_gl->gl_object = NULL;
gfs2_glock_put(ip->i_gl);
fail:
- iget_failed(inode);
+ if (inode->i_state & I_NEW)
+ iget_failed(inode);
+ else
+ iput(inode);
return ERR_PTR(error);
}
+/**
+ * gfs2_unlinked_inode_lookup - Lookup an unlinked inode for reclamation
+ * @sb: The super block
+ * no_addr: The inode number
+ * @@inode: A pointer to the inode found, if any
+ *
+ * Returns: 0 and *inode if no errors occurred. If an error occurs,
+ * the resulting *inode may or may not be NULL.
+ */
+
+int gfs2_unlinked_inode_lookup(struct super_block *sb, u64 no_addr,
+ struct inode **inode)
+{
+ struct gfs2_sbd *sdp;
+ struct gfs2_inode *ip;
+ struct gfs2_glock *io_gl;
+ int error;
+ struct gfs2_holder gh;
+
+ *inode = gfs2_iget_skip(sb, no_addr);
+
+ if (!(*inode))
+ return -ENOBUFS;
+
+ if (!((*inode)->i_state & I_NEW))
+ return -ENOBUFS;
+
+ ip = GFS2_I(*inode);
+ sdp = GFS2_SB(*inode);
+ ip->i_no_formal_ino = -1;
+
+ error = gfs2_glock_get(sdp, no_addr, &gfs2_inode_glops, CREATE, &ip->i_gl);
+ if (unlikely(error))
+ goto fail;
+ ip->i_gl->gl_object = ip;
+
+ error = gfs2_glock_get(sdp, no_addr, &gfs2_iopen_glops, CREATE, &io_gl);
+ if (unlikely(error))
+ goto fail_put;
+
+ set_bit(GIF_INVALID, &ip->i_flags);
+ error = gfs2_glock_nq_init(io_gl, LM_ST_SHARED, LM_FLAG_TRY | GL_EXACT,
+ &ip->i_iopen_gh);
+ if (unlikely(error)) {
+ if (error == GLR_TRYFAILED)
+ error = 0;
+ goto fail_iopen;
+ }
+ ip->i_iopen_gh.gh_gl->gl_object = ip;
+ gfs2_glock_put(io_gl);
+
+ (*inode)->i_mode = DT2IF(DT_UNKNOWN);
+
+ /*
+ * We must read the inode in order to work out its type in
+ * this case. Note that this doesn't happen often as we normally
+ * know the type beforehand. This code path only occurs during
+ * unlinked inode recovery (where it is safe to do this glock,
+ * which is not true in the general case).
+ */
+ error = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, LM_FLAG_TRY,
+ &gh);
+ if (unlikely(error)) {
+ if (error == GLR_TRYFAILED)
+ error = 0;
+ goto fail_glock;
+ }
+ /* Inode is now uptodate */
+ gfs2_glock_dq_uninit(&gh);
+ gfs2_set_iop(*inode);
+
+ return 0;
+fail_glock:
+ gfs2_glock_dq(&ip->i_iopen_gh);
+fail_iopen:
+ gfs2_glock_put(io_gl);
+fail_put:
+ ip->i_gl->gl_object = NULL;
+ gfs2_glock_put(ip->i_gl);
+fail:
+ return error;
+}
+
static int gfs2_dinode_in(struct gfs2_inode *ip, const void *buf)
{
const struct gfs2_dinode *str = buf;
goto fail_gunlock2;
inode = gfs2_inode_lookup(dir->i_sb, IF2DT(mode), inum.no_addr,
- inum.no_formal_ino, 0);
+ inum.no_formal_ino);
if (IS_ERR(inode))
goto fail_gunlock2;
extern void gfs2_set_iop(struct inode *inode);
extern struct inode *gfs2_inode_lookup(struct super_block *sb, unsigned type,
- u64 no_addr, u64 no_formal_ino,
- int skip_freeing);
+ u64 no_addr, u64 no_formal_ino);
+extern int gfs2_unlinked_inode_lookup(struct super_block *sb, u64 no_addr,
+ struct inode **inode);
extern struct inode *gfs2_ilookup(struct super_block *sb, u64 no_addr);
extern int gfs2_inode_refresh(struct gfs2_inode *ip);
return list_empty(&ai->ai_ail1_list);
}
-static void gfs2_ail1_start(struct gfs2_sbd *sdp, int flags)
+static void gfs2_ail1_start(struct gfs2_sbd *sdp)
{
struct list_head *head;
u64 sync_gen;
- struct list_head *first;
- struct gfs2_ail *first_ai, *ai, *tmp;
+ struct gfs2_ail *ai;
int done = 0;
gfs2_log_lock(sdp);
}
sync_gen = sdp->sd_ail_sync_gen++;
- first = head->prev;
- first_ai = list_entry(first, struct gfs2_ail, ai_list);
- first_ai->ai_sync_gen = sync_gen;
- gfs2_ail1_start_one(sdp, first_ai); /* This may drop log lock */
-
- if (flags & DIO_ALL)
- first = NULL;
-
while(!done) {
- if (first && (head->prev != first ||
- gfs2_ail1_empty_one(sdp, first_ai, 0)))
- break;
-
done = 1;
- list_for_each_entry_safe_reverse(ai, tmp, head, ai_list) {
+ list_for_each_entry_reverse(ai, head, ai_list) {
if (ai->ai_sync_gen >= sync_gen)
continue;
ai->ai_sync_gen = sync_gen;
* flush time, so we ensure that we have just enough free blocks at all
* times to avoid running out during a log flush.
*
+ * We no longer flush the log here, instead we wake up logd to do that
+ * for us. To avoid the thundering herd and to ensure that we deal fairly
+ * with queued waiters, we use an exclusive wait. This means that when we
+ * get woken with enough journal space to get our reservation, we need to
+ * wake the next waiter on the list.
+ *
* Returns: errno
*/
int gfs2_log_reserve(struct gfs2_sbd *sdp, unsigned int blks)
{
- unsigned int try = 0;
unsigned reserved_blks = 6 * (4096 / sdp->sd_vfs->s_blocksize);
+ unsigned wanted = blks + reserved_blks;
+ DEFINE_WAIT(wait);
+ int did_wait = 0;
+ unsigned int free_blocks;
if (gfs2_assert_warn(sdp, blks) ||
gfs2_assert_warn(sdp, blks <= sdp->sd_jdesc->jd_blocks))
return -EINVAL;
-
- mutex_lock(&sdp->sd_log_reserve_mutex);
- gfs2_log_lock(sdp);
- while(atomic_read(&sdp->sd_log_blks_free) <= (blks + reserved_blks)) {
- gfs2_log_unlock(sdp);
- gfs2_ail1_empty(sdp, 0);
- gfs2_log_flush(sdp, NULL);
-
- if (try++)
- gfs2_ail1_start(sdp, 0);
- gfs2_log_lock(sdp);
+retry:
+ free_blocks = atomic_read(&sdp->sd_log_blks_free);
+ if (unlikely(free_blocks <= wanted)) {
+ do {
+ prepare_to_wait_exclusive(&sdp->sd_log_waitq, &wait,
+ TASK_UNINTERRUPTIBLE);
+ wake_up(&sdp->sd_logd_waitq);
+ did_wait = 1;
+ if (atomic_read(&sdp->sd_log_blks_free) <= wanted)
+ io_schedule();
+ free_blocks = atomic_read(&sdp->sd_log_blks_free);
+ } while(free_blocks <= wanted);
+ finish_wait(&sdp->sd_log_waitq, &wait);
}
- atomic_sub(blks, &sdp->sd_log_blks_free);
+ if (atomic_cmpxchg(&sdp->sd_log_blks_free, free_blocks,
+ free_blocks - blks) != free_blocks)
+ goto retry;
trace_gfs2_log_blocks(sdp, -blks);
- gfs2_log_unlock(sdp);
- mutex_unlock(&sdp->sd_log_reserve_mutex);
+
+ /*
+ * If we waited, then so might others, wake them up _after_ we get
+ * our share of the log.
+ */
+ if (unlikely(did_wait))
+ wake_up(&sdp->sd_log_waitq);
down_read(&sdp->sd_log_flush_lock);
return 0;
}
-/**
- * gfs2_log_release - Release a given number of log blocks
- * @sdp: The GFS2 superblock
- * @blks: The number of blocks
- *
- */
-
-void gfs2_log_release(struct gfs2_sbd *sdp, unsigned int blks)
-{
-
- gfs2_log_lock(sdp);
- atomic_add(blks, &sdp->sd_log_blks_free);
- trace_gfs2_log_blocks(sdp, blks);
- gfs2_assert_withdraw(sdp,
- atomic_read(&sdp->sd_log_blks_free) <= sdp->sd_jdesc->jd_blocks);
- gfs2_log_unlock(sdp);
- up_read(&sdp->sd_log_flush_lock);
-}
-
static u64 log_bmap(struct gfs2_sbd *sdp, unsigned int lbn)
{
struct gfs2_journal_extent *je;
ail2_empty(sdp, new_tail);
- gfs2_log_lock(sdp);
atomic_add(dist, &sdp->sd_log_blks_free);
trace_gfs2_log_blocks(sdp, dist);
- gfs2_assert_withdraw(sdp, atomic_read(&sdp->sd_log_blks_free) <= sdp->sd_jdesc->jd_blocks);
- gfs2_log_unlock(sdp);
+ gfs2_assert_withdraw(sdp, atomic_read(&sdp->sd_log_blks_free) <=
+ sdp->sd_jdesc->jd_blocks);
sdp->sd_log_tail = new_tail;
}
if (buffer_eopnotsupp(bh)) {
clear_buffer_eopnotsupp(bh);
set_buffer_uptodate(bh);
+ fs_info(sdp, "barrier sync failed - disabling barriers\n");
set_bit(SDF_NOBARRIERS, &sdp->sd_flags);
lock_buffer(bh);
skip_barrier:
* @sdp: the filesystem
* @tr: the transaction
*
+ * We wake up gfs2_logd if the number of pinned blocks exceed thresh1
+ * or the total number of used blocks (pinned blocks plus AIL blocks)
+ * is greater than thresh2.
+ *
+ * At mount time thresh1 is 1/3rd of journal size, thresh2 is 2/3rd of
+ * journal size.
+ *
* Returns: errno
*/
up_read(&sdp->sd_log_flush_lock);
- gfs2_log_lock(sdp);
- if (sdp->sd_log_num_buf > gfs2_tune_get(sdp, gt_incore_log_blocks))
- wake_up_process(sdp->sd_logd_process);
- gfs2_log_unlock(sdp);
+ if (atomic_read(&sdp->sd_log_pinned) > atomic_read(&sdp->sd_log_thresh1) ||
+ ((sdp->sd_jdesc->jd_blocks - atomic_read(&sdp->sd_log_blks_free)) >
+ atomic_read(&sdp->sd_log_thresh2)))
+ wake_up(&sdp->sd_logd_waitq);
}
/**
{
gfs2_log_flush(sdp, NULL);
for (;;) {
- gfs2_ail1_start(sdp, DIO_ALL);
+ gfs2_ail1_start(sdp);
if (gfs2_ail1_empty(sdp, DIO_ALL))
break;
msleep(10);
}
}
+static inline int gfs2_jrnl_flush_reqd(struct gfs2_sbd *sdp)
+{
+ return (atomic_read(&sdp->sd_log_pinned) >= atomic_read(&sdp->sd_log_thresh1));
+}
+
+static inline int gfs2_ail_flush_reqd(struct gfs2_sbd *sdp)
+{
+ unsigned int used_blocks = sdp->sd_jdesc->jd_blocks - atomic_read(&sdp->sd_log_blks_free);
+ return used_blocks >= atomic_read(&sdp->sd_log_thresh2);
+}
/**
* gfs2_logd - Update log tail as Active Items get flushed to in-place blocks
int gfs2_logd(void *data)
{
struct gfs2_sbd *sdp = data;
- unsigned long t;
- int need_flush;
+ unsigned long t = 1;
+ DEFINE_WAIT(wait);
+ unsigned preflush;
while (!kthread_should_stop()) {
- /* Advance the log tail */
- t = sdp->sd_log_flush_time +
- gfs2_tune_get(sdp, gt_log_flush_secs) * HZ;
+ preflush = atomic_read(&sdp->sd_log_pinned);
+ if (gfs2_jrnl_flush_reqd(sdp) || t == 0) {
+ gfs2_ail1_empty(sdp, DIO_ALL);
+ gfs2_log_flush(sdp, NULL);
+ gfs2_ail1_empty(sdp, DIO_ALL);
+ }
- gfs2_ail1_empty(sdp, DIO_ALL);
- gfs2_log_lock(sdp);
- need_flush = sdp->sd_log_num_buf > gfs2_tune_get(sdp, gt_incore_log_blocks);
- gfs2_log_unlock(sdp);
- if (need_flush || time_after_eq(jiffies, t)) {
+ if (gfs2_ail_flush_reqd(sdp)) {
+ gfs2_ail1_start(sdp);
+ io_schedule();
+ gfs2_ail1_empty(sdp, 0);
gfs2_log_flush(sdp, NULL);
- sdp->sd_log_flush_time = jiffies;
+ gfs2_ail1_empty(sdp, DIO_ALL);
}
+ wake_up(&sdp->sd_log_waitq);
t = gfs2_tune_get(sdp, gt_logd_secs) * HZ;
if (freezing(current))
refrigerator();
- schedule_timeout_interruptible(t);
+
+ do {
+ prepare_to_wait(&sdp->sd_logd_waitq, &wait,
+ TASK_UNINTERRUPTIBLE);
+ if (!gfs2_ail_flush_reqd(sdp) &&
+ !gfs2_jrnl_flush_reqd(sdp) &&
+ !kthread_should_stop())
+ t = schedule_timeout(t);
+ } while(t && !gfs2_ail_flush_reqd(sdp) &&
+ !gfs2_jrnl_flush_reqd(sdp) &&
+ !kthread_should_stop());
+ finish_wait(&sdp->sd_logd_waitq, &wait);
}
return 0;
unsigned int ssize);
int gfs2_log_reserve(struct gfs2_sbd *sdp, unsigned int blks);
-void gfs2_log_release(struct gfs2_sbd *sdp, unsigned int blks);
void gfs2_log_incr_head(struct gfs2_sbd *sdp);
struct buffer_head *gfs2_log_get_buf(struct gfs2_sbd *sdp);
if (bd->bd_ail)
list_move(&bd->bd_ail_st_list, &bd->bd_ail->ai_ail2_list);
get_bh(bh);
+ atomic_inc(&sdp->sd_log_pinned);
trace_gfs2_pin(bd, 1);
}
trace_gfs2_pin(bd, 0);
gfs2_log_unlock(sdp);
unlock_buffer(bh);
+ atomic_dec(&sdp->sd_log_pinned);
}
if (!gfs2_glock_cachep)
goto fail;
- gfs2_glock_aspace_cachep = kmem_cache_create("gfs2_glock (aspace)",
+ gfs2_glock_aspace_cachep = kmem_cache_create("gfs2_glock(aspace)",
sizeof(struct gfs2_glock) +
sizeof(struct address_space),
0, 0, gfs2_init_gl_aspace_once);
static int gfs2_aspace_writepage(struct page *page, struct writeback_control *wbc)
{
- int err;
struct buffer_head *bh, *head;
int nr_underway = 0;
int write_op = (1 << BIO_RW_META) | ((wbc->sync_mode == WB_SYNC_ALL ?
} while (bh != head);
unlock_page(page);
- err = 0;
if (nr_underway == 0)
end_page_writeback(page);
- return err;
+ return 0;
}
const struct address_space_operations gfs2_meta_aops = {
struct gfs2_bufdata *bd = bh->b_private;
if (test_clear_buffer_pinned(bh)) {
+ atomic_dec(&sdp->sd_log_pinned);
list_del_init(&bd->bd_le.le_list);
if (meta) {
gfs2_assert_warn(sdp, sdp->sd_log_num_buf);
{
spin_lock_init(>->gt_spin);
- gt->gt_incore_log_blocks = 1024;
- gt->gt_logd_secs = 1;
gt->gt_quota_simul_sync = 64;
gt->gt_quota_warn_period = 10;
gt->gt_quota_scale_num = 1;
spin_lock_init(&sdp->sd_trunc_lock);
spin_lock_init(&sdp->sd_log_lock);
-
+ atomic_set(&sdp->sd_log_pinned, 0);
INIT_LIST_HEAD(&sdp->sd_log_le_buf);
INIT_LIST_HEAD(&sdp->sd_log_le_revoke);
INIT_LIST_HEAD(&sdp->sd_log_le_rg);
INIT_LIST_HEAD(&sdp->sd_log_le_databuf);
INIT_LIST_HEAD(&sdp->sd_log_le_ordered);
- mutex_init(&sdp->sd_log_reserve_mutex);
+ init_waitqueue_head(&sdp->sd_log_waitq);
+ init_waitqueue_head(&sdp->sd_logd_waitq);
INIT_LIST_HEAD(&sdp->sd_ail1_list);
INIT_LIST_HEAD(&sdp->sd_ail2_list);
struct dentry *dentry;
struct inode *inode;
- inode = gfs2_inode_lookup(sb, DT_DIR, no_addr, 0, 0);
+ inode = gfs2_inode_lookup(sb, DT_DIR, no_addr, 0);
if (IS_ERR(inode)) {
fs_err(sdp, "can't read in %s inode: %ld\n", name, PTR_ERR(inode));
return PTR_ERR(inode);
if (sdp->sd_args.ar_spectator) {
sdp->sd_jdesc = gfs2_jdesc_find(sdp, 0);
atomic_set(&sdp->sd_log_blks_free, sdp->sd_jdesc->jd_blocks);
+ atomic_set(&sdp->sd_log_thresh1, 2*sdp->sd_jdesc->jd_blocks/5);
+ atomic_set(&sdp->sd_log_thresh2, 4*sdp->sd_jdesc->jd_blocks/5);
} else {
if (sdp->sd_lockstruct.ls_jid >= gfs2_jindex_size(sdp)) {
fs_err(sdp, "can't mount journal #%u\n",
goto fail_jinode_gh;
}
atomic_set(&sdp->sd_log_blks_free, sdp->sd_jdesc->jd_blocks);
+ atomic_set(&sdp->sd_log_thresh1, 2*sdp->sd_jdesc->jd_blocks/5);
+ atomic_set(&sdp->sd_log_thresh2, 4*sdp->sd_jdesc->jd_blocks/5);
/* Map the extents for this journal's blocks */
map_journal_extents(sdp);
if (undo)
goto fail_quotad;
- sdp->sd_log_flush_time = jiffies;
-
p = kthread_run(gfs2_logd, sdp, "gfs2_logd");
error = IS_ERR(p);
if (error) {
GFS2_BASIC_BLOCK_SHIFT;
sdp->sd_fsb2bb = 1 << sdp->sd_fsb2bb_shift;
- sdp->sd_tune.gt_log_flush_secs = sdp->sd_args.ar_commit;
+ sdp->sd_tune.gt_logd_secs = sdp->sd_args.ar_commit;
sdp->sd_tune.gt_quota_quantum = sdp->sd_args.ar_quota_quantum;
if (sdp->sd_args.ar_statfs_quantum) {
sdp->sd_tune.gt_statfs_slow = 0;
memset(&args, 0, sizeof(args));
args.ar_quota = GFS2_QUOTA_DEFAULT;
args.ar_data = GFS2_DATA_DEFAULT;
- args.ar_commit = 60;
+ args.ar_commit = 30;
args.ar_statfs_quantum = 30;
args.ar_quota_quantum = 60;
args.ar_errors = GFS2_ERRORS_DEFAULT;
unsigned blocksize, iblock, pos;
struct buffer_head *bh, *dibh;
struct page *page;
- void *kaddr;
- struct gfs2_quota *qp;
- s64 value;
- int err = -EIO;
+ void *kaddr, *ptr;
+ struct gfs2_quota q, *qp;
+ int err, nbytes;
u64 size;
if (gfs2_is_stuffed(ip))
gfs2_unstuff_dinode(ip, NULL);
-
+
+ memset(&q, 0, sizeof(struct gfs2_quota));
+ err = gfs2_internal_read(ip, NULL, (char *)&q, &loc, sizeof(q));
+ if (err < 0)
+ return err;
+
+ err = -EIO;
+ qp = &q;
+ qp->qu_value = be64_to_cpu(qp->qu_value);
+ qp->qu_value += change;
+ qp->qu_value = cpu_to_be64(qp->qu_value);
+ qd->qd_qb.qb_value = qp->qu_value;
+ if (fdq) {
+ if (fdq->d_fieldmask & FS_DQ_BSOFT) {
+ qp->qu_warn = cpu_to_be64(fdq->d_blk_softlimit);
+ qd->qd_qb.qb_warn = qp->qu_warn;
+ }
+ if (fdq->d_fieldmask & FS_DQ_BHARD) {
+ qp->qu_limit = cpu_to_be64(fdq->d_blk_hardlimit);
+ qd->qd_qb.qb_limit = qp->qu_limit;
+ }
+ }
+
+ /* Write the quota into the quota file on disk */
+ ptr = qp;
+ nbytes = sizeof(struct gfs2_quota);
+get_a_page:
page = grab_cache_page(mapping, index);
if (!page)
return -ENOMEM;
if (!buffer_mapped(bh)) {
gfs2_block_map(inode, iblock, bh, 1);
if (!buffer_mapped(bh))
- goto unlock;
+ goto unlock_out;
+ /* If it's a newly allocated disk block for quota, zero it */
+ if (buffer_new(bh)) {
+ memset(bh->b_data, 0, bh->b_size);
+ set_buffer_uptodate(bh);
+ }
}
if (PageUptodate(page))
ll_rw_block(READ_META, 1, &bh);
wait_on_buffer(bh);
if (!buffer_uptodate(bh))
- goto unlock;
+ goto unlock_out;
}
gfs2_trans_add_bh(ip->i_gl, bh, 0);
kaddr = kmap_atomic(page, KM_USER0);
- qp = kaddr + offset;
- value = (s64)be64_to_cpu(qp->qu_value) + change;
- qp->qu_value = cpu_to_be64(value);
- qd->qd_qb.qb_value = qp->qu_value;
- if (fdq) {
- if (fdq->d_fieldmask & FS_DQ_BSOFT) {
- qp->qu_warn = cpu_to_be64(fdq->d_blk_softlimit);
- qd->qd_qb.qb_warn = qp->qu_warn;
- }
- if (fdq->d_fieldmask & FS_DQ_BHARD) {
- qp->qu_limit = cpu_to_be64(fdq->d_blk_hardlimit);
- qd->qd_qb.qb_limit = qp->qu_limit;
- }
- }
+ if (offset + sizeof(struct gfs2_quota) > PAGE_CACHE_SIZE)
+ nbytes = PAGE_CACHE_SIZE - offset;
+ memcpy(kaddr + offset, ptr, nbytes);
flush_dcache_page(page);
kunmap_atomic(kaddr, KM_USER0);
+ unlock_page(page);
+ page_cache_release(page);
+ /* If quota straddles page boundary, we need to update the rest of the
+ * quota at the beginning of the next page */
+ if (offset != 0) { /* first page, offset is closer to PAGE_CACHE_SIZE */
+ ptr = ptr + nbytes;
+ nbytes = sizeof(struct gfs2_quota) - nbytes;
+ offset = 0;
+ index++;
+ goto get_a_page;
+ }
+
+ /* Update the disk inode timestamp and size (if extended) */
err = gfs2_meta_inode_buffer(ip, &dibh);
if (err)
- goto unlock;
+ goto out;
size = loc + sizeof(struct gfs2_quota);
if (size > inode->i_size) {
brelse(dibh);
mark_inode_dirty(inode);
-unlock:
+out:
+ return err;
+unlock_out:
unlock_page(page);
page_cache_release(page);
return err;
* rgrp since it won't be allocated during the transaction
*/
al->al_requested = 1;
- /* +1 in the end for block requested above for unstuffing */
- blocks = num_qd * data_blocks + RES_DINODE + num_qd + 1;
+ /* +3 in the end for unstuffing block, inode size update block
+ * and another block in case quota straddles page boundary and
+ * two blocks need to be updated instead of 1 */
+ blocks = num_qd * data_blocks + RES_DINODE + num_qd + 3;
if (nalloc)
al->al_requested += nalloc * (data_blocks + ind_blocks);
memset(fqs, 0, sizeof(struct fs_quota_stat));
fqs->qs_version = FS_QSTAT_VERSION;
- if (sdp->sd_args.ar_quota == GFS2_QUOTA_ON)
- fqs->qs_flags = (XFS_QUOTA_UDQ_ENFD | XFS_QUOTA_GDQ_ENFD);
- else if (sdp->sd_args.ar_quota == GFS2_QUOTA_ACCOUNT)
- fqs->qs_flags = (XFS_QUOTA_UDQ_ACCT | XFS_QUOTA_GDQ_ACCT);
+
+ switch (sdp->sd_args.ar_quota) {
+ case GFS2_QUOTA_ON:
+ fqs->qs_flags |= (XFS_QUOTA_UDQ_ENFD | XFS_QUOTA_GDQ_ENFD);
+ /*FALLTHRU*/
+ case GFS2_QUOTA_ACCOUNT:
+ fqs->qs_flags |= (XFS_QUOTA_UDQ_ACCT | XFS_QUOTA_GDQ_ACCT);
+ break;
+ case GFS2_QUOTA_OFF:
+ break;
+ }
+
if (sdp->sd_quota_inode) {
fqs->qs_uquota.qfs_ino = GFS2_I(sdp->sd_quota_inode)->i_no_addr;
fqs->qs_uquota.qfs_nblks = sdp->sd_quota_inode->i_blocks;
* try_rgrp_unlink - Look for any unlinked, allocated, but unused inodes
* @rgd: The rgrp
*
- * Returns: The inode, if one has been found
+ * Returns: 0 if no error
+ * The inode, if one has been found, in inode.
*/
-static struct inode *try_rgrp_unlink(struct gfs2_rgrpd *rgd, u64 *last_unlinked,
- u64 skip)
+static u64 try_rgrp_unlink(struct gfs2_rgrpd *rgd, u64 *last_unlinked,
+ u64 skip)
{
- struct inode *inode;
u32 goal = 0, block;
u64 no_addr;
struct gfs2_sbd *sdp = rgd->rd_sbd;
if (no_addr == skip)
continue;
*last_unlinked = no_addr;
- inode = gfs2_inode_lookup(rgd->rd_sbd->sd_vfs, DT_UNKNOWN,
- no_addr, -1, 1);
- if (!IS_ERR(inode))
- return inode;
+ return no_addr;
}
rgd->rd_flags &= ~GFS2_RDF_CHECK;
- return NULL;
+ return 0;
}
/**
* Try to acquire rgrp in way which avoids contending with others.
*
* Returns: errno
+ * unlinked: the block address of an unlinked block to be reclaimed
*/
-static struct inode *get_local_rgrp(struct gfs2_inode *ip, u64 *last_unlinked)
+static int get_local_rgrp(struct gfs2_inode *ip, u64 *unlinked,
+ u64 *last_unlinked)
{
- struct inode *inode = NULL;
struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
struct gfs2_rgrpd *rgd, *begin = NULL;
struct gfs2_alloc *al = ip->i_alloc;
int loops = 0;
int error, rg_locked;
+ *unlinked = 0;
rgd = gfs2_blk2rgrpd(sdp, ip->i_goal);
while (rgd) {
case 0:
if (try_rgrp_fit(rgd, al))
goto out;
- if (rgd->rd_flags & GFS2_RDF_CHECK)
- inode = try_rgrp_unlink(rgd, last_unlinked, ip->i_no_addr);
+ /* If the rg came in already locked, there's no
+ way we can recover from a failed try_rgrp_unlink
+ because that would require an iput which can only
+ happen after the rgrp is unlocked. */
+ if (!rg_locked && rgd->rd_flags & GFS2_RDF_CHECK)
+ *unlinked = try_rgrp_unlink(rgd, last_unlinked,
+ ip->i_no_addr);
if (!rg_locked)
gfs2_glock_dq_uninit(&al->al_rgd_gh);
- if (inode)
- return inode;
+ if (*unlinked)
+ return -EAGAIN;
/* fall through */
case GLR_TRYFAILED:
rgd = recent_rgrp_next(rgd);
break;
default:
- return ERR_PTR(error);
+ return error;
}
}
case 0:
if (try_rgrp_fit(rgd, al))
goto out;
- if (rgd->rd_flags & GFS2_RDF_CHECK)
- inode = try_rgrp_unlink(rgd, last_unlinked, ip->i_no_addr);
+ if (!rg_locked && rgd->rd_flags & GFS2_RDF_CHECK)
+ *unlinked = try_rgrp_unlink(rgd, last_unlinked,
+ ip->i_no_addr);
if (!rg_locked)
gfs2_glock_dq_uninit(&al->al_rgd_gh);
- if (inode)
- return inode;
+ if (*unlinked)
+ return -EAGAIN;
break;
case GLR_TRYFAILED:
break;
default:
- return ERR_PTR(error);
+ return error;
}
rgd = gfs2_rgrpd_get_next(rgd);
if (rgd == begin) {
if (++loops >= 3)
- return ERR_PTR(-ENOSPC);
+ return -ENOSPC;
if (!skipped)
loops++;
flags = 0;
forward_rgrp_set(sdp, rgd);
}
- return NULL;
+ return 0;
}
/**
struct gfs2_alloc *al = ip->i_alloc;
struct inode *inode;
int error = 0;
- u64 last_unlinked = NO_BLOCK;
+ u64 last_unlinked = NO_BLOCK, unlinked;
if (gfs2_assert_warn(sdp, al->al_requested))
return -EINVAL;
if (error)
return error;
- inode = get_local_rgrp(ip, &last_unlinked);
- if (inode) {
+ error = get_local_rgrp(ip, &unlinked, &last_unlinked);
+ if (error) {
if (ip != GFS2_I(sdp->sd_rindex))
gfs2_glock_dq_uninit(&al->al_ri_gh);
- if (IS_ERR(inode))
- return PTR_ERR(inode);
- iput(inode);
+ if (error != -EAGAIN)
+ return error;
+ error = gfs2_unlinked_inode_lookup(ip->i_inode.i_sb,
+ unlinked, &inode);
+ if (inode)
+ iput(inode);
gfs2_log_flush(sdp, NULL);
+ if (error == GLR_TRYFAILED)
+ error = 0;
goto try_again;
}
int error;
spin_lock(>->gt_spin);
- args.ar_commit = gt->gt_log_flush_secs;
+ args.ar_commit = gt->gt_logd_secs;
args.ar_quota_quantum = gt->gt_quota_quantum;
if (gt->gt_statfs_slow)
args.ar_statfs_quantum = 0;
else
clear_bit(SDF_NOBARRIERS, &sdp->sd_flags);
spin_lock(>->gt_spin);
- gt->gt_log_flush_secs = args.ar_commit;
+ gt->gt_logd_secs = args.ar_commit;
gt->gt_quota_quantum = args.ar_quota_quantum;
if (args.ar_statfs_quantum) {
gt->gt_statfs_slow = 0;
}
if (args->ar_discard)
seq_printf(s, ",discard");
- val = sdp->sd_tune.gt_log_flush_secs;
- if (val != 60)
+ val = sdp->sd_tune.gt_logd_secs;
+ if (val != 30)
seq_printf(s, ",commit=%d", val);
val = sdp->sd_tune.gt_statfs_quantum;
if (val != 30)
}
if (test_bit(SDF_NOBARRIERS, &sdp->sd_flags))
seq_printf(s, ",nobarrier");
-
+ if (test_bit(SDF_DEMOTE, &sdp->sd_flags))
+ seq_printf(s, ",demote_interface_used");
return 0;
}
glops = gfs2_glops_list[gltype];
if (glops == NULL)
return -EINVAL;
+ if (!test_and_set_bit(SDF_DEMOTE, &sdp->sd_flags))
+ fs_info(sdp, "demote interface used\n");
rv = gfs2_glock_get(sdp, glnum, glops, 0, &gl);
if (rv)
return rv;
} \
TUNE_ATTR_2(name, name##_store)
-TUNE_ATTR(incore_log_blocks, 0);
-TUNE_ATTR(log_flush_secs, 0);
TUNE_ATTR(quota_warn_period, 0);
TUNE_ATTR(quota_quantum, 0);
TUNE_ATTR(max_readahead, 0);
TUNE_ATTR_3(quota_scale, quota_scale_show, quota_scale_store);
static struct attribute *tune_attrs[] = {
- &tune_attr_incore_log_blocks.attr,
- &tune_attr_log_flush_secs.attr,
&tune_attr_quota_warn_period.attr,
&tune_attr_quota_quantum.attr,
&tune_attr_max_readahead.attr,
#include "meta_io.h"
#include "trans.h"
#include "util.h"
+#include "trace_gfs2.h"
int gfs2_trans_begin(struct gfs2_sbd *sdp, unsigned int blocks,
unsigned int revokes)
return error;
}
+/**
+ * gfs2_log_release - Release a given number of log blocks
+ * @sdp: The GFS2 superblock
+ * @blks: The number of blocks
+ *
+ */
+
+static void gfs2_log_release(struct gfs2_sbd *sdp, unsigned int blks)
+{
+
+ atomic_add(blks, &sdp->sd_log_blks_free);
+ trace_gfs2_log_blocks(sdp, blks);
+ gfs2_assert_withdraw(sdp, atomic_read(&sdp->sd_log_blks_free) <=
+ sdp->sd_jdesc->jd_blocks);
+ up_read(&sdp->sd_log_flush_lock);
+}
+
void gfs2_trans_end(struct gfs2_sbd *sdp)
{
struct gfs2_trans *tr = current->journal_info;
projects / firefly-linux-kernel-4.4.55.git / commitdiff