aboutsummaryrefslogtreecommitdiff
path: root/fs/ocfs2/journal.c
diff options
context:
space:
mode:
Diffstat (limited to 'fs/ocfs2/journal.c')
-rw-r--r--fs/ocfs2/journal.c123
1 files changed, 93 insertions, 30 deletions
diff --git a/fs/ocfs2/journal.c b/fs/ocfs2/journal.c
index b141a44605c..4b0c68849b3 100644
--- a/fs/ocfs2/journal.c
+++ b/fs/ocfs2/journal.c
@@ -30,6 +30,7 @@
#include <linux/kthread.h>
#include <linux/time.h>
#include <linux/random.h>
+#include <linux/delay.h>
#include <cluster/masklog.h>
@@ -355,11 +356,14 @@ handle_t *ocfs2_start_trans(struct ocfs2_super *osb, int max_buffs)
if (journal_current_handle())
return jbd2_journal_start(journal, max_buffs);
+ sb_start_intwrite(osb->sb);
+
down_read(&osb->journal->j_trans_barrier);
handle = jbd2_journal_start(journal, max_buffs);
if (IS_ERR(handle)) {
up_read(&osb->journal->j_trans_barrier);
+ sb_end_intwrite(osb->sb);
mlog_errno(PTR_ERR(handle));
@@ -388,8 +392,10 @@ int ocfs2_commit_trans(struct ocfs2_super *osb,
if (ret < 0)
mlog_errno(ret);
- if (!nested)
+ if (!nested) {
up_read(&journal->j_trans_barrier);
+ sb_end_intwrite(osb->sb);
+ }
return ret;
}
@@ -450,6 +456,41 @@ bail:
return status;
}
+/*
+ * If we have fewer than thresh credits, extend by OCFS2_MAX_TRANS_DATA.
+ * If that fails, restart the transaction & regain write access for the
+ * buffer head which is used for metadata modifications.
+ * Taken from Ext4: extend_or_restart_transaction()
+ */
+int ocfs2_allocate_extend_trans(handle_t *handle, int thresh)
+{
+ int status, old_nblks;
+
+ BUG_ON(!handle);
+
+ old_nblks = handle->h_buffer_credits;
+ trace_ocfs2_allocate_extend_trans(old_nblks, thresh);
+
+ if (old_nblks < thresh)
+ return 0;
+
+ status = jbd2_journal_extend(handle, OCFS2_MAX_TRANS_DATA);
+ if (status < 0) {
+ mlog_errno(status);
+ goto bail;
+ }
+
+ if (status > 0) {
+ status = jbd2_journal_restart(handle, OCFS2_MAX_TRANS_DATA);
+ if (status < 0)
+ mlog_errno(status);
+ }
+
+bail:
+ return status;
+}
+
+
struct ocfs2_triggers {
struct jbd2_buffer_trigger_type ot_triggers;
int ot_offset;
@@ -796,14 +837,14 @@ int ocfs2_journal_init(struct ocfs2_journal *journal, int *dirty)
inode_lock = 1;
di = (struct ocfs2_dinode *)bh->b_data;
- if (inode->i_size < OCFS2_MIN_JOURNAL_SIZE) {
+ if (i_size_read(inode) < OCFS2_MIN_JOURNAL_SIZE) {
mlog(ML_ERROR, "Journal file size (%lld) is too small!\n",
- inode->i_size);
+ i_size_read(inode));
status = -EINVAL;
goto done;
}
- trace_ocfs2_journal_init(inode->i_size,
+ trace_ocfs2_journal_init(i_size_read(inode),
(unsigned long long)inode->i_blocks,
OCFS2_I(inode)->ip_clusters);
@@ -1091,7 +1132,7 @@ static int ocfs2_force_read_journal(struct inode *inode)
memset(bhs, 0, sizeof(struct buffer_head *) * CONCURRENT_JOURNAL_FILL);
- num_blocks = ocfs2_blocks_for_bytes(inode->i_sb, inode->i_size);
+ num_blocks = ocfs2_blocks_for_bytes(inode->i_sb, i_size_read(inode));
v_blkno = 0;
while (v_blkno < num_blocks) {
status = ocfs2_extent_map_get_blocks(inode, v_blkno,
@@ -1229,11 +1270,8 @@ static void ocfs2_queue_recovery_completion(struct ocfs2_journal *journal,
/* Though we wish to avoid it, we are in fact safe in
* skipping local alloc cleanup as fsck.ocfs2 is more
* than capable of reclaiming unused space. */
- if (la_dinode)
- kfree(la_dinode);
-
- if (tl_dinode)
- kfree(tl_dinode);
+ kfree(la_dinode);
+ kfree(tl_dinode);
if (qrec)
ocfs2_free_quota_recovery(qrec);
@@ -1260,6 +1298,9 @@ void ocfs2_complete_mount_recovery(struct ocfs2_super *osb)
{
struct ocfs2_journal *journal = osb->journal;
+ if (ocfs2_is_hard_readonly(osb))
+ return;
+
/* No need to queue up our truncate_log as regular cleanup will catch
* that */
ocfs2_queue_recovery_completion(journal, osb->slot_num,
@@ -1400,8 +1441,7 @@ bail:
mutex_unlock(&osb->recovery_lock);
- if (rm_quota)
- kfree(rm_quota);
+ kfree(rm_quota);
/* no one is callint kthread_stop() for us so the kthread() api
* requires that we call do_exit(). And it isn't exported, but
@@ -1541,9 +1581,9 @@ static int ocfs2_replay_journal(struct ocfs2_super *osb,
/* we need to run complete recovery for offline orphan slots */
ocfs2_replay_map_set_state(osb, REPLAY_NEEDED);
- mlog(ML_NOTICE, "Recovering node %d from slot %d on device (%u,%u)\n",
- node_num, slot_num,
- MAJOR(osb->sb->s_dev), MINOR(osb->sb->s_dev));
+ printk(KERN_NOTICE "ocfs2: Begin replay journal (node %d, slot %d) on "\
+ "device (%u,%u)\n", node_num, slot_num, MAJOR(osb->sb->s_dev),
+ MINOR(osb->sb->s_dev));
OCFS2_I(inode)->ip_clusters = le32_to_cpu(fe->i_clusters);
@@ -1598,6 +1638,9 @@ static int ocfs2_replay_journal(struct ocfs2_super *osb,
jbd2_journal_destroy(journal);
+ printk(KERN_NOTICE "ocfs2: End replay journal (node %d, slot %d) on "\
+ "device (%u,%u)\n", node_num, slot_num, MAJOR(osb->sb->s_dev),
+ MINOR(osb->sb->s_dev));
done:
/* drop the lock on this nodes journal */
if (got_lock)
@@ -1805,6 +1848,20 @@ static inline unsigned long ocfs2_orphan_scan_timeout(void)
* every slot, queuing a recovery of the slot on the ocfs2_wq thread. This
* is done to catch any orphans that are left over in orphan directories.
*
+ * It scans all slots, even ones that are in use. It does so to handle the
+ * case described below:
+ *
+ * Node 1 has an inode it was using. The dentry went away due to memory
+ * pressure. Node 1 closes the inode, but it's on the free list. The node
+ * has the open lock.
+ * Node 2 unlinks the inode. It grabs the dentry lock to notify others,
+ * but node 1 has no dentry and doesn't get the message. It trylocks the
+ * open lock, sees that another node has a PR, and does nothing.
+ * Later node 2 runs its orphan dir. It igets the inode, trylocks the
+ * open lock, sees the PR still, and does nothing.
+ * Basically, we have to trigger an orphan iput on node 1. The only way
+ * for this to happen is if node 1 runs node 2's orphan dir.
+ *
* ocfs2_queue_orphan_scan gets called every ORPHAN_SCAN_SCHEDULE_TIMEOUT
* seconds. It gets an EX lock on os_lockres and checks sequence number
* stored in LVB. If the sequence number has changed, it means some other
@@ -1920,6 +1977,7 @@ void ocfs2_orphan_scan_start(struct ocfs2_super *osb)
}
struct ocfs2_orphan_filldir_priv {
+ struct dir_context ctx;
struct inode *head;
struct ocfs2_super *osb;
};
@@ -1956,11 +2014,11 @@ static int ocfs2_queue_orphans(struct ocfs2_super *osb,
{
int status;
struct inode *orphan_dir_inode = NULL;
- struct ocfs2_orphan_filldir_priv priv;
- loff_t pos = 0;
-
- priv.osb = osb;
- priv.head = *head;
+ struct ocfs2_orphan_filldir_priv priv = {
+ .ctx.actor = ocfs2_orphan_filldir,
+ .osb = osb,
+ .head = *head
+ };
orphan_dir_inode = ocfs2_get_system_file_inode(osb,
ORPHAN_DIR_SYSTEM_INODE,
@@ -1978,8 +2036,7 @@ static int ocfs2_queue_orphans(struct ocfs2_super *osb,
goto out;
}
- status = ocfs2_dir_foreach(orphan_dir_inode, &pos, &priv,
- ocfs2_orphan_filldir);
+ status = ocfs2_dir_foreach(orphan_dir_inode, &priv.ctx);
if (status) {
mlog_errno(status);
goto out_cluster;
@@ -2076,12 +2133,6 @@ static int ocfs2_recover_orphans(struct ocfs2_super *osb,
iter = oi->ip_next_orphan;
spin_lock(&oi->ip_lock);
- /* The remote delete code may have set these on the
- * assumption that the other node would wipe them
- * successfully. If they are still in the node's
- * orphan dir, we need to reset that state. */
- oi->ip_flags &= ~(OCFS2_INODE_DELETED|OCFS2_INODE_SKIP_DELETE);
-
/* Set the proper information to get us going into
* ocfs2_delete_inode. */
oi->ip_flags |= OCFS2_INODE_MAYBE_ORPHANED;
@@ -2135,8 +2186,20 @@ static int ocfs2_commit_thread(void *arg)
|| kthread_should_stop());
status = ocfs2_commit_cache(osb);
- if (status < 0)
- mlog_errno(status);
+ if (status < 0) {
+ static unsigned long abort_warn_time;
+
+ /* Warn about this once per minute */
+ if (printk_timed_ratelimit(&abort_warn_time, 60*HZ))
+ mlog(ML_ERROR, "status = %d, journal is "
+ "already aborted.\n", status);
+ /*
+ * After ocfs2_commit_cache() fails, j_num_trans has a
+ * non-zero value. Sleep here to avoid a busy-wait
+ * loop.
+ */
+ msleep_interruptible(1000);
+ }
if (kthread_should_stop() && atomic_read(&journal->j_num_trans)){
mlog(ML_KTHREAD,
generated by cgit v1.2.3-70-g09d2 (git 2.46.0) at 2025-11-13 22:23:41 +0000