summaryrefslogtreecommitdiffstats
path: root/fs/jbd2/journal.c
diff options
context:
space:
mode:
authorLinus Torvalds <torvalds@linux-foundation.org>2010-08-07 13:03:53 -0700
committerLinus Torvalds <torvalds@linux-foundation.org>2010-08-07 13:03:53 -0700
commit09dc942c2a767e2d298f1cc9294bc19c7d7208c5 (patch)
treed310c118467c90c264e953bdc320ae08394c662a /fs/jbd2/journal.c
parent90e0c225968f0878e090c7ff3f88323973476cee (diff)
parent6c7a120ac6c62316ab1fc78dfc0a7b13f3bfcbff (diff)
downloadop-kernel-dev-09dc942c2a767e2d298f1cc9294bc19c7d7208c5.zip
op-kernel-dev-09dc942c2a767e2d298f1cc9294bc19c7d7208c5.tar.gz
Merge branch 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4
* 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4: (40 commits) ext4: Adding error check after calling ext4_mb_regular_allocator() ext4: Fix dirtying of journalled buffers in data=journal mode ext4: re-inline ext4_rec_len_(to|from)_disk functions jbd2: Remove t_handle_lock from start_this_handle() jbd2: Change j_state_lock to be a rwlock_t jbd2: Use atomic variables to avoid taking t_handle_lock in jbd2_journal_stop ext4: Add mount options in superblock ext4: force block allocation on quota_off ext4: fix freeze deadlock under IO ext4: drop inode from orphan list if ext4_delete_inode() fails ext4: check to make make sure bd_dev is set before dereferencing it jbd2: Make barrier messages less scary ext4: don't print scary messages for allocation failures post-abort ext4: fix EFBIG edge case when writing to large non-extent file ext4: fix ext4_get_blocks references ext4: Always journal quota file modifications ext4: Fix potential memory leak in ext4_fill_super ext4: Don't error out the fs if the user tries to make a file too big ext4: allocate stripe-multiple IOs on stripe boundaries ext4: move aio completion after unwritten extent conversion ... Fix up conflicts in fs/ext4/inode.c as per Ted. Fix up xfs conflicts as per earlier xfs merge.
Diffstat (limited to 'fs/jbd2/journal.c')
-rw-r--r--fs/jbd2/journal.c121
1 files changed, 58 insertions, 63 deletions
diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c
index 0368808..ad5866a 100644
--- a/fs/jbd2/journal.c
+++ b/fs/jbd2/journal.c
@@ -41,6 +41,7 @@
#include <linux/hash.h>
#include <linux/log2.h>
#include <linux/vmalloc.h>
+#include <linux/backing-dev.h>
#define CREATE_TRACE_POINTS
#include <trace/events/jbd2.h>
@@ -48,8 +49,6 @@
#include <asm/uaccess.h>
#include <asm/page.h>
-EXPORT_SYMBOL(jbd2_journal_start);
-EXPORT_SYMBOL(jbd2_journal_restart);
EXPORT_SYMBOL(jbd2_journal_extend);
EXPORT_SYMBOL(jbd2_journal_stop);
EXPORT_SYMBOL(jbd2_journal_lock_updates);
@@ -143,7 +142,7 @@ static int kjournald2(void *arg)
/*
* And now, wait forever for commit wakeup events.
*/
- spin_lock(&journal->j_state_lock);
+ write_lock(&journal->j_state_lock);
loop:
if (journal->j_flags & JBD2_UNMOUNT)
@@ -154,10 +153,10 @@ loop:
if (journal->j_commit_sequence != journal->j_commit_request) {
jbd_debug(1, "OK, requests differ\n");
- spin_unlock(&journal->j_state_lock);
+ write_unlock(&journal->j_state_lock);
del_timer_sync(&journal->j_commit_timer);
jbd2_journal_commit_transaction(journal);
- spin_lock(&journal->j_state_lock);
+ write_lock(&journal->j_state_lock);
goto loop;
}
@@ -169,9 +168,9 @@ loop:
* be already stopped.
*/
jbd_debug(1, "Now suspending kjournald2\n");
- spin_unlock(&journal->j_state_lock);
+ write_unlock(&journal->j_state_lock);
refrigerator();
- spin_lock(&journal->j_state_lock);
+ write_lock(&journal->j_state_lock);
} else {
/*
* We assume on resume that commits are already there,
@@ -191,9 +190,9 @@ loop:
if (journal->j_flags & JBD2_UNMOUNT)
should_sleep = 0;
if (should_sleep) {
- spin_unlock(&journal->j_state_lock);
+ write_unlock(&journal->j_state_lock);
schedule();
- spin_lock(&journal->j_state_lock);
+ write_lock(&journal->j_state_lock);
}
finish_wait(&journal->j_wait_commit, &wait);
}
@@ -211,7 +210,7 @@ loop:
goto loop;
end_loop:
- spin_unlock(&journal->j_state_lock);
+ write_unlock(&journal->j_state_lock);
del_timer_sync(&journal->j_commit_timer);
journal->j_task = NULL;
wake_up(&journal->j_wait_done_commit);
@@ -234,16 +233,16 @@ static int jbd2_journal_start_thread(journal_t *journal)
static void journal_kill_thread(journal_t *journal)
{
- spin_lock(&journal->j_state_lock);
+ write_lock(&journal->j_state_lock);
journal->j_flags |= JBD2_UNMOUNT;
while (journal->j_task) {
wake_up(&journal->j_wait_commit);
- spin_unlock(&journal->j_state_lock);
+ write_unlock(&journal->j_state_lock);
wait_event(journal->j_wait_done_commit, journal->j_task == NULL);
- spin_lock(&journal->j_state_lock);
+ write_lock(&journal->j_state_lock);
}
- spin_unlock(&journal->j_state_lock);
+ write_unlock(&journal->j_state_lock);
}
/*
@@ -310,7 +309,17 @@ int jbd2_journal_write_metadata_buffer(transaction_t *transaction,
*/
J_ASSERT_BH(bh_in, buffer_jbddirty(bh_in));
- new_bh = alloc_buffer_head(GFP_NOFS|__GFP_NOFAIL);
+retry_alloc:
+ new_bh = alloc_buffer_head(GFP_NOFS);
+ if (!new_bh) {
+ /*
+ * Failure is not an option, but __GFP_NOFAIL is going
+ * away; so we retry ourselves here.
+ */
+ congestion_wait(BLK_RW_ASYNC, HZ/50);
+ goto retry_alloc;
+ }
+
/* keep subsequent assertions sane */
new_bh->b_state = 0;
init_buffer(new_bh, NULL, NULL);
@@ -442,7 +451,7 @@ int __jbd2_log_space_left(journal_t *journal)
{
int left = journal->j_free;
- assert_spin_locked(&journal->j_state_lock);
+ /* assert_spin_locked(&journal->j_state_lock); */
/*
* Be pessimistic here about the number of those free blocks which
@@ -487,9 +496,9 @@ int jbd2_log_start_commit(journal_t *journal, tid_t tid)
{
int ret;
- spin_lock(&journal->j_state_lock);
+ write_lock(&journal->j_state_lock);
ret = __jbd2_log_start_commit(journal, tid);
- spin_unlock(&journal->j_state_lock);
+ write_unlock(&journal->j_state_lock);
return ret;
}
@@ -508,7 +517,7 @@ int jbd2_journal_force_commit_nested(journal_t *journal)
transaction_t *transaction = NULL;
tid_t tid;
- spin_lock(&journal->j_state_lock);
+ read_lock(&journal->j_state_lock);
if (journal->j_running_transaction && !current->journal_info) {
transaction = journal->j_running_transaction;
__jbd2_log_start_commit(journal, transaction->t_tid);
@@ -516,12 +525,12 @@ int jbd2_journal_force_commit_nested(journal_t *journal)
transaction = journal->j_committing_transaction;
if (!transaction) {
- spin_unlock(&journal->j_state_lock);
+ read_unlock(&journal->j_state_lock);
return 0; /* Nothing to retry */
}
tid = transaction->t_tid;
- spin_unlock(&journal->j_state_lock);
+ read_unlock(&journal->j_state_lock);
jbd2_log_wait_commit(journal, tid);
return 1;
}
@@ -535,7 +544,7 @@ int jbd2_journal_start_commit(journal_t *journal, tid_t *ptid)
{
int ret = 0;
- spin_lock(&journal->j_state_lock);
+ write_lock(&journal->j_state_lock);
if (journal->j_running_transaction) {
tid_t tid = journal->j_running_transaction->t_tid;
@@ -554,7 +563,7 @@ int jbd2_journal_start_commit(journal_t *journal, tid_t *ptid)
*ptid = journal->j_committing_transaction->t_tid;
ret = 1;
}
- spin_unlock(&journal->j_state_lock);
+ write_unlock(&journal->j_state_lock);
return ret;
}
@@ -566,26 +575,24 @@ int jbd2_log_wait_commit(journal_t *journal, tid_t tid)
{
int err = 0;
+ read_lock(&journal->j_state_lock);
#ifdef CONFIG_JBD2_DEBUG
- spin_lock(&journal->j_state_lock);
if (!tid_geq(journal->j_commit_request, tid)) {
printk(KERN_EMERG
"%s: error: j_commit_request=%d, tid=%d\n",
__func__, journal->j_commit_request, tid);
}
- spin_unlock(&journal->j_state_lock);
#endif
- spin_lock(&journal->j_state_lock);
while (tid_gt(tid, journal->j_commit_sequence)) {
jbd_debug(1, "JBD: want %d, j_commit_sequence=%d\n",
tid, journal->j_commit_sequence);
wake_up(&journal->j_wait_commit);
- spin_unlock(&journal->j_state_lock);
+ read_unlock(&journal->j_state_lock);
wait_event(journal->j_wait_done_commit,
!tid_gt(tid, journal->j_commit_sequence));
- spin_lock(&journal->j_state_lock);
+ read_lock(&journal->j_state_lock);
}
- spin_unlock(&journal->j_state_lock);
+ read_unlock(&journal->j_state_lock);
if (unlikely(is_journal_aborted(journal))) {
printk(KERN_EMERG "journal commit I/O error\n");
@@ -602,7 +609,7 @@ int jbd2_journal_next_log_block(journal_t *journal, unsigned long long *retp)
{
unsigned long blocknr;
- spin_lock(&journal->j_state_lock);
+ write_lock(&journal->j_state_lock);
J_ASSERT(journal->j_free > 1);
blocknr = journal->j_head;
@@ -610,7 +617,7 @@ int jbd2_journal_next_log_block(journal_t *journal, unsigned long long *retp)
journal->j_free--;
if (journal->j_head == journal->j_last)
journal->j_head = journal->j_first;
- spin_unlock(&journal->j_state_lock);
+ write_unlock(&journal->j_state_lock);
return jbd2_journal_bmap(journal, blocknr, retp);
}
@@ -830,7 +837,7 @@ static journal_t * journal_init_common (void)
mutex_init(&journal->j_checkpoint_mutex);
spin_lock_init(&journal->j_revoke_lock);
spin_lock_init(&journal->j_list_lock);
- spin_lock_init(&journal->j_state_lock);
+ rwlock_init(&journal->j_state_lock);
journal->j_commit_interval = (HZ * JBD2_DEFAULT_MAX_COMMIT_AGE);
journal->j_min_batch_time = 0;
@@ -1096,14 +1103,14 @@ void jbd2_journal_update_superblock(journal_t *journal, int wait)
set_buffer_uptodate(bh);
}
- spin_lock(&journal->j_state_lock);
+ read_lock(&journal->j_state_lock);
jbd_debug(1,"JBD: updating superblock (start %ld, seq %d, errno %d)\n",
journal->j_tail, journal->j_tail_sequence, journal->j_errno);
sb->s_sequence = cpu_to_be32(journal->j_tail_sequence);
sb->s_start = cpu_to_be32(journal->j_tail);
sb->s_errno = cpu_to_be32(journal->j_errno);
- spin_unlock(&journal->j_state_lock);
+ read_unlock(&journal->j_state_lock);
BUFFER_TRACE(bh, "marking dirty");
mark_buffer_dirty(bh);
@@ -1124,12 +1131,12 @@ out:
* any future commit will have to be careful to update the
* superblock again to re-record the true start of the log. */
- spin_lock(&journal->j_state_lock);
+ write_lock(&journal->j_state_lock);
if (sb->s_start)
journal->j_flags &= ~JBD2_FLUSHED;
else
journal->j_flags |= JBD2_FLUSHED;
- spin_unlock(&journal->j_state_lock);
+ write_unlock(&journal->j_state_lock);
}
/*
@@ -1391,13 +1398,9 @@ int jbd2_journal_check_used_features (journal_t *journal, unsigned long compat,
int jbd2_journal_check_available_features (journal_t *journal, unsigned long compat,
unsigned long ro, unsigned long incompat)
{
- journal_superblock_t *sb;
-
if (!compat && !ro && !incompat)
return 1;
- sb = journal->j_superblock;
-
/* We can support any known requested features iff the
* superblock is in version 2. Otherwise we fail to support any
* extended sb features. */
@@ -1545,7 +1548,7 @@ int jbd2_journal_flush(journal_t *journal)
transaction_t *transaction = NULL;
unsigned long old_tail;
- spin_lock(&journal->j_state_lock);
+ write_lock(&journal->j_state_lock);
/* Force everything buffered to the log... */
if (journal->j_running_transaction) {
@@ -1558,10 +1561,10 @@ int jbd2_journal_flush(journal_t *journal)
if (transaction) {
tid_t tid = transaction->t_tid;
- spin_unlock(&journal->j_state_lock);
+ write_unlock(&journal->j_state_lock);
jbd2_log_wait_commit(journal, tid);
} else {
- spin_unlock(&journal->j_state_lock);
+ write_unlock(&journal->j_state_lock);
}
/* ...and flush everything in the log out to disk. */
@@ -1585,12 +1588,12 @@ int jbd2_journal_flush(journal_t *journal)
* the magic code for a fully-recovered superblock. Any future
* commits of data to the journal will restore the current
* s_start value. */
- spin_lock(&journal->j_state_lock);
+ write_lock(&journal->j_state_lock);
old_tail = journal->j_tail;
journal->j_tail = 0;
- spin_unlock(&journal->j_state_lock);
+ write_unlock(&journal->j_state_lock);
jbd2_journal_update_superblock(journal, 1);
- spin_lock(&journal->j_state_lock);
+ write_lock(&journal->j_state_lock);
journal->j_tail = old_tail;
J_ASSERT(!journal->j_running_transaction);
@@ -1598,7 +1601,7 @@ int jbd2_journal_flush(journal_t *journal)
J_ASSERT(!journal->j_checkpoint_transactions);
J_ASSERT(journal->j_head == journal->j_tail);
J_ASSERT(journal->j_tail_sequence == journal->j_transaction_sequence);
- spin_unlock(&journal->j_state_lock);
+ write_unlock(&journal->j_state_lock);
return 0;
}
@@ -1617,7 +1620,6 @@ int jbd2_journal_flush(journal_t *journal)
int jbd2_journal_wipe(journal_t *journal, int write)
{
- journal_superblock_t *sb;
int err = 0;
J_ASSERT (!(journal->j_flags & JBD2_LOADED));
@@ -1626,8 +1628,6 @@ int jbd2_journal_wipe(journal_t *journal, int write)
if (err)
return err;
- sb = journal->j_superblock;
-
if (!journal->j_tail)
goto no_recovery;
@@ -1665,12 +1665,12 @@ void __jbd2_journal_abort_hard(journal_t *journal)
printk(KERN_ERR "Aborting journal on device %s.\n",
journal->j_devname);
- spin_lock(&journal->j_state_lock);
+ write_lock(&journal->j_state_lock);
journal->j_flags |= JBD2_ABORT;
transaction = journal->j_running_transaction;
if (transaction)
__jbd2_log_start_commit(journal, transaction->t_tid);
- spin_unlock(&journal->j_state_lock);
+ write_unlock(&journal->j_state_lock);
}
/* Soft abort: record the abort error status in the journal superblock,
@@ -1755,12 +1755,12 @@ int jbd2_journal_errno(journal_t *journal)
{
int err;
- spin_lock(&journal->j_state_lock);
+ read_lock(&journal->j_state_lock);
if (journal->j_flags & JBD2_ABORT)
err = -EROFS;
else
err = journal->j_errno;
- spin_unlock(&journal->j_state_lock);
+ read_unlock(&journal->j_state_lock);
return err;
}
@@ -1775,12 +1775,12 @@ int jbd2_journal_clear_err(journal_t *journal)
{
int err = 0;
- spin_lock(&journal->j_state_lock);
+ write_lock(&journal->j_state_lock);
if (journal->j_flags & JBD2_ABORT)
err = -EROFS;
else
journal->j_errno = 0;
- spin_unlock(&journal->j_state_lock);
+ write_unlock(&journal->j_state_lock);
return err;
}
@@ -1793,10 +1793,10 @@ int jbd2_journal_clear_err(journal_t *journal)
*/
void jbd2_journal_ack_err(journal_t *journal)
{
- spin_lock(&journal->j_state_lock);
+ write_lock(&journal->j_state_lock);
if (journal->j_errno)
journal->j_flags |= JBD2_ACK_ERR;
- spin_unlock(&journal->j_state_lock);
+ write_unlock(&journal->j_state_lock);
}
int jbd2_journal_blocks_per_page(struct inode *inode)
@@ -2201,8 +2201,6 @@ void jbd2_journal_init_jbd_inode(struct jbd2_inode *jinode, struct inode *inode)
void jbd2_journal_release_jbd_inode(journal_t *journal,
struct jbd2_inode *jinode)
{
- int writeout = 0;
-
if (!journal)
return;
restart:
@@ -2219,9 +2217,6 @@ restart:
goto restart;
}
- /* Do we need to wait for data writeback? */
- if (journal->j_committing_transaction == jinode->i_transaction)
- writeout = 1;
if (jinode->i_transaction) {
list_del(&jinode->i_list);
jinode->i_transaction = NULL;
OpenPOWER on IntegriCloud