/*
- * linux/fs/revoke.c
+ * linux/fs/jbd2/revoke.c
*
* Written by Stephen C. Tweedie <sct@redhat.com>, 2000
*
* need do nothing.
* RevokeValid set, Revoked set:
* buffer has been revoked.
+ *
+ * Locking rules:
+ * We keep two hash tables of revoke records. One hashtable belongs to the
+ * running transaction (is pointed to by journal->j_revoke), the other one
+ * belongs to the committing transaction. Accesses to the second hash table
+ * happen only from the kjournald and no other thread touches this table. Also
+ * journal_switch_revoke_table() which switches which hashtable belongs to the
+ * running and which to the committing transaction is called only from
+ * kjournald. Therefore we need no locks when accessing the hashtable belonging
+ * to the committing transaction.
+ *
+ * All users operating on the hash table belonging to the running transaction
+ * have a handle to the transaction. Therefore they are safe from kjournald
+ * switching hash tables under them. For operations on the lists of entries in
+ * the hash table j_revoke_lock is used.
+ *
+ * Finally, also replay code uses the hash tables but at this moment noone else
+ * can touch them (filesystem isn't mounted yet) and hence no locking is
+ * needed.
*/
#ifndef __KERNEL__
#else
#include <linux/time.h>
#include <linux/fs.h>
-#include <linux/jbd.h>
+#include <linux/jbd2.h>
#include <linux/errno.h>
#include <linux/slab.h>
#include <linux/list.h>
-#include <linux/smp_lock.h>
#include <linux/init.h>
+#include <linux/bio.h>
#endif
+#include <linux/log2.h>
-static kmem_cache_t *revoke_record_cache;
-static kmem_cache_t *revoke_table_cache;
+static struct kmem_cache *jbd2_revoke_record_cache;
+static struct kmem_cache *jbd2_revoke_table_cache;
/* Each revoke record represents one single revoked block. During
journal replay, this involves recording the transaction ID of the
last transaction to revoke this block. */
-struct jbd_revoke_record_s
+struct jbd2_revoke_record_s
{
struct list_head hash;
tid_t sequence; /* Used for recovery only */
- unsigned long blocknr;
+ unsigned long long blocknr;
};
/* The revoke table is just a simple hash table of revoke records. */
-struct jbd_revoke_table_s
+struct jbd2_revoke_table_s
{
/* It is conceivable that we might want a larger hash table
* for recovery. Must be a power of two. */
#ifdef __KERNEL__
static void write_one_revoke_record(journal_t *, transaction_t *,
struct journal_head **, int *,
- struct jbd_revoke_record_s *);
-static void flush_descriptor(journal_t *, struct journal_head *, int);
+ struct jbd2_revoke_record_s *, int);
+static void flush_descriptor(journal_t *, struct journal_head *, int, int);
#endif
/* Utility functions to maintain the revoke table */
/* Borrowed from buffer.c: this is a tried and tested block hash function */
-static inline int hash(journal_t *journal, unsigned long block)
+static inline int hash(journal_t *journal, unsigned long long block)
{
- struct jbd_revoke_table_s *table = journal->j_revoke;
+ struct jbd2_revoke_table_s *table = journal->j_revoke;
int hash_shift = table->hash_shift;
+ int hash = (int)block ^ (int)((block >> 31) >> 1);
- return ((block << (hash_shift - 6)) ^
- (block >> 13) ^
- (block << (hash_shift - 12))) & (table->hash_size - 1);
+ return ((hash << (hash_shift - 6)) ^
+ (hash >> 13) ^
+ (hash << (hash_shift - 12))) & (table->hash_size - 1);
}
-static int insert_revoke_hash(journal_t *journal, unsigned long blocknr,
+static int insert_revoke_hash(journal_t *journal, unsigned long long blocknr,
tid_t seq)
{
struct list_head *hash_list;
- struct jbd_revoke_record_s *record;
+ struct jbd2_revoke_record_s *record;
repeat:
- record = kmem_cache_alloc(revoke_record_cache, GFP_NOFS);
+ record = kmem_cache_alloc(jbd2_revoke_record_cache, GFP_NOFS);
if (!record)
goto oom;
oom:
if (!journal_oom_retry)
return -ENOMEM;
- jbd_debug(1, "ENOMEM in %s, retrying\n", __FUNCTION__);
+ jbd_debug(1, "ENOMEM in %s, retrying\n", __func__);
yield();
goto repeat;
}
/* Find a revoke record in the journal's hash table. */
-static struct jbd_revoke_record_s *find_revoke_record(journal_t *journal,
- unsigned long blocknr)
+static struct jbd2_revoke_record_s *find_revoke_record(journal_t *journal,
+ unsigned long long blocknr)
{
struct list_head *hash_list;
- struct jbd_revoke_record_s *record;
+ struct jbd2_revoke_record_s *record;
hash_list = &journal->j_revoke->hash_table[hash(journal, blocknr)];
spin_lock(&journal->j_revoke_lock);
- record = (struct jbd_revoke_record_s *) hash_list->next;
+ record = (struct jbd2_revoke_record_s *) hash_list->next;
while (&(record->hash) != hash_list) {
if (record->blocknr == blocknr) {
spin_unlock(&journal->j_revoke_lock);
return record;
}
- record = (struct jbd_revoke_record_s *) record->hash.next;
+ record = (struct jbd2_revoke_record_s *) record->hash.next;
}
spin_unlock(&journal->j_revoke_lock);
return NULL;
}
-int __init journal_init_revoke_caches(void)
+void jbd2_journal_destroy_revoke_caches(void)
{
- revoke_record_cache = kmem_cache_create("revoke_record",
- sizeof(struct jbd_revoke_record_s),
- 0, SLAB_HWCACHE_ALIGN, NULL, NULL);
- if (revoke_record_cache == 0)
- return -ENOMEM;
-
- revoke_table_cache = kmem_cache_create("revoke_table",
- sizeof(struct jbd_revoke_table_s),
- 0, 0, NULL, NULL);
- if (revoke_table_cache == 0) {
- kmem_cache_destroy(revoke_record_cache);
- revoke_record_cache = NULL;
- return -ENOMEM;
+ if (jbd2_revoke_record_cache) {
+ kmem_cache_destroy(jbd2_revoke_record_cache);
+ jbd2_revoke_record_cache = NULL;
+ }
+ if (jbd2_revoke_table_cache) {
+ kmem_cache_destroy(jbd2_revoke_table_cache);
+ jbd2_revoke_table_cache = NULL;
}
- return 0;
}
-void journal_destroy_revoke_caches(void)
+int __init jbd2_journal_init_revoke_caches(void)
{
- kmem_cache_destroy(revoke_record_cache);
- revoke_record_cache = NULL;
- kmem_cache_destroy(revoke_table_cache);
- revoke_table_cache = NULL;
+ J_ASSERT(!jbd2_revoke_record_cache);
+ J_ASSERT(!jbd2_revoke_table_cache);
+
+ jbd2_revoke_record_cache = kmem_cache_create("jbd2_revoke_record",
+ sizeof(struct jbd2_revoke_record_s),
+ 0,
+ SLAB_HWCACHE_ALIGN|SLAB_TEMPORARY,
+ NULL);
+ if (!jbd2_revoke_record_cache)
+ goto record_cache_failure;
+
+ jbd2_revoke_table_cache = kmem_cache_create("jbd2_revoke_table",
+ sizeof(struct jbd2_revoke_table_s),
+ 0, SLAB_TEMPORARY, NULL);
+ if (!jbd2_revoke_table_cache)
+ goto table_cache_failure;
+ return 0;
+table_cache_failure:
+ jbd2_journal_destroy_revoke_caches();
+record_cache_failure:
+ return -ENOMEM;
}
-/* Initialise the revoke table for a given journal to a given size. */
-
-int journal_init_revoke(journal_t *journal, int hash_size)
+static struct jbd2_revoke_table_s *jbd2_journal_init_revoke_table(int hash_size)
{
- int shift, tmp;
+ int shift = 0;
+ int tmp = hash_size;
+ struct jbd2_revoke_table_s *table;
- J_ASSERT (journal->j_revoke_table[0] == NULL);
+ table = kmem_cache_alloc(jbd2_revoke_table_cache, GFP_KERNEL);
+ if (!table)
+ goto out;
- shift = 0;
- tmp = hash_size;
while((tmp >>= 1UL) != 0UL)
shift++;
- journal->j_revoke_table[0] = kmem_cache_alloc(revoke_table_cache, GFP_KERNEL);
- if (!journal->j_revoke_table[0])
- return -ENOMEM;
- journal->j_revoke = journal->j_revoke_table[0];
-
- /* Check that the hash_size is a power of two */
- J_ASSERT ((hash_size & (hash_size-1)) == 0);
-
- journal->j_revoke->hash_size = hash_size;
-
- journal->j_revoke->hash_shift = shift;
-
- journal->j_revoke->hash_table =
+ table->hash_size = hash_size;
+ table->hash_shift = shift;
+ table->hash_table =
kmalloc(hash_size * sizeof(struct list_head), GFP_KERNEL);
- if (!journal->j_revoke->hash_table) {
- kmem_cache_free(revoke_table_cache, journal->j_revoke_table[0]);
- journal->j_revoke = NULL;
- return -ENOMEM;
+ if (!table->hash_table) {
+ kmem_cache_free(jbd2_revoke_table_cache, table);
+ table = NULL;
+ goto out;
}
for (tmp = 0; tmp < hash_size; tmp++)
- INIT_LIST_HEAD(&journal->j_revoke->hash_table[tmp]);
-
- journal->j_revoke_table[1] = kmem_cache_alloc(revoke_table_cache, GFP_KERNEL);
- if (!journal->j_revoke_table[1]) {
- kfree(journal->j_revoke_table[0]->hash_table);
- kmem_cache_free(revoke_table_cache, journal->j_revoke_table[0]);
- return -ENOMEM;
- }
-
- journal->j_revoke = journal->j_revoke_table[1];
-
- /* Check that the hash_size is a power of two */
- J_ASSERT ((hash_size & (hash_size-1)) == 0);
+ INIT_LIST_HEAD(&table->hash_table[tmp]);
- journal->j_revoke->hash_size = hash_size;
+out:
+ return table;
+}
- journal->j_revoke->hash_shift = shift;
+static void jbd2_journal_destroy_revoke_table(struct jbd2_revoke_table_s *table)
+{
+ int i;
+ struct list_head *hash_list;
- journal->j_revoke->hash_table =
- kmalloc(hash_size * sizeof(struct list_head), GFP_KERNEL);
- if (!journal->j_revoke->hash_table) {
- kfree(journal->j_revoke_table[0]->hash_table);
- kmem_cache_free(revoke_table_cache, journal->j_revoke_table[0]);
- kmem_cache_free(revoke_table_cache, journal->j_revoke_table[1]);
- journal->j_revoke = NULL;
- return -ENOMEM;
+ for (i = 0; i < table->hash_size; i++) {
+ hash_list = &table->hash_table[i];
+ J_ASSERT(list_empty(hash_list));
}
- for (tmp = 0; tmp < hash_size; tmp++)
- INIT_LIST_HEAD(&journal->j_revoke->hash_table[tmp]);
-
- spin_lock_init(&journal->j_revoke_lock);
-
- return 0;
+ kfree(table->hash_table);
+ kmem_cache_free(jbd2_revoke_table_cache, table);
}
-/* Destoy a journal's revoke table. The table must already be empty! */
-
-void journal_destroy_revoke(journal_t *journal)
+/* Initialise the revoke table for a given journal to a given size. */
+int jbd2_journal_init_revoke(journal_t *journal, int hash_size)
{
- struct jbd_revoke_table_s *table;
- struct list_head *hash_list;
- int i;
+ J_ASSERT(journal->j_revoke_table[0] == NULL);
+ J_ASSERT(is_power_of_2(hash_size));
- table = journal->j_revoke_table[0];
- if (!table)
- return;
+ journal->j_revoke_table[0] = jbd2_journal_init_revoke_table(hash_size);
+ if (!journal->j_revoke_table[0])
+ goto fail0;
- for (i=0; i<table->hash_size; i++) {
- hash_list = &table->hash_table[i];
- J_ASSERT (list_empty(hash_list));
- }
+ journal->j_revoke_table[1] = jbd2_journal_init_revoke_table(hash_size);
+ if (!journal->j_revoke_table[1])
+ goto fail1;
- kfree(table->hash_table);
- kmem_cache_free(revoke_table_cache, table);
- journal->j_revoke = NULL;
+ journal->j_revoke = journal->j_revoke_table[1];
- table = journal->j_revoke_table[1];
- if (!table)
- return;
+ spin_lock_init(&journal->j_revoke_lock);
- for (i=0; i<table->hash_size; i++) {
- hash_list = &table->hash_table[i];
- J_ASSERT (list_empty(hash_list));
- }
+ return 0;
- kfree(table->hash_table);
- kmem_cache_free(revoke_table_cache, table);
+fail1:
+ jbd2_journal_destroy_revoke_table(journal->j_revoke_table[0]);
+fail0:
+ return -ENOMEM;
+}
+
+/* Destroy a journal's revoke table. The table must already be empty! */
+void jbd2_journal_destroy_revoke(journal_t *journal)
+{
journal->j_revoke = NULL;
+ if (journal->j_revoke_table[0])
+ jbd2_journal_destroy_revoke_table(journal->j_revoke_table[0]);
+ if (journal->j_revoke_table[1])
+ jbd2_journal_destroy_revoke_table(journal->j_revoke_table[1]);
}
#ifdef __KERNEL__
/*
- * journal_revoke: revoke a given buffer_head from the journal. This
+ * jbd2_journal_revoke: revoke a given buffer_head from the journal. This
* prevents the block from being replayed during recovery if we take a
* crash after this current transaction commits. Any subsequent
* metadata writes of the buffer in this transaction cancel the
* revoke before clearing the block bitmap when we are deleting
* metadata.
*
- * Revoke performs a journal_forget on any buffer_head passed in as a
+ * Revoke performs a jbd2_journal_forget on any buffer_head passed in as a
* parameter, but does _not_ forget the buffer_head if the bh was only
* found implicitly.
*
* bh_in may not be a journalled buffer - it may have come off
* the hash tables without an attached journal_head.
*
- * If bh_in is non-zero, journal_revoke() will decrement its b_count
+ * If bh_in is non-zero, jbd2_journal_revoke() will decrement its b_count
* by one.
*/
-int journal_revoke(handle_t *handle, unsigned long blocknr,
+int jbd2_journal_revoke(handle_t *handle, unsigned long long blocknr,
struct buffer_head *bh_in)
{
struct buffer_head *bh = NULL;
BUFFER_TRACE(bh_in, "enter");
journal = handle->h_transaction->t_journal;
- if (!journal_set_features(journal, 0, 0, JFS_FEATURE_INCOMPAT_REVOKE)){
+ if (!jbd2_journal_set_features(journal, 0, 0, JBD2_FEATURE_INCOMPAT_REVOKE)){
J_ASSERT (!"Cannot set revoke feature!");
return -EINVAL;
}
if (bh)
BUFFER_TRACE(bh, "found on hash");
}
-#ifdef JBD_EXPENSIVE_CHECKING
+#ifdef JBD2_EXPENSIVE_CHECKING
else {
struct buffer_head *bh2;
set_buffer_revoked(bh);
set_buffer_revokevalid(bh);
if (bh_in) {
- BUFFER_TRACE(bh_in, "call journal_forget");
- journal_forget(handle, bh_in);
+ BUFFER_TRACE(bh_in, "call jbd2_journal_forget");
+ jbd2_journal_forget(handle, bh_in);
} else {
BUFFER_TRACE(bh, "call brelse");
__brelse(bh);
}
}
- jbd_debug(2, "insert revoke for block %lu, bh_in=%p\n", blocknr, bh_in);
+ jbd_debug(2, "insert revoke for block %llu, bh_in=%p\n",blocknr, bh_in);
err = insert_revoke_hash(journal, blocknr,
handle->h_transaction->t_tid);
BUFFER_TRACE(bh_in, "exit");
/*
* Cancel an outstanding revoke. For use only internally by the
- * journaling code (called from journal_get_write_access).
+ * journaling code (called from jbd2_journal_get_write_access).
*
* We trust buffer_revoked() on the buffer if the buffer is already
* being journaled: if there is no revoke pending on the buffer, then we
* the second time we would still have a pending revoke to cancel. So,
* do not trust the Revoked bit on buffers unless RevokeValid is also
* set.
- *
- * The caller must have the journal locked.
*/
-int journal_cancel_revoke(handle_t *handle, struct journal_head *jh)
+int jbd2_journal_cancel_revoke(handle_t *handle, struct journal_head *jh)
{
- struct jbd_revoke_record_s *record;
+ struct jbd2_revoke_record_s *record;
journal_t *journal = handle->h_transaction->t_journal;
int need_cancel;
int did_revoke = 0; /* akpm: debug */
spin_lock(&journal->j_revoke_lock);
list_del(&record->hash);
spin_unlock(&journal->j_revoke_lock);
- kmem_cache_free(revoke_record_cache, record);
+ kmem_cache_free(jbd2_revoke_record_cache, record);
did_revoke = 1;
}
}
-#ifdef JBD_EXPENSIVE_CHECKING
+#ifdef JBD2_EXPENSIVE_CHECKING
/* There better not be one left behind by now! */
record = find_revoke_record(journal, bh->b_blocknr);
J_ASSERT_JH(jh, record == NULL);
* we do not want to suspend any processing until all revokes are
* written -bzzz
*/
-void journal_switch_revoke_table(journal_t *journal)
+void jbd2_journal_switch_revoke_table(journal_t *journal)
{
int i;
/*
* Write revoke records to the journal for all entries in the current
* revoke hash, deleting the entries as we go.
- *
- * Called with the journal lock held.
*/
-
-void journal_write_revoke_records(journal_t *journal,
- transaction_t *transaction)
+void jbd2_journal_write_revoke_records(journal_t *journal,
+ transaction_t *transaction,
+ int write_op)
{
struct journal_head *descriptor;
- struct jbd_revoke_record_s *record;
- struct jbd_revoke_table_s *revoke;
+ struct jbd2_revoke_record_s *record;
+ struct jbd2_revoke_table_s *revoke;
struct list_head *hash_list;
int i, offset, count;
hash_list = &revoke->hash_table[i];
while (!list_empty(hash_list)) {
- record = (struct jbd_revoke_record_s *)
+ record = (struct jbd2_revoke_record_s *)
hash_list->next;
write_one_revoke_record(journal, transaction,
&descriptor, &offset,
- record);
+ record, write_op);
count++;
list_del(&record->hash);
- kmem_cache_free(revoke_record_cache, record);
+ kmem_cache_free(jbd2_revoke_record_cache, record);
}
}
if (descriptor)
- flush_descriptor(journal, descriptor, offset);
+ flush_descriptor(journal, descriptor, offset, write_op);
jbd_debug(1, "Wrote %d revoke records\n", count);
}
transaction_t *transaction,
struct journal_head **descriptorp,
int *offsetp,
- struct jbd_revoke_record_s *record)
+ struct jbd2_revoke_record_s *record,
+ int write_op)
{
struct journal_head *descriptor;
int offset;
/* If we are already aborting, this all becomes a noop. We
still need to go round the loop in
- journal_write_revoke_records in order to free all of the
+ jbd2_journal_write_revoke_records in order to free all of the
revoke records: only the IO to the journal is omitted. */
if (is_journal_aborted(journal))
return;
/* Make sure we have a descriptor with space left for the record */
if (descriptor) {
if (offset == journal->j_blocksize) {
- flush_descriptor(journal, descriptor, offset);
+ flush_descriptor(journal, descriptor, offset, write_op);
descriptor = NULL;
}
}
if (!descriptor) {
- descriptor = journal_get_descriptor_buffer(journal);
+ descriptor = jbd2_journal_get_descriptor_buffer(journal);
if (!descriptor)
return;
header = (journal_header_t *) &jh2bh(descriptor)->b_data[0];
- header->h_magic = cpu_to_be32(JFS_MAGIC_NUMBER);
- header->h_blocktype = cpu_to_be32(JFS_REVOKE_BLOCK);
+ header->h_magic = cpu_to_be32(JBD2_MAGIC_NUMBER);
+ header->h_blocktype = cpu_to_be32(JBD2_REVOKE_BLOCK);
header->h_sequence = cpu_to_be32(transaction->t_tid);
/* Record it so that we can wait for IO completion later */
JBUFFER_TRACE(descriptor, "file as BJ_LogCtl");
- journal_file_buffer(descriptor, transaction, BJ_LogCtl);
+ jbd2_journal_file_buffer(descriptor, transaction, BJ_LogCtl);
- offset = sizeof(journal_revoke_header_t);
+ offset = sizeof(jbd2_journal_revoke_header_t);
*descriptorp = descriptor;
}
- * ((__be32 *)(&jh2bh(descriptor)->b_data[offset])) =
- cpu_to_be32(record->blocknr);
- offset += 4;
+ if (JBD2_HAS_INCOMPAT_FEATURE(journal, JBD2_FEATURE_INCOMPAT_64BIT)) {
+ * ((__be64 *)(&jh2bh(descriptor)->b_data[offset])) =
+ cpu_to_be64(record->blocknr);
+ offset += 8;
+
+ } else {
+ * ((__be32 *)(&jh2bh(descriptor)->b_data[offset])) =
+ cpu_to_be32(record->blocknr);
+ offset += 4;
+ }
+
*offsetp = offset;
}
static void flush_descriptor(journal_t *journal,
struct journal_head *descriptor,
- int offset)
+ int offset, int write_op)
{
- journal_revoke_header_t *header;
+ jbd2_journal_revoke_header_t *header;
struct buffer_head *bh = jh2bh(descriptor);
if (is_journal_aborted(journal)) {
return;
}
- header = (journal_revoke_header_t *) jh2bh(descriptor)->b_data;
+ header = (jbd2_journal_revoke_header_t *) jh2bh(descriptor)->b_data;
header->r_count = cpu_to_be32(offset);
set_buffer_jwrite(bh);
BUFFER_TRACE(bh, "write");
set_buffer_dirty(bh);
- ll_rw_block(SWRITE, 1, &bh);
+ ll_rw_block((write_op == WRITE) ? SWRITE : SWRITE_SYNC_PLUG, 1, &bh);
}
#endif
* single block.
*/
-int journal_set_revoke(journal_t *journal,
- unsigned long blocknr,
+int jbd2_journal_set_revoke(journal_t *journal,
+ unsigned long long blocknr,
tid_t sequence)
{
- struct jbd_revoke_record_s *record;
+ struct jbd2_revoke_record_s *record;
record = find_revoke_record(journal, blocknr);
if (record) {
* ones, but later transactions still need replayed.
*/
-int journal_test_revoke(journal_t *journal,
- unsigned long blocknr,
+int jbd2_journal_test_revoke(journal_t *journal,
+ unsigned long long blocknr,
tid_t sequence)
{
- struct jbd_revoke_record_s *record;
+ struct jbd2_revoke_record_s *record;
record = find_revoke_record(journal, blocknr);
if (!record)
* that it can be reused by the running filesystem.
*/
-void journal_clear_revoke(journal_t *journal)
+void jbd2_journal_clear_revoke(journal_t *journal)
{
int i;
struct list_head *hash_list;
- struct jbd_revoke_record_s *record;
- struct jbd_revoke_table_s *revoke;
+ struct jbd2_revoke_record_s *record;
+ struct jbd2_revoke_table_s *revoke;
revoke = journal->j_revoke;
for (i = 0; i < revoke->hash_size; i++) {
hash_list = &revoke->hash_table[i];
while (!list_empty(hash_list)) {
- record = (struct jbd_revoke_record_s*) hash_list->next;
+ record = (struct jbd2_revoke_record_s*) hash_list->next;
list_del(&record->hash);
- kmem_cache_free(revoke_record_cache, record);
+ kmem_cache_free(jbd2_revoke_record_cache, record);
}
}
}