Lines Matching defs:transaction
20 * transaction's revoked blocks to the journal
22 * + Recovery: during recovery we record the transaction ID of all
29 * single transaction:
33 * cancel the revoke before the transaction commits.
40 * in the current transaction, so any revoke for that block in the
41 * transaction must have happened after the block was journaled and so
61 * running transaction (is pointed to by journal->j_revoke), the other one
62 * belongs to the committing transaction. Accesses to the second hash table
65 * running and which to the committing transaction is called only from
67 * to the committing transaction.
69 * All users operating on the hash table belonging to the running transaction
70 * have a handle to the transaction. Therefore they are safe from kjournald
97 journal replay, this involves recording the transaction ID of the
98 last transaction to revoke this block. */
314 * crash after this current transaction commits. Any subsequent
315 * metadata writes of the buffer in this transaction cancel the
420 * discarded, and then reallocated within the same transaction. In such
482 /* journal_switch_revoke table select j_revoke for next transaction
504 transaction_t *transaction, int write_op)
516 /* select revoke table for committing transaction */
526 write_one_revoke_record(journal, transaction,
545 transaction_t *transaction,
580 header->h_sequence = cpu_to_be32(transaction->t_tid);
584 journal_file_buffer(descriptor, transaction, BJ_LogCtl);
632 * check whether a given block in a given transaction should be replayed
634 * transaction)
665 * that block been revoked? A revoke record with a given transaction
666 * sequence number revokes all blocks in that transaction and earlier