branches/zip: internal mantis issue#49

This is to fix an unintended side effect of file_format_check changes.
We were reading the trx system page (TRX_SYS_PAGE_NO) before starting
recovery and that resulted in redo logs not being applied to the page.
The fix is to force a reread by calling buf_pool_invalidate().
This, however, made necessary that any writes to page are deferred to
until after the redo log application.

Reviewed by: Sunny

include/trx0sys.h

@@ -310,6 +310,15 @@ UNIV_INTERN
 void
 trx_sys_file_format_close(void);
 /*===========================*/
+/************************************************************************
+Tags the system table space with minimum format id if it has not been
+tagged yet.
+WARNING: This function is only called during the startup and AFTER the
+redo log application during recovery has finished. */
+UNIV_INTERN
+void
+trx_sys_file_format_tag_init(void);
+/*==============================*/
 /*********************************************************************
 Get the name representation of the file format from its id. */
 UNIV_INTERN
@@ -319,14 +328,16 @@ trx_sys_file_format_id_to_name(
 					/* out: pointer to the name */
 	const uint	id);		/* in: id of the file format */
 /*********************************************************************
-Set the file format tag unconditonally. */
+Set the file format id unconditionally except if it's already the
+same value. */
 UNIV_INTERN
 ibool
 trx_sys_file_format_max_set(
-/*===========================*/
+/*========================*/
 					/* out: TRUE if value updated */
-	ulint		file_format,	/* in: file format id */
-	char**		name);		/* out: max format name */
+	ulint		format_id,	/* in: file format id */
+	char**		name);		/* out: max file format name or
+					NULL if not needed. */
 /*********************************************************************
 Get the name representation of the file format from its id. */
 UNIV_INTERN

srv/srv0start.c

@@ -1462,7 +1462,17 @@ innobase_start_or_create_for_mysql(void)
 	} else {
 
 		/* Check if we support the max format that is stamped
-		on the system tablespace. */
+		on the system tablespace. 
+		Note:  We are NOT allowed to make any modifications to
+		the TRX_SYS_PAGE_NO page before recovery  because this
+		page also contains the max_trx_id etc. important system
+		variables that are required for recovery.  We need to
+		ensure that we return the system to a state where normal
+		recovery is guaranteed to work. We do this by
+		invalidating the buffer cache, this will force the
+		reread of the page and restoration to it's last known
+		consistent state, this is REQUIRED for the recovery
+		process to work. */
 		err = trx_sys_file_format_max_check(
 			srv_check_file_format_at_startup);
 
@@ -1470,6 +1480,13 @@ innobase_start_or_create_for_mysql(void)
 			return(err);
 		}
 
+		/* Invalidate the buffer pool to ensure that we reread
+		the page that we read above, during recovery.
+		Note that this is not as heavy weight as it seems. At
+		this point there will be only ONE page in the buf_LRU
+		and there must be no page in the buf_flush list. */
+		buf_pool_invalidate();
+
 		/* We always try to do a recovery, even if the database had
 		been shut down normally: this is the normal startup path */
 
@@ -1526,6 +1543,13 @@ innobase_start_or_create_for_mysql(void)
 		are initialized in trx_sys_init_at_db_start(). */
 
 		recv_recovery_from_checkpoint_finish();
+
+		/* It is possible that file_format tag has never
+		been set. In this case we initialize it to minimum
+		value.  Important to note that we can do it ONLY after
+		we have finished the recovery process so that the
+		image of TRX_SYS_PAGE_NO is not stale. */
+		trx_sys_file_format_tag_init();
 	}
 
 	if (!create_new_db && sum_of_new_sizes > 0) {

trx/trx0sys.c

@@ -1106,7 +1106,8 @@ static
 ulint
 trx_sys_file_format_max_read(void)
 /*==============================*/
-				/* out: the file format */
+				/* out: the file format or
+				ULINT_UNDEFINED if not set. */
 {
 	mtr_t			mtr;
 	const byte*		ptr;
@@ -1131,10 +1132,8 @@ trx_sys_file_format_max_read(void)
 	if (file_format_id.high != TRX_SYS_FILE_FORMAT_TAG_MAGIC_N_HIGH
 	    || format_id >= FILE_FORMAT_NAME_N) {
 
-		/* Either it has never been tagged, or garbage in it.
-		Reset the tag in either case. */
-		format_id = DICT_TF_FORMAT_51;
-		trx_sys_file_format_max_write(format_id, NULL);
+		/* Either it has never been tagged, or garbage in it. */
+		return(ULINT_UNDEFINED);
 	}
 
 	return(format_id);
@@ -1170,6 +1169,11 @@ trx_sys_file_format_max_check(
 	recover if the file format is not supported by the engine
 	unless forced by the user. */
 	format_id = trx_sys_file_format_max_read();
+	if (format_id == ULINT_UNDEFINED) {
+		/* Format ID was not set. Set it to minimum possible
+		value. */
+		format_id = DICT_TF_FORMAT_51;
+	}
 
 	ut_print_timestamp(stderr);
 	fprintf(stderr,
@@ -1212,11 +1216,11 @@ trx_sys_file_format_max_set(
 /*========================*/
 					/* out: TRUE if value updated */
 	ulint		format_id,	/* in: file format id */
-	char**		name)		/* out: max file format name */
+	char**		name)		/* out: max file format name or
+					NULL if not needed. */
 {
 	ibool		ret = FALSE;
 
-	ut_a(name);
 	ut_a(format_id <= DICT_TF_FORMAT_MAX);
 
 	mutex_enter(&file_format_max.mutex);
@@ -1232,6 +1236,26 @@ trx_sys_file_format_max_set(
 	return(ret);
 }
 
+/************************************************************************
+Tags the system table space with minimum format id if it has not been
+tagged yet.
+WARNING: This function is only called during the startup and AFTER the
+redo log application during recovery has finished. */
+UNIV_INTERN
+void
+trx_sys_file_format_tag_init(void)
+/*==============================*/
+{
+	ulint	format_id;
+
+	format_id = trx_sys_file_format_max_read();
+
+	/* If format_id is not set then set it to the minimum. */
+	if (format_id == ULINT_UNDEFINED) {
+		trx_sys_file_format_max_set(DICT_TF_FORMAT_51, NULL);
+	}
+}
+
 /************************************************************************
 Update the file format tag in the tablespace only if the given format id
 is greater than the known max id. */