branches/zip rb://138

The current implementation is to try to flush the neighbors of every
page that we flush. This patch makes the following distinction:

1) If the flush is from flush_list AND
2) If the flush is intended to move the oldest_modification LSN ahead
(this happens when a user thread sees little space in the log file and
attempts to flush pages from the buffer pool so that a checkpoint can
be made)

THEN

Do not try to flush the neighbors. Just focus on flushing dirty pages at
the end of flush_list

Approved by: Heikki

buf/buf0flu.c

@@ -979,6 +979,7 @@ buf_flush_batch(
 	ulint		old_page_count;
 	ulint		space;
 	ulint		offset;
+	ibool		try_neighbors = TRUE;
 
 	ut_ad((flush_type == BUF_FLUSH_LRU)
 	      || (flush_type == BUF_FLUSH_LIST));
@@ -986,6 +987,17 @@ buf_flush_batch(
 	ut_ad((flush_type != BUF_FLUSH_LIST)
 	      || sync_thread_levels_empty_gen(TRUE));
 #endif /* UNIV_SYNC_DEBUG */
+
+	/* If we are being asked to do a BUF_FLUSH_LIST flush and
+	min_n is ULINT_MAX and lsn_limit is provided then we are doing
+	this flush from within a query thread i.e.: not in background
+	and therefore we should not try to flush the neighbors and just
+	focus on getting the flushed LSN to the lsn_limit. */
+	if (flush_type == BUF_FLUSH_LIST && min_n == ULINT_MAX
+	    && lsn_limit != IB_ULONGLONG_MAX) {
+		try_neighbors = FALSE;
+	}
+
 	buf_pool_mutex_enter();
 
 	if ((buf_pool->n_flush[flush_type] > 0)
@@ -1045,19 +1057,37 @@ flush_next:
 			if (ready) {
 				space = buf_page_get_space(bpage);
 				offset = buf_page_get_page_no(bpage);
+				old_page_count = page_count;
 
-				buf_pool_mutex_exit();
+				if (try_neighbors) {
 
-				old_page_count = page_count;
+					buf_pool_mutex_exit();
 
-				/* Try to flush also all the neighbors */
+					/* Try to flush also all the
+					neighbors */
 					page_count += buf_flush_try_neighbors(
 						space, offset, flush_type);
 					/* fprintf(stderr,
-				"Flush type %lu, page no %lu, neighb %lu\n",
+					"Flush type %lu, page no %lu,"
+					" neighb %lu\n",
 					flush_type, offset,
 					page_count - old_page_count); */
 
+				} else {
+
+					/* Just flush this page. */
+					mutex_enter(block_mutex);
+
+					ut_a(buf_page_in_file(bpage));
+					ut_ad(bpage->in_page_hash);
+
+					/* buf_pool and block mutexes are
+					released inside the following
+					function. */
+					buf_flush_page(bpage, flush_type);
+					++page_count;
+				}
+
 				buf_pool_mutex_enter();
 				goto flush_next;