aboutsummaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
-rw-r--r--Documentation/sysctl/vm.txt12
1 files changed, 10 insertions, 2 deletions
diff --git a/Documentation/sysctl/vm.txt b/Documentation/sysctl/vm.txt
index 96f0ee825bed..84eb25cd69aa 100644
--- a/Documentation/sysctl/vm.txt
+++ b/Documentation/sysctl/vm.txt
@@ -574,16 +574,24 @@ of physical RAM. See above.
574 574
575page-cluster 575page-cluster
576 576
577page-cluster controls the number of pages which are written to swap in 577page-cluster controls the number of pages up to which consecutive pages
578a single attempt. The swap I/O size. 578are read in from swap in a single attempt. This is the swap counterpart
579to page cache readahead.
580The mentioned consecutivity is not in terms of virtual/physical addresses,
581but consecutive on swap space - that means they were swapped out together.
579 582
580It is a logarithmic value - setting it to zero means "1 page", setting 583It is a logarithmic value - setting it to zero means "1 page", setting
581it to 1 means "2 pages", setting it to 2 means "4 pages", etc. 584it to 1 means "2 pages", setting it to 2 means "4 pages", etc.
585Zero disables swap readahead completely.
582 586
583The default value is three (eight pages at a time). There may be some 587The default value is three (eight pages at a time). There may be some
584small benefits in tuning this to a different value if your workload is 588small benefits in tuning this to a different value if your workload is
585swap-intensive. 589swap-intensive.
586 590
591Lower values mean lower latencies for initial faults, but at the same time
592extra faults and I/O delays for following faults if they would have been part of
593that consecutive pages readahead would have brought in.
594
587============================================================= 595=============================================================
588 596
589panic_on_oom 597panic_on_oom