page-allocator: reset wmark_min and inactive ratio of zone when hotplug happens

Solve two problems.

Whenever memory hotplug sucessfully happens, zone->present_pages
have to be changed.

1) Now memory hotplug calls setup_per_zone_wmark_min only when
   online_pages called, not offline_pages.

   It breaks balance.

2) If zone->present_pages is changed, we also have to change
   zone->inactive_ratio.  That's because inactive_ratio depends on
   zone->present_pages.

Signed-off-by: Minchan Kim <minchan.kim@gmail.com>
Acked-by: Yasunori Goto <y-goto@jp.fujitsu.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c
index 037291e..e4412a6 100644 (file)
--- a/mm/memory_hotplug.c
+++ b/mm/memory_hotplug.c
@@ -423,6 +423,7 @@ int online_pages(unsigned long pfn, unsigned long nr_pages)
        zone->zone_pgdat->node_present_pages += onlined_pages;
 
        setup_per_zone_wmarks();
+       calculate_zone_inactive_ratio(zone);
        if (onlined_pages) {
                kswapd_run(zone_to_nid(zone));
                node_set_state(zone_to_nid(zone), N_HIGH_MEMORY);
@@ -832,6 +833,9 @@ repeat:
        totalram_pages -= offlined_pages;
        num_physpages -= offlined_pages;
 
+       setup_per_zone_wmarks();
+       calculate_zone_inactive_ratio(zone);
+
        vm_total_pages = nr_free_pagecache_pages();
        writeback_set_ratelimit();
 

diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index 303607f..00e2937 100644 (file)
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -4397,7 +4397,7 @@ static void setup_per_zone_lowmem_reserve(void)
 
 /**
  * setup_per_zone_wmarks - called when min_free_kbytes changes
- * or when memory is hot-added
+ * or when memory is hot-{added|removed}
  *
  * Ensures that the watermark[min,low,high] values for each zone are set
  * correctly with respect to min_free_kbytes.