Re: [PATCH][BUGFIX] memcg fix zone congestion

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 





On Thu, May 12, 2011 at 8:10 PM, KAMEZAWA Hiroyuki <kamezawa.hiroyu@xxxxxxxxxxxxxx> wrote:

ZONE_CONGESTED should be a state of global memory reclaim.
If not, a busy memcg sets this and give unnecessary throttoling in
wait_iff_congested() against memory recalim in other contexts. This makes
system performance bad.

I'll think about "memcg is congested!" flag is required or not, later.
But this fix is required 1st.
 
Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@xxxxxxxxxxxxxx>
---
 mm/vmscan.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

Index: mmotm-May11/mm/vmscan.c
===================================================================
--- mmotm-May11.orig/mm/vmscan.c
+++ mmotm-May11/mm/vmscan.c
@@ -941,7 +941,8 @@ keep_lumpy:
        * back off and wait for congestion to clear because further reclaim
        * will encounter the same problem
        */
-       if (nr_dirty == nr_congested && nr_dirty != 0)
+       if (scanning_global_lru(sc) &&
+           nr_dirty == nr_congested && nr_dirty != 0)
               zone_set_flag(zone, ZONE_CONGESTED);

       free_page_list(&free_pages);

For memcg, wonder if we should make it per-memcg-per-zone congested.
 
Acked-by: Ying Han <yinghan@xxxxxxxxxx>

--Ying

[Index of Archives]     [Linux ARM Kernel]     [Linux ARM]     [Linux Omap]     [Fedora ARM]     [IETF Annouce]     [Bugtraq]     [Linux]     [Linux OMAP]     [Linux MIPS]     [ECOS]     [Asterisk Internet PBX]     [Linux API]