problem with ZONE_MOVABLE.

[Balbir Singh]

KAMEZAWA Hiroyuki wrote:
> Hi, 
> 
> While I'm playing with memory controller of 2.6.23-rc4-mm1, I met following.
> 
> ==
> [root at drpq test-2.6.23-rc4-mm1]# echo $$ > /opt/mem_control/group_1/tasks
> [root at drpq test-2.6.23-rc4-mm1]# cat /opt/mem_control/group_1/memory.limit
> 32768
> [root at drpq test-2.6.23-rc4-mm1]# cat /opt/mem_control/group_1/memory.usage
> 286
> // Memory is limited to 512 GiB. try "dd" 1GiB (page size is 16KB)
> 
> [root at drpq test-2.6.23-rc4-mm1]# dd if=/dev/zero of=/tmp/tmpfile bs=1024 count=1048576
> Killed
> [root at drpq test-2.6.23-rc4-mm1]# ls
> Killed
> //above are caused by OOM.
> [root at drpq test-2.6.23-rc4-mm1]# cat /opt/mem_control/group_1/memory.usage
> 32763
> [root at drpq test-2.6.23-rc4-mm1]# cat /opt/mem_control/group_1/memory.limit
> 32768
> // fully filled by page cache. no reclaim run.
> ==
> 
> The reason  this happens is  because I used kernelcore= boot option, i.e
> ZONE_MOVABLE. Seems try_to_free_mem_container_pages() ignores ZONE_MOVABLE.
> 
> Quick fix is attached, but Mel's one-zonelist-pernode patch may change this.
> I'll continue to watch.
> 
> Thanks,
> -Kame
> ==
> Now, there is ZONE_MOVABLE...
> 
> page cache and user pages are allocated from gfp_zone(GFP_HIGHUSER_MOVABLE)
> 
> Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu at jp.fujitsu.com>
> ---
>  mm/vmscan.c |    9 ++-------
>  1 file changed, 2 insertions(+), 7 deletions(-)
> 
> Index: linux-2.6.23-rc4-mm1.bak/mm/vmscan.c
> ===================================================================
> --- linux-2.6.23-rc4-mm1.bak.orig/mm/vmscan.c
> +++ linux-2.6.23-rc4-mm1.bak/mm/vmscan.c
> @@ -1351,12 +1351,6 @@ unsigned long try_to_free_pages(struct z
> 
>  #ifdef CONFIG_CONTAINER_MEM_CONT
> 
> -#ifdef CONFIG_HIGHMEM
> -#define ZONE_USERPAGES ZONE_HIGHMEM
> -#else
> -#define ZONE_USERPAGES ZONE_NORMAL
> -#endif
> -
>  unsigned long try_to_free_mem_container_pages(struct mem_container *mem_cont)
>  {
>  	struct scan_control sc = {
> @@ -1371,9 +1365,10 @@ unsigned long try_to_free_mem_container_
>  	};
>  	int node;
>  	struct zone **zones;
> +	int target_zone = gfp_zone(GFP_HIGHUSER_MOVABLE);
> 
>  	for_each_online_node(node) {
> -		zones = NODE_DATA(node)->node_zonelists[ZONE_USERPAGES].zones;
> +		zones = NODE_DATA(node)->node_zonelists[target_zone].zones;
>  		if (do_try_to_free_pages(zones, sc.gfp_mask, &sc))
>  			return 1;
>  	}

Mel, has sent out a fix (for the single zonelist) that conflicts with
this one. Your fix looks correct to me, but it will be over ridden
by Mel's fix (once those patches are in -mm).

-- 
	Warm Regards,
	Balbir Singh
	Linux Technology Center
	IBM, ISTL