On Mon, Mar 28, 2011 at 12:26 PM, Ying Han <yinghan@xxxxxxxxxx> wrote:
> Two new stats in per-memcg memory.stat which tracks the number of
> page faults and number of major page faults.
>
> "pgfault"
> "pgmajfault"
>
> They are different from "pgpgin"/"pgpgout" stat which count number of
> pages charged/discharged to the cgroup and have no meaning of reading/
> writing page to disk.
>
> It is valuable to track the two stats for both measuring application's
> performance as well as the efficiency of the kernel page reclaim path.
> Counting pagefaults per process is useful, but we also need the aggregated
> value since processes are monitored and controlled in cgroup basis in memcg.
>
> Functional test: check the total number of pgfault/pgmajfault of all
> memcgs and compare with global vmstat value:
>
> $ cat /proc/vmstat | grep fault
> pgfault 1070751
> pgmajfault 553
>
> $ cat /dev/cgroup/memory.stat | grep fault
> pgfault 1071138
> pgmajfault 553
> total_pgfault 1071142
> total_pgmajfault 553
>
> $ cat /dev/cgroup/A/memory.stat | grep fault
> pgfault 199
> pgmajfault 0
> total_pgfault 199
> total_pgmajfault 0
>
> Performance test: run page fault test(pft) wit 16 thread on faulting in 15G
> anon pages in 16G container. There is no regression noticed on the "flt/cpu/s"
>
> Sample output from pft:
> TAG pft:anon-sys-default:
> ÂGb ÂThr CLine  User   System   Wall  Âflt/cpu/s fault/wsec
> Â15  16  1   0.69s  230.99s  Â14.62s  16972.539 268876.196
>
> +-------------------------------------------------------------------------+
>  ÂN      Min      Max    ÂMedian      Avg    ÂStddev
> x Â10 Â Â 16682.962 Â Â 17344.027 Â Â 16913.524 Â Â 16928.812 Â Â Â166.5362
> + Â10 Â Â Â16718.92 Â Â 17023.453 Â Â 16907.164 Â Â 16902.399 Â Â 88.468851
> No difference proven at 95.0% confidence
>
> Signed-off-by: Ying Han <yinghan@xxxxxxxxxx>
> ---
> ÂDocumentation/cgroups/memory.txt | Â Â4 +++
> Âfs/ncpfs/mmap.c         Â|  Â2 +
> Âinclude/linux/memcontrol.h    |  18 ++++++++++++++
> Âmm/filemap.c           |  Â1 +
> Âmm/memcontrol.c         Â|  47 ++++++++++++++++++++++++++++++++++++++
> Âmm/memory.c           Â|  Â2 +
> Âmm/shmem.c            |  Â2 +
> Â7 files changed, 76 insertions(+), 0 deletions(-)
>
> diff --git a/Documentation/cgroups/memory.txt b/Documentation/cgroups/memory.txt
> index b6ed61c..2db6103 100644
> --- a/Documentation/cgroups/memory.txt
> +++ b/Documentation/cgroups/memory.txt
> @@ -385,6 +385,8 @@ mapped_file - # of bytes of mapped file (includes tmpfs/shmem)
> Âpgpgin     - # of pages paged in (equivalent to # of charging events).
> Âpgpgout        Â- # of pages paged out (equivalent to # of uncharging events).
> Âswap      - # of bytes of swap usage
> +pgfault        Â- # of page faults.
> +pgmajfault   - # of major page faults.
> Âinactive_anon Â- # of bytes of anonymous memory and swap cache memory on
> Â Â Â Â Â Â Â ÂLRU list.
> Âactive_anon  Â- # of bytes of anonymous and swap cache memory on active
> @@ -406,6 +408,8 @@ total_mapped_file  - sum of all children's "cache"
> Âtotal_pgpgin      - sum of all children's "pgpgin"
> Âtotal_pgpgout     Â- sum of all children's "pgpgout"
> Âtotal_swap       - sum of all children's "swap"
> +total_pgfault     Â- sum of all children's "pgfault"
> +total_pgmajfault    - sum of all children's "pgmajfault"
> Âtotal_inactive_anon  Â- sum of all children's "inactive_anon"
> Âtotal_active_anon   Â- sum of all children's "active_anon"
> Âtotal_inactive_file  Â- sum of all children's "inactive_file"
> diff --git a/fs/ncpfs/mmap.c b/fs/ncpfs/mmap.c
> index a7c07b4..e5d71b2 100644
> --- a/fs/ncpfs/mmap.c
> +++ b/fs/ncpfs/mmap.c
> @@ -16,6 +16,7 @@
> Â#include <linux/mman.h>
> Â#include <linux/string.h>
> Â#include <linux/fcntl.h>
> +#include <linux/memcontrol.h>
>
> Â#include <asm/uaccess.h>
> Â#include <asm/system.h>
> @@ -92,6 +93,7 @@ static int ncp_file_mmap_fault(struct vm_area_struct *area,
> Â Â Â Â * -- wli
> Â Â Â Â */
> Â Â Â Âcount_vm_event(PGMAJFAULT);
> + Â Â Â mem_cgroup_count_vm_event(area->vm_mm, PGMAJFAULT);
> Â Â Â Âreturn VM_FAULT_MAJOR;
> Â}
>
> diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
> index 5a5ce70..45e5268 100644
> --- a/include/linux/memcontrol.h
> +++ b/include/linux/memcontrol.h
> @@ -24,6 +24,7 @@ struct mem_cgroup;
> Âstruct page_cgroup;
> Âstruct page;
> Âstruct mm_struct;
> +enum vm_event_item;
>
> Â/* Stats that can be updated by kernel. */
> Âenum mem_cgroup_page_stat_item {
> @@ -147,6 +148,9 @@ unsigned long mem_cgroup_soft_limit_reclaim(struct zone *zone, int order,
> Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Âgfp_t gfp_mask);
> Âu64 mem_cgroup_get_limit(struct mem_cgroup *mem);
>
> +void mem_cgroup_pgfault(struct mem_cgroup *memcg, int val);
> +void mem_cgroup_pgmajfault(struct mem_cgroup *memcg, int val);
Do we have to expose above two functions?
Isn't it enough to only mem_cgroup_count_vm_event?
> +void mem_cgroup_count_vm_event(struct mm_struct *mm, enum vm_event_item idx);
--
Kind regards,
Minchan Kim
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxxx For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href