Re: [PATCH] Add the pagefault count into memcg stats.

[Ying Han <yinghan@google.com>]

On Tue, Mar 29, 2011 at 8:30 AM, Balbir Singh <balbir@linux.vnet.ibm.com> wrote:
> * Ying Han <yinghan@google.com> [2011-03-26 17:14:44]:
>
>> Two new stats in per-memcg memory.stat which tracks the number of
>> page faults and number of major page faults.
>>
>> "pgfault"
>> "pgmajfault"
>>
>> 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.
>>
>> 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 1069962
>> pgmajfault 553
>> total_pgfault 1069966
>> 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:
>>   Gb  Thr CLine   User     System     Wall    flt/cpu/s fault/wsec
>>   15   16   1     0.67s   232.11s    14.68s   16892.130 267796.518
>>
>> $ ./ministat mmotm.txt mmotm_fault.txt
>> x mmotm.txt (w/o patch)
>> + mmotm_fault.txt (w/ patch)
>> +-------------------------------------------------------------------------+
>>     N           Min           Max        Median           Avg        Stddev
>> x  10     16682.962     17344.027     16913.524     16928.812      166.5362
>> +  10      16696.49      17480.09     16949.143     16951.448     223.56288
>> No difference proven at 95.0% confidence
>>
>> Signed-off-by: Ying Han <yinghan@google.com>
>> ---
>>  Documentation/cgroups/memory.txt |    4 +++
>>  fs/ncpfs/mmap.c                  |    2 +
>>  include/linux/memcontrol.h       |   22 +++++++++++++++
>>  mm/filemap.c                     |    1 +
>>  mm/memcontrol.c                  |   54 ++++++++++++++++++++++++++++++++++++++
>>  mm/memory.c                      |    2 +
>>  mm/shmem.c                       |    1 +
>>  7 files changed, 86 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)
>>  pgpgin               - # of pages paged in (equivalent to # of charging events).
>>  pgpgout              - # of pages paged out (equivalent to # of uncharging events).
>>  swap         - # of bytes of swap usage
>> +pgfault              - # of page faults.
>> +pgmajfault   - # of major page faults.
>>  inactive_anon        - # of bytes of anonymous memory and swap cache memory on
>>               LRU list.
>>  active_anon  - # of bytes of anonymous and swap cache memory on active
>> @@ -406,6 +408,8 @@ total_mapped_file - sum of all children's "cache"
>>  total_pgpgin         - sum of all children's "pgpgin"
>>  total_pgpgout                - sum of all children's "pgpgout"
>>  total_swap           - sum of all children's "swap"
>> +total_pgfault                - sum of all children's "pgfault"
>> +total_pgmajfault     - sum of all children's "pgmajfault"
>>  total_inactive_anon  - sum of all children's "inactive_anon"
>>  total_active_anon    - sum of all children's "active_anon"
>>  total_inactive_file  - sum of all children's "inactive_file"
>> diff --git a/fs/ncpfs/mmap.c b/fs/ncpfs/mmap.c
>> index a7c07b4..adb3f45 100644
>> --- a/fs/ncpfs/mmap.c
>> +++ b/fs/ncpfs/mmap.c
>> @@ -16,6 +16,7 @@
>>  #include <linux/mman.h>
>>  #include <linux/string.h>
>>  #include <linux/fcntl.h>
>> +#include <linux/memcontrol.h>
>>
>>  #include <asm/uaccess.h>
>>  #include <asm/system.h>
>> @@ -92,6 +93,7 @@ static int ncp_file_mmap_fault(struct vm_area_struct *area,
>>        * -- wli
>>        */
>>       count_vm_event(PGMAJFAULT);
>> +     mem_cgroup_pgmajfault_from_mm(area->vm_mm);
>>       return VM_FAULT_MAJOR;
>>  }
>>
>> diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
>> index 5a5ce70..f771fc1 100644
>> --- a/include/linux/memcontrol.h
>> +++ b/include/linux/memcontrol.h
>> @@ -147,6 +147,11 @@ unsigned long mem_cgroup_soft_limit_reclaim(struct zone *zone, int order,
>>                                               gfp_t gfp_mask);
>>  u64 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);
>> +void mem_cgroup_pgfault_from_mm(struct mm_struct *mm);
>> +void mem_cgroup_pgmajfault_from_mm(struct mm_struct *mm);
>> +
>>  #ifdef CONFIG_TRANSPARENT_HUGEPAGE
>>  void mem_cgroup_split_huge_fixup(struct page *head, struct page *tail);
>>  #endif
>> @@ -354,6 +359,23 @@ static inline void mem_cgroup_split_huge_fixup(struct page *head,
>>  {
>>  }
>>
>> +static inline void mem_cgroup_pgfault(struct mem_cgroup *memcg,
>> +                                   int val)
>> +{
>> +}
>> +
>> +static inline void mem_cgroup_pgmajfault(struct mem_cgroup *memcg,
>> +                                      int val)
>> +{
>> +}
>> +
>> +static inline void mem_cgroup_pgfault_from_mm(struct mm_struct *mm)
>> +{
>> +}
>> +
>> +static inline void mem_cgroup_pgmajfault_from_mm(struct mm_struct *mm)
>> +{
>> +}
>>  #endif /* CONFIG_CGROUP_MEM_CONT */
>>
>>  #if !defined(CONFIG_CGROUP_MEM_RES_CTLR) || !defined(CONFIG_DEBUG_VM)
>> diff --git a/mm/filemap.c b/mm/filemap.c
>> index a6cfecf..5dc5401 100644
>> --- a/mm/filemap.c
>> +++ b/mm/filemap.c
>> @@ -1683,6 +1683,7 @@ int filemap_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
>>               /* No page in the page cache at all */
>>               do_sync_mmap_readahead(vma, ra, file, offset);
>>               count_vm_event(PGMAJFAULT);
>> +             mem_cgroup_pgmajfault_from_mm(vma->vm_mm);
>>               ret = VM_FAULT_MAJOR;
>>  retry_find:
>>               page = find_get_page(mapping, offset);
>> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
>> index 4407dd0..63d66f1 100644
>> --- a/mm/memcontrol.c
>> +++ b/mm/memcontrol.c
>> @@ -94,6 +94,8 @@ enum mem_cgroup_events_index {
>>       MEM_CGROUP_EVENTS_PGPGIN,       /* # of pages paged in */
>>       MEM_CGROUP_EVENTS_PGPGOUT,      /* # of pages paged out */
>>       MEM_CGROUP_EVENTS_COUNT,        /* # of pages paged in/out */
>> +     MEM_CGROUP_EVENTS_PGFAULT,      /* # of page-faults */
>> +     MEM_CGROUP_EVENTS_PGMAJFAULT,   /* # of major page-faults */
>>       MEM_CGROUP_EVENTS_NSTATS,
>>  };
>>  /*
>> @@ -585,6 +587,16 @@ static void mem_cgroup_swap_statistics(struct mem_cgroup *mem,
>>       this_cpu_add(mem->stat->count[MEM_CGROUP_STAT_SWAPOUT], val);
>>  }
>>
>> +void mem_cgroup_pgfault(struct mem_cgroup *mem, int val)
>> +{
>> +     this_cpu_add(mem->stat->events[MEM_CGROUP_EVENTS_PGFAULT], val);
>> +}
>> +
>> +void mem_cgroup_pgmajfault(struct mem_cgroup *mem, int val)
>> +{
>> +     this_cpu_add(mem->stat->events[MEM_CGROUP_EVENTS_PGMAJFAULT], val);
>> +}
>> +
>>  static unsigned long mem_cgroup_read_events(struct mem_cgroup *mem,
>>                                           enum mem_cgroup_events_index idx)
>>  {
>> @@ -813,6 +825,40 @@ static inline bool mem_cgroup_is_root(struct mem_cgroup *mem)
>>       return (mem == root_mem_cgroup);
>>  }
>>
>> +void mem_cgroup_pgfault_from_mm(struct mm_struct *mm)
>> +{
>> +     struct mem_cgroup *mem;
>> +
>> +     if (!mm)
>> +             return;
>> +
>> +     rcu_read_lock();
>> +     mem = mem_cgroup_from_task(rcu_dereference(mm->owner));
>> +     if (unlikely(!mem))
>> +             goto out;
>
> A lot of this can be reused, just a minor nitpick. May be you can
> combine this function and the one below

This has been fixed in V3 :)

--Ying
>
>> +     mem_cgroup_pgfault(mem, 1);
>> +
>> +out:
>> +     rcu_read_unlock();
>> +}
>> +
>> +void mem_cgroup_pgmajfault_from_mm(struct mm_struct *mm)
>> +{
>> +     struct mem_cgroup *mem;
>> +
>> +     if (!mm)
>> +             return;
>> +
>> +     rcu_read_lock();
>> +     mem = mem_cgroup_from_task(rcu_dereference(mm->owner));
>> +     if (unlikely(!mem))
>> +             goto out;
>> +     mem_cgroup_pgmajfault(mem, 1);
>> +out:
>> +     rcu_read_unlock();
>> +}
>> +EXPORT_SYMBOL(mem_cgroup_pgmajfault_from_mm);
>> +
>>  /*
>>   * Following LRU functions are allowed to be used without PCG_LOCK.
>>   * Operations are called by routine of global LRU independently from memcg.
>> @@ -3772,6 +3818,8 @@ enum {
>>       MCS_PGPGIN,
>>       MCS_PGPGOUT,
>>       MCS_SWAP,
>> +     MCS_PGFAULT,
>> +     MCS_PGMAJFAULT,
>>       MCS_INACTIVE_ANON,
>>       MCS_ACTIVE_ANON,
>>       MCS_INACTIVE_FILE,
>> @@ -3794,6 +3842,8 @@ struct {
>>       {"pgpgin", "total_pgpgin"},
>>       {"pgpgout", "total_pgpgout"},
>>       {"swap", "total_swap"},
>> +     {"pgfault", "total_pgfault"},
>> +     {"pgmajfault", "total_pgmajfault"},
>>       {"inactive_anon", "total_inactive_anon"},
>>       {"active_anon", "total_active_anon"},
>>       {"inactive_file", "total_inactive_file"},
>> @@ -3822,6 +3872,10 @@ mem_cgroup_get_local_stat(struct mem_cgroup *mem, struct mcs_total_stat *s)
>>               val = mem_cgroup_read_stat(mem, MEM_CGROUP_STAT_SWAPOUT);
>>               s->stat[MCS_SWAP] += val * PAGE_SIZE;
>>       }
>> +     val = mem_cgroup_read_events(mem, MEM_CGROUP_EVENTS_PGFAULT);
>> +     s->stat[MCS_PGFAULT] += val;
>> +     val = mem_cgroup_read_events(mem, MEM_CGROUP_EVENTS_PGMAJFAULT);
>> +     s->stat[MCS_PGMAJFAULT] += val;
>>
>>       /* per zone stat */
>>       val = mem_cgroup_get_local_zonestat(mem, LRU_INACTIVE_ANON);
>> diff --git a/mm/memory.c b/mm/memory.c
>> index 8617d39..0f7ebc9 100644
>> --- a/mm/memory.c
>> +++ b/mm/memory.c
>> @@ -2836,6 +2836,7 @@ static int do_swap_page(struct mm_struct *mm, struct vm_area_struct *vma,
>>               /* Had to read the page from swap area: Major fault */
>>               ret = VM_FAULT_MAJOR;
>>               count_vm_event(PGMAJFAULT);
>> +             mem_cgroup_pgmajfault_from_mm(mm);
>>       } else if (PageHWPoison(page)) {
>>               /*
>>                * hwpoisoned dirty swapcache pages are kept for killing
>> @@ -3375,6 +3376,7 @@ int handle_mm_fault(struct mm_struct *mm, struct vm_area_struct *vma,
>>       __set_current_state(TASK_RUNNING);
>>
>>       count_vm_event(PGFAULT);
>> +     mem_cgroup_pgfault_from_mm(mm);
>>
>>       /* do counter updates before entering really critical section. */
>>       check_sync_rss_stat(current);
>> diff --git a/mm/shmem.c b/mm/shmem.c
>> index ad8346b..5a82674 100644
>> --- a/mm/shmem.c
>> +++ b/mm/shmem.c
>> @@ -1289,6 +1289,7 @@ repeat:
>>                       /* here we actually do the io */
>>                       if (type && !(*type & VM_FAULT_MAJOR)) {
>>                               __count_vm_event(PGMAJFAULT);
>> +                             mem_cgroup_pgmajfault_from_mm(current->mm);
>>                               *type |= VM_FAULT_MAJOR;
>>                       }
>>                       spin_unlock(&info->lock);
>> --
>> 1.7.3.1
>>
>
> --
>        Three Cheers,
>        Balbir
>

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  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=mailto:"dont@kvack.org"> email@kvack.org </a>