SLUB: include lifetime stats and sets of cpus / nodes in tracking output
authorChristoph Lameter <clameter@sgi.com>
Wed, 9 May 2007 09:32:45 +0000 (02:32 -0700)
committerLinus Torvalds <torvalds@woody.linux-foundation.org>
Wed, 9 May 2007 19:30:46 +0000 (12:30 -0700)
We have information about how long an object existed and about the nodes and
cpus where the allocations and frees took place.  Add that information to the
tracking output in /sys/slab/xx/alloc_calls and /sys/slab/free_calls

This will then enable slabinfo to output nice reports like this:

  christoph@qirst:~/slub$ ./slabinfo kmalloc-128

  Slabcache: kmalloc-128           Aliases:  0 Order :  0

  Sizes (bytes)     Slabs              Debug                Memory
  ------------------------------------------------------------------------
  Object :     128  Total  :      12   Sanity Checks : On   Total:   49152
  SlabObj:     200  Full   :       7   Redzoning     : On   Used :   24832
  SlabSiz:    4096  Partial:       4   Poisoning     : On   Loss :   24320
  Loss   :      72  CpuSlab:       1   Tracking      : On   Lalig:   13968
  Align  :       8  Objects:      20   Tracing       : Off  Lpadd:    1152

  kmalloc-128 has no kmem_cache operations

  kmalloc-128: Kernel object allocation
  -----------------------------------------------------------------------
        6 param_sysfs_setup+0x71/0x130 age=284512/284512/284512 pid=1 nodes=0-1,3
       11 percpu_populate+0x39/0x80 age=283914/284428/284512 pid=1 nodes=0
       21 __register_chrdev_region+0x31/0x170 age=282896/284347/284473 pid=1-1705 nodes=0-2
        1 sys_inotify_init+0x76/0x1c0 age=283423 pid=1004 nodes=0
       19 as_get_io_context+0x32/0xd0 age=6/247567/283988 pid=1-11782 nodes=0,2
       10 ida_pre_get+0x4a/0x80 age=277666/283773/284526 pid=0-2177 nodes=0,2
       24 kobject_kset_add_dir+0x37/0xb0 age=282727/283860/284472 pid=1-1723 nodes=0-2
        1 acpi_ds_build_internal_buffer_obj+0xd3/0x11d age=284508 pid=1 nodes=0
       24 con_insert_unipair+0xd7/0x110 age=284438/284438/284438 pid=1 nodes=0,2
        1 uart_open+0x2d2/0x4b0 age=283896 pid=1 nodes=0
       26 dma_pool_create+0x73/0x1a0 age=282762/282833/282916 pid=1705-1723 nodes=0
        1 neigh_table_init_no_netlink+0xd2/0x210 age=284461 pid=1 nodes=0
        2 neigh_parms_alloc+0x2b/0xe0 age=284410/284411/284412 pid=1 nodes=2
        2 neigh_resolve_output+0x1e1/0x280 age=276289/276291/276293 pid=0-2443 nodes=0
        1 netlink_kernel_create+0x90/0x170 age=284472 pid=1 nodes=0
        4 xt_alloc_table_info+0x39/0xf0 age=283958/283958/283959 pid=1 nodes=1
        3 fn_hash_insert+0x473/0x720 age=277653/277661/277666 pid=2177-2185 nodes=0
        1 get_mtrr_state+0x285/0x2a0 age=284526 pid=0 nodes=0
        1 cacheinfo_cpu_callback+0x26d/0x3e0 age=284458 pid=1 nodes=0
       29 kernel_param_sysfs_setup+0x25/0x90 age=284511/284511/284512 pid=1 nodes=0-1,3
        5 process_zones+0x5e/0x170 age=284546/284546/284546 pid=0 nodes=0
        1 drm_core_init+0x48/0x160 age=284421 pid=1 nodes=2

  kmalloc-128: Kernel object freeing
  ------------------------------------------------------------------------
      163 <not-available> age=4295176847 pid=0 nodes=0-3
        1 __vunmap+0x6e/0xf0 age=282907 pid=1723 nodes=0
       28 free_as_io_context+0x12/0x90 age=9243/262197/283474 pid=42-11754 nodes=0
        1 acpi_get_object_info+0x1b7/0x1d4 age=284475 pid=1 nodes=0
        1 do_acpi_find_child+0x45/0x4e age=284475 pid=1 nodes=0

  NUMA nodes           :    0    1    2    3
  ------------------------------------------
  All slabs                 7    2    2    1
  Partial slabs             2    2    0    0

Signed-off-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
mm/slub.c

index 3e614c1901b7e6696780ce3ba86235ce4c7766d7..ae28310278026938b87f7603067c4397b569eaae 100644 (file)
--- a/mm/slub.c
+++ b/mm/slub.c
@@ -1586,13 +1586,16 @@ static int calculate_order(int size)
                        order < MAX_ORDER; order++) {
                unsigned long slab_size = PAGE_SIZE << order;
 
-               if (slub_max_order > order &&
+               if (order < slub_max_order &&
                                slab_size < slub_min_objects * size)
                        continue;
 
                if (slab_size < size)
                        continue;
 
+               if (order >= slub_max_order)
+                       break;
+
                rem = slab_size % size;
 
                if (rem <= slab_size / 8)
@@ -2751,6 +2754,13 @@ static void resiliency_test(void) {};
 struct location {
        unsigned long count;
        void *addr;
+       long long sum_time;
+       long min_time;
+       long max_time;
+       long min_pid;
+       long max_pid;
+       cpumask_t cpus;
+       nodemask_t nodes;
 };
 
 struct loc_track {
@@ -2791,11 +2801,12 @@ static int alloc_loc_track(struct loc_track *t, unsigned long max)
 }
 
 static int add_location(struct loc_track *t, struct kmem_cache *s,
-                                               void *addr)
+                               const struct track *track)
 {
        long start, end, pos;
        struct location *l;
        void *caddr;
+       unsigned long age = jiffies - track->when;
 
        start = -1;
        end = t->count;
@@ -2811,12 +2822,29 @@ static int add_location(struct loc_track *t, struct kmem_cache *s,
                        break;
 
                caddr = t->loc[pos].addr;
-               if (addr == caddr) {
-                       t->loc[pos].count++;
+               if (track->addr == caddr) {
+
+                       l = &t->loc[pos];
+                       l->count++;
+                       if (track->when) {
+                               l->sum_time += age;
+                               if (age < l->min_time)
+                                       l->min_time = age;
+                               if (age > l->max_time)
+                                       l->max_time = age;
+
+                               if (track->pid < l->min_pid)
+                                       l->min_pid = track->pid;
+                               if (track->pid > l->max_pid)
+                                       l->max_pid = track->pid;
+
+                               cpu_set(track->cpu, l->cpus);
+                       }
+                       node_set(page_to_nid(virt_to_page(track)), l->nodes);
                        return 1;
                }
 
-               if (addr < caddr)
+               if (track->addr < caddr)
                        end = pos;
                else
                        start = pos;
@@ -2834,7 +2862,16 @@ static int add_location(struct loc_track *t, struct kmem_cache *s,
                        (t->count - pos) * sizeof(struct location));
        t->count++;
        l->count = 1;
-       l->addr = addr;
+       l->addr = track->addr;
+       l->sum_time = age;
+       l->min_time = age;
+       l->max_time = age;
+       l->min_pid = track->pid;
+       l->max_pid = track->pid;
+       cpus_clear(l->cpus);
+       cpu_set(track->cpu, l->cpus);
+       nodes_clear(l->nodes);
+       node_set(page_to_nid(virt_to_page(track)), l->nodes);
        return 1;
 }
 
@@ -2850,11 +2887,8 @@ static void process_slab(struct loc_track *t, struct kmem_cache *s,
                set_bit(slab_index(p, s, addr), map);
 
        for_each_object(p, s, addr)
-               if (!test_bit(slab_index(p, s, addr), map)) {
-                       void *addr = get_track(s, p, alloc)->addr;
-
-                       add_location(t, s, addr);
-               }
+               if (!test_bit(slab_index(p, s, addr), map))
+                       add_location(t, s, get_track(s, p, alloc));
 }
 
 static int list_locations(struct kmem_cache *s, char *buf,
@@ -2888,15 +2922,47 @@ static int list_locations(struct kmem_cache *s, char *buf,
        }
 
        for (i = 0; i < t.count; i++) {
-               void *addr = t.loc[i].addr;
+               struct location *l = &t.loc[i];
 
                if (n > PAGE_SIZE - 100)
                        break;
-               n += sprintf(buf + n, "%7ld ", t.loc[i].count);
-               if (addr)
-                       n += sprint_symbol(buf + n, (unsigned long)t.loc[i].addr);
+               n += sprintf(buf + n, "%7ld ", l->count);
+
+               if (l->addr)
+                       n += sprint_symbol(buf + n, (unsigned long)l->addr);
                else
                        n += sprintf(buf + n, "<not-available>");
+
+               if (l->sum_time != l->min_time) {
+                       unsigned long remainder;
+
+                       n += sprintf(buf + n, " age=%ld/%ld/%ld",
+                       l->min_time,
+                       div_long_long_rem(l->sum_time, l->count, &remainder),
+                       l->max_time);
+               } else
+                       n += sprintf(buf + n, " age=%ld",
+                               l->min_time);
+
+               if (l->min_pid != l->max_pid)
+                       n += sprintf(buf + n, " pid=%ld-%ld",
+                               l->min_pid, l->max_pid);
+               else
+                       n += sprintf(buf + n, " pid=%ld",
+                               l->min_pid);
+
+               if (num_online_cpus() > 1 && !cpus_empty(l->cpus)) {
+                       n += sprintf(buf + n, " cpus=");
+                       n += cpulist_scnprintf(buf + n, PAGE_SIZE - n - 50,
+                                       l->cpus);
+               }
+
+               if (num_online_nodes() > 1 && !nodes_empty(l->nodes)) {
+                       n += sprintf(buf + n, " nodes=");
+                       n += nodelist_scnprintf(buf + n, PAGE_SIZE - n - 50,
+                                       l->nodes);
+               }
+
                n += sprintf(buf + n, "\n");
        }