[PATCH] improving oom detection in rmap10c.

From: Ed Tomlinson (tomlins@cam.org)
Date: Sun Jan 06 2002 - 10:49:49 EST


Hi,

This patch should prevent oom situations where the vm does not see pages released
from the slab caches.

It adds a kmem_cache_shrink_nr function which returns the number of pages
released, and makes shrink_icache_memory and friends use the new call.
do_try_to_free_pages was modified so all calls that can release pages contribute
to ret. The changes in slab.c were done avoiding changes to the current interface.

It applies to 2.4.17 + rmap10c + hashed wait queues.

Virgin .17 or .18pre1 could also benifit from a varient of this patch.

Comments?
Ed Tomlinson

------------------
--- linux/fs/inode.c.orig Sat Jan 5 17:35:17 2002
+++ linux/fs/inode.c Sat Jan 5 17:36:06 2002
@@ -725,8 +725,7 @@
         count = inodes_stat.nr_unused / priority;
 
         prune_icache(count);
- kmem_cache_shrink(inode_cachep);
- return 0;
+ return kmem_cache_shrink_nr(inode_cachep);
 }
 
 /*
--- linux/fs/dcache.c.orig Sat Jan 5 17:37:02 2002
+++ linux/fs/dcache.c Sat Jan 5 17:37:57 2002
@@ -568,8 +568,7 @@
         count = dentry_stat.nr_unused / priority;
 
         prune_dcache(count);
- kmem_cache_shrink(dentry_cache);
- return 0;
+ return kmem_cache_shrink_nr(dentry_cache);
 }
 
 #define NAME_ALLOC_LEN(len) ((len+16) & ~15)
--- linux/fs/dquot.c.orig Sat Jan 5 17:38:21 2002
+++ linux/fs/dquot.c Sat Jan 5 17:38:57 2002
@@ -413,8 +413,7 @@
         lock_kernel();
         prune_dqcache(nr_free_dquots / (priority + 1));
         unlock_kernel();
- kmem_cache_shrink(dquot_cachep);
- return 0;
+ return kmem_cache_shrink_nr(dquot_cachep);
 }
 
 /* NOTE: If you change this function please check whether dqput_blocks() works right... */
--- linux/include/linux/slab.h.orig Sat Jan 5 17:27:13 2002
+++ linux/include/linux/slab.h Sat Jan 5 17:27:49 2002
@@ -55,6 +55,7 @@
                                        void (*)(void *, kmem_cache_t *, unsigned long));
 extern int kmem_cache_destroy(kmem_cache_t *);
 extern int kmem_cache_shrink(kmem_cache_t *);
+extern int kmem_cache_shrink_nr(kmem_cache_t *);
 extern void *kmem_cache_alloc(kmem_cache_t *, int);
 extern void kmem_cache_free(kmem_cache_t *, void *);
 
--- linux/mm/slab.c.orig Sat Jan 5 15:06:40 2002
+++ linux/mm/slab.c Sat Jan 5 18:54:40 2002
@@ -911,34 +911,45 @@
 #define drain_cpu_caches(cachep) do { } while (0)
 #endif
 
+/**
+ * Called with the &cachep->spinlock held, returns number of slabs released
+ */
+static int __kmem_cache_shrink_locked(kmem_cache_t *cachep)
+{
+ slab_t *slabp;
+ int ret = 0;
+
+ /* If the cache is growing, stop shrinking. */
+ while (!cachep->growing) {
+ struct list_head *p;
+
+ p = cachep->slabs_free.prev;
+ if (p == &cachep->slabs_free)
+ break;
+
+ slabp = list_entry(cachep->slabs_free.prev, slab_t, list);
+#if DEBUG
+ if (slabp->inuse)
+ BUG();
+#endif
+ list_del(&slabp->list);
+
+ spin_unlock_irq(&cachep->spinlock);
+ kmem_slab_destroy(cachep, slabp);
+ ret++;
+ spin_lock_irq(&cachep->spinlock);
+ }
+ return ret;
+}
+
 static int __kmem_cache_shrink(kmem_cache_t *cachep)
 {
- slab_t *slabp;
         int ret;
 
         drain_cpu_caches(cachep);
 
         spin_lock_irq(&cachep->spinlock);
-
- /* If the cache is growing, stop shrinking. */
- while (!cachep->growing) {
- struct list_head *p;
-
- p = cachep->slabs_free.prev;
- if (p == &cachep->slabs_free)
- break;
-
- slabp = list_entry(cachep->slabs_free.prev, slab_t, list);
-#if DEBUG
- if (slabp->inuse)
- BUG();
-#endif
- list_del(&slabp->list);
-
- spin_unlock_irq(&cachep->spinlock);
- kmem_slab_destroy(cachep, slabp);
- spin_lock_irq(&cachep->spinlock);
- }
+ __kmem_cache_shrink_locked(cachep);
         ret = !list_empty(&cachep->slabs_full) || !list_empty(&cachep->slabs_partial);
         spin_unlock_irq(&cachep->spinlock);
         return ret;
@@ -957,6 +968,24 @@
                 BUG();
 
         return __kmem_cache_shrink(cachep);
+}
+
+/**
+ * kmem_cache_shrink_nr - Shrink a cache returning pages released
+ */
+int kmem_cache_shrink_nr(kmem_cache_t *cachep)
+{
+ int ret;
+
+ if (!cachep || in_interrupt() || !is_chained_kmem_cache(cachep))
+ BUG();
+
+ drain_cpu_caches(cachep);
+
+ spin_lock_irq(&cachep->spinlock);
+ ret = __kmem_cache_shrink_locked(cachep);
+ spin_unlock_irq(&cachep->spinlock);
+ return ret<<(cachep->gfporder);
 }
 
 /**
--- linux/mm/vmscan.c.orig Sat Jan 5 17:39:19 2002
+++ linux/mm/vmscan.c Sat Jan 5 17:43:00 2002
@@ -559,10 +559,10 @@
          * dentry, inode and filesystem quota caches.
          */
         ret += page_launder(gfp_mask);
- shrink_dcache_memory(DEF_PRIORITY, gfp_mask);
- shrink_icache_memory(1, gfp_mask);
+ ret += shrink_dcache_memory(DEF_PRIORITY, gfp_mask);
+ ret += shrink_icache_memory(1, gfp_mask);
 #ifdef CONFIG_QUOTA
- shrink_dqcache_memory(DEF_PRIORITY, gfp_mask);
+ ret += shrink_dqcache_memory(DEF_PRIORITY, gfp_mask);
 #endif
 
         /*
@@ -573,7 +573,7 @@
         /*
          * Reclaim unused slab cache memory.
          */
- kmem_cache_reap(gfp_mask);
+ ret += kmem_cache_reap(gfp_mask);
 
         refill_freelist();
 
-----------------
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Mon Jan 07 2002 - 21:00:31 EST