[PATCH] fix potential stack overflow in mm/slab.c
authorSiddha, Suresh B <suresh.b.siddha@intel.com>
Tue, 26 Sep 2006 06:31:47 +0000 (23:31 -0700)
committerLinus Torvalds <torvalds@g5.osdl.org>
Tue, 26 Sep 2006 15:48:50 +0000 (08:48 -0700)
On High end systems (1024 or so cpus) this can potentially cause stack
overflow. Fix the stack usage.

Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
mm/slab.c

index 2b37a62..619337a 100644 (file)
--- a/mm/slab.c
+++ b/mm/slab.c
@@ -3725,22 +3725,26 @@ static void do_ccupdate_local(void *info)
 static int do_tune_cpucache(struct kmem_cache *cachep, int limit,
                                int batchcount, int shared)
 {
-       struct ccupdate_struct new;
+       struct ccupdate_struct *new;
        int i;
 
-       memset(&new.new, 0, sizeof(new.new));
+       new = kzalloc(sizeof(*new), GFP_KERNEL);
+       if (!new)
+               return -ENOMEM;
+
        for_each_online_cpu(i) {
-               new.new[i] = alloc_arraycache(cpu_to_node(i), limit,
+               new->new[i] = alloc_arraycache(cpu_to_node(i), limit,
                                                batchcount);
-               if (!new.new[i]) {
+               if (!new->new[i]) {
                        for (i--; i >= 0; i--)
-                               kfree(new.new[i]);
+                               kfree(new->new[i]);
+                       kfree(new);
                        return -ENOMEM;
                }
        }
-       new.cachep = cachep;
+       new->cachep = cachep;
 
-       on_each_cpu(do_ccupdate_local, (void *)&new, 1, 1);
+       on_each_cpu(do_ccupdate_local, (void *)new, 1, 1);
 
        check_irq_on();
        cachep->batchcount = batchcount;
@@ -3748,7 +3752,7 @@ static int do_tune_cpucache(struct kmem_cache *cachep, int limit,
        cachep->shared = shared;
 
        for_each_online_cpu(i) {
-               struct array_cache *ccold = new.new[i];
+               struct array_cache *ccold = new->new[i];
                if (!ccold)
                        continue;
                spin_lock_irq(&cachep->nodelists[cpu_to_node(i)]->list_lock);
@@ -3756,7 +3760,7 @@ static int do_tune_cpucache(struct kmem_cache *cachep, int limit,
                spin_unlock_irq(&cachep->nodelists[cpu_to_node(i)]->list_lock);
                kfree(ccold);
        }
-
+       kfree(new);
        return alloc_kmemlist(cachep);
 }
 
@@ -4274,6 +4278,7 @@ static int leaks_show(struct seq_file *m, void *p)
                show_symbol(m, n[2*i+2]);
                seq_putc(m, '\n');
        }
+
        return 0;
 }