2 files changed, 46 insertions, 22 deletions

diff --git a/include/linux/slab.h b/include/linux/slab.h
index 3e3c3ab8ff94..7d66385ae750 100644
--- a/include/linux/slab.h
+++ b/include/linux/slab.h
@@ -62,16 +62,9 @@ extern kmem_cache_t *kmem_cache_create(const char *, size_t, size_t, unsigned lo
 extern int kmem_cache_destroy(kmem_cache_t *);
 extern int kmem_cache_shrink(kmem_cache_t *);
 extern void *kmem_cache_alloc(kmem_cache_t *, unsigned int __nocast);
-#ifdef CONFIG_NUMA
-extern void *kmem_cache_alloc_node(kmem_cache_t *, int);
-#else
-static inline void *kmem_cache_alloc_node(kmem_cache_t *cachep, int node)
-{
-        return kmem_cache_alloc(cachep, GFP_KERNEL);
-}
-#endif
 extern void kmem_cache_free(kmem_cache_t *, void *);
 extern unsigned int kmem_cache_size(kmem_cache_t *);
+extern kmem_cache_t *kmem_find_general_cachep(size_t size, int gfpflags);
 
 /* Size description struct for general caches. */
 struct cache_sizes {
@@ -109,6 +102,20 @@ extern void *kcalloc(size_t, size_t, unsigned int __nocast);
 extern void kfree(const void *);
 extern unsigned int ksize(const void *);
 
+#ifdef CONFIG_NUMA
+extern void *kmem_cache_alloc_node(kmem_cache_t *, int flags, int node);
+extern void *kmalloc_node(size_t size, int flags, int node);
+#else
+static inline void *kmem_cache_alloc_node(kmem_cache_t *cachep, int flags, int node)
+{
+        return kmem_cache_alloc(cachep, flags);
+}
+static inline void *kmalloc_node(size_t size, int flags, int node)
+{
+        return kmalloc(size, flags);
+}
+#endif
+
 extern int FASTCALL(kmem_cache_reap(int));
 extern int FASTCALL(kmem_ptr_validate(kmem_cache_t *cachep, void *ptr));
 
diff --git a/mm/slab.c b/mm/slab.c
index ec660d85ddd7..771cc09f9f1a 100644
--- a/mm/slab.c
+++ b/mm/slab.c
@@ -583,7 +583,7 @@ static inline struct array_cache *ac_data(kmem_cache_t *cachep)
         return cachep->array[smp_processor_id()];
 }
 
-static inline kmem_cache_t *kmem_find_general_cachep(size_t size, int gfpflags)
+static inline kmem_cache_t *__find_general_cachep(size_t size, int gfpflags)
 {
         struct cache_sizes *csizep = malloc_sizes;
 
@@ -607,6 +607,12 @@ static inline kmem_cache_t *kmem_find_general_cachep(size_t size, int gfpflags)
         return csizep->cs_cachep;
 }
 
+kmem_cache_t *kmem_find_general_cachep(size_t size, int gfpflags)
+{
+        return __find_general_cachep(size, gfpflags);
+}
+EXPORT_SYMBOL(kmem_find_general_cachep);
+
 /* Cal the num objs, wastage, and bytes left over for a given slab size. */
 static void cache_estimate(unsigned long gfporder, size_t size, size_t align,
                  int flags, size_t *left_over, unsigned int *num)
@@ -672,14 +678,11 @@ static struct array_cache *alloc_arraycache(int cpu, int entries,
         int memsize = sizeof(void*)*entries+sizeof(struct array_cache);
         struct array_cache *nc = NULL;
 
-        if (cpu != -1) {
-                kmem_cache_t *cachep;
-                cachep = kmem_find_general_cachep(memsize, GFP_KERNEL);
-                if (cachep)
-                        nc = kmem_cache_alloc_node(cachep, cpu_to_node(cpu));
-        }
-        if (!nc)
+        if (cpu == -1)
                 nc = kmalloc(memsize, GFP_KERNEL);
+        else
+                nc = kmalloc_node(memsize, GFP_KERNEL, cpu_to_node(cpu));
+
         if (nc) {
                 nc->avail = 0;
                 nc->limit = entries;
@@ -2361,7 +2364,7 @@ out:
  * and can sleep. And it will allocate memory on the given node, which
  * can improve the performance for cpu bound structures.
  */
-void *kmem_cache_alloc_node(kmem_cache_t *cachep, int nodeid)
+void *kmem_cache_alloc_node(kmem_cache_t *cachep, int flags, int nodeid)
 {
         int loop;
         void *objp;
@@ -2393,7 +2396,7 @@ void *kmem_cache_alloc_node(kmem_cache_t *cachep, int nodeid)
                 spin_unlock_irq(&cachep->spinlock);
 
                 local_irq_disable();
-                if (!cache_grow(cachep, GFP_KERNEL, nodeid)) {
+                if (!cache_grow(cachep, flags, nodeid)) {
                         local_irq_enable();
                         return NULL;
                 }
@@ -2435,6 +2438,16 @@ got_slabp:
 }
 EXPORT_SYMBOL(kmem_cache_alloc_node);
 
+void *kmalloc_node(size_t size, int flags, int node)
+{
+        kmem_cache_t *cachep;
+
+        cachep = kmem_find_general_cachep(size, flags);
+        if (unlikely(cachep == NULL))
+                return NULL;
+        return kmem_cache_alloc_node(cachep, flags, node);
+}
+EXPORT_SYMBOL(kmalloc_node);
 #endif
 
 /**
@@ -2462,7 +2475,12 @@ void *__kmalloc(size_t size, unsigned int __nocast flags)
 {
         kmem_cache_t *cachep;
 
-        cachep = kmem_find_general_cachep(size, flags);
+        /* If you want to save a few bytes .text space: replace
+         * __ with kmem_.
+         * Then kmalloc uses the uninlined functions instead of the inline
+         * functions.
+         */
+        cachep = __find_general_cachep(size, flags);
         if (unlikely(cachep == NULL))
                 return NULL;
         return __cache_alloc(cachep, flags);
@@ -2489,9 +2507,8 @@ void *__alloc_percpu(size_t size, size_t align)
         for (i = 0; i < NR_CPUS; i++) {
                 if (!cpu_possible(i))
                         continue;
-                pdata->ptrs[i] = kmem_cache_alloc_node(
-                                kmem_find_general_cachep(size, GFP_KERNEL),
-                                cpu_to_node(i));
+                pdata->ptrs[i] = kmalloc_node(size, GFP_KERNEL,
+                                                cpu_to_node(i));
 
                 if (!pdata->ptrs[i])
                         goto unwind_oom;