[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20230413143452.211250-6-sxwjean@me.com>
Date: Thu, 13 Apr 2023 22:34:52 +0800
From: sxwjean@...com
To: cl@...ux.com, penberg@...nel.org, rientjes@...gle.com,
iamjoonsoo.kim@....com, akpm@...ux-foundation.org, vbabka@...e.cz,
roman.gushchin@...ux.dev, 42.hyeyoo@...il.com,
keescook@...omium.org
Cc: linux-mm@...ck.org, linux-kernel@...r.kernel.org,
linux-hardening@...r.kernel.org,
Xiongwei Song <xiongwei.song@...driver.com>
Subject: [PATCH 5/5] slub: Don't read nr_slabs and total_objects directly
From: Xiongwei Song <xiongwei.song@...driver.com>
We have node_nr_slabs() to read nr_slabs, node_nr_objs() to read
total_objects in a kmem_cache_node, so no need to access the two
members directly.
Signed-off-by: Xiongwei Song <xiongwei.song@...driver.com>
---
mm/slub.c | 13 ++++++-------
1 file changed, 6 insertions(+), 7 deletions(-)
diff --git a/mm/slub.c b/mm/slub.c
index 33b4fec6bc7a..2c3af399d9d6 100644
--- a/mm/slub.c
+++ b/mm/slub.c
@@ -5157,9 +5157,9 @@ static int validate_slab_node(struct kmem_cache *s,
validate_slab(s, slab, obj_map);
count++;
}
- if (count != atomic_long_read(&n->nr_slabs)) {
+ if (count != node_nr_slabs(n)) {
pr_err("SLUB: %s %ld slabs counted but counter=%ld\n",
- s->name, count, atomic_long_read(&n->nr_slabs));
+ s->name, count, node_nr_slabs(n));
slab_add_kunit_errors();
}
@@ -5443,12 +5443,11 @@ static ssize_t show_slab_objects(struct kmem_cache *s,
for_each_kmem_cache_node(s, node, n) {
if (flags & SO_TOTAL)
- x = atomic_long_read(&n->total_objects);
+ x = node_nr_objs(n);
else if (flags & SO_OBJECTS)
- x = atomic_long_read(&n->total_objects) -
- count_partial(n, count_free);
+ x = node_nr_objs(n) - count_partial(n, count_free);
else
- x = atomic_long_read(&n->nr_slabs);
+ x = node_nr_slabs(n);
total += x;
nodes[node] += x;
}
@@ -6387,7 +6386,7 @@ static int slab_debug_trace_open(struct inode *inode, struct file *filep)
unsigned long flags;
struct slab *slab;
- if (!atomic_long_read(&n->nr_slabs))
+ if (!node_nr_slabs(n))
continue;
spin_lock_irqsave(&n->list_lock, flags);
--
2.30.2
Powered by blists - more mailing lists