On Fri, Jul 26, 2019 at 10:12:47AM +0800, Miles Chen wrote:
> This patch is sent to report an use after free in mem_cgroup_iter()
> after merging commit: be2657752e9e "mm: memcg: fix use after free in
> mem_cgroup_iter()".
>
> I work with android kernel tree (4.9 & 4.14), and the commit:
> be2657752e9e "mm: memcg: fix use after free in mem_cgroup_iter()" has
> been merged to the trees. However, I can still observe use after free
> issues addressed in the commit be2657752e9e.
> (on low-end devices, a few times this month)
>
> backtrace:
> css_tryget <- crash here
> mem_cgroup_iter
> shrink_node
> shrink_zones
> do_try_to_free_pages
> try_to_free_pages
> __perform_reclaim
> __alloc_pages_direct_reclaim
> __alloc_pages_slowpath
> __alloc_pages_nodemask
>
> To debug, I poisoned mem_cgroup before freeing it:
>
> static void __mem_cgroup_free(struct mem_cgroup *memcg)
> for_each_node(node)
> free_mem_cgroup_per_node_info(memcg, node);
> free_percpu(memcg->stat);
> + /* poison memcg before freeing it */
> + memset(memcg, 0x78, sizeof(struct mem_cgroup));
> kfree(memcg);
> }
>
> The coredump shows the position=0xdbbc2a00 is freed.
>
> (gdb) p/x ((struct mem_cgroup_per_node *)0xe5009e00)->iter[8]
> $13 = {position = 0xdbbc2a00, generation = 0x2efd}
>
> 0xdbbc2a00: 0xdbbc2e00 0x00000000 0xdbbc2800 0x00000100
> 0xdbbc2a10: 0x00000200 0x78787878 0x00026218 0x00000000
> 0xdbbc2a20: 0xdcad6000 0x00000001 0x78787800 0x00000000
> 0xdbbc2a30: 0x78780000 0x00000000 0x0068fb84 0x78787878
> 0xdbbc2a40: 0x78787878 0x78787878 0x78787878 0xe3fa5cc0
> 0xdbbc2a50: 0x78787878 0x78787878 0x00000000 0x00000000
> 0xdbbc2a60: 0x00000000 0x00000000 0x00000000 0x00000000
> 0xdbbc2a70: 0x00000000 0x00000000 0x00000000 0x00000000
> 0xdbbc2a80: 0x00000000 0x00000000 0x00000000 0x00000000
> 0xdbbc2a90: 0x00000001 0x00000000 0x00000000 0x00100000
> 0xdbbc2aa0: 0x00000001 0xdbbc2ac8 0x00000000 0x00000000
> 0xdbbc2ab0: 0x00000000 0x00000000 0x00000000 0x00000000
> 0xdbbc2ac0: 0x00000000 0x00000000 0xe5b02618 0x00001000
> 0xdbbc2ad0: 0x00000000 0x78787878 0x78787878 0x78787878
> 0xdbbc2ae0: 0x78787878 0x78787878 0x78787878 0x78787878
> 0xdbbc2af0: 0x78787878 0x78787878 0x78787878 0x78787878
> 0xdbbc2b00: 0x78787878 0x78787878 0x78787878 0x78787878
> 0xdbbc2b10: 0x78787878 0x78787878 0x78787878 0x78787878
> 0xdbbc2b20: 0x78787878 0x78787878 0x78787878 0x78787878
> 0xdbbc2b30: 0x78787878 0x78787878 0x78787878 0x78787878
> 0xdbbc2b40: 0x78787878 0x78787878 0x78787878 0x78787878
> 0xdbbc2b50: 0x78787878 0x78787878 0x78787878 0x78787878
> 0xdbbc2b60: 0x78787878 0x78787878 0x78787878 0x78787878
> 0xdbbc2b70: 0x78787878 0x78787878 0x78787878 0x78787878
> 0xdbbc2b80: 0x78787878 0x78787878 0x00000000 0x78787878
> 0xdbbc2b90: 0x78787878 0x78787878 0x78787878 0x78787878
> 0xdbbc2ba0: 0x78787878 0x78787878 0x78787878 0x78787878
>
> In the reclaim path, try_to_free_pages() does not setup
> sc.target_mem_cgroup and sc is passed to do_try_to_free_pages(), ...,
> shrink_node().
>
> In mem_cgroup_iter(), root is set to root_mem_cgroup because
> sc->target_mem_cgroup is NULL.
> It is possible to assign a memcg to root_mem_cgroup.nodeinfo.iter in
> mem_cgroup_iter().
>
> try_to_free_pages
> struct scan_control sc = {...}, target_mem_cgroup is 0x0;
> do_try_to_free_pages
> shrink_zones
> shrink_node
> mem_cgroup *root = sc->target_mem_cgroup;
> memcg = mem_cgroup_iter(root, NULL, &reclaim);
> mem_cgroup_iter()
> if (!root)
> root = root_mem_cgroup;
> ...
>
> css = css_next_descendant_pre(css, &root->css);
> memcg = mem_cgroup_from_css(css);
> cmpxchg(&iter->position, pos, memcg);
>
> My device uses memcg non-hierarchical mode.
> When we release a memcg: invalidate_reclaim_iterators() reaches only
> dead_memcg and its parents. If non-hierarchical mode is used,
> invalidate_reclaim_iterators() never reaches root_mem_cgroup.
>
> static void invalidate_reclaim_iterators(struct mem_cgroup *dead_memcg)
> {
> struct mem_cgroup *memcg = dead_memcg;
>
> for (; memcg; memcg = parent_mem_cgroup(memcg)
> ...
> }
>
> So the use after free scenario looks like:
>
> CPU1 CPU2
>
> try_to_free_pages
> do_try_to_free_pages
> shrink_zones
> shrink_node
> mem_cgroup_iter()
> if (!root)
> root = root_mem_cgroup;
> ...
> css = css_next_descendant_pre(css, &root->css);
> memcg = mem_cgroup_from_css(css);
> cmpxchg(&iter->position, pos, memcg);
>
> invalidate_reclaim_iterators(memcg);
> ...
> __mem_cgroup_free()
> kfree(memcg);
>
> try_to_free_pages
> do_try_to_free_pages
> shrink_zones
> shrink_node
> mem_cgroup_iter()
> if (!root)
> root = root_mem_cgroup;
> ...
> mz = mem_cgroup_nodeinfo(root, reclaim->pgdat->node_id);
> iter = &mz->iter[reclaim->priority];
> pos = READ_ONCE(iter->position);
> css_tryget(&pos->css) <- use after free
>
> To avoid this, we should also invalidate root_mem_cgroup.nodeinfo.iter in
> invalidate_reclaim_iterators().
>
> Change since v1:
> Add a comment to explain why we need to handle root_mem_cgroup separately.
> Rename invalid_root to invalidate_root.
>
> Cc: Johannes Weiner <hannes@xxxxxxxxxxx>
> Signed-off-by: Miles Chen <miles.chen@xxxxxxxxxxxx>
> ---
> mm/memcontrol.c | 38 ++++++++++++++++++++++++++++----------
> 1 file changed, 28 insertions(+), 10 deletions(-)
>
> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> index cdbb7a84cb6e..09f2191f113b 100644
> --- a/mm/memcontrol.c
> +++ b/mm/memcontrol.c
> @@ -1130,26 +1130,44 @@ void mem_cgroup_iter_break(struct mem_cgroup *root,
> css_put(&prev->css);
> }
>
> -static void invalidate_reclaim_iterators(struct mem_cgroup *dead_memcg)
> +static void __invalidate_reclaim_iterators(struct mem_cgroup *from,
> + struct mem_cgroup *dead_memcg)
> {
> - struct mem_cgroup *memcg = dead_memcg;
> struct mem_cgroup_reclaim_iter *iter;
> struct mem_cgroup_per_node *mz;
> int nid;
> int i;
>
> - for (; memcg; memcg = parent_mem_cgroup(memcg)) {
> - for_each_node(nid) {
> - mz = mem_cgroup_nodeinfo(memcg, nid);
> - for (i = 0; i <= DEF_PRIORITY; i++) {
> - iter = &mz->iter[i];
> - cmpxchg(&iter->position,
> - dead_memcg, NULL);
> - }
> + for_each_node(nid) {
> + mz = mem_cgroup_nodeinfo(from, nid);
> + for (i = 0; i <= DEF_PRIORITY; i++) {
> + iter = &mz->iter[i];
> + cmpxchg(&iter->position,
> + dead_memcg, NULL);
> }
> }
> }
>
> +/*
> + * When cgruop1 non-hierarchy mode is used, parent_mem_cgroup() does
> + * not walk all the way up to the cgroup root (root_mem_cgroup). So
> + * we have to handle dead_memcg from cgroup root separately.
> + */
> +static void invalidate_reclaim_iterators(struct mem_cgroup *dead_memcg)
> +{
> + struct mem_cgroup *memcg = dead_memcg;
> + int invalidate_root = 0;
> +
> + for (; memcg; memcg = parent_mem_cgroup(memcg)) {
> + __invalidate_reclaim_iterators(memcg, dead_memcg);
> + if (memcg == root_mem_cgroup)
> + invalidate_root = 1;
> + }
> +
> + if (!invalidate_root)
> + __invalidate_reclaim_iterators(root_mem_cgroup, dead_memcg);
"invalidate_root" suggests we still have to invalidate the root, but
the variable works the opposite way. How about dropping it altogether
and moving the comment directly to where the decision is made:
struct mem_cgroup *memcg = dead_memcg;
do {
__invalidate_reclaim_iterators(memcg, dead_memcg);
last = memcg;
} while ((memcg = parent_mem_cgroup(memcg)));
/*
* When cgruop1 non-hierarchy mode is used,
* parent_mem_cgroup() does not walk all the way up to the
* cgroup root (root_mem_cgroup). So we have to handle
* dead_memcg from cgroup root separately.
*/
if (last != root_mem_cgroup)
__invalidate_reclaim_iterators(root_mem_cgroup, dead_memcg);
