[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20160122230823.GI17997@ZenIV.linux.org.uk>
Date: Fri, 22 Jan 2016 23:08:23 +0000
From: Al Viro <viro@...IV.linux.org.uk>
To: Dmitry Vyukov <dvyukov@...gle.com>
Cc: "linux-fsdevel@...r.kernel.org" <linux-fsdevel@...r.kernel.org>,
LKML <linux-kernel@...r.kernel.org>,
Hugh Dickins <hughd@...gle.com>,
"linux-mm@...ck.org" <linux-mm@...ck.org>,
syzkaller <syzkaller@...glegroups.com>,
Kostya Serebryany <kcc@...gle.com>,
Alexander Potapenko <glider@...gle.com>,
Sasha Levin <sasha.levin@...cle.com>,
Eric Dumazet <edumazet@...gle.com>
Subject: Re: fs: use-after-free in link_path_walk
On Fri, Jan 22, 2016 at 11:33:09PM +0100, Dmitry Vyukov wrote:
> Hello,
>
> The following program triggers a use-after-free in link_path_walk:
> https://gist.githubusercontent.com/dvyukov/fc0da4b914d607ba8129/raw/b761243c44106d74f2173745132c82d179cbdc58/gistfile1.txt
Hmm... Actually, I wonder if that had been triggerable since May. What
happens is that unlike struct inode itself, shmem info->symlink is
freed immediately, without an RCU delay. Easy to fix, fortunately...
Could you check if the patch below fixes that for you?
diff --git a/include/linux/shmem_fs.h b/include/linux/shmem_fs.h
index a43f41c..4d4780c 100644
--- a/include/linux/shmem_fs.h
+++ b/include/linux/shmem_fs.h
@@ -15,10 +15,7 @@ struct shmem_inode_info {
unsigned int seals; /* shmem seals */
unsigned long flags;
unsigned long alloced; /* data pages alloced to file */
- union {
- unsigned long swapped; /* subtotal assigned to swap */
- char *symlink; /* unswappable short symlink */
- };
+ unsigned long swapped; /* subtotal assigned to swap */
struct shared_policy policy; /* NUMA memory alloc policy */
struct list_head swaplist; /* chain of maybes on swap */
struct simple_xattrs xattrs; /* list of xattrs */
diff --git a/mm/shmem.c b/mm/shmem.c
index 38c5e72..440e2a7 100644
--- a/mm/shmem.c
+++ b/mm/shmem.c
@@ -701,8 +701,7 @@ static void shmem_evict_inode(struct inode *inode)
list_del_init(&info->swaplist);
mutex_unlock(&shmem_swaplist_mutex);
}
- } else
- kfree(info->symlink);
+ }
simple_xattrs_free(&info->xattrs);
WARN_ON(inode->i_blocks);
@@ -2549,13 +2548,12 @@ static int shmem_symlink(struct inode *dir, struct dentry *dentry, const char *s
info = SHMEM_I(inode);
inode->i_size = len-1;
if (len <= SHORT_SYMLINK_LEN) {
- info->symlink = kmemdup(symname, len, GFP_KERNEL);
- if (!info->symlink) {
+ inode->i_link = kmemdup(symname, len, GFP_KERNEL);
+ if (!inode->i_link) {
iput(inode);
return -ENOMEM;
}
inode->i_op = &shmem_short_symlink_operations;
- inode->i_link = info->symlink;
} else {
inode_nohighmem(inode);
error = shmem_getpage(inode, 0, &page, SGP_WRITE, NULL);
@@ -3132,6 +3130,7 @@ static struct inode *shmem_alloc_inode(struct super_block *sb)
static void shmem_destroy_callback(struct rcu_head *head)
{
struct inode *inode = container_of(head, struct inode, i_rcu);
+ kfree(inode->i_link);
kmem_cache_free(shmem_inode_cachep, SHMEM_I(inode));
}
Powered by blists - more mailing lists