ceph: do not access the kiocb after aio requests (d1d96550) · Commits · 戴 / test

In aio case, if the completion comes very fast just before the ceph_read_iter() returns to fs/aio.c, the kiocb will be freed in the completion callback, then if ceph_read_iter() access again we will potentially hit the use-after-free bug. [ jlayton: initialize direct_lock early, and use it everywhere ] URL: https://tracker.ceph.com/issues/45649 Signed-off-by:

Xiubo Li <xiubli@redhat.com> Signed-off-by:

Jeff Layton <jlayton@kernel.org> Signed-off-by:

Ilya Dryomov <idryomov@gmail.com>

fs/ceph/file.c

+3 −2

Original line number	Diff line number	Diff line
		@@ -1538,6 +1538,7 @@ static ssize_t ceph_read_iter(struct kiocb iocb, struct iov_iter to)
		struct inode *inode = file_inode(filp);
		struct ceph_inode_info *ci = ceph_inode(inode);
		struct page *pinned_page = NULL;
		bool direct_lock = iocb->ki_flags & IOCB_DIRECT;
		ssize_t ret;
		int want, got = 0;
		int retry_op = 0, read = 0;
		@@ -1546,7 +1547,7 @@ again:
		dout("aio_read %p %llx.%llx %llu~%u trying to get caps on %p\n",
		inode, ceph_vinop(inode), iocb->ki_pos, (unsigned)len, inode);

		if (iocb->ki_flags & IOCB_DIRECT)
		if (direct_lock)
		ceph_start_io_direct(inode);
		else
		ceph_start_io_read(inode);
		@@ -1603,7 +1604,7 @@ again:
		}
		ceph_put_cap_refs(ci, got);

		if (iocb->ki_flags & IOCB_DIRECT)
		if (direct_lock)
		ceph_end_io_direct(inode);
		else
		ceph_end_io_read(inode);

Admin message