From: WangJinchao <wangjinchao@xfusion.com>
Date: Mon, 16 Oct 2023 01:15:21 +0000 (+0800)
Subject: padata: Fix refcnt handling in padata_free_shell()
X-Git-Tag: v6.1.68~908
X-Git-Url: http://review.tizen.org/git/?a=commitdiff_plain;h=c7c26d0ef5d20f00dbb2ae3befcabbe0efa77275;p=platform%2Fkernel%2Flinux-starfive.git

padata: Fix refcnt handling in padata_free_shell()

[ Upstream commit 7ddc21e317b360c3444de3023bcc83b85fabae2f ]

In a high-load arm64 environment, the pcrypt_aead01 test in LTP can lead
to system UAF (Use-After-Free) issues. Due to the lengthy analysis of
the pcrypt_aead01 function call, I'll describe the problem scenario
using a simplified model:

Suppose there's a user of padata named `user_function` that adheres to
the padata requirement of calling `padata_free_shell` after `serial()`
has been invoked, as demonstrated in the following code:

```c
struct request {
    struct padata_priv padata;
    struct completion *done;
};

void parallel(struct padata_priv *padata) {
    do_something();
}

void serial(struct padata_priv *padata) {
    struct request *request = container_of(padata,
    				struct request,
				padata);
    complete(request->done);
}

void user_function() {
    DECLARE_COMPLETION(done)
    padata->parallel = parallel;
    padata->serial = serial;
    padata_do_parallel();
    wait_for_completion(&done);
    padata_free_shell();
}
```

In the corresponding padata.c file, there's the following code:

```c
static void padata_serial_worker(struct work_struct *serial_work) {
    ...
    cnt = 0;

    while (!list_empty(&local_list)) {
        ...
        padata->serial(padata);
        cnt++;
    }

    local_bh_enable();

    if (refcount_sub_and_test(cnt, &pd->refcnt))
        padata_free_pd(pd);
}
```

Because of the high system load and the accumulation of unexecuted
softirq at this moment, `local_bh_enable()` in padata takes longer
to execute than usual. Subsequently, when accessing `pd->refcnt`,
`pd` has already been released by `padata_free_shell()`, resulting
in a UAF issue with `pd->refcnt`.

The fix is straightforward: add `refcount_dec_and_test` before calling
`padata_free_pd` in `padata_free_shell`.

Fixes: 07928d9bfc81 ("padata: Remove broken queue flushing")

Signed-off-by: WangJinchao <wangjinchao@xfusion.com>
Acked-by: Daniel Jordan <daniel.m.jordan@oracle.com>
Acked-by: Daniel Jordan <daniel.m.jordan@oracle.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---

diff --git a/kernel/padata.c b/kernel/padata.c
index de90af5..791d9cb 100644
--- a/kernel/padata.c
+++ b/kernel/padata.c
@@ -1094,12 +1094,16 @@ EXPORT_SYMBOL(padata_alloc_shell);
  */
 void padata_free_shell(struct padata_shell *ps)
 {
+	struct parallel_data *pd;
+
 	if (!ps)
 		return;
 
 	mutex_lock(&ps->pinst->lock);
 	list_del(&ps->list);
-	padata_free_pd(rcu_dereference_protected(ps->pd, 1));
+	pd = rcu_dereference_protected(ps->pd, 1);
+	if (refcount_dec_and_test(&pd->refcnt))
+		padata_free_pd(pd);
 	mutex_unlock(&ps->pinst->lock);
 
 	kfree(ps);