[PATCH net] net: slip: serialize receive against buffer reallocation

[Sungmin Kang <726ksm@gmail.com>]

sl_realloc_bufs() replaces rbuff and updates buffsize while holding
sl->lock. slip_receive_buf() reads those fields and writes through rbuff
without holding the lock.

An MTU change can therefore race with receive processing. An MTU shrink
can expose the new smaller rbuff with the old larger bound, causing an
out-of-bounds write. A receive callback which already loaded the old
rbuff can instead continue writing after that buffer has been freed.

Serialize receive processing with sl_realloc_bufs() by holding sl->lock
while consuming each receive batch.

Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
Cc: stable@vger.kernel.org
Signed-off-by: Sungmin Kang <726ksm@gmail.com>
---
Testing:

Before this change, a generic-KASAN guest reported both a
slab-out-of-bounds write and a slab-use-after-free in slip_receive_buf()
while receive data was raced with MTU changes. A non-KASAN build with
SLUB redzone and poisoning also detected cross-object corruption.

With this change, the same KASAN stress test ran for 180 seconds and
completed 38,137,741 receive-writer iterations and 216,165 MTU changes
without a KASAN report, warning, or panic.

 drivers/net/slip/slip.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/drivers/net/slip/slip.c b/drivers/net/slip/slip.c
index 820e1a8fc..faae711cf 100644
--- a/drivers/net/slip/slip.c
+++ b/drivers/net/slip/slip.c
@@ -693,6 +693,8 @@ static void slip_receive_buf(struct tty_struct *tty, const u8 *cp, const u8 *fp,
 	if (!sl || sl->magic != SLIP_MAGIC || !netif_running(sl->dev))
 		return;
 
+	spin_lock_bh(&sl->lock);
+
 	/* Read the characters out of the buffer */
 	while (count--) {
 		if (fp && *fp++) {
@@ -708,6 +710,8 @@ static void slip_receive_buf(struct tty_struct *tty, const u8 *cp, const u8 *fp,
 #endif
 			slip_unesc(sl, *cp++);
 	}
+
+	spin_unlock_bh(&sl->lock);
 }
 
 /************************************
-- 
2.49.0.windows.1