io_uring/zcrx: fix overshooting recv limit

It's reported that sometimes a zcrx request can receive more than was
requested. It's caused by io_zcrx_recv_skb() adjusting desc->count for
all received buffers including frag lists, but then doing recursive
calls to process frag list skbs, which leads to desc->count double
accounting and underflow.

Reported-and-tested-by: Matthias Jasny <matthiasjasny@gmail.com>
Fixes: 6699ec9 ("io_uring/zcrx: add a read limit to recvzc requests")
Cc: stable@vger.kernel.org
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

Author: isilence

io_uring/zcrx.c

@@ -1236,12 +1236,16 @@ io_zcrx_recv_skb(read_descriptor_t *desc, struct sk_buff *skb,
 
 		end = start + frag_iter->len;
 		if (offset < end) {
+			size_t count;
+
 			copy = end - offset;
 			if (copy > len)
 				copy = len;
 
 			off = offset - start;
+			count = desc->count;
 			ret = io_zcrx_recv_skb(desc, frag_iter, off, copy);
+			desc->count = count;
 			if (ret < 0)
 				goto out;