Received-SPF: pass (google.com: domain of linux-kernel+bounces-177696-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.48.161 as permitted sender) client-ip=147.75.48.161;
Date: Mon, 13 May 2024 07:25:56 -0700
From: Jakub Kicinski <kuba@kernel.org>
To: Ryosuke Yasuoka <ryasuoka@redhat.com>
Cc: krzk@kernel.org, davem@davemloft.net, edumazet@google.com,
 pabeni@redhat.com, horms@kernel.org, netdev@vger.kernel.org,
 linux-kernel@vger.kernel.org, syoshida@redhat.com,
 syzbot+d7b4dc6cd50410152534@syzkaller.appspotmail.com
Subject: Re: [PATCH net v4] nfc: nci: Fix uninit-value in nci_rx_work
Message-ID: <20240513072556.278df9f4@kernel.org>
In-Reply-To: <Zj9QRIjGLbVdd7MX@zeus>
References: <20240509113036.362290-1-ryasuoka@redhat.com>
	<20240510190613.72838bf0@kernel.org>
	<Zj9QRIjGLbVdd7MX@zeus>
Precedence: bulk
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit

On Sat, 11 May 2024 20:02:28 +0900 Ryosuke Yasuoka wrote:
> Sorry I don't get it. Do you mean I just insert
> BUILD_BUG_ON(NCI_CTRL_HDR_SIZE != NCI_DATA_HDR_SIZE) or insert this and
> clean up the code duplication like this? (It is just a draft. I just
> share what I mean.) I can avoid to call nci_valid_size() repeatedly
> inside the switch statement.
> 
> static void nci_rx_work(struct work_struct *work)
> {
> ...
> 		if (!skb->len) {
> 			kfree_skb(skb);
> 			kcov_remote_stop();
> 			break;
> 		}
> 
> 		BUILD_BUG_ON(NCI_CTRL_HDR_SIZE != NCI_DATA_HDR_SIZE);
> 		unsigned int hdr_size = NCI_CTRL_HDR_SIZE;
> 
> 		if (!nci_valid_size(skb, hdr_size)) {
> 			kfree_skb(skb);
> 			continue;
> 		}
> 
> 		/* Process frame */
> 		switch (nci_mt(skb->data)) {
> 		case NCI_MT_RSP_PKT:
> 			nci_rsp_packet(ndev, skb);
> 			break;
> 
> 		case NCI_MT_NTF_PKT:
> 			nci_ntf_packet(ndev, skb);
> 			break;

Yes, that's what I meant. I'd probably merge the nci_valid_size()
check with the !skb->len check, too.