From: Pengpeng Hou <pengpeng@iscas.ac.cn>
To: "Toke Høiland-Jørgensen" <toke@toke.dk>
Cc: linux-wireless@vger.kernel.org, linux-kernel@vger.kernel.org,
Pengpeng Hou <pengpeng@iscas.ac.cn>
Subject: [PATCH v2] wifi: ath9k: validate USB RX stream lengths before copying
Date: Fri, 14 Aug 2026 15:57:16 +0800 [thread overview]
Message-ID: <20260814075716.18804-1-pengpeng@iscas.ac.cn> (raw)
ath9k_hif_usb_rx_stream() parses records from one USB receive buffer
and also completes records split across two buffers. It checks the stream
tag and a packet length ceiling, but not every fixed header, current
payload extent or continuation extent before reading or copying.
Validate the fixed header and complete-record extent, bound a split
record to the capacity of two receive buffers, and require the next
buffer to contain the expected continuation. Clear all saved continuation
state when a malformed record is dropped.
Fixes: fb9987d0f748 ("ath9k_htc: Support for AR9271 chipset.")
Assisted-by: Codex:gpt-5
Signed-off-by: Pengpeng Hou <pengpeng@iscas.ac.cn>
---
Changes since v1: https://lore.kernel.org/all/20260701053823.49801-1-pengpeng@iscas.ac.cn/
- validate fixed headers and complete records before copying
- prove the split-record path against two receive-buffer capacities
- clear every saved continuation field after success or rejection
Every RX parser branch was reviewed statically; malformed USB traffic was
not injected on ath9k hardware.
drivers/net/wireless/ath/ath9k/hif_usb.c | 33 ++++++++++++++++++++++--
1 file changed, 31 insertions(+), 2 deletions(-)
diff --git a/drivers/net/wireless/ath/ath9k/hif_usb.c b/drivers/net/wireless/ath/ath9k/hif_usb.c
index 47f904e7e652..8d60779165ab 100644
--- a/drivers/net/wireless/ath/ath9k/hif_usb.c
+++ b/drivers/net/wireless/ath/ath9k/hif_usb.c
@@ -566,6 +566,18 @@ static void ath9k_hif_usb_rx_stream(struct hif_device_usb *hif_dev,
if (rx_remain_len != 0) {
struct sk_buff *remain_skb = hif_dev->remain_skb;
+ if (rx_remain_len < hif_dev->rx_pad_len ||
+ rx_remain_len > len) {
+ dev_kfree_skb_any(remain_skb);
+ hif_dev->remain_skb = NULL;
+ hif_dev->rx_remain_len = 0;
+ hif_dev->rx_transfer_len = 0;
+ hif_dev->rx_pad_len = 0;
+ RX_STAT_INC(hif_dev, skb_dropped);
+ spin_unlock(&hif_dev->rx_lock);
+ return;
+ }
+
if (remain_skb) {
ptr = (u8 *) remain_skb->data;
@@ -579,11 +591,13 @@ static void ath9k_hif_usb_rx_stream(struct hif_device_usb *hif_dev,
skb_put(remain_skb, rx_pkt_len);
skb_pool[pool_index++] = remain_skb;
- hif_dev->remain_skb = NULL;
- hif_dev->rx_remain_len = 0;
} else {
index = rx_remain_len;
}
+ hif_dev->remain_skb = NULL;
+ hif_dev->rx_remain_len = 0;
+ hif_dev->rx_transfer_len = 0;
+ hif_dev->rx_pad_len = 0;
}
spin_unlock(&hif_dev->rx_lock);
@@ -596,6 +610,11 @@ static void ath9k_hif_usb_rx_stream(struct hif_device_usb *hif_dev,
ptr = (u8 *) skb->data;
+ if (len - index < 4) {
+ RX_STAT_INC(hif_dev, skb_dropped);
+ goto invalid_pkt;
+ }
+
pkt_len = get_unaligned_le16(ptr + index);
pkt_tag = get_unaligned_le16(ptr + index + 2);
@@ -624,6 +643,11 @@ static void ath9k_hif_usb_rx_stream(struct hif_device_usb *hif_dev,
index = index + 4 + pkt_len + pad_len;
if (index > MAX_RX_BUF_SIZE) {
+ if (len < MAX_RX_BUF_SIZE || index > 2 * MAX_RX_BUF_SIZE) {
+ RX_STAT_INC(hif_dev, skb_dropped);
+ goto invalid_pkt;
+ }
+
spin_lock(&hif_dev->rx_lock);
nskb = __dev_alloc_skb(pkt_len + 32, GFP_ATOMIC);
if (!nskb) {
@@ -648,6 +672,11 @@ static void ath9k_hif_usb_rx_stream(struct hif_device_usb *hif_dev,
hif_dev->remain_skb = nskb;
spin_unlock(&hif_dev->rx_lock);
} else {
+ if (index > len) {
+ RX_STAT_INC(hif_dev, skb_dropped);
+ goto invalid_pkt;
+ }
+
if (pool_index == MAX_PKT_NUM_IN_TRANSFER) {
dev_err(&hif_dev->udev->dev,
"ath9k_htc: over RX MAX_PKT_NUM\n");
--
2.50.1 (Apple Git-155)
reply other threads:[~2026-08-14 7:57 UTC|newest]
Thread overview: [no followups] expand[flat|nested] mbox.gz Atom feed
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20260814075716.18804-1-pengpeng@iscas.ac.cn \
--to=pengpeng@iscas.ac.cn \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-wireless@vger.kernel.org \
--cc=toke@toke.dk \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox