From: Maoyi Xie <maoyixie.tju@gmail.com>
To: Arend van Spriel <arend.vanspriel@broadcom.com>
Cc: linux-wireless@vger.kernel.org, brcm80211@lists.linux.dev,
brcm80211-dev-list.pdl@broadcom.com,
linux-kernel@vger.kernel.org, stable@vger.kernel.org
Subject: [PATCH wireless 1/2] wifi: brcmfmac: fwsignal: bound the AMPDU reorder indices
Date: Mon, 16 Nov 2026 02:58:42 +0800 [thread overview]
Message-ID: <20261115185843.64633-2-maoyixie.tju@gmail.com> (raw)
In-Reply-To: <20261115185843.64633-1-maoyixie.tju@gmail.com>
brcmf_fws_rxreorder() allocates max_idx + 1 slots when a reorder flow
starts. It then uses the indices of each frame without a check.
rfi->pktslots[rfi->cur_idx] = pkt;
A cur_idx above the allocated maximum makes the driver write and read
past pktslots[]. An exp_idx above max_idx makes
brcmf_rxreorder_get_skb_list() read past the array or loop forever.
Store the allocated maximum in the flow. Bound max_idx by it and the
other indices by max_idx. Deliver a frame that fails the check without
reordering.
This was found with a static analysis tool that we develop. A userspace
harness that compiles the fwsignal.c functions unchanged runs under
AddressSanitizer. It shows the out-of-bounds accesses and the endless
loop before this patch, but not after it. No hardware was available for
testing.
Fixes: 87edd8916ee2 ("brcmfmac: add AMPDU reordering functionality")
Cc: stable@vger.kernel.org
Signed-off-by: Maoyi Xie <maoyixie.tju@gmail.com>
---
.../broadcom/brcm80211/brcmfmac/core.h | 2 ++
.../broadcom/brcm80211/brcmfmac/fwsignal.c | 24 ++++++++++++++++---
2 files changed, 23 insertions(+), 3 deletions(-)
diff --git a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/core.h b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/core.h
index 399b6810e394de..19cecb987a87bc 100644
--- a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/core.h
+++ b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/core.h
@@ -52,6 +52,7 @@
* @cur_idx: last AMPDU index from firmware.
* @exp_idx: expected next AMPDU index.
* @max_idx: maximum amount of packets per AMPDU.
+ * @alloc_max_idx: highest index of @pktslots, set when the flow is created.
* @pend_pkts: number of packets currently in @pktslots.
* @pktslots: array for ordering AMPDU packets.
*/
@@ -60,6 +61,7 @@ struct brcmf_ampdu_rx_reorder {
u8 cur_idx;
u8 exp_idx;
u8 max_idx;
+ u8 alloc_max_idx;
u8 pend_pkts;
struct sk_buff *pktslots[];
};
diff --git a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/fwsignal.c b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/fwsignal.c
index a43f1a38b0e30e..e81105154d6543 100644
--- a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/fwsignal.c
+++ b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/fwsignal.c
@@ -1721,8 +1721,16 @@ void brcmf_fws_rxreorder(struct brcmf_if *ifp, struct sk_buff *pkt)
ifp->drvr->reorder_flows[flow_id] = rfi;
rfi->max_idx = max_idx;
+ rfi->alloc_max_idx = max_idx;
}
if (flags & BRCMF_RXREORDER_NEW_HOLE) {
+ cur_idx = reorder_data[BRCMF_RXREORDER_CURIDX_OFFSET];
+ exp_idx = reorder_data[BRCMF_RXREORDER_EXPIDX_OFFSET];
+ max_idx = reorder_data[BRCMF_RXREORDER_MAXIDX_OFFSET];
+ if (max_idx > rfi->alloc_max_idx || cur_idx > max_idx ||
+ exp_idx > max_idx)
+ goto bad_idx;
+
if (rfi->pend_pkts) {
brcmf_rxreorder_get_skb_list(rfi, rfi->exp_idx,
rfi->exp_idx,
@@ -1731,9 +1739,9 @@ void brcmf_fws_rxreorder(struct brcmf_if *ifp, struct sk_buff *pkt)
} else {
__skb_queue_head_init(&reorder_list);
}
- rfi->cur_idx = reorder_data[BRCMF_RXREORDER_CURIDX_OFFSET];
- rfi->exp_idx = reorder_data[BRCMF_RXREORDER_EXPIDX_OFFSET];
- rfi->max_idx = reorder_data[BRCMF_RXREORDER_MAXIDX_OFFSET];
+ rfi->cur_idx = cur_idx;
+ rfi->exp_idx = exp_idx;
+ rfi->max_idx = max_idx;
rfi->pktslots[rfi->cur_idx] = pkt;
rfi->pend_pkts++;
brcmf_dbg(DATA, "flow-%d: new hole %d (%d), pending %d\n",
@@ -1741,6 +1749,8 @@ void brcmf_fws_rxreorder(struct brcmf_if *ifp, struct sk_buff *pkt)
} else if (flags & BRCMF_RXREORDER_CURIDX_VALID) {
cur_idx = reorder_data[BRCMF_RXREORDER_CURIDX_OFFSET];
exp_idx = reorder_data[BRCMF_RXREORDER_EXPIDX_OFFSET];
+ if (cur_idx > rfi->max_idx || exp_idx > rfi->max_idx)
+ goto bad_idx;
if ((exp_idx == rfi->exp_idx) && (cur_idx != rfi->exp_idx)) {
/* still in the current hole */
@@ -1811,6 +1821,8 @@ void brcmf_fws_rxreorder(struct brcmf_if *ifp, struct sk_buff *pkt)
} else {
/* explicitly window move updating the expected index */
exp_idx = reorder_data[BRCMF_RXREORDER_EXPIDX_OFFSET];
+ if (exp_idx > rfi->max_idx)
+ goto bad_idx;
brcmf_dbg(DATA, "flow-%d (0x%x): change expected: %d -> %d\n",
flow_id, flags, rfi->exp_idx, exp_idx);
@@ -1830,6 +1842,12 @@ void brcmf_fws_rxreorder(struct brcmf_if *ifp, struct sk_buff *pkt)
__skb_unlink(pkt, &reorder_list);
brcmf_netif_rx(ifp, pkt);
}
+ return;
+
+bad_idx:
+ bphy_err(drvr, "flow-%d: invalid reorder index, ignore this packet\n",
+ flow_id);
+ brcmf_netif_rx(ifp, pkt);
}
void brcmf_fws_hdrpull(struct brcmf_if *ifp, s16 siglen, struct sk_buff *skb)
--
2.34.1
next prev parent reply other threads:[~2026-11-15 18:58 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-11-15 18:58 [PATCH wireless 0/2] wifi: brcmfmac: fwsignal: bound indices from firmware signals Maoyi Xie
2026-11-15 18:58 ` Maoyi Xie [this message]
2026-11-15 18:58 ` [PATCH wireless 2/2] wifi: brcmfmac: fwsignal: bound the MACDESC_ADD interface index Maoyi Xie
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=20261115185843.64633-2-maoyixie.tju@gmail.com \
--to=maoyixie.tju@gmail.com \
--cc=arend.vanspriel@broadcom.com \
--cc=brcm80211-dev-list.pdl@broadcom.com \
--cc=brcm80211@lists.linux.dev \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-wireless@vger.kernel.org \
--cc=stable@vger.kernel.org \
/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