* [PATCH v2] Bluetooth: hci_bcm4377: fix OOB read from firmware event length in completion ring
@ 2026-04-15 22:24 Tristan Madani
2026-04-15 22:30 ` [v2] " bluez.test.bot
0 siblings, 1 reply; 3+ messages in thread
From: Tristan Madani @ 2026-04-15 22:24 UTC (permalink / raw)
To: Sven Peter; +Cc: Hector Martin, Luiz Augusto von Dentz, linux-bluetooth, asahi
From: Tristan Madani <tristan@talencesecurity.com>
The firmware-controlled entry->len is used as the memcpy size for inline
payload data without bounds checking when the PAYLOAD_MAPPED flag is not
set. This causes out-of-bounds reads from the completion ring DMA memory
for the HCI_D2H and SCO_D2H transfer rings.
Add a length validation against the completion ring payload_size.
Fixes: 8a06127602de ("Bluetooth: hci_bcm4377: Add new driver for BCM4377 PCIe boards")
Signed-off-by: Tristan Madani <tristan@talencesecurity.com>
---
Note: v2 resubmission -- original sent via Gmail had HTML rendering
issues. This version uses git send-email for plain-text formatting.
drivers/bluetooth/hci_bcm4377.c | 8 ++++++++
1 file changed, 8 insertions(+)
diff --git a/drivers/bluetooth/hci_bcm4377.c b/drivers/bluetooth/hci_bcm4377.c
index XXXXXXX..XXXXXXX 100644
--- a/drivers/bluetooth/hci_bcm4377.c
+++ b/drivers/bluetooth/hci_bcm4377.c
@@ -754,6 +754,14 @@ static void bcm4377_handle_completion(struct bcm4377_data *bcm4377,
data_len = le32_to_cpu(entry->len);
msg_id = le16_to_cpu(entry->msg_id);
transfer_ring = le16_to_cpu(entry->ring_id);
+
+ if (data_len > ring->payload_size) {
+ dev_warn(&bcm4377->pdev->dev,
+ "event data len %zu exceeds payload size %zu for completion ring %d\n",
+ data_len, ring->payload_size, ring->ring_id);
+ return;
+ }
+
if ((ring->transfer_rings & BIT(transfer_ring)) == 0) {
dev_warn(
^ permalink raw reply [flat|nested] 3+ messages in thread* [PATCH v2] Bluetooth: hci_bcm4377: fix OOB read from firmware event length in completion ring
@ 2026-04-15 23:02 Tristan Madani
0 siblings, 0 replies; 3+ messages in thread
From: Tristan Madani @ 2026-04-15 23:02 UTC (permalink / raw)
To: sven; +Cc: marcan, luiz.dentz, linux-bluetooth, asahi
The firmware-controlled entry->len is used as the memcpy size for inline
payload data without bounds checking when the PAYLOAD_MAPPED flag is not
set. This causes out-of-bounds reads from the completion ring DMA memory
for the HCI_D2H and SCO_D2H transfer rings.
Add a length validation against the completion ring payload_size.
Fixes: 8a06127602de ("Bluetooth: hci_bcm4377: Add new driver for BCM4377 PCIe boards")
Signed-off-by: Tristan Madani <tristan@talencesecurity.com>
---
v2: Resend with corrected diff context against current bluetooth-next tree.
Fixed format specifiers for u16/size_t types. Original was sent via
Gmail which caused formatting issues and stale context lines.
drivers/bluetooth/hci_bcm4377.c | 7 +++++++
1 file changed, 7 insertions(+)
diff --git a/drivers/bluetooth/hci_bcm4377.c b/drivers/bluetooth/hci_bcm4377.c
index 925d0a635..5129708fd 100644
--- a/drivers/bluetooth/hci_bcm4377.c
+++ b/drivers/bluetooth/hci_bcm4377.c
@@ -755,6 +755,13 @@ static void bcm4377_handle_completion(struct bcm4377_data *bcm4377,
msg_id = le16_to_cpu(entry->msg_id);
transfer_ring = le16_to_cpu(entry->ring_id);
+ if (data_len > ring->payload_size) {
+ dev_warn(&bcm4377->pdev->dev,
+ "completion event data len %u exceeds payload size %u for ring %d\n",
+ (unsigned int)data_len, ring->payload_size, ring->ring_id);
+ return;
+ }
+
if ((ring->transfer_rings & BIT(transfer_ring)) == 0) {
dev_warn(
&bcm4377->pdev->dev,
--
2.47.3
^ permalink raw reply related [flat|nested] 3+ messages in thread
end of thread, other threads:[~2026-04-15 23:02 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-04-15 22:24 [PATCH v2] Bluetooth: hci_bcm4377: fix OOB read from firmware event length in completion ring Tristan Madani
2026-04-15 22:30 ` [v2] " bluez.test.bot
-- strict thread matches above, loose matches on Subject: below --
2026-04-15 23:02 [PATCH v2] " Tristan Madani
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox