CVE-2026-98730
NoneCVSS
—
EPSS
0.2%
Published
Oct 18, 2026
CISA due date
—
Cloudanix analysis
A flaw in Linux kernel Bluetooth handling for certain Intel hardware lets a local user send too much data at once and potentially take over or crash the whole computer.
- Exploitation
- Local Privilege
- Attacker needs
- Low Privilege
- Complexity
- Moderate
- Impact
- Privilege Gain
- Component
- OS kernel
- Class
- Buffer Overflow
- Remediation
- Patch
- Effort
- Low
- Patch risk
- Low
What it actually does
Buffer overflow in btintel_pcie driver when btintel_pcie_send_frame copies oversized skb data plus 4-byte header into a fixed 4096-byte DMA buffer via memcpy without prior size validation.
Why it matters to the business
Local attackers on affected Linux systems can crash the kernel or potentially execute arbitrary code through Bluetooth hardware access, leading to denial of service or full system compromise on devices using Intel PCIe Bluetooth controllers.
How an attack unfolds
An attacker with local user-space access obtains HCI_CHANNEL_USER privileges, crafts an oversized HCI packet exceeding the DMA buffer limit, and submits it via the Bluetooth stack, triggering the unchecked memcpy that overflows the buffer.
How to fix it
Update the Linux kernel to a version containing the btintel_pcie_send_frame size check rejecting packets larger than BTINTEL_PCIE_BUFFER_SIZE minus the 4-byte header length; rebuild and reboot affected systems.
How to confirm the fix worked
Confirm the patch is present by inspecting btintel_pcie.c for the early size rejection before prepare_tx, or verify the running kernel version includes the fix commit; test by attempting to send an oversized packet and confirming rejection.
Typical attack path
- Obtain local shell or process with HCI_CHANNEL_USER socket access
- Craft malicious oversized TX packet
- Inject packet into btintel_pcie_send_frame path
- Trigger memcpy overflow in btintel_pcie_prepare_tx
- Achieve kernel memory corruption or crash
How to detect it
Review kernel logs for btintel_pcie buffer overflow warnings or oops traces, audit socket creation attempts for HCI_CHANNEL_USER by untrusted processes, and inspect Bluetooth stack traffic for abnormally large TX frames.
- btintel_pcie_send_frame
- btintel_pcie_prepare_tx
- BTINTEL_PCIE_BUFFER_SIZE
- HCI_CHANNEL_USER oversized packet
- memcpy overflow in btintel_pcie
What your security team should do
- Apply kernel updates containing the size validation fix immediately on all systems with Intel PCIe Bluetooth
- Restrict untrusted local access to HCI_CHANNEL_USER sockets via AppArmor or SELinux policies
- Monitor for kernel oops or panic logs related to btintel_pcie
- Audit Bluetooth user-space applications for unnecessary privileged HCI access
The record
In the Linux kernel, the following vulnerability has been resolved: Bluetooth: btintel_pcie: reject oversized TX packets in send_frame() btintel_pcie_send_frame() eventually copies skb->data into a fixed BTINTEL_PCIE_BUFFER_SIZE (4096) DMA buffer via memcpy() in btintel_pcie_prepare_tx(), with a 4-byte PCIe type header prepended. A user-space process with HCI_CHANNEL_USER access can inject an oversized packet and overflow the buffer. Reject packets larger than BTINTEL_PCIE_BUFFER_SIZE - BTINTEL_PCIE_HCI_TYPE_LEN on entry.
Affected packages
| Ecosystem | Package | Introduced | Fixed |
|---|