Vulnerabilities > CVE-2017-0568 - Unspecified vulnerability in Linux Kernel 3.10/3.18
Attack vector
LOCAL Attack complexity
HIGH Privileges required
NONE Confidentiality impact
HIGH Integrity impact
HIGH Availability impact
HIGH Summary
An elevation of privilege vulnerability in the Broadcom Wi-Fi driver could enable a local malicious application to execute arbitrary code within the context of the kernel. This issue is rated as High because it first requires compromising a privileged process. Product: Android. Versions: Kernel-3.10, Kernel-3.18. Android ID: A-34197514. References: B-RB#112600.
Vulnerable Configurations
Part | Description | Count |
---|---|---|
OS | 2 |
Seebug
bulletinFamily | exploit |
description | Broadcom produces Wi-Fi HardMAC SoCs which are used to handle the PHY and MAC layer processing. These chips are present in both mobile devices and Wi-Fi routers, and are capable of handling many Wi-Fi related events without delegating to the host OS. On Android devices, the "bcmdhd" driver is used in order to communicate with the Wi-Fi SoC (also referred to as "dongle"). Along with the regular flow of frames transferred between the host and the dongle, the two communicate with one another via a set of "ioctls" which can be issued to read or write dongle configuration from the host. This information is exchanged using the SDIO "control" channel (`SDPCM_CONTROL_CHANNEL`) rather than the regular "data" and "glom" channels (which are used to transfer frames). When the "bcmdhd" driver performs a network scan, it does so by calling "`wl_run_escan`". In configurations where P2P scan is enabled (non-legacy configurations), the function first fetches the list of allowed channels in the current regulatory domain. The channels are fetched by issuing the `WLC_GET_VALID_CHANNELS` ioctl to the dongle. This ioctl's results are interpreted as an array of 16-bit values representing channels and a single 32-bit bit value denoting the length of the channel array. Here is a short snippet from the code handling the P2P scan's logic: ``` 1. u16* default_chan_list = kzalloc(num_chans * sizeof(*default_chan_list), GFP_KERNEL); 2. u32 n_nodfs = 0 3. ... 4. if (!wl_get_valid_channels(ndev, chan_buf, sizeof(chan_buf))) { 5. list = (wl_uint32_list_t *) chan_buf; 6. n_valid_chan = dtoh32(list->count); 7. for (i = 0; i < num_chans; i++) 8. { 9. _freq = request->channels[i]->center_freq; 10. channel = ieee80211_frequency_to_channel(_freq); 11. ... 12. for (j = 0; j < n_valid_chan; j++) { 13. if (channel == (dtoh32(list->element[j]))) 14. default_chan_list[n_nodfs++] = channel; 15. } 16. } 17. } ``` Where "`wl_get_valid_channels`" is a simple wrapper around the `WLC_GET_VALID_CHANNELS` ioctl. An attacker controlling the dongle can re-write the ioctl handling function (since it is entirely RAM-resident), in order to control the results of the ioctl above. This would allow the attacker to return an arbitrarily large value for "`list->count`". Doing so would cause the internal loop (lines 12-15) to iterate many times, and each time a value matching "channel" is encountered, line 14 would be executed. Since there is no validation to make sure that "`n_nodfs`" does not exceed the size of "`num_chans`", this would cause the loop to overflow the allocated chunk for "`default_chan_list`", corrupting the kernel's memory. I've been able to statically verify this issue on the "bcmdhd-3.10" driver, and in the corresponding "bcmdhd" driver on the Nexus 5 (hammerhead) and Nexus 6P's (angler) kernels. |
id | SSV:92901 |
last seen | 2017-11-19 |
modified | 2017-04-05 |
published | 2017-04-05 |
reporter | Root |
title | Broadcom: Heap overflow in "wl_run_escan" when handling WLC_GET_VALID_CHANNELS ioctl results(CVE-2017-0568) |