DHCP Client Remote Code Execution Vulnerability Demystified


CVE-2019-0547 was the first vulnerability patched by Microsoft this year. The dynamic link library, dhcpcore.dll, which is responsible for DHCP client services in a system, is vulnerable to malicious DHCP reply packets.

This vulnerability allows remote code execution if the user tries to connect to a network with a rogue DHCP Server, hence making it a critical vulnerability.

DHCP protocol overview

DHCP is a client-server protocol used to dynamically assign IP address when a computer connects to a network. DHCP server listens on port 67 and is responsible for distributing IP addresses to DHCP clients and allocating TCP/IP configuration to endpoints.

The DHCP hand shake is represented below:

During DHCP Offer and DHCP Ack, the packet contains all the TCP/IP configuration information required for a client to join the network. The structure of a DHCP Ack packet is shown below:

The options field holds several parameters required for basic DHCP operation. One of the options in the Options field is Domain Search (type field is 119).

Domain Search Option field (RFC 3397)

This option is passed along with OFFER and ACK packets to the client to specify the domain search list used when resolving hostnames using DNS. The format of the DHCP option field is as follows:

To enable the searchlist to be encoded compactly, searchstrings in the searchlist are concatenated and encoded.

A list of domain names, such as  www.example.com and dns.example.com are encoded thus:


There is a vulnerability in the DecodeDomainSearchListData function of dhcpcore.dll.

The DecodeDomainSearchListData function decodes the encoded search list option field value. While decoding, the function calculates the length of the decoded domain name list and allocates memory and copies the decoded list.

A malicious user can create an encoded search list, such that when DecodeDomainSearchListData function decodes, the resulting length is zero. This will lead to heapalloc with zero memory, resulting in an out-of-bound write.


The patch includes a check which ensures the size argument to HeapAlloc is not zero. If zero, the function exits.


A rogue DHCP server in the network can exploit this vulnerability, by replying to the DHCP request from the clients. This rogue DHCP server can also be a wireless access point which a user connects. Successful exploitation of this vulnerability can trigger a code execution in the client and take control of the system.

McAfee NSP customers are protected from this attack by signature “0x42602000”.

The post DHCP Client Remote Code Execution Vulnerability Demystified appeared first on McAfee Blogs.