The vulnerability stems from the OFPMultipartReply parser's handling of message body length. The code shown in GitHub issue #191 demonstrates a while loop that parses message parts using b.length for offset calculation. When an attacker provides b.length=0, the offset increment becomes zero, making the loop condition (offset < msg_len) never false. This matches the CWE-835 infinite loop pattern described in the advisory. The code location is explicitly referenced in both the CVE description and GitHub issue, with a clear PoC demonstrating the exploit mechanism.