CVE-2026-107576 in hMailServer
Summary
by MITRE • 10/08/2026
Inefficient algorithmic complexity in the inbound DKIM and ARC signature verification of Progressive Robot hMailServer 6.0.0 through 6.3.5 allows a remote unauthenticated attacker to make the mail services unavailable by sending a message. Building the canonical header and choosing the header fields named in a signature's h= tag took time growing with the square of the message's header: the 'simple' canonicalisation prepended each continuation line of a folded field to the lines already gathered, and both canonicalisations searched the gathered fields from the bottom for each h= name and erased the match from the middle of the list. A message whose header holds very many fields, or a field folded over very many lines, with a DKIM-Signature the attacker signs for a domain they control, keeps a worker thread busy for tens of seconds per signature; up to ten signatures are evaluated per message by each of the DKIM and DMARC tests, on the threads that serve delivery and SMTP.
If you want to get best quality of vulnerability data, you may have to visit VulDB.
Analysis
by VulDB Data Team • 10/08/2026
The vulnerability identified in Progressive Robot hMailServer versions 6.0.0 through 6.3.5 represents a critical denial-of-service risk stemming from inefficient algorithmic complexity within the inbound DomainKeys Identified Mail and Authenticated Received Chain signature verification processes. This flaw allows remote unauthenticated attackers to exhaust server resources by crafting maliciously structured email messages that trigger excessive computational overhead during validation checks. The core technical issue lies in how the mail server handles header canonicalization, a necessary step for verifying cryptographic signatures against RFC standards. Specifically, the implementation of both simple and relaxed canonicalization methods exhibits quadratic time complexity relative to the size of the message headers rather than linear or logarithmic growth expected in optimized implementations.
The operational mechanism of this flaw involves two primary inefficiencies during the processing of DKIM-Signature fields containing an h= tag which lists specific header names for verification. First, when constructing canonicalized headers from folded lines, the algorithm prepends each continuation line to the previously gathered lines rather than appending them or using a more efficient data structure like a linked list with proper indexing. This operation requires shifting existing memory blocks repeatedly as new content is added, leading to significant performance degradation as header size increases. Second, for every header name specified in the h= tag of a DKIM signature, the system searches through the already gathered fields from the bottom up and then removes the matched field from the middle of the list. Removing an element from the middle of an array-based structure necessitates shifting all subsequent elements to fill the gap, further compounding the computational cost with each verification step.
The impact of this vulnerability is severe for mail server operators relying on hMailServer for inbound email processing. An attacker can craft a message containing numerous header fields or headers folded across many lines and sign it using a domain they control. When such a message arrives, the DKIM and ARC verification modules are invoked to validate the signature. Because up to ten signatures may be evaluated per message during combined DKIM and DMARC testing phases, the cumulative effect of these inefficient operations can keep worker threads busy for tens of seconds each. Since hMailServer uses dedicated threads for serving SMTP connections and handling mail delivery, tying up multiple workers with computationally expensive tasks effectively blocks legitimate email traffic. This results in a service availability issue where new messages cannot be processed or delivered until the malicious verification processes complete, potentially causing timeouts and connection drops for other users.
From a classification perspective, this vulnerability aligns with CWE-400, which describes uncontrolled resource consumption due to inefficient algorithms. It also maps to MITRE ATT&CK technique T1496, Resource Hijacking, specifically under the sub-technique of Denial of Service via computational exhaustion. The attack vector is remote and requires no authentication, making it particularly dangerous for publicly accessible mail servers that accept inbound connections from any source on port 25 or submission ports.
Mitigation strategies should focus primarily on upgrading to a patched version of hMailServer where the canonicalization logic has been optimized to use efficient data structures such as linked lists or hash maps with O(1) lookup times and minimal overhead for insertion and deletion operations. Until an update is available, administrators can implement rate limiting at the network perimeter using firewalls or intrusion prevention systems to restrict the number of SMTP connections from single IP addresses within a given time window. Additionally, deploying anti-spam solutions that perform preliminary header analysis before handing off messages to the mail server for full DKIM/ARC verification can help filter out excessively large headers or suspiciously signed messages originating from external domains not trusted by the organization. Monitoring worker thread utilization and SMTP connection durations may also provide early detection of such attacks in progress, allowing for rapid isolation of affected services.