Skip to main content
Vulnerability Database/CVE-2026-112278

CVE-2026-112278: rdesktop Heap Out-of-Bounds Read DoS

CVE-2026-112278 is a heap out-of-bounds read denial of service vulnerability in rdesktop that allows malicious RDP servers to crash clients. This post explains its impact, affected versions, and mitigation steps.

Published:

Disclaimer: This content was generated using AI. While we strive for accuracy, please verify critical information with official sources.

Sources: National Vulnerability Database, CISA KEV, Exploit-DB

CVE-2026-112278 Overview

CVE-2026-112278 is a heap out-of-bounds read vulnerability in rdesktop through version 1.9.0. The flaw resides in the process_raw_bmpcache() function within orders.c. A malicious Remote Desktop Protocol (RDP) server can send a crafted raw bitmap cache order that declares large width, height, and bits-per-pixel (bpp) values while providing a very small payload. The resulting memcpy() reads past the receive buffer boundary and crashes the client. The vulnerability is tracked as [CWE-125] Out-of-Bounds Read and affects users who connect rdesktop clients to untrusted RDP endpoints.

Critical Impact

A malicious RDP server can crash a connected rdesktop client, producing a denial-of-service condition and potentially leaking adjacent heap memory contents.

Affected Products

  • rdesktop versions up to and including 1.9.0
  • Linux and Unix distributions shipping rdesktop as the default RDP client
  • Any downstream package or appliance that embeds the rdesktop codebase

Discovery Timeline

  • 2026-10-19 - CVE-2026-112278 published to the National Vulnerability Database (NVD)
  • 2026-10-22 - Last updated in NVD database

Technical Details for CVE-2026-112278

Vulnerability Analysis

The vulnerability occurs during the parsing of raw bitmap cache orders received from an RDP server. The process_raw_bmpcache() function in orders.c reads width, height, and bpp fields from the server-controlled stream. These fields determine how many bytes the function expects to copy from the receive buffer into a newly allocated bitmap structure.

The function does not validate that the declared size matches the actual payload remaining in the receive buffer. When a server advertises a large bitmap but supplies a tiny amount of pixel data, the subsequent memcpy() reads bytes beyond the end of the receive buffer. The read crosses into unmapped memory or adjacent heap allocations, triggering a client crash. In some cases the out-of-bounds data may propagate into the bitmap cache and leak heap contents through later rendering operations.

Root Cause

The root cause is missing bounds validation between the attacker-declared bitmap dimensions and the length of the received payload. The code trusts the width, height, and bpp fields transmitted by the server. The resulting size calculation (width * height * (bpp / 8)) is passed directly to memcpy() without comparing it against the remaining readable bytes in the stream.

Attack Vector

Exploitation requires a victim to initiate an RDP connection to a server controlled by an attacker. User interaction is required to launch rdesktop and connect to the malicious endpoint. Phishing lures, poisoned bookmarks, and compromised .rdp configuration files are typical delivery mechanisms. Once the TLS handshake completes, the server responds with a crafted raw bitmap cache order that triggers the out-of-bounds read on the client.

For technical analysis of the flaw, see the HackMD OOB Read Analysis and the VulnCheck Advisory for rdesktop.

Detection Methods for CVE-2026-112278

Indicators of Compromise

  • Repeated rdesktop process crashes or segmentation faults logged shortly after an outbound RDP session is established
  • Core dumps referencing process_raw_bmpcache or orders.c on the call stack
  • Outbound RDP (TCP/3389) connections from workstations to unexpected or untrusted external hosts

Detection Strategies

  • Inspect host telemetry for abnormal termination of the rdesktop binary and correlate with active network sessions on port 3389
  • Hunt for .rdp files or shortcuts delivered through email, chat, or file shares that point to external IP addresses
  • Monitor network flows for RDP traffic originating from users or systems that do not normally use the protocol

Monitoring Recommendations

  • Enable application crash reporting on Linux endpoints running rdesktop and forward core dump metadata to a central log store
  • Alert on new outbound RDP connections to Internet-facing destinations outside approved jump hosts
  • Track package inventory for installed rdesktop versions at or below 1.9.0 across the fleet

How to Mitigate CVE-2026-112278

Immediate Actions Required

  • Inventory all systems running rdesktop and identify versions at or below 1.9.0
  • Restrict outbound RDP connections from endpoints to a vetted list of internal destinations
  • Advise users to avoid connecting to untrusted or unknown RDP servers until a patched build is deployed

Patch Information

At publication, upstream rdesktop has not shipped a tagged release that addresses CVE-2026-112278. A proposed fix is tracked in GitHub Pull Request #437 and the vulnerable code path can be reviewed in the GitHub rdesktop Source Code. Downstream distributions may ship backported patches ahead of an upstream release; monitor your distribution's security advisories and apply the vendor-supplied package once available.

Workarounds

  • Replace rdesktop with an actively maintained RDP client such as FreeRDP where feasible
  • Block outbound TCP/3389 at the perimeter and require RDP sessions to traverse an authenticated gateway
  • Apply the patch from GitHub Pull Request #437 and rebuild rdesktop from source for environments that cannot migrate clients
bash
# Identify installed rdesktop version and block outbound RDP
rdesktop --version
sudo iptables -A OUTPUT -p tcp --dport 3389 -j REJECT

Disclaimer: This content was generated using AI. While we strive for accuracy, please verify critical information with official sources.

Sources: National Vulnerability Database, CISA KEV, Exploit-DB

Experience the Most Advanced Cybersecurity Platform

See how the world’s most intelligent, autonomous cybersecurity platform can protect your organization today and into the future.