FreeRDP CVE-2026-110837

Exploitation requires the FreeRDP client to be built with the specific non-default combination WITH_LAME=ON and WITH_DSP_FFMPEG=OFF, so that the vulnerable LAME-based freerdp_dsp_decode_mp3 path is compiled in rather than the FFMPEG DSP backend. The victim must then initiate or maintain an RDP session to an attacker-controlled or compromised server that streams malicious MP3 audio over the rdpsnd channel. Key limiting factors: the attack originates from the server side (victim must connect out to a malicious endpoint), the build configuration is not the default for most distributions, and the documented impact is a client crash (denial of service); the high attack complexity (AC:H) and attack requirement (AT:P) reflect the need for both the specific build and a crafted audio stream that forces samples across the allocation boundary.

This is a moderate, conditional client-side denial-of-service issue rather than a top-tier priority. The supplied CVSS 4.0 vector (AV:N/AC:H/AT:P/PR:L/UI:N, score 6.1) already encodes two major dampeners: high attack complexity (AC:H) and a present attack requirement (AT:P), which reflect the fact that exploitation only works against a non-default build (WITH_LAME=ON and WITH_DSP_FFMPEG=OFF) and requires decoded MP3 samples to cross the heap allocation. There is no EPSS score provided, no CISA KEV listing, and no public exploit identified at time of analysis, so all three exploitation-likelihood signals are either low or absent and point to no evidence of widespread or opportunistic exploitation. One internal inconsistency should be flagged for verification: the input vector rates VI:H/VA:H (high integrity and availability impact), but the description only substantiates a client crash ('decoded samples cross the heap allocation, crashing the client'), which supports an availability impact but not clearly a high integrity/code-execution impact - the heap-overflow primitive (CWE-122) makes memory corruption plausible but it is not confirmed by the description, so the high integrity rating should be verified with the FreeRDP/VulnCheck advisories before treating code execution as established. Also note the attack is inverted from a typical RDP flaw: the threat actor is a malicious or compromised RDP server attacking connecting clients, so PR:L in the input does not mean the attacker needs victim credentials.

A victim using a FreeRDP-based client that was compiled with the LAME MP3 decoder enabled and FFMPEG DSP disabled connects to an attacker-controlled or compromised RDP server. The server streams a crafted MP3 frame over the rdpsnd audio channel, and the right-channel PCM pointer miscalculation causes decoded samples to overrun the heap allocation, crashing the client. No public exploit code has been identified at time of analysis, and the attack is limited to this specific non-default build configuration.