CVE-2019-12480 is an out-of-bounds read (CWE-125) in the BACnet Protocol Stack, the open-source bacnet-stack C library, through 0.8.6: malformed DeviceCommunicationControl, AtomicReadFile or AtomicWriteFile requests make the decoder read invalid memory and crash the bacserv server, a remote unauthenticated denial of service. NVD scores it 7.5. Version 0.8.7 adds strict decoding.
Vulnerability summary
CVSS base score
7.5
High severity
EPSS score (FIRST)
33.9%
Probability of exploitation in the wild in the next 30 days
CISA KEV catalog
No
Not listed
Published
Vulnerability ID: CVE-2019-12480 · Source: NVD
Affected components
| Component | CPE | Version range |
|---|---|---|
| BACnet Protocol Stack / bacnet-stack (APDU decoders, bacserv demo server) | cpe:2.3:a:bacnet_protocol_stack_project:bacnet_protocol_stack:*:*:*:*:*:*:*:* | through 0.8.6 (NVD); strict decoders for DCC, AtomicReadFile and AtomicWriteFile in 0.8.7 (bacnet-stack 0.8.6 to 0.8.7 compare, researcher disclosure) |
| Embedded BACnet devices whose firmware is built on bacnet-stack 0.8.6 or older | — | per vendor firmware; check the bundled stack version |
What does this mean for affected devices?
CVE-2019-12480 hits the APDU decoders of the BACnet Protocol Stack, the project now published as bacnet-stack. NVD's description: the stack "through 0.8.6 has a segmentation fault leading to denial of service in BACnet APDU Layer because a malformed DCC in AtomicWriteFile, AtomicReadFile and DeviceCommunicationControl services. An unauthenticated remote attacker could cause a denial of service (bacserv daemon crash) because there is an invalid read in bacdcode.c during parsing of alarm tag numbers." NVD records CWE-125 and its own Primary CVSS v3.0 score of 7.5 (high), vector CVSS:3.0/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H (NVD).
How was it found and reported?
Through fuzzing of the demo server. The researcher's disclosure write-up gives single-packet UDP payloads for port 47808 that crash bacserv in the DeviceCommunicationControl, AtomicReadFile and AtomicWriteFile handlers, and dates public disclosure to 20/08/2019. The project's bug #62, created 2019-05-28, shows a related crash in alarm_ack_decode_service_request from one malformed packet; the project's SECURITY.md files CVE-2019-12480 under that bug. These services are exactly the ones a device on a BACnet network will answer without any credential.
Which version fixes it?
The 0.8.7 release. The researcher states the fix shipped in 0.8.7, and the 0.8.6 to 0.8.7 comparison contains the commits "Change DCC to have strict decoding to prevent stack overflow", "Fix AtomicReadFile decoding to strictly enforce decoding rules" and the matching AtomicWriteFile fix, which rewrite dcc.c, arf.c, awf.c and bacdcode.c. SECURITY.md lists every branch before 1.4.x as unsupported today, so a product still on 0.8.x gets no upstream fixes for the later decoder bugs either.
No device vendor or CISA ICS advisory naming a product affected by this CVE was found (CISA's CSAF advisory set checked 2026-11-17). bacnet-stack is built into embedded BACnet devices, so the exposure sits in whichever firmware forked an old version. NVD files this CVE under the CPE vendor bacnet_protocol_stack_project, while newer bacnet-stack CVEs use bacnetstack:bacnet_stack; an SBOM matcher needs both names to catch every record.
What should teams do about it now?
Check for old forks. Our BACnet protocol security guide lists this CVE first among the stack's decoder bugs; the newer ones are CVE-2026-26264 and CVE-2026-41475. A firmware SBOM shows which version each product carries. For the weakness class, see our memory corruption fix guide.
CVE-2019-12480 is not in the CISA KEV catalog as of 2026-11-17 (catalog version 2026.11.13). FIRST EPSS gives it 0.3389, the 98th percentile, in the 2026-11-16 dataset, helped by the public proof-of-concept payloads (CISA KEV, FIRST EPSS).
To fix it, move firmware off bacnet-stack 0.8.6 and older to the newest release of a supported branch (1.4.x or 1.6.x and later per SECURITY.md), which also covers the later decoder CVEs. Until then, keep BACnet/IP reachable only from the building management network.
Evidence
- Through 0.8.6: segfault from malformed DCC, AtomicWriteFile, AtomicReadFile; unauthenticated remote DoS (bacserv crash); invalid read in bacdcode.c parsing alarm tag numbers; NVD Primary CVSS v3.0 7.5; CWE-125; CPE bacnet_protocol_stack_project:bacnet_protocol_stack through 0.8.6. — NVD
- Bug #62 Invalid read in bacserv when decoding alarm tags, created 2019-05-28; one UDP packet to 47808 crashes in alarm_ack_decode_service_request. — SourceForge bacnet bug #62
- Fuzzing crashes in DCC, AtomicReadFile and AtomicWriteFile with UDP payloads; public disclosure 20/08/2019; fix released for 0.8.7. — Researcher disclosure
- 0.8.6...0.8.7 includes strict decoding commits for DCC, AtomicReadFile, AtomicWriteFile; dcc.c, arf.c, awf.c, bacdcode.c modified. — bacnet-stack GitHub compare
- CVE-2019-12480 listed under bug #62; 1.4.x, 1.6.x, 1.7.x, 1.8.x supported; 1.3.x and older unsupported. — bacnet-stack SECURITY.md
- Not in the CISA KEV catalog as of 2026-11-17 (catalog version 2026.11.13). — CISA KEV
- EPSS 0.3389 (98th percentile), dataset 2026-11-16, read 2026-11-17. — FIRST EPSS

Related advisories
Sources
Frequently asked questions
Which BACnet services trigger CVE-2019-12480?
Malformed DeviceCommunicationControl, AtomicReadFile and AtomicWriteFile requests, with an invalid read in bacdcode.c while parsing tags. One UDP packet is enough (per NVD and the researcher's disclosure).
Which version fixes CVE-2019-12480?
0.8.7, which adds strict decoding for those services. That branch is now unsupported, so move to a supported branch such as 1.4.x or 1.6.x and later (per the bacnet-stack compare view and SECURITY.md).
Why can an SBOM scan miss CVE-2019-12480?
NVD files it under the CPE vendor bacnet_protocol_stack_project, while newer bacnet-stack CVEs use bacnetstack. A matcher that knows only one name misses the other records (per NVD).
Is CVE-2019-12480 in the CISA KEV catalog?
No, not as of 2026-11-17 (catalog version 2026.11.13). FIRST EPSS gives it 0.3389, the 98th percentile, in the 2026-11-16 dataset (per CISA KEV and FIRST EPSS).
Questions about this page?