Skip to main content
CVE Vulnerability Database
Vulnerability Database/CVE-2026-19024

CVE-2026-19024: HDF5 Denial of Service Vulnerability

CVE-2026-19024 is a denial of service vulnerability in HDF5 caused by NULL pointer dereference in H5Pget_fill_value. Attackers exploit malformed fill value messages to crash applications. Learn about affected versions and fixes.

Published:

CVE-2026-19024 Overview

CVE-2026-19024 is a NULL pointer dereference vulnerability in the H5Pget_fill_value function of the HDF5 library before version 2.1.1. The flaw is triggered when a dataset contains a version 1 or version 2 fill value message with the defined flag set alongside a negative size field. The library fails to normalize this malformed combination to its internal undefined sentinel value. The malformed input reaches H5T_path_find with a NULL datatype pointer, causing a crash. Attackers can weaponize a crafted HDF5 file to induce denial of service in applications that parse untrusted data.

Critical Impact

A crafted HDF5 file causes any application using HDF5 before 2.1.1 to terminate through a NULL pointer dereference in the fill-value processing path.

Affected Products

  • HDF5 library versions prior to 2.1.1
  • Applications linking against vulnerable HDF5 releases that call H5Pget_fill_value
  • Scientific, engineering, and data pipelines that ingest HDF5 datasets from untrusted sources

Discovery Timeline

  • 2026-08-05 - CVE CVE-2026-19024 published to NVD
  • 2026-08-05 - Last updated in NVD database

Technical Details for CVE-2026-19024

Vulnerability Analysis

The vulnerability lives in HDF5's property list handling for dataset fill values. HDF5 datasets can carry a fill-value message describing the default byte pattern for unwritten elements. Two on-disk encodings, version 1 and version 2, express whether the fill value is defined or undefined and record a size field. When the defined flag is asserted and the size field is set to a negative value, the parser fails to reconcile the contradiction. The library should coerce this state into its internal undefined sentinel but does not. The malformed size then propagates through H5Pget_fill_value into H5T_path_find, which is called with a NULL datatype pointer. The subsequent dereference terminates the host process. The classification maps to [CWE-476] NULL Pointer Dereference.

Root Cause

The root cause is missing input normalization in the fill-value message decoder. The decoder accepts a negative size while the defined flag remains set, producing an inconsistent internal state. Downstream consumers assume that a defined fill value carries a valid datatype pointer. That assumption breaks, and H5T_path_find receives NULL for its datatype argument.

Attack Vector

Exploitation requires an attacker to supply a crafted HDF5 file to a target application and induce the application to read its fill-value properties. The attack vector is local and requires user interaction, such as opening an attacker-supplied dataset. Impact is limited to availability; confidentiality and integrity are not affected. The vulnerability enables denial of service against pipelines, viewers, and analytics tools that process third-party HDF5 files. See the HDF5 issue tracker entry for technical discussion.

No verified public exploit is available; the mechanism is documented in prose only.

Detection Methods for CVE-2026-19024

Indicators of Compromise

  • Unexpected termination or segmentation faults in processes linked against libhdf5 when opening third-party datasets
  • HDF5 files whose fill-value message headers contain the defined flag together with a size field encoded as a negative integer
  • Crash dumps referencing H5Pget_fill_value or H5T_path_find in the call stack

Detection Strategies

  • Scan repositories of HDF5 files with a validator that inspects version 1 and version 2 fill-value messages for negative size fields
  • Instrument HDF5-consuming applications with crash reporting to capture stack frames in the fill-value code path
  • Correlate file open events with subsequent process termination in endpoint telemetry to surface anomalous parser failures

Monitoring Recommendations

  • Track versions of the HDF5 library across build environments and enforce a minimum of 2.1.1
  • Log all HDF5 file ingestion in analytics and scientific computing pipelines, including source origin
  • Alert on repeated process crashes tied to the same HDF5 input file, which may indicate exploitation attempts

How to Mitigate CVE-2026-19024

Immediate Actions Required

  • Upgrade the HDF5 library to version 2.1.1 or later across all systems that parse HDF5 data
  • Rebuild and redeploy applications that statically link HDF5 against the patched release
  • Restrict ingestion of HDF5 files to trusted sources until the upgrade is complete

Patch Information

The HDF Group addressed the issue in HDF5 2.1.1. The fix normalizes malformed fill-value messages so that a defined flag paired with a negative size collapses to the library's undefined sentinel before reaching H5T_path_find. Details are tracked in the HDF5 GitHub issue.

Workarounds

  • Pre-validate HDF5 files with a hardened parser that rejects fill-value messages carrying negative size fields
  • Run HDF5 parsing in a sandboxed or containerized process so that a crash does not disrupt the parent workflow
  • Disable automatic ingestion of externally supplied HDF5 datasets in production pipelines until patches are deployed
bash
# Configuration example
# Verify installed HDF5 version and upgrade if below 2.1.1
h5cc -showconfig | grep -i version
# Example package upgrade (adjust for your distribution or build system)
# apt-get install --only-upgrade libhdf5-dev
# pip install --upgrade h5py  # if using the Python binding, rebuild against HDF5 >= 2.1.1

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

Default Legacy - Prefooter | Experience the World’s Most Advanced Cybersecurity Platform

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.