CVE-2026-124024: IBM1390/IBM1399 encoders may read past input
Converting crafted input from UCS-4 to IBM1390 or IBM1399 character encoding, in the GNU C Library version 2.3.3 and later may result in the converter reading past the end of the input buffer, causing the calling application to crash.
When converting from UCS-4, these encoders may encounter a combining base such as U+00E6 and jump ahead by 4 bytes to check the next UCS-4 code point. The check for end of buffer however does not account for this, which may result in a buffer overread if the following code point has less than 4 bytes remaining. Exploitation of this bug not only requires attacker control of input and encoding, but also strict control over layout of the input in memory.
Affected Software
Event History
Frequently Asked Questions
What conditions are required for exploitation?
An attacker needs control of both the input and the selected encoding, and the application must convert crafted UCS-4 input to IBM1390 or IBM1399. Exploitation also requires strict control of the input buffer's memory layout.
What is the practical impact on an affected application?
The converter can read beyond the end of the input buffer, causing the calling application to crash. The provided information identifies availability impact only; it does not identify confidentiality or integrity impact.
How can I determine whether an application is exposed?
Identify applications using GNU C Library 2.3.3 or later that perform UCS-4 conversion to IBM1390 or IBM1399. Exposure further depends on whether untrusted parties can influence both the converted input and the encoding choice.