CVE-2026-6368

The iconv() function in the GNU C Library versions 2.39 and older may overflow the output buffer by up to 4 bytes when converting strings to the ISO-2022-CN-EXT character set. This may lead to application crashes or overwrite a neighbouring variable.

In GNU C Library versions 2.13 to 2.40, when the assert() function fails, it does not allocate enough space for the assertion failure message string and size information, which may lead to a buffer overflow if the message string size aligns to page size.

In GNU C Library versions 2.27 to 2.38, an untrusted LD_LIBRARY_PATH environment variable vulnerability allows attacker-controlled loading of dynamically shared libraries in statically compiled setuid binaries that call dlopen (including internal dlopen calls after setlocale or calls to NSS functions).

Vulnerability in the GNU C Library (glibc) versions 2.7 through 2.43. Calling scanf family functions with %mc (malloc'd character match) and an explicit width greater than 1024 can cause a one-byte heap buffer overflow.

The strncasecmp function in the GNU C Library 2.24 and later optimized for the Power8 architecture may read one byte beyond the input size limit, which may crash a program when that byte is not readable. This occurs when attacker-controlled input strings match up to the edge of their page and the neighboring page is either not mapped or not readable.

Initializing the DNS stub resolver from /etc/resolv.conf or LOCALDOMAIN with a search list containing a domain of roughly 200 characters or more in GNU C Library versions 2.26 to 2.44 results in an assertion failure that aborts the process. The resolver truncates the search list when copying into the fixed-size _res.defdname buffer, but the consistency check compares against the wrong size and does not handle a first entry that does not fit, causing the assertion to fail.

Converting crafted EUC_JISX0213 input to UCS-4 or the internal wide character encoding, for example with iconv, in the GNU C Library version 2.3 to 2.44 may result in the converter making no progress, causing the calling application to hang. Some EUC_JISX0213 sequences decode to two code points. If the output buffer has room for only the first one, the converter stores the second in the conversion state and returns E2BIG, but it never clears that pending character after emitting it on the next call. The converter then keeps emitting the pending character without consuming further input, so an application that retries the conversion loops forever. The input must be attacker controlled and the application must convert it with an output buffer small enough to split the two code points. Only the EUC_JISX0213 character set is affected, which is not commonly used.

Converting crafted SHIFT_JISX0213 input to UCS-4 or the internal wide character encoding, for example with iconv, in the GNU C Library version 2.3 to 2.44 may result in the converter making no progress, causing the calling application to hang. Some SHIFT_JISX0213 sequences decode to two code points. If the output buffer has room for only the first one, the converter stores the second in the conversion state and returns E2BIG, but it never clears that pending character after emitting it on the next call. The converter then keeps emitting the pending character without consuming further input, so an application that retries the conversion loops forever. The input must be attacker controlled and the application must convert it with an output buffer small enough to split the two code points. Only the SHIFT_JISX0213 character set is affected, which is not commonly used.

In the GNU C Library versions 2.1 to 2.44, calling tdelete on a sufficiently deep tree may write one pointer past the end of an alloca-allocated array on the stack, which may crash the application. The tdelete implementation keeps an explicit stack of parent nodes for rebalancing, which is grown as needed while descending the tree; two rebalancing branches push an additional entry without checking the capacity.