{"id":153405,"date":"2026-01-05T17:12:15","date_gmt":"2026-01-05T15:12:15","guid":{"rendered":"https:\/\/prohoster.info\/blog\/novosti-interneta\/v-gnu-wget2-2-2-1-ustranena-uyazvimost-dopuskayushhaya-perezapis-proizvolnyh-fajlov"},"modified":"2026-01-05T17:12:15","modified_gmt":"2026-01-05T15:12:15","slug":"v-gnu-wget2-2-2-1-ustranena-uyazvimost-dopuskayushhaya-perezapis-proizvolnyh-fajlov","status":"publish","type":"post","link":"https:\/\/prohoster.info\/en\/blog\/news\/v-gnu-wget2-2-2-1-ustranena-uyazvimost-dopuskayushhaya-perezapis-proizvolnyh-fajlov","title":{"rendered":"A vulnerability allowing arbitrary file overwriting has been eliminated in GNU Wget2 2.2.1.","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"
The release of GNU Wget2 version 2.2.1 is available, which develops a completely rewritten and revamped version of the program for automating recursive content downloads. Wget2 offers a set of additional options, supports multi-threaded downloads, allows access to functionality through the libwget library, supports HTTP\/2 and TLS 1.3 protocols, enables downloading only changed data, can save data from streaming servers, properly handles internationalized domain names, and can re-encode downloaded content. The wget2 utility is provided under the GPLv3+ license, while the library is under the LGPLv3+. <\/p>\n
The new version addresses two vulnerabilities: <\/p>\n
\n
CVE-2025-69194 \u2014 Lack of proper file path validation when processing content in the Metalink format, used for describing links to files for download. By using the sequence '..\/' in file paths within the block, an attacker can create, delete, or overwrite arbitrary files outside the base directory into which the download occurs. For instance, the attacker could overwrite the contents of ~\/.ssh\/authorized_keys or ~\/.bashrc and execute their code on the system.\n
CVE-2025-69195 \u2014 Buffer overflow in the file name cleanup code in the get_local_filename_real() function, potentially leading to code execution when processing specially crafted URLs on download pages or handling redirects. The issue occurs when the '--restrict-file-names=windows|unix|ascii' option is enabled and is caused by allocating a fixed 1024-byte buffer without checking the actual size of the data being written. <\/ul>\n