Improper Output Neutralization for Logs vulnerability in the FileAppender, RollingFileAppender, UdpAppender and SmtpAppender of Apache log4net. Characters the configured encoding could not represent were converted by that encoding's best-fit fallback, which turns fullwidth characters such as U+FF1C, U+FF1E and U+FF02 into <, > and ". The conversion happened after XmlLayout or XmlLayoutSchemaLog4J had escaped the text, so a party whose data reaches a log message could forge complete log events in an XML log, for example a fake audit entry. On .NET Framework this needs no configuration where the ANSI code page is single-byte, since that is the default encoding; on .NET it needs a configured single-byte encoding such as ISO-8859-1. This is the best-fit conversion described by the WorstFit research. This issue affects Apache log4net: from 1.2.9 before 3.6.0. Users are recommended to upgrade to version 3.6.0, which fixes the issue.
https://lists.apache.org/thread.html/v2oljwjcboxk0tvyy5v86o2fjlslx1zt
https://github.com/apache/logging-log4net/pull/355
https://github.com/apache/logging-log4net/commit/735db80ab163ef08c54d2cc4d3d576b22b77f4bc