LTS and ELTS report December 2025 ================================= Debian LTS ========== 1. I have been LTS frontdesk from 22.12.2025 to 28.12.2025. During this time I was responsible for triaging and analyzing newly discovered security vulnerabilities (CVE) in the following Debian packages: avahi, fluidsynth, imagemagick, net-snmp, python-filelock, qemu, gimp, gdcm, nbconvert, gnupg2, dcmtk, apache-log4j2, direwolf, fonttools, freedombox, igmpproxy, node-nodemailer, proxychains-ng, python-marshmallow, robocode, rtl-433 and ruby-httparty. The package with the most severe security impact appeared to be net-snmp because of a remotely exploitable buffer overflow vulnerability. (CVE-2025-68615). In the beginning there were not enough information to make a decision. I investigated the problem and requested more details from upstream and fortunately it turned out they had already contacted Craig Small, Debian maintainer of net-snmp, via private email and provided a targeted solution for the problem but didn't want to make it public. This issue was later addressed by Andreas Henriksson for bullseye via DLA-4430-1. 2. dcmtk: I prepared a security update for dcmtk fixing 2 CVE in bullseye (CVE-2025-14607 and CVE-2025-14841) because I had previously worked on the same package. There may be a follow-up update for bookworm and trixie in the future but the detected issues are not critical and a fix can be postponed for now. 3. apache-log4j: I prepared a security update for apache-log4j, a popular logging framework for Java and addressed a possible man-in-the-middle attack. (CVE-2025-68161) ELTS ==== 1. I have been the ELTS frontdesk from 22.12.2025 to 28.12.2025. During this time I was responsible for triaging and analyzing newly discovered security vulnerabilities (CVE) in the following Debian packages: avahi, fluidsynth, imagemagick, python-filelock, qemu, gimp, u-boot, gdcm, net-snmp, dcmtk, apache-log4j and gnupg2. 2. dcmtk and apache-log4j: Both packages went out of support in January 2026 but I decided to finish the update because the CVE were reported shortly before the EOL date. In addition apache-log4j in bullseye and buster are identical and the differences in dcmtk are also quite minor and a lot of additional time was not required. Hence the main focus was on QA and testing the packages for the supported ELTS distributions. 3. tomcat8: I continued to work on tomcat8 based on the latest available release 8.5.100. So far I could only address half of the 19 open CVE for tomcat8 and I intend to release a partial security update now because the new version is already a great improvement over the status quo. Backporting the patches for tomcat9 is non-trivial since both versions have diverged significantly over the past years. Not all of the remaining issues may affect current users of tomcat8 though because of unlikely configuration options, use cases or general server environments which make it very difficult to exploit the reported vulnerabilities. Thus we will review the remaining problems later and decide on a case-by-case basis wether it makes sense to backport more tomcat9 patches to tomcat8.