Fedora has issued an advisory today (July 4): https://lists.fedoraproject.org/archives/list/package-announce@lists.fedoraproject.org/thread/SGPQZFVJCFGDSISFXPCQTTBBD7QZLJKI/ The issues are fixed upstream in 4.6.0. Mageia 8 is also affected.
CC: (none) => geiger.david68210, mageiaStatus comment: (none) => Fixed upstream in 4.6.0Whiteboard: (none) => MGA8TOO
This SRPM has been comitted by various people, so assigning this globally.
Assignee: bugsquad => pkg-bugs
Suggested advisory: ======================== The updated packages fix a security vulnerability: An invalid pointer initialization issue was found in the SLiRP networking implementation of QEMU. The flaw exists in the bootp_input() function and could occur while processing a udp packet that is smaller than the size of the 'bootp_t' structure. A malicious guest could use this flaw to leak 10 bytes of uninitialized heap memory from the host. The highest threat from this vulnerability is to data confidentiality. This flaw affects libslirp versions prior to 4.6.0. (CVE-2021-3592) An invalid pointer initialization issue was found in the SLiRP networking implementation of QEMU. The flaw exists in the udp6_input() function and could occur while processing a udp packet that is smaller than the size of the 'udphdr' structure. This issue may lead to out-of-bounds read access or indirect host memory disclosure to the guest. The highest threat from this vulnerability is to data confidentiality. This flaw affects libslirp versions prior to 4.6.0. (CVE-2021-3593) An invalid pointer initialization issue was found in the SLiRP networking implementation of QEMU. The flaw exists in the udp_input() function and could occur while processing a udp packet that is smaller than the size of the 'udphdr' structure. This issue may lead to out-of-bounds read access or indirect host memory disclosure to the guest. The highest threat from this vulnerability is to data confidentiality. This flaw affects libslirp versions prior to 4.6.0. (CVE-2021-3594) An invalid pointer initialization issue was found in the SLiRP networking implementation of QEMU. The flaw exists in the tftp_input() function and could occur while processing a udp packet that is smaller than the size of the 'tftp_t' structure. This issue may lead to out-of-bounds read access or indirect host memory disclosure to the guest. The highest threat from this vulnerability is to data confidentiality. This flaw affects libslirp versions prior to 4.6.0. (CVE-2021-3595) References: https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2021-3592 https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2021-3593 https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2021-3594 https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2021-3595 https://lists.fedoraproject.org/archives/list/package-announce@lists.fedoraproject.org/thread/SGPQZFVJCFGDSISFXPCQTTBBD7QZLJKI/ ======================== Updated packages in core/updates_testing: ======================== lib(64)slirp-devel-4.4.0-1.1.mga8 lib(64)slirp0-4.4.0-1.1.mga8 from SRPM: libslirp-4.4.0-1.1.mga8.src.rpm
Status: NEW => ASSIGNEDVersion: Cauldron => 8CC: (none) => nicolas.salgueroWhiteboard: MGA8TOO => (none)Status comment: Fixed upstream in 4.6.0 => (none)Assignee: pkg-bugs => qa-bugs
MGA8-64 Plasma on Lenovo B50 No installation issues No previous updates shown so # urpmq --whatrequires lib64slirp0 lib64slirp0 qemu-system-aarch64-core qemu-system-alpha-core qemu-system-arm-core and more moreof thos packages slirp4netns. Looked a thelast one in https://github.com/rootless-containers/slirp4netns and it looks like all I need to upset my wifi connection. Leaving for others.....
CC: (none) => herman.viaene
Tried on the ML to get a qemu user to test this, with no response, so I decided to take the plunge and try to learn to use qemu myself, with mixed success. Installed virt-manager, which drew in a number of qemu packages, and lib64slirp0. Managed to create a basic Mageia 8 VM which did boot up. However, whilr the VM said the Internet connection was active, it wasn't communicating with the host wifi connection the way VirtualBox does. Perhaps a wired connection would have worked better. Moving on, Using qarepo, I updated lib64slirp0 with no issues. Running the VM again, there was no change. Other than the Internet not properly connecting, everything else worked OK. This update has been sitting here for over 2 months. We have two clean installs, and an incompetent qemu user seeing no changes, so at least it doesn't seem to have broken anything. I'm going to send it on its way. Validating. Advisory in Comment 2.
CC: (none) => andrewsfarm, sysadmin-bugsKeywords: (none) => validated_updateWhiteboard: (none) => MGA8-64-OK
Keywords: (none) => advisoryCC: (none) => davidwhodgins
An update for this issue has been pushed to the Mageia Updates repository. https://advisories.mageia.org/MGASA-2021-0480.html
Resolution: (none) => FIXEDStatus: ASSIGNED => RESOLVED