In the Linux kernel, the following vulnerability has been resolved: nvme-tcp: fix UAF when detecting digest errors We should also bail from the io_work loop when we set rd_enabled to true, so we don't attempt to read data from the socket when the TCP stream is already out-of-sync or corrupted.
https://git.kernel.org/stable/c/c3eb461aa56e6fa94fb80442ba2586bd223a8886
https://git.kernel.org/stable/c/5914fa32ef1b7766fea933f9eed94ac5c00aa7ff
https://git.kernel.org/stable/c/19816a0214684f70b49b25075ff8c402fdd611d3
https://git.kernel.org/stable/c/160f3549a907a50e51a8518678ba2dcf2541abea
https://git.kernel.org/stable/c/13c80a6c112467bab5e44d090767930555fc17a5