{"id":25547,"date":"2026-09-18T04:31:32","date_gmt":"2026-09-17T20:31:32","guid":{"rendered":"https:\/\/www.54iter.cn\/?vuln=cve-2026-90138"},"modified":"2026-09-18T04:31:32","modified_gmt":"2026-09-17T20:31:32","slug":"cve-2026-90138","status":"publish","type":"vuln","link":"https:\/\/www.54iter.cn\/?vuln=cve-2026-90138","title":{"rendered":"CVE-2026-90138 &#8211; In the Linux kernel, the following vulnerability has been resolved:\n\nvsock: don&#8217;t check th"},"content":{"rendered":"<p>In the Linux kernel, the following vulnerability has been resolved:<\/p>\n<p>vsock: don&#8217;t check the listener&#8217;s sk_err in vsock_accept()<\/p>\n<p>Syzbot reported an issue which can be reproduced with these steps:<br \/>\n\tr0 = socket(AF_VSOCK, SOCK_STREAM, 0)<br \/>\n\tbind(r0, {VMADDR_CID_ANY, PORT})<br \/>\n\tconnect(r0, {VMADDR_CID_LOCAL, PORT}) -&gt; -1, EPROTO (self-connect)<br \/>\n\tlisten(r0, backlog)                   -&gt; 0<br \/>\n\tr1 = socket(AF_VSOCK, SOCK_STREAM, 0)<br \/>\n\tconnect(r1, {VMADDR_CID_LOCAL, PORT}) -&gt; 0<br \/>\n\taccept(r0)                            -&gt; -1, EPROTO (stale sk_err)<\/p>\n<p>Basically, it creates a socket (r0) and triggers a self-connect after<br \/>\nbinding it. This self-connect fails with EPROTO because it loops back<br \/>\nto r0 while the socket is still in the TCP_SYN_SENT state, causing it<br \/>\nto be incorrectly dispatched to the connecting-client path. The<br \/>\nunexpected packet type encountered there sets sk_err to EPROTO.<\/p>\n<p>After that, it invokes a listen() call on the same socket. This<br \/>\nlisten() call succeeds because the kernel&#8217;s listening path never<br \/>\ninspects or clears sk_err. Then, a new socket (r1) is created as a<br \/>\nnormal client and connects to r0. However, vsock_accept() rejects this<br \/>\nincoming connection because the listener&#8217;s sk_err still holds the<br \/>\nEPROTO error from the earlier failed self-connect.<\/p>\n<p>This rejection causes the child socket created for r1&#8217;s connection to<br \/>\nnever be freed on virtio or hyperv transports; only the VMCI transport<br \/>\nimplements pending_work to revisit and clean up a rejected socket.<\/p>\n<p>For a non-blocking connect(), vsock_connect() may return -EINPROGRESS<br \/>\nimmediately, and vsock_connect_timeout() can later set sk-&gt;sk_err<br \/>\nasynchronously.<\/p>\n<p>Since no vsock transport ever sets sk_err on a socket while it is in<br \/>\nTCP_LISTEN state, checking it in vsock_accept() serves no purpose and<br \/>\nonly carries forward errors left behind by earlier, unrelated<br \/>\nconnection attempts on the same socket. Remove the checks so accept()<br \/>\nno longer rejects valid incoming connections because of a stale<br \/>\nerror, which also avoids the resource leak described above.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>In the Linux kernel, the following vulnerability has been resolved:<\/p>\n<p>vsock: don&#8217;t check the listener&#8217;s sk_err in vsock_accept()<\/p>\n<p>Syzbot reported an issue which can be reproduced with these steps:<br \/>\n\tr0 <\/p>\n","protected":false},"template":"","meta":[],"categories":[],"tags":[],"class_list":["post-25547","vuln","type-vuln","status-publish","hentry"],"_links":{"self":[{"href":"https:\/\/www.54iter.cn\/index.php?rest_route=\/wp\/v2\/vuln\/25547","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.54iter.cn\/index.php?rest_route=\/wp\/v2\/vuln"}],"about":[{"href":"https:\/\/www.54iter.cn\/index.php?rest_route=\/wp\/v2\/types\/vuln"}],"wp:attachment":[{"href":"https:\/\/www.54iter.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=25547"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.54iter.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=25547"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.54iter.cn\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=25547"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}