{"id":21568,"date":"2026-09-12T04:31:21","date_gmt":"2026-09-11T20:31:21","guid":{"rendered":"https:\/\/www.54iter.cn\/?vuln=cve-2026-89646"},"modified":"2026-09-12T04:31:21","modified_gmt":"2026-09-11T20:31:21","slug":"cve-2026-89646","status":"publish","type":"vuln","link":"https:\/\/www.54iter.cn\/?vuln=cve-2026-89646","title":{"rendered":"CVE-2026-89646 &#8211; In the Linux kernel, the following vulnerability has been resolved:\n\nceph: fix leaked inod"},"content":{"rendered":"<p>In the Linux kernel, the following vulnerability has been resolved:<\/p>\n<p>ceph: fix leaked inode reference on writeback abort at umount<\/p>\n<p>ceph_dirty_folio() takes a wrbuffer claim on each newly dirtied folio: it<br \/>\nbumps i_wrbuffer_ref (taking an ihold() on the 0-&gt;1 transition) and<br \/>\nattaches the snap_context to folio-&gt;private.  That claim is released only<br \/>\nby ceph_put_wrbuffer_cap_refs(), which for a submitted write runs from<br \/>\nwritepages_finish().<\/p>\n<p>In ceph_submit_write(), if ceph_inc_osd_stopping_blocker() fails &#8212; which<br \/>\nhappens during umount &#8212; the request is aborted before submission: the<br \/>\nalready-collected folios are only redirtied and unlocked, so<br \/>\nwritepages_finish() never runs and the claim is leaked.<br \/>\nredirty_page_for_writepage() -&gt; folio_redirty_for_writepage() -&gt;<br \/>\nfilemap_dirty_folio() sets PG_dirty directly and does not go through<br \/>\n-&gt;dirty_folio, so ceph_dirty_folio() is not re-entered to rebalance it.<br \/>\nBecause every subsequent writeback also fails the osd_stopping_blocker,<br \/>\ni_wrbuffer_ref never returns to 0, the ihold() is never dropped, and the<br \/>\ninode cannot be evicted:<\/p>\n<p>  VFS: Busy inodes after unmount of ceph<br \/>\n  kernel BUG at fs\/super.c:650!<\/p>\n<p>Release the orphaned claim in the abort path before redirtying, via<br \/>\nceph_undo_wrbuffer_claim(): detach the snap_context, drop the wrbuffer<br \/>\nreference (letting i_wrbuffer_ref reach 0 and iput() the inode), and drop<br \/>\nthe snap_context reference &#8212; i.e. do what writepages_finish() would have<br \/>\ndone for these never-submitted folios.<\/p>\n<p>Only the locked_pages entries are undone; folios still in the fbatch were<br \/>\nnever dirty-cleared by this call (folio_clear_dirty_for_io() is the<br \/>\nownership-transfer point, and a successful move NULLs the fbatch slot), so<br \/>\nthey hold no claim this call owns.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>In the Linux kernel, the following vulnerability has been resolved:<\/p>\n<p>ceph: fix leaked inode reference on writeback abort at umount<\/p>\n<p>ceph_dirty_folio() takes a wrbuffer claim on each newly dirtied foli<\/p>\n","protected":false},"template":"","meta":[],"categories":[],"tags":[],"class_list":["post-21568","vuln","type-vuln","status-publish","hentry"],"_links":{"self":[{"href":"https:\/\/www.54iter.cn\/index.php?rest_route=\/wp\/v2\/vuln\/21568","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=21568"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.54iter.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=21568"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.54iter.cn\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=21568"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}