{"id":7592,"date":"2026-09-04T18:49:36","date_gmt":"2026-09-04T10:49:36","guid":{"rendered":"https:\/\/www.54iter.cn\/?vuln=cve-2026-74572"},"modified":"2026-09-04T18:49:36","modified_gmt":"2026-09-04T10:49:36","slug":"cve-2026-74572","status":"publish","type":"vuln","link":"https:\/\/www.54iter.cn\/?vuln=cve-2026-74572","title":{"rendered":"CVE-2026-74572 [HIGH] &#8211; In the Linux kernel, the following vulnerability has been resolved:\n\nbtrfs: zoned: fix dea"},"content":{"rendered":"<p>In the Linux kernel, the following vulnerability has been resolved:<\/p>\n<p>btrfs: zoned: fix deadlock between metadata writeback and transaction commit<\/p>\n<p>When writing out metadata extent buffers in a zoned filesystem,<br \/>\nbtree_writepages() holds fs_info-&gt;zoned_meta_io_lock across the whole<br \/>\nwriteback loop, including the call to btrfs_check_meta_write_pointer() -&gt;<br \/>\ncheck_bg_is_active().<\/p>\n<p>For the tree-log block group, check_bg_is_active() may fail to activate<br \/>\nthe zone and fall back to btrfs_zone_finish_one_bg() to free an active<br \/>\nzone. That path waits for the running transaction to commit while still<br \/>\nholding zoned_meta_io_lock, but the committer needs that same lock to<br \/>\nwrite out the tree extents, so the two tasks deadlock:<\/p>\n<p>  Task A (kworker, metadata writeback)      Task B (fsstress, transaction commit)<br \/>\n  &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;      &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;-<br \/>\n  wb_workfn()                               btrfs_commit_transaction(T)<br \/>\n   btree_writepages()                        btrfs_write_and_wait_transaction()<br \/>\n    btrfs_zoned_meta_io_lock()                btrfs_write_marked_extents()<br \/>\n    btrfs_check_meta_write_pointer()           btree_writepages()<br \/>\n     check_bg_is_active() [treelog_bg]          btrfs_zoned_meta_io_lock()<br \/>\n      btrfs_zone_finish_one_bg()<br \/>\n        do_zone_finish()<br \/>\n         btrfs_inc_block_group_ro()<br \/>\n          btrfs_wait_for_commit()<\/p>\n<p>The sibling branch in check_bg_is_active() already drops zoned_meta_io_lock<br \/>\naround do_zone_finish() for this exact reason. Do the same in the tree-log<br \/>\nbranch: release the lock around btrfs_zone_finish_one_bg() and re-acquire<br \/>\nit afterwards. The lock only protects fs_info-&gt;active_{meta,system}_bg,<br \/>\nwhich this branch does not touch, and ctx-&gt;zoned_bg keeps a reference to<br \/>\nthe block group across the unlock, so nothing is lost while the lock<br \/>\nis dropped.<\/p>\n<p>This hang occasionally reproduces with fstests generic\/475 on a zoned<br \/>\nbtrfs filesystem.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>In the Linux kernel, the following vulnerability has been resolved:<\/p>\n<p>btrfs: zoned: fix deadlock between metadata writeback and transaction commit<\/p>\n<p>When writing out metadata extent buffers in a zoned f<\/p>\n","protected":false},"template":"","meta":[],"categories":[],"tags":[],"class_list":["post-7592","vuln","type-vuln","status-publish","hentry"],"_links":{"self":[{"href":"https:\/\/www.54iter.cn\/index.php?rest_route=\/wp\/v2\/vuln\/7592","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=7592"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.54iter.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=7592"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.54iter.cn\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=7592"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}