Fix Change Details ------------------ (Compare with comment #3)
The xattr dir/child inodes are evicted/disposed immediately on file removal, not on shrinker. $ touch /zfs/file <...>-23940 [006] d... 10808.746798: p_zfs_mknode_0: (zfs_mknode+0x0/0xe00 [zfs]) flag=0x0 dzp=0xffff88011aa94400 <...>-23940 [006] d... 10808.746925: p_zfs_znode_alloc_0: (zfs_znode_alloc+0x0/0x520 [zfs]) obj=0xa <...>-23940 [006] d... 10808.746938: r_zfs_znode_alloc_0: (zfs_mknode+0x8a3/0xe00 [zfs] <- zfs_znode_alloc) zpp=0xffff88011aa90880 $ setfattr -n user.debug -v 1 /zfs/file <...>-23941 [006] d... 10818.945733: p_zfs_mknode_0: (zfs_mknode+0x0/0xe00 [zfs]) flag=0x2 dzp=0xffff88011aa90880 <...>-23941 [006] d... 10818.945892: p_zfs_znode_alloc_0: (zfs_znode_alloc+0x0/0x520 [zfs]) obj=0xb <...>-23941 [006] d... 10818.945905: r_zfs_znode_alloc_0: (zfs_mknode+0x8a3/0xe00 [zfs] <- zfs_znode_alloc) zpp=0xffff88011aa961c0 <...>-23941 [006] d... 10818.945925: p_zfs_zget_0: (zfs_zget+0x0/0x230 [zfs]) zsb=0xffff880016f00000 obj=0xa setfattr-23941 [006] d... 10818.945929: r_zfs_zget_0: (zfs_zaccess+0x12b/0x220 [zfs] <- zfs_zget) setfattr-23941 [006] d... 10818.945933: p_zfs_zget_0: (zfs_zget+0x0/0x230 [zfs]) zsb=0xffff880016f00000 obj=0xa setfattr-23941 [006] d... 10818.945936: r_zfs_zget_0: (zfs_zaccess+0x12b/0x220 [zfs] <- zfs_zget) setfattr-23941 [006] d... 10818.945948: p_zfs_zget_0: (zfs_zget+0x0/0x230 [zfs]) zsb=0xffff880016f00000 obj=0xa setfattr-23941 [006] d... 10818.945951: r_zfs_zget_0: (zfs_zaccess+0x12b/0x220 [zfs] <- zfs_zget) setfattr-23941 [006] d... 10818.945966: p_zfs_mknode_0: (zfs_mknode+0x0/0xe00 [zfs]) flag=0x0 dzp=0xffff88011aa961c0 setfattr-23941 [006] d... 10818.945990: p_zfs_znode_alloc_0: (zfs_znode_alloc+0x0/0x520 [zfs]) obj=0xc setfattr-23941 [006] d... 10818.945996: r_zfs_znode_alloc_0: (zfs_mknode+0x8a3/0xe00 [zfs] <- zfs_znode_alloc) zpp=0xffff88011aa90440 $ rm /zfs/file <...>-23942 [006] d... 10822.453456: p_zfs_zget_0: (zfs_zget+0x0/0x230 [zfs]) zsb=0xffff880016f00000 obj=0xa <...>-23942 [006] d... 10822.453469: r_zfs_zget_0: (zfs_dirent_lock+0x56c/0x6c0 [zfs] <- zfs_zget) rm-23942 [006] d... 10822.453494: p_zfs_zget_0: (zfs_zget+0x0/0x230 [zfs]) zsb=0xffff880016f00000 obj=0xb rm-23942 [006] d... 10822.453498: r_zfs_zget_0: (zfs_remove+0x22b/0x4c0 [zfs] <- zfs_zget) rm-23942 [006] d... 10822.453597: p_zpl_evict_inode_0: (zpl_evict_inode+0x0/0x60 [zfs]) inode=0xffff88011aa90a90 rm-23942 [006] d... 10822.453602: p_zfs_inactive_0: (zfs_inactive+0x0/0x270 [zfs]) inode=0xffff88011aa90a90 rm-23942 [006] d... 10822.453603: p_zfs_zinactive_0: (zfs_zinactive+0x0/0xe0 [zfs]) znode=0xffff88011aa90880 obj=0xa rm-23942 [006] d... 10822.453607: p_zfs_rmnode_0: (zfs_rmnode+0x0/0x360 [zfs]) znode=0xffff88011aa90880 rm-23942 [006] d... 10822.453619: p_zfs_zget_0: (zfs_zget+0x0/0x230 [zfs]) zsb=0xffff880016f00000 obj=0xb rm-23942 [006] d... 10822.453623: r_zfs_zget_0: (zfs_rmnode+0x25b/0x360 [zfs] <- zfs_zget) rm-23942 [006] d... 10822.453656: p_zfs_iput_async_0: (zfs_iput_async+0x0/0x60 [zfs]) inode=0xffff88011aa963d0 obj=0x0 z_iput-23904 [006] d... 10822.453675: p_zpl_evict_inode_0: (zpl_evict_inode+0x0/0x60 [zfs]) inode=0xffff88011aa963d0 z_iput-23904 [006] d... 10822.453676: p_zfs_inactive_0: (zfs_inactive+0x0/0x270 [zfs]) inode=0xffff88011aa963d0 z_iput-23904 [006] d... 10822.453677: p_zfs_zinactive_0: (zfs_zinactive+0x0/0xe0 [zfs]) znode=0xffff88011aa961c0 obj=0xb z_iput-23904 [006] d... 10822.453679: p_zfs_rmnode_0: (zfs_rmnode+0x0/0x360 [zfs]) znode=0xffff88011aa961c0 z_iput-23904 [006] d... 10822.453681: p_zfs_purgedir_0: (zfs_purgedir+0x0/0x230 [zfs]) znode=0xffff88011aa961c0 z_iput-23904 [006] d... 10822.453686: p_zfs_zget_0: (zfs_zget+0x0/0x230 [zfs]) zsb=0xffff880016f00000 obj=0xc z_iput-23904 [006] d... 10822.453690: r_zfs_zget_0: (zfs_purgedir+0xb4/0x230 [zfs] <- zfs_zget) z_iput-23904 [006] d... 10822.453713: p_zfs_iput_async_0: (zfs_iput_async+0x0/0x60 [zfs]) inode=0xffff88011aa90650 obj=0x0 z_iput-23904 [006] d... 10822.453719: r_zfs_purgedir_0: (zfs_rmnode+0x272/0x360 [zfs] <- zfs_purgedir) z_iput-23904 [006] d... 10822.453755: p_zpl_evict_inode_0: (zpl_evict_inode+0x0/0x60 [zfs]) inode=0xffff88011aa90650 z_iput-23904 [006] d... 10822.453756: p_zfs_inactive_0: (zfs_inactive+0x0/0x270 [zfs]) inode=0xffff88011aa90650 z_iput-23904 [006] d... 10822.453757: p_zfs_zinactive_0: (zfs_zinactive+0x0/0xe0 [zfs]) znode=0xffff88011aa90440 obj=0xc z_iput-23904 [006] d... 10822.453759: p_zfs_rmnode_0: (zfs_rmnode+0x0/0x360 [zfs]) znode=0xffff88011aa90440 $ echo 2 | sudo tee /proc/sys/vm/drop_caches 2 <...>-23945 [005] d... 10836.555639: p_dispose_list_0: (dispose_list+0x0/0x50) <...>-23945 [005] d... 10836.555775: p_dispose_list_0: (dispose_list+0x0/0x50) <...>-23945 [005] d... 10836.560865: p_dispose_list_0: (dispose_list+0x0/0x50) <...>-23945 [005] d... 10836.560993: p_dispose_list_0: (dispose_list+0x0/0x50) <...>-23945 [005] d... 10836.567297: p_dispose_list_0: (dispose_list+0x0/0x50) <...>-23945 [005] d... 10836.567371: p_dispose_list_0: (dispose_list+0x0/0x50) -- You received this bug notification because you are a member of Kernel Packages, which is subscribed to zfs-linux in Ubuntu. https://bugs.launchpad.net/bugs/1839521 Title: Xenial: ZFS deadlock in shrinker path with xattrs Status in zfs-linux package in Ubuntu: Invalid Status in zfs-linux source package in Xenial: In Progress Status in zfs-linux source package in Bionic: Invalid Status in zfs-linux source package in Disco: Invalid Status in zfs-linux source package in Eoan: Invalid Bug description: [Impact] * Xenial's ZFS can deadlock in the memory shrinker path after removing files with extended attributes (xattr). * Extended attributes are enabled by default, but are _not_ used by default, which reduces the likelyhood. * It's very difficult/rare to reproduce this problem, due to file/xattr/remove/shrinker/lru order/timing circumstances required. (weeks for a reporter user) but a synthetic test-case has been found for tests. [Test Case] * A synthetic reproducer is available for this LP, with a few steps to touch/setfattr/rm/drop_caches plus a kernel module to massage the disposal list. * In the original ZFS module: the xattr dir inode is not purged immediately on file removal, but possibly purged _two_ shrinker invocations later. This allows for other thread started before file remove to call zfs_zget() on the xattr child inode and iput() it, so it makes to the same disposal list as the xattr dir inode. * In the modified ZFS module: the xattr dir inode is purged immediately on file removal not possibly later on shrinker invocation, so the problem window above doesn't exist anymore. [Regression Potential] * Low. The patches are confined to extended attributes in ZFS, specifically node removal/purge, and another change how an xattr child inode tracks its xattr dir (parent) inode, so that it can be purged immediately on removal. * The ZFS test-suite has been run on original/modified zfs-dkms package/kernel modules, with no regressions. To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/zfs-linux/+bug/1839521/+subscriptions -- Mailing list: https://launchpad.net/~kernel-packages Post to : kernel-packages@lists.launchpad.net Unsubscribe : https://launchpad.net/~kernel-packages More help : https://help.launchpad.net/ListHelp