https://bugzilla.kernel.org/show_bug.cgi?id=217965 Eduard Kohler (glandvador@xxxxxxxxx) changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |glandvador@xxxxxxxxx --- Comment #15 from Eduard Kohler (glandvador@xxxxxxxxx) --- I encounter the same issue on a stock fedora 37 kernel # uname -a Linux xxx.net 6.4.16-100.fc37.x86_64 #1 SMP PREEMPT_DYNAMIC Wed Sep 13 18:22:38 UTC 2023 x86_64 x86_64 x86_64 GNU/Linux The issue seems to be present also on brtfs partitions, as mentioned in https://bugzilla.redhat.com/show_bug.cgi?id=2242391 In my case, I have an EXT4 partition over an mdadm raid 1 array of two HDD. # mdadm --misc --detail /dev/md0 /dev/md0: Version : 1.2 Creation Time : Sun Feb 24 21:40:33 2013 Raid Level : raid1 Array Size : 3906118928 (3.64 TiB 4.00 TB) Used Dev Size : 3906118928 (3.64 TiB 4.00 TB) Raid Devices : 2 Total Devices : 2 Persistence : Superblock is persistent Update Time : Tue Oct 17 12:05:22 2023 State : clean Active Devices : 2 Working Devices : 2 Failed Devices : 0 Spare Devices : 0 Consistency Policy : resync Name : xxx.net:0 (local to host xxx.net) UUID : 25070e74:b2ac1695:46ee20f9:7e8d1e05 Events : 402440 Number Major Minor RaidDevice State 3 8 17 0 active sync /dev/sdb1 2 8 33 1 active sync /dev/sdc1 dumpe2fs -h /dev/md0p1 dumpe2fs 1.46.5 (30-Dec-2021) Filesystem volume name: <none> Last mounted on: /home Filesystem UUID: 4df09a03-d8ef-4cef-a64d-cefde4823fc2 Filesystem magic number: 0xEF53 Filesystem revision #: 1 (dynamic) Filesystem features: has_journal ext_attr resize_inode dir_index filetype needs_recovery extent flex_bg sparse_super large_file huge_file uninit_bg dir_nlink extra_isize Filesystem flags: signed_directory_hash Default mount options: user_xattr acl Filesystem state: clean Errors behavior: Continue Filesystem OS type: Linux Inode count: 244137984 Block count: 976529232 Reserved block count: 48826457 Overhead clusters: 15373546 Free blocks: 545717412 Free inodes: 243040909 First block: 0 Block size: 4096 Fragment size: 4096 Reserved GDT blocks: 791 Blocks per group: 32768 Fragments per group: 32768 Inodes per group: 8192 Inode blocks per group: 512 RAID stride: 32745 Flex block group size: 16 Filesystem created: Sun Feb 24 22:13:52 2013 Last mount time: Mon Oct 16 21:06:57 2023 Last write time: Mon Oct 16 21:06:57 2023 Mount count: 6 Maximum mount count: -1 Last checked: Sat Oct 7 16:27:02 2023 Check interval: 0 (<none>) Lifetime writes: 238 GB Reserved blocks uid: 0 (user root) Reserved blocks gid: 0 (group root) First inode: 11 Inode size: 256 Required extra isize: 28 Desired extra isize: 28 Journal inode: 8 First orphan inode: 13509593 Default directory hash: half_md4 Directory Hash Seed: af37e79f-0457-4318-9c5d-840dc2f60bce Journal backup: inode blocks Journal features: journal_incompat_revoke Total journal size: 128M Total journal blocks: 32768 Max transaction length: 32768 Fast commit length: 0 Journal sequence: 0x003bc716 Journal start: 1 # df /dev/md0p1 Filesystem 1K-blocks Used Available Use% Mounted on /dev/md0p1 3844622744 1661757340 1987543192 46% /home # mount |grep md0 /dev/md0p1 on /home type ext4 (rw,relatime,seclabel,stripe=32745) In order to trig the 100% CPU on one core, a simple # for i in {0001..0200}; echo "some text" > "file_${i}.txt" is enough. This will bring the kworker/flush at 100 % for around 90 seconds. This time is dependent of the number of created files. Trying to delete these files is more or less impossible. On the same setup I have a small SSD used for system installation. The same test doesn't bring the kworker/flush behaviour. # dumpe2fs -h /dev/sda3 dumpe2fs 1.46.5 (30-Dec-2021) Filesystem volume name: <none> Last mounted on: / Filesystem UUID: 589968d6-2283-4903-8699-1b23f5c341a5 Filesystem magic number: 0xEF53 Filesystem revision #: 1 (dynamic) Filesystem features: has_journal ext_attr resize_inode dir_index filetype needs_recovery extent flex_bg sparse_super large_file huge_file uninit_bg dir_nlink extra_isize Filesystem flags: signed_directory_hash Default mount options: user_xattr acl Filesystem state: clean Errors behavior: Continue Filesystem OS type: Linux Inode count: 848640 Block count: 3390208 Reserved block count: 169510 Free blocks: 1823000 Free inodes: 735742 First block: 0 Block size: 4096 Fragment size: 4096 Reserved GDT blocks: 827 Blocks per group: 32768 Fragments per group: 32768 Inodes per group: 8160 Inode blocks per group: 510 Flex block group size: 16 Filesystem created: Sun Apr 6 00:00:12 2014 Last mount time: Mon Oct 16 21:06:53 2023 Last write time: Mon Oct 16 21:06:47 2023 Mount count: 437 Maximum mount count: -1 Last checked: Sun Apr 6 00:00:12 2014 Check interval: 0 (<none>) Lifetime writes: 1791 GB Reserved blocks uid: 0 (user root) Reserved blocks gid: 0 (group root) First inode: 11 Inode size: 256 Required extra isize: 28 Desired extra isize: 28 Journal inode: 8 First orphan inode: 130778 Default directory hash: half_md4 Directory Hash Seed: 7d17214b-9585-4370-8c06-4236e449fa7f Journal backup: inode blocks Journal features: journal_incompat_revoke Total journal size: 128M Total journal blocks: 32768 Max transaction length: 32768 Fast commit length: 0 Journal sequence: 0x0114205c Journal start: 8194 # mount |grep sda3 /dev/sda3 on / type ext4 (rw,noatime,seclabel) -- You may reply to this email to add a comment. You are receiving this mail because: You are watching the assignee of the bug.
