https://bugzilla.kernel.org/show_bug.cgi?id=201685 Bug ID: 201685 Summary: ext4 file system corruption Product: File System Version: 2.5 Kernel Version: maybe one of 4.18.18 4.19.1 4.20-rc2 Hardware: All OS: Linux Tree: Mainline Status: NEW Severity: normal Priority: P1 Component: ext4 Assignee: fs_ext4@xxxxxxxxxxxxxxxxxxxx Reporter: claude@xxxxxxxxxxx Regression: No Created attachment 279431 --> https://bugzilla.kernel.org/attachment.cgi?id=279431&action=edit dmesg 4.18.18 amdgpu.dc=0 My system was fine when I shut it down on Sunday Nov 11. Today Nov 13 I booted 4.19.1, built two new kernels 4.20-rc2 4.18.18 (using a tmpfs, not the SSD or HDD), then booted into those kernels briefly (to test if a different bug had been fixed). Finally I booted into 4.18.18 (setting amdgpu.dc=0 to workaround my other bug), and after some moments experienced symptoms of filesystem corruption on opening an xterm: sed: error while loading shared libraries: /lib/x86_64-linux-gnu/libattr.so.1: unexpected PLT reloc type 0x00000107 sed: error while loading shared libraries: /lib/x86_64-linux-gnu/libattr.so.1: unexpected PLT reloc type 0x00000107 claude@eiskaffee:~$ I fixed it by extracting the relevant file from the Debian archive on a different machine and using `cat` with `bash` shell IO redirection to overwrite the corrupted shared library file on my problem machine. Here are the relevant versions extracted from my syslog: Nov 13 15:45:49 eiskaffee kernel: [ 0.000000] Linux version 4.19.1 (claude@eiskaffee) (gcc version 8.2.0 (Debian 8.2.0-9)) #1 SMP Tue Nov 6 14:58:04 GMT 2018 Nov 13 18:44:12 eiskaffee kernel: [ 0.000000] Linux version 4.20.0-rc2 (claude@eiskaffee) (gcc version 8.2.0 (Debian 8.2.0-9)) #1 SMP Tue Nov 13 16:38:55 GMT 2018 Nov 13 18:45:00 eiskaffee kernel: [ 0.000000] Linux version 4.18.18 (claude@eiskaffee) (gcc version 8.2.0 (Debian 8.2.0-9)) #1 SMP Tue Nov 13 16:23:11 GMT 2018 Nov 13 18:46:13 eiskaffee kernel: [ 0.000000] Linux version 4.18.18 (claude@eiskaffee) (gcc version 8.2.0 (Debian 8.2.0-9)) #1 SMP Tue Nov 13 16:23:11 GMT 2018 mount says: /dev/nvme0n1p2 on / type ext4 (rw,relatime,errors=remount-ro) The machine in question is my production workstation, so I don't feel like testing anything that might result in data loss. -- You are receiving this mail because: You are watching the assignee of the bug.
