Bug 215879
Summary: | EXT4-fs error - __ext4_find_entry:1612: inode #2: comm systemd: reading directory lblock 0 | ||
---|---|---|---|
Product: | File System | Reporter: | sander44 (ionut_n2001) |
Component: | ext4 | Assignee: | fs_ext4 (fs_ext4) |
Status: | RESOLVED INVALID | ||
Severity: | high | CC: | tytso, vladimir |
Priority: | P1 | ||
Hardware: | x86-64 | ||
OS: | Linux | ||
Kernel Version: | 5.18.0-rc3 | Subsystem: | |
Regression: | No | Bisected commit-id: | |
Attachments: |
config kernel
photo bug |
Description
sander44
2022-04-24 15:20:35 UTC
Created attachment 300795 [details]
config kernel
Created attachment 300796 [details]
photo bug
We should really improve the error message; what this indicates is that while trying to read from a directory, ext4 received an I/O error from the storage device: wait_on_buffer(bh); if (!buffer_uptodate(bh)) { EXT4_ERROR_INODE_ERR(dir, EIO, "reading directory lblock %lu", (unsigned long) block); brelse(bh); ret = ERR_PTR(-EIO); goto cleanup_and_exit; } I'm not sure why systemd is trying to read so many directories, and why you aren't seeing any I/O error messages logged on the console. At a guess, the block device has completely failed, and there were messages about the unfolding disaster that has since scrolled off your screen. Bottom line, this is very likely an hardware problem of some sort. I am getting the same error as sander44 (the OP) reported except the name of the program generating the error is different. EXT4-fs error (device sdb1): __ext4_find_entry:1612: inode #2: comm vorta: reading directory lblock 0 The behavior of the Samsung 1TB SDXC card is predictable. It works for minutes/hour, then throws the above error. This particular SD card is used for borg backup which occur every 15 minutes (and also every 3 hours). My (temporary) workaround is to unmount the SD card, run e2fsck, then re-mount the card. I'm suspicious of this being exclusively a hardware error because it is always the same error at the same spot. Might it be an unexpected call or unexpected function arguments? (I have no knowledge of ext4 internals, so it's likely I'm talking nonsense.) I'm happy to provide more detail or even roll my own kernel to provide debugging symbols. (I'm using Arch 5.15.43-1-lts, but the error is independent of what kernel I'm using.) — Vladimir Hi, I notice today this issue with 6.1.12 kernel version. On Ubuntu Team, i view this: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1993205 But with my search i view workaround: https://unix.stackexchange.com/questions/375450/random-ssd-turn-off-ext4-find-entry-reading-directory-lblock0 https://askubuntu.com/questions/905710/ext4-fs-error-after-ubuntu-17-04-upgrade Will try with this: nvme_core.default_ps_max_latency_us=200 for testing this issue and to see if it reproduces. Note that what you found in your stack exchange search was from five years ago, and described a workaround in a Linux kernel versiojn 4.10. In addition to manually disabling APST (a quirk for a very specific Samsung SSD which has since been added to newer kernels), other suggestions in the stack exchange or linked web pages included " removing SDD, blowing air into M.2 connector and reinserting it back" and "switching off the 'UEFI Secure Boot' setting in the BIOS" All of which is to say that the symptom is caused by an I/O error, and there are many potential causes for an I/O error --- everything from missing quirks (to work around broken firmware / hardware design) to bad connections to misconfigured BIOS settings to just plain broken hardware. This is why blindly web searching based on symptoms can often lead to misleading results; an abdominal pain could mean anything from indigestion, to a pulled muscle, to an infected appendix, to a heart attack. It's also why I am not fond of people finding bug reports on the web and assuming that anything that has the same symptom must have the same root cause..... |