Conversation
sync() writes dirty data back, but it does not leave the root filesystem clean: a journalling filesystem only commits its superblock on remount or unmount. On ext4 this means every boot replays the journal and e2fsck finds the free block and inode counts stale, even after an orderly reboot. The shutdown scripts cannot fix this. "umount -a -r" does attempt the read-only remount, but it runs while procd and everything it supervises are still alive, and the remount fails with EBUSY as long as any process holds a file open for writing. The only point at which the root is idle is after the final SIGKILL, so remount it read-only there, right before the kernel is asked to reboot. A remount of an overlay root only syncs its upper filesystem, and a swap file on the root filesystem still keeps it busy; both are left as they are. The root of a container belongs to the host and is skipped. logd is stopped long before this point, so anything logged from STATE_HALT went to a syslog socket nobody reads. Switch to the console that set_console() already prepared, so that a failed remount can be seen. Signed-off-by: John Audia <therealgraysky@proton.me>
Author
|
Closing/superseded by #44 |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
sync() writes dirty data back, but it does not leave the root filesystem clean: a journalling filesystem only commits its superblock on remount or unmount. On ext4 this means every boot replays the journal and e2fsck finds the free block and inode counts stale, even after an orderly reboot.
The shutdown scripts cannot fix this. "umount -a -r" does attempt the read-only remount, but it runs while procd and everything it supervises are still alive, and the remount fails with EBUSY as long as any process holds a file open for writing. The only point at which the root is idle is after the final SIGKILL, so remount it read-only there, right before the kernel is asked to reboot.
A remount of an overlay root only syncs its upper filesystem, and a swap file on the root filesystem still keeps it busy; both are left as they are. The root of a container belongs to the host and is skipped.
logd is stopped long before this point, so anything logged from STATE_HALT went to a syslog socket nobody reads. Switch to the console that set_console() already prepared, so that a failed remount can be seen.