Gabriel de Perthuis
Gabriel de Perthuis
This is where going read-write is coming from: ``` Thread 1 "bcachefs" hit Breakpoint 3, bch2_fs_read_write_early (c=0x7ffff7cfa000) at libbcachefs/super.c:523 523 { (gdb) bt #0 bch2_fs_read_write_early (c=0x7ffff7cfa000) at libbcachefs/super.c:523 #1 0x0000555555742825...
Somehow `make debug` (with CFLAGS+=-DCONFIG_BCACHEFS_DEBUG=y) produced a command that does dump.
Decompression is done in memory, so maybe it is better to keep a length limit. I've reformatted with black and added a test case. It is kind of a drive-by...
It will blow up, as it did in some cases before the patch (systemd coredump on oracular). The problem is that callers aren't expecting errors, for example when going through...
Here is some list_journal output, after I got this message in dmesg: ``` 2025-11-26T11:14:47.546105+00:00 moulinex kernel: bcachefs (2c9b10da-e32f-44ab-b303-d2cd1005acf2): inode 36894648 i_sectors underflow: 216 + -224 < 0 2025-11-26T11:14:47.546307+00:00 moulinex kernel:...
Here is a journal dump 8 minutes after the log message: https://sebsauvage.net/paste/?902c87fec9c65720#eG8SLjA7pgawxqece6KrWQ0hnR86ObOt2/UkdAGGqFA= ``` 2025-11-28T14:01:00.438102+00:00 moulinex kernel: bcachefs (2c9b10da-e32f-44ab-b303-d2cd1005acf2): inode 36894648 i_sectors underflow: 272 + -280 < 0 2025-11-28T14:01:00.456635+00:00 moulinex kernel:...
Attempting to summarize from IRC: the fs has metadata_replicas=2 (metadata_replicas_required=1), my intent is to have one replica on ssd and one on hdd. There is tiering: foreground_target=ssd, background_target=hdd, promote_target=ssd, but...
> nothing looks off here, possibly space being taken up by snapshots you said - so we can't do anything with this until we have per-snapshot accounting exposed I've deleted...
Oh, should have added the important part: ``` LC_ALL=C df ~ Filesystem 1K-blocks Used Available Use% Mounted on /dev/Consolidated/Home:/dev/Consolidated/HomeCache 3708243148 3708243148 0 100% /home/g2p ``` which is completely unchanging, despite...
`/sys/kernel/debug/bcachefs/2c9b10da-e32f-44ab-b303-d2cd1005acf2/btrees/snapshots/keys` has 3200 lines which seems high, 2314 grep for delete, 993 for no_keys, 123 match neither. The number I would expect is `ls ~/.snapshots |wc -l` which is 65....