Hello
I'm sorry for the delay.
> Are you able to reliably preoeduce the issue and can bisect it to the
> introducing commit?
I faced this issue on real data but I struggled to find a reliable scenario to
reproduce it. Here is what I just came up with:
sudo mkfs -t ext4 -O fast_commit,inline_data /dev/sdb
sudo mount /dev/sdb /mnt/
sudo install -d -o myuser /mnt/annex
cd /mnt/annex
git init && git annex init
for i in {1..2}; do
for i in {1..10000}; do
dd if=/dev/urandom of=file-${i} bs=1K count=1 2>/dev/null
done
git annex add -J cpus . >/dev/null && git annex sync -J cpus && git annex
fsck -J cpus >/dev/null
git rm * && git annex sync && git annex dropunused all
done
Then at some point the following error appears:
EXT4-fs error (device sdb): ext4_map_blocks:577: inode #3942343: block 4:
comm git-annex:w: lblock 1 mapped to illegal pblock 4 (length 1)
Some observations:
- Contrary to what I implied in my first message, using the option
inline_data alone doesn't seem to trigger the error.
- Working with less than 10K files doesn't seem to trigger the error
- Sometimes the commands just run fine that's why I run the whole thing twice
- I also tried with various file size, in particular 1B, but it does not seem
to have any impact
- The disk I was using is a 5400RPM HDD
I'm running the kernel 6.5.0-3-amd64.
I'm also getting this issue on Debian stable's kernel (6.1.0-13-amd64) so I'm
not aware of any working kernel version.
Regards
Hervé
