inode incorrect and can't be repaired.

[zhoucheng361]

What happened:

root@bench-02:~/jfs# ./juicefs format sqlite3://load.db test-volume --enable-acl --trash-days 0 --storage minio --bucket http://localhost:9000/test --access-key minioadmin --secret-key minioadmin
root@bench-02:~/jfs# ./juicefs mount sqlite3://load.db /tmp/jfs --enable-xattr -d --cache-dir memory
root@bench-02:~/jfs#./random-test runOp -baseDir /tmp/jfs/test -files 500000 -ops 5000000 -threads 50 -dirSize 10 -duration 30s -createOp 30,uniform -deleteOp 5,end --linkOp 10,uniform --removeLinkOp 1,end --symlinkOp 20,uniform --numSymlinks 5000 --setXattrOp 10,uniform --truncateOp 10,uniform  
root@bench-02:~/jfs# df -i 
Filesystem            Inodes IUsed    IFree IUse% Mounted on
JuiceFS:test-volume 10501023 15263 10485760    1% /tmp/jfs

root@bench-02:~/jfs# ./juicefs dump sqlite3://load.db dump.json
root@bench-02:~/jfs# grep inode dump.json  | wc -l
15278

inode incorrect after repair:
root@bench-02:~/jfs# ./juicefs fsck  --sync-dir-stat sqlite3://load.db --path / -r --repair
root@bench-02:~/jfs# df -i 
Filesystem            Inodes IUsed    IFree IUse% Mounted on
JuiceFS:test-volume 10501023 15263 10485760    1% /tmp/jfs

What you expected to happen:

How to reproduce it (as minimally and precisely as possible):

Anything else we need to know?

Environment:

• JuiceFS version (use juicefs --version) or Hadoop Java SDK version:
• Cloud provider or hardware configuration running JuiceFS:
• OS (e.g cat /etc/os-release):
• Kernel (e.g. uname -a):
• Object storage (cloud provider and region, or self maintained):
• Metadata engine info (version, cloud provider managed or self maintained):
• Network connectivity (JuiceFS to metadata engine, JuiceFS to object storage):
• Others:

[jiefenghuang]

没重现出来，用random-test生成，dump之后多出来的数量是刚好是硬链接，不计入inode