Hi wread,
I give you an example:
Code: Select all
# mkdir -p binder/usr binder/var
# touch binder/usr/file1.txt
# touch binder/var/file2.txt
So the contents of the binder directory is:
Code: Select all
[root:~]# ls -R binder
binder:
usr var
binder/usr:
file1.txt
binder/var:
file2.txt
[root:~]#
Now, I duplicate the binder directory as:
Now, if I create modules from these 2 directories:
Code: Select all
dir2xzm binder binder.xzm
dir2xzm grinder grinder.xzm
This is the content of the 2 modules:
Code: Select all
[root:~]# unsquashfs -l binder.xzm
Parallel unsquashfs: Using 8 processors
2 inodes (0 blocks) to write
squashfs-root
squashfs-root/binder
squashfs-root/binder/usr
squashfs-root/binder/usr/file1.txt
squashfs-root/binder/var
squashfs-root/binder/var/file2.txt
[root:~]# unsquashfs -l grinder.xzm
Parallel unsquashfs: Using 8 processors
2 inodes (0 blocks) to write
squashfs-root
squashfs-root/usr
squashfs-root/usr/file1.txt
squashfs-root/var
squashfs-root/var/file2.txt
You can see that the top level directory for binder.xzm is binder (and then usr and var as subdirectories), while for the grinder.xzm there is no top level grinder directory. This is because, "binder" has the substring "bin", a top level filesystem directory in the beginning.
So, I guess this should also be the same for directories like "variable*", "libreoffice*" etc.