Why does JBD2 store its structures in Big Endian format ?
 

I am going through the JBD2 in linux. I am wondering why are the data structures in JBD stored in big endian format ? Is there some reason or just a matter of choice. I see this as a performance loss as every time there needs to be conversion before assignment.

Most of these deals come from some person. That person may have started the work on some other system and OS or even under some different hardware. So they just continued to linux with their work.

I guess there could be some performance or other issue. It doesn't seem to be directly or easily found as to why that was done.