Questions about Built-in Storage file system
Not sure what's the best place to ask this, let's try here.
I have a linux PDA where all design decisions are open to see and well documented. In the following I'm only interested in file systems for flash storage. The file system used on the Linux PDA (jffs2) is a journalling file system (cannot get corrupted), does wear levelling (does not wear out flash storage by repeatedly writing to the same position) and uses compression (saves space, reduces flash wear, is usually (for slow flash and fast cpu) faster than uncompressed storage).
I cannot find any comparable information about these points for the Built-in Storage (or similar) on Windows Mobile devices. I would really appreciate answers to the following questions:
- Which file system is used?
- Is it journalling? Can it get corrupted? If so, how to fix it?
- Does it do wear levelling?
- Does the wear levelling still work for almost full file systems?
- The wear levelling might also be implemented at a lower hardware level, is that the case?
- Does the file system (or some lower layer) use compression? (It appears that it does not, which would be a severe design flaw. Has anything changed for WM5 where this would be even more important?)
- If it does not use compression, would the use of third-party utilities which promise compression be a good idea? (Obviously yes, if these utilities could be trusted never to corrupt the data, but can they?)
Concerning the ramdisk Main storage, it does not need wear levelling and it does use compression, but I always wondered:
- Which file system is used?
- Is it journalling? Can it get corrupted? If so, how to fix it?
Thanks in advance for any helpful insights.
|