[MTD] Rework the out of band handling completely

http://lists.infradead.org/pipermail/linux-mtd-cvs/2006-May/005519.html
官方介绍有关oob这部分的知识，特别是MTD_OOB_RAW, MTD_OOB_PLACE, MTD_OOB_AUTO的含义。

    [MTD] Rework the out of band handling completely
   
    Hopefully the last iteration on this!
   
    The handling of out of band data on NAND was accompanied by tons of fruitless
    discussions and halfarsed patches to make it work for a particular
    problem. Sufficiently annoyed by I all those "I know it better" mails and the
    resonable amount of discarded "it solves my problem" patches, I finally decided
    to go for the big rework. After removing the _ecc variants of mtd read/write
    functions the solution to satisfy the various requirements was to refactor the
    read/write _oob functions in mtd.
   
    The major change is that read/write_oob now takes a pointer to an operation
    descriptor structure "struct mtd_oob_ops".instead of having a function with at
    least seven arguments.
   
    read/write_oob which should probably renamed to a more descriptive name, can do
    the following tasks:
   
    - read/write out of band data
    - read/write data content and out of band data
    - read/write raw data content and out of band data (ecc disabled)
   
    struct mtd_oob_ops has a mode field, which determines the oob handling mode.
   
    Aside of the MTD_OOB_RAW mode, which is intended to be especially for
    diagnostic purposes and some internal functions e.g. bad block table creation,
    the other two modes are for mtd clients:
   
    MTD_OOB_PLACE puts/gets the given oob data exactly to/from the place which is
    described by the ooboffs and ooblen fields of the mtd_oob_ops strcuture. It's
    up to the caller to make sure that the byte positions are not used by the ECC
    placement algorithms.
   
    MTD_OOB_AUTO puts/gets the given oob data automaticaly to/from the places in
    the out of band area which are described by the oobfree tuples in the ecclayout
    data structre which is associated to the devicee.
   
    The decision whether data plus oob or oob only handling is done depends on the
    setting of the datbuf member of the data structure. When datbuf == NULL then
    the internal read/write_oob functions are selected, otherwise the read/write
    data routines are invoked.
   
    Tested on a few platforms with all variants. Please be aware of possible
    regressions for your particular device / application scenario
   
    Disclaimer: Any whining will be ignored from those who just contributed "hot
    air blurb" and never sat down to tackle the underlying problem of the mess in
    the NAND driver grown over time and the big chunk of work to fix up the
    existing users. The problem was not the holiness of the existing MTD
    interfaces. The problems was the lack of time to go for the big overhaul. It's
    easy to add more mess to the existing one, but it takes alot of effort to go
    for a real solution.
   
    Improvements and bugfixes are welcome!