[PATCH] avoid unnnecessary mtd read when read can be satisfied by write buffer

[Kevin Vigor]

ghannon at cspi.com wrote:

>However when programming a flash device and then reading back
>to verify that it was programmed correctly,  I think this write
>cache would defeat the purpose.
>
>  
>
This problem already exists with the wbuf code since a call to 
jffs2_flash_read will return the write buffer contents regardless of the 
actual contents of the media (unless you call  jffs2_flush_wbuf_pad() 
between the write and the read).

So the semantics could remain the same in that an explicit flush would 
be necessary in order to validate the data on flash.