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

[David Woodhouse]

On Thu, 2006-06-01 at 16:49 +0200, Jörn Engel wrote:
> One of the observations was that a lot of flash is being read multiple
> times.  Reads on NAND happen in page granularity, so first we have to
> read pretty much everything during scan, later again when accessing
> the actual files.  There is some room for savings, as RAM is faster
> than flash even in the real world. 

My point is that for any given problem, raw caching of flash is not the
answer -- something more cunning is going to be better.

The specific cases of dirents and symlinks I already talked about. Any
other time we keep reading the same bit of flash over and over again is
also a problem. The one you mention is fixed by using eraseblock
summary.

I need to finish what I was doing yesterday, which halves the amount of
space taken by summaries. We should probably also remove the dependency
on EXPERIMENTAL.

-- 
dwmw2