svn commit: r184251 - in head/sys: conf dev/cfi sys

[Bruce M. Simpson]

Warner Losh wrote:
>> Thanks for bringing this code in. Hopefully it shouldn't be too hard to 
>> add MTD-like support on top, and/or port it to other parts.
>>     
>
> What, exactly, do you mean here?  The partitioning stuff, the flash
> erase stuff, moving all things through the /dev/mt* or something else?
>   

Well, everything... it's going to take a while to sort through all this 
without a commercial entity driving things (got bills to pay after all).

I guess we need to deal with stuff like block erase at the BIO layer.

I know that in previous discussions it has been by suggested by others 
that GEOM isn't the right place for flash partitioning to be handled, 
I'm inclined to disagree, as it takes care of a lot of the rigmarole of 
creating device nodes for each slice, etc, but whoever gets it done, 
wins the argument, really.