Re: single device to replace ADSL router, WiFi/Ethernet router, SIP router?

[Luke Kenneth Casson Leighton <lkcl@lkcl.net>]

On Thu, Dec 22, 2011 at 8:01 PM, Bill Gatliff <bgat@billgatliff.com> wrote:

> In addition, I really, really hate the quality levels I am seeing with
> uSD/eMMC devices right now.  I know they are all internally based on
> NAND, why not ditch the little microcontroller they must also include
> and talk to a NAND directly?

 because if you're using the NAND for regular writes, you know that
it's going to break at some point in the future.

> Finally, obviously what you are calling the "bootloader" in your board
> is merely a second-stage one, as the CPU almost certainly does not
> natively know how to read from uSD.

 ah that's ... ok, the TI OMAPs datasheets talk about "starting up"
from a memory location and running an on-board program, implication
being somehow that that program is modifiable, and that program _does_
have enough logic to detect certain GPIO pins high or low, and decide
on that basis to start directly from a pre-programmed memory address
(published in the datasheet) or a couple of other options _or_ whether
to read from SD/MMC a fixed amount of "stuff" into RAM and then
execute that.

 on enquiring with TI they said "actually that lot is in a
pre-programmed ROM that's hard-coded into the masks, so yes sure you
could get that reprogrammed but it'd cost you a set of mask charges".

 then there's the S3C6410, S5PC1xx series etc. which play the same
trick: again, an on-board ROM is capable of loading up from a wide
range of options.

 so yes, you're right: the CPU *itself* doesn't know - it starts up
and executes from a fixed address, but depending on the CPU, the
[unchangeable] on-board micro-bootloader does "know".

 l.