Olmari from irc tested it, but I don't remember, which device he had.
But I don't see, how this could be a problem. The bootloader only
calculates the crc over trx->len, which gets also fixed up. I don't
think he runs any additional checks or uses the value in some way
The only place I know of, where len actually matters is during
flashing with the original firmware, which complains if the filesize
and the trx len values are different.
Bernhard
2010/5/3 Felix Fietkau <[email protected]>:
> On 2010-05-03 1:27 AM, Bernhard Loos wrote:
>> Ok, I did split the patch in two parts.
>> The first one introduces libmtd.c and moves the mtd open/close
>> functions there and the second one contains the rest of the rework.
>> The second patch is still rather big, but it's not so easy to separate
>> it further, so I would like to avoid it, if possible.
>> This also fixes a small problem on the wrt160nl.
> Haven't reviewed all of it yet, but I did notice something in the first
> patch. You add the following line to trx_fixup():
> + trx->len = block->info.erasesize;
>
> Wouldn't this break broadcom? Where has this been tested?
>
> - Felix
>
_______________________________________________
openwrt-devel mailing list
[email protected]
https://lists.openwrt.org/mailman/listinfo/openwrt-devel
