[PATCH 2/2] ath79: mojo c-75: use second flash chip

[Tomasz Maciej Nowak]

W dniu 14.12.2020 o 17:30, Adrian Schmutzler pisze:
> Hi,
> 
>> -----Original Message-----
>> From: Tomasz Maciej Nowak [mailto:tmn505 at gmail.com]
>> Sent: Dienstag, 8. Dezember 2020 18:05
>> To: Adrian Schmutzler <mail at adrianschmutzler.de>; openwrt-
>> devel at lists.openwrt.org
>> Cc: 'Vladimir Georgievsky' <vladimir.georgievsky at yahoo.com>
>> Subject: Re: [PATCH 2/2] ath79: mojo c-75: use second flash chip
>>
>> W dniu 01.12.2020 o 20:50, Adrian Schmutzler pisze:
>>>> diff --git a/target/linux/ath79/image/generic.mk
>>>> b/target/linux/ath79/image/generic.mk
>>>> index 0f90c00b9c43..b7618a8ab33c 100644
>>>> --- a/target/linux/ath79/image/generic.mk
>>>> +++ b/target/linux/ath79/image/generic.mk
>>>> @@ -1199,7 +1199,7 @@ define Device/mojo_c-75
>>>>    DEVICE_ALT0_VENDOR := WatchGuard
>>>>    DEVICE_ALT0_MODEL := AP320
>>>>    DEVICE_PACKAGES := ath10k-firmware-qca988x kmod-ath10k-ct kmod-
>>>> usb2
>>>> -  IMAGE_SIZE := 15936k
>>>> +  IMAGE_SIZE := 32320k
>>>
>>> I'm not sure whether that's possible or whether the image itself is
>> restricted to a single consecutive area to be written to (i.e. the flash can be
>> concatenated and used, but the _image_ is limited to the first partition _in_
>> it)
>>
>> The mtd-concat creates virtual mtd device, which whole space we name as
>> 'firmware' partiton, and for kernel it's completely transparent, as for mtd-
>> splitter and mtd command.
> 
> That might hold for OpenWrt-internal sysupgrade (I'm actually not sure here), but will it also be true for initial flashing?

That's why the installation is done from initramfs image, where everything is properly set-up.
Thinking about it more, the U-Boot might be not capable of loading kernel image spanned through two flash chips, so I'll add KERNEL_SIZE variable.

> 
> Best
> 
> Adrian
> 

Regards.

-- 
TMN