The Ben NanoNote is running OpenWrt! :)

Xiangfu Liu xiangfu at qi-hardware.com
Sat Aug 8 10:13:14 EDT 2009


Mirko Vogt wrote:
> Hello Qi!
>
> As you may have noticed - OpenWrt is running on the Ben NanoNote! :)
>   
great.
> Based on a newer patch[1] provided by ingenic (for kernel 2.6.24.4) we
> ported the patch to kernel 2.6.25.20, cleaned it up, cleared unneeded
> stuff out created a little list what needs to be done on kernel level.
>
> The userspace of OpenWrt is running fine, I'll play around with Xorg
> today but 32MB of RAM is a real limitation.
>
> During reading the code / testing / disassembling the device some
> questions raised about the installed hardware:
>
>  - which SoC is the Ben NanoNote using?
> Regarding the code it's very similar to he Pavo / Leo / Lyra boards
> which all have a JZ4740 built-in. But the schematics in the git of Adam
> are about a JZ4720 (which is listed as discontinued on the
> ingenic-website).
> The chip on the Ben itself seems to be "soused" with black
> gum/plastic[2] (btw., is that necessary?).
>
>  - what display is used within the Ben NanoNote?
> We experienced some discrepancy in the qi_lb60-code about the used
> display / displaydriver (Foxconn-PT035TN01 / standard 8-bit-serial-tft /
> sth. else?) While disassembling it, there are just chains of numbers
> written on LCM/LCD[3,4] - while searching for them we got to lcdsolar[5]
> - which is not really verbose.
> @Xiang Fu: Any news about this?
>   
I will update, when make progress on LCD.
> Looking at the turned-on display through a magnifier (even in u-boot),
> it seems the pixels within every other line are shifted. Any idea
> whether it's hardware/software?
>   
will try to make sure later.
>  - any news about the proprietary mtblock-jz?
> Xiang Fu told me they won't release the source of the object file within
> the original kernel-patch.
> While reading/writing to the NAND with the standard mtdblock-driver we
> experienced lot's of problems. It should not be too difficult getting a
> working MTD-driver done, but having their code might ease ours workload.
>   
we can try to talk with ingenic people.  :-)
>
> I'm currently in Hamburg - hacking together with Lars and John (also
> from OpenWrt) at the Ben NanoNote. 
>
> How to proceed? Our roadmap for the next days:
>
>  - level up the new patchset to 2.6.31-rc5
> there's is a new GPIO driven matrix keyboard driver within[6], which
> should make it pretty easy to get a keyboard driver done. An example
> implementation can be found here[7].
> @Xiang Fu: As far as I know you were interested in doing so?.
>   
yes. I have interesting, I will try.
> Also getting NAND-flashes > 4 GB are now nativly supported and do not
> need the ingenic hacks anymore - also many other things which then can
> be dropped out of the patch.
>
>  - (re)writing MMC-/SD-card- / GPIO- / IRQ- / and-maybe-others- /
> drivers
> In our ongoing effort to get the device support by Linux 2.6.31 and to
> cleanup the patchset we're properly going to (re)write some of the
> needed device drivers.
> This should hopefully purge out some issues, also that the kernel
> currently not always recognizes the SD-card while booting.
>   
sounds good.
>
> Have a nice weekend!
>
> Greetings from Hamburg!
>
> mirko
>
>
> [1]
> ftp://ftp.ingenic.cn/3sw/01linux/02kernel/linux-2.6.24/linux-2.6.24.3-jz-20090506.patch.gz
> [2] http://nanl.de/files/qi/ben_soc.jpg
> [3] http://nanl.de/files/qi/ben_lcd.jpg
> [4] http://nanl.de/files/qi/ben_lcm.jpg
> [5] http://www.lcdsolar.com/products/All/AUO/A030DN01V0.html
> [6] http://kerneltrap.org/mailarchive/git-commits-head/2009/7/8/6161203
> [7] http://www.spinics.net/lists/linux-input/msg04115.html
>
>
>   





More information about the discussion mailing list


interactive