[LEDE-DEV] NAND JFFS2 question

Thu Mar 9 02:01:45 PST 2017

Hi!

The UBI is working right - indeed. Thanks for the hint of the bad block table.
I'll have a look at it. The background for this is the update routine [0], which does
the following:

-----------------------
      memset(&opts, 0, sizeof(opts));
          opts.offset = destAddr;
          opts.length = rootfs_size;
          opts.jffs2  = 1;
          opts.quiet  = 0;
          printf("erasing 0xx size 0xxn",destAddr, rootfs_size);
ret = nand_erase_opts(&nand_info[0], &opts);

 .. and then writes the image ...

ret = nand_write_skip_bad(&nand_info[0], destAddr, &srcLen64,
(u_char *)srcAddr);
-----------------------

With the current development state, I use a usb memory stick for the rootfs (formatted with f2fs).
Yesterday I tried the following:
- Built a image via "make".
- Then moved the lede-lantiq-xrx200-VGV952CJW33-ubifs-fullimage-ubinized.bin as fullimage.img to my tftp server.
- Switched off the Easybox
- Held the reset button while switched on the device.
- The update routine starts: It loads the fullimage.img from the server and updates the rootfs and kernel section.
  The rootfs section is crap (of course - this is the reason of my help mail). But at least the kernel boots and I point the rootfs to /dev/sda3 so LEDE starts booting.

My goal was to have people running this firmware without the need of a serial cable.

> Vitaly Chekryzhev send patch to add MDIO support to the rtl8367b used by
> the 904 [0]. This patch wasn't merged since it caused issues. My last
> update about this was that he is going to fix the issues and will send a
> new patch.

I'm also in touch with vitaly lately to test the current staging tree (was successful, but not incorporated in my tree yet).
I hope this merge would be soon active.

> Long story short, I got the following working:
> - LAN
> - ethernet WAN (the DSL port is ethernet and dsl at the same time)
> - LEDs
> - Buttons
> - Flash without bad block support
> - usb port on the back
> - ram boot u-boot for recovery

Thanks for your work, I if I found your work earlier I would be faster by port LEDE to the device =)
I also have some things done (in parallel of your work):
- LAN
- ethernet WAN (the DSL port is ethernet and dsl at the same time)
- LEDs
- usb port on the back (and the upper one)
- Telephony via Asterix (analogue ports ready)
- Display as framebuffer, see it in action (LCD4Linux) [1]
- Touchpad as input device

WLAN is really missing. No open source driver - not wlan. But it is no use case for me, so I resigned to it.

Here's my work [2].

[0] https://github.com/uwehermann/easybox-904-xdsl-firmware/blob/master/package/infineon-utilities/feeds/ifx_feeds_uboot/open_uboot/src.904dsl/common/upgrade.c#L130
[1] http://picpaste.de/pics/EB904.1488320671.JPG
[2] https://github.com/Quallenauge/Easybox-904-XDSL

Best regards,
Quallenauge

> -----Ursprüngliche Nachricht-----
> Von: Mathias Kresin
> Gesendet: Do. 09.03.2017 08:34
> An: hamsi2k at freenet.de, lede-dev at lists.infradead.org
> Betreff: Re: [LEDE-DEV] NAND JFFS2 question
>
> 08.03.2017 23:01, hamsi2k at freenet.de:
> > Hi!
> >
> > I currently work to assembly a fullimage.img for the Easybox 904xdsl.
> > Actually there are some differences to upstream made by the vendor.
> > I take over some code and now I'm able to update the firmware, or at
> > least the kernel via the recovery method within the uboot.
> >
> > As I wrote, the kernel and also the rootfs is flashed without errors.
> > When I try to boot the image, or mount the partition it is not possible
> > due some strange ECC errors.
> >
> > So I did some investigations: When I boot into the target system
> > via sdcard as rootfs and then I perform a
> >
> > ------------------------------------------------------------------------
> > flash_eraseall /dev/mtd1
> > nand_write /dev/mtd1 /root/image.jffs2
> >
> > mkdir /tmp/disk
> > mount -t jffs2 /dev/mtdblock1 /tmp/disk
> > ------------------------------------------------------------------------
>
>
> Using jffs2 on NAND flash isn't the best idea. jffs2 doesn't work that
> good with NAND flash. Use ubi instead!
>
> I worked on the Easybox 904xdsl as well but stopped after realising that:
>
> - Arcadian decided to use their own bad block table patterns instead of
> the ones which are used by the kernel and an unmodified u-boots. Means a
> kernel patch is required just for this board.
>
> - to support the wireless a complete protocol needs to be reverse
> engineered and a lot of missing code needs to be added to the rt2x00 driver
>
> The rt3883 wireless chip of the Easybox 904xdsl is not a usual wireless
> chip, it is a full SoC which is supported as own suptarget in LEDE
> (ramips/rt3883). In case of the 904 a complete realtime operating system
> is uploaded/runs on the rt3883 instead of a "normal" Operating System
> like OpenWrt/LEDE. Since it is a full SoC it has subsystems like PCI and
> so on. The rt5392 wireless is attached via PCI to the rt3883 SoC.
>
> The internal ethernet of the rt3883 SoC is connected to the internal
> lantiq switch via MDIO/RGMII. The whole communication and configuration
> of the rt3883/rt5392 is done via a proprietary protocol, which is based
> on ethernet frames. The way the protocol really works is unknown and
> there is no support for that protocol in the rt2x00 wireless driver.
>
> Vitaly Chekryzhev send patch to add MDIO support to the rtl8367b used by
> the 904 [0]. This patch wasn't merged since it caused issues. My last
> update about this was that he is going to fix the issues and will send a
> new patch.
>
> Long story short, I got the following working:
>
> - LAN
> - ethernet WAN (the DSL port is ethernet and dsl at the same time)
> - LEDs
> - Buttons
> - Flash without bad block support
> - usb port on the back
> - ram boot u-boot for recovery
>
> You can find my code at https://github.com/mkresin/lede/tree/904xdsl.
>
> Feel free to use it or to rebase your work on it. Would be nice if you
> publish what you have so far as well.
>
> Supporting the build in display should be possible. The driver for the
> display is in the staging section of the kernel for a while now [1].
>
> Mathias
>
> [0] https://github.com/lede-project/source/pull/537
> [1]
> https://git.kernel.org/cgit/linux/kernel/git/stable/linux-stable.git/tree/drivers/staging/fbtft?h=linux-4.4.y
>
>
> -----Ursprüngliche Nachricht Ende----- 

---
https://email.freenet.de/emig/index.html?utm_medium=Text&utm_source=Footersatz&utm_campaign=Footersatz_Sicherheit170207&epid=e9900000699&utm_content=Text Wir garantieren Ihnen verschlüsselte Datenübertragung & Datenspeicherung auf deutschen Servern - E-Mail made in Germany!