Multi-boot for 7000 - public beta

[noggie]

This is my multi-boot/miniroot solution for the 7000. I strongly believe that images built on OpenEmbedded technology have a brighter long-term future than the traditional build environment for the the 7000. The primary purpose of this software is to enable 7000 owners to run 7020 images. I've personally used internal versions of this software to run OE-based images for the 7020 since before the 7020 was officially released. I store my images on USB sticks (yes, I have two of them), and that's obviously the storage medium that has received the most testing.

Based on kernel and "head.ko" from latest official DMM release (2.6.9, FP 1.05). Other binaries has also been "stolen" from OE or the tuxbox CVS.

Even though this is an image and needs to be flashed as any other image on the 7000, it's a very radical kind of image. It does not contain a copy of enigma, and it is utterly useless on it's own for watching TV. Think of it as a base image for multi-boot without internal TV-watching functionality. It will need the help of other images for watching TV.

Contains

• the same WEB interface for installing/removing/administrating images etc. that is used on the 7020 version
• the same bootmenu to select which image to start, displayed on the TV and controlled by the remote control
• the same options for external storage, i.e. USB/HDD/CF/NFS/CIFS
• supports both 7000 images and 7020 images

Installation: Read the included "README7000.txt" file. Make sure you really want to install it. Be aware that for best results you should repartition/reformat your USB stick, so if you want the option to go back to your previous environemnt, you probably want backups of everything before installing this. Then flash the .img file. Use whatever flash method that works best for you. Personally, I've had great success with using the "root.cramfs" method run from the latest official DMM image. The boot menu is displayed on your TV when the box reboots after flashing. Use your remote control to set your IP address (default is to use DHCP, so no need for this if you have a DHCP server), then go to WEB mode . Now use your WEB browser on your PC/MAC/Linux/whatever to install images (and to read more documentation).

Uninstallation: Flash the latest official DMM image. If you have installed the latest official DMM image (for the 7000) on external media, you can boot this image and use it's "root.cramfs" method for flashing.

A big thanks to the testers of the 7020 version! This version has taken advantage of the bugs uncovered in that process.

Because it is based on the 7020 version (which has received quite a bit of testing) and because for the 7000 version I have been able to do some testing on my own machine, I consider this to be "public beta" level. I have no objections to this version being spread around on other boards, provided that the uploader has personal experience installing and running it. Basically I want it to be "I tried it, and it works for me". That should be enough to stop it from spreading in case I goofed up in any last-minute change or in the packaging itself.

To update from an earlier version: backup your config file (from WEB mode) and re-flash. I've personally had success by using the "root.cramfs" method from an official DMM image on multiboot.

WARNING! For some hardware, it seems that the partitioning/formatting function in WEB mode makes incorrect file systems. The new version 0.18 fixes that bug (I hope), but it will also have the effect that some devices can't be automatically formatted and partitioned. It is now recommened that partitioning and formatting is done by hand, from the command line using "telnet".

Update for 0.18:
- Important update: Attempts to fix partitioning/formatting bug. Still, it is now recommended that the user performs the partitioning/formatting by hand. (thanks and my sincere apologies to those who got into trouble)
- Fixes message on TV when fsck fails during image mount.
- Fixes bug in naming partitioning schemes, with and without swap on USB and CF (thanks to grandi).
- German translation of README file for 7000 (thanks to zoppa).
- French translation of README file for 7020 (thanks to mimi74).
- Changes to /etc/httpd.conf (thanks to SADER).
- Kernel update for 7000 version. Tentative, will go away again if it creates more problems than it solves (thanks to Epmak).

Same conditions apply: Please don't upload this to other places until you have personally tested it. This time, the fixed bug was so important that I've skipped on my own testing. It should work, though, but please make sure before spreading it around.

[Epmak]

Hi noggie!

I tried to run miniroot on DM7000.
Unfortunatelly without success yet. What I did and what was strange:

1) I backup my images from flash & USB using FlashWizard (I suppose it a must for everyone)
2) I flashed miniroot with Flaswizard (no problems there)
3) When miniroot loaded I got screen about network. There was first small problem for me. As I understand that screen is displayed using SVHS mode.
But my TV set have 3 inputs: 2 allow composite&SVHS and one allow Composite&RGB. I use RGB output option for Dreambox, so screen looked not nice for me. I suppose that it could be better to set composite output as default.
4) Then I opened browser and connected to DM7000 trying to make initial
partitioning. (I have USB stick & HDD drive)
I was realy surprised that I didnt have USB as recognized device (I have USB as partitioned scheme and IDE1 as device in comboboxes). That is realy
confusing becouse majority will just click and lost their hdd data.
5) I went to telnet and found that there is nothing in /dev/scsi
6) dmesg shows lines that usb stick was recognized
7) Then I tried to restore flash using FlashWizard. Flashing was successfull but Dreambox does not rebooted automatically. I used '3 magic keys' for that.

Maybe I missunderstood something?

If you have any ideas I can repeat flashing.

With best wishes
Epmak

[rsrasc]

Hi Noggie,

Would you please tell me how long it will take to install a 7020 image. I'm installing now the base image and it's approaching now an hour since I hit the buttom. Is this normal? What is the approximate time? Your advise is appreciated.

Thanks,
rsrasc

[zoppa]

I try to install miniroot7000 as miniroot7020 image but mounting squashfs fails. This is the ouput from dmesg:

SQUASHFS error: Can't find a SQUASHFS superblock on loop1

[noggie]

Originally Posted by zoppa

I try to install miniroot7000 as miniroot7020 image but mounting squashfs fails. This is the ouput from dmesg:

SQUASHFS error: Can't find a SQUASHFS superblock on loop1

Yes. That's one of the radical bits about it. It doesn't use quashfs for the root partition, instead it's using jffs2. So no standard way of unpacking it will work. If anybody wants to peek inside without actually installing it, the best bet is to download the "build" pack, which contains the directory structure that's used to make the jffs2 part of the image.

[noggie]

Originally Posted by Epmak

Hi noggie!

I tried to run miniroot on DM7000.
Unfortunatelly without success yet. What I did and what was strange:

1) I backup my images from flash & USB using FlashWizard (I suppose it a must for everyone)
2) I flashed miniroot with Flaswizard (no problems there)
3) When miniroot loaded I got screen about network. There was first small problem for me. As I understand that screen is displayed using SVHS mode.
But my TV set have 3 inputs: 2 allow composite&SVHS and one allow Composite&RGB. I use RGB output option for Dreambox, so screen looked not nice for me. I suppose that it could be better to set composite output as default.
4) Then I opened browser and connected to DM7000 trying to make initial
partitioning. (I have USB stick & HDD drive)
I was realy surprised that I didnt have USB as recognized device (I have USB as partitioned scheme and IDE1 as device in comboboxes). That is realy
confusing becouse majority will just click and lost their hdd data.
5) I went to telnet and found that there is nothing in /dev/scsi
6) dmesg shows lines that usb stick was recognized
7) Then I tried to restore flash using FlashWizard. Flashing was successfull but Dreambox does not rebooted automatically. I used '3 magic keys' for that.

Maybe I missunderstood something?

If you have any ideas I can repeat flashing.

With best wishes
Epmak

Thanks for an excellent report, Epmak! Glad to hear that you managed to get back to your old set-up without too much trouble or any lasting ill effects.

1. I agree 100%. It takes a very special dedication to untested software to leave yourself without a retreat.
2. Good. I didn't expect any problems since I've tested flashing it myself, but it's always nice to have it confirmed by someone else.
3. If it's any consolation, it looks shitty on my TV also. But at least I've chosen the colors so that it's still readable. My problem is that I don't know how to initialize whatever it is that isn't correctly initialized, and I haven't spent the necessary time hunting around to try to find out, since it's not critical. It anybody has a snippet of code or a reference to some source that does it correctly, I'd be more than happy to remove this blemish.
4-6. This is the worrying part. I've read that some people have had problems with USB when running the latest ofiicial kernel. I've been fortunate enough to have had my hardware recognized without any problems. Did you have problems when running 109official? The observation you had that /dev/scsi was empty and that the device list didn't include the USB stick make sense. The software is picking up the list of possible devices and partitions from /proc/partitions, so it should recognize everything that the kernel recognizes, and the kernel is reponsible (through devfs) for populating /dev. The confusing part is that "dmesg" shows that the kernel found the hardware... But then again, I never had these problems with the 109 kernel and USB myself, so I don't know the details of how this is reported.
7. Hey! Good news! I didn't expect that using FW would work at all for restoring a different image, especially when running off miniroot itself. I would have though that dreamUP was the only way back in that situation. That you needed the three-fingered salute to reboot doesn't surprise me the least. I haven't tried to reverse-engineer FW to try find out the details of what it is doing, so I have no idea what the missing component is that makes the reboot fail.

No, I don't think you have misunderstood something. The main problem is that somehow your USB stick is not properly recognized.

I may have to ask (beg, really) to try again later, but before that, there's a few other questions/tests that are more relevant.

The most pressing one is: Could you please say if you had any problems with USB when running official 109 from flash? And what image are you running from flash normally (presumably without USB problems)?

[noggie]

Originally Posted by rsrasc

Hi Noggie,

Would you please tell me how long it will take to install a 7020 image. I'm installing now the base image and it's approaching now an hour since I hit the buttom. Is this normal? What is the approximate time? Your advise is appreciated.

Thanks,
rsrasc

An hour is far too much. Didn't you get any feedback from your browser at all? Nothing on the TV about a problem during fsck? It's extremely difficult to come up with any suggestions as long as I have no idea where in the process something went wrong...

Maybe I'll just confuse the issue by shooting off in the dark, but the first thing that strikes me is that maybe the file system you're installing on is rotten, and that installation hit problems when doing the fsck. Did you format the media before installation, or at least a "check-and-tune"?

[Epmak]

Originally Posted by noggie

I've read that some people have had problems with USB when running the latest ofiicial kernel.

Did you have problems when running 109official? The observation you had that /dev/scsi was empty and that the device list didn't include the USB stick make sense. The confusing part is that "dmesg" shows that the kernel found the hardware... But then again, I never had these problems with the 109 kernel and USB myself, so I don't know the details of how this is reported.

I may have to ask (beg, really) to try again later, but before that, there's a few other questions/tests that are more relevant.

The most pressing one is: Could you please say if you had any problems with USB when running official 109 from flash? And what image are you running from flash normally (presumably without USB problems)?

I just tried to flash official 1.09 release (till today I had Hydra Phoenix in flash).
My USB drive works fine with 1.09.
But I must say that I had problems with some beta versions of official firmware with 2.6 kernel.

Now I flashed miniroot. There is dmesg output:

/dev/scsi # dmesg
Linux version 2.6.9 (ghost@Client) (gcc version 3.4.3) #1 Mon Mar 14 15:53:41 CET 2005
On node 0 totalpages: 8192
DMA zone: 8192 pages, LIFO batch:2
Normal zone: 0 pages, LIFO batch:1
HighMem zone: 0 pages, LIFO batch:1
Built 1 zonelists
Kernel command line: console=null root=/dev/mtdblock5 rootfstype=jffs2 ro
PID hash table entries: 256 (order: 8, 4096 bytes)
Console: colour dummy device 80x25
Dentry cache hash table entries: 8192 (order: 3, 32768 bytes)
Inode-cache hash table entries: 4096 (order: 2, 16384 bytes)
Memory: 29904k available (1900k kernel code, 520k data, 80k init, 0k highmem)
Calibrating delay loop... 250.36 BogoMIPS (lpj=125184)
Mount-cache hash table entries: 512 (order: 0, 4096 bytes)
NET: Registered protocol family 16
SCSI subsystem initialized
usbcore: registered new driver usbfs
usbcore: registered new driver hub
devfs: 2004-01-31 Richard Gooch (rgooch@atnf.csiro.au)
devfs: boot_options: 0x1
JFFS2 version 2.2. (C) 2001-2003 Red Hat, Inc.
Serial: 8250/16550 driver $Revision: 1.90 $ 7 ports, IRQ sharing disabled
ttyS0 at MMIO 0x0 (irq = 20) is a 16550A
ttyS1 at MMIO 0x0 (irq = 21) is a 16550A
ttyS2 at MMIO 0x0 (irq = 31) is a 16550A
loop: loaded (max 8 devices)
smc91x.c: v1.1, sep 22 2004 by Nicolas Pitre <nico@cam.org>
eth0: SMC91C11xFD (rev 1) at 0xc306a300 IRQ 28 [nowait]
eth0: Ethernet addr: 00:09:34:10:04:a2
eth0: PHY LAN83C183 (LAN91C111 Internal)
Uniform Multi-Platform E-IDE driver Revision: 7.00alpha2
ide: Assuming 50MHz system bus speed for PIO modes; override with idebus=xx
IBM STB04xxx IDE driver version 2.0
setup bus 0
Probing IDE interface ide0...
hda: SAMSUNG SV1604N, ATA DISK drive
Unhandled interrupt 19, disabled
set drive!
Using deadline io scheduler
ide0 at 0xc306c1f8-0xc306c1ff,0xc306c1f6 on irq 25
hda: max request size: 1024KiB
hda: 312581808 sectors (160041 MB) w/2048KiB Cache, CHS=19457/255/63, UDMA(33)
hda: cache flushes supported
/dev/ide/host0/bus0/target0/lun0: p1
dreambox: flash mapping: 800000 at 7f800000
DreamBOX rev3+: Found 1 x16 devices at 0x0 in 16-bit bank
Intel/Sharp Extended Query Table at 0x0031
cfi_cmdset_0001: Erase suspend on write enabled
0: offset=0x0,size=0x20000,blocks=64
Using buffer write method
disable flash VPP
Creating 7 MTD partitions on "DreamBOX rev3+":
0x00000000-0x00600000 : "DreamBOX cramfs+squashfs"
0x00600000-0x007c0000 : "DreamBOX jffs2"
0x007c0000-0x00800000 : "DreamBOX OpenBIOS"
0x00000000-0x007c0000 : "DreamBOX (w/o bootloader)"
0x00000000-0x00800000 : "DreamBOX (w/ bootloader)"
0x00120000-0x00600000 : "DreamBOX SquashedFS"
0x00000000-0x00120000 : "DreamBOX Cramfs"
drivers/usb/host/ohci-ocp.c: 2004 Feb 02 USB 1.1 'Open' Host Controller (OHCI) Driver (OCP)
drivers/usb/host/ohci-ocp.c: block sizes: ed 64 td 64
drivers/usb/host/ohci-ocp.c: starting OCP OHCI USB Controller
drivers/usb/host/ohci-ocp.c: ocp_ohci (OCP OHCI) at 0xe0010000, irq 18
ocp-ohci 07: new USB bus registered, assigned bus number 1
hub 1-0:1.0: USB hub found
hub 1-0:1.0: 2 ports detected
Initializing USB Mass Storage driver...
usbcore: registered new driver usb-storage
USB Mass Storage support registered.
mice: PS/2 mouse device common for all mice
i2c /dev entries driver
IBM IIC driver v2.1
ibm-iic0: using standard (100 kHz) mode
ibm-iic1: using standard (100 kHz) mode
NET: Registered protocol family 2
IP: routing cache hash table of 512 buckets, 4Kbytes
TCP: Hash tables configured (established 2048 bind 4096)
NET: Registered protocol family 1
NET: Registered protocol family 17
VFS: Mounted root (jffs2 filesystem) readonly.
Mounted devfs on /dev
Freeing unused kernel memory: 80k init
Warning: unable to open an initial console.
usb 1-1: new full speed USB device using address 2
hub 1-1:1.0: USB hub found
hub 1-1:1.0: 1 port detected
usb 1-1.1: new full speed USB device using address 3
scsi0 : SCSI emulation for USB Mass Storage devices
Vendor: Model: Rev:
Type: Direct-Access ANSI SCSI revision: 02
head: module license 'Proprietary' taints kernel.
fp: attaching fp at 0xee to adapter IBM IIC
FPrev 0.05
valid rtc time.. 1111769213
using a0900000 as framebuffer base
VID_DISP_MODE set to 0088e6c0
registered framebuffer
allocated major 251 for denc.
demux set source
core: Loaded IBM STB04xxx core driver
write 1 event's ...
audio init
AUDIO_CLEAR_BUFFER
res: 0
demux: init
xp0_clk_stop
registering frontend...
disable continuous tone
already got a frontend.
detected tuner.
stb04_pig: $Id: stb04_pig.c,v 1.10 2005/02/22 23:07:45 obi Exp $
lcd.o: LCD driver (KS0713) module
avs: attaching avs at 0x94 to adapter IBM IIC
write 1 event's ...
VIDEO_SELECT_SOURCE
open video clip-mode
VIDEO_CLEAR_BUFFER
VIDEO_PLAY / VIDEO_CONTINUE, state 1 0
write 1 event's ...
VIDEO_STOP, state 0
setting 414720 bytes to 0
setting 414720 bytes to 0
eth0: link down
eth0: link up, 100Mbps, full-duplex, lpa 0x45E1
Unable to find swap-space signature
kjournald starting. Commit interval 5 seconds
EXT3-fs warning: maximal mount count reached, running e2fsck is recommended
EXT3 FS on hda1, internal journal
EXT3-fs: recovery complete.
EXT3-fs: mounted filesystem with ordered data mode.

[grandi]

Works fine here.

- Installed as "root.cramfs" to the flash
- After reboot getting an IP-address from my dhcp
- Repart. the usb-stick through the web-interface
- I've choosen to make an swap-partition - but only one partition was made. I think the usb-stick is too small -> only 64 MB
- installing the last dmm-release for the 7020 on my 7000 (took one or two minutes) and choosing the hdd for temp. data.
- Rebooting and starting the 7020 image
- For now no problems to report but i think it's better to have an swap-partition ...

Thank you noggie for this great improvement

[noggie]

Originally Posted by Epmak

I just tried to flash official 1.09 release (till today I had Hydra Phoenix in flash).
My USB drive works fine with 1.09.
But I must say that I had problems with some beta versions of official firmware with 2.6 kernel.

Now I flashed miniroot. There is dmesg output:

Excellent. If you still have it loaded, could you also report the contents of /proc/partitions? And just to make sure, the contents of /tmp/partitions?