|
Factorite (Junior) Posts: | 34 |
Location: | Castle Rock, Colorado | Status: | |
| Quote oldman - 2007-10-31 11:45 AM
First of all, I saw some members thought the kernel would be part of the distributions. It isn't! The kernel and the initrd always come from the boot image (nBkProOs.img).
Ok, things are starting to make sense. That means the JFFS2 files are the "userspace"... correct?
This guide you put together is gold. Thanks! As we get a little further on, maybe we can get a semi-formal guide put together. Wiki somewhere, perhaps?
This bit is of interest to me (and quite possibly salmonito... ) :
Quote
In the running distribution: Making the correct kernel modules available. IMPORTANT!
This gives access to the additional kernel drivers, provided by the initrd!!! 
In a root shell type:
# cd /lib/modules
# mv 2.6.9-rc1-bk18-nb1 2.6.9-rc1-bk18-nb1.old
# ln -s /initrd/lib/modules/2.6.9-rc1-bk18-nb1 .
# halt
Then wait for black screen, reset the machine and re-start as usual with Ctrl+D+ESC (or without Ctrl+D if installed to the flash ROM's ). If you enter a root shell again and type lsmod, you should see a lot of modules loaded.
Going to work on this bit tonight and see if I can get one of my wifi cards to work.
@mrkrupa: Have you tried erasing the NAND (from the regular BooSt menu ) before trying to boot linux? I had to do that. Also, you'll get the black screen for a minute or so before you get into the linux environment. On a serial connection, you can see that the NBP is writing the info to NAND (unless you follow oldman's instructions for CF-only booting ). |
|
| |
|
Factor Fanatic Posts: | 64 |
Location: | Washington, DC USA | Status: | |
| So this is the emergency shell and I don't see why it doesn't boot like Salimino. I used the same files. What is the Boost version?
Paul Krupa |
|
| |
|
Factor Fanatic Posts: | 61 |
Location: | Vermont | Status: | |
| Quote mrkrupa - 2007-10-31 3:20 PM
So this is the emergency shell and I don't see why it doesn't boot like Salimino. I used the same files. What is the Boost version?
Paul Krupa
Paul its good that you can see output through your serial cable I did not have any. Just keep three files on your CF those are two img files and sh unmodified from ~anonymous, and then try different jffs2s there are two from Florian in addition. Tonight I will try booting Angstrom jffs2...
Eddie yes that would be interesting  So it means that even if i boot jffs2 from Angstrom i will still use old nBkProOS.img and it will result in me having 2.6.9 kernel right?...
Edited by salmonito 2007-10-31 4:37 PM
|
|
| |
|
| @mrkrupa: The fatal error in your output seems to be the message: mount: /dev/hda1 is not a valid block device. Because this means the script NBLXLoad.sh will not be executed. Perhaps it could help to partition and format the CF card again. Make sure the parition type is 06h (use linux fdisk), and format it FAT16 (use linux mkdosfs without any options). Maybe the partition must not be too large (<32MB or so). |
|
| |
|
| @Eddit & salmonito: Yes, the jffs2 images are userspace and do not contain the kernel. I saw the latest angstrom images are made for kernel 2.6.17. I tried one yesterday using the flash-less boot procedere, but it did not work (don't know why, because I still have no serial cable). I think we would have to create a nBkProOs.img with a 2.6.17 kernel first.
But maybe one should try to prepare an up-to-date kernel (2.6.23.1 or so). I saw that several parts of the driver patches have already gone into the official kernel sources (because the chips are even used in other machines). This could reduce the effort of updating the patches. I'm still confused about the nBkProOs.img. Does it contain just the kernel and the initrd, or is there some extra boot-up code which tells the kernel about the initrd and other params. This could be figured out by compiling 2.6.9-rc1-bk18-nbp0 and comparing the disassemblings. But I did not yet get a GCC-3.44 based arm-linux toolchain running here, just 2.95 and 4.2 which are refusing to compile kernel 2.6.9. Does anybody know where to download such a toolchain in binary form? (I'm still too lazy to compile it myself...)
Edited by oldman 2007-10-31 5:54 PM
|
|
| |
|
H/PC Elite Posts: | 639 |
Location: | Green Bay, WI | Status: | |
| Probably the easiest would be to download and setup OE ( open embedded ) as thats whats used to build jlime, openzaurus, familiar, angstrom etc etc.
http://www.openembedded.org/wiki/GettingStarted
Cheers. |
|
| |
|
Factor Fanatic Posts: | 64 |
Location: | Washington, DC USA | Status: | |
| I did a dmesg at the emergency shell:
/bin # dmesg
00
Bad eraseblock 115 at 0x001cc000
Bad eraseblock 116 at 0x001d0000
Bad eraseblock 117 at 0x001d4000
Bad eraseblock 118 at 0x001d8000
Bad eraseblock 119 at 0x001dc000
Bad eraseblock 120 at 0x001e0000
Bad eraseblock 121 at 0x001e4000
Bad eraseblock 122 at 0x001e8000
Bad eraseblock 123 at 0x001ec000
Bad eraseblock 124 at 0x001f0000
Bad eraseblock 125 at 0x001f4000
Bad eraseblock 126 at 0x001f8000
Bad eraseblock 127 at 0x001fc000
Bad eraseblock 128 at 0x00200000
Bad eraseblock 129 at 0x00204000
Bad eraseblock 130 at 0x00208000
Bad eraseblock 131 at 0x0020c000
Bad eraseblock 132 at 0x00210000
Bad eraseblock 133 at 0x00214000
Bad eraseblock 134 at 0x00218000
Bad eraseblock 135 at 0x0021c000
Bad eraseblock 136 at 0x00220000
Bad eraseblock 137 at 0x00224000
Bad eraseblock 138 at 0x00228000
Bad eraseblock 139 at 0x0022c000
Bad eraseblock 140 at 0x00230000
Bad eraseblock 141 at 0x00234000
Bad eraseblock 142 at 0x00238000
Bad eraseblock 143 at 0x0023c000
Bad eraseblock 144 at 0x00240000
Bad eraseblock 145 at 0x00244000
Bad eraseblock 146 at 0x00248000
Bad eraseblock 147 at 0x0024c000
Bad eraseblock 148 at 0x00250000
Bad eraseblock 149 at 0x00254000
Bad eraseblock 150 at 0x00258000
Bad eraseblock 151 at 0x0025c000
Bad eraseblock 152 at 0x00260000
Bad eraseblock 153 at 0x00264000
Bad eraseblock 154 at 0x00268000
Bad eraseblock 155 at 0x0026c000
Bad eraseblock 156 at 0x00270000
Bad eraseblock 157 at 0x00274000
Bad eraseblock 158 at 0x00278000
Bad eraseblock 159 at 0x0027c000
Bad eraseblock 160 at 0x00280000
Bad eraseblock 161 at 0x00284000
Bad eraseblock 162 at 0x00288000
Bad eraseblock 163 at 0x0028c000
Bad eraseblock 164 at 0x00290000
Bad eraseblock 165 at 0x00294000
Bad eraseblock 166 at 0x00298000
Bad eraseblock 167 at 0x0029c000
Bad eraseblock 168 at 0x002a0000
Bad eraseblock 169 at 0x002a4000
Bad eraseblock 170 at 0x002a8000
Bad eraseblock 171 at 0x002ac000
Bad eraseblock 172 at 0x002b0000
Bad eraseblock 174 at 0x002b8000
Bad eraseblock 175 at 0x002bc000
Bad eraseblock 176 at 0x002c0000
Bad eraseblock 177 at 0x002c4000
Bad eraseblock 178 at 0x002c8000
Bad eraseblock 179 at 0x002cc000
Bad eraseblock 180 at 0x002d0000
Bad eraseblock 181 at 0x002d4000
Bad eraseblock 182 at 0x002d8000
Bad eraseblock 183 at 0x002dc000
Bad eraseblock 184 at 0x002e0000
Bad eraseblock 185 at 0x002e4000
Bad eraseblock 186 at 0x002e8000
Bad eraseblock 187 at 0x002ec000
Bad eraseblock 188 at 0x002f0000
Bad eraseblock 189 at 0x002f4000
Bad eraseblock 190 at 0x002f8000
Bad eraseblock 191 at 0x002fc000
Bad eraseblock 192 at 0x00300000
Bad eraseblock 193 at 0x00304000
Bad eraseblock 194 at 0x00308000
Bad eraseblock 195 at 0x0030c000
Bad eraseblock 196 at 0x00310000
Bad eraseblock 197 at 0x00314000
Bad eraseblock 198 at 0x00318000
Bad eraseblock 199 at 0x0031c000
Bad eraseblock 200 at 0x00320000
Bad eraseblock 201 at 0x00324000
Bad eraseblock 202 at 0x00328000
Bad eraseblock 203 at 0x0032c000
Bad eraseblock 204 at 0x00330000
Bad eraseblock 205 at 0x00334000
Bad eraseblock 206 at 0x00338000
Bad eraseblock 207 at 0x0033c000
Bad eraseblock 208 at 0x00340000
Bad eraseblock 209 at 0x00344000
Bad eraseblock 210 at 0x00348000
Bad eraseblock 211 at 0x0034c000
Bad eraseblock 212 at 0x00350000
Bad eraseblock 213 at 0x00354000
Bad eraseblock 214 at 0x00358000
Bad eraseblock 215 at 0x0035c000
Bad eraseblock 216 at 0x00360000
Bad eraseblock 217 at 0x00364000
Bad eraseblock 218 at 0x00368000
Bad eraseblock 219 at 0x0036c000
Bad eraseblock 220 at 0x00370000
Bad eraseblock 221 at 0x00374000
Bad eraseblock 222 at 0x00378000
Bad eraseblock 223 at 0x0037c000
Bad eraseblock 224 at 0x00380000
Bad eraseblock 225 at 0x00384000
Bad eraseblock 226 at 0x00388000
Bad eraseblock 227 at 0x0038c000
Bad eraseblock 228 at 0x00390000
Bad eraseblock 229 at 0x00394000
Bad eraseblock 230 at 0x00398000
Bad eraseblock 231 at 0x0039c000
Bad eraseblock 232 at 0x003a0000
Bad eraseblock 233 at 0x003a4000
Bad eraseblock 234 at 0x003a8000
Bad eraseblock 235 at 0x003ac000
Bad eraseblock 236 at 0x003b0000
Bad eraseblock 237 at 0x003b4000
Bad eraseblock 238 at 0x003b8000
Bad eraseblock 239 at 0x003bc000
Bad eraseblock 240 at 0x003c0000
Bad eraseblock 241 at 0x003c4000
Bad eraseblock 242 at 0x003c8000
Bad eraseblock 243 at 0x003cc000
Bad eraseblock 244 at 0x003d0000
Bad eraseblock 245 at 0x003d4000
Bad eraseblock 246 at 0x003d8000
Bad eraseblock 247 at 0x003dc000
Bad eraseblock 248 at 0x003e0000
Bad eraseblock 249 at 0x003e4000
Bad eraseblock 250 at 0x003e8000
Bad eraseblock 255 at 0x003fc000
Adding partition at 00400000, size 01800000
Creating 2 MTD partitions on "NAND 32MiB 3,3V 16-bit":
0x00400000-0x01c00000 : "NetbookPro nand 1"
0x03c00000-0x04000000 : "NetbookPro nand 2"
mtd: partition "NetbookPro nand 2" is out of reach -- disabled
NetBookPro: installing PCMICA driver device
usbcore: registered new driver usbhid
drivers/usb/input/hid-core.c: v2.0:USB HID core driver
mice: PS/2 mouse device common for all mice
NetBookPro Keyboard Driver, (c) 2004 Simtec Electronics
i2c /dev entries driver
Resetting I2C Controller Unit
I2C: Slave address 112
I2C: Enabling slave mode
I2C: NVRAM at bus address 0x53
I2C: i2c-0: PXA I2C adapter
Ricoh RTC Driver, (c) 2004 Simtec Electronics
I2C: Ricoh R2051 RTC, 12Hr mode, oscillator restarted
Moving power state to Run
input: NetBookPro Keyboard
i2c_eeprom: adding watcher c0620bd8 to 0x00
i2c_eeprom: adding watcher c061dab0 to 0x05
ac97_codec: AC97 Audio codec, id: WML5 (Wolfson WM9705/WM9710)
wolfson_init05: read 00000410
ac97: enabling WM9705 Variable-Rate sampling
wm97xx: Wolfson WM9705/WM9712 Touchscreen Controller driver
wm97xx: Version 0.23 liam.girdwood@wolfsonmicro.com
wm97xx: Detected a WM9705 codec connected to AC97 controller 0
NET: Registered protocol family 2
Power state is now Run (8)
starting power monitor
nbp-battmon: issuing start...
nbp-battmon: thread woken
nbp-battmon: nbp_battmon_setstate: 1
IP: routing cache hash table of 1024 buckets, 8Kbytes
TCP: Hash tables configured (established 8192 bind 16384)
NET: Registered protocol family 1
NET: Registered protocol family 17
Epson S1D13XXX FB Driver, (c) 2004 Simtec Electronics
s1d138xx: s1d13806: revision 0
s1dfb-nbp: s1d13806: initialsing netbookpro settings
Console: switching to colour frame buffer device 100x37
fb0: s1d13xxx framebuffer device
RAMDISK: ext2 filesystem found at block 0
RAMDISK: Loading 8192KiB [1 disk] into ram disk... <7>nbp-battmon: nbp_battmon_setstate: returned 0
nbp-battmon: Battery monitor ready
<6>nbp-battmon: system restarted
done.
VFS: Mounted root (ext2 filesystem).
Freeing init memory: 72K
JFFS2 version 2.2. (NAND) (C) 2001-2003 Red Hat, Inc.
/bin #
|
|
| |
|
Factorite (Junior) Posts: | 34 |
Location: | Castle Rock, Colorado | Status: | |
| @oldman:
Tried your module activation instructions. Got mixed results. The WiFi drivers appear to be available now. The Socket CF WiFi and the Proxima PCMCIA WiFi didn't work -- the system logs state that there aren't any compatible drivers for the Intersil chipset. I had the CF/PC Card utility opened and inserted the Linksys PCMCIA WiFi. Detected it right away and assigned it the wavelan driver automatically.
Then after a few seconds the whole thing froze solid. Had to reset. Tried inserting the card at various times (before turning on, before a hard reset, etc ) and it always hangs. When I insert the card before a hard reset, I get the usual serial output. Here's the final few lines before it freezes:
Pivoting filesystem
INIT: version 2.86 booting
Setting up IP spoofing protection: rp_filter.
Configuring network interfaces... done.
Starting portmap daemon: portmap.
Wed Oct 31 16:46:00 UTC 2007
Nothing to be done
INIT: Entering runlevel: 5
Starting Dropbear SSH server: dropbear.
Starting system message bus: dbus-1.
Starting IrDA: .
Starting PCMCIA services: cardmgr[4758]: watching 2 sockets
done.
cardmgr[4760]: socket 1: The Linksys Group, Inc. Instant Wireless Network PC Card
Starting syslogd/klogd: done
Starting Bluetooth subsystem: hcid sdpd hid2hci rfcomm.
Starting at daemon: atd.
Starting GPE display manager: gpe-dm
OpenZaurus 3.5.4.1 netbook-pro ttySA0
netbook-pro login:
At this point (unless Maris is having success ), I have to agree with oldman: we need to start looking at compiling a fresh kernel/initrd (and probably JFFS2 ).
*Sigh* Time to read up on getting the tool chain set up...  |
|
| |
|
Factorite (Junior) Posts: | 34 |
Location: | Castle Rock, Colorado | Status: | |
| Maris, you having any luck with the Angstrom JFFS2? I tried the Angstrom-x11-image-glibc-test-20070710-netbook-pro.rootfs.jffs2 file, but couldn't get it to work. Started off well (watching on serial connection ) but seg-faulted near the end.
...
Writing data to block 15f0000
completed - remove card
JFFS2 version 2.2. (NAND) (C) 2001-2003 Red Hat, Inc.
mounting root fs writable
Pivoting filesystem
Segmentation fault
|
|
| |
|
H/PC Sensei Posts: | 1,330 |
Location: | North of England | Status: | |
| I think that's where we are at - new kernel needed.
I know Koen looked at what was necessary and I think there's some backporting of code involved: IIRC it's something around the PCON interface that needs porting to the right level.
update - here's an extract of Koen's mail:
Quote Once we can succesfully boot a 2.6.9 kernel image with that, we can look at
updating the kernel port (mostly forward porting the pcon driver to the new i2c code).
|
|
| |
|
Factor Fanatic Posts: | 64 |
Location: | Washington, DC USA | Status: | |
| You guys that have this running. Can you look at dmesg and see if you see the same as I posted above. I assume that is because the file has not been modded for the smaller ram size. It does not seem to affect the ability to load.
Paul
PS I have reformatted and tried numerous cf's with linux and windows and netbook formatting. Widows seems to give the best results, the ext2 format had additional IDE errors. |
|
| |
|
Factor Fanatic Posts: | 61 |
Location: | Vermont | Status: | |
| Quote Intrepid-Eddie - 2007-11-01 12:20 AM
Maris, you having any luck with the Angstrom JFFS2? I tried the Angstrom-x11-image-glibc-test-20070710-netbook-pro.rootfs.jffs2 file, but couldn't get it to work. Started off well (watching on serial connection ) but seg-faulted near the end.
...
Writing data to block 15f0000
completed - remove card
JFFS2 version 2.2. (NAND) (C) 2001-2003 Red Hat, Inc.
mounting root fs writable
Pivoting filesystem
Segmentation fault
No luck here too... I tried nBkProOS.img from LX project did not work too, here is the point where trying will not work and I am lacking knowledge.  My next step i think is to install OE and all it needs. Decompress os imagefile see what's inside. Compile kernel myself. Just to know what goes where. It sucks not to know enough. |
|
| |
|
Factor Fanatic Posts: | 61 |
Location: | Vermont | Status: | |
| Paul :mtd: partition "NetbookPro nand 2" is out of reach -- disabled ?
no idea why it can be so...
Edited by salmonito 2007-11-01 8:08 AM
|
|
| |
|
Factor Fanatic Posts: | 61 |
Location: | Vermont | Status: | |
| Quote chiark - 2007-11-01 3:50 AM
I think that's where we are at - new kernel needed.
I know Koen looked at what was necessary and I think there's some backporting of code involved: IIRC it's something around the PCON interface that needs porting to the right level.
update - here's an extract of Koen's mail:
Quote Once we can succesfully boot a 2.6.9 kernel image with that, we can look at
updating the kernel port (mostly forward porting the pcon driver to the new i2c code).
And in simpler terms it would mean? |
|
| |
|
H/PC Sensei Posts: | 1,330 |
Location: | North of England | Status: | |
| ...that to move on, we need to help work on the kernel. Good news is that we have the source to the old kernel though which I think was missing before!
I've received the BoOst header calculator code - it's written as something for the author to use themself so could do with a bit of work to make into a user tool, and I've said to the author that I'll do that before I release it. Looks like we don't need it yet, anyhow |
|
| |
