I-Appliance BBS
The Official Source for Internet Appliance Upgrades and Mods
Amazon Honor System Click Here to Pay Learn More
BBS Main List | Sign In | Sign Up | Search | Help | Linux-Hacker.netReply to Thread | Printer |

Home / Other I-Appliances / WebPal
Had to cut JP1

New MessageHad to cut JP1 (modified 0 times) newell
Profile
One of my webpals didn't want to run reliably in gdb or comm with wpflash. Cut the trace bridging JP1 and everything is fine. Suspect that the micro's port pin is not tristated, so it's conflicting with the console serial port input (J11-3).

BTW, my old 486-33 wasn't swift enough to deal with the data rate of wpflash.

01-01-2003 00:29:27

New MessageRE:Had to cut JP1 (modified 0 times) Signals
Profile
I think I might be having the same problem, but I don't understand your solution. Where would I find JP1? I'm not seeing anything screened on the motherboard...

-Kevin

01-18-2003 12:40:07

New MessageRE:Had to cut JP1 (modified 0 times) Signals
Profile
Nevermind, I found it under the memory. I cut JP1 and I still am having trouble. ONCE I got it to load prog.srec, but then the next "set" instruction hung indefinately. I've not gotten it to even load prog.srec again. I usually can't get it to sync.

Should the Easyrider message says ER_NBOARD in it? Looks like an error message to me.

-Signals

01-18-2003 13:08:04

New MessageRE:Had to cut JP1 (modified 0 times) newell
Profile
Both my Webpals give the message "Easy Rider [64MB] ER_NBOARD (Newcom) (bootrom v3.07beta)", so I don't think the ER_NBOARD is a error. (ER may stand for _E_asy _R_ider?)

It's not unusual for me to reset three or four times before gdb syncs-up.

01-18-2003 16:55:41

New MessageRE:Had to cut JP1 (modified 0 times) Signals
Profile
I got it... I had forgotten about my Gnome based Palm sync software. (I don't use it.) It was set to listen on the same serial line as I was trying to use for the Webpal. After turning it off, gdb was able to sync immediately. I've been able to wpflash as much as I want, and I haven't had another failure yet!

Whew, tough weekend. Finally got the kernel loaded on a webpal, got a cross-compiler working, patched the kernel source for IR keyboard support, built my own kernel, and made my own 352K root filesystem. Now all I have to do is figure out what to do with this laptop drive I have sitting here. I can't seem to find a power connector on it, just the ribbon cable connector on the back. Looks like it has a few more pins than a normal IDE drive does. It's a Quantum Go*Drive Daytona, which is unfortunate since there's no information about HDs at Quantum anymore, and I can't find any at Maxtor. Oh well, I'm sure I'll figure it out....

Thanks for all of the help.
-Signals

01-20-2003 06:53:44

New MessageRE:Had to cut JP1 (modified 0 times) newell
Profile
Now all I have to do is figure out what to do with this laptop drive I have sitting here. I can't seem to find a power connector on it, just the ribbon cable connector on the back. Looks like it has a few more pins than a normal IDE drive does.

Laptop IDE drives are 44 pin. The extra pins provide +5V and ground, hence no power connector. They're nice because they don't need +12V, which isn't easily available in a stock Webpal.

BTW, what's in your ~300kB root filesystem? (I'm still NFS mounting root.)

01-20-2003 08:26:31

New MessageRE:Had to cut JP1 (modified 0 times) Signals
Profile
> BTW, what's in your ~300kB root filesystem? (I'm still NFS mounting root.)

Right now, pretty much nothing but busybox and some kernel modules. I just spent an evening trying to PLIP the Webpal to my Linux machine only to find out that PLIP is broken in the 2.4.18 kernel. Now I have to figure out how to apply the Webpal patch to 2.4.19 or I have to figure out how to patch the PLIP code in 2.4.18 or something... Doh! Maybe I should just break down and order some of those cheap NE2000's

I tried SLIP yesterday, so I had 'slattach' in my root partition, but I had to gzip it and expand it in a ram disk to make it fit. Unfortunately, I had no luck with slip either.

-Signals

01-20-2003 19:48:00

New MessageRE:Had to cut JP1 (modified 0 times) Ifaistos
Profile
I am also having problems using the gdb and wpflash, especially the second.
The gdb connects ok (after a couple of resets), prog.srec loads but then when i run wpflash nothing happens. (wpflash read all wepbal.bin)
I get a message "Reset webpal to continue programming" but reseting the webpal puts it back to "normal" mode...
Tried to build and install emlilo but webpal goes into "never-never land" after the emlilo-rom is flashed (The ram based rdp emlilo loads and executes ok)

Everything is connected on ttyS0 and no other apps claim it.

Any idea what i am doing wrong ?

Ifaistos

P.S I have build a "stand alone" programmer to do all the flash programming, so when the orginal webpal rom gets flashed and does not work, i reflash the original code back

01-21-2003 13:08:12

New MessageRE:Had to cut JP1 (modified 0 times) newell
Profile
The gdb connects ok (after a couple of resets), prog.srec loads but then when i run wpflash nothing happens. (wpflash read all wepbal.bin)
I get a message "Reset webpal to continue programming" but reseting the webpal puts it back to "normal" mode...

Sounds to me like you need to add the -nowait flag to your wpflash command line:
wpflash -nowait read all webpal.bin

01-21-2003 13:58:56

New MessageRE:Had to cut JP1 (modified 0 times) mevanson
Profile
Signals --

What kind of busybox configuration did you use? I used the toolset referenced on bigbrd's site to build busybox, but the resulting images (static build) were too large regardless of the functionality I switched off.

So I build uclibc (generic ARM) and then build busybox with uclibc. When I boot with the resulting disk image, I get some sort of stack dump after it loads the executable (I know it's after the executable because when I first build busybox, I compiled it dynamically but did not include the necessary libraries and I'm getting to the same point now).

Thanks,
--matt

01-26-2003 14:00:27

New MessageRE:Had to cut JP1 (modified 0 times) bigbrd
Profile
If you built a uclibc compiler using the 3.2.x version of the compiler, I discovered this weekend that the default for that version of the compiler seems to generate code for a newer ARM processor than the webpal has. If you use the compiler flags "-march=armv3 -mtune=arm710, it should probalby generate code with the correct ARM instruction subset.

The main instructions missing from the CS7500 are the halfword instructions. You can use the arm version of objdump -d binary.file to disassemble a binary that panic'd and then search for "ldrh". If you see any, that's what probably caused the crash. "ldrhi" is ok, that's a "ldr if hi" instruction.

I'm in the process of building a 3.2.1 compiler that has the CS7500 instruction set as the default. I changed "gcc/config/arm/unknown-elf.h" to "#define SUBTARGET_CPU_DEFAULT TARGET_CPU_arm7500fe".

Or, build a version of uclibc with the 2.95.4 compiler which seems to default to something that the webpal can handle. I've successfully used uclibc with 2.95.4 to build busybox and other things.

01-26-2003 23:55:50

New MessageRE:Had to cut JP1 (modified 0 times) mevanson
Profile
Thanks for the the response bigbrd. The information you've provided over time to this group has been invaluable.

Shortly after I posted this last message, I'd re-read your webpal page and discovered that I should have been using cramfs. Once I did that, I was able to build busybox with the 2.95.3 tool chain without the need for uclibc. Unfortunately, there's still not enough room to include all the functionality.

With the 2.95.3 toolchain, it seems that the only tool I'm not able to build is route. It seems definitions from the stdlib include with 2.95.3 tool chain are conflicting with the kernel headers provided with it. Haven't yet spent much time investigating it, but ultimately I may as I need the ability to define at least a default route.

01-27-2003 10:38:20

New MessageRE:Had to cut JP1 (modified 0 times) mevanson
Profile
Well, I came back to seeing what I could do with busybox and uClibc.

I discovered that the version that I had made things virtually unusable for ARM processesors. I was using 0.9.16. They just released 0.9.17 on January 25th.

Using that version I was successfully able to build a working busybox with nearly all the options. A little work and I could perhaps build more.

As for bigbrd's comments from earlier: I first built uClibc and busybox without the parameters you recommended. I then performed the objdump -d and there were a number of ldrh instructions. I built with the 2.95.3 toolchain reference on your site.

I didn't try the resulting binary. Installed I recompiled with your flags.

All works beautiful. It is amazing what they've accomplished with uClibc and the extra capacity it gives me.

01-30-2003 17:08:38

Reply to Thread | Printer |
All times are PSTPowered by UltraBoard v1.62



Copyright © 2000, Netmake Inc. All Rights Reserved.
See Terms and Conditions for more information.




i-opener opener laptop notebook computer help drivers dll free windows dos repair fix linux mac macintosh 2000 95 98 nt pc configure hardware software sound video netscape explorer network networking lan wan software cmos fat bios printer card mouse modem ide scsi cd rom controllers scanner tape hard drive cgi scripts source code mp3