| 00:07:12 | andersee | Electricity over IP |
| 00:07:12 | andersee | ftp://ftp.isi.edu/in-notes/rfc3251.txt |
| 00:08:02 | TomW | Russ: hello? |
| 00:10:35 | cpenguin_away | is away: Happy April Prime! |
| 00:18:03 | hevdevdisplay | hello all .. quick question on the linux progress patch ... |
| 00:19:05 | hevdevdisplay | the buildroot make file states that the kernel commandline should only include /dev/ttyS2 set as console ... |
| 00:19:59 | hevdevdisplay | but if the ttySA0 is taken out of the commandline how can you connect to the tuxscreen to update ramdisk etc? |
| 00:20:50 | hevdevdisplay | .. /dev/ttyS2 should have read /dev/tty2 =) |
| 00:21:44 | sjm | hevdevdisplay: you update using blob - before the kernel boots, so it doesn't matter what consoles you do/do not have enabled |
| 00:22:33 | hevdevdisplay | ahh .. i see that now .. too much coffee i think .. thanks =) |
| 00:23:13 | TomW | hevdevdisplay: too many trees in the way, eh? |
| 00:23:56 | sjm | TomW: you get nanox running? |
| 00:24:39 | TomW | Yeah, I did. Then I got wrapped up with upgrading the DRAM to 64M. |
| 00:24:51 | sjm | TomW: cool |
| 00:25:52 | TomW | I did have one board with 32M on it, it reported 16M on boot. Before I could get a new blob together, I toasted the board by lying it down on some solder wick. :-( |
| 00:47:42 | jacques | hi |
| 00:49:50 | TomW | hello |
| 00:51:01 | jacques | TomW: how's it going? |
| 00:51:14 | TomW | slowly. |
| 00:52:31 | TomW | cannot seem to find too many good tux CPU boards.. One I toasted, another I cannabalbized, the one I am using now has problems... |
| 00:53:53 | TomW | ... One I won't take apart (still works), only one left: it crashes on bootup but will work on a manual reset. guess that I need to purchase a couple more unit? ;-) |
| 00:58:47 | jacques | TomW: heh, get em while you can :-) |
| 00:59:53 | TomW | jacques: Yes! One of the reasons that I am working at the Tux and dragging feet on getting some work done for one customer.. |
| 01:01:42 | TomW | heh. Wonder if we'll be seeing used / slightly modified Shannons for sale on EBay a year from now? |
| 01:04:01 | jacques | almost certainly |
| 01:04:29 | jacques | if bzflag wasn't so determined to sell to developers we would be seeing them now |
| 01:04:42 | jacques | ppl were trying to buy his whole stock to resell |
| 01:04:57 | TomW | Really?!! |
| 01:05:12 | jacques | yep |
| 01:05:29 | jacques | the display panel alone is worth > $100 |
| 01:08:20 | TomW | Well, I am certainly thankfull that Tim went in that direction. I have been looking at ARM processors for a while and have not found the kernel-arm list to be the "best' place for a newbie to get started from ... |
| 01:09:02 | jacques | heh i agree totally |
| 01:10:07 | TomW | I have had a CLPS-7500FE development board here for almost two years and never got linux to run on it. Queries to the linux-arm list always resulted in something like "that is an older chip, but you should get it working with no trouble", DUH! That is why I asked for help, I couldn't. I finally gave up. |
| 01:10:50 | jacques | heh yeah I've seen things like that on there |
| 01:11:09 | jacques | one reason why I'm hesitant to post |
| 01:11:18 | TomW | Anyway. The Tux is very affordable and it was ready to rock-n-roll, all I had to do was to start hacking away! :-) |
| 01:11:42 | jacques | :-) |
| 01:12:16 | TomW | BTW, XMODEM is nicer. |
| 01:12:34 | TomW | esp with 1k blocks. |
| 01:13:19 | jacques | ah you are running the new blob |
| 01:13:25 | jacques | how is it? |
| 01:15:08 | TomW | quirky as all hell.. |
| 01:15:21 | jacques | hmmm, not for me yet then :-) |
| 01:19:10 | TomW | 2.0.5-pre3 wants you to manually erase the flash, or it won't program it properly from a 'flash x' command. |
| 01:19:41 | TomW | now I am having the devil's own time of putting a ramdisk into flash. Something not being initailized someplace.. |
| 01:24:40 | jacques | :-\ |
| 01:25:22 | russ_ | TomW? |
| 01:25:40 | Russ|werk | I was off watching my boards get stuffed |
| 01:27:34 | jacques | boards? |
| 01:27:54 | jacques | ah you designed your own? cool |
| 01:28:06 | Russ|werk | ya, designed them about a year ago and did 3 proto's |
| 01:28:12 | Russ|werk | 25 are running right now |
| 01:28:13 | TomW | Russ: Hey! I saw the results of that portscan I ran on that russ.blah.blah site. What are you doing, running a honeypot? |
| 01:29:03 | Russ|werk | dtk |
| 01:29:23 | TomW | dtk? |
| 01:29:25 | Russ|werk | it has a bunch of perl scripts that provide fake vunerable services, and log activity |
| 01:29:44 | TomW | Oh cool! Get many "takers"? |
| 01:30:45 | Russ|werk | just some very mild stuff |
| 01:30:50 | Russ|werk | attempted ftp, etc |
| 01:31:02 | Russ|werk | we get regular port scans though being at a uni |
| 01:31:47 | Russ|werk | anyone know if white wine cleans red wine off a PCB?= |
| 01:34:01 | TomW | carbon tet |
| 01:34:32 | TomW | carbon tetrichloride should work well. ;-) |
| 01:35:09 | TomW | All else fails, soak the fiberglass board in Acetone. |
| 01:35:22 | Russ|werk | hmm...don't think I have any of that |
| 01:35:26 | Russ|werk | its someones laptop |
| 01:35:26 | TomW | Heh. Believe that and you won't have a board for very long! |
| 01:36:02 | TomW | Acetone is a great PCB cleaner but it breaks down the fiberglass glue. |
| 01:36:29 | Russ|werk | TomW: just as long as you use it only in the center then |
| 01:38:11 | TomW | Yeah, I would polish the copper blanks with ScotchBrite, wash it with acetone, then rinse in water. Let it dry, then wash it with alcohol. |
| 01:38:21 | TomW | pain. |
| 01:39:25 | TomW | AHA! The new blob (2.0.5-pre3) won't program the ramdisk if you only have 8 meg of RAM (SODIMM removed). |
| 01:39:43 | jacques | ugh |
| 01:39:47 | TomW | It goes through the motions but does nothing. |
| 01:39:56 | TomW | or only does a partial. |
| 01:40:03 | jacques | that's just nasty |
| 01:40:12 | TomW | let me stick the SODIMM back in and retry. |
| 01:41:39 | TomW | They probably were thinking of a sequence of: xdo kernel, xdo ramdisk, fla kernel, fla ramdisk. Without 16M ram, they had no place to hold it all. Regardless if you only do: xdo ramdisk, fla ramdisk, it still uses the upper 8Meg to keep the ramdisk image? |
| 01:42:41 | Russ|werk | TomW: and thinking of what happens if you have a 32M SODIMM in and the first 8M is in 1M chunks |
| 01:44:43 | TomW | Hmm. That is true, they won't have the benefit of a MMU at the time that the blob is running? ARM memory fragmentation / consolidation is still beyond my understanding at this point. |
| 01:48:22 | Sammy | morning.... |
| 01:49:05 | TomW | good evening |
| 01:49:36 | jacques | hi Sammy |
| 01:51:01 | TomW | Yup. You need the SODIMM installed to program the ramdisk image when using blob-2.0.5-pre3. Well, that means that one more board hits the pile, until I can find the time to get the SODIMM working on it. |
| 01:51:19 | jacques | :-( |
| 01:53:09 | TomW | Ok. Got one more Tux left to go. Two out of five (so far) are on the pile, one I won't touch, another is for a customer proposal, .. leaves me with the unit with the RESET problem. But, it does boot with 16Meg. Wonder if this is going to give me grief when I upgrade the mainboard RAM to 32Meg? |
| 02:01:09 | jacques | im drooling over onboard 32MB |
| 03:53:54 | kergoth`zzz | is away: sleep |
| 04:42:50 | jacques | so quiet |
| 04:46:58 | TomW | busy |
| 04:48:49 | jacques | that's a good thing I suppose |
| 04:54:29 | TomW | getting better, only took half hour to remove the 4 RAM chips and clean the board in preparation for the new chips. |
| 05:19:46 | TomW | jacques: chips in, gotta clean the board and check for shorts / opens. |
| 05:28:40 | scanline | is back (gone 11:42:36) |
| 05:29:51 | jacques | TomW: :-) |
| 05:30:33 | TomW | Okay, it says that I now have 16M on the board. Now to change the value for MDCNFG and see if I can get 32Meg. |
| 06:20:19 | scanline | AFAICT busybox has no telnetd... what telnetd works well with uClibc? |
| 06:21:28 | LoneTech | could add a telnetd though. |
| 06:21:36 | jacques | one was added recently |
| 06:21:47 | LoneTech | there you have it (: |
| 06:21:51 | jacques | checkout latest buildroot |
| 06:21:53 | scanline | aha |
| 06:22:07 | scanline | well, this isn't for tuxscreen, it's for my dreamcast.. but i'll snag the CVS busybox |
| 06:22:46 | jacques | scanline: a telnetd was added to buildroot and I know telnetd patches have been submitted for busybox - HOWEVER |
| 06:23:01 | jacques | I am not sure the telnetd stuff is in busybox cvs at this time |
| 06:23:06 | scanline | ok |
| 06:23:16 | scanline | so i'll see what buildroot does :) |
| 06:23:19 | jacques | wish andersee was here - he would know - BZFlag too |
| 06:23:29 | LoneTech | which inetd is used? |
| 06:24:01 | LoneTech | makes sense to have telnetd in buxybox.. after all, it already has telnet |
| 06:24:10 | scanline | nods |
| 06:24:20 | LoneTech | I recall that telnetd was one of the largest things in an embedded system I built a few years ago |
| 06:24:50 | scanline | and to be usable, telnetd shouldn't need to be very big |
| 06:25:24 | LoneTech | busybox telnet source: /* use globals to reduce size ??? */ /* test this hypothesis later */ |
| 06:25:46 | jacques | looks like stuff from netkitbase :-\ |
| 06:25:52 | sjm | -rwxr--r-- 1 sjm sjm 18600 Nov 9 21:54 romfs/bin/telnetd* |
| 06:25:54 | LoneTech | I wonder about that. It goes in the uninitialized globals block, so the executable size isn't increased, anyway |
| 06:26:33 | jacques | yeah the stuff in the new buildroot is from netkitbase |
| 06:27:11 | scanline | hm.... is the DNS for ftp.us.kernel.org broke for anyone else here? |
| 06:27:24 | scanline | ibot: nslookup ftp.us.kernel.org |
| 06:27:24 | | ftp.us.kernel.org is 128.193.4.3 |
| 06:27:38 | scanline | i guess it's just me :( |
| 06:27:48 | jacques | ive had problems with their round robin or whatever it is before |
| 06:30:18 | scanline | that's better, switched DNS servers.. |
| 06:36:14 | scanline | makes a fresh buildroot... and watches his Athlon melt down |
| 06:36:15 | LoneTech | eww |
| 06:36:26 | LoneTech | busybox telnet changes 0xff to 0x7f |
| 06:37:52 | scanline | aak |
| 06:38:00 | scanline | now you'll need busybox uuencode ;P |
| 06:38:24 | LoneTech | notices nc.c |
| 06:38:38 | LoneTech | guess I can tolerate it with nc.c available |
| 06:42:38 | scanline | micah@yoshi:~/tuxscreen/buildroot-tux$ find | grep telnetd |
| 06:42:38 | scanline | micah@yoshi:~/tuxscreen/buildroot-tux$ |
| 06:43:28 | scanline | micah@yoshi:~/tuxscreen/buildroot-tux$ make telnetd |
| 06:43:28 | scanline | make: Nothing to be done for `telnetd'. |
| 06:43:33 | scanline | hmm |
| 06:48:47 | scanline | ok, found the makefile for telnetd |
| 06:49:15 | scanline | it just uses netkit-telnet |
| 06:49:41 | scanline | my this is hackish... |
| 06:49:45 | scanline | :) |
| 06:51:33 | LoneTech | is inetutils telnetd bigger than netkit telnetd? |
| 06:53:54 | scanline | good, followed the buildroot's makefile manually and got it compiled |
| 06:53:56 | scanline | -rwxr-xr-x 1 micah micah 65696 Apr 1 23:56 telnetd |
| 06:54:10 | LoneTech | strip it? |
| 06:54:17 | scanline | it is |
| 06:54:33 | scanline | micah@yoshi:~/src/netkit-telnet-0.17/telnetd$ file telnetd |
| 06:54:33 | scanline | telnetd: ELF 32-bit LSB executable, Hitachi SH, version 1 (SYSV), statically linked, stripped |
| 06:54:33 | LoneTech | shudders |
| 06:54:37 | scanline | but that's statically linked with uClibc |
| 06:54:49 | LoneTech | even so, telnetd should not be big |
| 06:54:56 | scanline | nods |
| 06:55:12 | LoneTech | bear in mind that it only uses sockets (tcp and anonymous unix) |
| 06:56:11 | scanline | yea |
| 06:56:56 | scanline | hmm... |
| 06:57:03 | scanline | telnetd: getpeername: Socket operation on non-socket |
| 06:57:42 | LoneTech | seems it wants to log the connecting end |
| 06:57:53 | LoneTech | and it's an inetd-based telnetd |
| 07:00:04 | scanline | no inetd in busybox... |
| 07:00:06 | scanline | grumbles |
| 07:01:19 | LoneTech | inetd is important imho. |
| 07:01:22 | scanline | yea |
| 07:01:42 | LoneTech | inetd+getty=emergency telnetd |
| 07:01:46 | scanline | briefly considers running pgserver from inetd ;) |
| 07:02:02 | LoneTech | doable. |
| 07:02:07 | LoneTech | tcp nowait |
| 07:02:12 | LoneTech | err |
| 07:02:15 | LoneTech | wait, I mean |
| 07:02:35 | scanline | that'd be kinda funny... on-demand GUI |
| 07:02:48 | LoneTech | can be done |
| 07:03:14 | LoneTech | only need to revamp the net code to use stdin for the listening socket |
| 07:04:21 | LoneTech | but I think it more interesting to have cli_c start a new pgserver for each picogui app run outside of picogui |
| 07:04:32 | LoneTech | with its current arch. |
| 07:04:49 | LoneTech | when it has "virtual screens" (remembered layouts) I'll probably change that opinion |
| 07:06:39 | scanline | depending on how well thought out eks' current project is, that might handle it |
| 07:07:23 | LoneTech | otherwise I'll just have to bang on it or him until it does ;) |
| 07:07:32 | scanline | yep |
| 07:09:18 | scanline | compiles netkit-base.. |
| 07:11:21 | scanline | -rwxr-xr-x 1 micah micah 58472 Apr 2 00:14 inetd |
| 07:11:43 | LoneTech | sighs |
| 07:11:56 | LoneTech | I could write one. |
| 07:12:09 | scanline | all i really need is netcat + a shell probably |
| 07:12:31 | LoneTech | netcat is in busybox |
| 07:12:47 | LoneTech | I want a bidirectional pipe redirector in sh |
| 07:13:17 | scanline | :) |
| 07:13:23 | scanline | you could use a fifo i suppose |
| 07:13:24 | LoneTech | something with a similar effect to mkfifo foo ; cmd1 < foo | cmd2 > foo |
| 07:13:30 | scanline | :) |
| 07:13:47 | scanline | cmd1 <> cmd2 |
| 07:13:52 | scanline | if only... |
| 07:14:03 | LoneTech | with that, and netcat, you can make a single shot net logon thingy. |
| 07:15:18 | jacques | heh |
| 07:15:27 | jacques | how does netcat compare with ncat ? |
| 07:15:33 | LoneTech | netcat=nc |
| 07:15:37 | LoneTech | dunno what ncat is |
| 07:15:47 | LoneTech | but nc is in busybox |
| 07:17:06 | scanline | Connected to twisty. |
| 07:17:06 | scanline | Escape character is '^]'. |
| 07:17:06 | scanline | Connection closed by foreign host. |
| 07:17:07 | jacques | ah |
| 07:17:34 | LoneTech | http://www.busybox.net/lists/busybox/2001-July/009493.html |
| 07:19:33 | scanline | i wonder if andersee knows about that |
| 07:22:08 | scanline | grr.. kernel oops |
| 07:24:31 | LoneTech | not good |
| 07:24:55 | scanline | ok.. sanity check- i have this line in my inetd.conf: |
| 07:24:59 | scanline | 24 stream tcp nowait /bin/echo echo foo |
| 07:25:09 | scanline | i'd think that "telnet twisty 24" should give me "foo" |
| 07:25:37 | scanline | it connects, but closes immediately |
| 07:25:38 | LoneTech | user /bin/echo? |
| 07:26:06 | scanline | it's in /bin/echo, a symlink to /bin/busybox |
| 07:26:18 | LoneTech | you have forgotten the user field |
| 07:26:54 | scanline | ohh |
| 07:27:01 | scanline | thanks |
| 07:27:44 | scanline | ok good, another kernel oops... |
| 07:27:50 | scanline | dusts off the System.map |
| 07:28:22 | jacques | does anyone know what archiver is used in most .exe files? |
| 07:28:28 | scanline | in __wake_up.. i'll have to bug the linuxdc folks |
| 07:28:34 | jacques | i used to know its like lharc or something |
| 07:29:22 | jacques | it allows decompressing of most .exe archives |
| 07:29:59 | LoneTech | jacques: .exe is one extension for M$-DOS executables. pretty much any archive can be made into .exe SFXes |
| 07:30:18 | scanline | the most common ones are zip |
| 07:30:39 | LoneTech | nowadays, InstallShield Cabinets are all too common. they suck in all ways. |
| 07:31:45 | jacques | LoneTech: i know that |
| 07:32:00 | jacques | i have ome oold files from like the late 80's |
| 07:32:11 | jacques | i have done tihs years ago but i forgot how i did it |
| 07:33:07 | LoneTech | well, lha handles lzh and lharc files, and probably even arc files. |
| 07:33:15 | LoneTech | unzip can deal with zip files |
| 07:33:24 | LoneTech | you may have to look ooz up for zoo files |
| 07:33:37 | LoneTech | "unp" is a simple frontend that tries them all, basically |
| 07:33:50 | LoneTech | and unsqueeze should handle sqz files |
| 07:35:29 | jacques | LoneTech: i think lha is what i want |
| 07:38:02 | jacques | hmm i guess these are not any common format |
| 07:38:31 | jacques | I remember thinking I was going to have to install a dos emu to uncompress some .eve files ten finding out I could just use lha or something on them |
| 07:38:33 | LoneTech | unarj and unrar may also interest you |
| 07:38:44 | scanline | oh.. it does work |
| 07:38:47 | LoneTech | lha is free including source |
| 07:38:47 | scanline | micah@yoshi:~/dreamcast/kernel-boot.nfs$ nc twisty 24 |
| 07:38:47 | scanline | foo |
| 07:39:14 | scanline | telnetd must be missing whatever it's trying to log in with |
| 07:40:40 | scanline | throws security to the wind |
| 07:41:12 | LoneTech | telnetd would use login.telnetd or something like that |
| 07:41:36 | scanline | i tried to set the login program to /bin/sh |
| 07:41:51 | scanline | didn't change the behavior |
| 07:42:11 | LoneTech | I don't know what it does |
| 07:42:27 | scanline | k |
| 07:43:45 | scanline | makes do with sticking /bin/sh directly in inetd |
| 08:22:11 | TomW | LoneTech: Are you trying to get a particular telnetd working, or *any* telnetd would do? |
| 08:22:38 | LoneTech | ask scanline - he wanted it. |
| 08:23:09 | scanline | TomW: any telnetd.. though i've for the moment given up. I just wanted to see the stdout of pgserver, but i'm using graphical debugging instead |
| 08:23:10 | TomW | scanline: Micah, are you particular as to which telnetd you want to get working? |
| 08:23:57 | TomW | scanline: didn't you see my post the TuxScreen list? Complete with source(s) for inetd, telnetd, inetd.conf, etc. |
| 08:24:23 | scanline | mm... nyet. I'll look at that though when i get sick of watching guru messages |
| 08:25:36 | TomW | Okay, the stuff is in the list. I had patched the buildroot-tux/Makefile (that was before BZFlag modified the Makefile system of the current CVS). |
| 08:25:41 | LoneTech | picogui's guru messages are much friendlier than amiga's.. bloat :P |
| 08:26:03 | | BZFlag was last seen on #bzflag 5 hours, 12 minutes and 9 seconds ago, saying: later [Tue Apr 2 04:13:54 2002] |
| 08:26:03 | TomW | ibot: seen BZFlag |
| 08:26:31 | LoneTech | my friend who fried his tuxscreen has decided to order two more |
| 08:27:37 | TomW | BZFlag: did you happen to get any gerber data on the Shannon CPU board? Or gifs of the foil layers? |
| 09:07:39 | BZFlag | drops in while boxing phones. |
| 09:08:01 | BZFlag | TomW: no gerber data at all. can't seem to find anyone that has them |
| 09:08:03 | LoneTech | hi BZFlag (: |
| 09:08:25 | BZFlag | hey, you were trying out the inetd and telnetd I added to buildroot? |
| 09:08:41 | LoneTech | no, scanline wanted one for his dreamcast |
| 09:09:07 | LoneTech | my friend fried his tuxscreen by getting the power supply mirrored |
| 09:09:19 | LoneTech | he blames poor english skills |
| 09:10:03 | LoneTech | perhaps you could add a warning on http://tuxscreen.net/wiki/view/Main/PowerSupply clarifying that the pictures are seen as when looking into the male connector on the cable? |
| 09:11:13 | LoneTech | I hope I'm not being too annoying. I find it quite sad that his broke :( |
| 09:18:26 | LoneTech | wb scanline |
| 09:18:59 | scanline | thanks |
| 09:19:03 | scanline | nvidia's fault |
| 09:20:51 | LoneTech | as usual |
| 09:27:13 | BZFlag | LoneTech: bummer. you can edit the page ;-) |
| 09:28:06 | BZFlag | edits the page anyway |
| 09:28:18 | LoneTech | oh |
| 09:34:52 | LoneTech | thank you |
| 10:23:20 | TomW | scanline: you still here? |
| 10:23:28 | scanline | yepp |
| 10:23:43 | TomW | # cat /proc/meminfo |
| 10:23:43 | TomW | total: used: free: shared: buffers: cached: |
| 10:23:43 | TomW | Mem: 37081088 3665920 33415168 0 4096 888832 |
| 10:23:43 | TomW | Swap: 0 0 0 |
| 10:23:43 | TomW | MemTotal: 36212 kB |
| 10:23:44 | TomW | MemFree: 32632 kB |
| 10:23:46 | TomW | MemShared: 0 kB |
| 10:23:48 | TomW | Buffers: 4 kB |
| 10:23:50 | TomW | Cached: 868 kB |
| 10:23:52 | TomW | SwapCached: 0 kB |
| 10:23:54 | TomW | Active: 584 kB |
| 10:23:56 | TomW | Inactive: 1048 kB |
| 10:23:58 | TomW | HighTotal: 0 kB |
| 10:24:00 | TomW | HighFree: 0 kB |
| 10:24:02 | TomW | LowTotal: 36212 kB |
| 10:24:04 | TomW | LowFree: 32632 kB |
| 10:24:06 | TomW | SwapTotal: 0 kB |
| 10:24:08 | TomW | SwapFree: 0 kB |
| 10:24:34 | TomW | Lot of work, but it will be easy for others to duplicate it. 32 Meg of DRAM on the mainboard, now all I need is a 32Meg SODIMM and I will have 64Meg in the Tux! |
| 10:24:49 | scanline | wow! |
| 10:24:55 | scanline | too much :P |
| 10:25:29 | TomW | MT4LC4M16R6-TG6 rams, replace the ones on the board, then 3 cuts & 3 jumpers. |
| 10:25:46 | TomW | Oh, an a single byte change to blob. |
| 10:26:28 | TomW | Time for me to put this unit together before it gets hurt like the last one. (stray peice of solderwick). |
| 10:27:03 | scanline | ow |
| 10:28:05 | TomW | Well, it worked out for the best. I had to trace two signals out to see where they went and decide the best place to make the cuts. That board doesn't have many chips left on it. :-D |
| 10:43:26 | scanline | hehe |
| 10:43:32 | TomW | ?? |
| 10:43:45 | scanline | 16-minute delayed laugh :P |
| 10:44:14 | scanline | is trying to figure out how to do conditionals in automake... |
| 10:44:38 | TomW | wishes Micah Luck |
| 10:44:51 | scanline | thanks |
| 10:45:47 | scanline | ah... |
| 10:45:52 | scanline | there's a manual for automake :) |
| 10:48:31 | TomW | depends how cryptic the thing is. I hope it is not written by Dan Bernstein (QMail author), his manpages can be absolutely cryptic. |
| 10:51:21 | scanline | it's not bad |
| 11:06:50 | LoneTech | TomW: cool (: |
| 11:07:16 | TomW | Yes, now if I can only document what it was that I had to do.. |
| 11:07:57 | LoneTech | canonical cryptic man page section: |
| 11:08:05 | LoneTech | The backreference \n, where n is a single digit, matches |
| 11:08:05 | LoneTech | the substring previously matched by the nth parenthesized |
| 11:08:06 | LoneTech | subexpression of the regular expression. |
| 11:08:17 | scanline | :) |
| 11:33:33 | fontenot | TomW: document it before you forget! |
| 11:34:23 | fontenot | TomW: how hard are those chips to get? and how expensive? |
| 11:34:39 | TomW | Doing it now. Those chips are used in the PalmIIIxe |
| 11:34:46 | TomW | 7.35 apeice. |
| 11:35:52 | fontenot | wowee |
| 11:36:09 | TomW | 3 cuts, 3 jumpers. |
| 11:36:30 | fontenot | I'm more worried about the smt work :-) |
| 11:36:52 | fontenot | i wonder if anyone will setup shop doing conversions |
| 11:37:06 | fontenot | for us poor smt-challenged folk |
| 11:39:24 | TomW | how do you get <br> (new lines) into the stupid wiki? |
| 11:41:06 | TomW | http://TuxScreen.net/wiki/view/Main/32Meg |
| 11:41:11 | fontenot | checking |
| 11:42:23 | fontenot | TomW: see http://tuxscreen.net/wiki/view/Main/WikiFormat |
| 11:44:13 | fontenot | you would probably want to use CODE or PRE to format yourboot logs |
| 11:45:16 | TomW | very clumsy. |
| 11:47:42 | fontenot | yeah i remember thinking that when i did the xfree howto |
| 11:52:45 | TomW | !@$#!@!23% Wiki!!! |
| 11:53:17 | TomW | I can do shit faster using vi and html. |
| 11:54:05 | fontenot | that's what i did |
| 11:54:17 | fontenot | then i just pasted the result into the wiki and fine tuned it |
| 11:56:08 | fontenot | of course the fine tuning took longer than everything else put together, including actually building xfree :-) |
| 11:56:11 | TomW | Ummm, you added the <HTML><HEAD></HEAD> to the doc? |
| 11:56:29 | fontenot | not that im aware of |
| 11:56:40 | fontenot | i hit the edit button but didnt do anything |
| 11:58:22 | fontenot | i thought i could get to the edit help from edit |
| 11:58:22 | TomW | 5,114c3 Paragraphs |
| 11:58:23 | TomW | Do not indent paragraphs |
| 11:58:23 | TomW | Hard-wrap to window width and fill as needed |
| 12:07:44 | TomW | I have no intention of fscking with this peice of bizzare crap any longer. I am putting it on a webpage at my site. >:O |
| 13:15:56 | TomW | fontenot: http://www.openhardware.net/TuxScreen/32MegDram.html |
| 13:16:58 | TomW | fontenot: I put this board back into the plastic case, once I had taken the photo. There is no way I want to blow this board up after changing those DRAM chips. |
| 14:51:27 | cpenguin_away | is back (gone 14:40:50) |
| 15:29:30 | pattieja | morning all |
| 15:29:34 | scanline | is away: I'm busy |
| 15:29:47 | scanline | is back (gone 00:00:12) |
| 15:32:56 | pattieja | anybody here worked with the TransNote? |
| 16:12:47 | GPSFan | TomW: could you look at ftp://www.desertweyr.com/pub/Image02c.jpg and ftp://www.desertweyr.com/pub/Image03b.jpg and see if those are the correct loactions for the cuts & jumpers for your 32M mod. |
| 16:13:29 | TomW | Look at my site again, I just finished up the section on cuts & jumpers. |
| 16:13:46 | GPSFan | ah, will do.. |
| 16:14:43 | TomW | can I have the un-altered photo? My focal range is not as good as your camera. |
| 16:15:15 | GPSFan | sure, I'll upload them in a bit. |
| 16:16:16 | TomW | You have everything correct. |
| 16:16:32 | TomW | Are you building your own blob? |
| 16:17:00 | scanline | is away: C++ for dummies |
| 16:17:04 | GPSFan | last one I built was 2.05 pre-2, worked fine, where did you get pre-3? |
| 16:17:24 | TomW | from blob cvs. |
| 16:19:10 | GPSFan | k, the images are up ftp://www.desertweyr.com/pub/Image02.jpg & ftp://www.desertweyr.com/pub/Image03.jpg |
| 16:20:34 | GPSFan | TomW: I use a closeup lens & a "ring light" sometimes the focus isn't perfect but usually good enough. |
| 16:25:12 | TomW | The blob changes are now on the webpage, there is just one byte that needs to be changes. |
| 16:25:18 | TomW | s/changes/changed/ |
| 16:26:18 | TomW | It took me four hours to track down where the cuts had to be done. |
| 16:26:48 | TomW | good feeling though, I have a 40Meg Tux, now to go look for a 32M SODIMM. |
| 16:27:52 | GPSFan | looks really good, any new features in pre-3, I have some 32M SODIMMS comming tomorrow, don't know wheather they will work, but they were cheap, if they do, or can be made to I'll post the results. |
| 16:29:29 | GPSFan | I'll try to incorporate your mods into a modified schematic for the memory section. |
| 16:30:37 | TomW | Can you give me a photo of the area that I am showing on my page? I would attempt to paste the two images you had, but.. |
| 16:31:02 | TomW | That is such an excellent camera setup you have there. |
| 16:31:26 | GPSFan | sure, I just took those to verify the placement, I can try to duplicate your image.. standby.. |
| 16:31:39 | TomW | great! |
| 16:38:46 | GPSFan | TomW: try this, it's full res, so you might want to reduce the jpeg quality a bit to reduce the file size after you have added your text etc. ftp://www.desertweyr.com/pub/Image05.jpg |
| 16:40:01 | TomW | THAT is just great! Now I can do a decent cut & jumper illustration and put it on the Wiki. Thank you! |
| 16:40:09 | TomW | Are you going to mod your unit? |
| 16:41:55 | GPSFan | np, the camera & setup are something anyone can duplicate eaisly, the real trick is enough light and the closeup lens. Yes I will mod my tux after I see if the SODIMMS I'm getting will work. |
| 16:42:48 | TomW | I think that you will have to modify the blob so that the SA-1100 will output the 12 columns instead of the default of 10 columns. |
| 16:43:02 | TomW | Then your 32Meg SODIMMs will work. |
| 16:43:28 | TomW | brb |
| 16:44:51 | TomW | I had to shut X down. When I do a lot of graphics work, it chews the RAM up and I start running in swap all the time. |
| 16:44:53 | GPSFan | depends on the SODIMM, if it's built with 4x16 chips it probably will work eaisly, 8x16 maybe tricky. I'll know tomorrow. |
| 16:45:16 | TomW | Where did you order the SODIMMs from? |
| 16:45:46 | scanline | is back (gone 00:28:46) |
| 16:46:39 | GPSFan | http://www.computersurplusoutlet.com/showproduct.asp?category=6&subcategory=30 at 6.99 each it was hard to resist, if my Tux doesn't like them, I have a notebook that takes 72pin EDO. |
| 17:01:21 | TomW | With our luck, if you ask the salesman if they are 4K refresh the only answer you would get is "Huh?". |
| 17:02:01 | GPSFan | yep, I figured it was worth the 6.99 each to know the truth. |
| 17:11:29 | TomW | I see Yahoo had some 32M 72pin 4K refresh on an Auction, I wonder about EBay? |
| 17:28:38 | TomW | GPSFan: here is your 32M SODIMM, $15 per. http://www.18004memory.com/category.asp?catid=2 |
| 17:31:58 | TomW | GPSFan: I just ordered 6 of them. |
| 17:32:34 | GPSFan | these? SNED32U/032/4K |
| 17:33:25 | TomW | right, they say they have "unlimited" stock. He said that they get orders for them so often they have someone manufacturing the SODIMMs for them. |
| 17:34:10 | TomW | Be careful to insist on the 4K refresh, he tried to sell me 2K refresh. |
| 17:35:30 | GPSFan | so they are sodimms, the page says simm. yeah, others have had problems getting substitute parts from various memory suppliers. I'll get some ot those if the ones I ordered don't pan out. |
| 17:35:48 | TomW | cool! maybe by the end of the week I can have two 64M tuxes. |
| 17:36:19 | TomW | ah shit, I'd better call to make sure of that partnumber. brb |
| 17:36:32 | GPSFan | if they are 4x16 chips it owuld be cheaper to buy them and pull of the chips 4 for $15 vs $7.35 each. |
| 17:38:53 | TomW | you ever try to salvage smt parts? it is a pain. |
| 17:40:28 | GPSFan | a hot air desoldering tool does wonders, I once took off a 208 pin flatpack and was able to re-use it. An extreme case to be sure. but I've pulled off that style memoy chip before with fairly good results. |
| 17:43:22 | TomW | I have a zephrytronics hot air, but I never thought to use the pencil to take parts off the board. I usually use their low-temp alloy. Hmmm. |
| 17:44:11 | TomW | BTW: LGED32U/032/G3V60 @ 29.95 per. More expensive, but they had them. |
| 17:44:51 | GPSFan | where I used to work, I had a Liester, it had special tips for SMT devices, heated all pins at once. Hako makes a cheaper, I now have an Edysen which works ok too. |
| 17:44:52 | TomW | Thanks for catching that for me. I am trying to get back on a day schedule and have been awake for 28 hours.. |
| 17:45:22 | GPSFan | so ie that the correct p/n for the Tux compatible 4k 32M 72pin SODIMMs? |
| 17:45:32 | GPSFan | s/ie/is |
| 17:46:09 | TomW | Yeah, I saw some of those units, I didn't like the amount of heat that it threw onto the board. was afraid of de-laminating the fiberglass. |
| 17:46:19 | TomW | yes that is the number he gave me. |
| 17:46:49 | TomW | I expect that they will be compatible. I shall find out. |
| 17:46:53 | GPSFan | so SNED32U/032/4K really is a simm then, bummer. |
| 17:47:04 | TomW | yeah, cheap too. |
| 17:47:49 | TomW | Oh well, seems like any project I undertake anymore always costs me at least $1000. |
| 17:48:58 | GPSFan | yea, at $30 each, the $7.35 from Arrow isn't bad, I wouldn't bother taking off chips for that. These projects are all money sinks, but hey, they are fun.. |
| 17:50:19 | LoneTech | seems the touchscreen is out of order |
| 17:50:30 | TomW | I was thinking of taking the chips off of an 8M SODIMM, mounting the 32M chips on it, and running additional 6 mires. But, that is a lot of work just trying to save money. |
| 17:50:44 | TomW | LoneTech: ?? you broke it? |
| 17:51:10 | LoneTech | don't think so |
| 17:51:25 | LoneTech | the kernel just doesn't send any data through /dev/usb1x00-ts |
| 17:51:39 | LoneTech | will look into it later |
| 17:52:10 | GPSFan | I'd do it if there was no other way, but there seem to be other options. |
| 17:52:36 | TomW | I think that you have to use the IOCTL to get data from it. I don't think that it will stream data. (cat /dev/ucb1x00-ts don't work) |
| 17:53:51 | TomW | I was even considering piggy-backing 4 more ram chips on the cpu board over top of the four that were there. :-/ |
| 17:56:54 | GPSFan | bbl |
| 18:50:19 | scanline | is away: circuits lab |
| 19:03:28 | andersee | Morning Tim |
| 19:05:13 | TomW | BZFlag: 40Meg of ram on my tux. |
| 19:06:33 | andersee | TomW: I have 256 Meg ram on my netwinder. ;-) |
| 19:06:47 | TomW | :-P |
| 19:07:17 | andersee | TomW: I had to hack a SO-DIMM for that too... |
| 19:08:17 | TomW | heh, yeah, I was considering soldering another 32 Meg on top of the existing chips, then I thought better about it. |
| 19:08:49 | andersee | TomW: Admittedly, the SO-DIMM hack I did looks much easier than your hack... |
| 19:09:38 | TomW | no, not at all. GPSFan gave me a better photo to work with, i haven't posted it yet. |
| 19:11:06 | TomW | It is a simplle hack, it just took many hours to work it out so that others can duplicate it. |
| 19:12:52 | TomW | andersee: I haven't tried to figure out how to get images on the Wiki. |
| 19:30:42 | BZFlag | TomW: you have a sf account? has it been added? |
| 19:31:00 | BZFlag | then you scp your files to the server and add the url in a wiki page. |
| 19:31:07 | TomW | sf? no. I am linking to my server. |
| 19:31:24 | BZFlag | course the url could point to anywhere on the net, but uploadind to sf is preferred. |
| 19:32:31 | BZFlag | what's your sf account name? (you have one, I presume) |
| 19:33:43 | TomW | No, I have only have account for Wiki. |
| 19:36:42 | BZFlag | get a free sf account, and tell me the name. I'll add you to the project. |
| 19:37:19 | BZFlag | your comments on: http://www.openhardware.net/TuxScreen/32MegDram.html on the bottom have the same comment for both memory configs. |
| 19:38:05 | TomW | TomWalsh |
| 19:38:16 | BZFlag | did you change the kernel to use more mem regions? |
| 19:38:40 | TomW | No, all you have to do is set mdcnfg in blob. Then it goes. |
| 19:38:47 | BZFlag | else it looks like it would only see like 35M or so. |
| 19:39:14 | BZFlag | I think there is a max number of regions, and you have 10 of them which iirc is over the default max. |
| 19:39:56 | TomW | Memory map: |
| 19:39:56 | TomW | 0x01000000 @ 0xc0000000 (16 MB) |
| 19:39:56 | TomW | 0x01000000 @ 0xc8000000 (16 MB) |
| 19:39:56 | TomW | 0x00100000 @ 0xd0000000 (1 MB) |
| 19:39:56 | TomW | 0x00100000 @ 0xd0400000 (1 MB) |
| 19:39:57 | TomW | 0x00100000 @ 0xd0800000 (1 MB) |
| 19:39:58 | BZFlag | TomWalsh is a new sf account? |
| 19:39:59 | TomW | 0x00100000 @ 0xd0c00000 (1 MB) |
| 19:40:01 | TomW | 0x00100000 @ 0xd8000000 (1 MB) |
| 19:40:03 | TomW | 0x00100000 @ 0xd8400000 (1 MB) |
| 19:40:05 | TomW | 0x00100000 @ 0xd8800000 (1 MB) |
| 19:40:07 | TomW | 0x00100000 @ 0xd8c00000 (1 MB) |
| 19:40:09 | TomW | yes. |
| 19:40:18 | BZFlag | yeah, I wsaw that. blob handles it. |
| 19:40:24 | BZFlag | it's the kernel I'm talkign about. |
| 19:41:00 | TomW | ok, I will be getting 32Meg 4K refresh SODIMMs by this weekend. I will try that out then (64M). |
| 19:41:27 | BZFlag | remind me to add your sf account later. looks like it needs some time to settle in. |
| 19:41:55 | BZFlag | with the 32M sodimm you should have 4 16M mem regions and everything should be fine. |
| 19:41:59 | TomW | The key is that you have to insist on 4K refresh on the SODIMM, they are 12 columns by 10 rows. The organization matches the settings of the DRAM controller. |
| 19:43:08 | BZFlag | /* 0x19A cnt refresh, 1 cycle CAS latch, 10 CASb4RAS refresh, */ |
| 19:43:08 | BZFlag | /* 4 RAS precharge, memory clock == cpu clock, 12 row address, */ |
| 19:43:08 | BZFlag | /* enable all banks */ |
| 19:43:08 | TomW | sf account is now activated. |
| 19:43:18 | BZFlag | is you comment for both configs, no? |
| 19:44:16 | TomW | Are you sure abt the 0x194 I thought that it would be 0xcd. |
| 19:44:57 | BZFlag | added account |
| 19:45:49 | BZFlag | not sure, no. other than it works as is. |
| 19:46:10 | BZFlag | should be able to run at 206mhz, but it does not at present. feel free to mess with it. |
| 19:46:39 | BZFlag | course 194mhz uses faster bus rate than 206 iirc. |
| 19:47:02 | TomW | doh! same number of rows to refresh, refresh rates stay the same. |
| 19:48:04 | TomW | just looking around http://sourceforge.net/projects/tuxscreen to see the info on scp. |
| 19:48:19 | BZFlag | both comments you list have "12 row addresses" in the comments. that's not right is it? |
| 19:48:40 | BZFlag | ssh tomwalsh@tuxscreen.net |
| 19:48:47 | TomW | No, it is 12 columns + 10 rows. |
| 19:49:02 | BZFlag | it? I'm talking about "they" |
| 19:49:15 | TomW | Ok, the memsetup-sa1100.S file, let me look. |
| 19:50:09 | BZFlag | 1f vs 3f should have different comments |
| 19:50:48 | TomW | fixed |
| 19:51:07 | BZFlag | sweet. |
| 19:51:22 | BZFlag | so 2 phones for one upgraded phone? ;-) |
| 19:51:34 | TomW | Can do that. yes. |
| 19:51:44 | BZFlag | cool. let's do it. |
| 19:51:44 | TomW | Just send the CPU board. |
| 19:52:12 | TomW | I still have 12 RAMs left. I set 4 aside for your board. |
| 19:52:21 | TomW | s/I/I'll/ |
| 19:53:10 | BZFlag | cool. I'll send 1 phone + one main board on to ya then. |
| 19:53:37 | TomW | Do you need a 32Meg SODIMM? |
| 19:53:45 | BZFlag | got one. and a 64M too. |
| 19:56:04 | TomW | I don't think the 64M will work for you. Means more cuts & jumpers to get it working, that is if the ARM can be programmed to 13x10 or 12x11 organization. |
| 19:56:56 | TomW | Wait! The SA1100 manual does say that the DRAM Controller can handle 4 banks of 128 meg, doesn't it? |
| 19:57:44 | sjhill | andersee: hey Erik |
| 19:58:18 | TomW | RAS lines are there... |
| 19:59:15 | TomW | BZFlag: yes, a 64Meg SODIMM would work in the Tux! It may work for you now, just set the mdcnfg register and try it. |
| 19:59:42 | BZFlag | won;t work at present |
| 20:00:14 | BZFlag | the next address pin is reported to map to a reset line or something. I expect that one line would have to be cut and jumpered. |
| 20:00:45 | BZFlag | the box won't boot at all with the sodimm installed. |
| 20:01:09 | BZFlag | look at the difference between the 32M E8 and E4 pinouts. |
| 20:01:26 | BZFlag | the 64M is mapped like the E8 |
| 20:07:41 | andersee | sjhill: hi |
| 20:07:50 | sjhill | ELF header=0x2aaa8000 |
| 20:07:51 | sjhill | First Dynamic section entry=0x2aaa80cc |
| 20:07:51 | sjhill | About to do library loader relocations. |
| 20:07:51 | sjhill | Done relocating library loader, so we can now |
| 20:07:51 | sjhill | use globals and make function calls! |
| 20:07:52 | sjhill | GOT found at 0x2aaae000 |
| 20:07:54 | sjhill | Lib Loader:(0x2aaa8000) /usr/mipsel-linux-uclibc/lib/ld-uClibc.so.0 |
| 20:07:56 | sjhill | searching for library: 'libc.so.0' |
| 20:07:58 | sjhill | searching in ldso dir: /usr/mipsel-linux-uclibc/lib |
| 20:08:00 | sjhill | Segmentation fault (core dumped) |
| 20:08:05 | sjhill | getting close :) |
| 20:08:06 | andersee | sjhill: Very cool! |
| 20:08:25 | andersee | sjhill: BTW, dunno if you saw, I checked in a fix for a buffer overflow in ld_string.h |
| 20:08:45 | sjhill | i'm about 3 days behind on cvs, i will check out now |
| 20:09:18 | andersee | sjhill: The only other main fix is a printf threading deadlock (i.e. not related) |
| 20:10:27 | sjhill | k |
| 20:11:18 | andersee | sjhill: I'm planning on cutting a bug-fix release later today. So if you have anything you want in... |
| 20:11:45 | sjhill | yup, i should have something on the list in about an hour |
| 20:19:43 | TomW | BZFlag: Ok, I am starting to figure out sf. The picture is now on sf site. |
| 20:19:56 | BZFlag | cool. |
| 20:21:06 | TomW | /usr/bin/locate helps. |
| 20:25:06 | BZFlag | heh ;-) |
| 20:26:25 | BZFlag | TomW: http://tuxscreen.net/download/CT06.pdf |
| 20:28:21 | TomW | tnx |
| 20:30:01 | scanline | is back (gone 01:39:42) |
| 20:33:46 | BZFlag | contrast with: http://www.avanttechnology.com/pdf's/TNC3208N12AG6A0.PDF |
| 20:35:18 | sjhill | andersee: it should be on the mailing list by now. i'm getting ready to take off here in about an half hour. how about looking the patch over pls? |
| 20:46:35 | andersee | k, will do |
| 20:47:37 | BZFlag | TomW: pins 33,45,46 differ from what I can see. |
| 20:59:34 | TomW | 33 & 45 and 34 & 44 are correct. I suspect that pin 46 is the #0E line. On a 12 column RAM, A12 is actually the #OE (on 13 column RAM, it is located elsewhere). |
| 21:01:13 | TomW | I really don't see any problems here. Other than the PRDx lines, not familiar with those. I suspect these are the sense lines for the size of the memory, they may be shoring the Address lines on them to gnd / Vdd. |
| 21:01:51 | TomW | buzz the PRDx lines of the SODIMM out and make sure they are NOT connected to either vss / vdd. |
| 21:01:59 | BZFlag | is there an address line on the main board on pin 46? |
| 21:02:11 | andersee | sjhill: I don't see any problems with your patch. Applied. |
| 21:03:10 | BZFlag | TomW: I don't have the equipment here to do that with. =( |
| 21:03:13 | sjhill | andersee: thx! |
| 21:03:27 | sjhill | finishes his reservations for OLS |
| 21:03:37 | TomW | No, it is #OE (output enable), look at the datasheet for the devices on the card. NO, just an ohmmeter. |
| 21:03:40 | BZFlag | I could send along both sodimms to you with the board to modify if you want to play with em and send em back |
| 21:03:57 | andersee | OLS? |
| 21:03:57 | TomW | send 'em. |
| 21:04:02 | BZFlag | ok |
| 21:04:26 | TomW | BZFlag: I am pretty sure that the problem is with the PRD lines. |
| 21:04:49 | sjhill | andersee: Ottawa Linux Symposium |
| 21:05:18 | TomW | IIRC, the PRDx lines are selectively tied to gnd so that the processor can read those lines on a port and see what size it is. |
| 21:05:24 | BZFlag | TomW: the machine would at least boot if those were wrong, no? |
| 21:05:44 | TomW | No, it wouldn't if they were hardwired to gnd. |
| 21:05:54 | BZFlag | might not find the right mem config, but blob would not be running there anyway |
| 21:06:14 | BZFlag | pokes around fo the sodimms |
| 21:06:14 | TomW | Think abt it, you would be shorting some / one of the address lines to gnd. |
| 21:06:33 | BZFlag | ah, true. |
| 21:08:30 | BZFlag | finds stacks of 8M sodimms, but no 32/64 |
| 21:08:34 | TomW | Look at the 8Meg SODIMM for the Tux, see those three surface mount jumpers? One 0 ohm jumper is installed. Short the next (center) jumper's pads together, if I am correct, the Tux won't boot. |
| 21:08:41 | BZFlag | I think my 32 and 64 are at the office. |
| 21:11:55 | TomW | I just did it, I shorted the center PRData jumper and the Tux refused to boot, when I removed the jumper, it booted. |
| 21:13:33 | TomW | The problem with your 64 meg SODIMM is that it is configured for laptop use, not for the Tux. locate the PRD jumpers and remove them from the board. The Shannon CPU "discovers" the memory by running a quick & dirty pattern test as it cannot read the PRData lines. |
| 21:14:04 | BZFlag | considers a run to the office to grab the sodimms |
| 21:14:34 | BZFlag | what are the prd lines mapped to on the main board? |
| 21:15:07 | TomW | Address bus signals! heh |
| 21:15:53 | TomW | before they had EEPROMs on SDRAMs, they had PRDATA lines on DRAMs. |
| 21:17:34 | BZFlag | so cutting the address line on the mainboard should work as well? |
| 21:17:50 | BZFlag | well, cutting it from the sodimm slot... ;-) |
| 21:18:47 | BZFlag | throws a board on the scope... |
| 21:18:55 | TomW | BZFlag: BTW, we have old parts here in the Tux, they may have used jumpers at the time. Your more modern part may have only pcb traces that are grounding the PRDATA lines. Hey! Go to Radio Shack, spend $15 and get an Ohmmeter. ;-) |
| 21:19:20 | TomW | NO, DON"T CUT THE MAINBOARD!!!!!! |
| 21:20:07 | TomW | You're obviously a programmer, am I right? |
| 21:20:38 | BZFlag | heh, yeah. |
| 21:21:14 | TomW | heh, thought so. I am 60/40 (like solder), 60% software, 40% hardware. |
| 21:21:38 | TomW | ... some years it goes 40/60. |
| 21:22:26 | BZFlag | ok, so why not cut the trace to pin 69 on the sodimm socket? |
| 21:22:37 | BZFlag | shouldn't that accomplish the same thing? |
| 21:23:14 | TomW | BZFlag: Tim, tell you what, keep the X-acto knife in the drawer, send the SODIMMs along and I'll see what I can do to get 'em working. I would rather do it than have you cut the wrong traces. |
| 21:23:59 | BZFlag | I'm not sitting here with knife in hand. just asking conceptually. |
| 21:25:49 | TomW | heh, when you said something about cutting address lines off the mainboard you had me worried. ;-) |
| 21:26:24 | BZFlag | I was just asking, not planning on doing it. |
| 21:27:01 | BZFlag | well at least not until we're fairly certian it would work. |
| 21:27:02 | TomW | ... but then, you have replacement boards. Say, how many of those tux'es you still have there, just curious. |
| 21:27:21 | BZFlag | glances at the topic... over 300 |
| 21:27:57 | BZFlag | I'd guess 320 or so at present. can count if you like. /me goes to count. |
| 21:29:10 | scanline | is away: ZzZzZz.... |
| 21:29:37 | | 331 |
| 21:29:37 | BZFlag | ibot 27 * 12 + 7 |
| 21:30:15 | BZFlag | 331 plus some without boxes. |
| 21:30:54 | TomW | Same pinout on the datasheet that you sent over, isn't it? |
| 21:31:49 | TomW | umm. did you just send another PDF or is my mind losing traction? |
| 21:32:14 | BZFlag | yeah one for a 32M and one for the 64M |
| 21:32:34 | BZFlag | http://tuxscreen.net/download/CT06.pdf http://www.avanttechnology.com/pdf's/TNC3208N12AG6A0.PDF |
| 21:32:54 | BZFlag | well, that's a 64 and 32 I think. ;-) |
| 21:33:23 | TomW | They are both the same pinouts. The only thing that would change is the number of columns & rows on the RAMs / number of RAMs. |
| 21:33:40 | BZFlag | http://www.tuxscreen.net/download/schematics/dram.pdf |
| 21:34:19 | BZFlag | well, something causes the machine to lockup. could be a jumper as you suspect. |
| 21:34:41 | TomW | yes, already got that one. |
| 21:34:44 | BZFlag | ok, I'm going to run to the office and pick up the sodimms. back in 30 or so. |
| 21:34:52 | TomW | later. |
| 22:09:50 | byron1 | hello all .. having trouble getting microwin to compile in buildroot .. if i compile with sharedlibs on then i get undefined reference errors ... |
| 22:10:32 | byron1 | with sharedlibs off it compiles ok, but image does not work on tuxscreen .. get 'unable to open graphics server' error |
| 22:10:41 | byron1 | any info on this on the net somewhere? |
| 22:24:41 | CosmicPenguin | Is anyone else getting the shared lib problems? |
| 22:24:54 | CosmicPenguin | THat seems to be a fairly new phenonomon... perhaps its a gcc 3.0 problem? |
| 22:25:01 | CosmicPenguin | I mean, on other packages |
| 22:28:49 | byron1 | currently using 2.95.3 toolchain |
| 22:29:24 | BZFlag | TomW: back. |
| 22:29:57 | TomW | here, just keeping my eyes open. I am attempting to get back on a daytime schedule. |
| 22:34:39 | TomW | I've been thinking about a 10G notebook drive I have here doing nothing. I believe that I could add some modest electronics to J602 and add that IDE drive to the Tux. |
| 22:34:46 | BZFlag | ok 32M that works has jumpers open for 69,70,71 one jumper installed for a pin I can't trace (well, can't see the trace) |
| 22:36:02 | BZFlag | 64M has 4 jumpers, one closed. can't see the tracings any of them. |
| 22:37:06 | TomW | ohm out the PRDATA lines, ins 11, 66, 67, 68, 69, 70 and 71. See if any of them go to ground. |
| 22:38:29 | TomW | REVISE that list to: pins 11, 69, 70 and 71. See if any of them go to ground. |
| 22:40:21 | TomW | Hint, use straight pins (the kind you get with a new dress shirt) and some alligator leads to grasp the pins. Makes it easier to do. |
| 22:41:36 | BZFlag | http://www.avanttechnology.com/pdf's/TNC3208N12AG6C0.PDF claims 32M should be pd1 and 2 gnd |
| 22:44:08 | TomW | yes, that is what it is, density signalling lines for laptops to read from. |
| 22:48:18 | BZFlag | 67,68 on the 32; 11,66,68 on the 64 |
| 22:49:08 | BZFlag | which does not match up... =( |
| 22:49:55 | TomW | What do you have grounded on those pins? |
| 22:50:16 | BZFlag | huh? what do you mean what do I have? |
| 22:50:31 | BZFlag | those pins ring through to ground. |
| 22:50:47 | BZFlag | I'm using pin1 and/or ping 72 for gnd |
| 22:51:25 | TomW | All of them? all pins 11, 69, 70 and 71 on the 32M are grounded? |
| 22:57:34 | BZFlag | 67,68 on the 32; 11,66,68 on the 64 |
| 22:59:18 | TomW | See if you can remove the gnd from pin 11 on the 64Meg, the other pins (66 & 68) you can ignore. Then, I suspect that your Tux will boot with that RAM. |
| 23:06:32 | TomW | Here is a definition I found: PD: Presence Detect. Indicator pins or jumpers on a SIMM or DIMM module that provide information about the capacity, speed, or configuration of a module to the computer. Can be used by the computer to set refresh speed, but most commonly used determine compatibility and to prevent 3<SUP>rd</SUP> party companies from using generic SIMMs. :) |
| 23:10:11 | BZFlag | heh |
| 23:10:24 | TomW | ?? |
| 23:12:18 | BZFlag | your definition |
| 23:13:41 | BZFlag | well, the only place I can sever pin 11 is right next to the pad. the trace must go down a foil from there, neither side of the board shows it. |
| 23:13:41 | BZFlag | and no xray hw here. |
| 23:13:48 | TomW | most of those DRAM boards are Multi-layer. |
| 23:16:51 | TomW | One trick I used to use a lot was to take cellophane tape and cover the edge finger. You lay the tape over the finger, then trim & peel with an X-acto knife, leaving the pad covered in tape. |
| 23:17:51 | TomW | I forgot about that, it has been many years since I troubleshot boards for a living. Now I get to design them. :-) |
| 23:43:02 | BZFlag | phone, another head hunter. and taking care of kids... /me returns and looks for appropriate tape. |
| 23:45:31 | TomW | LOL, yes, the headhunters! It is a challange to find one that knows what it is that you're talking about. |
| 23:46:08 | BZFlag | heh, yep. |
| 23:46:40 | andersee | TomW: I usually start off that I don't negotiate with companies for less than $250k... That shuts em up quickly. :-) |
| 23:46:58 | TomW | I recall telling one that I did embedded systems design and software. He then asked me if that had anything to do with database. |
| 23:47:13 | TomW | haha! |
| 23:47:42 | TomW | ... helps filter out the rash. |
| 23:47:49 | TomW | s/rash/trash/ |
| 23:48:52 | BZFlag | I normally start with "does this have anything to do with sql?" and then hang up if yes. |
| 23:49:59 | BZFlag | you've seen the disclaimer on my resume? /me needs to update if for folks that want hime to move to CA. http://rikers.org/resume.htm |
| 23:51:35 | andersee | hehe |
| 23:52:30 | andersee | is away: I'm busy |
| 23:52:59 | BZFlag | cool! covering pin 11 shows 40M under inferno. it did not boot before. |
| 23:53:54 | BZFlag | grabbed the first phone is site which happened to still have inferno on it, so keep your comments to yourself, andersee. |
| 23:53:59 | BZFlag | er sight. |
| 23:55:39 | TomW | With having done what you have on your resume, I estimate that you are in your 40's. |
| 23:55:54 | BZFlag | turning 39 in august. |
| 23:56:12 | TomW | Mmm.. 49 in July. |
| 23:57:28 | BZFlag | finds a jtag equipped linux phone... |
| 23:57:37 | TomW | Ok, got down to the work history. |
| 23:59:32 | BZFlag | Memory map: |
| 23:59:32 | BZFlag | 0x00400000 @ 0xc0000000 (4 MB) |
| 23:59:32 | BZFlag | 0x00400000 @ 0xc8000000 (4 MB) |
| 23:59:33 | BZFlag | 0x00400000 @ 0xd0000000 (4 MB) |
| 23:59:33 | BZFlag | 0x00400000 @ 0xd0800000 (4 MB) |
| 23:59:33 | BZFlag | 0x00400000 @ 0xd8000000 (4 MB) |
| 23:59:35 | BZFlag | 0x00400000 @ 0xd8800000 (4 MB) |
| 23:59:50 | BZFlag | that's without changing blob. changing blob now... |
