| 02:12:37 | nelson | Anybody reflashed blob from blob yet? |
| 03:52:36 | Russ | nelson? |
| 03:53:19 | Sammy | Russ : watch blob ... |
| 03:55:24 | Sammy | plz #blob |
| 04:06:19 | | nelson was last seen on #tuxscreen 1 hours, 53 minutes and 42 seconds ago, saying: Anybody reflashed blob from blob yet? [Thu Aug 16 03:12:37 2001] |
| 04:06:19 | Russ | ibot: seen nelson |
| 05:58:22 | Russ | hey tim, wb |
| 06:00:57 | BZFlag | Russ: thanx. Thoughts on RussNelson's crash report? |
| 06:01:06 | BZFlag | I'll do md5's for the images... |
| 06:01:57 | Russ | I would like the full output |
| 06:02:02 | Russ | so I could se gcc, etc |
| 06:02:12 | Russ | er, see |
| 06:02:18 | BZFlag | yep. |
| 06:02:20 | Russ | does he have the 64 in there? |
| 06:03:19 | Russ | oh, I bet he has the old blob |
| 06:03:27 | Russ | that doesn't set the CPU speed rigth |
| 06:03:38 | Russ | which is why the pcmcia timings are off |
| 06:03:48 | BZFlag | ahh! Yes I'll bet you are right |
| 06:08:38 | Russ | do I get schematics? |
| 06:12:25 | BZFlag | hmmm... my imap copy just lost the message with the scans... recovering (I hope) |
| 06:12:46 | Russ | heh...sure... |
| 06:13:09 | Russ | there *are* no schematics, its a vast lineo conpirasy |
| 06:13:16 | BZFlag | hehehe |
| 06:14:33 | BZFlag | how are you feeling about the current patches? Once the ts driver issue is resolved should we seek to get them included in np or rmk's? |
| 06:15:06 | BZFlag | I imagine the DSP work will be in the form of a user space daemon. You? |
| 06:15:27 | Russ | thats the plan (with the DSP) |
| 06:15:36 | Russ | I can start breaking stuff off and feeding it |
| 06:15:56 | BZFlag | sounds good. |
| 06:15:58 | Russ | ucb1x00, pcmcia, fb |
| 06:22:58 | BZFlag | http://TuxScreen.net/img/schematics/ |
| 06:29:17 | Russ | some of these may need to be scanned at a higher res |
| 06:29:31 | Russ | I plan to draw up some schematics in gschem |
| 06:30:25 | Russ | just those 6{? |
| 06:31:39 | Russ | I like the hand sketched notes, helps to know what they are talking about in the source |
| 06:31:55 | BZFlag | that's it. the others were data sheets that are already linked to from the site, and a few "welcome to the WebPhone beta program" letters. |
| 06:32:47 | BZFlag | the scans are actually about as clear as the originals. Remember the originals are copies. |
| 06:33:02 | BZFlag | who knows how many generations later we have gotten them. |
| 06:33:20 | Russ | ah, suppose I'll have to guess the words |
| 06:33:59 | BZFlag | I think there should be a 2b that covers the other half of that image. |
| 06:34:20 | BZFlag | I'll get that side done tomorrow. |
| 06:34:39 | Russ | so 2a/2b is mongoose at a higher res |
| 06:38:12 | BZFlag | different paper, but yeah. |
| 06:38:32 | BZFlag | note that 2a has more chips than mongoose |
| 06:39:02 | BZFlag | and a different part number on wheaties |
| 06:39:07 | Russ | the footprints are there from what I can tell |
| 06:44:55 | Russ | http://support.tandy.com/support_phones/doc44/44618.htm |
| 06:45:25 | Russ | aparentyl, you can order the DSP from radio shack, but lucent would like to pretend that is doesn't exist |
| 06:47:44 | Russ | http://www.google.com/search?q=cache:IhXccPZFUAk:www.lucent.com/micro/dsp/docs/TN98006-1.2.pdf+DSP1653&hl=en |
| 06:47:55 | Russ | clearly lucent had a DSP16XX line |
| 06:48:01 | Russ | but has remove all data sheets |
| 06:53:03 | BZFlag | replaced by the dsp16xxx |
| 06:53:26 | Russ | I know |
| 06:53:31 | Russ | but why remove all the docs |
| 06:58:55 | Russ | the commboard is a mongoose 3 |
| 06:59:05 | Russ | (taken from comcode printed on board) |
| 07:02:18 | Russ | doc request sent to lucent |
| 07:05:37 | BZFlag | http://www.st.com/stonline/books/pdf/docs/1653.pdf |
| 07:05:46 | BZFlag | not the same part? |
| 07:06:40 | Russ | nope |
| 07:08:05 | BZFlag | http://www.bdti.com/procsum/dsp16xx.htm |
| 07:10:03 | Russ | the best doc I have is a google cache of a lucent app note for compiling for DSP16XX |
| 07:25:49 | BZFlag | http://e-insite.freetradezone.com/cgi-bin/ncommerce3/ExecMacro/partsearch.d2w/report?part=DSP165&styp=generic&mfr=1661&status=all&db=all&path=RES0&srn= |
| 07:29:00 | Russ | close |
| 07:30:43 | Russ | it has functional differences from 1653 in the doc |
| 07:32:09 | Russ | should be enough |
| 07:33:18 | BZFlag | which of the 3 did you start with? |
| 07:33:32 | Russ | ummm...3 what? |
| 07:34:02 | BZFlag | oh, I assumed you started with DSP1650T24* data sheet? |
| 07:34:20 | Russ | think so |
| 07:36:39 | BZFlag | ok, I'm uploading a copy to TuxScreen.net/download/ |
| 07:40:00 | Russ | cool |
| 07:40:30 | BZFlag | rom - diffusion? |
| 07:41:47 | Russ | ? |
| 07:44:42 | BZFlag | 1650 is rom = flash, but 1653 is rom diffusion. what is a diffusion rom? |
| 07:44:53 | Russ | don't know |
| 07:44:59 | BZFlag | me neither |
| 07:45:04 | Russ | what I really wanted was the pinout |
| 07:45:15 | BZFlag | I'm hoping it's not a one time programable. |
| 07:46:36 | Russ | it would be a nice learning experience for people |
| 07:52:17 | Russ | If I can get schematics done, and the DSP properly documented |
| 07:52:26 | Russ | I think that would be a great package for universities |
| 07:53:36 | Russ | although it would be very linux-contric |
| 07:53:40 | Russ | er, centric |
| 07:58:15 | BZFlag | to bad for them to have to learn a stable flexible OS. I feel for them. ;-) |
| 08:00:05 | BZFlag | http://TuxScreen.net/wiki/view//LucentDSP updated. time for bed. happy hacking! |
| 11:10:08 | mallum | hi bworth |
| 11:10:33 | bworth | hey mallum |
| 11:10:46 | bworth | how's your tux? |
| 11:14:38 | mallum | ok, still trying to build uCLibc based crossed compiler .... |
| 11:16:05 | nelson | woo hoo, thanks Russ / BZFlag, upgrading to the blob did the trick! |
| 11:16:43 | mallum | cool nelson :) |
| 11:18:06 | nelson | Hey, and I got an Ethernet address off my Socket LP-E Ethernet, so maybe it'll work on 3.3V. |
| 11:18:50 | mallum | yeah I think they do work ... |
| 11:33:16 | mallum | arg, my machine was locked up by a make menuconfig ???? |
| 21:01:27 | mallum | hi all |
| 21:03:39 | mallum | microdrive working ..... ? |
| 21:04:20 | BZFlag | yes. I have a 1G in my phone here with debian potato booting. udrive is root on reiserfs |
| 21:06:17 | mallum | I tried mine , with no luck ... it was asking for 5v |
| 21:07:01 | BZFlag | mine asked too, but is happy with the 3.3v that is got. |
| 21:07:53 | BZFlag | you running the latest kernel with the pcmcia-ts patches? there were irq issues with the earlier ones that hung the udrive of long periods as polling was the only thing working. |
| 21:08:55 | BZFlag | Russ's 2.4.7 has the same fix. |
| 21:09:08 | finalfantasydog | Bzflag are you still selling them and if so is linux up on the ones you sell now? |
| 21:09:58 | BZFlag | I am still selling them. $99 gets just the Inferno flash locked version. I suppose I could charge more and do a linux install if you like. have not thought on how much to charge. |
| 21:10:31 | finalfantasydog | yes, i'd like that |
| 21:10:37 | finalfantasydog | how much for the linux install? |
| 21:10:50 | finalfantasydog | well when you come up witha price please tell me |
| 21:10:54 | Russ|werk | probably would just be a blob install |
| 21:10:56 | BZFlag | what's it worth to ya? ;-) |
| 21:11:21 | Russ|werk | its fun to see inferno running on it |
| 21:11:24 | BZFlag | Russ|werk: hmmm... Just the flash unlock might be enough. |
| 21:11:35 | mallum | BZFlag: I have 0,3 installed ... is there a newer version now ? |
| 21:11:50 | BZFlag | mallum: there is a new kernel in the v0.3 directory |
| 21:12:20 | finalfantasydog | heh depends how difficult it is too do it, and time-consuming |
| 21:12:33 | mallum | BZFlag: cool ... does that that have ide_cs built in ? |
| 21:13:29 | mallum | Russ|werk: sorry, but can I hassle you more about uClibc ... ? |
| 21:14:42 | mallum | Russ|werk: did you catch my messages earlier about the problems I was having ? |
| 21:14:52 | Aelhaz | I've been considering offering to unlock 'em and install the blob for $20 . . . |
| 21:15:55 | Russ|werk | mallum: you don't want it as your gcc |
| 21:16:05 | Russ|werk | you want to run arm-uclibc-gcc to invoke it |
| 21:16:12 | Russ|werk | gcc should run your native gcc |
| 21:16:44 | mallum | Russ|werk: I set the uclibc to br first in my path |
| 21:17:18 | mallum | Russ|werk: 'which gcc' returns the path to /usr/uClibc-arm/bin/gcc |
| 21:17:22 | BZFlag | mallum: yes. it has everything the config up there for the non -pcmcia-ts version has. jsut a few code changes that are in Russ' 2.4.7 release backported. |
| 21:17:26 | Russ|werk | mallum: thats bad |
| 21:17:47 | Russ|werk | mallum: I make symlinks to arm-uclibc-* in /usr/local/bin |
| 21:18:12 | BZFlag | finalfantasydog: there ya go. $119 and I'll have Aelhaz unlock and blob it for ya. ;-) |
| 21:18:41 | Russ|werk | BZFlag: how much extra for a null modem cable |
| 21:19:28 | mallum | Russ|werk: Im sure its using the uClibc gcc , did you see the errors I was getting ? |
| 21:19:31 | BZFlag | Russ|werk: andersee has both options working. the broken-just-gcc option is to use with broken Makefile's |
| 21:20:19 | mallum | BZFlag: was that to me ? |
| 21:20:21 | Russ|werk | mallum: what errors |
| 21:20:38 | mallum | Russ|werk: one sec I show you what I mean .. |
| 21:20:40 | Russ|werk | BZFlag: I prefer fixing the makefiles |
| 21:21:20 | Russ|werk | BZFlag: since most of the programs I've seen with b0rked makefiles also need to compile some code generation stuff with native gcc |
| 21:21:49 | mallum | Russ|werk: ok heres what I get from trying to compile w3m : |
| 21:21:58 | mallum | indep.h:25: conflicting types for `strcasestr' |
| 21:21:58 | mallum | /usr/arm-linux-uclibc/include/string.h:98: previous declaration of `strcasest |
| 21:22:41 | mallum | when ever I try to compile anything I seem to get lots of 'previous delaration' and redefined errors |
| 21:22:42 | BZFlag | Russ|werk: heh, no extra cables here, I doubt I'll order a bunch to resell either. |
| 21:22:53 | Aelhaz | BZFlag: I wonder how hard it would be to modify the 12v connector to get it to hold onto the edge of the board |
| 21:23:01 | Russ|werk | mallum: that pataticular error seems to be in your program |
| 21:23:10 | BZFlag | Russ|werk: I prefer fixing the makefiles too. ;-) |
| 21:23:34 | Russ|werk | are there #ifdef's in indep.h around the strcasestr declaration? |
| 21:23:50 | BZFlag | Aelhaz: moberg was working on that, but I have not heard from him. Shaggy has one that snaps into the holes. |
| 21:24:35 | Russ|werk | BZFlag: a little dot of solder works well |
| 21:25:21 | BZFlag | yeah, but if I'm going to do 1000 phones, I'll build a snap on connector. |
| 21:25:30 | mallum | Russ|werk: only the usual #ifndef INDEP_H |
| 21:25:31 | mallum | #define INDEP_H |
| 21:25:32 | mallum | .. |
| 21:25:52 | Russ|werk | mallum: are there multiple indep.h files? one for each arch? |
| 21:26:29 | BZFlag | then get a box to send the IR commands ESC, ctrl-a, P/m, P/u 0, P/m over and over. ;-) |
| 21:27:09 | Russ|werk | BZFlag: might as well flash blob as well |
| 21:27:22 | BZFlag | then they won't get to play with Inferno forst. |
| 21:27:59 | Russ|werk | later, you might have someone doing pcmcia hacks for $$$ |
| 21:28:00 | BZFlag | er first. better to build a small app that does the full install instead of inferno-mincom-uuencode |
| 21:28:11 | mallum | Russ|werk: Could you point me towards something you have compiled , then I can see if I can compile it too - and I'll know if its my setup or not ? |
| 21:28:25 | Russ|werk | mallum: what are you trying to compile |
| 21:28:43 | BZFlag | then put the app on the site. ie: dl nukeinferno,blob,zImage,root.jffs2 and run nukeinferno. ;-) |
| 21:29:38 | BZFlag | nukeinferno could even check for protected flash etc. |
| 21:31:29 | mallum | Russ|werk: so far I have just tried random stuff with out too many dependancies - expat, w3m, wget, wireless tools ( iwconfig ) |
| 21:31:42 | mallum | Russ|werk: all have failed :( |
| 21:32:07 | Russ|werk | back to this indep.h |
| 21:32:29 | Russ|werk | from the name, it looks like there should be more than one, each for a different arch |
| 21:32:36 | Russ|werk | and you are usung the wrong one |
| 21:34:43 | mallum | there seems to be only 1 |
| 21:35:07 | mallum | I compiled tinylogin ok |
| 21:35:09 | Russ|werk | fix it then |
| 21:35:20 | Russ|werk | which app is that for? |
| 21:35:25 | mallum | w3m |
| 21:35:39 | Russ|werk | is indep.h generated? |
| 21:36:18 | mallum | dont think so - indep.h seems pretty simple |
| 21:36:46 | Russ|werk | I would imagine its automatically generated based on you machin |
| 21:37:22 | Russ|werk | in any case, take out the supplemental strcasestr, because uclibc already has it |
| 21:38:22 | mallum | ok will try |
| 21:39:52 | mallum | sorry to be a hassle, I just dont understand why Im having so much bad luck with it :( |
| 21:39:59 | Russ|werk | what other conflicting/duplicate declarations do you have? |
| 21:40:16 | Russ|werk | oh, also, one more thing to change in uClibc |
| 21:40:45 | mallum | well when I try and compile expat I get the following ... |
| 21:41:11 | Russ|werk | SHARED_LIB_LOADER_PATH=/lib |
| 21:42:22 | Russ|werk | ? |
| 21:43:02 | mallum | expat dont compile with normal gcc ... so we can forget that one |
| 21:44:09 | mallum | maybe I've just been unlucky with the stuff I've tried to compile |
| 21:45:07 | Russ|werk | what messages does it give you? |
| 21:45:18 | Russ|werk | I usually have little quirks to work out |
| 21:45:36 | Russ|werk | things like UTMP_FILE not being defined, etc |
| 21:50:01 | mallum | ok, can you give me an example of something that compiles clean, so I can test it - and I'll know that it is or is not my setup ? |
| 21:50:15 | mallum | sorry, I dont mean to be a pain ... |
| 21:50:23 | finalfantasydog | can you ship internatinoally bz? |
| 21:51:04 | Russ|werk | tinylogin |
| 21:51:15 | mallum | I compiled that though :) |
| 21:51:18 | Russ|werk | mallum: just let me know what errors you have |
| 21:51:27 | Russ|werk | its ok to flood a bit |
| 21:52:24 | mallum | Russ|werk: also I had to patch the stock kernel with the rmk patch to get it to make config |
| 21:53:06 | mallum | Russ|werk: and I think you have to make dep to get the soft links after that |
| 21:53:54 | BZFlag | finalfantasydog: yes. I've shipped to the ca, uk, and nl so far |
| 21:54:35 | mallum | Russ|werk: I'd really like to get tiny-x compiled |
| 21:55:41 | CosmicPenguin | So, I guess I need to go get a microdrive now, huh? |
| 21:56:17 | CosmicPenguin | There is something strange about paying 4 times as much for storage as you did for the base unit.... :0 |
| 21:56:52 | CosmicPenguin | (Replace the :0 with a :) ) |
| 21:58:50 | Aelhaz | CosmicPenguin: I just picked up a generic 64 meg cf card for $35 |
| 21:59:12 | mallum | Aelhaz: nice |
| 21:59:15 | Aelhaz | tho i coulda just done nfsroot |
| 21:59:41 | CosmicPenguin | Aelhaz: Thats not bad. Are you doing a journaling filesystem on it? |
| 21:59:44 | Aelhaz | have to pick up that card at the fedex dock tonight |
| 21:59:56 | Aelhaz | CosmicPenguin: Probably. I haven't finalized my plans, really |
| 22:03:15 | Russ|werk | CosmicPenguin: there are pcmcia->laptop IDE adaptors available |
| 22:03:26 | Russ|werk | so you can use a normal 2.5" drive |
| 22:03:54 | Russ|werk | run the cable back around and mount it on the inside |
| 22:05:27 | Russ|werk | I saw 128M CF the other day for $53, thats enough for a small distro |
| 22:05:33 | Russ|werk | too bad its flsah |
| 22:05:43 | CosmicPenguin | Russ: I have a uClibc patch that does a quick and dirty implemntation of the IO (inb/outb, etc...) for the ARM. Is there anywhere in particular you want it on the server? |
| 22:09:44 | kR_ | Russ: what's wrong with flash? |
| 22:13:00 | Russ|werk | CosmicPenguin: why do you need it? |
| 22:13:15 | Russ|werk | kR: it erases slowly in large sectors |
| 22:13:50 | Russ|werk | kR: and only sustains 100,000 erase cycles per sector |
| 22:14:35 | Aelhaz | 100,000 cycles is a guesstimate, lowest common denominator even |
| 22:14:50 | Aelhaz | but it's still not good to hammer your flash with i/o |
| 22:15:48 | Russ|werk | 100,000 happens pretty quite when you have log files and internet cache/history |
| 22:15:53 | Russ|werk | er, quick |
| 22:15:55 | Russ|werk | not quite |
| 22:16:14 | Aelhaz | our testing manager here used to be a QA manager at philips semiconductor - says that they would rate their flash chips at 100k 'cause everyone else did |
| 22:16:42 | Russ|werk | heh |
| 22:16:44 | Russ|werk | makes a note not to buy philips flash |
| 22:16:44 | Aelhaz | on rare occasions they tested samples that did less than 50k before flaking out, so they tossed those in the wood chipper |
| 22:17:16 | Aelhaz | Usually, the test batches lasted 200-400k cycles, but these are just samples from a batch, could have a few not as good |
| 22:17:27 | Aelhaz | and on rare occasions they had samples that went into the millions of cycles |
| 22:17:59 | Aelhaz | so, it varies, but the point is that it's got a limited lifespan if you reprogram it a lot |
| 22:18:28 | Aelhaz | and yeah, log files, caches, etc, will do that pretty rapidly |
| 22:19:11 | Aelhaz | I'm gonna go visit FedEx . . . |
| 22:22:26 | CosmicPenguin | Russ: In order to build Microwindows |
| 22:31:35 | CosmicPenguin | Don't worry about it, I'll put it somewhere else |
| 22:38:19 | mallum | CosmicPenguin: I thought you'd built microwindows ...? |
| 22:39:32 | CosmicPenguin | I did, but to build it with uClibc, I had to do a little magic. |
| 22:45:17 | mallum | ok |
| 22:45:26 | mallum | bed time for me - night all |
| 22:51:46 | BZFlag | CosmicPenguin: going to fit in rootfs? |
| 22:55:16 | Russ|werk | CosmicPenguin: I would just make a .o and link it in |
| 22:55:35 | Russ|werk | CosmicPenguin: where does microwindows do inb, outb? |
| 22:55:41 | Russ|werk | for what purpose? |
| 23:30:26 | CosmicPenguin | Russ: There is some code in the planar VGA driver. I really don't need it, but I figured that it should compile with everything..... |
| 23:31:10 | CosmicPenguin | Sorry, I took so long, I was adding a page to the Wiki.... |
| 23:34:15 | CosmicPenguin | Russ: I'll take out that dependancy, so we don't have to patch everybody's uClibc.... |