irclog2html for #openjtag on 20050614

05:05.56*** join/#openjtag jamie (~jamie@softmodem.org) [NETSPLIT VICTIM]
05:05.56*** join/#openjtag Tiersten (tman@Tiersten.nslu2-linux) [NETSPLIT VICTIM]
05:05.56*** join/#openjtag pjames (~jcouzens@142.179.94.230) [NETSPLIT VICTIM]
06:43.24*** join/#openjtag jamie (~jamie@82.152.157.198)
08:45.47*** join/#openjtag JMunakra (~JMunakra@dsl-084-059-206-162.arcor-ip.net)
09:32.43*** join/#openjtag ep1220 (~NN@gate.epatec.at)
09:52.05*** join/#openjtag pjames (~jcouzens@S0106000024c23f2c.vs.shawcable.net)
11:41.30*** join/#openjtag prpplague (~billybob@72.22.135.51)
15:23.28*** join/#openjtag ka6sox (~ka6sox@netblock-66-159-209-52.dslextreme.com)
19:04.44*** join/#openjtag jacmet (~jacmet@d51A52AE6.access.telenet.be)
19:57.31*** join/#openjtag jacmet (~jacmet@d51A52AE6.access.telenet.be)
20:14.37*** join/#openjtag [g2] (~g2@g2.nslu2-linux)
20:27.02*** join/#openjtag prpplague (~billybob@72.22.132.158)
20:38.08*** join/#openjtag prpplague^2 (~billybob@216.110.100.26)
21:20.30*** join/#openjtag JMunakra (~JMunakra@dsl-084-059-206-162.arcor-ip.net)
21:45.58*** join/#openjtag beewoolie-afk (~beewoolie@florence.buici.com)
22:09.57*** join/#openjtag purl (ibot@apt.bot.TimRiker.active.supporter.pdpc)
22:09.57*** topic/#openjtag is This is the place to discuss development of Open JTAG tools and Bootloaders
22:33.05[g2]beewoolie, DUDE you rock!
22:33.18beewoolie[g2]: it doesn't work for you yet.
22:33.21beewoolieOr does it?
22:33.32[g2]well it mostly works
22:33.39[g2]it doesn't fully work
22:33.51[g2]but you still ROCK
22:33.55beewoolieI know that you were having problems with listing the /boot directory.
22:34.02[g2]nod.
22:34.03beewoolieAnd then reading from an offset didn't work.
22:34.11beewoolieCan you load a kernel anyway?
22:34.13[g2]nod.
22:34.23[g2]haven't tried loading the kernel
22:34.40[g2]I was trying to load a file first and then I was going to dump memory and verify the contents
22:35.23beewoolieI haven't tested the code on a slug yet.  I've been working with one of my dev boards instead since it is hooked up and going.
22:36.01[g2]this was running on the fatslug
22:36.11beewoolieThat ought not matter.
22:36.27[g2]nod... that's why we try it right ? :)
22:36.36beewoolieIndeed.
22:36.50[g2]is there anything I can try for you ?
22:37.15beewoolieI'm thinking on it right now.
22:37.22[g2]OK
22:37.26beewoolie...actually, you could try one thing tho.
22:37.30[g2]sure name it
22:38.08beewoolieIf you build the loader with the #define TALK defined in src/drivers/drv-jffs2.c and then try info on the /boot directory, it will write a bunch of stuff to the console.
22:38.15beewoolieCapture it and send it by email.
22:38.39beewoolieAlso, try the copy command and send that output.
22:40.45[g2]you mean the last one from the e-mail I sent you ?
22:40.50[g2]I tried that with foo
22:40.55beewoolieyeah, that's it.
22:41.04[g2]that was in the last e-mail
22:41.10beewoolieThe TALK output will tell me what it's doing.
22:41.17[g2]ah
22:41.23beewoolieYou sent me the console output with TALK enabled?
22:41.37[g2]it's built and scp'd to the host, I"m verifying the MD5 then I'll reflash
22:41.59beewoolieYou can run it without reflashing.
22:42.03beewoolieThat's what I do.
22:42.13beewoolieJust load it to 0x0 and then g 0x0
22:42.20beewoolieSafer than reflashing.
22:42.41[g2]you load over serial ?
22:42.47beewoolieUsually.
22:42.50beewoolieIt doesn't take long.
22:43.01beewoolieapex> xr 0
22:44.10[g2]or so just xr 0 and then xmodem the file ?
22:44.16beewoolieYep.
22:44.21[g2]I've been wanting to do this
22:44.31beewoolieit takes a couple of seconds for the two ends to sync.
22:47.00[g2]I'm trying the xmodem send from minicom which isn't clear
22:47.13beewoolieSend-> Xmodem
22:47.25beewoolieThe interface for browsing file sux.
22:47.49[g2]I figured that much ... I didn't see the tabs at the bottom :)
22:50.35[g2]minicom still isn't playing well :*
22:50.51beewoolieDo you have rzsz installed?
22:51.03[g2]dunno
22:51.13beewoolietry 'which sx"
22:51.26[g2]no sz
22:51.28[g2]no sx
22:51.34beewoolieThere's your problem.
22:51.44beewoolieThe package on debian is lrzsz.
22:53.14[g2]ok it's emerged
22:53.50[g2]Sending apex.bin, 369 blocks: Give your local XMODEM receive ?
22:53.50[g2]stopped k?command now.
22:54.12beewoolieHmm/
22:54.14beewoolie<PROTECTED>
22:54.23[g2]it might be sending
22:54.39[g2]it's certainly better after the emerge :)
22:55.00[g2]Ok I think it worked
22:55.07[g2]I'm back at the APEX prompt
22:55.15beewoolieNice, eh?
22:55.24[g2]yeah very nice
22:55.29beewoolieI think there is a bug in it, tho.  it doesn't want to write directly to flash at the moment.
22:55.35beewoolieIt should be possible to do it.
22:55.43[g2]so go 0
22:56.02beewoolieYep.
22:56.39[g2]it's unhappy
22:56.49[g2]x-mas tree lights all lit up
22:56.56beewoolieHmm.
22:57.01[g2]go 0
22:57.01[g2]Jumping to 0x00000000...
22:57.08beewoolieCrash?
22:57.16beewoolieHmm.
22:57.22[g2]that'd be my guess
22:57.25[g2]or lockup
22:57.43[g2]power-cycle fixes it
22:57.55beewoolieWell, you can try writing it to flash to see if it's happier.  I'll hook it up in a little while to see if I can determine the cause of the failure.
22:58.03beewoolieAll on means something, too.
22:58.20[g2]I'd imagine TALK is pretty safe right ?
22:58.46beewoolieIt will make it more verbose, that's all.
22:59.01beewoolieAs you have JTAG, I'm not too worried.  :-)
22:59.21[g2]well JTAG hasn't been verified on this unit
22:59.32beewoolieHmm.
22:59.36beewoolieIs this the avila?
22:59.46[g2]no this is on the fatsly
22:59.49[g2]fatslug
23:00.03beewoolieis there a way to try the latest APEX on the avila?
23:00.16[g2]Although I got the UBER-DOOBER fat slug on loan today
23:00.41[g2]I'll have it for 3-4 weeks
23:00.42beewoolieWhat does that mean?
23:00.48beewoolieRAM and flash upgrades?
23:01.06[g2]IXP425 Intel Dev board 256MB 425 @ 533 I think
23:01.14[g2]I know its got 256MB
23:01.39beewoolieNice.
23:01.48[g2]and one of my most favorite pieces of hw
23:01.50beewoolieSeems like testing JTAG would be a good idea.
23:02.07[g2]on the fat slug or the 425 ?
23:02.59[g2]it's got the little plastic coffin that you can switch the flash parts in and out
23:03.09beewoolieI like those.
23:03.20beewoolieI think those sockets are $40 or so.
23:03.25[g2]that's "one of my favorite hw parts"
23:03.45[g2]no if I only had a flash programmer I'd really be set :)
23:04.26[g2]come to think of it those would be super handy to have on the slug
23:04.33[g2]and a programmer of course :)
23:05.14beewoolieIt would be slick for development, yeah.
23:06.30[g2]the IXP425 kit or the little coffin ?
23:07.05beewoolieWell, both, but a coffin'd slug could mean easier development.
23:07.25[g2]if we had a flash programmer
23:07.44[g2]I've got some 16MB samples it would be nice to just plug them in
23:09.52[g2]those would be *way* cool for memory chips :)
23:15.19[g2]beewoolie, you've got mail :)
23:15.23beewoolieKO.
23:17.10[g2]beewoolie, BTW I hope I'm not bugging you too much with this stuff!
23:17.15beewoolieNope.
23:17.26beewoolieI'm doing several other things at the same time, so you'll have to be patient.
23:17.28[g2]I hope I'm the helpful tester
23:18.17[g2]NP, me too :)
23:18.22beewoolieYeah.  This output is helpfil.  The i -1 in the read is important.
23:18.55[g2]diagnostic's are good.  It's great to be able to just turn some on
23:19.10[g2]or even diagnostics
23:27.03beewoolieHmm.  I think I might have found the problem.
23:27.13[g2]excellent
23:27.19[g2]you're fast
23:27.25[g2]what do you think it is ?
23:27.55beewoolieI changed one of the routines.  I optimized it.  And I think that that optimization doesn't work as I thought.
23:28.19[g2]with that cache ?
23:28.19beewoolieIt seems to be that the jffs2 filesystem you sent has important differences from the one I have in flash.
23:28.41[g2]how's that ?
23:28.43beewoolieIt's the find_cached_inode() function.  I changed it to do a complete binary search instead of a bsearch just on the inode number.
23:29.04beewoolieI don't know what is different about your data, but it certainly helps with finding bugs.
23:29.54beewoolieOh yeah.  I think I found it.
23:30.00[g2]COOL..
23:30.10[g2]I told you beewoolie ROCKS
23:30.45[g2]BTW did you see that article about the virtual ARM simulator ? it's supposed to run at 217Mhz on a 2-3G pentium
23:30.56beewoolieThat could be handy.
23:31.01beewoolieWhere was the article?
23:31.19[g2]it was about a week or so ago... I can probably dig it up
23:31.47jacqueslinked from linuxdevices
23:31.55jacquesvir-someting
23:32.14[g2]http://www.linuxdevices.com/news/NS4665550218.html
23:32.24[g2]that's where I thought I saw it
23:32.32jacqueshttp://www.virtera.com/
23:32.35[g2]they were running OPIE in a simulator
23:33.05jacquesno BE tho
23:33.16[g2]yeah 217Mhz on a 2.6 Pentium
23:33.25[g2]well we can test in LE right :)
23:33.35[g2]it'll work for jacmet
23:33.44beewoolie[g2]: I can now read the /boot directory.
23:33.56[g2]go beewoolie go beewoolie
23:34.00beewoolieDon't quite know why I didn't catch that one myself.
23:34.08beewoolieI need to do some testing here before I can call it OK.
23:34.26[g2]OK...
23:34.42[g2]I'm gonna have to try building my 425 kernel sometime
23:38.46beewooliebugs bugs bugs
23:39.28[g2]I'm guessing you're squashing bugs bugs bugs
23:39.55beewoolie...finding...and then SMASHING!  Beewoolie SMASH!
23:40.17jacqueslol

Generated by irclog2html.pl by Jeff Waugh - find it at freshmeat.net! Modified by Tim Riker to work with blootbot logs, split per channel, etc.