**** BEGIN LOGGING AT Wed Mar 02 23:59:56 2005 Mar 03 00:16:52 morning Mar 03 00:43:08 how is kernel 2.6.11 going? Mar 03 00:43:22 working yet? Mar 03 00:43:37 * dyoung mumbles random obscenities. Mar 03 00:43:55 a, it's going great i see ;) Mar 03 02:00:09 using xchat on windows via ssl-tunnel and miau ;) Mar 03 02:00:58 [g2], when I try booting the unmodified ixp4xx kernel, it gets slightly farthur and says "Error: a" before hanging. Mar 03 02:01:23 [g2], what happens when your machine id is incorrect? Mar 03 02:03:47 * perlguru-work byebye Mar 03 02:04:42 bye bye? Mar 03 02:04:56 bye.. Mar 03 02:05:24 who is leaving? Mar 03 02:05:47 you are? Mar 03 02:05:59 [02:03] * perlguru-work byebye Mar 03 02:06:17 no, i restarted putty (now with keepalive) Mar 03 02:06:25 ah..understood Mar 03 02:06:38 i use putty for ssltunnel Mar 03 02:07:01 need keepalive or else proxy will disconnect the session Mar 03 02:08:21 dyoung, have you scanned the changelog of 2.6.11? Mar 03 02:11:36 Yes. Mar 03 02:14:01 any hints in there? Mar 03 02:25:15 There were a lot of arm changes. so the short answer is "not really". Mar 03 02:28:43 hi all. openslug gone beta ylready? Mar 03 02:30:05 not yet Mar 03 02:30:17 there is a release candidate Mar 03 02:30:24 any you are number #20 nice :) Mar 03 02:30:28 but you have to make it yourself Mar 03 02:30:50 ok, you know when its going to be a binary? Mar 03 02:31:00 yep, but i reverted back to unslung, because of some strange AMD64 related issues Mar 03 02:31:13 oh Mar 03 02:31:19 ask [g2], he is the release manager ;) Mar 03 02:31:27 k Mar 03 02:31:53 were u able to try some irc or torrent on openslug? Mar 03 02:32:10 It wont be release until we have some resolution on what intel's position is on their licensing for the NPE drivers. Mar 03 02:32:15 Wipmac, http://www.nslu2-linux.org/wiki/OpenSlug/BetaRelease Mar 03 02:38:38 Wipmac, bitchx works... obviously. ;) Mar 03 02:38:49 bittorrent is a bit unstable. Mar 03 02:39:00 u must be on bitchx now i guess :) Mar 03 02:39:04 which bt client? Mar 03 02:39:33 perlguru-work, I tried ctorrent, btget and original bittorrent. all of them lead to kernel panic after some hours of downloading. Mar 03 02:40:09 mmm, strange. I have no issues with Unslung Mar 03 02:40:21 I hope that issue disappears with 2.6.11. Mar 03 02:40:40 I hope so too Mar 03 03:09:26 I expect a lot of people will be disappointed when 2.6.11 doesn't solve all the troubles of the world ... Mar 03 03:10:05 the next version will solve all your problems ;) Mar 03 03:10:14 hmmm, I updated the puppy package from 1.6 to 1.8 at least 36 hours ago, but it's still not available when I do ipkg update; ipkg upgrade puppy Mar 03 03:10:18 any ideas? Mar 03 03:11:15 Unslung package? Mar 03 03:11:58 yes Mar 03 03:12:32 asked jp30? Mar 03 03:16:13 rwhitby, then it would be time to start debugging 2.6.11. I was not in the mood to debug a 6 months old kernel when someone else potentially already did that. Mar 03 03:16:19 * siddy-away is lazy! Mar 03 03:18:08 siddy-away: fully agree. there's just a lot of hope around that number, and I hope [:-)] that it's not dashed hope Mar 03 03:18:38 It doesnt need to solve all the problems in the world. It just needs to solve the irq26/ehci problem Mar 03 03:19:08 and the pl2303 problem, and the bittorrent problem, .... Mar 03 03:19:29 there was a bulk USB-serial fix. Mar 03 03:19:40 but I dotn get why that ATEN works for me. Mar 03 03:19:46 under 2.6.8.1 . Mar 03 03:19:52 (on the desktop) Mar 03 03:20:41 :-) Mar 03 03:20:49 Well, where its at now is it compiles. Mar 03 03:20:52 but thats about it. Mar 03 03:21:23 And who turned on USB sound in the kernel? Mar 03 03:21:47 lunacy! Mar 03 03:22:03 I mean geez, at least make it a module. Mar 03 03:22:04 Are .8 and .10 supposed to be the stable versions, and .9 and .11 the unstable versions ? Or does that only work for the second digit ? Mar 03 03:22:19 that sonly the 2nd digit. Mar 03 03:23:07 rwhitby, dyoung: linus actually started a discussion on lkml today to transfer that scheme to the third digit. Mar 03 03:24:32 so, maybe we'll have new features in 2.6.odd and bugfixes in 2.6.even in the near future. Mar 03 03:25:32 dyoung: the serial cable works when plugged directly into the slug, but doesn't work when plugged into the 7 port hub. Mar 03 03:26:18 Wierd. Mar 03 03:26:27 I bought a 4 port hub today, so I'm just about to try that one out ... Mar 03 03:27:24 Why dont you boot.... Mar 03 03:27:37 If I suck up the current BSR, will that help anyone debug this problem? Mar 03 03:31:02 why don't I boot what? Mar 03 03:32:13 Why doesnt devslug boot 2.6.11 Mar 03 03:33:03 cause I don't do bleeding edge kernel ... Mar 03 03:33:30 I was talking about my devslug. Mar 03 03:34:05 oh, you were talking to yourself about your devslug .. :-) Mar 03 03:34:16 yeah, it was a rhetorical question. Mar 03 03:35:27 well, if someone wants the BSR at the time the crash happens, I ihave it. Mar 03 03:35:36 I'm too lazy to decode this. Mar 03 03:35:42 and I'm giving up. Mar 03 03:36:13 I need to talk with Lennert before proceeding. Mar 03 03:45:10 dyoung-2611canbi: serial cable fails with 4 port hub to - perhaps it's an ehci problem too ? Mar 03 03:45:23 Probably. Mar 03 03:45:40 There were numerous problems. Mar 03 03:45:50 I wonder if 2.6.10 fixes them. Mar 03 03:46:03 many of the ehci problems were fixed in 2.6.10... Mar 03 03:46:17 I still have my 2.6.10 patchfile lying around. Mar 03 03:46:26 did you get 2.6.10 working? Mar 03 03:46:34 Actually I didnt try. Mar 03 03:46:38 it compiles. Mar 03 03:47:03 okay I'll try it. Mar 03 03:59:32 mgetty finds the modem ... Mar 03 04:14:07 dont want to see if 2.6.10 boots? Mar 03 04:14:50 I can wait to find out tomorrow. Should get to sleep before midnight at least once a week .... Mar 03 04:15:00 ok Mar 03 04:15:10 I was about to sleep too Mar 03 04:15:19 here goes. Mar 03 04:15:52 nope. Mar 03 04:15:59 same symptom. Mar 03 04:17:54 that was a good excercise, it gives me someplace to start looking. Mar 03 04:34:17 Christ. Mar 03 04:34:33 a bug they fixed in 2.6.10 creeped back into 2.6.11. Mar 03 04:34:46 in ixp4xx-regs.h Mar 03 04:35:01 <[g2]> Doh! Mar 03 04:47:10 <[g2]> dyoung, So running the plain vanilla gets further ? Mar 03 04:48:09 Yeah, it says "Error: a" Mar 03 04:48:30 <[g2]> that's with no patches Mar 03 04:48:38 right. Mar 03 04:49:05 <[g2]> I'm sorry this is so frustrating Mar 03 04:49:22 <[g2]> But I do appreciate all you attempts and efforts Mar 03 04:49:34 I need professional help. Mar 03 04:49:50 <[g2]> On the kernel ? Mar 03 04:49:58 yeah that too. ;-) Mar 03 04:50:28 <[g2]> humor keeps us all sane Mar 03 04:50:33 I sent lennert the latest working set of stuff. Mar 03 04:50:50 <[g2]> I mean it's been 24 hours right ? :) Mar 03 04:51:02 working as in we're working onit; not that it works. Mar 03 04:51:17 Oh, and you have mail too. Mar 03 04:52:47 <[g2]> thx Mar 03 04:54:18 2.6.10 does the exact same thing btw. Mar 03 04:54:35 <[g2]> nod. I saw that in the log Mar 03 04:56:06 if you have time, it might be nice to dtermine what the salient differences between the 2.6.9 and 2.6.10 loader stuff is/was. Mar 03 04:56:35 <[g2]> loader stuff ? Mar 03 04:56:47 ARM assembly is too complicated for my feeble brain. Mar 03 04:57:29 I'm pretty sure its barfing on something related to that. asm/arch/foo.S Mar 03 04:57:53 or possibly arch/arm/boot/compressed/foo.S Mar 03 04:58:22 <[g2]> The old patch got through the decompression Mar 03 04:58:43 Yeah, misc.c seems to be the last thing that gets run. Mar 03 04:59:01 <[g2]> does the new patch do the same or does it die earlier ? Mar 03 04:59:09 it dies in the same place. Mar 03 04:59:46 My batteries ran out 45min ago. Mar 03 04:59:47 night! Mar 03 04:59:58 <[g2]> sweet dreams! Mar 03 05:50:28 morning gents Mar 03 05:50:49 <[g2]> VoodooZ_work, morning Mar 03 05:55:18 VoodooZ_work, morning! Mar 03 05:56:43 morning! Mar 03 05:57:10 * VoodooZ_work finished reading the log. Mar 03 05:57:41 poor dyoung seems to be having a hard time with 2.6.11. Thank god he's doing it though. Mar 03 06:00:09 VoodooZ_work, any closer on finding the rootcause of our amd64 problem? Mar 03 06:02:14 not really. it seems nobody has clue or don't give a damn (and they are all smarter than me) so I give up for now. Mar 03 06:08:47 I just noticed that somebody has added a uboot package to OE. Mar 03 06:12:22 VoodooZ_work, do you have a unslung dev. env.? Mar 03 06:12:45 nope. I have absolutely no use for it. Mar 03 06:12:57 plus it's tainted by linksys. :) Mar 03 06:13:12 too bad. or else you could compare binaries made by gcc 3.3.5 Mar 03 06:18:22 * perlguru-work byebye Mar 03 07:07:04 going home, bye Mar 03 07:07:08 * perlguru-work Leaving Mar 03 07:42:01 <[g2]> giel, VoodooZ_work jacques perlguru OK thanks to dyoung I've got a running 2.6.11 kernel Mar 03 07:42:19 <[g2]> bad news is that ehci nobody cares remains an open issue Mar 03 07:42:53 <[g2]> the 2.6.11 kernel is larger and I needed to make ext3/jbd as modules Mar 03 07:43:58 <[g2]> I'm ixp still works and everything seems pretty similar to before (except it is the new kernel which has some known fixes built in) Mar 03 07:44:56 <[g2]> Do you guys want me to push these changes or stay on the 2.6.9 kernel ? Mar 03 07:46:30 cool! Mar 03 07:47:27 I would push the new bb files so we can test it out first but if you want to make it the default I don't mind either. Mar 03 07:52:11 [g2]: Please push the new pack/bb files for openslug-kernel_2.6.11 at a minimum. Thanks. Mar 03 07:54:35 <[g2]> doh! I forgot to put rwhitby on that list Mar 03 07:54:39 hehe Mar 03 08:23:38 morning rwhitby-away Mar 03 08:23:51 oops. Wrong dude Mar 03 08:27:21 [g2]: pong? Mar 03 08:48:00 <[g2]> giel, ping Mar 03 08:52:22 <[g2]> siddy, ping Mar 03 08:52:57 [g2], hi! Mar 03 08:53:12 <[g2]> siddy, did you see the e-mail ? Mar 03 08:53:38 oh, reading... Mar 03 08:54:22 hey, cool! Mar 03 08:54:26 ~hail dyoung Mar 03 08:54:28 * jbot bows down to dyoung and chants, "I'M NOT WORTHY!!" Mar 03 08:54:34 ~hail [g2] Mar 03 08:54:36 * jbot bows down to [g2] and chants, "I'M NOT WORTHY!!" Mar 03 08:54:47 dyoung rules! Mar 03 08:55:04 is it already in the repository? Mar 03 08:55:11 <[g2]> he's sleeping WAIT till he's in the channel! Mar 03 08:55:31 hehehe Mar 03 08:55:34 <[g2]> siddy, not yet that was my question I was curious if you guys wanted to move to 2.6.11 Mar 03 08:55:42 <[g2]> or stay on 2.6.9 Mar 03 08:56:01 I'd love to have 2.6.11! Mar 03 08:56:29 <[g2]> ok so far VoodooZ_lunch jbowler and siddy are want 2.6.11 Mar 03 08:56:36 <[g2]> s/are/all Mar 03 08:56:56 Well, either way. as long as the .bb files are on the repo so I can try it out. Mar 03 08:56:57 <[g2]> I'm doing a build after clearing tmp Mar 03 08:57:12 <[g2]> and the latest pull Mar 03 08:57:23 put my vote in there Mar 03 08:57:35 ~praise [g2] too! Mar 03 08:57:37 All hail [g2] too!! Mar 03 08:58:25 <[g2]> gcc-cross is builing now, I'm gonna head out to lunch Mar 03 08:58:35 me too. good appettite. Mar 03 08:58:46 <[g2]> I'll flash the scratch build when I get back Mar 03 08:59:07 <[g2]> If it works, I'll probably just tweak the defconfig to match the latest 2.6.9 and then push Mar 03 08:59:20 <[g2]> after a rebuild of the kernel and flashing of course Mar 03 08:59:45 <[g2]> I'll keep you guys posted Mar 03 09:01:26 [g2]-away: i definitely want to move to 2.6.11 Mar 03 10:00:39 03paulhargreaves * 10unslung/Makefile: Mar 03 11:46:21 03g2 07 * r1.2948 10openembedded/ (11 files in 3 dirs): Move OpenSlug to 2.6.11 kernel Mar 03 11:55:25 ~praise [g2] Mar 03 11:55:28 All hail [g2]! Mar 03 11:56:03 yay Mar 03 12:04:15 <[g2]> kergoth, are we the first OE distro to go to 2.6.11 ? Mar 03 12:54:01 I just started reading the yahoo group's processor clock speed discussion and it scares me to think that it's only runnign at 133Mhz. Mar 03 12:54:25 I'm tempted to try changing that bit to see if it changes the speed. Mar 03 12:54:52 <[g2]> VoodooZ_lunch, I think it be way cool if we were only running 1/2 speed Mar 03 12:55:44 me too but the specs seem to say we can't change it in software so it could also be a bad thing. Mar 03 12:58:14 Somebody also mentionned something about the /proc/cpu/alignment not being set properly in message 4741 Mar 03 12:58:49 <[g2]> I saw taht Mar 03 12:58:53 <[g2]> I saw that Mar 03 12:59:04 <[g2]> system stuff happens automatically Mar 03 12:59:08 * VoodooZ_work should start reading the yahoo group more often. Mar 03 12:59:16 <[g2]> user space is the issue Mar 03 13:00:24 yes but what if it's not happening properly. like non-critical tweaks. Mar 03 13:01:41 <[g2]> VoodooZ_work, I had back-ported the fix for 2.6.9, it's in the base release for 2.6.11 Mar 03 13:02:06 <[g2]> we currently generat 2706 halfword errors on startup which used to be like 11 or 12 Mar 03 13:02:15 Mar 03 13:19:18 drat. Mar 03 13:19:35 ~praise [g2] Mar 03 13:19:37 All hail [g2]! Mar 03 13:19:59 <[g2]> dyoung, is that a drat Mar 03 13:20:08 <[g2]> morning sunshine Mar 03 13:20:16 <[g2]> that 1-line drat ? Mar 03 13:21:45 a drat because I see what was missed! Mar 03 13:21:56 and in retrospect its obvious! Mar 03 13:22:18 <[g2]> a little sleep makes things clearer :) Mar 03 13:22:36 <[g2]> I know you're must making me feel useful Mar 03 13:22:49 <[g2]> I hope I didn't steal your glory Mar 03 13:23:25 What Glory? Mar 03 13:23:26 <[g2]> I made the change Lennert suggested about the ARCH/MACH too Mar 03 13:23:47 Yeah, I got that one too. Mar 03 13:24:04 <[g2]> That's in the pushed code Mar 03 13:24:13 And in the meantime, he sent me some good tips on how to debug the early "lkernel dont start" type problems. Mar 03 13:24:23 <[g2]> cool Mar 03 13:25:02 <[g2]> Sounds interesting, I'll have to check if I got those too Mar 03 13:25:43 <[g2]> ~praise dyoung Mar 03 13:25:45 All hail dyoung! Mar 03 13:26:01 the extra modification to head.S may not be neccessary. I added that i an effort to get early serial working. Mar 03 13:26:13 in debug mode. Mar 03 13:26:18 <[g2]> Ok, we can back that out Mar 03 13:26:39 <[g2]> did you see kas11's thoughts in openjtag ? Mar 03 13:26:43 You need the first .endm but everything after that in that elif stanza is probably not neccessary. Mar 03 13:27:00 <[g2]> about the n2 driver and the pci irq stuff Mar 03 13:33:17 Yeah, I just read that. Mar 03 13:43:48 ~praise dyoung Mar 03 13:43:51 All hail dyoung! Mar 03 13:44:12 my 2.6.11 desktop just spontaniously rebooted Mar 03 13:44:15 gr Mar 03 13:46:24 03jp30 * 10unslung/Makefile: promote xmail, demote vim Mar 03 13:52:20 later Mar 03 14:02:11 <[g2]> ls Mar 03 14:04:40 ~praise dyoung Mar 03 14:04:42 All hail dyoung! Mar 03 14:05:51 ~praise [g2] Mar 03 14:05:53 All hail [g2]! Mar 03 14:06:08 so, 2.6.11 is in the repository? Mar 03 14:06:25 <[g2]> siddy-away, yes Mar 03 14:06:45 cool, I will give it a try tomorrow! Mar 03 14:06:45 <[g2]> I'm having an issue with ehci right now Mar 03 14:06:56 a new issue? Mar 03 14:07:35 <[g2]> well things are a little different due to kernel size Mar 03 14:09:20 I guess we need some changes in switchbox, right? Mar 03 14:18:51 time to sleep, guys. have a good night and thx for your work on 2.6.11!! Mar 03 15:27:47 03dyoung 07 * r1.2951 10openembedded/packages/meta/openslug-packages.bb: add mt-daapd to feed list Mar 03 15:30:43 <[g2]> dyoung ping Mar 03 15:30:59 I was just about to head out. Mar 03 15:31:12 ~emulate rwhitby Mar 03 15:31:14 Well, my work here is done. Mar 03 15:31:21 <[g2]> heads up the ehci driver may not be loading Mar 03 15:31:32 its not? Mar 03 15:31:39 <[g2]> *may* Mar 03 15:31:50 <[g2]> I'm not sure if it's my local build or not Mar 03 15:32:09 <[g2]> have you pulled ? Or are you running locally Mar 03 15:32:17 I pulled. Mar 03 15:32:37 <[g2]> and you've built and flashed ? Mar 03 15:32:45 Yes. Mar 03 15:32:53 <[g2]> and ehci gets loaded ? Mar 03 15:33:04 <[g2]> or are you just running jffs2 ? Mar 03 15:33:22 just jffs2 Mar 03 15:33:35 but I dont see EHCI whiz by in the dmesg either. Mar 03 15:33:36 <[g2]> can you peek at the dmesg or boot up Mar 03 15:37:00 I think I tracked down the issue with the month being off by 1 too. Mar 03 15:37:20 <[g2]> awesome Mar 03 15:38:12 So why doesnt EHCI get loaded, I wonder. Mar 03 15:38:20 oh well, I'll keep wondering, I gotta get outta here. Mar 03 15:38:46 back later Mar 03 15:39:56 <[g2]> see you later Mar 03 18:36:29 * bigEE need assistance in #nslu2-general, anybody? Mar 03 19:47:51 03g2 07 * r1.2954 10openembedded/packages/linux/ (2 files in 2 dirs): Beware of last-minute changes ..... The one that got away.... I changed this at the last moment it breaks ehci startup and other stuff for reasons unknown to me. Mar 03 20:20:43 03g2 07 * r1.2955 10openembedded/conf/distro/openslug.conf: Allow PREFERRED_VERSION to be overridden in local.conf Mar 03 20:20:58 [g2]: thx Mar 03 20:23:33 <[g2]> rwhitby-web, My pleasure Mar 03 20:24:19 <[g2]> I tried to add in one to many changes and the one that I thought was most benign was the bad guy Mar 03 20:24:22 <[g2]> :) Mar 03 20:24:36 it's alpha software :-) Mar 03 20:39:33 scanline: ping Mar 03 20:41:07 pong Mar 03 20:41:17 Hi there, Mar 03 20:41:30 hello Mar 03 20:41:31 can CIA see into bk repo's for changsets? Mar 03 20:41:42 CIA only sees what the client scripts send it Mar 03 20:41:59 okay is there a client script for BK repos? Mar 03 20:42:03 yes Mar 03 20:42:14 where would I find that? Mar 03 20:42:17 kergoth wrote it, I don't really know anything about it :) Mar 03 20:42:21 k Mar 03 20:42:24 it's in the CIA documentation Mar 03 20:42:32 okay I'll check there...thanks. Mar 03 20:42:35 http://cia.navi.cx/doc/clients Mar 03 20:42:36 np Mar 03 20:42:42 scanline: do you know whether the current script for the nslu2-linux bk repo on bkbits.net sees the full changeset, or just the log message? Mar 03 20:42:48 hehe Mar 03 20:43:02 no idea Mar 03 20:43:19 I'm going to look at the doco and try to get this working. Mar 03 20:44:11 Look in triggers Mar 03 20:44:21 BitKeeper/triggers Mar 03 20:45:46 rwhitby-web: "the full changeset"? Mar 03 20:45:59 i dont send a bunch of bk binary data to CIA Mar 03 20:46:12 kergoth: sorry, the output of "bk send" Mar 03 20:46:22 ? Mar 03 20:46:29 bk send is bk specific format Mar 03 20:46:34 nothing cia would understand Mar 03 20:46:37 do you mean bk export -tpatch? Mar 03 20:46:44 we want to have the changesets emails to nslu-changesets@lists.sf.net Mar 03 20:46:46 no, i dont send bk screwy formatted data Mar 03 20:46:52 and no, i dont send diffs Mar 03 20:47:00 we send files and logs, thats it Mar 03 20:47:06 you probably want a new trigger. Mar 03 20:47:13 we currently say "bk send -r nslu-changesets@lists.sf.net" manually, and we're looking for a way to automate that Mar 03 20:47:31 yeah - looks like we need a new trigger Mar 03 20:48:01 i've been thinking about taking scanline's python svn trigger, with my modifications, and making a generic source notification trigger that can send to cia or email or whatever, but i'm lazy Mar 03 20:48:28 yeah,something like that Mar 03 20:48:49 zecke was working on something like that Mar 03 21:05:15 03dyoung 07 * r1.2956 10openembedded/packages/linux/openslug-kernel-2.6.11/defconfig: Added raid modules to defconfig **** ENDING LOGGING AT Thu Mar 03 23:59:56 2005