**** BEGIN LOGGING AT Thu Jul 21 23:59:57 2005 Jul 22 00:39:41 Made tha latest Openslug image, but it fails to come up. Symptoms are: Beeps, but LED goes red, Ethernet LED works, Power off button works, DHCP not started so cannot ping. I have no serial cable (though I've brought the pins out). I would guess the problem to be in the init scripts somewhere, as it looks as the kernel works. Is that a reasonable conclusion? Jul 22 00:40:44 To be more presise: Power LED is initially green, it seems to boot, then it beeps one time and the power LED goes red. Jul 22 03:43:46 jbowler: ah, missed that in vlan, did you do anything about it, or should I? Jul 22 05:56:14 jbowler: I just commited a fix for vlan-1.8, hopefully that will work Jul 22 06:49:13 <[g2]> DaKa2, hey change boy! Jul 22 06:49:23 hey :-) Jul 22 06:49:25 <[g2]> Or is that Mr. DaKa2 Dev now :) Jul 22 06:49:56 :) Jul 22 06:50:13 <[g2]> congrats.... Jul 22 06:50:15 my first commit, and nothing broke (yet) Jul 22 06:51:35 and the u-a thing is slowly getting done, a few more packages, a few more rounds in #OE and its could be committed to Jul 22 06:52:28 <[g2]> one step at at time Jul 22 06:53:10 yup... Jul 22 06:54:11 then there is the native build stuff, that probably needs a bit more talking, and then someone just have to put the pieces together Jul 22 06:54:50 <[g2]> rwhitby-asleep, mentioned updating the NativeCompile wiki page Jul 22 06:55:05 * dyoung-zzzz cant sleep. Jul 22 06:55:29 [g2]: I updated HowTo/OpenSlugNativeCompileEnvironment Jul 22 06:55:58 <[g2]> yeah... that's what I'm looking at now Jul 22 06:56:10 dyoung-zzzz: thats no good, how about counting slugs or something, bug fluffy white slugs Jul 22 06:58:08 <[g2]> what's on your mind dyoung-zzzz ? Jul 22 06:59:42 DaKa2, do you read stuff in #oe ? Jul 22 06:59:51 dyoung-zzzz: some Jul 22 07:00:27 something was said there I should read? Jul 22 07:00:33 well the ciabot's database got polluted/infected with that nightmare 3-head unmergable bit they experienced around 6 housr ago. Jul 22 07:00:52 so I'm sorry to say that your first commit will not be reported by CIA unless I enter it manually. Jul 22 07:01:07 and in reality, its easier for me just to say, Jul 22 07:01:23 hehe, well, no problem, just it got in the db everything is ok Jul 22 07:02:28 DaKa2 org.openembedded.nslu2-linux * foomatic openembedded/packages/vlan-foo bar changed some foo stuff Jul 22 07:02:42 their cia got it atleast :-) Jul 22 07:11:48 Fluffy white slugs.... Jul 22 07:32:20 Is BitBake Build Tool Core version 1.3.0, bitbake version 1.3.1 good enough to build OpenSlug? Jul 22 07:34:03 Ok: Let's put it like this: Does anyone use these versions to build openslug now and are able to build bootable images of R13? Jul 22 07:37:24 ingeba-toandfrom: thats what Im using, dunno about builting R13, but it has worked so far.. Jul 22 07:37:48 thats what I think Im using atleast.. Jul 22 07:38:05 Ok - when did you last build an image that booted? Jul 22 07:38:18 Sorry: Thanks DaKa2 ! :-) Jul 22 07:38:29 Just stressed getting things up and running Jul 22 07:43:15 :) Jul 22 07:53:51 DaKa2: Do you know it there are there any newer binary images or tested snapshots that are not on Berlios newer than 2.0beta from 22/6? Jul 22 07:55:09 2.3beta was just made, and put in svn, but I dont think there are any binary images yet Jul 22 07:55:13 There are no official binary releases apart from the ones that are at openslug.org. Any unofficial ones shouldn't be distributed. Jul 22 07:55:34 Fair enough, but all svn snapshots are on the dead Berlios site? Jul 22 07:55:43 Or? Jul 22 07:56:12 you can pull the 2.3beta from monotone somehow.. don't rememeber how... Jul 22 07:56:25 inegba, do you have a devel setup with monotone? Jul 22 07:56:27 Yes Jul 22 07:56:38 whats up with berlios anyway? Jul 22 07:57:04 Donno - "connection refused" since yesterday morning, but I hope it is only bitbake that resides there. Jul 22 07:57:56 monotone --db=monotone/nslu2-linux.db co --revision=t:OpenSlug-2.3-beta --branch=org.openembedded.nslu2-linux OpenSlug-2.3-beta Jul 22 07:58:06 that has worked for me Jul 22 07:58:13 hanjo: You're the best! Jul 22 07:58:37 no, the thanks have to go to rwhitby Jul 22 07:58:59 I think there was a problem with the source for something in there, that was only fixed in svn Jul 22 07:59:00 i just acted as a temp repository of his wisdom :) Jul 22 07:59:11 :-) Jul 22 08:00:41 DaKa2: Do you remember what the problem was? Jul 22 08:01:54 I think some of the nis utilities, yp-something Jul 22 08:02:08 Ah - ok - good. np - :-) Thanks! Jul 22 10:46:15 <[g2]> DaKa2, NAiL ping Jul 22 10:46:40 <[g2]> we should move some extra stuff out of openslug packages Jul 22 10:46:55 <[g2]> and create an openslug-extra Jul 22 10:47:03 <[g2]> stuff like monotone Jul 22 10:47:19 <[g2]> and possibly perl Jul 22 10:48:22 well, perl shouldn't be in openslug-packages, not for cross build, why move monotone away? Jul 22 10:49:34 <[g2]> first off it's the wrong version 19 Jul 22 10:49:47 :) well, that should be fixed Jul 22 10:49:52 <[g2]> second I don't think anyone uses it Jul 22 10:50:00 <[g2]> it take forever on my P4 Jul 22 10:50:09 we will, for the native build environ, probably, maybe Jul 22 10:50:37 <[g2]> maybe only witht he loft that's got the 422 (md5/sha1 in hw) Jul 22 10:50:56 wget the db and use monotone on that, no revs to check, just unpack the source Jul 22 10:51:15 <[g2]> I really doubt it Jul 22 10:51:33 <[g2]> you'd update the db then wget the file Jul 22 10:52:04 <[g2]> If we had a tiny monotone db for natve then maybe Jul 22 10:52:23 on some other host, update the db and gzip every 10 minutes, on the native wget that, unpack and run monotone Jul 22 10:52:35 might work, or not Jul 22 10:53:12 <[g2]> the packages are getting to a stable point Jul 22 10:53:24 <[g2]> it's not like we are change the be stuff that often Jul 22 10:53:50 <[g2]> and at a 4-5 hour turn time for the whole she'bang Jul 22 10:54:00 <[g2]> I don't know how useful it is Jul 22 10:54:05 <[g2]> btw it just upgraded to 21 Jul 22 10:55:11 <[g2]> I think on the slug using a checked out known db is the most likely thing for non-hardcore like us Jul 22 10:55:41 <[g2]> it makes me laught those guys trying to run qemu for the i386 on the slug Jul 22 10:55:45 <[g2]> laugn Jul 22 10:55:49 <[g2]> laugh Jul 22 10:56:58 hehe, well, they will get a ~20-30Mhz i386 Jul 22 10:57:10 That's a very generous estimate Jul 22 10:57:14 yes.. Jul 22 10:57:57 <[g2]> a much better idea in my mind were the guys gettin a 200Mhz ARM on the 2.8/3G P4 Jul 22 10:58:14 <[g2]> 233 iirc Jul 22 10:58:24 hehe, why not just buy a slug and get 266? :-) Jul 22 10:58:33 <[g2]> for debugging Jul 22 10:58:49 <[g2]> fully synchronous Jul 22 10:59:05 oh, that might be good... Jul 22 10:59:10 <[g2]> and full machine state/ core dump all in one :) Jul 22 10:59:44 <[g2]> for driver debugging oh yeah.... Jul 22 10:59:48 <[g2]> and kernel testing Jul 22 10:59:51 What is the point of qemu on the slug actually? I can't think of anything useful that you'd run on it Jul 22 11:00:00 it's neat but.... Jul 22 11:00:07 <[g2]> Tiersten, yes neat Jul 22 11:00:29 You wouldn't get alignment problems I guess :) Jul 22 11:01:43 <[g2]> depends on how good the emulation is Jul 22 11:01:51 <[g2]> I think you would get those Jul 22 11:02:25 It should work with no alignment problems since if it did then qemu would fail very early on Jul 22 11:03:51 <[g2]> I don't know what the status of the arm system emulation is. I know they've been working on it Jul 22 11:04:08 <[g2]> I'd imagine it'd trap out just like in the real hw Jul 22 11:04:18 yeah Jul 22 11:04:52 Okay. No endian issues since it would appear to be a x86 and most code is written assuming that :) Jul 22 11:04:55 <[g2]> but they'd be a *lot* easier to find :) Jul 22 11:05:13 woohoo. 1 reason! :) Jul 22 11:05:30 <[g2]> but that's for ARM on the x86 not the other way around Jul 22 11:28:26 <[g2]> hanjo, are you serving up nfs3+tcp from the slugs ? Jul 22 11:28:46 <[g2]> last I heard nfsv4 was still green Jul 22 11:29:08 <[g2]> Ill check the logs later Jul 22 11:29:09 <[g2]> cheers Jul 22 11:45:35 NAiL , I just sent you a patch to cross-compile screen Jul 22 15:40:37 DaKa2: I didn't do anything to vlan in the end, it builds ok from the non-safe make target (the default), just with 'safe' I set HOME=/dev/null and cc doesn't like -I/dev/null/linux/include... Jul 22 15:51:04 [g2]: perl should be removed from openslug-packages, but only cause it doesn't work and has been replaced by the native build. Nothing else which builds and runs should *ever* be removed from openslug-packages. If you need a stable subset, then create a new .bb - openslug-packages.bb is defined to be *all* packages which we know at least build (and preferably run) on openslug. Jul 22 19:29:29 rwhitby-away: oh, forgot to say, u-a is ready for the next inspection and round of review, all needed packages for openslug atleast are fixed Jul 22 19:31:33 jbowler: well, I ended up cleaning up the vlan bb a bit anyway, it had some more problems too. It was my one of my first bb packages made :-) Jul 22 19:32:31 DaKa2: perhaps we can do that at the same 24hr time as we last did (tonight my time) Jul 22 19:33:36 rwhitby-away: that would work for me Jul 22 19:35:42 ok - feel free to start it without me if I'm not around. Just make sure some of the main #oe guys are there Jul 22 19:39:18 rwhitby-away: ok, but I'll probably wait for you, somehow it feels better having you around :-) Jul 22 19:39:34 safety in numbers :-) Jul 22 19:40:00 hehe, yeah, bring in the nslu2 army :) **** ENDING LOGGING AT Fri Jul 22 23:59:56 2005