**** BEGIN LOGGING AT Tue Jul 12 23:59:57 2005 **** BEGIN LOGGING AT Wed Jul 13 07:58:57 2005 Jul 13 14:26:55 [g2]: Hey man Jul 13 14:27:01 dyoung: yo Jul 13 14:27:12 Any samba users out there? Jul 13 14:27:23 It looks like it *almost* works. Jul 13 14:30:52 Yo G Jul 13 14:30:59 <[g2]> beewoolie, Hey! Jul 13 14:31:12 <[g2]> excellent what's up ? Jul 13 14:32:28 doodz. Jul 13 14:32:42 I tried samba just now. I get an...interesting result... Jul 13 14:32:55 mb: \> ls Jul 13 14:32:55 D 0 Thu Jun 1 02:34:13 2000 Jul 13 14:32:55 D 0 Thu Jun 1 02:34:13 2000 Jul 13 14:32:55 D 0 Thu Jun 1 02:34:13 2000 Jul 13 14:32:57 D 0 Thu Jun 1 02:34:13 2000 Jul 13 14:32:59 56259 blocks of size 2097152. 25439 blocks available Jul 13 14:33:09 If this isn't clear, the point I'm trying to make is that there are no filenames. Jul 13 14:33:29 there Jul 13 14:33:33 Hey Jul 13 14:33:37 :) Jul 13 14:33:46 I was just saying that I tried samba, not working as expected. Jul 13 14:33:55 It's magic Jul 13 14:33:57 smbclient (and windows xp) see no file names. Jul 13 14:34:00 (or [g2]) Jul 13 14:34:05 huh? Jul 13 14:34:10 <[g2]> windows xp worked for me Jul 13 14:34:13 What feed did you pull from? Jul 13 14:34:23 <[g2]> did you set your workgroup ? Jul 13 14:34:24 I love that rejoinder. Really I do. Jul 13 14:34:34 Yep. that's not the problem, let me assure you. Jul 13 14:34:39 I'm playing all my mp3s from my slug running samba Jul 13 14:34:42 <[g2]> samba on windows xp worked for me :) Jul 13 14:34:44 The lack of filenames has nothing to do with the group. Jul 13 14:34:50 Oh, I believe you. Jul 13 14:34:56 no, if you don't see any filenames, I know what the problem is Jul 13 14:35:02 Share? Jul 13 14:35:06 but that is supposedly fixed in both stable and unstable Jul 13 14:35:12 Ah. Jul 13 14:35:16 Did you build it yourself? Jul 13 14:35:22 I'm using version -r8 IIRC. I'll check. Jul 13 14:35:28 No. I pulled from the feed. Jul 13 14:35:32 Remember, that's my job. Jul 13 14:35:36 * NAiL goes WTF? Jul 13 14:35:44 Also, there still isn't an update for nfs-utils. Jul 13 14:35:57 * NAiL repeats himself Jul 13 14:36:06 * beewoolie repeats NAiL Jul 13 14:36:20 * beewoolie 's lunch repeats Jul 13 14:36:22 * dyoung dunnos Jul 13 14:36:33 What version of nfs-utils should be out there? Jul 13 14:36:40 ...on the feed, that is. Jul 13 14:36:41 beewoolie: Hang on a sec, and I'll install samba on my other slug Jul 13 14:36:51 I tested it a few days ago Jul 13 14:37:04 The thing that caused the problems were some samba_cv_ vars in the site file Jul 13 14:37:07 I can uninstall it and reinstall. Jul 13 14:37:20 go ahead Jul 13 14:37:20 As I'm not building, that shouldn't be the issue for me. Jul 13 14:37:37 no, but it should be fixed in the feed Jul 13 14:37:42 s Jul 13 14:38:51 * beewoolie is Downloading http://ipkg.nslu2-linux.org/feeds/openslug/oe/samba_3.0.14a-r8_armeb.ipk Jul 13 14:39:33 Well, it's the last rev. It really *should* work. Jul 13 14:40:02 Unless there is a problem with the way that ipkg performs installs. Jul 13 14:40:29 you're running from disk, no? Jul 13 14:40:41 No dice...and yeah. Jul 13 14:41:20 <[g2]> beewoolie, this isn't IPv6 is it ? Jul 13 14:41:32 I don't yet have ipv6 in the kernel. Jul 13 14:41:45 Besides, NAiL says he knows where this comes from. Jul 13 14:42:04 My md5sum on smbd is 6bf270c4274dd7d3db5156cb7820b7c8 smbd Jul 13 14:42:13 <[g2]> nod. Jul 13 14:42:41 ugh Jul 13 14:42:49 my slug just broke on something Jul 13 14:42:49 root@gunslinger:~# dmesg Jul 13 14:42:50 -sh: dmesg: Input/output error Jul 13 14:42:51 Nu? Jul 13 14:43:14 Probably doesn't like existential literature. Jul 13 14:43:44 root@gunslinger:~# reboot Jul 13 14:43:45 -sh: reboot: Input/output error Jul 13 14:43:47 Strange Jul 13 14:43:59 That means you have lost access to / Jul 13 14:44:03 Check the mount table. Jul 13 14:44:13 /dev/sda2 on / type ext2 (rw) Jul 13 14:44:22 Perhaps /tmp ? Jul 13 14:44:27 heh Jul 13 14:44:31 or /dev Jul 13 14:44:35 there's absolutely *nothing* in / Jul 13 14:44:46 * NAiL restarts slug Jul 13 14:47:11 hmm Jul 13 14:47:19 * NAiL suspects his iPod just died Jul 13 14:48:00 NAiL, OT, but what kind of mains power do you have there? Jul 13 14:48:06 220 Jul 13 14:49:33 you're thinking brick for enclosure? Jul 13 14:50:13 Yeah, I need to fish it out from behind this pile of slugs and '244's... Jul 13 14:51:07 What does the enclosure need? I have a universal one Jul 13 14:52:02 well, thats the choice. The one thats already in the case and sound like a chainsaw has this weird 4-pin din connector. Jul 13 14:52:09 oh Jul 13 14:52:26 I think the PS was universal, but I need to get it. Jul 13 14:54:57 Make a new one out of 244s? Jul 13 14:55:49 I was waiting for someone to make a '244 comment. :-) Jul 13 14:57:06 Does TI automatically send a truck around with 244s to your house? :) Jul 13 14:57:08 Is it possible that I'm the only person pulling files from the feed? Jul 13 14:57:22 beewoolie: I do too Jul 13 14:57:32 My slug isn't booting Jul 13 14:57:39 Wierd. Let me see if I can do a complete update. Jul 13 14:58:05 What does this mean: Jul 13 14:58:10 root@rojo:/# ipkg upgrade Jul 13 14:58:10 Upgrading ¾ÿþ| base-passwd on root from 3.5.9-r0 to 3.5.9-r1... Jul 13 14:58:23 What's with the wierd characters? Jul 13 14:58:56 I'm rebooting for good measure. Jul 13 14:58:58 You're running OpenSlug 2.0-beta? Jul 13 14:59:05 Yep. Jul 13 14:59:09 ie, you didn't build openslug yourself either? Jul 13 14:59:16 Blame [g2]. He told me to. Jul 13 14:59:20 Nope. Jul 13 14:59:23 I don't know what the wierd chars are, but I've seen them myself Jul 13 14:59:32 They don't seem to affect anything though Jul 13 14:59:36 ..and it doesn't bother you? Jul 13 14:59:42 yes, it does Jul 13 15:01:20 What's the name of that pl chip again? Need to find that driver... Jul 13 15:01:48 pl2303 Jul 13 15:01:53 thx Jul 13 15:02:50 <[g2]> beewoolie, are you looking on the slug side or PC side ? Jul 13 15:03:16 [g2]: I don't think you understand the problem. Jul 13 15:03:32 I connect to the smb server with smbclient. From anywhere. Jul 13 15:03:38 There are no file names. Jul 13 15:03:51 Also, this .recovery thing is really annoyng. Jul 13 15:04:07 Every time I reboot, I have to remove the .recovery file. Jul 13 15:05:26 <[g2]> I didn't used smbclient I just did a Windows Explorer and openend the Network folder and logged in to the machine (... apex) Jul 13 15:05:48 Same result with windows xp. Jul 13 15:05:50 <[g2]> I was able to create files and remove files Jul 13 15:05:58 We've been over this before. Jul 13 15:06:06 Try the versions in the feed. Jul 13 15:06:20 <[g2]> I did need to make the directory 777 Jul 13 15:06:36 That's not the problem, believe me. Jul 13 15:07:12 beewoolie: I'll look at it if I get my slug back up Jul 13 15:07:20 We have been having this problem for time time. The feed is out of sync with the repo. I don't know how it gets updated, but I think something ought to be done. Jul 13 15:07:38 <[g2]> beewoolie, NAiL has been working on it Jul 13 15:07:49 <[g2]> he's the guy that builds and releases Jul 13 15:08:04 <[g2]> we are working on getting all the packages statused Jul 13 15:08:20 And this .recovery thing is really inconvenient. If I'm using ext3, there is no reason for it. Jul 13 15:08:59 <[g2]> well the current situation is in flux. I'm unhappy about it. Jul 13 15:09:58 On another topic...I've been working to get back in the kernel dev saddle. Jul 13 15:10:11 The BK switch is causing lots of chaos. Jul 13 15:10:14 Good chaos. Jul 13 15:10:35 In spite of the warts, I think that git and the derivatives are a huge improvement over BK. Jul 13 15:11:06 Yeah. I agree. Chaos now but once it gets all sorted it should be good. Jul 13 15:11:07 What the ? Jul 13 15:11:20 Ok, my slug isn't booting. I can't get serial output. Jul 13 15:11:34 It died in the middle of starting samba Jul 13 15:11:37 joy Jul 13 15:11:44 nail, if thats 2.6.12.2, you dont have serial output until after init starts Jul 13 15:12:04 dyoung: I should see redboot anyhoo Jul 13 15:12:05 <[g2]> and preferred kernel 2.6.11.2 doesn't work either Jul 13 15:12:47 [g2] it should. the preferred kernel provider changed back to openslug-kernel. Jul 13 15:13:19 <[g2]> it doesn't the kernel modules except ixp wind up in 2.6.12.2 Jul 13 15:13:20 I admit I havnt tried it myself though. Jul 13 15:13:30 NAiL: BTW, samba didn't install a link in rc5.d. It doesn't start the server when the system boots. Jul 13 15:13:43 <[g2]> beewoolie, nod. Jul 13 15:13:54 oh well. Jul 13 15:14:07 <[g2]> and the depmods are fsck'ed Jul 13 15:14:23 [g2] I'm not smart enough to make it work. Jul 13 15:14:28 And, still no dice. Reboot doesn't clear it. I guess this really *isn't* windows. Jul 13 15:14:36 <[g2]> and you won't be able to mount ext2/3 cause they are modules Jul 13 15:15:24 ext2 is in the kernel i thought... becuase I could turnup it. ext3 is modular though. Jul 13 15:15:55 Ether way though, I'm not smart enough to make it work. Jul 13 15:16:29 <[g2]> dyoung, is quite smart Jul 13 15:16:50 <[g2]> maybe not motivated to make it work, but certainly smart enough Jul 13 15:17:17 Youre smarter than me, so I left those files in your capable hands. :-) Jul 13 15:17:25 (secretly hoping you'd magically fix them) Jul 13 15:18:07 <[g2]> actually, there's a later patch that's required and some stuff for the early printk Jul 13 15:18:40 what the? Jul 13 15:18:59 * NAiL beats Windows to a pulp Jul 13 15:19:06 <[g2]> I think beewoolie is the only one here that would probably really understand it (I didn't get it and lennert gave some crib notes) Jul 13 15:19:35 can you forward me some info to this patch? Jul 13 15:19:46 <[g2]> sure Jul 13 15:20:13 <[g2]> hey beewoolie do you think you could help us get the kernel debug working in the 2.6.12.x kernel ? Jul 13 15:22:04 <[g2]> dyoung, beewoolie sent Jul 13 15:22:14 ah well, I'll check back later when not as cranky. Jul 13 15:23:05 finally Jul 13 15:23:28 * NAiL adds a note on the wiki that pl2303 does not have a driver for Windows XP x64 Jul 13 15:23:50 Windows can Bite Me. Jul 13 15:24:02 ... and it has, many a time. Jul 13 15:24:20 bbiaw Jul 13 15:24:36 EXT2-fs warning: mounting unchecked fs, running e2fsck is recommended Jul 13 15:24:36 /boot/disk: 47: cannot create /mnt/.reKernel panic - not syncing: Attempted to kill init! Jul 13 15:24:39 covery: Input/output error Jul 13 15:24:55 It didn't even finish printing the error before panic Jul 13 15:26:50 Every time I reboot, I have to remove the .recovery file. Jul 13 15:27:14 Once in a while, I have to remove the .recovery as well Jul 13 15:27:24 jbowler: I'm going to add a boot-time script to remove it. Jul 13 15:27:33 <[g2]> I've never had to rm .recover Jul 13 15:27:35 If you don't get through to the user rmnologin step you have to remove it. Jul 13 15:27:52 beewoolie: there is a boot time script to remove it. Jul 13 15:28:06 jbowler: not on boot-up there isn't Jul 13 15:28:19 Or if there is, there's something broken. Jul 13 15:28:25 It's there after my system boots. Jul 13 15:28:48 The script is /etc/rc5.d/S99rmrecovery Jul 13 15:29:21 That's not what's in my version. Jul 13 15:29:29 My version removes /etc/nologin.boot. Jul 13 15:29:41 [g2] you gotta be kidding me.... Jul 13 15:29:47 That's /etc/rc5.d/S99rmnologin Jul 13 15:29:56 That is a different script. Jul 13 15:29:57 It actually needs the empty structure?! Jul 13 15:29:57 I don't have any other. Jul 13 15:30:04 <[g2]> dyoung, kidding you about ? Jul 13 15:30:13 nooo Jul 13 15:30:15 (It's hard to read the file names as they're in cyan. Jul 13 15:30:16 argh Jul 13 15:30:20 my iPod has died Jul 13 15:30:32 beewoolie: what arguments did you provide to turnup to make the root partition? Jul 13 15:30:36 It looks like there are more problems with the openslug distro that we thought. Jul 13 15:30:38 { { .foo = } , { }, } Jul 13 15:30:59 jbowler: I didn't make the partition anew as it had /home which I needed. Jul 13 15:31:16 <[g2]> , {} Jul 13 15:31:20 So what is creating the .rmrecovery file then? Jul 13 15:31:22 I copied the files from flash. Jul 13 15:31:22 scsi0 (0:0): rejecting I/O to offline device Jul 13 15:31:23 scsi0 (0:0): rejecting I/O to offline device Jul 13 15:31:31 Thats easily adjusted for. I'll try it later. Jul 13 15:31:37 And wrote your own /linuxrc right? Jul 13 15:31:45 No, I didn't write anything. Jul 13 15:31:54 <[g2]> dyoung, I tried it and it still did'nt boot all the way Jul 13 15:32:05 Turnup doesn't have an option to preserve the /home directory. Jul 13 15:32:10 I had to do the setup by hand. Jul 13 15:32:34 So you got the /linuxrc wrong! You shouldn't be creating the .recovery file from /linuxrc if you don Jul 13 15:32:37 't remove it! Jul 13 15:32:54 Well, turnup doesn't work, then, does it. Jul 13 15:33:09 You didn't use it, it works fine if you use it. Jul 13 15:33:10 If we have to use turnup, then it has to allow users to keep the files on the HD. Jul 13 15:33:40 I'll be happy to try a new version. Jul 13 15:33:52 No user has to use turnup. You can construct your own root partition and pivot to it, but if you use bits of the existing boot system you have to get it right. Jul 13 15:34:17 I'll be glad to follow some instructions if you have them. Jul 13 15:34:30 all I've been told is what I see when the system boots. it says to use turnup. Jul 13 15:34:34 Flash the system with OpenSlug-2.0 Jul 13 15:34:38 Boot Jul 13 15:34:41 Run turnup init Jul 13 15:34:45 Reboot Jul 13 15:34:54 Doesn't init erase the HD? Jul 13 15:34:59 mkfs a new disk partition Jul 13 15:35:25 run turnup disk -i Jul 13 15:35:33 yeah, that erases the partition. Jul 13 15:35:38 Not going to happen. Jul 13 15:36:32 Then skip the mkfs step. Jul 13 15:38:20 Hmm. Let's read the turnup documentation... Jul 13 15:38:28 with Jul 13 15:38:28 -i it must be completely empty, Jul 13 15:38:43 Mine wasn't Jul 13 15:38:50 and it worked fine Jul 13 15:38:51 I think there is something awry here. Either the doc is wrong, or ... Jul 13 15:39:07 I moved all the stuff to a temporary dir though Jul 13 15:39:15 I don't have such an option. Jul 13 15:39:28 jbowler: So, which is it? Does -i require the partition to be empty? Jul 13 15:39:54 Here's more from the turnup script... Jul 13 15:39:56 # check_rootfs [-i] Jul 13 15:39:57 # Make sure the candidate rootfs is empty Jul 13 15:40:00 beewoolie: I can confirm that samba is broken in -r8... This is *very* odd Jul 13 15:40:14 NAiL: which version are you running? Jul 13 15:40:21 Imean, which version is working for you? Jul 13 15:40:33 beewoolie: no Jul 13 15:40:55 beewoolie: -r8 Jul 13 15:41:10 Give me half an hour, and I think it should be fixed. Jul 13 15:41:12 NAiL: That's odd, then. Jul 13 15:41:36 jbowler: Then why does the script say it has to be? Jul 13 15:41:50 I think you can work that one out yourself ;-) Jul 13 15:42:09 jbowler: Not a very helpful answer. Jul 13 15:42:27 I'm actually suprised that you care so little to get this stuff working correctly. Jul 13 15:43:13 It's thirteen lines down from the line you just quoted. Jul 13 15:43:37 I care deeply about not overwriting valuable data. Jul 13 15:44:20 beewoolie: I see where the error was. I hadn't set up the push correctly. It did svn up in the wrong place (So the packages got updated, but not the site file) Jul 13 15:44:27 I saw that. It still isn't clear what it is going to do. if it is going to overwrite, how do I know it isn't going to overwrite something I care about. Jul 13 15:44:27 I have curious feelings about someone who has a disk with data on it, gets a new NSLU2 system, and expects to be able to install a beta openslug on top of that data. Jul 13 15:44:38 beewoolie: So the *correct* samba will be in the feed shortly. Same rev though Jul 13 15:44:55 I didn't run turnup. Jul 13 15:44:59 i don't trust it. Jul 13 15:45:03 beewoolie: ! you how do you know? If you don't know how does anyone know! Jul 13 15:45:14 I trust the rest of GNU/Linux since I've been using it for years. Jul 13 15:45:57 Yes - then don't use it. You don't need to. But don't try to pick holes in it by taking a part of it and using it wrongly. Jul 13 15:46:13 Um, openslug is broken without it, ATM. Jul 13 15:46:19 See the catch 22? Jul 13 15:46:32 Why? Jul 13 15:46:47 I've lost interest in this conversation. Jul 13 15:46:56 Why do you need to use turnup? Jul 13 16:00:50 beewoolie: samba_3.0.14a-r8 has been fixed and is in the feed now. I've removed/reinstalled and know it works now Jul 13 16:01:05 Sorry for the inconvenience, and thank you for riding with OpenSlug airlines. Jul 13 16:01:38 so, if I uninstall and reinstall, it will grab the new version? Jul 13 16:01:44 correct Jul 13 16:03:56 I'm playing mp3's off r8 now. Jul 13 16:05:36 NAiL: Yipee! Jul 13 16:05:44 OK. How about getting the nfs-utils updated? Jul 13 16:05:45 np, thanks for pointing it out Jul 13 16:05:49 working on that as well Jul 13 16:05:59 NAiL: Thanks for being so quick about fixing it. Jul 13 16:06:04 Excellent. Jul 13 16:06:09 NOTE: package nfs-utils-1.0.6-r2: task do_unpack: failed Jul 13 16:06:16 It fails while... unpacking? Jul 13 16:06:17 Hmm. Jul 13 16:06:36 I haven't uninstalled it, so ipkg install doesn't even try to download it. Jul 13 16:06:42 I thought that there would be a new version. Jul 13 16:08:26 jbowler: ping? Jul 13 16:08:35 I'm getting this... any idea? Jul 13 16:08:35 cp: cannot stat `/home/repvik/stable/OpenSlug-2.0-beta/openembedded/packages/nfs-utils/nfs-utils-1.0.6/nfsserver': No such file or directory Jul 13 16:09:08 NAiL: samba works great. Thanks a bunch. Jul 13 16:09:12 np Jul 13 16:10:02 <[g2]> beewoolie, so samba's happy now ? Jul 13 16:10:17 [g2]: indeed. Jul 13 16:10:25 <[g2]> excellent Jul 13 16:10:28 <[g2]> NAiL, THX Jul 13 16:10:29 It doesn't startup by default, but I'll fix that. Jul 13 16:10:34 <[g2]> right Jul 13 16:10:36 NAiL: missing file Jul 13 16:11:27 jbowler: yes. any idea why? Jul 13 16:11:30 the initfile should be in 'files' Jul 13 16:12:22 Sorry, nfs-utils-1.0.6 Jul 13 16:12:24 aha.., forgot to do svn add before committing earlier Jul 13 16:14:53 Anyone who knows makefiles... Why doesn't this exit with an error? Jul 13 16:14:54 ( cd OpenSlug-2.0-beta ; source setup-env ; bitbake openslug-packages | tee ~/stable/make.log ) || ( mv make.log BUILD-ERRORS && false ) Jul 13 16:14:59 when bb fails Jul 13 16:19:37 beewoolie: Any other packages you're having problems with? :) Jul 13 16:19:48 Not right now. Jul 13 16:19:55 I'm looking at your fragment... Jul 13 16:20:27 Are you sure you want it do that? Jul 13 16:20:39 exit code of bitbake is irrelevant - the tee always succeeds Jul 13 16:20:52 If either of the clauses succeeds, it will not fail. Jul 13 16:20:59 jbowler: aha, that's what's doing it Jul 13 16:21:00 || means either can be true Jul 13 16:21:22 NAiL: it's difficult to do, you need to get the exit code out of bitbake (e.g. a temp file) Jul 13 16:21:39 well.. the output to console isn't really needed Jul 13 16:21:45 so I could just drop the tee Jul 13 16:22:07 bb || touch /tmp/failed Jul 13 16:22:51 <[g2]> that's a bad idea for threading Jul 13 16:23:04 tmp/failed.$$ Jul 13 16:23:18 <[g2]> or .failed Jul 13 16:23:39 The $$ version is probably safer. Jul 13 16:24:37 <[g2]> yeah **** ENDING LOGGING AT Wed Jul 13 23:59:56 2005