00:28.18 | *** join/#nslu2-linux thraxisp (n=thraxisp@24.139.16.154) |
00:43.57 | *** join/#nslu2-linux yvasilev (n=yvasilev@189.140.150.8) |
01:07.14 | rwhitby | Annirak: what's the problem you're having compiling Apex? |
01:07.30 | Annirak | march=arm |
01:07.45 | Annirak | Lame, I realise, but that is it. |
01:10.55 | rwhitby | so you've compiled it now? |
01:12.40 | Annirak | no... I've just run make setup-host-debian |
01:12.45 | Annirak | it's nearly done |
01:13.33 | rwhitby | you don't need any of that to simply build apex |
01:13.41 | rwhitby | you do need a cross compiler though. |
01:13.58 | rwhitby | and yes, the easiest way to get one of those is via the master makefile. |
01:14.06 | Annirak | <PROTECTED> |
01:14.12 | Annirak | after running the master makefile |
01:14.14 | Annirak | *sigh* |
01:14.23 | *** join/#nslu2-linux kolla_ (i=kolla@firda.kolla.no) |
01:14.28 | Annirak | I gotta run |
01:14.31 | Annirak | or I will be late |
01:14.37 | Annirak | I'll be on tomorrow |
01:14.46 | Annirak | maybe I'll figure it out then |
01:15.20 | rwhitby | right, you need to build a cross compiler. either via building a firmware image (e.g. slugosle) or some other means |
01:18.45 | rwhitby | 2.6.24 is released |
01:57.30 | joshin | Yay. Maybe that oom-killer problem will be fixed by 2.6.9876 |
01:57.53 | ka6sox-laptop | joshin, heh |
02:29.56 | *** join/#nslu2-linux scant (n=scant@24.174.0.137) |
02:49.38 | *** join/#nslu2-linux Annirak_ (n=chatzill@S0106000129d15bac.vc.shawcable.net) |
03:14.01 | Annirak_ | I'm a noob at crosscompiling. How do I get arm-linux-gcc installed on ubuntu? |
03:21.04 | mwester | http://www.nslu2-linux.org/wiki/Development/MasterMakefile |
03:31.53 | CIA-15 | 03bzhou * r7653 10optware/trunk/platforms/toolchain-openwrt-ixp4xx.mk: toolchain-openwrt-ixp4xx: switch to toolchain of kamikaze 7.09 releases |
04:39.00 | *** join/#nslu2-linux DrZimmerman (n=theo@2001:5c0:8fff:fffe:0:0:0:6af1) |
05:13.38 | *** join/#nslu2-linux osas (n=nnnnnnnn@nslu2-linux/osas) |
05:42.48 | *** join/#nslu2-linux jacques (n=jacques@nslu2-linux/jacques) |
05:49.50 | *** join/#nslu2-linux Zug\hO_BokPATN^ (n=Sinclair@p57B49C73.dip0.t-ipconnect.de) |
07:08.30 | *** join/#nslu2-linux drath_ (i=vmaster@p5B07FF9C.dip.t-dialin.net) |
07:39.14 | *** join/#nslu2-linux kolla (n=kolla@haugum.uninett.no) |
07:45.43 | *** join/#nslu2-linux mithro (n=tim@ppp246-117.static.internode.on.net) |
08:00.05 | *** join/#nslu2-linux eiffel (n=eiffel@p5B050D83.dip0.t-ipconnect.de) |
08:10.29 | *** join/#nslu2-linux ch4os_ (n=ch4os@unaffiliated/ch4os/x-059673) |
09:44.37 | *** join/#nslu2-linux mithro (n=tim@ppp246-117.static.internode.on.net) |
11:41.25 | *** join/#nslu2-linux erikja (n=erikja@0x57314066.abnxx10.adsl-dhcp.tele.dk) |
12:30.23 | *** join/#nslu2-linux SyNko (n=SyNko@79.0.132.253) |
12:38.58 | *** join/#nslu2-linux lyakh (n=lyakh@p57BD3065.dip0.t-ipconnect.de) |
12:58.08 | *** join/#nslu2-linux ^sn[MsKGwdTn^Es (n=Sinclair@p57B49C73.dip0.t-ipconnect.de) |
14:23.12 | *** join/#nslu2-linux Ragnor (n=Ragnor@dslb-084-060-135-221.pools.arcor-ip.net) |
14:26.39 | *** join/#nslu2-linux GPSFan (n=kenm@12.10.255.246) |
15:27.03 | *** join/#nslu2-linux HopsNBarley (n=hops@nslu2-linux/HopsNBarley) |
15:28.30 | *** join/#nslu2-linux yvasilev (n=yvasilev@dune.math.cinvestav.mx) |
15:31.13 | *** join/#nslu2-linux yvasilev (n=yvasilev@dune.math.cinvestav.mx) |
15:48.03 | *** join/#nslu2-linux ch4os_ (n=ch4os@unaffiliated/ch4os/x-059673) |
16:02.54 | *** part/#nslu2-linux SyNko (n=SyNko@79.0.132.253) |
16:24.18 | *** join/#nslu2-linux yvasilev (n=yvasilev@dune.math.cinvestav.mx) |
17:25.33 | *** join/#nslu2-linux hillct (n=H@cpe-024-211-242-232.nc.res.rr.com) |
17:50.43 | Annirak | Do I have to go through the master makefile to build apex? |
17:50.46 | *** join/#nslu2-linux gouki (n=gouki@ubuntu/member/gouki) |
17:51.23 | Annirak | How would I install a cross-compilation environment without the master makefile |
17:52.41 | eno | Annirak: read master makefile and you will understand |
17:52.57 | *** join/#nslu2-linux sdm485 (n=steve@S01060012171db828.cg.shawcable.net) |
17:53.03 | eno | it really depends on which cross compilation env you're talking about |
17:53.53 | Annirak | eno: I'm trying to buid apex for debian on an ubuntu-i686 platform. |
17:53.53 | eno | but why not just use the master makefile ? |
17:54.09 | Annirak | Updating the monotone caches is taking forever |
17:54.36 | Annirak | it's been running for ~20 minutes and it's only at 900/21000 |
17:54.38 | eno | that's certainly true |
17:55.11 | eno | i vaguely remember that you can copy a snapshot of mtn db from somewhere |
17:55.13 | sdm485 | Annirak: Go to venge.net and get monotone-0.28. You can also tell Ubuntu package manager to lock the version and stop trying to update it. |
17:55.46 | sdm485 | It takes my 2.4G celeron about 10 hours to migrate a db. Not worth the time. |
17:56.08 | Annirak | I have a better idea: sdm485, I know you have an apex image with 128/256MB support. Would you be willing to send that to me? |
17:57.27 | sdm485 | Sure. |
17:57.53 | Annirak | =) |
17:58.01 | Annirak | Dcc? |
17:58.10 | sdm485 | This will be the second stage loader? |
17:58.14 | Annirak | Yes |
17:58.27 | Annirak | Redboot->Apex->Debian |
17:58.54 | sdm485 | Working..:) |
18:01.02 | Annirak | I want to use the installer image, which means 2.6.18 |
18:01.15 | Annirak | I was figuring on upgrading to 2.6.22-3 afterwards |
18:02.12 | Annirak | Will that be a problem with apex? |
18:04.29 | sdm485 | I doubt it. Can you send me a post with your address? yours appears blocked by Yahoo and I can't attach anything to it. |
18:05.06 | Annirak | Ah. Yeah, that address is bad. One sec |
18:10.32 | *** join/#nslu2-linux dwery (n=dwery@nslu2-linux/dwery) |
18:15.00 | Annirak | sdm485: did you get my message? |
18:15.25 | sdm485 | I have a question about the mlock function call. When I launch memtester as a user, it refuses to lock memory which makes sense. When I run memtester as root, it will happily allocate whatever I ask to the point of crashing the system. Does this make sense? |
18:15.32 | sdm485 | Annirak: not yet |
18:15.53 | Annirak | You don't see the PM I sent you? |
18:16.15 | sdm485 | Oops. didn't see it.. |
18:16.31 | Annirak | np |
18:17.56 | sdm485 | On it's way. |
18:18.48 | Annirak | :) |
18:19.54 | sdm485 | Good. I think apex defaults to copying the maximum possible kernel size so about the only problem I see you having is if the location of the start of the kernel is different in a debian image |
18:21.37 | Annirak | I think it should be the same |
18:21.45 | Annirak | I hope it shoudl be the same |
18:22.00 | sdm485 | :) |
18:23.57 | sdm485 | It is also possible the area of the flash used for storing any additional env strings is actually used by debian. But if it makes the RAM available, there is no need for anything else. |
18:24.07 | *** join/#nslu2-linux thraxisp (n=thraxisp@ottgate.precidia.com) |
18:24.48 | Annirak | Do I need to pad the apex.bin or anything? |
18:24.53 | Annirak | The old one was 127k |
18:24.59 | Annirak | the one you sent me is 48x |
18:25.03 | Annirak | 48k |
18:26.03 | sdm485 | Hmm. Not sure. That is the problem. I would try it without padding first myself as I think that is how it is pieced together in slugosbe |
18:26.28 | Annirak | ok |
18:26.39 | Annirak | Maybe slugimage deals with that internally |
18:27.12 | sdm485 | I am going to look through the logs on slugimage to see if there are any clues. |
18:27.37 | Annirak | So what I've done here is taken the debian-installer image, and replaced the apex.bin portion and replaced it with your apex.bin then reassembled it with slugimage |
18:27.56 | *** join/#nslu2-linux pleemans_ (n=peter@d51A5E76A.access.telenet.be) |
18:34.38 | sdm485 | Annirak: It looks like 1 128k flash block is allocated for the <loader> and that is where apex is put. The kernel is loaded into the 5th block. Probably padding makes no difference as long as the total size is less than 1 block. |
18:36.11 | Annirak | Ok. I'm loading the new image in now. |
18:39.30 | Annirak | rebooted now... waiting for the beep |
18:39.59 | Annirak | there's the beeps |
18:40.08 | sdm485 | cool :) |
18:40.50 | Annirak | well, that doesn't work. |
18:41.01 | Annirak | The kernel still only sees 32MB |
18:41.17 | Annirak | Looks like I'll have to compile a kernel too if I want to go any further with that. |
18:42.05 | sdm485 | I bet the kernel has a hardcoded command line. You need the version that will accept command lines from the bootloader. |
18:42.24 | Annirak | And that version won't be in the debian installer image |
18:42.47 | sdm485 | Probably not. |
18:43.09 | Annirak | I wonder what would happen if I took the 2.6.22-3 out of the current di package and used its vmlinuz in the image |
18:43.14 | sdm485 | Can you change the kernel? |
18:43.18 | Annirak | yeah |
18:43.20 | Annirak | easily |
18:43.37 | Annirak | the only problem is that I may loose ethernet support if I do |
18:47.02 | sdm485 | If you need modules, you will run into a problem. |
18:47.11 | Annirak | I've been using the unofficial image with the IXP4xx |
18:50.12 | sdm485 | Have you thought of loading a slugos image (with the apex I sent) just to make absolutely sure the RAM is in order. I understand you aren't using a serial port. |
18:51.47 | Annirak | I had thought of it... I understand that slugosle works with debian... Hmm. That sounds like a viable option. |
18:51.48 | *** join/#nslu2-linux dwery (n=dwery@nslu2-linux/dwery) |
18:52.09 | Annirak | I can always reflash later |
18:52.31 | sdm485 | The apex version I sent you is from slugosbe. |
18:53.42 | Annirak | and yet it still booted. Weird. |
18:56.13 | sdm485 | Redboot is big-endian and apex is big-endian so the kernel must set the endian bit itself I guess. |
18:58.11 | Annirak | Well, that works. So there are conventional slugos images, right? |
18:58.34 | sdm485 | conventional? |
19:00.03 | Annirak | And by conventional, I mean binary |
19:00.32 | sdm485 | Ah. You can download them from the website. |
19:02.13 | sdm485 | Got to go. Have fun. |
19:02.26 | Annirak | =) |
19:24.48 | *** join/#nslu2-linux yvasilev (n=yvasilev@189.140.150.8) |
19:54.44 | *** join/#nslu2-linux Reedy_Boy (n=Reedy_bo@wikimedia/Reedy-Boy) |
20:13.58 | rwhitby | morning |
20:16.03 | Reedy_Boy | morning rwhitby |
20:16.34 | Reedy_Boy | having a VMWare PC that is on the same subnet as the rest of your network, would that have the same affect as me booting a ubuntu live cd? |
20:18.14 | rwhitby | Reedy_Boy: upgrade protocol does not work in vmware |
20:18.25 | rwhitby | (cause it's a non-standard, non-IP ethernet protocol) |
20:18.52 | Reedy_Boy | err.. |
20:19.09 | Reedy_Boy | If copy paste would work |
20:19.17 | Reedy_Boy | I've got both my NSLU2's in upgrade mode |
20:19.31 | Reedy_Boy | the debian VMWare PC has deteced the debian NSLU2 in upgrade mode |
20:19.52 | *** join/#nslu2-linux aadamson (n=adamson_@c-76-97-53-219.hsd1.ga.comcast.net) |
20:19.53 | Annirak | yo |
20:20.59 | Annirak | So, I've been having no end of trouble getting debian on to my slug |
20:21.02 | Annirak | Currently, I have Slugos/LE installed |
20:21.08 | Annirak | and that's working fine. |
20:21.39 | Annirak | But debootstrap is having some problems with me not having perl. |
20:22.31 | Reedy_Boy | rwhitby, http://images.reedyboy.net/VMWare.jpg |
20:24.08 | Annirak | So I'm installing perl. I think I'm getting further now, but it's hard to tell. I'll know for sure once debootstrap runs |
20:24.40 | Reedy_Boy | You can use a premade image... |
20:25.09 | Annirak | I haven't been able to get di-nslu2 to work |
20:25.19 | Annirak | I've tried all the workarounds I've seen. |
20:25.20 | rwhitby | Reedy_Boy: wow - what version of vmware is that? |
20:25.26 | aadamson | while we are talking about images... I've got slugos beta 4.8 image installed just fine and built the latest off the monotone tree (4.9) |
20:25.31 | Reedy_Boy | 6.0.2 |
20:25.40 | Reedy_Boy | rwhitby, i think, its the networking method thats used.. |
20:25.42 | rwhitby | must be fixed in 6 then - I only have 5.x |
20:25.47 | aadamson | when I installed it, the memory footprint was dramatically larger in 4.9 than 4.8 is that to be expected? |
20:26.04 | Reedy_Boy | rwhitby, ahhh.. I have it taking an IP via DHCP from my router |
20:26.27 | rwhitby | Reedy_Boy: yeah, I've tried bridged and nat before with 5.x - no go. |
20:26.27 | Reedy_Boy | so it acts like its a pc on my network at home... not like a subnet only my pc can access |
20:26.46 | Reedy_Boy | This is bridged :) |
20:26.57 | rwhitby | They've fixed usb2 in 6 too |
20:27.49 | Reedy_Boy | rwhitby, i think that proves that my unslung NSLU2 is at fault... |
20:28.55 | Reedy_Boy | Its already in the wiki... |
20:28.56 | Reedy_Boy | "Last solution, Upslug2 does work with VMware? Workstation. Note that the user must configure the VM to have access on the ethernet in Bridge mode. (Upslug2 requires layer 2 access on the network.) " |
20:30.46 | rwhitby | Reedy_Boy: note the version requirement for vmware |
20:31.12 | Reedy_Boy | Good idea, i'll put it on that it works in v6 |
20:31.27 | rwhitby | Reedy_Boy: one last test - unplug the slug that it found and see if it finds the other one |
20:31.39 | Reedy_Boy | it didnt find it first |
20:31.42 | Reedy_Boy | but yeah, sure |
20:31.56 | rwhitby | if not, then I agree the slug in question is not operating correctly |
20:32.15 | *** join/#nslu2-linux H4rvy (n=norm@62-30-143-108.cable.ubr05.pres.blueyonder.co.uk) |
20:32.15 | Reedy_Boy | [no NSLU2 machines..... |
20:33.40 | Reedy_Boy | That might become quite a popular method with there being the free VMWare player... for those in windows at least |
20:34.11 | Annirak | Anyone here have deboostrap experience? |
20:34.56 | Annirak | Nevermind, I'm being an idiot. |
20:35.18 | Reedy_Boy | I'd just use http://www.slug-firmware.net/d-dls.php |
20:35.19 | Reedy_Boy | tbh |
20:35.24 | Annirak | I omitted the http:// from the address. |
20:35.32 | Annirak | Reedy_Boy: didn't work for me |
20:36.05 | rwhitby | Reedy_Boy: windows users can just use the sercomm utility |
20:36.24 | Reedy_Boy | yeah, just it can be a bit arkward at times |
20:36.26 | Annirak | The installer crashes continuously. That's supposed to be a symptom of low memory. So I fattened my slug. Except that now I need a build environment to get a kernel that supports a fatslug. |
20:36.56 | Reedy_Boy | During the drive partition/format |
20:37.11 | Annirak | Yes. |
20:37.15 | rwhitby | Annirak: the crashes are a symptom of large drives and mkfs.ext2 needing lots of memory to format large drives |
20:37.43 | rwhitby | I believe if you get in and add swap before that step (by manually partitioning the disk on another machine first) then you can get it to work |
20:37.44 | Annirak | rwhitby: I'm using reiser, with keep existing data specified. |
20:38.15 | Annirak | rwhitby: I have turned swap on before that. When partman runs, it unmounts your swap. |
20:38.31 | Reedy_Boy | That you can, or at least, was Martins work around on the debian-arm mailing list |
20:38.53 | Annirak | The workaround is to manually enable swap, then partition manually, then let partman run, but tell it to keep existing data and not format. |
20:39.02 | scant | rwhitby: are you here? |
20:39.17 | Annirak | At the end of that sequence, my install hangs indefinitely, and eventually drops my connection. |
20:39.33 | rwhitby | scant: yep |
20:39.41 | rwhitby | Annirak: just put a serial port on it. |
20:40.01 | Annirak | heh |
20:40.04 | scant | rwhitby: i'm asking for help in #oe, could you look? |
20:40.16 | Annirak | rwhitby: how's that going to help? |
20:41.52 | rwhitby | Annirak: it may be only the ssh connection that drops, at which point you can complete the installation using the serial console |
20:55.01 | Annirak | rwhitby: That may be. So far, the slugos/le based debian install has been far less annoying, if far more involved. |
20:55.37 | Annirak | (annoyance happens when my slug kicks me out) |
20:56.30 | rwhitby | Annirak: yep, that's how I used to do my debian nslu2 installs. |
20:57.48 | Annirak | rwhitby: once I have my slugos/le+debian install done, I'm going to try to work out how to get a kernel with a larger memory support in there. |
20:58.36 | Annirak | I understand it's purely a matter of the kernel arguments, since I have a modified apex loader installed already |
21:08.20 | rwhitby | yep |
21:10.50 | Annirak | I think I have a problem. I just tried to unzip a file and got "tar: invalid gzip magic" |
21:11.38 | Reedy_Boy | missing dependancies? |
21:11.39 | *** join/#nslu2-linux bpmurray (n=chatzill@user12.monasterboice.net1.ie) |
21:11.50 | *** part/#nslu2-linux bpmurray (n=chatzill@user12.monasterboice.net1.ie) |
21:11.56 | Annirak | I don't think so. It had worked earlier |
21:12.17 | Reedy_Boy | lol |
21:13.04 | Annirak | I'm going to see what it does with a reboot |
21:15.31 | Annirak | Here's a new one for me: mount: Cannot allocate memory |
21:17.03 | Annirak | MemFree: 18704 kB |
21:23.52 | *** join/#nslu2-linux kolla (n=kolla@drift1.uninett.no) |
21:24.17 | Reedy_Boy | rwhitby, was looking into some filesystem stuff last night... They reccommeneded XFS/ReiserFS over ext3... |
21:24.57 | Reedy_Boy | A bit of googling showed Martin Michlmayr saying no to XFS - some debian kernel problem |
21:25.07 | Reedy_Boy | but JFS would be ok instead... |
21:25.31 | Reedy_Boy | Would a different FS make any difference |
21:26.08 | Reedy_Boy | i was looking into 25GB of missing space on my drive - the 5% reserved space... which was 25GB |
21:26.16 | Reedy_Boy | i've taken that downt o 1% = 5GB... |
21:40.47 | Annirak | I think when I installed perl and reiserfs on my slugos/le it corrupted the onboard flash. |
22:16.28 | *** part/#nslu2-linux Annirak (n=bmoran@h2-cpe102.insinc.com) |