**** BEGIN LOGGING AT Sun Mar 04 02:59:58 2007 Mar 04 03:28:13 pavlov * r6499 /packages/net/wol/ (. Makefile): wol contributed by mpapamanz@gmail.com -- ticket: 1265 Mar 04 03:40:00 nbd: I love your pkg metadata/info caching, makes everything so much quicker :-) Mar 04 03:40:26 you must symlink everything.. :) Mar 04 03:41:06 i'm about to spam a huge commit to packages Mar 04 03:41:23 propsets Mar 04 03:41:30 pavlov * r6500 /packages/ (482 files in 345 dirs): huge propset, fixes eol-style on everything and sets keywords and copyright on all Makefiles Mar 04 03:41:37 lst: yes, I have all packages symlinked in :-) Mar 04 03:41:46 o.O Mar 04 03:41:56 * Bartman007 cancels his building to update again. Mar 04 03:42:21 it's just svn properties, not code changes Mar 04 03:42:40 fixes id expansion and such Mar 04 03:42:46 so it is, but r6500 just looks a lot cleaner too :-) Mar 04 03:43:13 yeah, props to me for getting a fun commit revision for that propset ;) Mar 04 04:02:21 pavlov * r6501 /trunk/package/ (376 files in 173 dirs): same propset here as changeset:6500 Mar 04 04:17:32 pavlov * r6502 /trunk/ (386 files in 119 dirs): more propsets... i'm done for now. **** ENDING LOGGING AT Sun Mar 04 09:44:25 2007 **** BEGIN LOGGING AT Sun Mar 04 09:44:57 2007 Mar 04 13:22:07 nbd * r6503 /trunk/scripts/ (7 files in 2 dirs): force a config reset when the user changes the target platform in menuconfig Mar 04 13:40:09 hmm, mount_root now mounts sysfs twice Mar 04 13:40:56 * nbd checks... Mar 04 13:41:31 is mount_root called twice? Mar 04 13:41:46 i can only find one place where it mounts sysfs Mar 04 13:42:18 11th and the last line Mar 04 13:44:11 huh? which file are you talking about? Mar 04 13:44:24 oh, sorryu Mar 04 13:44:28 i was looking at an old copy Mar 04 13:44:29 package/base-files/files/sbin/mount_root Mar 04 13:45:23 nbd * r6504 /trunk/package/base-files/files/sbin/mount_root: only mount sysfs once (thx, ejka) Mar 04 13:45:25 :) Mar 04 13:54:04 nbd * r6505 /trunk/package/busybox/patches/450-truncated_ping_results.patch: hopefully fix ping bugs introduced in [6443] Mar 04 15:20:11 nbd * r6506 /trunk/scripts/ (5 files in 2 dirs): disable the automatic config reset if 'Advanced configuration options' is selected Mar 04 15:44:41 nbd * r6507 /trunk/include/unpack.mk: add a crlf workaround similar to the one from freewrt Mar 04 16:36:12 h3sp4wn: that mtd weirdness seems to be there, still.. we'll try to track it down Mar 04 16:36:39 h3sp4wn: can you check if both minipci cards are detected? because i would commit a fix for it then Mar 04 16:37:07 Kaloz: I can if you can give me the command to do a netboot Mar 04 16:37:57 there is no netboot image generated in trunk/bin - so I presume I take vmlinux from the linux dir ? Mar 04 16:38:40 h3sp4wn: "lo -r -b %{FREEMEMLO} openwrt-gateway7001-2.6-zImage" and "exec" if you compile a ramdisk image Mar 04 16:38:51 h3sp4wn: let me create an image for you Mar 04 16:40:43 I thought it would be named something else but the file size indicates its right Mar 04 16:46:34 wifi0: Atheros 5212: mem=0x48000000, irq=28 Mar 04 16:46:37 wifi1: Atheros 5212: mem=0x48010000, irq=0 Mar 04 16:46:54 looks to be still the same unless you have something uncommited Mar 04 16:47:20 hcg * r6508 /trunk/target/linux/at91-2.6/ (9 files in 5 dirs): Added initial support for at91 plaform (HCG) Mar 04 16:47:34 i've said let me create an image because i'm not sure the fix is right, so it's not yet commited Mar 04 16:47:37 :) Mar 04 16:47:46 np Mar 04 16:52:56 h3sp4wn: http://kaloz.rulez.org/files/openwrt-gateway7001-2.6-zImage Mar 04 16:54:58 Kaloz: I will test it right now Mar 04 17:02:35 Kaloz: http://paste.debian.net/23057 (halts at uncompressing linux - tried twice) Mar 04 17:03:16 h3sp4wn: wait at least 2 minutes Mar 04 17:03:27 h3sp4wn: mdev seems to go crazy and starts the OOM killer Mar 04 17:03:41 h3sp4wn: probably we will change it to hotplug2 Mar 04 17:06:27 Kaloz: another 3 minute nothing happened (the initramfs I built from trunk did boot though - after a shorter wait) Mar 04 17:06:52 yeah, it's peretty stupid bug, and pretty random Mar 04 17:06:56 let me make the kernel smaller Mar 04 18:04:10 h3sp4wn: *sigh* maybe it would have been faster to get hotplug2 working.. trying to make a dman image that boots most of the time at least Mar 04 18:13:06 Kaloz: Its strange it seems like good quality hardware should be easier to get to work Mar 04 18:14:08 h3sp4wn: nah, currently the problem is with the damn mdev Mar 04 18:14:28 h3sp4wn: on initramfs it seems to fsck up otplug and eats all the memory Mar 04 18:14:32 hotplug* Mar 04 18:14:46 Is this with devfs removed also ? Mar 04 18:14:56 yeah, this is without devfs Mar 04 18:15:09 h3sp4wn: but i'm sick of it now, so i'm creating a devfs based image for you to check Mar 04 18:15:18 backported the patch i've made Mar 04 18:15:39 ok thanks Mar 04 18:47:03 h3sp4wn: same location Mar 04 18:47:34 h3sp4wn: this is older code, but if i'm right, both cards should be working (please verify with an iwlist scan if possible) Mar 04 19:07:18 nbd * r6509 /trunk/package/udev/Makefile: upgrade udev to latest version and add separate package for udevtrigger (will be used by hotplug2) Mar 04 19:12:38 Kaloz: Ok I will check (just had to eat) Mar 04 19:12:53 h3sp4wn: np :) Mar 04 19:13:14 nbd * r6510 /trunk/package/hotplug2/ (. Makefile): add hotplug2 Mar 04 19:15:26 * Bartman007 drools over the quad-core machine he now has access to. Mar 04 19:20:48 h3sp4wn: btw, any idea when could that breakage come up? Mar 04 19:23:02 Kaloz: As in which svn revision ? somewhere between 2.6.17 and 2.6.19.2 Mar 04 19:23:58 that doesn't help much :) anyways, let me know if both minipcis are working or not :) theoritically the first will have irq 28 the second irq 27 Mar 04 19:24:54 Kaloz: It works so far as the irq's are correct Mar 04 19:25:27 and there is no complaints about it but I cannot actually see anything using wifi1 Mar 04 19:25:52 (no scan results - with exactly the same method as getting results with wifi0 Mar 04 19:26:05 hmz, can you destroy both ath0 and ath1 and recreate them as stations? Mar 04 19:26:58 That is already what I did Mar 04 19:27:35 then leave ath0 (the b/g one) as master, and scan with the other one for it :) Mar 04 19:28:08 wait for the crash... Mar 04 19:28:22 or are the atheros drivers more stable now days? Mar 04 19:29:41 http://paste.debian.net/23077 - I will try that Mar 04 19:30:57 h3sp4wn: iwlist ath1 freq? Mar 04 19:32:09 Frequency:5.18 GHz - I must need to set it to the right frequency Mar 04 19:32:48 http://paste.debian.net/23078 Mar 04 19:32:59 nbd * r6511 /trunk/package/hotplug2/Makefile: make hotplug2 depend on udevtrigger Mar 04 19:33:33 ah, it's programmed for 5ghz only Mar 04 19:33:42 do you have any 802.11a gear? Mar 04 19:34:02 Not other than this no Mar 04 19:34:40 Bartman007: didn't you have the same box? Mar 04 19:35:13 Kaloz: the 7001 a/b/g ? yeah. Mar 04 19:35:35 and you have some other 802.11a gear, right? Mar 04 19:35:45 oh, sure you do, those dlinks are still there i guess :) Mar 04 19:35:47 that I don't think I can get an antenna capable of picking up h3sp4wn's signal :-P Mar 04 19:35:53 could you check that initramfs image? Mar 04 19:36:22 Kaloz: yes, though I've had your two all packaged up sitting in a box in my trunk for over a month.... Mar 04 19:36:41 :) Mar 04 19:36:47 any other 802.11a gear? Mar 04 19:36:50 sure, I'll do that now Mar 04 19:37:15 yeah, the third DWL-7000AP Mar 04 19:37:51 * Bartman007 looks for one of his 5V 2.5A bricks Mar 04 19:40:52 :) Mar 04 19:41:42 I think I should switch everything to the a band though (there is no other networks using it around here) Mar 04 19:49:16 here it comes ... Mar 04 19:49:22 nbd * r6512 /trunk/ (19 files in 13 dirs): nuke mdev and replace it with hotplug2 :) Mar 04 19:49:27 :P Mar 04 19:50:45 hehe Mar 04 19:51:11 ... mdev got nuked already? Mar 04 19:51:14 yeah Mar 04 19:51:27 it was too inflexible Mar 04 19:51:27 * Bartman007 stops the build on his new machine and updates Mar 04 19:53:27 Bartman007: did you find a psu? :) Mar 04 19:53:33 Kaloz: yeah Mar 04 19:53:45 then I lost the first one momentarily :-0 Mar 04 19:54:22 :)) Mar 04 19:57:12 wow, my 7001 has r4779 on it.... Mar 04 19:57:23 nbd * r6513 /trunk/Makefile: add some missing deps for the metadata files Mar 04 19:57:36 :) Mar 04 19:57:58 Bartman007: just netboot the image i've put online, and the a radio works as well, i'll commit the fix Mar 04 20:01:07 nbd * r6514 /trunk/target/linux/ (12 files in 2 dirs): reorder target profiles Mar 04 20:24:57 nbd * r6515 /trunk/ (Makefile include/scan.mk): fix metadata scan for profile split Mar 04 20:27:42 it can see the 11a radio, but I feel like an idiot because I can't get either card to see any available networks Mar 04 20:28:58 use the dlink as the ap and the gateway's radio as a client Mar 04 20:29:13 and if it's set to open network, probably you can associate Mar 04 20:30:14 I can set the essid to an open network, but it doesn't associate with either card Mar 04 20:31:01 iwlist ath1 scan Mar 04 20:31:02 ath1 Interface doesn't support scanning : Invalid argument Mar 04 20:31:51 I've destroyed both ath devices and created them with wlanmode sta Mar 04 20:31:59 ifconfig ath1 up Mar 04 20:31:59 :) Mar 04 20:32:28 oy. Mar 04 20:33:05 http://pastebin.ca/381432 Mar 04 20:33:08 :-) Mar 04 20:33:48 yay :) Mar 04 20:34:44 I'm in heaven building on this quad machine :-) Mar 04 20:34:44 kaloz * r6516 /trunk/target/linux/ixp4xx-2.6/patches/200-gateway_7001.patch: Mar 04 20:34:57 fuck, forgot the commit messgae Mar 04 20:34:58 :P Mar 04 20:35:01 hehe. Mar 04 20:35:08 whatever :) Mar 04 20:36:11 Bartman007: btw, ethernet is working as well if upload the microcode right Mar 04 20:36:24 Bartman007: still have to make some crap to extract the mac addresses Mar 04 20:38:12 Kaloz: you can probably use awk for that Mar 04 20:38:44 or do you want me to hack up an awk script to do that? Mar 04 20:38:52 nbd: nah, i want to make it somehow like it's done for the nslu and the nas100d Mar 04 20:39:01 how do they do it? Mar 04 20:39:26 https://dev.openwrt.org/browser/trunk/target/linux/ixp4xx-2.6/patches/152-nas100d_mtd_load_mac.patch Mar 04 20:39:37 https://dev.openwrt.org/browser/trunk/target/linux/ixp4xx-2.6/patches/153-nslu2_mtd_load_mac.patch Mar 04 20:40:08 but if you want to do it, i can upload that mtd partition dump that hold the data :) Mar 04 20:40:32 shouldn't be too hard Mar 04 20:40:37 gimme the dump :) Mar 04 20:40:54 thanks :) I'm still doing the other stuff, so I cna use any help :)) Mar 04 20:40:57 just a sec Mar 04 20:42:06 http://kaloz.rulez.org/files/gw7k1-mtdblock3-default.bin Mar 04 20:42:36 nbd: the name of the partition is "default", and probably it's enough to extract the first mac and use +1 for the second Mar 04 20:42:54 nbd * r6517 /trunk/target/linux/x86-2.6/Makefile: enable ne2k-pci by default Mar 04 20:43:13 ok Mar 04 20:43:33 nbd: btw, I had an idea - what if we create Qemu target images (depending on devleoper stuff or so)? Mar 04 20:43:42 why? Mar 04 20:43:42 nbd: i guess quite some people could be interested in that Mar 04 20:43:48 the regular images run in qemu just fine Mar 04 20:43:54 i just enabled ne2k-pci by default Mar 04 20:43:57 so networking runs as well Mar 04 20:44:10 yeah, but i mena quemu sparc, qemu arm10 and so Mar 04 20:44:15 hmm Mar 04 20:44:17 no idae Mar 04 20:44:32 so actually it cna be tested on stuff we don't have yet :) Mar 04 20:51:44 Kaloz: http://paste.biz/paste-829.html Mar 04 20:52:07 Kaloz: this prints both macs Mar 04 20:52:18 nbd: fine, but we need to do that in kernel space :) Mar 04 20:52:28 the mac could come up as 0.0.0.0 Mar 04 20:52:37 i mean 0:0:0:0:0:0 Mar 04 20:52:44 and an init script can initialize it properly Mar 04 20:53:24 i know, but that's not too proper :) i had something like that laready (okay, it wa sugly), but want to make it something like the two patches i've showed Mar 04 20:53:25 just do . /etc/functions.sh; set $(sh mac.sh $(find_mtd_part default)) from a script Mar 04 20:53:32 and $1 will be mac1 and $2 will be mac2 Mar 04 20:53:47 i think doing it from user space is better Mar 04 20:53:51 because we don't rely on magic offsets Mar 04 20:54:14 i hate magic offsets in kernel space even more than in user space Mar 04 20:54:56 what if you want nfsroot? :) Mar 04 20:55:17 use randomized macs for the early init Mar 04 20:55:27 ugly :)) Mar 04 20:55:35 btw, trying to find the damn thing that broke jffs2 on xscale Mar 04 20:55:38 but better than magic offsets for reading the mac from redboot Mar 04 20:55:40 it's really annoying me Mar 04 20:55:53 it's a pain in the ass to find a new offset for every new target Mar 04 20:55:58 and if they change the boot loader, it breaks Mar 04 20:56:00 that sucks Mar 04 20:56:20 not too much people are changing the bootloader :) Mar 04 20:56:32 new revisions of the hardware might change it Mar 04 20:57:27 and nfsroot will need a minimal userspace anyway Mar 04 20:57:30 because of firmware Mar 04 20:57:30 imho Mar 04 20:57:43 unless it's thrown into its own mtd partitions Mar 04 20:59:53 nope, you tftpboot a kernel and use nfs as root, that doesn't need any userspace thingie Mar 04 21:00:55 i mean if you don't have firmware for the ethernet.. Mar 04 21:33:58 Hi all, Mar 04 21:33:58 I just noticed that Sveasoft operates open-wrt.com, open-wrt.net, and open-wrt.org. Does that mean that Sveasoft is working with OpenWRT again? Mar 04 21:34:02 Thanks! Mar 04 21:34:04 HAHAHA Mar 04 21:34:05 hello Mar 04 21:34:06 P.S. The same thing goes for free-wrt.com and willybox.com. Are these new releases of Sveasoft software as well? Mar 04 21:34:47 wtf? that can only be a joke right? Mar 04 21:35:14 nah, he probably registered those domain for real Mar 04 21:37:35 he forgot 0penwrt.org Mar 04 21:37:50 yeah, and i-have-no-life.org Mar 04 21:37:51 :9 Mar 04 21:38:00 no, that one is mine Mar 04 21:38:01 :D Mar 04 21:38:21 :P Mar 04 21:39:49 http://sveasoft.blogspot.com/ Mar 04 21:39:57 this one makes me laugh even more Mar 04 21:40:00 :)) Mar 04 21:40:38 "P.S. Next time I call .sa, I'll tell her that you've been touching Lucas and Ryan in naughty places. But that idle threat pales in comparison to what you've done to hundreds of former customers." :DDD Mar 04 21:40:59 no idea who's that guy is, but he definitely annoys James :D Mar 04 21:41:54 I think I miss some background Mar 04 21:43:00 this james guy == sveasoft? Mar 04 21:43:07 i think so Mar 04 21:43:17 yep Mar 04 21:43:35 and quite some people hate him if you take a look at that blog Mar 04 21:44:22 well, I heard there is some reason to Mar 04 21:44:33 I'm also thinking I'm being polite here Mar 04 21:44:38 hehe Mar 04 21:45:25 no worries, I'm not james...I think.... Mar 04 21:46:09 but after the palinka I never can tell Mar 04 22:30:53 http://sveasoft.blogspot.com/2006_03_01_archive.html Mar 04 22:31:12 i hope he registred them for 10 years at least Mar 04 22:31:26 why? Mar 04 22:31:56 apart from the fact that it costs money Mar 04 22:33:51 just for that, to make him cost more money Mar 04 22:34:12 hell, maybe I visit Sweden just to go there and make fun of him once Mar 04 22:34:43 I wouldn't do that Mar 04 22:34:46 booze in .se is expensive Mar 04 22:35:22 if i would like to put that moron on fire, palinka could be exported Mar 04 22:35:23 :P Mar 04 22:51:31 lol, think he can't hold the domains any longer if someone goes to the court and points to his blog :D poor reaction..... Mar 04 23:27:29 nbd * r6518 /trunk/target/linux/rb532-2.6/patches/120-cf.patch: fix a compile error caused by the devfs removal Mar 04 23:47:45 nbd * r6519 /trunk/target/linux/brcm-2.4/base-files/default/ (etc/preinit sbin/mount_root): more hotplug2/mdev/devfs related fixups, this time for broadcom Mar 04 23:48:20 florian * r6520 /packages/net/freeradius/Makefile: Fix compilation when modules are selected or not, should fix buildbot spam :p Mar 05 00:10:14 florian * r6521 /packages/net/freeradius/Makefile: libradius is expected to be in /usr/lib/ Mar 05 00:14:08 nbd * r6522 /trunk/target/linux/ixp4xx-2.6/ (Makefile config/default image/Makefile): add squashfs support for ixp4xx Mar 05 00:31:14 nbd * r6523 /trunk/target/linux/x86-2.6/image/Makefile: add missing init=/etc/preinit cmdline parameter Mar 05 00:39:51 kaloz * r6524 /trunk/target/linux/ixp4xx-2.6/config/default: fix squashfs for ixp4xx Mar 05 00:40:13 h3sp4wn, Bartman007: squashfs is working now on the gigabyte :) Mar 05 00:44:56 pavlov * r6525 /packages/lang/php5/patches/01-configure_cross.patch: fixes compile for ldap module in php5 Mar 05 00:48:27 gigabyte? (not gateway?) Mar 05 00:48:39 gateway, right Mar 05 00:48:42 i'm tired Mar 05 00:48:43 :) Mar 05 00:48:51 yay! Mar 05 00:49:27 hmmm, I haven't checked if ethernet works in quite a while (since r4779) :-) Mar 05 00:49:38 guess I'll do that once I confirm that my new buildbot works. Mar 05 00:50:24 s/bot/box/ Mar 05 00:50:40 if you upload the firmware and set mac addresses, they do Mar 05 00:50:56 i'll try to figure a nice way to do it later Mar 05 00:51:15 was observing massive breakage when I told buildroot-ng to use 4 concurrent jobs. Mar 05 00:51:59 Kaloz: oh really? cool, I think I'll have to make my 7001 a functional device finally. I assume mac addys are defined the typical linux way? Mar 05 00:53:20 well, you have to manually do a "ifconfig ethX hw ether 11:22:33:44:55:66" for both eth0 and eth1 Mar 05 00:53:32 the mac addresses are stored in the flash partition named "default" Mar 05 00:54:13 and for the microcode, for now the easiest thing is if you extract them (look into the kernel sources, Documentation/networking/ixp4xx or smg like that) Mar 05 00:54:40 and create a directory named "files" in the top directory, and place them there , so they will be copied into the firmware Mar 05 00:55:29 but i'll try to make some userfriendly way for these tomorrow Mar 05 00:56:34 Bartman007: i think i know what will i do :) Mar 05 00:59:53 Kaloz: yeah, ifconfig is what I meant by the typical linux way :-) Mar 05 01:00:22 Bartman007: or just use /etc/config/network for that Mar 05 01:00:24 hmz Mar 05 01:00:57 nbd: can't we put that small script into netconfig somehow? probably it could check if the cpuinfo says it's a gateway or so Mar 05 01:02:26 sure Mar 05 01:02:51 damn, I wish I could get my hands on another 7001 for $50... Mar 05 01:03:09 then it has to be run only once :) can you integrate it? Mar 05 01:04:03 (for the microcode, i think i'll use the unallocated space on the gateway to create an additional partition for it..) Mar 05 01:04:42 that way it can be loaded as easily without hotplug and will still enable anyone to flahs back to the factory stuff if he/she wants Mar 05 01:05:01 (aslo, it will be a one-time "problem") Mar 05 01:28:37 Kaloz: are you aware of how we have handled the IXP microcode on the ixp4xx devices? Is this a similar problem? Mar 05 01:29:30 (we put it in spare space in an existing block in the flash, and then our kernel searches for a magic string in that block and loads it directly from there. no userspace intervention is required. Mar 05 01:31:44 pavlov * r6527 /trunk/package/ (40 files in 9 dirs): trunk.. same deal as changeset:6526 Mar 05 01:36:02 6526 was so much bigger :-p Mar 05 01:36:08 * lst punts the bot Mar 05 01:37:29 lst: heh, yeah, I had to look up revision up on trac because the bot coughed Mar 05 01:52:55 lst: did you script those renames or do it by hand? Mar 05 01:54:04 re **** ENDING LOGGING AT Mon Mar 05 02:59:57 2007