**** BEGIN LOGGING AT Tue Oct 27 02:59:56 2009 Oct 27 08:36:17 sn9: ? Oct 27 08:36:28 sn9: its orking ootb with supported devices Oct 27 08:43:08 blogic: what is? Oct 27 08:43:28 try a full sentence then i know what you mean :) Oct 27 08:43:35 the 3g thing Oct 27 08:43:44 21:31 < sn9> blogic: he said you have not had extensive testing of the 3g firmware enhancements yet Oct 27 08:43:56 in the fonera firmware Oct 27 08:44:25 yes Oct 27 08:44:57 are there supported devices for the fonera firmware other than la fonera? Oct 27 08:45:23 ah no Oct 27 08:45:26 not yet Oct 27 08:45:32 what are you after ? Oct 27 08:45:36 the umts feature ? Oct 27 08:45:43 it is handled by umtsd Oct 27 08:45:48 which is a lua/luci app Oct 27 08:45:52 hmm Oct 27 08:45:59 hmmm, maybe I was unclear. I meant to indicate that he has it working quite well in the fonera firmware but has not had time to commit it to openwrt yet Oct 27 08:46:00 i guess we can merge it in owrt Oct 27 08:46:07 yes Oct 27 08:46:11 i can do that Oct 27 08:46:24 i have been busy last week working on making owrt boot on fon20n Oct 27 08:46:31 next will be merging some of the luci workflow Oct 27 08:46:38 luci-owrt is a huge uci editor Oct 27 08:46:46 luci-fon is a webui with workflow Oct 27 08:46:56 Bartman007: near as i can figure, 2.0g never arrived on these shores Oct 27 08:47:16 sn9: fon20n is avail in us though Oct 27 08:47:25 sn9: i still have a few 50% discount voucher :) Oct 27 08:47:30 2 are left Oct 27 08:47:49 how quickly are they shipped? Oct 27 08:48:06 3-4 days in europe Oct 27 08:48:06 sn9: 150Mbit shipping. Oct 27 08:48:13 no idea how long shipment to us takes Oct 27 08:48:24 but as far as i am aware, fon has a warehouse in the us Oct 27 08:48:30 i might have ordered the mzk-w04nu, but that is 5-9 days shipping Oct 27 08:49:35 my ticket shows a departure the morning of the 29th, and i might actually use it Oct 27 08:49:56 blogic: hmmm, is the fon20 (g) no longer available? I don't see it in the fon store. Oct 27 08:50:03 phew Oct 27 08:50:05 no idea Oct 27 08:50:08 i only hack for fon Oct 27 08:50:16 i keep well out of the business crap Oct 27 08:50:17 :) Oct 27 08:50:19 sorry Oct 27 08:50:37 Bartman007: if you really need a 20g let me know, if they are no longer sold i am sure i can still get some Oct 27 08:51:17 they are no longer sold, but when they were, they were europe-only Oct 27 08:51:32 blogic: I have no need, just trying to figure out ways to help sn9 Oct 27 08:52:33 of course, the cheapest thing i could do is use a device i already have Oct 27 08:53:04 the only device i have with usb that i can think of is the wrt350n Oct 27 08:53:21 actually, no Oct 27 08:53:32 the anas350 has usb also Oct 27 08:53:42 use comgt Oct 27 08:53:47 it is already in svn Oct 27 08:53:51 yes Oct 27 08:53:57 sn9: well, I do have an mzk-w04nu :) Oct 27 08:54:21 and a fon20, but it's packed up, I'd have to find it. Oct 27 08:54:26 Bartman007: there was a report a few hours ago that mini_fo is b0rked on it Oct 27 08:54:38 on the planex Oct 27 08:55:10 the wrt350n can, in theory, use 2.4 or 2.6 Oct 27 08:55:23 2.4 does not like the evdo dongle Oct 27 08:55:31 2.6 does not boot, at all Oct 27 08:59:52 do either of you have any spare broadcom device with ready serial where you can try this build? http://downloads.openwrt.org/snapshots/trunk/brcm47xx/ Oct 27 09:00:43 i have verified that that does not boot, either Oct 27 09:00:47 sn9: I need to finish some classwork first. Oct 27 09:02:52 sn9: r17465 runs fine here (WGT634U) Oct 27 09:03:09 loswillios: try a newer rev? Oct 27 09:03:58 17465 was well before 2.6.30 Oct 27 09:04:27 and yes, i did try 2.6.28; no diff Oct 27 09:05:03 (after i nuked a newer patch in generic-2.6, that is) Oct 27 09:05:19 @sn9: you can give a try this one: http://wifi.ozo.com/airo/openwrt/firmware/kamikaze/2.6/17585/ Oct 27 09:05:35 haven't tried yet Oct 27 09:05:51 acoul: still well before 2.6.30 Oct 27 09:06:58 @sn9: actually it's 2.6.31 Oct 27 09:07:20 hmm; ok. have you tried it yourself? Oct 27 09:07:46 let me check Oct 27 09:07:53 hi sn9, hi loswillios Oct 27 09:07:54 I forgot. my wgt634u is currently running r18115. Oct 27 09:08:04 Linux OpenWrt 2.6.30.8 #1 Thu Oct 22 07:48:07 UTC 2009 mips GNU/Linux Oct 27 09:08:37 hi Bartman007 Oct 27 09:09:14 danage: hey Oct 27 09:09:24 Bartman007: so, something between 18115 and now broke it Oct 27 09:09:26 @sn9: last working one (from my repository) I checked is this: http://openwrt.pastebin.com/d340e2f92 Oct 27 09:10:40 sn9: I'll try upgrading it, see if it explodes. Oct 27 09:11:21 acoul: http://wifi.ozo.com/airo/openwrt/firmware/kamikaze/2.6/17585/openwrt-wrt350n_v1-jffs2.bin does not even start the kernel Oct 27 09:11:56 @Bartman007: gcc-4.4.2/linux-2.6.31.5 should provide a nice explosion ;-) Oct 27 09:12:24 i am still suspecting some sort of loader borkage Oct 27 09:13:04 @sn9: I am sorry my test devices are: Broadcom BCM4710 V0.0 Oct 27 09:13:58 acoul: i will try your r17324 Oct 27 09:14:10 sn9: root@OpenWrt:/# uname -a Oct 27 09:14:10 Linux OpenWrt 2.6.30.9 #1 Mon Oct 26 11:51:33 CDT 2009 mips GNU/Linux Oct 27 09:14:21 latest snapshot works on the wgt634u Oct 27 09:15:17 Bartman007: the one from here? http://downloads.openwrt.org/snapshots/trunk/brcm47xx/ Oct 27 09:15:39 ack Oct 27 09:16:05 then there is something specific to the wrt350n Oct 27 09:16:28 @Bartman007: if you want to see nice smoke colors from that explosion you can always try this: https://dev.openwrt.org/ticket/5954 Oct 27 09:16:34 the wgt is bcm5350, (bcm3302 v0.7 core) IIRC. Oct 27 09:16:51 something like that Oct 27 09:18:47 this would suggest that the problem might be in platform support, rather than the loader Oct 27 09:19:30 @sn9: it may simply be a bad/wrong network script ... Oct 27 09:19:41 acoul: it never gets that far Oct 27 09:19:51 if your device has a wan/eth1 port Oct 27 09:19:56 acoul: it never gets that far Oct 27 09:20:14 you can try to see if it gets an IP from a dhcp server Oct 27 09:20:31 no: http://openwrt.pastebin.com/f436fad58 Oct 27 09:20:32 do you have a serial console Oct 27 09:20:49 see paste Oct 27 09:22:25 can you also paste o good boot? Oct 27 09:22:38 good boot of 2.6? Oct 27 09:22:42 yes Oct 27 09:23:18 i'd have to dig up a 2.6 image that actually boots, and i don't remember where i last had one Oct 27 09:24:20 acoul: with your image, the last line before autoreboot, btw, is: Oct 27 09:24:22 Starting program at 0x80001000 Oct 27 09:31:00 @sn9: so when you boot a latest trunk you just get reboots instead of a prompt? Oct 27 09:31:19 i get what you saw in the paste Oct 27 09:33:28 @sn9: this is a nice toy to play, can you compile your own custom trunk and start testing various settings? Oct 27 09:33:51 i did; nothing changed Oct 27 09:35:31 did you try gcc-4.4.2? Oct 27 09:37:30 yes Oct 27 09:38:15 no difference Oct 27 09:39:43 @sn9: what is the actual error/issue the paste is not that clear Oct 27 09:42:20 acoul: because it's unknown Oct 27 09:44:51 why are there packages that are on the trunk repository and not all in packages repository? Oct 27 09:49:37 nunojpg: if it's required to boot a device, or part of the "core" package set, the package sits in trunk. otherwise it resides in packages Oct 27 11:29:48 {Nico}: we have some problem with mod_curl Oct 27 11:30:01 we are trying to compile openwrt for x86 Oct 27 11:33:24 {Nico}: did the freeswitch package you made work also with openwrt for x86? Oct 27 11:35:10 oh, {Nico} did that? Oct 27 11:38:53 what do you mean? Oct 27 11:39:10 it does not compile mod_curl Oct 27 11:39:40 anyway i'll start testing freeswitch-minimal Oct 27 11:40:40 <_trine> hello Lalloso Oct 27 11:41:20 hi john Oct 27 11:41:32 i'm trying the nice packages for freeswitch openwrt Oct 27 11:41:38 go on and try them yourself! Oct 27 11:41:46 {Nico} has done a great work Oct 27 11:42:00 i think it's only missing a freeswitch core package Oct 27 11:42:11 <_trine> I tried FS on my router but there was a problem with deps in relation to it installing Oct 27 11:42:33 <_trine> but that was over a week ago Oct 27 11:42:43 <_trine> it might have changed again Oct 27 11:46:49 what kind of problem? we must help nico with the testing Oct 27 11:48:58 <_trine> I was discussing it with him but he seemed to think it was a problem of mixing and using different sources but as far as I know I didn't do that then he stopped responding maybe he got called away, but the conversation ended at that point Oct 27 11:59:11 Bartman007: # cat /proc/cpuinfo Oct 27 11:59:11 cpu model : Broadcom BCM3302 V0.7 Oct 27 11:59:44 russell_: what device? Oct 27 12:00:41 wgt634u Oct 27 12:00:54 hmm Oct 27 12:01:19 then what is it about the wrt350n that causes it not to work? Oct 27 12:01:40 nfi, i'm still back on r16538 ... embarrassing! Oct 27 12:04:39 btw, sn9, i still have that airlink 300n Oct 27 12:04:58 russell_: as i just mentioned to Kaloz a bit ago Oct 27 12:05:25 have you gotten jtag going on any device since? Oct 27 12:05:49 i have a jtag interface thingie, but haven't tried it on that thing Oct 27 12:06:02 has it worked on other devices? Oct 27 12:06:05 i've been playing with a peplink board, arm922t-based Oct 27 12:06:07 yeah Oct 27 12:06:24 does it have separate srst and trst? Oct 27 12:06:39 the peplink has those tied together Oct 27 12:06:47 the jtag interface Oct 27 12:06:54 i think so Oct 27 12:07:08 which adapter is it? Oct 27 12:07:19 it's the olimex arm-usb-ocd Oct 27 12:07:57 i've forgotten if the 300n is arm-based or something else Oct 27 12:09:16 bbiab Oct 27 12:09:37 kaloz * r18180 /trunk/target/linux/generic-2.6/ (config-2.6.30 config-2.6.31): add missing kconfig symbol Oct 27 12:12:50 kaloz * r18181 /trunk/target/linux/generic-2.6/ (6 files): sort kconfig symbols in generic configs Oct 27 12:45:30 re Oct 27 12:51:04 kaloz * r18182 /trunk/target/linux/iop32x/Makefile: upgrade iop32x to 2.6.30 Oct 27 12:54:14 russell_: orion is based on arm926ejs, yes Oct 27 12:54:32 not really Oct 27 12:54:43 :) Oct 27 12:54:43 ok, "inspired by" Oct 27 12:54:48 ok :) Oct 27 12:55:02 it's like arm926ejs grown up and taking steroids :) Oct 27 12:55:22 Kaloz: i think they changed the whole pipelining etc Oct 27 12:55:38 i remember when the first feroceons came out and no JTAG dongles worked Oct 27 12:55:45 they just looked lik arm926ejs Oct 27 12:55:46 yes, they did.. first arm with variable pipelines Oct 27 12:55:47 :) Oct 27 12:56:13 the design (accidently) is closer to xscale than to 926ejs Oct 27 12:56:33 blogic: hardware-wise, that hasn't really changed Oct 27 12:57:32 i think i was the first person to get halt-on-reset working on feroceon in openocd Oct 27 12:58:05 hehe Oct 27 12:58:50 i submitted a couple of kludgey patches that made it work for everyone Oct 27 13:00:11 russell_: you still there? Oct 27 13:03:34 /etc/init.d/led have start 60 Oct 27 13:03:47 /etc/init.d/done have start 95 Oct 27 13:04:05 but done is changing some leds to advise that boot is...done Oct 27 13:04:39 I think the led configuration is what we want for system default and not only for the time between boot 60 and 95 Oct 27 13:04:50 so I think it should be change to 96 Oct 27 13:04:56 nunojpg: why? Oct 27 13:06:03 /etc/init.d/done overrides WPS led set at /etc/config/system Oct 27 13:06:31 i don't think so Oct 27 13:06:43 why? Oct 27 13:07:02 /etc/config should not touch the led Oct 27 13:07:26 so what is the porpuose of /etc/init.d/led? Oct 27 13:07:43 boot Oct 27 13:07:56 boot? Oct 27 13:08:05 60-95 Oct 27 13:08:38 /etc/config/system let's configure hostname, timezone and led state Oct 27 13:08:59 led state? Oct 27 13:09:01 I think the intention is what to set the leds forever Oct 27 13:09:19 why would anyone do that? Oct 27 13:09:45 an led should mean something Oct 27 13:09:55 To save power, to put it a little darker, whatever Oct 27 13:10:27 ok, so you think /etc/init.d/led should not exist Oct 27 13:10:33 no Oct 27 13:11:02 it sets the diag led, cleared in done Oct 27 13:11:08 negative Oct 27 13:11:36 it only applies the led settings at /etc/config/system(none by default) Oct 27 13:17:25 nunojpg: sorry, /etc/preinit sets it Oct 27 13:18:55 so yes, you're right Oct 27 13:20:19 Hauke: current brcm47xx trunk confirmed booting on wgt634u, but not wrt350n Oct 27 13:21:18 did openwrt brcm47xx ever worked on wrt350n? Oct 27 13:21:24 yes Oct 27 13:21:43 before i flashed it, it had a booting brcm47xx Oct 27 13:22:42 i don't remember the revision Oct 27 13:23:02 i think it was 13xxx or 14xxx Oct 27 13:23:24 possibly 15xxx Oct 27 13:23:42 and now kernel 2.6.28 does not compile any more? Oct 27 13:24:08 correct, but i got it to compile by nuking a newer patch in generic-2.6 Oct 27 13:25:33 does it boot? Oct 27 13:25:50 no; same result Oct 27 13:27:36 just submitted a patch, although its strange to have a script running after "done" Oct 27 13:27:49 maybe the solution is to integrate led with done Oct 27 13:27:50 florian * r18183 /trunk/target/linux/rdc/config-2.6.30: [rdc] resync kernel configuration, re-enable printk and make the kernel fit under 768KB Oct 27 13:28:23 hmm, [florian] is around Oct 27 13:28:37 sn9: so the problem was not caused by the kernel upgrade from 28 -> 30 Oct 27 13:28:44 Hauke: correct Oct 27 13:29:25 either the breakage is earlier, or it affects all kernels Oct 27 13:31:16 I doen't know where vthe probelm is Oct 27 13:31:48 i think if any of us did, we'd have solved it by now Oct 27 13:32:36 I think the only option is to try out some older versions to narrow the problem down to a change Oct 27 13:32:53 hmm Oct 27 13:33:23 sn9: you could also review the patches that succeeded the one that booted fine for you, to narrow the search Oct 27 13:33:47 danage: i don't know which one that was anymore Oct 27 13:33:58 rats Oct 27 13:34:05 mice Oct 27 13:34:10 drat Oct 27 13:34:17 dmouse Oct 27 13:35:08 all is broken today. first my tooth, then my ikea armchair, and now this Oct 27 13:35:46 no, this was just as broken yesterday Oct 27 13:36:29 my tooth too, the doctor told me, i just didn't notice. the armchair is debatable Oct 27 13:36:41 oh Oct 27 13:36:55 how i love the sound of dental equipment in the morning Oct 27 13:37:27 ok but i will shut up now, because this all won't fix bcm47xx either Oct 27 13:37:35 danage: sounds like a day to take a break and step away from it all, go to the beach or out for a ride Oct 27 13:38:14 or out for a ride to the beach Oct 27 13:38:39 or a ride on the beach... if you got the right vehicle Oct 27 13:39:08 OutBackDingo: :) i wish there were beaches around Oct 27 13:39:10 * sn9 does not know the closest beach to kassel, anyway Oct 27 13:39:50 danage: sorry to hear beach here is 1 block from me, but i live on an island :) LOL Oct 27 13:40:41 sn9: you know, unlike with teeth, with routers there's always the eazy-e option: "throw it in the gutter, go buy another" Oct 27 13:41:14 another will have its own headaches Oct 27 13:42:21 and eazy-e was hardly a role model Oct 27 14:01:54 florian * r18184 /trunk/target/linux/ifxmips/Makefile: [ifxmips] remove the 2.6.28 specific Kernel/Prepare hack Oct 27 15:12:27 florian * r18185 /trunk/target/linux/generic-2.6/ (config-2.6.30 config-2.6.31): [kernel] add more missing configuration symbols Oct 27 17:25:37 xMff: alive ? Oct 27 19:18:04 kaloz * r18186 /trunk/target/linux/ (ppc40x/config-2.6.30 ppc44x/config-default): disable ppc4xx watchdog, it's unreliable Oct 27 19:19:08 kaloz * r18187 /trunk/target/linux/ppc40x/ (config-default patches/): remove 2.6.28 support from ppc40x Oct 27 19:20:33 kaloz * r18188 /trunk/target/linux/ppc40x/ (config-2.6.30 config-default patches/ patches-2.6.30/): ppc40x: use version number-less patches and config Oct 27 19:53:25 r15000 boots; r16000 does not. gonna try r15500 Oct 27 19:57:25 * maligor hugs git bisect Oct 27 19:58:21 * xMff bisects maligor Oct 27 20:00:20 sn9: :) Oct 27 20:00:51 sherlock n9 Oct 27 20:03:11 xMff, You forgot to specify the commit hashes Oct 27 20:03:34 damn Oct 27 20:14:58 OutBackDingo: now yes Oct 27 20:19:22 interesting: r15000 used 2.6.28.8, r15500 used 2.6.28.9, and r16000 used 2.6.28.10 Oct 27 20:26:38 OutBackDingo: OT: i notice you're in the tampa area; would you by any chance be at the ubuntu-florida local community team get-together on 11/7? Oct 27 20:30:40 sn9: HELL NO Oct 27 20:30:44 hahahha Oct 27 20:30:51 sn9: where you at ? Oct 27 20:31:08 san francisco (for the moment) Oct 27 20:31:28 sn9: got freinds in florida i guess ? Oct 27 20:31:36 im a BSD person by trade :) Oct 27 20:31:43 haha Oct 27 20:31:49 atm, still planning to go to FL shortly Oct 27 20:32:06 hoping to have a working usb-3g router there Oct 27 20:32:17 danage: whats funny ? Oct 27 20:32:28 sn9: shouldnt be too hard Oct 27 20:33:06 the only device i have that i can immediately think of as having usb is the wrt350n Oct 27 20:33:18 although, the anas350 also has it Oct 27 20:35:45 getting that device ready would likely be even more work atm (i should know; i did the initial port to it) Oct 27 20:37:46 for some bizarre reason, download.pl is taking an ungodly long time to fetch kernel 2.6.28.9 Oct 27 20:38:45 hrmm seems everything in californias gotten extremely slow Oct 27 20:39:26 90% done after 36 min Oct 27 20:39:50 sn9: jeeez OUCH! Oct 27 20:40:21 92% done after 37 min Oct 27 20:40:24 OutBackDingo: the fact that you associate "hell" with "ubuntu" because you are a "BSD person" :P Oct 27 20:40:50 94% done after 38 min Oct 27 20:41:12 danage: LOL Oct 27 20:41:37 96% done after 39 min Oct 27 20:41:41 danage: i actually think ubuntus ok for a desktop Oct 27 20:42:02 but the words linux and server dont go in the same sentence with me Oct 27 20:42:10 phew, so i might stand a chance of not rotting in hell :) Oct 27 20:42:25 ^^ but they did! ;) Oct 27 20:42:32 could be worse; could be os x Oct 27 20:43:00 sn9: hahaha OS X is ok on the desktop Oct 27 20:43:34 i was referring to being in the same sentence with server Oct 27 20:45:19 hahahaha no server i know needs a GUI Oct 27 20:45:38 right, and that includes openwrt Oct 27 20:45:45 osx is probably as good at being a server as openbsd is at being a desktop ;) Oct 27 20:46:00 no, not even Oct 27 20:47:09 openbsd actually makes a tolerable desktop; osx does not do that for servers Oct 27 20:47:59 why not? Oct 27 20:48:13 you can compile services from ports if you want to ;) Oct 27 20:48:35 if they work; the xnu kernel is severely brain-damaged Oct 27 20:48:46 i'm using osx on my main dev machine Oct 27 20:48:51 and i haven't noticed many xnu fuckups yet Oct 27 20:49:07 i used to use it for main dev Oct 27 20:49:26 got sick of it, and went to a real os Oct 27 20:50:09 fuck theo Oct 27 20:50:15 excuse the french but Oct 27 20:51:03 you're inhuman! :D *scnr* Oct 27 20:51:37 nbd: lol all i needed to say on that topic Oct 27 20:52:30 nbd: too bad their dumping ZFS though Oct 27 21:19:00 nbd * r18189 /trunk/package/lua/ (7 files in 2 dirs): lua: re-enable host builds with some more fixes Oct 27 21:20:53 nbd: so who is actually the luci team other then xMff and maybe jow Oct 27 21:21:06 jow == xMff Oct 27 21:21:35 ahhhh Oct 27 21:21:42 is he the only luci dev ? Oct 27 21:21:57 no, there's cyrus as well, but he hasn't been active in luci trunk for a while Oct 27 21:22:12 the project was started by cyrus Oct 27 21:22:17 ahh got it Oct 27 21:28:17 nbd * r18190 /trunk/package/lua/Makefile: lua: fix a typo Oct 27 21:38:51 r15500 does not boot, either Oct 27 21:43:21 you're close Oct 27 21:45:48 trying r15017 next Oct 27 22:34:38 hmm, that boots -- same kernel Oct 27 22:41:02 building r15250... Oct 27 22:50:36 florian * r18191 /trunk/package/siit/src/siit.c: [package] fix siit compile with 2.6.30 kernels and superior, when CONFIG_COMPAT_NET_DEV_OPS is not set Oct 27 23:17:01 that boots, too; on to r15375... Oct 27 23:43:25 jow * r18192 /packages/lang/luaexpat/ (Makefile patches/001-compile-fix.patch): [packages] luaexpat: fix compile error triggered by cs toolchain Oct 28 00:01:54 that boots, too; on to r15438... Oct 28 00:12:01 jow * r18193 /trunk/package/lua/Makefile: [package] lua: the variable expansion to detect the host os does not work with GNU Make 3.81, replace it with a nested ifeq Oct 28 00:28:57 jow * r18194 /packages/lang/luaexpat/Makefile: [packages] luaexpat: compile with -fPIC Oct 28 00:33:14 that boots, too; on to r15469... Oct 28 01:07:33 that boots, too; on to r15485... Oct 28 01:34:46 that boots, too; on to r15493... **** ENDING LOGGING AT Wed Oct 28 02:59:58 2009