**** BEGIN LOGGING AT Thu Aug 06 02:59:57 2009 Aug 06 05:57:32 anybody familiar with quilt? need pointers in how to ammend a particular patch. Aug 06 06:17:27 good morning juhosg Aug 06 07:24:38 danage: I will probably arrive late at tonights meeting, can you do the practicals Aug 06 07:28:36 glp: i won't be there :( Aug 06 07:28:43 glp: xMff said he would though... Aug 06 07:29:39 danage: can you ask him? I have to leave now Aug 06 07:29:56 will do. i won't be back until about 2 weeks from today either Aug 06 08:39:13 hi I have made some patches for libuci Aug 06 08:39:50 to offer a "better" interface for C programs. who should I submit the patch to for consideration if it is useful? Aug 06 08:40:14 <[florian]> hi alesan Aug 06 08:40:19 hi Aug 06 08:40:22 <[florian]> alesan: send them to nbd or xMff Aug 06 08:40:31 like in an email? Aug 06 08:40:56 well I will contact him Aug 06 08:41:06 xMff, are you awake? Aug 06 08:41:44 <[florian]> alesan: I think email to both is fine Aug 06 08:42:08 yes Aug 06 08:42:36 listen, I have a board I would like to port to openwrt. Aug 06 08:44:33 what is my best bet: is there a guide or can I send one to you or somebody. Unfortunately, the board is no public yet but will be in the merket in a few months Aug 06 08:45:14 alesan: -devel mailing list is a good place Aug 06 08:45:20 <[florian]> write down an email to openwrt-hackers/devel Aug 06 08:45:51 where do I subscribe? let me see Aug 06 08:46:52 alesan: lists.openwrt.org Aug 06 08:47:10 yeah Aug 06 08:47:49 [florian], is "-hackers" a "hidden" mailing list :) ?? Aug 06 08:49:17 ^^ oops :P Aug 06 08:50:08 <[florian]> it is an internal mailing-list, archives are not public, but still one can write down an email to it for a specific need like this ;) Aug 06 08:53:31 well ok I will try with -devel at first Aug 06 08:55:04 actually I need an advice first Aug 06 08:57:25 this board is not released yet. if I send some free samples to "core" people in the openwrt community, do you think the port can be just prepared but not made public until the release of the product on the market? Aug 06 08:58:16 Can I create a LuCI module for edit some files in /etc/ folder? Aug 06 08:58:58 dkostousov: http://luci.freifunk-halle.net/Documentation Aug 06 08:59:05 <[florian]> alesan: of course Aug 06 09:01:07 moonflux: I know about this doc. I have written simple LuCI module for /etc/config/mycfg edititon. Aug 06 09:01:33 LuCI CBI was used for that. Aug 06 09:02:03 But I can't use CBI for edit any file. Aug 06 09:02:46 dkostousov: ah, you mean /etc Aug 06 09:03:33 dkostousov: you can edit any file ,but not with cbi of course Aug 06 09:15:07 juhosg * r17140 /trunk/tools/firmware-utils/src/addpattern.c: [tools] firmware-utils/addpattern: add support for the WRT160NL Aug 06 09:32:18 juhosg * r17141 /trunk/target/linux/ar71xx/files/arch/mips/ar71xx/irq.c: [ar71xx] get rid of some ifdefs in the IRQ code Aug 06 09:42:30 juhosg * r17142 /trunk/target/linux/ar71xx/files/arch/mips/ar71xx/irq.c: [ar71xx] oops, add missing semicolons Aug 06 10:07:38 I have write my own config file in eth /etc/config folder. How I cat access to this file from my init script? Aug 06 10:12:28 dkostousov: http://kamikaze.openwrt.org/docs/openwrt.html#x1-310001.3 Aug 06 10:12:40 :) Aug 06 10:14:32 thx :-) Aug 06 10:56:51 juhosg * r17143 /trunk/target/linux/ar71xx/files/drivers/spi/ar71xx_spi.c: [ar71xx] update ar71xx_spi driver Aug 06 11:11:39 xMff: ping Aug 06 11:12:44 omg, building openwrt on 64bit is still a bit patchy. Aug 06 11:14:54 46 packages failed. Aug 06 11:15:34 hm, it's been a while since last time i did a full build Aug 06 11:15:37 which is about 5% . Aug 06 11:15:37 maybe i should do it again Aug 06 11:15:49 but this shouldn't be 64 bit specific Aug 06 11:15:57 usually many openwrt developers build stuff on 64 bit Aug 06 11:15:58 iirc Aug 06 11:16:09 nbd: it might also be that I'm trying to compile 64bit x86 on 64bit x86 Aug 06 11:16:09 i certainly do (whenever i'm not building on mac os x) Aug 06 11:16:15 hmm. ok Aug 06 11:16:18 yes, that's a different matter Aug 06 11:16:58 i seem to be getting more luck with building 64bit x86 on 32 bit x86 Aug 06 11:17:12 interesting Aug 06 11:17:20 maybe some cross compiling checks are failing? Aug 06 11:17:27 .. which unfortunately is slower by a goodly order of magnitude Aug 06 11:18:04 yes, asterisk ./configure is making a meal of the checks. Aug 06 11:18:44 first it was failing 'cause pwlib was using stroq and it wasn't linking. Aug 06 11:19:02 hmm.. I wonder if it's looking in the wrong place for strtoq Aug 06 11:19:32 the asterisk configure seems also to be failing to find it. Aug 06 11:19:54 (linking against the wrong library, perhaps) Aug 06 11:20:22 danage: pong Aug 06 11:21:20 nbd: I could give you my configs for 64bit x86 ... Aug 06 11:21:41 i don't have time to dig through this currently Aug 06 11:22:31 xMff: glp and i won't make the meeting today Aug 06 11:22:32 np. Aug 06 11:22:35 can you moderate? Aug 06 11:22:57 I'll pastebin a config.log .. maybe somebody can help. Aug 06 11:23:09 danage: hm, I'll try Aug 06 11:23:14 thanks Aug 06 11:23:37 how many hours till the meeting starts xMff? Aug 06 11:24:27 frogonwheels: 20:00 CEST iirc that's +6h30m from now on Aug 06 11:25:01 4 in the morning. think I'll pass. Aug 06 11:25:29 I'll publish the log Aug 06 11:35:24 strange. stroq is defined in stdlib.h but has no implementation in ucLibc.. /usr/lib/* Aug 06 11:36:29 which seems to be the battle. Aug 06 12:02:53 jow * r17144 /trunk/package/broadcom-wl/ (Makefile src/wlc/wlc.c): [package] broadcom-wl: add "assoclist" command to wlc Aug 06 12:13:59 xMff, I made some patches to libuci to offer a kind of "easier" interface for C programs Aug 06 12:14:37 xMff, [florian] suggested to send the patch either to you or nbd ... do you think I can send you an email? Aug 06 12:14:56 please send it to me Aug 06 12:14:57 alesan: nice, send it to nbd please since he is the one maintaining it in the end Aug 06 12:15:19 oh ok Aug 06 12:15:34 nbd, I guess it's.... nbd@openwrt.org Aug 06 12:15:39 yes Aug 06 12:15:45 well there must be the email in the soueces Aug 06 12:16:19 ok thanks I will do later tonight when I finish my work here Aug 06 12:17:29 oh I have a more generic question: the moment one writes a partition with mtd write - and that partition is the rootfs... who guarantees that the system that is currently running will not access the flash and either crash or pollute the flash? Aug 06 12:18:08 sysupgrade tries to unmount as much as possible from the filesystem before writing its stuff Aug 06 12:19:04 shouldn't a remount,ro solve most of this? Aug 06 12:19:30 yeah Aug 06 12:19:43 yes I agree Aug 06 12:19:45 ok thanks Aug 06 12:31:39 nbd: atheros build with 2.6.30.4 fails at the end: http://pastebin.com/m1568afb3 Aug 06 12:32:19 you have a 16 core machine? Aug 06 12:32:30 yes Aug 06 12:32:49 i'll check the source, just a sec Aug 06 12:37:12 loswillios: that's really odd Aug 06 12:38:51 hm Aug 06 12:39:04 i have an idea Aug 06 12:39:24 it was working fine a few days ago Aug 06 12:39:51 r17099 built successfully Aug 06 12:40:37 until recently atheros didn't default to using linux 2.6.30 Aug 06 12:40:42 and thus squashfs 4.0 Aug 06 12:41:26 please try adding -processors 1 to SQUASHFS_OPTS in include/image.mk, then run make target/install V=99 Aug 06 12:43:46 works Aug 06 12:44:10 ok, so mksquashfs has a threading problem Aug 06 12:44:27 * loswillios nods Aug 06 12:45:38 ah Aug 06 12:45:41 i have another idea Aug 06 12:45:51 sh00t Aug 06 12:46:02 please revert that change in image.mk Aug 06 12:46:06 i'll send you a patch Aug 06 12:48:11 please apply http://nbd.name/squashfs.patch - then run make tools/squashfs4/install target/install V=99 Aug 06 12:49:48 nope, doesn't work Aug 06 12:50:18 did it recompile squashfs4? Aug 06 12:50:34 yes Aug 06 12:50:36 if hm Aug 06 12:50:37 hm Aug 06 12:51:38 and it added -pthread to the cflags correctly? Aug 06 12:51:52 yes Aug 06 12:51:59 hm Aug 06 12:52:08 ok, i guess i'll have to add the workaround then Aug 06 12:52:32 are you building on UP systems? Aug 06 12:52:38 no, dual core Aug 06 12:52:40 works there Aug 06 12:52:45 hm Aug 06 12:52:57 though i think i've seen a similar error at least once Aug 06 12:55:24 nbd * r17145 /trunk/target/linux/ (5 files in 3 dirs): atheros: rename CONFIG_ATHEROS to CONFIG_ATHEROS_AR231X Aug 06 12:55:42 nbd * r17146 /trunk/include/image.mk: work around a threading bug in mksquashfs4 by limiting the number of threads to 1 Aug 06 12:55:57 thanks Aug 06 12:56:30 nbd, will you send the bug upstream in such case? Aug 06 12:57:28 first i need to make sure it's not something that i caused Aug 06 12:57:45 because i added lzma support Aug 06 12:58:05 in fact i think it's likely that the problem is in the lzma support Aug 06 13:58:35 I tought squashfs was already lzma based Aug 06 13:58:37 isn't it?? Aug 06 13:59:18 no Aug 06 14:00:34 what was it using then? Aug 06 14:00:41 gzip/zlib Aug 06 14:01:22 hi all! Aug 06 14:01:27 well, we've always used lzma, but the official squashfs has never supported it. Aug 06 14:01:32 [florian] ping Aug 06 14:01:45 I see Bartman007 Aug 06 14:05:56 <[florian]> AndyI: pon Aug 06 14:06:28 florian * r17147 /trunk/target/linux/generic-2.6/patches-2.6.30/027-mips_module_reloc.patch: [kernel] make 64-bits MIPS kernel compile again on arch/mips/kernel/module.c Aug 06 14:06:36 <[florian]> There was time to look SPI? Aug 06 14:06:45 <[florian]> AndyI: not yet, I will do later tonight, I promise Aug 06 14:08:09 <[florian]> ok Aug 06 14:14:17 there is also this: https://lists.openwrt.org/pipermail/openwrt-devel/2009-February/003886.html Aug 06 14:35:15 <[florian]> AndyI: what is the flash on your board? Aug 06 14:46:58 <[florian]> on 2 board parallel on 1 board spi Aug 06 14:47:15 <[florian]> AndyI: what is the exact model of the SPI flash? Aug 06 14:47:38 shared spi bus? Aug 06 14:47:38 <[florian]> AndyI: can you show me the CFE bootlog of the SPI flash-based device Aug 06 14:48:06 <[florian]> pardon ? Aug 06 14:48:23 <[florian]> macronix mx25l1605a Aug 06 14:48:33 <[florian]> SPI is almost always shared as you can have multiple slaves on the same SPI bus Aug 06 14:48:36 <[florian]> AndyI: thanks Aug 06 14:49:13 yeah, and ar71xx spi bus driver does not support shared bus :( Aug 06 14:49:36 <[florian]> Dogge_: maybe because it cannot drive the slave select signals Aug 06 14:50:08 [florian] no idea Aug 06 16:38:06 xMff: hello, did you try lua in windows? Aug 06 16:38:48 a friend of mine did to make some of the luci C libs cross-platform, why? Aug 06 16:39:55 no, because I'm download it but I did read have some issues and only run under Administrator account Aug 06 16:40:21 it should run from within any location Aug 06 16:40:35 maybe the installer only runs as administrator because it wants to register file extensions Aug 06 16:40:46 and put various paths in the registrs Aug 06 16:40:59 yes, is logic Aug 06 16:41:06 registry* Aug 06 16:42:44 any way is only to try it and use some script of socket from my pc Aug 06 16:43:27 install it somehwere and only copy the installed directory over? Aug 06 16:43:39 maybe you can even just extract the installer Aug 06 16:43:50 or maybe there's even a simple .zip file of luaforwindows Aug 06 16:45:29 yea I did download it and now it's installing Aug 06 16:47:42 ok, it's installed and working Aug 06 16:47:51 nice Aug 06 16:55:28 jejeje, now i need update my paths but look like work ok Aug 06 17:01:23 hi Aug 06 17:12:22 hi Aug 06 17:26:10 [florian] fix bcm6345 watchdog reboot Aug 06 17:27:04 danage ping Aug 06 17:32:50 any atheros experts around Aug 06 17:34:14 https://forum.openwrt.org/viewtopic.php?id=21193 Aug 06 17:34:25 nbd: you hack on madwifi alot right? Aug 06 17:34:46 is it possible or do i need to find a broadcom unit Aug 06 17:35:25 forget client bridging Aug 06 17:35:30 it's a stupid idea, no matter which driver Aug 06 17:35:49 just make the device use the client mode interface as wan Aug 06 17:36:04 wow. people still mess with client bridged-mode :( Aug 06 17:36:14 at least some are considering it Aug 06 17:36:19 bad enough ;) Aug 06 17:41:44 hi Aug 06 17:42:02 WRT160NL flashing supported now? Aug 06 17:42:37 nbd: but can i make another ap so i don't loose wifi? Aug 06 17:42:50 or do i need 2 radios Aug 06 17:43:00 you can do ap + station in madwifi Aug 06 17:43:10 the ap will be brought up as soon as the station has associated Aug 06 17:43:16 got an example config? Aug 06 17:43:27 or i'll just screw around when i get home Aug 06 17:43:34 just make one wifi-iface for sta mode and one for ap mode Aug 06 17:43:39 nothing complicated Aug 06 17:43:45 hi AndyI Aug 06 17:44:04 freezer: i think so, groz made that happen Aug 06 17:44:21 ok thanks Aug 06 17:44:22 ya talked to him yesterday Aug 06 17:47:57 your bcm6345 have watchdog problem? Aug 06 17:48:22 AndyI: it crashes, but i don't know watchdog is the reason. why? Aug 06 17:48:49 please send my bootlog Aug 06 17:51:14 [florian]: do you still have one of my bootlogs? Aug 06 17:58:10 And for you it was not saved? Aug 06 18:13:21 AndyI: i put into pastebin, but i lost the backlog, it's on another machine Aug 06 18:23:53 ok Aug 06 18:24:10 Howdy.. .sorry I am late... slept in ;) Aug 06 18:24:29 What did I miss? Aug 06 18:25:49 Is anyone here???? Aug 06 18:29:13 test Aug 06 18:29:41 ok... so my firewall is working, but no one is here... Aug 06 18:41:44 <[florian]> AndyI: fix watchdog, is it a question or an order ? Aug 06 18:44:22 florian * r17148 /trunk/target/linux/brcm63xx/files/arch/mips/bcm63xx/dev-wdt.c: [brcm63xx] disable watchdog for bcm6345 until it gets fixed Aug 06 18:46:50 lol Aug 06 18:48:26 <[florian]> this solved problem "bcm_writel(0xffffffff-1, bcm63xx_wdt_device.regs + WDT_DEFVAL_REG);" Aug 06 18:48:55 <[florian]> AndyI: oh, ok, should be safe for others too Aug 06 18:49:26 <[florian]> 6338 & 6332 work fine with this patch Aug 06 18:49:40 <[florian]> AndyI: ok Aug 06 18:51:14 <[florian]> AndyI: about SPI, I will send you a patch to test after dinner Aug 06 18:51:19 [florian]: this fixes watchdog? Aug 06 18:51:47 <[florian]> danage: apparently Aug 06 18:51:59 <[florian]> danage: hold on I commit the complete fix Aug 06 18:52:04 interesting Aug 06 18:52:54 yes Aug 06 18:53:45 florian * r17149 /trunk/target/linux/brcm63xx/files/ (arch/mips/bcm63xx/dev-wdt.c drivers/watchdog/bcm63xx_wdt.c): [brcm63xx] fix off-by-one maximum timeout in the wathchdog default counter for bcm6345, thanks AndyI Aug 06 18:53:58 <[florian]> AndyI, danage I prepare a dinner, brb Aug 06 18:54:06 :) Aug 06 18:54:20 k folks, got what I _hope_ is a simple question for ar71xx Aug 06 18:54:40 it appears that when I insert an adapter into a bridge, eth0 in this case, i'm not getting stp Aug 06 18:54:57 adapter is not automatically placed in promiscuous mode, I manually set promiscuous Aug 06 18:55:08 and i'm still nto seeing stp on that adapter Aug 06 18:55:35 is there a trick to getting ar71xx into promiscuous mode for bridging to do stp properly ? Aug 06 18:55:45 or is it more likely the switch on this box ? Aug 06 18:57:40 <[florian]> ok waiting spi patch (Bon appetit!) Aug 06 18:58:38 hi groz Aug 06 18:58:58 groz, just compiling the trunk Aug 06 18:59:10 oh, you got it to patch now ? Aug 06 18:59:15 <[florian]> AndyI: aligato Aug 06 18:59:30 <[florian]> this not time-out waiting register Aug 06 18:59:33 groz, it has been added the the trunk some hours ago?! Aug 06 18:59:40 *to the Aug 06 18:59:43 ahh, ok Aug 06 18:59:51 <[florian]> AndyI: this is not what my comment says, this is the default counter Aug 06 18:59:55 * groz been madly packing the van and prepping to leave Aug 06 19:00:05 now i'm just trying to solve one more little issue before we hit the road Aug 06 19:00:24 <[florian]> ;-) Aug 06 19:00:48 that is? Aug 06 19:01:03 not getting stp between my 160nl and another box Aug 06 19:01:11 so, bridging not completely right Aug 06 19:01:17 would like to get that fixed if I can Aug 06 19:02:29 for some reason, it wont run stp over a wired link to a wrt54gl I have running here Aug 06 19:02:50 which is running 2.4 with all the broadcom stuff Aug 06 19:04:46 groz: imho the realtek Aug 06 19:04:47 or has Europe gone dst?? Aug 06 19:05:38 i'm thinking probably kaloz, so, just testing it by plugging a usb eth in Aug 06 19:05:39 did the wiki meeting happen. or not ? Aug 06 19:06:56 juhosg * r17150 /trunk/target/linux/ar71xx/ (2 files in 2 dirs): [ar71xx] add try partition parser for the WRT160NL board (thanks to Christian Daniel and Gerry Rozema) Aug 06 19:07:49 juhosg * r17151 /trunk/target/linux/ar71xx/image/Makefile: [ar71xx] create firmware image for the WRT160NL board (thanks to Gerry Rozema) Aug 06 19:09:24 aah Aug 06 19:09:34 :) Aug 06 19:09:52 now i built without those :( Aug 06 19:11:13 svn up; make target/linux/clean; make world Aug 06 19:13:43 okay thx Aug 06 19:17:06 groz: i did not try stp yet, so it may be a problem also with the ag71xx driver Aug 06 19:17:25 that's why i bypassed the ag71 completely ju Aug 06 19:17:40 and stuck in a usb attach ethernet adapter known to work on another box Aug 06 19:17:44 and pass stp properly Aug 06 19:17:45 the ethernet driver/mac isn't supposed to interfere with STP, i think the switch is a more likely cause Aug 06 19:18:11 because some switches try to be 'smart' about handling STP Aug 06 19:18:14 I bypassed everything, and still have the problem, so, gonna resync with trunk Aug 06 19:18:20 and rebuild the world Aug 06 19:18:36 thinking something I've done in the kernel over the last few days may have broke the bridge Aug 06 19:18:53 are you sure stp was working before? Aug 06 19:19:07 i know it's working on an x86 box with basically all the same Aug 06 19:19:18 and using this same usb ethernet device Aug 06 19:19:42 the only thing different is that i've got it running now on the wrt160nl Aug 06 19:19:55 and i've bypassed all the on board ethernet stuff with the usb gadget Aug 06 19:19:58 still got the issue Aug 06 19:20:13 ah Aug 06 19:20:36 so that eliminates a lot me thinks Aug 06 19:21:59 besides, somebody should do a fresh build out of trunk, and test all the stuff ju just committed Aug 06 19:22:18 make sure it survived the trip from patches to commits ok Aug 06 19:29:56 how to flash the WRT160NL image? Aug 06 19:30:04 upgrade code.bin? Aug 06 19:30:15 from uboot, yes Aug 06 19:30:18 upgrade code.bin Aug 06 19:30:26 or, upload to the web interface of linksys code Aug 06 19:33:00 and if openwrt is already running, i can use mtd for flashing? Aug 06 19:33:08 should be able to Aug 06 19:33:19 :<> Aug 06 19:33:24 i've done it here with mtd write xxx.bin firmware Aug 06 19:33:58 xxx = code? Aug 06 19:34:09 yah, whatever the filename is Aug 06 19:34:18 ok Aug 06 19:39:13 so i got openwrt-ar71xx-wrt160nl-squashfs.bin Aug 06 19:39:29 reeeee Aug 06 19:39:33 in theory, it should just flash in and work Aug 06 19:39:36 rename it to code.bin and flash it? Aug 06 19:39:40 and, i'll know here in a bit Aug 06 19:39:46 groz: whow... you really did it Aug 06 19:39:52 for tftp I've been renaiming it Aug 06 19:40:02 code.bin then upgrade code.bin from uboot Aug 06 19:40:06 simplest and easiest way Aug 06 19:40:22 ya hopscotch, was hollering yesterday to see if you wanted patches Aug 06 19:40:26 but, they are in trunk now Aug 06 19:40:32 cool Aug 06 19:40:41 the stuff from you modified a bit Aug 06 19:40:46 some stuff from juhosg Aug 06 19:40:51 and i added a bit for bins etc Aug 06 19:40:52 i've been burried in work for the last week Aug 06 19:40:56 and voila, it's working Aug 06 19:41:02 *pull* Aug 06 19:41:06 i just did a fresh checkout from trunk Aug 06 19:41:13 and i'm building to do a test flash on it now Aug 06 19:41:24 me too Aug 06 19:42:45 its flashing now Aug 06 19:43:36 ok Aug 06 19:43:41 i'm hoping this build finishes before I gotta hit the road Aug 06 19:43:42 now comes the magic moment Aug 06 19:43:42 haha Aug 06 19:44:21 you flashed the squash, expect to see it detect properly, and do the mini_fo: magic on the rest of free flash Aug 06 19:44:56 looks good Aug 06 19:45:00 openwrt prompt! Aug 06 19:45:08 cool Aug 06 19:45:11 ok, ifconfig -a Aug 06 19:45:14 then mount Aug 06 19:45:17 wifi and eth1,eth0 Aug 06 19:45:20 all up Aug 06 19:45:20 check the mounts Aug 06 19:46:53 root@OpenWrt:/# mount Aug 06 19:46:53 rootfs on / type rootfs (rw) Aug 06 19:46:53 /dev/root on / type jffs2 (rw) Aug 06 19:46:53 proc on /proc type proc (rw) Aug 06 19:46:53 sysfs on /sys type sysfs (rw) Aug 06 19:46:54 tmpfs on /tmp type tmpfs (rw,nosuid,nodev,size=14916k) Aug 06 19:46:56 tmpfs on /dev type tmpfs (rw,size=512k) Aug 06 19:46:58 devpts on /dev/pts type devpts (rw,mode=600) Aug 06 19:47:00 flood, sorry Aug 06 19:47:02 :P Aug 06 19:47:38 ah Aug 06 19:47:44 i flashed the jffs2 Aug 06 19:47:53 ok, it worked Aug 06 19:48:06 juhosg is spi shared bus working now? Aug 06 19:48:16 if you got the prompt flashing the jffs2, then the partitions are correct, and it found the root file system Aug 06 19:48:31 ya Aug 06 19:48:46 telnet connect to 192.168.1.1 works too Aug 06 19:50:04 back in 5 Aug 06 20:01:09 Dogge_: i don't know, but probably not. i can't check it, because my RB-433AH has been died :/ Aug 06 20:01:30 [florian] look e-mail Aug 06 20:02:06 <[florian]> AndyI: ok Aug 06 20:05:56 <[florian]> AndyI: sounds good Aug 06 20:08:29 kaloz * r17152 /trunk/package/opkg/ (6 files in 2 dirs): upgrade opkg to r215 from the googlecode svn Aug 06 20:25:14 ok, just did a fresh checkout from trunk, configured for wrt160nl, built the whole mess, and flashed the resulting bin for squashfs Aug 06 20:25:20 it flashed ok, and came up correctly Aug 06 20:33:44 florian * r17153 /trunk/target/linux/brcm63xx/files/ (4 files in 3 dirs): [brcm63xx] bcm6345 fixes from AndyI Aug 06 20:59:41 [florian] about spi? Aug 06 20:59:58 <[florian]> AndyI: hold on Aug 06 21:10:19 Worth while writing in support for a config option for a custom opkg repo source? Aug 06 21:10:26 I've got a patch (just testing) Aug 06 21:10:36 <[florian]> AndyI: can you try that: http://openwrt.pastebin.com/m3aa7dab9 Aug 06 21:10:47 <[florian]> AndyI: no guarantee :p Aug 06 21:11:16 <[florian]> ok Aug 06 21:21:14 <[florian]> this only add flash definition Aug 06 21:21:47 questions about why driver hang on loading? Aug 06 21:21:57 <[florian]> questions about why driver hang on loading? Aug 06 21:23:39 <[florian]> and compile error Aug 06 21:31:57 I'm having some interesting differences compiling 64bit x86 on 64bit x86 vs compiliing it on 32bit x86.. Aug 06 21:35:24 <[florian]> AndyI: ok, I did not compile test this Aug 06 21:35:36 <[florian]> AndyI: I go sleeping now Aug 06 21:36:01 <[florian]> AndyI: will provide you a better patch later Aug 06 21:36:19 <[florian]> ok waiting! send on my email Aug 06 21:43:13 submitted Aug 06 21:45:37 I'm looking at asterisk - and its ./configure script is not working particularly well... or something Aug 06 21:46:17 on a 32 bit host it works fine, but on the 64bit host, it's having issues with strtoq Aug 06 21:56:03 trac (dev.openwrt.org) is diving db errors Aug 06 22:02:06 does anybod know what the make target is that builds the kernel images Aug 06 22:02:12 ? Aug 06 22:07:03 hmm.. target/linux-install probably. Aug 06 22:11:31 frogonwheels: yes, that shold do it. Aug 06 22:13:22 kupesoft: microsoft seems to think it's ok to throw 8+ spiders at dev.openwrt.org at once Aug 06 22:13:33 Bartman007: :) hey, I'm having problems booting my new image on a new target. Aug 06 22:13:54 Bartman007: That's not as bad as when spiders lied eggs in my apartment :( Aug 06 22:14:02 s/lied/laid/ Aug 06 22:14:08 Bartman007: It's not recognising the drives (gets a kernel panic, unable to mount root fs.) Aug 06 22:14:48 kupesoft: not to mention yahoo, google and a couple others, but the rest of them seem to be acting rather politely in comparsion to the msnbot*.search.msn.com Aug 06 22:14:52 Bartman007: .. which I believe should be /dev/sda2 .. normally just above that message, it should actually tell you what drivers are available, no? Aug 06 22:15:10 Bartman007: Maybe this is part of an marking campaign for Microsoft's unannounced new router? Aug 06 22:15:21 s/marking/marketing/ -- wtf is wrong me Aug 06 22:15:30 Bartman007: (I've got this machine booting under Ubuntu btw) Aug 06 22:16:19 frogonwheels: OpenWrt on x86 is always a headache. What board? Aug 06 22:16:28 .. but it's going to be pretty much a router-on-steroids / asterisk server and hard-drive Aug 06 22:16:32 kupesoft: kontron. Aug 06 22:16:57 frogonwheels: openwrt can only boot if the driver is compiled into the kernel as we don't use an initramfs on disk based images Aug 06 22:17:07 I've built in what I beleive is the correct drivers for the device Aug 06 22:17:51 if you can pastebin a dmesg I'll take a look at it later, but I'm running out the door at the moment. Aug 06 22:18:10 Bartman007: no probs Aug 06 22:18:30 dmesg of a working ubuntu, or non-working openwrt? Aug 06 22:19:54 I suppose both, of course, though getting the bootlog of the non-working openwrt will have to be done via serial Aug 06 22:20:20 pastebin kernel boot log up until the panic if you can Aug 06 22:20:29 yep. unfortunately I'm flat-outa null-modems. Aug 06 22:20:34 :| Aug 06 22:20:41 wow, that sentence is quite bad... I suppose both, though you will of course need to get the openwrt bootload via serial. (much better) Aug 06 22:20:43 i'll get there Aug 06 22:21:33 frogonwheels: your trying to build 64bit for this? have you tried booting in vmware, vbox, kvm, etc for 64bit? Aug 06 22:21:45 you're* Aug 06 22:21:47 nope Aug 06 22:21:55 I mean yes I am, and no I haven't Aug 06 22:22:22 does virtualbox do 64bit? Aug 06 22:22:58 yes Aug 06 22:23:41 frogonwheels: Don't make me pull out lmgtfy.com :P Aug 06 22:23:51 ;) Aug 06 22:23:55 just running it now. Aug 06 22:24:05 frogonwheels: good luck Aug 06 22:24:08 * Bartman007 departs Aug 06 22:24:25 au revoir Aug 06 22:30:31 wow Aug 06 22:30:40 WRT160NL works pretty well Aug 06 22:30:50 wifi with psk2 up n running Aug 06 22:31:22 That's it, I'm buying one tomorrow. Aug 06 22:31:23 :P Aug 06 22:31:38 :D Aug 06 22:32:05 I saw them on the shelves at a store, they're a little expensive though Aug 06 22:32:22 where r u from Aug 06 22:33:45 here they are available from 70eur Aug 06 22:34:32 The one on top of America Aug 06 22:34:33 but 400mhz and 11n is quite nice Aug 06 22:34:36 $118 CAD here Aug 06 22:34:58 Wow, I just went to xe.net, our dollar is strong right now Aug 06 22:35:30 wondering how good the wifi works in noisy environment Aug 06 22:35:40 It's not gigabit, is it? Aug 06 22:35:56 nope Aug 06 22:36:04 meh, they screwed up there IMO Aug 06 22:37:51 maybe Aug 06 22:38:05 only if you really need it Aug 06 22:38:22 its a nice asus wl500gP replacement Aug 06 22:39:04 could have more flash/ram Aug 06 22:41:38 Anything inside? Serial port, jtag, any other headers? Aug 06 22:44:57 http://nanl.de/blog/wp-content/wrt160nl.jpg :) Aug 06 22:45:16 serial is even with header ya Aug 06 22:46:01 i mean with pins attached Aug 06 22:46:42 below the serial is jtag?! Aug 06 22:48:56 No idea, got a multimeter? Aug 06 22:50:39 ya..but not now Aug 06 22:53:18 looks like there are quite some ar71xx routers to come Aug 06 22:53:23 nice to see the progress Aug 06 23:17:09 Wasting time? Aug 06 23:17:09 :D Aug 06 23:17:09 I really don't have a good answer for that one Aug 06 23:17:17 err, sorry Aug 07 00:07:20 nbd * r17154 /trunk/tools/ (5 files in 3 dirs): add mklibs (based on a patch by florian) in preparation for optional library symbol stripping Aug 07 00:07:44 nbd * r17155 /trunk/package/base-files/files/etc/functions.sh: Aug 07 00:07:44 base-files: allow default value for config_get Aug 07 00:07:44 Signed-off-by: Malte S. Stretz Aug 07 00:07:44 Signed-off-by: Felix Fietkau Aug 07 00:07:48 nbd * r17156 /trunk/package/base-files/files/etc/functions.sh: Aug 07 00:07:48 base-files: port functions.sh to new config_get Aug 07 00:07:48 Signed-off-by: Malte S. Stretz Aug 07 00:07:53 nbd * r17157 /trunk/package/base-files/files/etc/init.d/boot: Aug 07 00:07:53 base-files: port init.d/boot to new config_get Aug 07 00:07:54 ... and localize some variables. Aug 07 00:07:56 Signed-off-by: Malte S. Stretz Aug 07 00:07:58 nbd * r17158 /trunk/package/base-files/files/etc/init.d/led: Aug 07 00:08:00 base-files: port init.d/led to new config_get Aug 07 00:08:04 ... and don't call logger directly, rcS takes care of the correct redirection Aug 07 00:08:06 Signed-off-by: Malte S. Stretz Aug 07 00:08:08 nbd * r17159 /trunk/package/base-files/files/etc/init.d/fstab: Aug 07 00:08:10 base-files: clean up init.d/fstab and port it to the new config_get Aug 07 00:08:12 ... also remove the unused noauto variable in swapon. Aug 07 00:08:14 Signed-off-by: Malte S. Stretz Aug 07 00:08:18 nbd * r17160 /trunk/package/base-files/files/lib/network/config.sh: Aug 07 00:08:20 base-files: some /lib/network/config.sh refactoring Aug 07 00:08:22 * Localize a lot more variables than before. Aug 07 00:08:24 * Do stuff or localize variables at later points or inner blocks Aug 07 00:08:26 so less commands are executed in some cases. Aug 07 00:08:28 * Use the default option of config_get where it makes sense. Aug 07 00:08:32 Signed-off-by: Malte S. Stretz Aug 07 00:10:51 kupesoft: I think I've found the problem with booting. forgot to include scsi disk support :| Aug 07 00:12:30 kupesoft: I think I've made the same mistake b4 years ago when CD Writable drivers were only in scsi system. Aug 07 01:57:15 This works fine on a 32bit host: host=Ubuntu 64bit target=x86 64Bit make package/grub-compile fails at configure. config.log snippet: http://pastebin.com/d6a6510b5 Aug 07 01:57:46 (I've made my own 64bit x86 sub-target) Aug 07 02:00:07 given that it's built the other 800+ packages, the compiler works fine.. Aug 07 02:02:29 "cannot find -lgcc" Aug 07 02:02:46 yep Aug 07 02:03:06 why would the other 800+ packages compile then? Aug 07 02:03:12 or is it the path perhaps? Aug 07 02:03:27 -lgcc.. thats libgcc.so or something? Aug 07 02:04:47 .. no .a ? Aug 07 02:04:49 you said you build for x86_64 target? Aug 07 02:04:55 yep Aug 07 02:05:11 i've got it building under 32bit host Aug 07 02:05:27 the actual issue seems to be "skipping incompatible [...]/libgcc.a when searching for -lgcc" Aug 07 02:05:31 cause by "-m32" Aug 07 02:05:35 frogonwheels: good, I'm glad. :) Aug 07 02:05:48 ah Aug 07 02:06:03 huh Aug 07 02:06:28 now why would _that_ happen :| Aug 07 02:08:26 try passing "-static-libgcc" in the ldflags/cflags Aug 07 02:09:55 oh wait. something strange. on the 32bit target it didn't even build grub. but still managed to make a grub image. Aug 07 02:10:05 (just a random guess, I have no real clue about this) Aug 07 02:10:07 I've missed somethng. Aug 07 02:10:32 maybe it used the host-grub? Aug 07 02:11:56 hmm.. which it would hav eto really. Aug 07 02:12:30 ok. there's some grub files it was after. will work out what's causing them not tobe copied. Aug 07 02:13:48 make target/linux-install is looking for staging_dir/host/usr/lib/grub/i386-*/stage1 - which isn't there. Aug 07 02:14:03 actually usr/lib isn't there even. Aug 07 02:14:03 package/grub honours some config option NATIVEGRUB Aug 07 02:15:10 http://pastebin.ca/1520929 Aug 07 02:15:33 maybe adding "NATIVEGRUB-x86_64:=1" does the trick Aug 07 02:53:20 xMff: how do I find out what $(HOST_OS) is set to in the makefile? **** ENDING LOGGING AT Fri Aug 07 02:59:57 2009