**** BEGIN LOGGING AT Fri Feb 06 02:59:59 2015 Feb 06 05:19:39 build #803 of avr32 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/avr32/builds/803 Feb 06 06:24:42 Akagi201: was this yours? https://forum.openwrt.org/viewtopic.php?id=52345 Feb 06 06:25:16 DonkeyHotei: call me? no Feb 06 06:26:01 == got the content Feb 06 06:26:32 the link in the post Feb 06 06:27:04 yes Feb 06 06:27:29 was that for http://www.oyewifi.com/? Feb 06 06:28:12 it's for hiwifi 1s Feb 06 06:28:45 are they different? Feb 06 06:31:10 yes, but you may have a try. Feb 06 06:31:40 and correct the errors. Feb 06 09:35:44 mmc-spi as well as sdhci? seems odd? Feb 06 09:36:10 thanks for the code at least, I've got one of the oye-0001 on the way, may have some adventures with it. Feb 06 09:56:47 build #286 of omap is complete: Failure [failed compile_4] Build details are at http://buildbot.openwrt.org:8010/builders/omap/builds/286 Feb 06 10:32:41 jow r44289 trunk/package/libs/libevent2/Makefile * libevent2: update to v2.0.22 Feb 06 11:18:54 kaloz r44290 trunk/target/ (8 files in 2 dirs) * mvebu: support building images for a single board Feb 06 11:21:34 kaloz r44291 trunk/target/linux/mvebu/image/Makefile * mvebu: add note about the move to single firmware files Feb 06 11:52:43 jow r44292 trunk/include/prereq-build.mk * include: rework git prereq test to not require 'man' Feb 06 12:34:10 jow r44293 trunk/target/imagebuilder/Makefile * imagebuilder: don't pregenerate package index, strip host binaries Feb 06 12:38:11 jow r44294 trunk/rules.mk * rules.mk: use relative paths for BASH, TAR, FIND, PATCH, PYTHON Feb 06 12:48:38 build #770 of ar71xx is complete: Failure [failed compile_2] Build details are at http://buildbot.openwrt.org:8010/builders/ar71xx/builds/770 Feb 06 12:54:12 build #201 of lantiq.xrx200 is complete: Failure [failed compile_2] Build details are at http://buildbot.openwrt.org:8010/builders/lantiq.xrx200/builds/201 Feb 06 12:59:44 build #884 of at91 is complete: Failure [failed compile_2] Build details are at http://buildbot.openwrt.org:8010/builders/at91/builds/884 Feb 06 13:05:17 build #212 of brcm47xx.mips74k is complete: Failure [failed compile_2] Build details are at http://buildbot.openwrt.org:8010/builders/brcm47xx.mips74k/builds/212 Feb 06 13:10:41 build #743 of au1000 is complete: Failure [failed compile_2] Build details are at http://buildbot.openwrt.org:8010/builders/au1000/builds/743 Feb 06 13:16:04 build #913 of brcm63xx is complete: Failure [failed compile_2] Build details are at http://buildbot.openwrt.org:8010/builders/brcm63xx/builds/913 Feb 06 13:17:05 jow r44295 trunk/include/host-build.mk * include: use SHELL, not BASH for HOST_CONFIGURE_VARS Feb 06 13:19:00 <_trine> jow_laptop, is that last commit why I have a read-only file system Feb 06 13:19:13 <_trine> on my nane which I have just flashed Feb 06 13:19:26 no Feb 06 13:19:46 <_trine> any idea why I have a read only file system Feb 06 13:20:23 <_trine> I can get to the telnet session but I cannot mount_root Feb 06 13:20:36 maybe oyur image is too big, leaving no room for jffs2 Feb 06 13:20:41 dmesg will tell Feb 06 13:21:27 build #242 of brcm47xx.legacy is complete: Failure [failed compile_2] Build details are at http://buildbot.openwrt.org:8010/builders/brcm47xx.legacy/builds/242 Feb 06 13:21:48 <_trine> I cannot see anything obvious in dmesg Feb 06 13:22:04 dmesg | grep few Feb 06 13:23:12 <_trine> nothing returned and with df -h the rootfs shows as 4.5M Feb 06 13:24:07 you said you "cannot mount_root", what does it mean? Feb 06 13:24:45 <_trine> well in a telnet session should I not be able to mount_root Feb 06 13:25:13 <_trine> because I cannot set the passwd I cannot get ssh Feb 06 13:25:15 sigh Feb 06 13:25:27 what happens if you type "mount_root" ? Feb 06 13:26:15 <_trine> mount_root: not found Feb 06 13:27:06 <_trine> I have got the ability to connect a serial line Feb 06 13:27:21 well thats likely the cause of your readonly fs too Feb 06 13:27:32 you probably removed some essential package from your build Feb 06 13:27:50 <_trine> well I could have Feb 06 13:28:04 <_trine> but which one is the question Feb 06 13:29:11 pastebin the output of ./scripts/diffconfig.sh Feb 06 13:30:47 <_trine> http://paste.debian.net/144272/ Feb 06 13:33:03 CONFIG_PACKAGE_fstools Feb 06 13:34:23 <_trine> yes I remember removing that when I was trying to make more space Feb 06 13:34:35 <_trine> mia culpa Feb 06 13:35:02 maybe consider openssl-polarssl Feb 06 13:36:37 <_trine> well I was trying to fit python in as well so I could run a python script for aircrack and I need openssl for aircrack Feb 06 13:37:29 <_trine> but I gave up and left out python and aircrack because they would not fit with openvpn as well Feb 06 13:37:52 did you try python2 instead of 3? Feb 06 13:38:29 <_trine> python2 ? Feb 06 13:38:39 <_trine> do you mean 2.7 Feb 06 13:38:47 <_trine> if so then yes Feb 06 13:38:57 <_trine> and it still did not foit Feb 06 13:39:01 <_trine> fit Feb 06 13:39:32 there's also a python3-base which appears to be way smaller Feb 06 13:39:40 python3 is the include-everything variant Feb 06 13:39:41 <_trine> it all did fit in at one point about 18 months ago but it all seems to have got bigger recently Feb 06 13:40:44 <_trine> would python3-base run a python script ,, I ask because I did try that and the script I was using errored Feb 06 13:41:15 sure Feb 06 13:41:30 your script probably errored because it required a python module not provided by base Feb 06 13:42:07 <_trine> yes I think so Feb 06 13:42:43 <_trine> I cannot remember the exact error but it did mention dependencies Feb 06 13:43:17 <_trine> is there a way I can recover this nano without the serial line Feb 06 13:43:53 sysupgrade might still worke Feb 06 13:44:12 if not you can likely write the sysupgrade image using mtd into the firmware partition Feb 06 13:48:31 <_trine> jow_laptop, is there a way to start ssh without a passwd being set Feb 06 13:48:45 not easily Feb 06 13:48:56 better is to pull the file from your pc Feb 06 13:49:02 by running scp on the router Feb 06 13:49:21 <_trine> yes ok Feb 06 13:49:25 <_trine> I will do that Feb 06 14:02:28 DonkeyHotei: I took a look at the route thing and it appears to work normally here Feb 06 14:03:11 DonkeyHotei: if I specify config route; option interface wan; option target 8.8.8.8 I get a route entry like: Feb 06 14:03:14 8.8.8.8 0.0.0.0 255.255.255.255 UH 0 0 0 eth0 Feb 06 14:03:29 if I change "wan" to another iface, the eth0 changes accordingly Feb 06 14:04:20 but try specifying an interface whose protocol is dhcp but is unplugged Feb 06 14:04:43 then the route is not installed Feb 06 14:04:50 it is here' Feb 06 14:05:00 pastebin the route -n output Feb 06 14:05:27 and the corresponding network config Feb 06 14:05:56 i will need to reboot it, will lose connectivity for a few min Feb 06 14:08:42 also what kind of vpn do you use? Feb 06 14:16:19 0.0.0.0 0.0.0.0 0.0.0.0 U 0 0 0 vpn-wan2 Feb 06 14:16:21 0.0.0.0 50.0.149.1 0.0.0.0 UG 1 0 0 br-wan Feb 06 14:16:25 50.0.149.0 0.0.0.0 255.255.255.0 U 1 0 0 br-wan Feb 06 14:16:27 192.168.83.0 0.0.0.0 255.255.255.0 U 0 0 0 br-lan Feb 06 14:16:29 208.201.249.242 50.0.149.1 255.255.255.255 UGH 1 0 0 br-wan Feb 06 14:18:47 jow_laptop: http://pastebin.com/5LV7aSgc Feb 06 14:22:23 and what kind of vpn? Feb 06 14:23:33 openvpn? cisco? Feb 06 14:45:59 bleh, upgraded laptop now has lua5.2 and of course I'm hitting stuff that doesn't work the same way :) Feb 06 14:46:41 I dislike the direction lua 5.x is heading Feb 06 14:47:02 feels a bit like PHP using \ for package separators Feb 06 14:47:23 and the new scope block syntax as replacement for get/setfenv is strange too Feb 06 14:47:38 not mentioning the ugly goto label syntax Feb 06 14:48:07 well, until I get around to replacing openwrt on tinyu devices with desktop linux on bigger android stbs, where I could use python and friends, it's lua or C :) Feb 06 14:48:29 if I could easily install lua51 on my laptop again, that would have been fine, Feb 06 14:54:22 jow_laptop: https://github.com/openwrt/packages/pull/863 Feb 06 14:58:13 DonkeyHotei: I think your route is installed by proto_add_host_dependency within openconnect.sh Feb 06 14:58:25 try commenting out that call Feb 06 14:59:20 (/lib/netifd/proto/openconnect.sh) Feb 06 15:00:01 rebooting to test, bbiab Feb 06 15:06:02 jow_laptop: seems you're right. how do i fix the call? Feb 06 15:09:07 it needs a logic change in interface_ip_add_target_route() within netifd for the iface == NULL case Feb 06 15:09:08 checking for the suffix of module shared libraries... ./configure: ./libtool: bash: bad interpreter: No such file or directory Feb 06 15:09:13 nice buildbot breakage :) Feb 06 15:09:28 more like autoconf braindeadness Feb 06 15:09:33 but already fixed Feb 06 15:11:52 DonkeyHotei: actually the logic is fine I guess, it would already skip that route if it finds a subnet route matching the host Feb 06 15:12:25 DonkeyHotei: in your unique case it fails because the host route is not present due to the base itnerface being down Feb 06 15:13:06 another possible fix is adding a 3rd argument to the proto_add_host_dependency call which specifies the ifname to bind to Feb 06 15:13:32 jow_laptop: is that implemented? Feb 06 15:13:33 but to properly solve that the openconnect.sh needs changes to accept an ifname parameter to specify the lower dev Feb 06 15:13:45 that part is no biggie Feb 06 15:13:47 similar to how pppd has an ifname that specifies the l2 iface Feb 06 15:14:24 <_trine> thanks jow_laptop its recovered now Feb 06 15:14:33 does proto_add_host_dependency currently accept a 3rd arg? Feb 06 15:14:40 yes Feb 06 15:14:52 local interface="$1" Feb 06 15:14:52 local host="$2" Feb 06 15:14:52 local ifname="$3" Feb 06 15:15:08 $1 is the logical name of the parent iface it belongs to Feb 06 15:15:21 $2 is the ip address of the target you depend on (typically the vpn server) Feb 06 15:15:24 is it the interface name from network config, or from ifconfig? Feb 06 15:15:30 $3 is the netdev you want to bind to Feb 06 15:15:37 which is currently unset Feb 06 15:17:58 both $1 and $3 are logical names from /e/c/network Feb 06 15:18:04 ok Feb 06 15:19:03 as a simple test you can try hardcoding "wan2_phys" as 3rd param Feb 06 15:19:14 and remove the route definition from /etc/config/network Feb 06 15:21:36 rebooting again Feb 06 15:26:20 build #876 of orion is complete: Failure [failed compile_1] Build details are at http://buildbot.openwrt.org:8010/builders/orion/builds/876 Feb 06 15:28:04 what was the fix/workaround for this binary grep thing? just trying to rebuild an older AA tree. Feb 06 15:28:22 something with LC_ALL afair Feb 06 15:28:57 thanks, that works. Feb 06 15:29:00 $ LC_ALL=C ./scripts/feeds update -a Feb 06 15:48:52 jow_laptop: now that that's solved, any idea why 6rd breaks if wan2_phys is up on boot? Feb 06 16:15:50 build #61 of ramips.mt7621 is complete: Failure [failed compile_4] Build details are at http://buildbot.openwrt.org:8010/builders/ramips.mt7621/builds/61 Feb 06 16:19:45 build #329 of pxa is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/pxa/builds/329 Feb 06 16:27:09 DonkeyHotei: nope Feb 06 16:27:18 DonkeyHotei: I suggest to talk to cyrusff Feb 06 16:27:31 ok Feb 06 17:41:30 jow r44296 trunk/include/prereq-build.mk * include: fix prereq tests for perl and file (#18911) Feb 06 18:41:58 Hmm? Feb 06 18:42:41 cyrusff: DonkeyHotei has some problem with 6rd being defunct after a changing wan connection (?) Feb 06 18:42:49 and I'm not into the 6rd implementation Feb 06 18:43:21 Well guess i need some details Feb 06 18:43:34 I guess so too :) Feb 06 18:44:15 cyrusff: http://pastebin.com/5LV7aSgc Feb 06 18:45:01 if wan2_phys is plugged in before boot, 6rd fails until ifdown wan2_phys Feb 06 18:45:29 ifup wan2_phys after that does not make it fail Feb 06 18:47:33 You can remove option ifname for 6rd Feb 06 18:49:26 should that fix it? Feb 06 18:50:51 Maybe Feb 06 18:50:55 Don't know Feb 06 18:51:07 Wort a try Feb 06 19:35:22 openwrt supports building on systems other than linux? Feb 06 19:40:33 {Nico}: thanks for the review of my PR, but I'm more than a little confused by the OSNAME bit. Openwrt buildroot works on systems other than linux? Feb 06 19:41:06 it can be built on a mac Feb 06 19:41:32 groz: and buildroot doesn't intercept calls to uname so uname would show it's a mac? Feb 06 19:42:00 I dont remember all the details, just remember going thru the process of getting things built on a mac some time back Feb 06 19:42:17 involved setting up a case sensative file system and getting all the macports stuff installed Feb 06 19:42:29 then a couple minor other things, and off it went Feb 06 19:42:53 groz: yeah, a PR of mine just rejected because "this break when not run on linux" and I didn't realize that people were actualy using openwrt buildroot on mac. Feb 06 19:42:56 meh, easy enough to fix Feb 06 19:48:04 <- dnsmasq patch. Please help me make it acceptable. Feb 06 19:49:10 ManoftheSea: pastebin Feb 06 20:04:18 Zero_Chaos: yes, it also supports building on OS X Feb 06 20:04:28 Zero_Chaos: and FreeBSD (in theory) Feb 06 20:05:02 jow_laptop: cool, I removed something thinking "well this is a pointless override" but now I've learned better. thanks Feb 06 20:05:20 we don't intercept uname Feb 06 20:05:26 since its required in some cases Feb 06 20:05:31 for host builds Feb 06 20:05:41 jow_laptop: yeah, no problems. easy enough to unbreak my "fix" Feb 06 20:12:33 PKG_SOURCE:=$(PKG_NAME)-$(PKG_VERSION).tar.gz Feb 06 20:12:33 PKG_SOURCE_URL:=http://download.aircrack-ng.org/$(PKG_SOURCE) Feb 06 20:12:42 this is clearly wrong, but I'm not sure how. anyone help? Feb 06 20:12:48 _trine maybe? Feb 06 20:14:24 it is trying to download http://download.aircrack-ng.org/aircrack-ng-1.2-rc1.tar.gz/aircrack-ng-1.2-rc1.tar.gz Feb 06 20:14:28 which is... weird? Feb 06 20:15:22 Zero_Chaos: remove $(PKG_SOURCE) Feb 06 20:15:30 jow_laptop: that much is automatic? Feb 06 20:15:46 Zero_Chaos: openwrt automatically appends /$(PKG_SOURCE) to $(PKG_SOURCE_URL) Feb 06 20:16:12 jow_laptop: excellent, thanks Feb 06 20:16:14 Zero_Chaos: I suggest to read include/package-defaults.mk, this will you what stuff is default already Feb 06 20:16:18 bbl Feb 06 20:16:27 jow_laptop: thanks, I'll do that Feb 06 20:23:18 er, why do I need the openssl-dev to make clean? Feb 06 20:34:57 anyone have a minute to review my updated PR? I think I got it right now: https://github.com/openwrt/packages/pull/866 Feb 06 20:41:33 <__trine> Zero_Chaos, it would be nice if you could make the aircrack package configurable so only the parts a user wanted would be compiled Feb 06 20:41:54 <__trine> I have to do this Feb 06 20:41:57 <__trine> $(CP) $(PKG_INSTALL_DIR)/usr/bin/aircrack-ng $(1)/usr/bin/ Feb 06 20:41:57 <__trine> $(INSTALL_DIR) $(1)/usr/sbin Feb 06 20:41:57 <__trine> $(CP) $(PKG_INSTALL_DIR)/usr/sbin/{aireplay-ng,airmon-ng,airodump-ng} $(1)/usr/sbin/ Feb 06 20:41:57 <__trine> endef Feb 06 20:42:17 <__trine> so I only have the bits I want to keep the code as small as possible Feb 06 20:42:21 __trine: lol, do that, and it will be broken on openwrt if you try to use airmon-ng most likely Feb 06 20:42:34 <__trine> no it works Feb 06 20:42:53 <__trine> I have been using it like this for a long time now Feb 06 20:42:53 __trine: also I agree entirely, but I have a lot of packages to get in, and then I'll worry about learning how to make things configurable Feb 06 20:43:04 __trine: some of it will work, some of it won't Feb 06 20:43:17 <__trine> the bits I want do work Feb 06 20:43:40 <__trine> but things get very tight for space in 8M flash Feb 06 20:43:59 <__trine> just a thought anyway Feb 06 20:44:01 __trine: well, airmon-zc replaces airmon-ng in the next release, so maybe you should be testing it ;-)_ Feb 06 20:44:17 __trine: I agree, I want it configurable, but that's beyond my skill level at present. Feb 06 20:45:10 <__trine> it needs to be made selectable in menuconfig, but all will come with time Feb 06 20:46:39 <__trine> if you have a look at the FreeSwich package for example every part is selectable Feb 06 20:46:49 __trine: honestly I'd like it selectable for nearly every executable, but as I have no clue how to do that right now, it will wait :-) Feb 06 20:46:50 <__trine> FreeSwitch* Feb 06 20:47:03 <__trine> yes of course Feb 06 20:47:27 <__trine> its just an idea Feb 06 21:00:06 <__trine> Zero_Chaos, actually I rarely use airmon-ng as mostly I configure the interfaces in openwrt myself Feb 06 21:00:49 <__trine> its largely unnecessary Feb 06 21:01:11 __trine: I would imagine a lot of users are uncomfortable with using openwrt like that. Your way is better, but not sure people understand it Feb 06 21:02:15 I never use uci. I don't know how. Feb 06 21:02:47 trine_: well the freeswitch package fails at the buildbots somehow - so thats probably a bad example Feb 06 21:03:15 <__trine> plntyk, now now Feb 06 21:03:24 is there a way other than uci for configuration ?? Feb 06 21:03:59 <__trine> something has upset Mazilo so he seems not to be doing anything on FreeSwitch now Feb 06 21:04:39 <__trine> iw phy phy0 interface add mon0 type monitor flags otherbss active Feb 06 21:10:21 <__trine> plntyk, Mazilo does not have access to github as far as I know so he could not do anything with FS even if he wanted to Feb 06 21:12:11 Makefile:94: *** Download/default is missing the SUBDIR field.. Stop. Feb 06 21:12:11 but i have a PKG_SUBDIR:=foo line Feb 06 21:13:15 how do i fix? Feb 06 21:30:19 nice, a coloured shell for testing: https://twitter.com/rundfreifunk/status/563811676127248387 Feb 06 21:33:41 <__trine> rundfreifunk, it's bad for old eyes Feb 06 21:34:11 <__trine> do you have one in Braille Feb 06 21:34:14 <__trine> :) Feb 06 21:36:41 oh, i'm an idiot, it's PKG_SOURCE_SUBDIR Feb 06 21:44:52 trine, yes.. my eyes also aching. Feb 06 22:13:02 git.openwrt.org appears to be down Feb 06 22:20:39 weird, trouble updating luci feed from git://git.openwrt.org/projects/luci.git Feb 06 22:20:43 fatal: unable to connect to git.openwrt.org: Feb 06 22:20:47 git.openwrt.org[0: 217.115.15.20]: errno=Connection timed out Feb 06 22:21:12 and yet: Feb 06 22:21:15 Updating feed 'oldpackages' from 'git://git.openwrt.org/packages.git' ... Feb 06 22:21:15 Already up-to-date. Feb 07 02:29:55 build #804 of avr32 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/avr32/builds/804 **** ENDING LOGGING AT Sat Feb 07 02:59:59 2015