**** BEGIN LOGGING AT Tue Jun 24 02:59:59 2014 Jun 24 04:46:58 build #99 of pxa is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/pxa/builds/99 Jun 24 05:10:43 build #1 of brcm47xx.mips74k is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/brcm47xx.mips74k/builds/1 Jun 24 05:11:57 build #1 of brcm47xx.legacy is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/brcm47xx.legacy/builds/1 Jun 24 07:27:47 build #193 of mvebu is complete: Failure [failed compile_8] Build details are at http://buildbot.openwrt.org:8010/builders/mvebu/builds/193 Jun 24 08:17:38 build #100 of x86_64 is complete: Failure [failed compile_1] Build details are at http://buildbot.openwrt.org:8010/builders/x86_64/builds/100 Jun 24 08:51:25 build #99 of adm8668 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/adm8668/builds/99 Jun 24 08:58:30 build #639 of lantiq is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/lantiq/builds/639 Jun 24 09:31:50 build #503 of adm5120 is complete: Failure [failed shell_15] Build details are at http://buildbot.openwrt.org:8010/builders/adm5120/builds/503 Jun 24 10:14:56 zajec: I think your RFC patch to simply drop the network config is too intrusive Jun 24 10:15:52 zajec: you should consider an aproach like that: https://dev.openwrt.org/browser/trunk/target/linux/ar71xx/base-files/etc/uci-defaults/03_network-vlan-migration Jun 24 10:16:32 the advantage of this approach is that it only touches network configs which are more or less default Jun 24 10:27:06 will see Jun 24 10:29:46 well, that is simple Jun 24 10:30:01 simple VLAN change (however there should be check for VID 1 being already in use) Jun 24 10:30:07 true Jun 24 10:30:13 so even this simple case is not handled 100% safe Jun 24 10:30:31 I have no idea how we could nicely migrate old Broadcom switch config :| Jun 24 10:30:43 it extremely complicated Jun 24 10:30:51 thats the reason why it is not done yet Jun 24 10:30:52 we should change interface names, VIDs, ports Jun 24 10:33:09 jow_laptop: i think you could append "exit 0" to this ar71xx uci default Jun 24 10:33:27 (and I probably should make my script a uci-default with exit 0 as well) Jun 24 11:19:29 is there "nice" / "right" way of defaulting the host name to something based on the mac address? I've traditionally had a boot script that runs late and reconfigures wifi and networking and hostnames and so on, and eventually reboots to make everything restart with new configs, but I'm trying to put it into base-files instead, avoid the reboot and service restarting Jun 24 11:20:38 /etc/uci-defaults/99_calculate-hostname Jun 24 11:22:07 that's too late to get into the /etc/config/wireless SSID that mac80211/files/lib/wifi/mac80211.sh hardcodes as just "OpenWrt" Jun 24 11:22:44 uci set wireless.@wifi-iface[0].ssid="$(magic)" Jun 24 11:22:48 uci commit wireless Jun 24 11:27:22 will /etc/init.d/network restart then fix that properly? Jun 24 11:27:37 uci-defaults run before the first init script is launched Jun 24 11:28:06 will ifconfig still work to see eth0 at that point? Jun 24 11:28:12 yes Jun 24 11:28:22 I suggest not parsing ifconfig though Jun 24 11:28:53 # read mac < /sys/class/net/eth0/address Jun 24 11:28:54 well, it's always felt ugly, but it works well enough, what's a better way? Jun 24 11:28:57 # echo $mac Jun 24 11:29:02 c2:3f:0e:7a:a0:94 Jun 24 11:29:34 heh, always finding more useful stuff in /sys Jun 24 11:30:01 /sys/class/net is great in general, especially if you want to checxk for the existance of ifaces Jun 24 11:30:20 like test -e /sys/class/net/eth1 && ifname=eth1 || ifname=eth0 Jun 24 11:30:52 this actually means I can do most of my customization in a package that provides a uci-defaults file, rather than hacking on base-files and mac80211 too. Jun 24 11:31:01 which certainly seems like a safer long term plan Jun 24 11:31:59 yes Jun 24 11:32:21 in general it probably makes sense to provide the ability to implement hooks for ssid and mac generation Jun 24 11:33:02 like if a procedure wifi_generate_ssid() exists it will be used to create ssids, if its undeclared "OpenWrt" is used Jun 24 11:33:24 but the question is whether implementing such subsets of functionality makes sense if you can do anything in uci-defaults Jun 24 11:33:42 which, after all, is intended to allow preparing configs Jun 24 11:33:57 mostly just discoverability of how to use features like this I guess. Jun 24 11:59:44 Kaloz, acoul: Is mpc83xx still been maintained? The builder for mpc83xx stops since months on the same error. I've posted according to the web and forum comments a (very small) github pull request (https://github.com/mirrors/openwrt/pull/12) in the hope that the builder could run again. Is there anyone who can check, accept, comment or improve the pull request? Jun 24 12:00:22 kako: the github mirror is inofficial and not maintained by openwrt Jun 24 12:00:31 kako: I think nobody even looks there for pull requests Jun 24 12:00:46 ok, where is the right place to make a pull request? Jun 24 12:00:59 build #461 of octeon is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/octeon/builds/461 Jun 24 12:01:10 nowhere, openwrt core is maintained in svn Jun 24 12:01:19 preferred way is git send-email to the development list Jun 24 12:01:54 jow_laptop: I've already send an email to Kaloz and acoul, but no reaction. Jun 24 12:01:58 important is to include a Signed-Off Jun 24 12:02:04 ok Jun 24 12:03:10 hm, I send the email directly using my email programm. Should I send the email using git? Jun 24 12:06:20 kako: depends on your mailer Jun 24 12:06:31 kako: if it attaches patch nicely, no need to Jun 24 12:07:08 kako: easy test: save the e-mail you sent as .patch file (plain text file) and try toa pply it with "git am" Jun 24 12:07:14 if it works, it's fine :) Jun 24 12:07:36 thanks, will try it again. Jun 24 12:07:42 mailers like to malform white spaces, lines (by adding line breaks), etc. Jun 24 12:09:04 (mailers are also pretty good at saving attachments....) Jun 24 12:28:57 kako: patch inline >>>> patch attached as a file. Jun 24 12:29:31 luka r41309 trunk/ (6 files in 4 dirs) * kernel: update 3.14 to 3.14.8 Jun 24 12:42:51 <__trine> russell--, was it you a few weeks ago who was having problems connecting with VPN and openssl but polarssl was working ? Jun 24 12:54:38 KanjiMonster: In https://dev.openwrt.org/wiki/SubmittingPatches ist written to send "only plain text", but if I understand you correct I should send the patch as attachement (like "file.patch"). What is now the best way, inline or attachment? Jun 24 12:55:30 kako: I mean inline patch "infinity better/greater than" patch as attachment ;) Jun 24 13:00:01 KanjiMonster: Ok, get it :-) Jun 24 13:00:55 but it's also harder to get right if you are not using git send-email, the majority of email clients break inline patches Jun 24 13:08:04 KanjiMonster: btw. in abscence of patchwork, how do i sanely receive patches sent by git send-email? ;) Jun 24 13:08:45 cyrusff: use an email client that allows you exporting/saving as mbox Jun 24 13:09:09 git am is capable of ignoring all unneeded/unknown email-headers Jun 24 13:09:21 yeah didn't really work well with thunderbird last time Jun 24 13:09:43 for wahtever reason Jun 24 13:09:57 strange Jun 24 13:10:27 KanjiMonster: ok, never used git send-email before, but will do it now :-) Jun 24 13:10:45 wonder whats the reasoning against attachments here Jun 24 13:10:50 it works for me even with the gmail web interface (-> show orginal) Jun 24 13:11:05 cyrusff: it's much easier to do code review if the patch is inline Jun 24 13:12:48 with the patch inline i can just hit reply(-all), with it as a patch I first need to save the .patch, open it in a text editor, then copy the contents into the reply and add quotemarks manually Jun 24 13:13:23 i see Jun 24 13:13:31 oh well, don't have to like it Jun 24 13:13:43 as long as patchwork runs i don't have to deal with it Jun 24 13:15:26 cyrusff: I wouldn't wonder if there are scripts for extracting patches from an inbox through imap/pop3 Jun 24 13:16:26 that would be pretty nice actually Jun 24 13:16:33 git receive-mail Jun 24 13:16:34 or so Jun 24 13:16:39 heh, yeah Jun 24 13:16:54 no, you're all meant to live in the dark ages because some kernel dudes insist on doing evyerhting in plain text :) Jun 24 13:17:07 so we use email for humans to send messages for machines to read :) Jun 24 13:18:20 but such is life, we just move with it, Jun 24 13:18:30 and git send-email means people don't complain (much) Jun 24 13:18:49 cyrusff: assuming you use imap, a sensible workflow could be tag emails in $favourite_email_client, then the script will extract patches from all tagged emails Jun 24 13:19:22 yeah just throw them in imap-folder Jun 24 13:32:40 KanjiMonster, jow_laptop , zajec: finally, the first patch sent with git send-email. Curious if I did everything right :-) Jun 24 13:33:32 kako: the usual advice is to send it to yourself, then save it from your email client and try to apply it Jun 24 13:36:39 KanjiMonster: I confess, I've sended the mail several times to my email before I finale call "git send-email --to ... --cc ..." Jun 24 13:38:14 hm, I don't see any new emails there Jun 24 13:38:25 are you subscribed to openwrt-devel? Jun 24 13:55:00 I've subscribed and confirmed today. Maybe the first email was to fast... Jun 24 13:55:33 should I recend it or wait? Jun 24 13:55:52 s/recend/resend/ Jun 24 14:01:39 if your email is not in the mail archive by now you should probably resend Jun 24 14:05:29 build #100 of mpc83xx is complete: Failure [failed compile_8] Build details are at http://buildbot.openwrt.org:8010/builders/mpc83xx/builds/100 Jun 24 14:21:45 kako: if you are using git send-email, make sure it uses the same from-address you used for registering to the ml Jun 24 15:22:16 KanjiMonster: the error was not my email adresse, the error was trying to send the email from my develop-system. It is configurated to redirect all emails to me (so that my programms in development do not send outgoing emails by mistake) :-( Jun 24 15:22:16 Not I see my email in the archive Jun 24 15:24:14 s/Not/Now/ Jun 24 15:28:08 kako: ah, nice one ;D Jun 24 15:39:57 jow_laptop: with using uci-defaults like this, if someone has customized their settings, and sysupgrade runs (choosing to _keep_ settings) will the uci-defaults scripts run again and trample over them? Jun 24 15:40:09 or only if they choose to not keep settings? Jun 24 16:56:32 __trine: yes, on brcm47xx (wgt634u was the specific device) Jun 24 16:57:08 <__trine> hmm I seem to be having openvpn problems on my wndr-3700 Jun 24 16:59:06 <__trine> and kernel bug problems on my nanostation Jun 24 16:59:15 <__trine> maybe its all related Jun 24 17:09:31 __trine: what's the problem? Jun 24 17:10:20 <__trine> zajec, I am not sure other than I cannot initiate a VPN connection Jun 24 17:10:36 <__trine> I feel like I am sinking into the mud right now Jun 24 17:11:09 oh, so nothing brcm47xx specific Jun 24 17:11:11 ok Jun 24 17:12:12 <__trine> no not broadcom Jun 24 17:21:11 anyone seen this squashfs problem? Only see it if I'm sysupgrading via serial console, and it doesn't _seem_ to have any sideaffects, it boots up fine with the new image after Jun 24 17:35:18 fwiw, openvpn-openssl works fine for me on ar71xx and x86 as of recently Jun 24 17:36:34 is there a neat trick to wait for jffs to be finished mounting and be properly writable? if I write in these uci-defaults I don't always see evyerthing written, and the kernel says "unlocking the mtd device" _way_ later Jun 24 17:36:39 karlp: uci-defaults are run on every new flash, regardless if settings were kept or not (some are even there to fixup old network configs), and have extra checks to not trample over existing configs Jun 24 17:38:19 e.g. 02_network won't do anything if there already is a /etc/config/network (one of the first lines is a [ -e /etc/config/network ] && exit 0 or something to that effect) Jun 24 17:39:18 where's 02_network? Jun 24 17:39:36 found it, nvm, not in AA Jun 24 17:40:37 just called network in AA, Jun 24 17:40:39 so there is no magic value you can check if to know if it was an image with settings kept or not Jun 24 17:41:11 would be nice :) Jun 24 17:41:29 build #605 of rb532 is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/rb532/builds/605 Jun 24 17:41:31 build #605 of ppc44x is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/ppc44x/builds/605 Jun 24 17:41:51 I guess I can look for a flag of my own. Jun 24 17:41:55 build #574 of sibyte is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/sibyte/builds/574 Jun 24 17:43:34 is my jffs problem possibly from calling uci commit in my uci-defaults files? Jun 24 17:43:49 i see it's also don in lib/functions.sh after running all the uci-defaults files, Jun 24 17:44:13 or am I just finding that something else is conflicting/overwriting the same things I'm setting? Jun 24 17:44:45 I can write a log to /tmp showing that I ran my commands, but the luci theme for instance isn't set Jun 24 17:49:54 <__trine> I think it must be some kind of FW mis configuration as I am not getting any vpn packets back Jun 24 17:56:22 karlp: quite likely not. As long as jffs2 is not ready a tmpfs overlay is used, and only if jffs2 is ready the changes are copied over to jffs2. Jun 24 17:57:28 so whateever I write will get there eventually, it isn't just disappearing when jffs2 finally finishes? Jun 24 17:57:59 or the opposite? and there's a window where I can "write" and it won't actually write anywhere useful? Jun 24 18:12:09 has there been any effort to port to realtek SoC's ? Jun 24 18:12:33 RTL8651B Jun 24 18:14:36 blogic r41310 trunk/target/linux/lantiq/image/ubinize-overlay.cfg * lantiq: ubinize autoresize flag for rootfs_data Jun 24 18:14:40 blogic r41311 trunk/include/image.mk * image.mk: introduce Image/Build/UbinizeImage make target Jun 24 18:14:44 blogic r41312 trunk/target/linux/ lantiq/xway/target.mk lantiq/image/Makefile * lantiq: remove deprecated jffs2_nand configuration Jun 24 18:14:49 blogic r41313 trunk/package/boot/uboot-envtools/patches/300-support-env-in-ubivol-chardev.patch * uboot-envtools: fix build if ubi volume support is disabled Jun 24 18:14:53 blogic r41314 trunk/tools/firmware-utils/src/mkdir615h1.c * firmware-utils/mkdir615h1: fix magic XOR calculation Jun 24 18:14:57 blogic r41315 trunk/package/system/fstools/Makefile * fstools: fix factory reset on ubifs overlay Jun 24 18:15:02 blogic r41316 trunk/target/linux/ atheros/patches-3.10/230-3_10-updates.patch atheros/patches-3.10/110-ar2313_ethernet.patch * atheros[ar231x-eth]: merge patch Jun 24 18:15:06 blogic r41317 trunk/target/linux/atheros/patches-3.10/100-board.patch * atheros[ar2315]: don't detect flash size during board initialization Jun 24 18:15:10 blogic r41318 trunk/target/linux/atheros/patches-3.10/120-spiflash.patch * atheros[ar2315-spiflash]: print detected flash size Jun 24 18:15:15 blogic r41319 trunk/target/linux/atheros/patches-3.10/105-ar2315_pci.patch * atheros[ar231x-pci]: do not touch iomem_resource ranges Jun 24 18:15:19 blogic r41320 trunk/target/linux/ atheros/patches-3.10/100-board.patch atheros/patches-3.10/120-spiflash.patch * atheros[ar2315-spiflash]: pass I/O mem regions via resources Jun 24 18:15:23 blogic r41321 trunk/target/linux/atheros/patches-3.10/120-spiflash.patch * atheros[ar2315-spiflash]: make spiflash header local Jun 24 18:15:28 blogic r41322 trunk/target/linux/atheros/patches-3.10/120-spiflash.patch * atheros[ar2315-spiflash]: I/O fixes Jun 24 18:15:32 blogic r41323 trunk/target/linux/atheros/patches-3.10/120-spiflash.patch * atheros[ar2315-spiflash]: make local variables and functions static Jun 24 18:15:36 blogic r41324 trunk/target/linux/ atheros/patches-3.10/120-spiflash.patch atheros/patches-3.10/100-board.patch * atheros[ar2315-spiflash]: update driver id Jun 24 18:15:40 blogic r41325 trunk/target/linux/ atheros/patches-3.10/120-spiflash.patch atheros/patches-3.10/100-board.patch * atheros[ar2315-spiflash]: remove duplicated definitions Jun 24 18:15:44 blogic r41326 trunk/ include/image.mk target/linux/ar71xx/image/Makefile * SysupgradeNAND: include rootfs-type in image name Jun 24 18:15:48 blogic r41327 trunk/package/system/procd/files/nand.sh * procd: fix nand.sh kernel_length typo Jun 24 18:15:53 blogic r41328 trunk/package/libs/libubox/Makefile * libubox: update to latest git head Jun 24 18:15:58 blogic r41329 trunk/config/Config-images.in * config/Config-images.in: enable zlib as the default ubi compression Jun 24 18:16:02 blogic r41330 trunk/target/linux/mpc83xx/Makefile * mpc83xx: add missing kernelnames into Makefile Jun 24 18:16:06 blogic r41331 trunk/target/linux/ramips/patches-3.10/0118-NET-MIPS-add-ralink-SoC-ethernet-driver.patch * ralink: Update MT7530 switch driver. Jun 24 18:16:10 blogic r41332 trunk/target/linux/ar71xx/image/ubinize-wndr4300.ini * ar71xx: remove old jffs2 marker from image generation on nand targets Jun 24 18:16:15 blogic r41333 trunk/target/linux/ar71xx/image/Makefile * ar71xx: fix wndr3700v4 image header Jun 24 18:20:42 luka r41334 trunk/package/base-files/files/etc/init.d/led * base-files: update copyright headers Jun 24 18:20:44 luka r41335 trunk/target/linux/ kirkwood/base-files/lib kirkwood/base-files/lib/functions/kirkwood.sh kirkwood/base-files/lib/functions * kirkwood: base-files: add script to detect board name and populate /tmp/sysinfo Jun 24 18:20:45 luka r41336 trunk/target/linux/ kirkwood/base-files/etc/uci-defaults/01_leds kirkwood/base-files/etc/uci-defaults/02_network * kirkwood: base-files: use /lib/functions/kirkwood.sh Jun 24 18:21:20 build #585 of avr32 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/avr32/builds/585 Jun 24 18:40:32 build #645 of cobalt is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/cobalt/builds/645 Jun 24 18:58:31 <__trine> whats all this about Jun 24 18:58:34 <__trine> http://pastebin.com/ekAAH9QN Jun 24 19:07:02 build #651 of orion is complete: Failure [failed compile_5 shell_15] Build details are at http://buildbot.openwrt.org:8010/builders/orion/builds/651 Jun 24 19:09:41 build #512 of iop32x is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/iop32x/builds/512 Jun 24 19:13:11 build #616 of uml is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/uml/builds/616 Jun 24 19:41:38 build #540 of au1000 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/au1000/builds/540 Jun 24 22:43:04 build #500 of mcs814x is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/mcs814x/builds/500 **** ENDING LOGGING AT Wed Jun 25 02:59:58 2014