**** BEGIN LOGGING AT Sun Oct 05 03:00:01 2014 Oct 05 04:16:33 build #195 of adm8668 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/adm8668/builds/195 Oct 05 07:18:54 build #707 of ppc44x is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/ppc44x/builds/707 Oct 05 07:59:01 karlp: "Package libubus is missing dependencies for the following libraries: Oct 05 07:59:01 libubox.so" Happened after a feeds update, and goes away if you 'make clean' Oct 05 08:47:50 movi: another option is to look here: https://dev.openwrt.org/browser/branches Oct 05 08:49:05 russell--: i already figures it out, thanks anyway :) Oct 05 09:01:59 <_trine> reaver isn't compiling anymore http://paste.debian.net/124543/ Oct 05 09:03:07 <_trine> gcc-4.8-linaro I might try 4.6 Oct 05 09:49:37 _trine: it is in oldpackage, afaict Oct 05 09:49:43 oldpackages* Oct 05 10:26:44 nbd r42758 trunk/package/network/utils/iwinfo/Makefile * iwinfo: enable nl80211 support based on cfg80211, not mac80211 Oct 05 10:26:54 nbd r42759 trunk/package/network/utils/iwinfo/ (5 files in 3 dirs) * iwinfo: fix handling of accessing nl80211 interfaces via radio* Oct 05 10:26:59 nbd r42760 trunk/package/network/utils/iwinfo/ src/api/broadcom.h src/iwinfo_wl.c * iwinfo: detect HT PHY for broadcom-wl Oct 05 10:27:06 nbd r42761 trunk/package/network/utils/iwinfo/ src/api/broadcom.h src/iwinfo_wl.c * iwinfo: detect AC PHY for broadcom-wl Oct 05 10:27:16 nbd r42762 trunk/package/network/services/hostapd/patches/001-fix_pmksa_cache.patch * hostapd: merge an upstream patch for pmksa cache Oct 05 10:40:41 nbd r42763 branches/barrier_breaker/package/network/services/hostapd/patches/001-fix_pmksa_cache.patch * hostapd: merge an upstream patch for pmksa cache Oct 05 10:48:25 <_trine> russell--, yes it is what you call an 'old package' Oct 05 10:49:22 <_trine> however nothing has changed in 'that' package Oct 05 10:51:53 you can probably get rid of the error by putting PKG_USE_MIPS16:=0 into its makefile Oct 05 10:52:13 <_trine> nbd, thanks I will try that Oct 05 10:52:45 <_trine> but the error seemed to be something to do with gcc 4.8 Oct 05 10:52:57 <_trine> anyway I will try what you suggest Oct 05 10:54:36 <_trine> nbd, I don’t use reaver now anyway and it was only in my compile as an artefact from previous compilations because reaver no longer works with mon0 Oct 05 10:56:44 nbd r42764 trunk/package/network/services/hostapd/files/netifd.sh * hostapd: allow using iapp for any encryption type (fixes #18022) Oct 05 11:02:37 nbd r42765 trunk/package/network/services/relayd/Makefile Oct 05 11:02:37 relayd: update to the latest version, fixes a build error with the new gcc (#18010) Oct 05 11:03:01 <_trine> nbd, you were correct it did compile thanks Oct 05 11:51:32 nbd r42766 trunk/package/kernel/mac80211/Makefile * mac80211_hwsim: select 802.11n support for hostapd Oct 05 12:52:50 russell--: how can you use the browser to find what svn rev BB final was? the barrier breaker branch has commits only 2 hours old. and https://dev.openwrt.org/log/branches/barrier_breaker doesn't seem to have any commit tagging anything, nor does https://dev.openwrt.org/log/tags have any tags for BB at this point (presumably then the final images aren't final?) Oct 05 13:39:09 is there any specific reason why the dnsmasq init script only register the lan ipv4 to the local hostname? Oct 05 13:46:39 karlp: x86 BB image is built from BARRIER BREAKER (14.07, r42625) revision Oct 05 13:47:16 karlp: the current git clone of BB says => ./scripts/getver.sh => r42763 Oct 05 13:49:32 slachta: that's what I was getting at. current code is not what the "final" images are labelled, so I didn't know how russel was finding 42625 from the svn browser at dev.openwrt.org. Oct 05 13:52:28 karlp: I think I have missed the discussion before Oct 05 13:54:15 someone had asked how to find the svn rev of the BB final releases Oct 05 13:54:26 I suggested installing one and looking at /etc/openwrt_release Oct 05 13:54:45 russle said you could find that on https://dev.openwrt.org/browser/branches Oct 05 13:54:56 I asked how Oct 05 13:55:47 you cannot find it here, since it does not depend on the svn revision Oct 05 13:57:47 and the commits does not explicitly say "We are bumping to newer release" Oct 05 13:58:53 ok, let me reword differently, if svn is still the official master, could we maybe please tag the svn repo? (as was done for AA releases) Oct 05 14:00:44 karlp: I think its worth the shot. Oct 05 14:01:15 branches for the continuous changes and tag for the latest final stable Oct 05 14:01:29 that would be pretty standard expectations yes :) Oct 05 14:01:50 there's often questions of what exactly was in which binaries Oct 05 14:02:16 * swalker is assuming there isn't a checklist for doing releases Oct 05 14:18:02 build #753 of orion is complete: Failure [failed compile_8] Build details are at http://buildbot.openwrt.org:8010/builders/orion/builds/753 Oct 05 14:19:52 build #676 of sibyte is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/sibyte/builds/676 Oct 05 14:49:17 nbd: running into missing headers (sanitizer_platform_limits_posix.cc:76:27: fatal error: netrom/netrom.h: No such file or directory) trying gcc-4.9-linaro on x86_64 Oct 05 14:52:26 context? Oct 05 14:53:58 nbd: I'll try http://patchwork.openwrt.org/patch/5734/ in a bit Oct 05 15:00:36 nbd r42767 trunk/toolchain/ gcc/minimal/Makefile gcc/final/Makefile * gcc: gcc 4.9.x disable libsanitize Oct 05 15:00:48 nbd r42768 trunk/package/kernel/mac80211/Makefile * ath10k: update firmware to the latest version Oct 05 15:01:08 oh, overlooked that i had queued this up already when preparing the ath10k change Oct 05 15:01:18 either way, this should probably fix it, if it's the same bug Oct 05 15:17:26 build #688 of avr32 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/avr32/builds/688 Oct 05 15:22:26 nbd r42769 trunk/toolchain/ gcc/patches/4.9-linaro/930-fix-mips-noexecstack.patch gcc/patches/4.8-linaro/930-fix-mips-noexecstack.patch Oct 05 15:22:26 Fix broken implementation of NOEXECSTACK flag for MIPS architectures for gcc-linaro-4.8 Oct 05 15:33:43 nbd: it does Oct 05 16:42:22 nbd r42770 trunk/ scripts/metadata.pl include/package-defaults.mk include/package-dumpinfo.mk scripts/metadata.pm Oct 05 16:42:22 build: add support for declaring package CONFLICTS which only affect selecting built-in packages Oct 05 16:42:27 nbd r42771 trunk/include/package-ipkg.mk * build: propagate package conflicts to opkg metadata Oct 05 16:42:39 nbd r42772 trunk/package/network/services/hostapd/Makefile * hostapd: add conflicts with wpad(-mini) to hostapd and wpa_supplicant Oct 05 16:59:05 nbd r42773 trunk/include/kernel-defaults.mk * build: make sure modules are modposted after initramfs image is built Oct 05 17:14:23 nbd r42774 trunk/package/network/utils/iproute2/Makefile * iproute2: add conflict for ip with ip-full Oct 05 17:23:23 build #779 of at91 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/at91/builds/779 Oct 05 17:26:14 build #610 of iop32x is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/iop32x/builds/610 Oct 05 17:33:25 wigyori r42775 trunk/target/linux/sunxi/image/Makefile * sunxi: re-add rootfs build when no profile is set, fixes #17980, thanks. Oct 05 18:19:58 updated openwrt/upstream, https://home.comcast.net/~sdwalker/uscan/index.html Oct 05 18:26:49 swalker: svn revision for the package chan-sccp-b is wrong on the page from uscan tool Oct 05 18:27:34 current revision is 5754 (see PKG_SOURCE_URL:=http://svn.code.sf.net/p/chan-sccp-b/code/tags/openwrt ) Oct 05 18:30:09 slachta: svn --non-interactive --trust-server-cert info http://svn.code.sf.net/p/chan-sccp-b/code/tags/openwrt; I'm using the Last Changed Rev rather than Revision Oct 05 18:30:29 which I'd argue makes more sense Oct 05 18:32:04 swalker: seems legit to me too. Oct 05 18:37:12 swalker: shame on me, I was updating the package even if the sources were not changed for this tag. Oct 05 18:38:27 build #201 of brcm2708 is complete: Failure [failed compile_4] Build details are at http://buildbot.openwrt.org:8010/builders/brcm2708/builds/201 Oct 05 18:38:40 iirc there's another package that's hit the same issue Oct 05 18:39:08 yes, the variant for asterisk11 Oct 05 18:41:28 nbd r42776 trunk/toolchain/ binutils/Makefile binutils/patches/2.22 binutils/patches/2.23.1 binutils/Config.in * binutils: remove old versions Oct 05 18:41:32 nbd r42777 trunk/toolchain/ gcc/patches/4.8.0 gcc/Config.in gcc/Config.version * gcc: remove version 4.8.0 Oct 05 18:43:13 I am running the script which watches the svn revisions Oct 05 18:43:17 I was thinking of crtmpserver ;) Oct 05 18:46:34 bash is also without cve coverage given 4.3.26 doesn't exist upstream and has unnamed patches Oct 05 18:49:45 swalker: how do you get the list of existing CVE's for packages in uscan? Oct 05 18:57:10 slachta: by giving the proper cpe string to (a patched) cvechecker, "echo cpe:/a:gnu:bash:4.3::: | cvechecker -C -r -w -" Oct 05 18:59:35 slachta: http://pastebin.com/D6eATCvJ is the output Oct 05 18:59:36 nbd r42778 trunk/ include/ltqtapi.mk package/kernel/lantiq/ltq-vmmc/Makefile package/kernel/lantiq/ltq-tapi/Makefile * lantiq: remove the useless ltqtapi.mk include file Oct 05 19:00:21 swalker: I suspected you from doing some black magic. I have not found the right tool or API for this. :-) I have not heard of cvechecker yet. Oct 05 19:02:35 slachta: 382 cpe strings so far, no idea how many I'm missing Oct 05 19:48:51 jow_laptop: ping Oct 05 19:54:24 slachta: pong Oct 05 19:56:52 I just wanted to ask you for recommendation. I am planning to write modules for luci that creates quite a big abstraction layer for the telephony packages (to make PBX distro based on OpenWrt - the luci-pbx is a small step, I am aiming to make something more powerful). Since it will be quite a long term task.. Oct 05 19:57:16 I've got several questions about the luci. Oct 05 19:58:27 I am not sure whether the way of creating modules for luci2 is the same as for luci.. I am not sure about the future of it at all. Oct 05 19:58:41 Where should I start? Oct 05 20:01:44 nbd r42779 packages/ lang/php5 lang/php5/pecl.mk Oct 05 20:01:44 php5: temporarily restore pecl.mk to reduce unnecessary oldpackages breakage from the move to github Oct 05 20:01:52 slachta: i think i started writing luci2 doc Oct 05 20:01:56 see wiki page Oct 05 20:02:05 slachta: i also wrote a module for p910nd Oct 05 20:02:09 see LuCI2 sources Oct 05 20:06:31 zajec: oh, thanks. I think I should write some foo module first to fully understand it. Oct 05 20:06:52 zajec: may I send you an email for the further questions about luci2? Oct 05 20:07:08 (it will not be today) Oct 05 20:08:58 no Oct 05 20:09:01 send it to the ML Oct 05 20:09:07 if i find I can help, i'll reply Oct 05 20:09:33 ok, thank you Oct 05 20:10:03 slachta: if its indeed a long term thing then I strongly suggest to start with luci2 Oct 05 20:10:40 slachta: you'll need it anyway I suppose because pbx stuff requires quite complicated widgets etc. which are easier done in plain js/html Oct 05 20:10:44 jow_laptop: what about my early wireless support? ;) Oct 05 20:11:09 zajec: its on the list Oct 05 20:11:12 http://patchwork.openwrt.org/patch/5738/ Oct 05 20:11:15 ok Oct 05 20:12:22 jow_laptop: that's what I'm targeting at the moment Oct 05 20:13:27 jow_laptop: any possibility for having luci2 modules on the separate repository (telephony) or should it all be located within luci repository? Oct 05 20:13:54 slachta: can be hosted anywhere Oct 05 20:14:10 there is a bit of an unsolved problem yet regarding package organization Oct 05 20:14:15 that affects luci1 too Oct 05 20:14:16 slachta: i think you didn't check my code Oct 05 20:14:30 its about internationalization Oct 05 20:14:32 slachta: see http://git.openwrt.org/?p=project/luci2/ui.git;a=tree , my module is a separated directory Oct 05 20:15:13 I think about removing language support from .ipk packages completely Oct 05 20:15:34 and set up a server somewhere where they can be downloaded separately Oct 05 20:15:49 s/they/the lang files/ Oct 05 20:16:32 the problem with out-of-tree luci apps is that they need to ship all langauge files with them because their strings cannot be part of luci-i18n-whatever obviously Oct 05 20:16:52 this makes them needlessly huge Oct 05 20:17:50 zajec: I did, I just wanted to have the information about package organization which jow_laptop just introduced Oct 05 20:18:08 nbd r42780 trunk/target/linux/ (26 files in 26 dirs) * targets: remove useless quotes around KERNELNAME Oct 05 20:18:13 nbd r42781 trunk/include/kernel-defaults.mk Oct 05 20:18:13 build: remove useless quote stripping for KERNELNAME, add fallback for it in the kernel build command line Oct 05 20:18:31 ok :) Oct 05 20:19:27 apart from that I see no problem with doing out-of-tree luci apps Oct 05 20:36:52 jow_laptop: thank you Oct 05 20:37:04 nbd: thank you for the ltq fix Oct 05 21:33:21 <_trine> does anyone know how to resolve this http://paste.debian.net/124674/ Oct 05 23:12:22 there is a strange path on the repros Oct 05 23:12:23 http://downloads.openwrt.org/snapshots/trunk/ar71xx/packages/packages/ Oct 05 23:12:33 why would there be two packages dirs ? Oct 05 23:30:24 there are the packages in the "packages" repository Oct 05 23:30:34 http://downloads.openwrt.org/snapshots/trunk/ar71xx/packages/ Oct 05 23:30:39 base, luci, packages Oct 06 00:35:42 build #553 of octeon is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/octeon/builds/553 Oct 06 02:33:06 build #94 of lantiq.xrx200 is complete: Failure [failed shell_14] Build details are at http://buildbot.openwrt.org:8010/builders/lantiq.xrx200/builds/94 Oct 06 02:36:12 build #661 of ar71xx is complete: Failure [failed shell_14] Build details are at http://buildbot.openwrt.org:8010/builders/ar71xx/builds/661 **** ENDING LOGGING AT Mon Oct 06 02:59:59 2014