**** BEGIN LOGGING AT Fri Jul 25 03:00:01 2014 Jul 25 03:00:35 ah, yes, vlans on awm003 are hosed because of the new name not being supported Jul 25 03:00:55 * russell-- submitted an earlier patch that fixed an earlier brokenness Jul 25 05:09:41 build #675 of cobalt is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/cobalt/builds/675 Jul 25 07:30:15 build #132 of bcm53xx is complete: Failure [failed shell_14] Build details are at http://buildbot.openwrt.org:8010/builders/bcm53xx/builds/132 Jul 25 08:05:13 build #130 of pxa is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/pxa/builds/130 Jul 25 09:16:55 build #586 of kirkwood is complete: Failure [failed shell] Build details are at http://buildbot.openwrt.org:8010/builders/kirkwood/builds/586 Jul 25 10:23:35 build #129 of cns3xxx is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/cns3xxx/builds/129 Jul 25 10:48:09 build #635 of ppc44x is complete: Failure [failed shell_9] Build details are at http://buildbot.openwrt.org:8010/builders/ppc44x/builds/635 Jul 25 12:22:44 build #646 of uml is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/uml/builds/646 Jul 25 12:24:59 build #604 of sibyte is complete: Failure [failed shell_9] Build details are at http://buildbot.openwrt.org:8010/builders/sibyte/builds/604 Jul 25 12:58:19 swalker: ping Jul 25 12:59:47 swalker: I wanted to inform you that upstream version in uscan for openwrt says that spandsp in pre21 version is newer than the final version. Would you be so kind to take a look at it? (telephony repo) Jul 25 13:19:15 build #616 of avr32 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/avr32/builds/616 Jul 25 14:04:39 build #705 of at91 is complete: Failure [failed shell] Build details are at http://buildbot.openwrt.org:8010/builders/at91/builds/705 Jul 25 14:14:13 jow r41826 trunk/package/network/utils/iwinfo/ src/iwinfo_nl80211.c src/iwinfo_utils.c Makefile * iwinfo: continue scanning even if temporary mac cannot be changed Jul 25 14:26:23 nbd r41827 trunk/package/kernel/mac80211/patches/300-pending_work.patch * mac80211: merge a few WDS-AP related fixes (#14756, #16802) Jul 25 14:32:18 nbd r41828 trunk/package/network/services/hostapd/patches/300-noscan.patch Jul 25 14:32:18 hostapd: prevent spurious 20/40 mhz channel bandwidth switches if noscan is enabled Jul 25 14:39:03 jow r41829 trunk/package/network/utils/iwinfo/src/iwinfo_utils.c * iwinfo: fix secondary radios being misreported as ralink device Jul 25 14:40:08 build #529 of mcs814x is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/mcs814x/builds/529 Jul 25 15:52:50 Mazilo: ping Jul 25 15:53:32 Mazilo: would you be so kind to check all dependencies for the freeswitch package (without the packages from old-packages repository installed)? Jul 25 15:54:38 jow r41830 trunk/package/network/utils/iwinfo/src/iwinfo_nl80211.c * iwinfo: avoid creating tmp.* ifaces for scanning Jul 25 16:23:04 slachta_: I am not sure exactly what you meant. If you found any unused dependencies, please do let me know. Jul 25 16:24:13 slachta_: R U talking about libedit? Jul 25 16:29:19 jow_laptop: do you know what are GPIOs? Jul 25 17:03:19 Mazilo: rather make sure all dependencies are available from base/packages and none to old-packages Jul 25 18:02:08 build #681 of orion is complete: Failure [failed compile_8] Build details are at http://buildbot.openwrt.org:8010/builders/orion/builds/681 Jul 25 19:07:38 KanjiMonster: R U referring to libedit as dependency? Jul 25 20:36:20 Hauke: hey, could you think about picking my b43 patches again? Jul 25 20:36:28 Hauke: to add support for 5 GHz to the OpenWrt Jul 25 20:36:39 yes will do that Jul 25 20:37:00 thx Jul 25 20:37:17 btw. is there some way to build wireless driver from external tree? Jul 25 20:37:34 like tell OpenWrt to use some directory or git repo for wireless drivers? Jul 25 20:37:46 i don't really have any better idea for developing drivers for OpenWrt Jul 25 20:41:07 uhm, how can i set a conf file permissions for those in files/ before building trunk ? Jul 25 20:41:27 "just do it" works for me. Jul 25 20:41:59 zajec: no I do not know of such a way Jul 25 20:42:50 nbd: ping? Jul 25 20:43:02 nbd: how do you develop/work on wireless drivers? :| Jul 25 20:54:24 karlp, nope, does not work, i guess i have to fiddle with the Makefile Jul 25 21:00:51 works for me with packages at least, Jul 25 21:01:08 I jsut chmod'd +x in my package and it worked fine nex ttime. Jul 25 21:04:41 zajec: ? Jul 25 21:08:05 nbd: see my question posted to Hauke Jul 25 21:08:16 nbd: i was wondering if I can replace compat-wireless with a git tree Jul 25 21:08:30 to make OpenWrt compile wireless drivers from pointed git tree (like a wireless-next.git) Jul 25 21:08:41 that would allow me to develop wireless driver and test them Jul 25 21:08:57 that's not going to work because of the backport stuff Jul 25 21:09:06 what i do is this: Jul 25 21:09:13 what kind of backport stuff? Jul 25 21:09:21 since I switch to wireless-next.git I don't need any backports Jul 25 21:09:31 ah Jul 25 21:09:35 so you want to replace the entire kernel Jul 25 21:09:47 well, you'd have to port the arch patches, some of the generic openwrt patches, etc. Jul 25 21:09:49 just wireless drivers would be nicer Jul 25 21:09:58 as i said, that's not going to work Jul 25 21:10:02 yeah, switching the wole kernel is a big effort Jul 25 21:10:02 since that needs backport stuff Jul 25 21:10:08 so here's how i do it: Jul 25 21:10:14 * zajec listening to Jul 25 21:10:25 i make patches in the compat-wireless build dir using quilt Jul 25 21:10:43 when those are done, i apply them to my wireless-testing checkout and make git commits out of them Jul 25 21:10:46 and submit them from there Jul 25 21:11:06 yeah, but first I'd need to backport patches that are already in wireless-next.git Jul 25 21:11:11 and that are not in compat-wireless yet Jul 25 21:11:19 right Jul 25 21:11:24 and that isn't easy as what is in build_dri doesn't accept there patches Jul 25 21:11:29 ? Jul 25 21:11:57 i can't do cd build_dir/..../compat_wireless; patch -p1 < 0001-b43-sent-mainline.patch Jul 25 21:12:03 because patches in mainline use CONFIG_ Jul 25 21:12:19 and what is in build_dir uses CNFCFG_ (or sth like that) Jul 25 21:12:30 right, that part doesn't work as is Jul 25 21:12:33 :( Jul 25 21:12:39 so still no clean solution Jul 25 21:12:41 ok Jul 25 21:12:44 i usually do search/replace when syncing patches with such stuff between owrt and the git tree Jul 25 21:12:57 yeah, i tried that so Jul 25 21:12:58 *too Jul 25 21:13:03 just a big annoying :) Jul 25 21:13:09 yep Jul 25 21:13:26 do you think we could create another package to solve this problem? Jul 25 21:13:33 something to replace compat-wireless Jul 25 21:13:40 i don't think that's going to work Jul 25 21:13:41 that will pick code from git tree Jul 25 21:13:43 :( Jul 25 21:13:51 if you say that, i won't even try ;) Jul 25 21:13:57 because the code from wireless-next won't build as-is Jul 25 21:14:02 because it's targetting a newer kernel version Jul 25 21:14:14 sec Jul 25 21:14:18 i mean picking only wireless drivers Jul 25 21:14:22 not the whole kernel Jul 25 21:14:27 i know Jul 25 21:14:30 could we pick wirelss drivers from wireless-next only? Jul 25 21:14:39 they should compile pretty cleanly, shouldn't they? Jul 25 21:14:41 but the wireless drivers depend on the kernel infrastructure and APIs Jul 25 21:14:47 and those APIs tend to change between kernel versions Jul 25 21:15:17 hm, i hoped cfg80211 + mac80211+ b43 won't really depen on too much other code Jul 25 21:15:35 it depends on quite a bit of api related stuff Jul 25 21:15:39 ok Jul 25 21:15:42 trusgting you Jul 25 21:16:00 i think i'll be back to my old method then... replacing the whole kernel with the wireless-next git and applying OpenWrt patches in it Jul 25 21:16:13 (and using in-kernel drivers instead of compat-wireless) Jul 25 21:16:22 is there really that much CONFIG_* stuff? Jul 25 21:16:48 because in most cases, dealing with manual CONFIG_* related hackery is a lot less work than replacing the kernel Jul 25 21:16:48 last time I tried it I got many patches touching code in/around CONFIG_* Jul 25 21:17:21 but that was a time there were some changes in b43 toucing Makefile + Kconfig + code around #ifdef Jul 25 21:17:26 maybe it'll try again Jul 25 21:17:28 doesn't our current code already have most of the interesting b43 upstream patches for your to continue your work? Jul 25 21:17:38 with all those backports that hauke made Jul 25 21:17:46 b43 changes upstream every day ;) Jul 25 21:54:41 uhm, is there any hidden reason for INSTALL_CONF not chmod 0600 the config files i instructed to copy to ipkg? Jul 25 23:42:09 harro Jul 25 23:42:34 i am wondering about the relationship of rpcd, and uhttpd-mod-ubus + ubus Jul 25 23:43:24 luci2 is backed by the latter, and i'm wondering if it's a good idea to write scripts for rpcd Jul 25 23:44:16 i basically wanna manage a process, i.e. update its state when configuration changes, and when network interfaces change Jul 25 23:44:39 for the network stuff, netifd is obvious Jul 25 23:45:16 ubus seems right for listening to config changes, and passing events between processes Jul 25 23:45:29 so what's rpcd's role in the wider system? Jul 26 00:36:06 i think i got it Jul 26 00:36:46 http://wiki.openwrt.org/doc/techref/ubus#namespaces.procedures was helpful Jul 26 00:38:36 http://wiki.openwrt.org/doc/techref/ubus#acls as well Jul 26 00:57:34 build #640 of x86 is complete: Failure [failed shell] Build details are at http://buildbot.openwrt.org:8010/builders/x86/builds/640 Jul 26 01:33:04 nbd, I've found a very nasty race condition in netifd Jul 26 01:33:18 neoraider: what kind of race condition? Jul 26 01:33:56 I have a bridge with proto dhcpv6 which gets a renew very quickly after the setup. As both use the same netifd_process, the setup is killed before it is finished Jul 26 01:35:20 ok Jul 26 01:35:38 shouldn't be too hard to fix Jul 26 01:37:25 i'll make a patch for you to test Jul 26 01:37:49 Thanks Jul 26 01:40:02 neoraider: http://nbd.name/netifd-renew.patch Jul 26 01:46:17 nbd, seems to work fine :) Jul 26 01:46:25 great Jul 26 01:49:09 nbd r41831 trunk/package/network/config/netifd/Makefile * netifd: update to the latest version, fixes a race condition with renew/setup **** ENDING LOGGING AT Sat Jul 26 03:00:00 2014