**** BEGIN LOGGING AT Fri Jan 22 02:59:58 2016 Jan 22 07:06:57 morning guys Jan 22 07:43:45 morning nitro Jan 22 09:23:20 rmilecki r48449 trunk/ (6 files in 5 dirs) * kernel: backport bcma stuff sent for 4.6 Jan 22 09:42:35 nbd r48450 trunk/include/target.mk * build: add uclient-fetch to the default packages, in case opkg is not selected Jan 22 09:42:41 nbd r48451 trunk/package/base-files/files/lib/upgrade/common.sh * base-files: fix sysupgrade 'wget' handling Jan 22 11:59:45 nbd r48452 trunk/package/network/config/qos-scripts/ files/usr/lib/qos/generate.sh files/usr/lib/qos/tcrules.awk * qos-scripts: Add IPv6 support Jan 22 12:02:49 nbd r48453 trunk/target/linux/ generic/config-4.4 x86/64/config-default * x86_64: Add missing symbols Jan 22 12:02:54 nbd r48454 trunk/package/kernel/linux/modules/other.mk * kernel: Add kernel module for Freescale SNVS RTC on chip module Jan 22 12:59:02 nbd: Could you also please bump the version or revision on qos-scripts? I would submit a patch, but I think it would be just as easy for you to do it. Jan 22 13:06:52 nbd r48455 trunk/package/network/config/qos-scripts/Makefile * qos-scripts: bump version Jan 22 13:08:46 Thanks! :) Jan 22 15:58:00 if my package _only_ has DEPENDS lines with +xxx in them, how do I end up with a generated Config.in that has a "depends on " in the middle of all the "select PACKAGE_" crufgt? Jan 22 15:58:32 it's luci-base causing problems, what should my luci apps do if not DEPENDS=+luci-base? Jan 22 15:59:03 I want to have my package do DEPENDS= +luci-mod-admin-full +luci-app-mystuff and not get circular dependency complaints with luci-base Jan 22 15:59:15 you inherit dependencies from packages you select Jan 22 15:59:35 maybe you need to add +luci-base as well Jan 22 15:59:57 brb Jan 22 15:59:58 https://paste.fedoraproject.org/313792/14534783/ Jan 22 16:00:18 why do I need to do the +luci-base as well, if luci-mod-admin-full already does that? as does luci-app-mystuff? Jan 22 16:00:24 hm Jan 22 16:01:24 what are the deps of luci-app-mystuff? Jan 22 16:01:44 bu tyeah, explicitly adding +luci-base to my top level "fixes" it Jan 22 16:02:03 my app had +luci-lib-json, and +luci-base Jan 22 18:12:54 Argh, I am getting the general protections faults from procd on my x86_64 trunk builds again. I had it working for a while, but I accidentally wiped my configuration and now it crashes every time again. Jan 22 18:13:26 never seen that Jan 22 18:14:20 It's the craziest thing. It only happens when I import my configuration from CC though, so there must be something in that configuration that is causing the problem. Jan 22 18:15:09 I would think it would be related to the toolchain or libc or something, but if I start a configuration from scratch and manually reconfigure it the same way it was in the imported configuration, I don't get the problem. Jan 22 18:30:27 I guess I will need to copy over the configuration bit-by-bit until I figure out which thing is causing the problem. This will be fun. Jan 22 19:50:52 nbd r48456 trunk/target/linux/ar71xx/image/Makefile * Revert "ar71xx: Use full 128MB flash on Netgear WNDR4300 and WNDR3700v4" Jan 23 00:07:08 nbd, jow_laptop: wrt uhttpd/polarssl/ustream-ssl issues - fresh, clean checkout, http://pastebin.com/raw/uYRSG7U0 Jan 23 00:27:18 swalker: please post the diff of the full changes that you've made Jan 23 00:27:55 it uses ECDHE agian, so it doesn't look like unmodified trunk to me Jan 23 00:31:51 nbd, https://patchwork.ozlabs.org/patch/571420/ ;) Jan 23 00:32:11 will look into that tomorrow Jan 23 00:32:15 going to need some sleep soon Jan 23 00:32:40 night Jan 23 00:32:46 works awesome btw Jan 23 00:34:03 nbd: that was from a clean tree, but it was a firewall redirect issue Jan 23 02:56:26 CONFIG_PACKAGE_less-wide=y apparently causes procd to crash immediately on startup. It took me a day of repeatedly compiling with different configurations to figure that out. **** ENDING LOGGING AT Sat Jan 23 02:59:58 2016