**** BEGIN LOGGING AT Sat Nov 02 02:59:57 2019 Nov 02 06:07:26 http://buildbot.openwrt.org/master/images/builders/ath79%2Fgeneric/builds/135/steps/signunpack/logs/stdio time is incorrect somewhere Nov 02 09:46:24 Kids part in the office building on Saturday. One out of two days when it is possible to get something done ... Nov 02 09:46:27 party Nov 02 09:46:33 bah Nov 02 13:42:08 Hi my luci is not working rite. I am getting Error Nov 02 13:42:08 XHR request timed out Nov 02 13:42:24 Am I missing a package from my build? Nov 02 13:49:49 Tapper: Someone posted to mailing list commenting about the same right now, so you are not alone at least :) Nov 02 14:17:29 kristrev OK Thanks mate. I have bin roped in to making a apple cake with my 9 year old girl now so I have know time to play with routers lol Nov 02 14:23:07 Tapper: Sounds like a nice Saturday, enjoy Nov 02 14:52:03 is there any kind of "wan watchdog" software available in openwrt? as of now it seems that openwrt doesn't allwayas reconnect WAN when it does become available (basic ethernet for openwrt perspective) Nov 02 14:52:07 upstream we have lte-modem with IP passthrough, and that sometimes fials, and comes back up, but openwrt behind it odes not Nov 02 15:00:00 is there a luci irc channel, who are the core developers for luci, i hope jow is not the only one Nov 02 15:00:25 stupid question i am going to check luci commit log to figure out Nov 02 15:01:29 https://github.com/openwrt/luci/graphs/contributors pretty much it is jow alone since 2010 :) Nov 02 15:02:58 inda sad... I'm happy user of (also) luci, it's awesome! Nov 02 15:06:31 mangix: thanks, the perl watch files have been updated Nov 02 15:08:06 jow: is luci.http.conditionals luci.http.date luci.http.mime going away? the doc page on github showed 404: Error: Cannot load https://raw.githubusercontent.com/openwrt/luci/master/documentation/api/modules/luci.http.conditionals.html: 404 Not Found Nov 02 15:47:43 since embedded lua is all good why not use it as the default for luci Nov 02 16:09:23 Hello where can I download the 4.14.151 kernel headers? Nov 02 16:10:57 for the OpenWrt 19.07-SNAPSHOT Nov 02 16:28:43 TheM4rco: when you want to build a kernel module for an openwrt snapshot I would suggest to build openwrt from sources Nov 02 16:29:16 I do not know if it is also possible to build kernel moduels with the SDK or if it only works for user space applications Nov 02 18:28:22 Hauke: https://patchwork.ozlabs.org/patch/1188122/ Nov 02 18:28:34 uff, why do we use longjmp() in the first place ?! Nov 02 18:28:46 i merged 1/2 but need to look at 2/2 closer Nov 02 18:38:59 ugh. This is going to hurt perfomance. Though I guess it doesn't really matter in this context ;) Nov 02 18:41:24 f00b4r0: yup Nov 02 18:41:32 I am wondering if it might make more sense to fix the cause Nov 02 18:42:33 is https://my.blinn.edu loading for anyone? Nov 02 18:42:54 Slimey: i am always scared to click on random links ppl post ;) Nov 02 18:43:08 well tbh just reading the commit message without reading the code made me cringe... This is the typical case of "wrong reason to use the volatile keyword" :) Nov 02 18:43:36 volatile makes me cringe full stop Nov 02 18:43:49 Slimey: https://downforeveryoneorjustme.com/my.blinn.edu Nov 02 18:44:06 blogic: good then ;) Nov 02 18:44:15 i get conflicting reports from that site its down for some and up for others Nov 02 19:10:39 OTOH volatile is useful with compiler bugs Nov 02 19:10:44 specifically with OpenCL Nov 02 19:11:13 where Nvidia OpenCL compiler works properly but not AMD... Nov 03 00:19:11 blogic: the error handling in the uci code uses longjmp all over the place Nov 03 00:19:55 when you jump back with a longjump the last setjump the state of non volatile variables is undefined Nov 03 00:20:20 this is what the manpage of longjump says Nov 03 00:20:56 volatile force the comoiler to put this onto the stack and then it has a defined value after the longjmp Nov 03 00:21:19 removing all the longjmps from the code is a big refactoring Nov 03 00:21:59 the volatile is not a workaround bt a fix for this code Nov 03 00:22:01 *but Nov 03 00:22:21 I also do not like volatile, but I also do not like longjmp Nov 03 00:23:04 makeing this volatile would also have prevented the CVE **** ENDING LOGGING AT Sun Nov 03 02:59:59 2019