**** BEGIN LOGGING AT Thu Jul 21 02:59:56 2011 Jul 21 03:28:28 build #49 of sibyte is complete: Exception [exception interrupted] Build details are at http://buildbot.openwrt.org:8010/builders/sibyte/builds/49 Jul 21 03:28:29 build #48 of xburst is complete: Exception [exception interrupted] Build details are at http://buildbot.openwrt.org:8010/builders/xburst/builds/48 Jul 21 03:28:31 build #39 of adm5120 is complete: Exception [exception interrupted] Build details are at http://buildbot.openwrt.org:8010/builders/adm5120/builds/39 Jul 21 03:33:06 hi Jul 21 03:33:18 I know C and Linux, but am new to firmware devel. Jul 21 03:33:57 I'm looking for a C API where I can return the IP Address of an interface, and make an led light/flicker. Jul 21 03:40:10 Also, I want to edit existing openwrt functionality (being the reset button NOT do anything when pressed). Where can I leard the relevent code is located ? Jul 21 03:52:26 check the wiki Jul 21 03:52:38 we already setup leds and buttons Jul 21 03:52:42 0 0 MARK all -- * * 0.0.0.0/0 0.0.0.0/0 mark match 0x0 LAYER7 l7proto http MARK set 0x3 Jul 21 03:52:57 jow_laptop: is l7 broken? Jul 21 03:59:20 Weedy: are there system calls in C for leds and buttons ? Jul 21 04:00:17 or are they scripts/commands Jul 21 04:04:12 both Jul 21 04:05:34 http://wiki.openwrt.org/doc/uci/system#leds Jul 21 04:10:44 Weedy: thanks. I get lost in big websites. Jul 21 04:10:47 * snax bookmarks Jul 21 09:59:50 Weedy: possibly Jul 21 13:48:37 obsy * r27718 /packages/net/transmission/Makefile: [packages] transmission: update to 2.33 Jul 21 15:01:02 sorry, here's probably more appropriate: Jul 21 15:01:03 on https://dev.openwrt.org/wiki/SubmittingPatches it mentions emailing the patch, with the right signed-off-by line, as an svn diff. I can do this, but we actually have multiple authors already, and it's in a hg repo at bitbucket. Is there a better way of submitting this? Jul 21 15:01:19 relevant code is: https://bitbucket.org/oojah/mosquitto-packaging/src/41fa53408b41/openwrt/ Jul 21 15:06:43 karlp: usually every commit/patch is done by one person, even if it changes from commit to commit, right? Afaik the usual kernel way is having a "From: " as the first line in the email, to make clear the submitter isn't the author. Jul 21 15:07:37 well, it was maintained for a while externally, Jul 21 15:07:53 so there's the openwrt packaging files that were used for mosquitto 0.10, Jul 21 15:08:07 then I expaned on them to support more options and the current release Jul 21 15:08:18 and the project owner merged all of them into his "upstream" Jul 21 15:08:42 as a group we'd like to get the current openwrt packaging into openwrt itself, Jul 21 15:09:29 seeing as there's no mosquitto package in openwrt at all at present, it could all be seen as a single commit into openwrt, probably by the project owner? Jul 21 15:09:31 well then there should be the appropriate copyrights in the source files; no need to have them in the commit-message (that should probably be the one that did the packaging) Jul 21 15:10:38 the only source file is the makefile though? Jul 21 15:10:53 (it does have a copyright notice from the first openwrt packager) Jul 21 15:11:19 well, the makefile will probably trigger a download of the actual sources, right? Jul 21 15:11:34 yes, but those are already properly licensed and copyright noticed.. Jul 21 15:12:20 am I correct in thinking that what we need to to get a package into the official "feeds" index is to submit the Makefile and patches that go into /packages ? Jul 21 15:12:46 well, you submit a patch for the packages feed that adds the Makefile and the patches Jul 21 15:13:14 so it's partially a patch of patches ;) Jul 21 15:13:20 yes :) Jul 21 15:13:37 I was just curious, as the Makefile is the work of multiple people already... Jul 21 15:14:24 you'll find plenty of examples on http://patchwork.openwrt.org/, e.g. http://patchwork.openwrt.org/patch/1069/ Jul 21 15:14:55 which mirko forgot to set to accepted ;p Jul 21 15:56:01 thanks KanjiMonster, I've got one in now: http://patchwork.openwrt.org/patch/1230/ Jul 21 16:39:29 mazilo: is there any way to patch libraries within freeswitch ? Jul 21 16:41:19 the tradition mechanism does not work because FS downloads the stuff later and unpacks it itself Jul 21 16:42:21 Is there an easy way to override gcc in the build? Mac OS X 10.7, gcc is now LLVM, and you have to specify gcc-4.2 is want gcc with no LLVM Jul 21 16:42:58 the toolchain compiler built with LLVM won't run and I am trying to figure out if its LLVM Jul 21 16:43:42 I've tried both gcc-4.6.1 and gcc with linaro, both seg fault when run with built with apple's default llvm (gcc) Jul 21 16:44:44 tried CC=gcc-4.2 ? Jul 21 16:46:01 no, just setting `$ CC=gcc-4.2 make' ? that would be easy ;-) Jul 21 16:46:09 let me try Jul 21 16:46:14 make CC=gcc-4.2 Jul 21 16:46:27 one sets a shell var, one tells make to override it's built in CC var Jul 21 16:46:57 oh yea, right, good point Jul 21 16:47:29 I know that some programs violate this by hardcoding "cc" Jul 21 16:47:41 but since those are usually host tools it should be no issue Jul 21 16:49:43 and there a few other 10.7 patches i have Jul 21 16:50:08 darwin finally updates libc and supports getline() and some others, which were previously worked around Jul 21 16:50:18 \o/ Jul 21 16:54:35 test Jul 21 16:56:41 [florian], Kaloz do you have any more input for the support of my board? Do you think it could be added to OpenWRT? Or maybe not Jul 21 17:12:52 well that worked Jul 21 17:13:04 * karlp cheers Jul 21 17:14:32 LLVM breaks gcc (in the same place too for both 4.6.1 and 4.5.x with linaro Jul 21 17:17:06 should i waste my time patching openwrt files to respect CC? or should I see if I can patch the problem in gcc? and make toolchain gcc work with LLVM? Jul 21 17:17:55 both! Jul 21 17:18:06 openwrt files not respecting CC is a separate and real bug Jul 21 17:18:22 and probably the fastest path to glory for yourself Jul 21 17:18:43 working out how to make gcc and LLVM play properly will help others in the longer term Jul 21 17:20:59 good points, i work on CC first, so I can continue building openwrt, then look at the LLVM bug Jul 21 17:22:44 perhaps file the LLVM bug right now, so that the gcc/llvm guys can get on it themselves in the meantime maybe? Jul 21 17:24:25 sure Jul 21 18:18:20 jow_laptop: well shit :/ anything I can do to fix it? Jul 21 19:00:45 Is there a way to create an OpenWRT that doesn't do Prepare, Configure, Configure, and Compile, but just package the source code into an OpenWRT IPKG file? Jul 21 19:13:38 mazilo: you mean something like this? -> https://dev.openwrt.org/browser/trunk/package/6in4/Makefile ;) Jul 21 19:25:11 KanjiMonster: Thanks. But, the 6in4 package does do the compilation. Jul 21 19:25:40 mazilo: no it doesn't, see line 31/32 Jul 21 19:33:37 KanjiMonster: You are right and that's what I am looking for. Thanks. Jul 21 19:40:13 mazilo: you're welcome Jul 21 21:38:53 Weedy: updates usually help Jul 21 21:40:55 jow_laptop: i'm at like last weeks trunk Jul 21 23:32:17 luka * r27719 /packages/utils/mc/Makefile: Jul 21 23:32:17 mc: prevent from picking up ncursesw Jul 21 23:32:17 mc will link against ncursesw instead of ncurses if it finds it. Make sure Jul 21 23:32:17 it always links only against ncurses. Jul 21 23:32:17 Closes #9225 and #9315. Jul 21 23:32:18 Signed-off-by: Jonas Gorski Jul 22 00:27:10 jow * r27720 /trunk/package/base-files/ (Makefile files/sbin/ifdown files/sbin/ifup): (log message trimmed) Jul 22 00:27:10 [package] base-files: attempt bring up related wifi devices when calling ifup Jul 22 00:27:10 If a user invoked /sbin/ifup to bring up an interface, the setup used to fail Jul 22 00:27:10 in case of wireless networks tied to a non-bridged interface definition. Jul 22 00:27:10 Likewise, the bringup of "lan" in the default configuration will reinitialize Jul 22 00:27:10 the bridge but do not re-join the wireless network to it, requiring an extra Jul 22 00:27:11 call to /sbin/wifi (which might not be possible anymore due to a severed link Jul 22 02:10:15 jow_laptop: how can review/commit the OCF version bump? Jul 22 02:10:19 *who **** ENDING LOGGING AT Fri Jul 22 02:59:57 2011