**** BEGIN LOGGING AT Mon Sep 05 02:59:57 2011 Sep 05 03:05:33 jow_laptop: modules/admin-full/ipkg/postinst is still around post r7388 Sep 05 04:21:03 swalker * r28172 /packages/net/uftp/Makefile: [packages] uftp: fix typo Sep 05 06:20:43 build #65 of ppc44x is complete: Failure [failed compile_4] Build details are at http://buildbot.openwrt.org:8010/builders/ppc44x/builds/65 Sep 05 07:11:36 build #72 of lantiq is complete: Failure [failed compile_6] Build details are at http://buildbot.openwrt.org:8010/builders/lantiq/builds/72 Sep 05 08:04:52 <_trine> how come the primary colour has changed from blue to green in make menuconfig Sep 05 08:10:20 depends on your terminal Sep 05 08:11:41 <_trine> why Sep 05 08:11:43 everybody wants to go green these days Sep 05 08:12:00 <_trine> it looks crap Sep 05 08:12:12 <_trine> green fanboys Sep 05 08:13:06 but as loswillios said it is probably your terminal emulator not make menuconfig Sep 05 08:18:18 <_trine> nope it was always blue before Sep 05 08:19:05 <_trine> now we have blue writing on a green background Sep 05 08:19:21 <_trine> blue and Green should never be seen except when worn by a fool Sep 05 08:19:29 <_trine> is the saying Sep 05 08:22:10 <_trine> the finished firmware has also significantly risen in size recently Sep 05 08:23:00 _trine: echo -e '\033[34mBLUE?\033[32mGREEN?\033[0m' Sep 05 08:24:56 <_trine> BLUE?GREEN? Sep 05 08:25:29 <_trine> thats just the asci code Sep 05 08:25:53 no colors? Sep 05 08:26:04 <_trine> yes blue and green Sep 05 08:26:27 hm, so it is make menuconfig and not your terminal emulator Sep 05 08:26:40 <_trine> of course Sep 05 08:26:40 awesome Sep 05 08:27:02 <_trine> I have 2 Dell monitors here Sep 05 08:27:05 we should add an option in the developers section Sep 05 08:27:10 lol Sep 05 08:27:14 " Turn off n00b mode" Sep 05 08:27:29 and leave "blue writing on a green background" as default Sep 05 08:27:35 <_trine> it looks shit at the moment its making me feel sick Sep 05 08:28:28 <_trine> it looks like some psychopath has been allowed to play Sep 05 08:28:41 check the svn logs Sep 05 08:28:59 <_trine> own up oh guilty one Sep 05 08:29:03 <_trine> :) Sep 05 08:30:14 <_trine> why has the code suddenly shot up in size? Sep 05 08:30:48 <_trine> now my normal configuration wont fit in my wndr3700 anymore Sep 05 08:31:33 <_trine> Luci is the first candidate to get the chop Sep 05 08:32:58 _trine: find build_dir/target-mips_r2_uClibc-0.9.32/root-ar71xx -type f | xargs ls -larS Sep 05 08:33:02 then see what the biggest files are Sep 05 08:33:34 _trine: do you use gcc 4.6? afaik it tends to build bigger binaries than 4.5 Sep 05 08:34:31 <_trine> I have not changed anything I always build the same make menuconfig but for some reason it just got bigger Sep 05 08:35:33 ar71xx defaults to 4.5, but check with grep GCC .config Sep 05 08:36:33 _trine: yeah, i got that part Sep 05 08:36:40 _trine: it would still be a good idea to figure out what's so big Sep 05 08:36:50 <_trine> let me look Sep 05 08:54:12 <_trine> this is the last bit of the file http://dpaste.com/608977/ Sep 05 08:59:22 looks perfectly normal Sep 05 08:59:34 sure you don't have anything in files/? Sep 05 09:02:08 <_trine> I didnt make a files folder Sep 05 09:06:48 <_trine> the top and bottom of it is this the same build has gone from 5.4 meg to 7.8meg Sep 05 09:07:15 <_trine> I will clear it all out and try a new compile Sep 05 09:08:02 hm, seems that the file list is too long for proper sorting Sep 05 09:09:46 _trine: find build_dir/target-mips_r2_uClibc-0.9.32/root-ar71xx -type f | xargs stat -c "%s %N" | sort -n Sep 05 09:12:27 <_trine> I have already deleted it all sorry Sep 05 09:12:34 ok Sep 05 09:12:40 <_trine> I have made a distclean and now make Sep 05 09:13:00 <_trine> it will tale a while to complete Sep 05 09:13:58 <_trine> I saved ./config and am reusing it Sep 05 09:45:57 _trine: status report capitan Sep 05 09:46:57 <_trine> Weedy, ship is steaming full speed ahead and weedy pirates are about to walk the plank Sep 05 09:47:26 watch out for that iceburg Sep 05 09:48:32 <_trine> ar ar mi hearty thars no icebergs in the Caribbean Sep 05 09:49:51 coral reef then Sep 05 09:50:00 off the port bow Sep 05 09:53:48 nbd: I get some of those with latest trunk + hp touchpad client: received EAPOL-Key 2/2 Group with unexpected replay counter Sep 05 09:54:04 those are fine Sep 05 09:54:15 i think i should probably hide them with the default verbosity level Sep 05 09:54:24 okies Sep 05 09:54:44 yeah Sep 05 09:55:26 > touchpad Sep 05 09:55:35 loswillios: YOU MOTHER FUCKER Sep 05 09:55:51 heh Sep 05 09:56:18 sorry I can't hear you over the noise how awesome my touchpad is Sep 05 09:56:36 well, or s/awesome/cheap/ Sep 05 09:56:49 > 10 inch IPS Sep 05 09:56:53 you all seen the hitler touchpad rant video yet? :P Sep 05 09:57:01 my hate, you personify all of it Sep 05 09:57:24 so is the android dump a firmware image yet? Sep 05 09:58:04 no, but cyanogen is working on it Sep 05 09:58:26 any idea whether it'll be able to fully run without proprietary driver? Sep 05 09:58:32 did the android touchpads ever get to devs? Sep 05 09:59:27 nbd: that's unknown yet (to me at least) Sep 05 09:59:48 Weedy: the dumps did, the pad was sold on ebay IIRC Sep 05 10:02:25 nbd: at least the wifi-driver seems to be AR6003X :) Sep 05 10:02:38 :) Sep 05 10:03:23 loswillios: cyanogen got lucky **** BEGIN LOGGING AT Mon Sep 05 11:28:16 2011 Sep 05 13:53:28 build #61 of avr32 is complete: Failure [failed compile_4] Build details are at http://buildbot.openwrt.org:8010/builders/avr32/builds/61 Sep 05 15:58:57 Hi, can someone give me a hint for fixing my package makefile? it is an "external" kernel module and depends on $(LINUX_DIR)/System.map .. i've tried to realize it the default way (BuildPackage ...) and as kernel module (KernelPackage...), both work, but not in a fresh buildroot, because my package then gets build before the kernel (and thus System.map is missing). How to add that dependency? Tried everything i could imagine, but nothing Sep 05 15:59:47 giz2: right now its more or less impossible because the kernel image is built as one of the very last steps of the build process Sep 05 16:02:50 Hmm... interesting and at least clarifying, would have expected the opposite (kernel gets build very early .. at least before the rest of the target software). Sep 05 16:23:08 the modules are Sep 05 16:23:13 but not the kernel image Sep 05 17:19:10 build #70 of pxcab is complete: Failure [failed compile_10] Build details are at http://buildbot.openwrt.org:8010/builders/pxcab/builds/70 Sep 05 18:15:35 I am trying to compile some packages that require "staging_dir/toolchain-mipsel_gcc-4.6-linaro_uClibc-0.9.32/mipsel-openwrt-linux-uclibc/include/c++/4.6.2/ext/mt_allocator.h", "staging_dir/toolchain-mipsel_gcc-4.6-linaro_uClibc-0.9.32/mipsel-openwrt-linux-uclibc/include/c++/4.6.2/ext/mt_allocator.h" , and "staging_dir/toolchain-mipsel_gcc-4.6-linaro_uClibc-0.9.32/mipsel-openwrt-linux-uclibc/include/c++/4.6.2/mipsel-openwrt-linux- Sep 05 18:15:37 uclibc/bits/c++config.h" header files. Does anyone here know how do I add these to an OpenWRT Makefile for the package so that it will compile for other different platforms? Sep 05 18:18:01 Basically, the the inclusion files are called like: #include . #include , and #include , respectively. Sep 05 18:20:04 You could try to resolve the include path with a glob Sep 05 18:20:33 jow_laptop: I am not familiar with 'glob'. Can you please kindly elaborate? Sep 05 18:21:12 e.g. TARGET_CXXFLAGS += -I$(firstword $(wildcard $(TOOLCHAIN_DIR)/*/include/c++/*)) Sep 05 18:21:31 jow_laptop: OK. Thanks. I will try that. Sep 05 18:33:13 hi, is there a reason that there is no profile for ubiquiti m devices besides "no one has done it"? i think it would be good to have one that includes ath9k Sep 05 18:34:56 <_trine> soma, m is compiled with the others Sep 05 18:35:28 yes. its compiled, but at least when i generate images with the imagebuilder with profile UBNTRS ath9k is missing Sep 05 18:35:57 <_trine> I use trunk Sep 05 18:36:12 <_trine> its all ok with trunk Sep 05 18:36:17 of course i can select that extra (and tell users to do see), but it would be more convenient to have another profile like UBNTM or so Sep 05 18:36:47 really? i checked the profiles and it seems ath9k is never included, i'll recheck Sep 05 18:38:37 ups, i meant "when i generate images with the imagebuilder with profile UBNT", not UBNTRS. Sep 05 18:39:49 no, using UBNT its not referenced anywhere, thats a bit misleading Sep 05 18:42:44 nbd * r28173 /trunk/target/linux/ar71xx/ (17 files in 3 dirs): ar71xx: clean up profiles, put in kmod-ath9k and wpad-mini by default (fixes #9954) Sep 05 18:43:05 thx nbd :) Sep 05 18:44:42 but imo thats not optimal, on some devices like RS it might not be needed Sep 05 19:08:06 soma: imho most people use RS with 11n cards Sep 05 19:08:13 soma: and the ones that don't can just deselect the driver Sep 05 19:09:00 ok. makes sense. do you intend to backport that to backfire too? Sep 05 19:09:26 i guess i could do that Sep 05 19:11:50 probably ;) i would like it, would somehow "fix" the imagebuilder i run (i din't need to tell people they need to select ath9k extra) Sep 05 19:12:13 sure Sep 05 19:17:09 nbd * r28174 /branches/backfire/target/linux/ar71xx/ (13 files in 2 dirs): ar71xx: sync profile/target default changes with trunk Sep 05 19:17:17 there you go Sep 05 19:17:27 thx again Sep 05 19:34:20 jow * r28175 /trunk/rules.mk: [buildroot] rules.mk: provide a LIBGCC_A variable Sep 05 19:36:59 jow * r28176 /branches/backfire/rules.mk: [backfire] backport r28175 Sep 05 19:37:35 jow * r28177 /packages/utils/valgrind/Makefile: [packages] valgrind: use LIBGCC_A Sep 05 19:38:11 jow * r28178 /branches/packages_10.03.1/utils/valgrind/Makefile: [backfire] merge r28177 Sep 05 20:05:24 heil * r28179 /packages/net/haproxy/Makefile: Sep 05 20:05:24 package: haproxy Sep 05 20:05:24 - bugfixes from upstream results in version 1.4.17 Sep 05 20:43:02 nbd: seems like the drivers for the touchpad are fully open, but some userspace stuff (ie graphics) is proprietary Sep 06 00:28:28 jow * r28180 /packages/utils/valgrind/ (Makefile patches/002-fix_load_address_substitution.patch): (log message trimmed) Sep 06 00:28:28 [packages] valgrind: fix load address handling Sep 06 00:28:28 The compile-time generation of the valgrind linker script broke due to toolchain updates Sep 06 00:28:28 resulting in valgrind binaries compiled with the standard load address of 0x08048000 which Sep 06 00:28:28 leads to memory conflicts with the debug client later on; valgrind will abort with the Sep 06 00:28:28 message below: Sep 06 00:28:29 valgrind: mmap(0x8048000, 348160) failed in UME with error 22 (Invalid argument). Sep 06 00:29:05 jow * r28181 /branches/packages_10.03.1/utils/valgrind/ (Makefile patches/002-fix_load_address_substitution.patch): [backfire] merge r28180 **** ENDING LOGGING AT Tue Sep 06 02:59:56 2011