**** BEGIN LOGGING AT Wed Mar 02 02:59:57 2011 Mar 02 05:12:28 build #75 of uml is complete: Failure [failed compile_10] Build details are at http://tksite.gotdns.org:8010/builders/uml/builds/75 Mar 02 07:00:54 swalker * r25819 /packages/net/tor/Makefile: [packages] tor: update to 0.2.1.30, add missing librt dependency after r25548, use secure urls Mar 02 07:43:48 oh... Android 3.0 Honeycomb is out :) Mar 02 07:53:11 oh, and Qt for android :) http://sourceforge.net/p/necessitas/home/ Mar 02 11:43:58 kaloz * r25820 /trunk/target/linux/gemini/patches/ (003-missing_from_upstream.patch 006-wbd111_usb.patch): [gemini]: add usb support for the wbd111, refresh patches Mar 02 11:46:02 gmorning Mar 02 12:01:54 nbd * r25821 /trunk/package/hostapd/Makefile: hostapd: use -ffunction-sections, -fdata-sections and --gc-sections, reduces binary size a bit Mar 02 12:04:27 nbd * r25822 /packages/libs/libtorrent/Makefile: libtorrent: fix libgcc linking command Mar 02 12:04:30 nbd * r25823 /packages/libs/ustl/files/Common.mk: ustl: fix libgcc linking Mar 02 12:04:31 nbd * r25824 /packages/utils/valgrind/Makefile: valgrind: fix libgcc linking Mar 02 12:04:33 nbd * r25825 /packages/utils/gnuplot/Makefile: gnuplot: fix libgcc linking Mar 02 12:23:41 build #81 of ps3 is complete: Failure [failed compile_3] Build details are at http://tksite.gotdns.org:8010/builders/ps3/builds/81 Mar 02 12:45:47 build #76 of x86 is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/x86/builds/76 Mar 02 12:48:39 nbd * r25826 /trunk/package/dropbear/Makefile: dropbear: use -ffunction-sections, -fdata-sections and --gc-sections, saves 19k uncompressed Mar 02 12:48:44 nbd * r25827 /trunk/package/dnsmasq/Makefile: dnsmasq: use -ffunction-sections, -fdata-sections and --gc-sections, saves 8k uncompressed Mar 02 12:48:46 nbd * r25828 /trunk/package/ppp/Makefile: pppd: use -ffunction-sections, -fdata-sections and --gc-sections, saves 5k uncompressed Mar 02 12:48:51 nbd * r25829 /trunk/package/opkg/Makefile: opkg: use -ffunction-sections, -fdata-sections and --gc-sections Mar 02 13:46:16 build #83 of pxcab is complete: Failure [failed shell compile_3] Build details are at http://tksite.gotdns.org:8010/builders/pxcab/builds/83 Mar 02 14:03:30 nbd: heh, I discovered the space-saving virtues of those args with u-boot :) Mar 02 14:21:37 kaloz * r25830 /trunk/package/dropbear/ (4 files in 2 dirs): [package/dropbear]: upgrade to 0.53.1 Mar 02 14:52:32 kaloz * r25831 /trunk/package/dropbear/ (3 files in 2 dirs): [package/dropbear]: tune some more options by default to decrease size Mar 02 16:21:45 mb * r25832 /trunk/target/linux/omap24xx/patches-2.6.38/ (3 files): cbus-retu-wdt: Some fixes Mar 02 16:23:41 mb * r25833 /trunk/target/linux/omap24xx/patches-2.6.38/900-n810-battery-management.patch: n810bm: Add charger_state attribute Mar 02 16:37:38 Does anyone here know if Mono package (http://mono-project.com) has been ported to OpenWRT? Mar 02 16:38:00 nope Mar 02 16:38:07 it has not been ported Mar 02 16:38:08 xMff: Thanks. Mar 02 16:54:37 Hi, could you tell me where the source from which 10.03.1-rc4 has been compiled is? Mar 02 16:54:52 It does not seem to be tagged in SVN. Mar 02 16:58:29 branches/backfire @ r24045 Mar 02 17:07:15 Thank you! Mar 02 17:07:34 If you have write permission to the repo, would it be difficult to tag it now? Mar 02 17:11:14 well apart from 8.09 we nevver tagged rc releases so I won't do that now Mar 02 17:21:07 Why is that, by the way? I'm just curious. Mar 02 17:22:54 there's little point in recompiling an rc release as it most likely is full of bugs which are already fixed in a later branch revision Mar 02 17:23:10 But at least it is stable. Mar 02 17:23:25 well Mar 02 17:23:47 the only thing that is not stable is packages/ and that is independant from the core system anyway Mar 02 17:24:00 having a tagged rc release won't help Mar 02 17:24:04 It should be tagged, too, IMHO. A tag does not cost anything. Mar 02 17:26:17 well feeds.conf should be fixed to a specific packages rev but I strongly oppose tagging the feed for an rc release Mar 02 17:26:38 Well, right now, my problem is that a certain hardware does not work with the 10.03 release but works with 10.03.1-rc4. But I would like to compile a firmware that has a package selection that is different from the binary release. Mar 02 17:26:52 Why the strong opposition, BTW? It does not cost anything. Mar 02 17:27:54 it generates the impression that it is stable while it is just some arbritary freeze, then you get bug reports against it which are worthless as nobody is going to maintain this tag anyway Mar 02 17:28:56 Aren't they release candidates? I mean do you not consider them to be release-worthy unless proven too buggy by testers? Mar 02 17:30:14 yeah, and usually very little changes between RCs so there's no point tagging packages for them as stuff is forward and backwards compatible Mar 02 17:30:16 Tags are not supposed to be maintained (those are branches). Tags are just reference points. Mar 02 17:30:30 is it? Mar 02 17:30:46 no, it's not. Mar 02 17:31:04 backwards and forward compatible i mean... hmm.... no, not always Mar 02 17:31:05 :p Mar 02 17:31:17 yes, recent changes where an exception Mar 02 17:31:55 The trunk often does not even build properly. Mar 02 17:32:06 don't get this wrong, better break compatibility for old stuff than working around limitations that create issues Mar 02 17:32:19 well, it IS work in progress Mar 02 17:32:33 nobody talked about trunk here Mar 02 17:32:48 it's not supposed to be safe Mar 02 17:33:18 or even working... at all... at any given time anything can be broken Mar 02 17:33:21 That's why I would like to work from the latest release candidate. Mar 02 17:33:32 Which is not tagged. :-( Mar 02 17:33:55 branches/backfire is not trunk Mar 02 17:34:02 it is very stable Mar 02 17:34:05 Sorry, I meant that one. Mar 02 17:34:07 RCs are not made from trunk Mar 02 17:34:16 I meant the backfire branch. Mar 02 17:34:52 what makes you think that an older version of the backfire branch compiles if the current one does not? Mar 02 17:35:05 indeed Mar 02 17:35:36 The fact that it has been compiled for the release candidate. Mar 02 17:35:47 it usually means missing dependencies or less likely an unreachable web address Mar 02 17:36:16 or you just forgot to clean the distr and update/install the packages Mar 02 17:37:07 cleanup everything, update/install feeds, config and compile, should work Mar 02 17:37:28 You mean the backfire branch? Mar 02 17:37:37 yeah Mar 02 17:40:52 if some package from the feeds was updated side by side with a package from trunk, forgetting to update both could result in issues Mar 02 17:41:11 like removed patches trying to be applied Mar 02 17:41:47 or new ones not being able to Mar 02 17:41:51 etc etc Mar 02 17:54:53 nbd * r25834 /trunk/toolchain/gcc/patches/linaro/860-disable_extension_elimination.patch: gcc: disable extension elimination, it causes hostapd to be miscompiled on mips Mar 02 18:24:05 build #94 of s3c24xx is complete: Failure [failed compile_10] Build details are at http://tksite.gotdns.org:8010/builders/s3c24xx/builds/94 Mar 02 19:21:18 jow * r25835 /trunk/package/firewall/ (Makefile files/lib/core_init.sh): [package] firewall: move include sourcing into a subshell, this makes the firewall init immune against exit in the include scripts Mar 02 19:21:54 jow * r25836 /branches/backfire/package/firewall/ (Makefile files/lib/core_init.sh): [backfire] merge r25835 Mar 02 20:30:09 got this bluetooth error when building for wzr-hp-g300nh (ar913x): http://openwrt.pastebin.com/Qay5pFsz Mar 02 20:30:24 perhaps the name of a kernel module has changed Mar 02 20:45:38 nope, the kernel module was just not built... Mar 02 20:46:00 maybe a new dependency Mar 02 20:47:16 ahh, I see 'depends on BT && BT_L2CAP && INPUT && HID_SUPPORT' Mar 02 20:50:02 hello, i got an error while compiling for brcm63xx -> http://pastebin.com/DwGEhy5G Mar 02 20:50:24 does anyone know a solution to this? Mar 02 20:51:00 you disabled large file support? Mar 02 20:51:09 solution would be to enable it Mar 02 20:52:52 xMff: Indeed, is it disabled by default? Because i didn't change much Mar 02 20:53:01 no its not Mar 02 21:02:11 xMff: it seems like CONFIG_HID_SUPPORT is not defined Mar 02 21:02:28 I will try to include usb-hid in my config file Mar 02 21:03:39 :q Mar 02 21:03:47 Mar 02 21:09:19 nbd * r25837 /trunk/package/mac80211/files/lib/wifi/mac80211.sh: (log message trimmed) Mar 02 21:09:20 mac80211: allow AP configuration of beacon interval, DTIM period, maximum permissible STA listen interval, and basic rates Mar 02 21:09:20 This introduces beacon_int and basic_rate (per wifi-device), and Mar 02 21:09:20 dtim_period and max_listen_int (per wifi-iface) for mac80211. These Mar 02 21:09:20 configure the beacon interval, basic rate specification, DTIM period Mar 02 21:09:20 (one DTIM per this many beacon frames), and maximum listen interval Mar 02 21:09:20 that a STA will be permitted to associate with. All of the new Mar 02 21:10:18 huge update Mar 02 21:12:53 xMff: should $(call AddDepends/hid) set CONFIG_HID_SUPPORT ? Mar 02 21:13:18 netprince: I think you should rather list it in KCONFIG Mar 02 21:16:06 do you mean in the file linux-2.6.37.1/net/bluetooth/hidp/Kconfig? I would could set 'select HID && HID_SUPPORT' Mar 02 21:16:29 no in package/kernel/modules/*.mk Mar 02 21:17:07 ahh yes, thats what I was looking at, it has $(call AddDepends/hid) in the bluetooth section Mar 02 21:18:52 ok, I think I see what you mean Mar 02 22:38:09 fyi: I'll be working on the miniupnp stuff next time I get a chance to work on OpenWRT...I've been sick, but did manage to get a couple of commits in on the weekend Mar 02 23:40:18 I'm getting some unknown symbols during boot, trunk r25837 on ar71xx/rspro, http://paste.openwrt.org/m62946377 Mar 02 23:40:39 when I manually insmod act_connmark after boot, that works, so that's probably an issue with module loading order ? Mar 03 00:18:00 nbd * r25838 /trunk/include/package.mk: add a stamp that can forcibly disable autorebuild for a specific package - useful for testing changes to packages with extremely long recompilation times Mar 03 00:19:13 nbd * r25839 /branches/backfire/include/package.mk: merge autorebuild override change from r25838 Mar 03 00:33:43 Did anyone notice that quilt is broken in the current head of backfire branch? Mar 03 00:39:11 epointsystem: did you ever figure out that mounting of whole block devices? I tried to get in touch with you, and tend to forget about it until I see your name come up. Mar 03 00:42:15 loswillios * r25840 /packages/sound/pulseaudio/ (Makefile files/ files/pulseaudio.init): [pulseaudio] add init script Mar 03 00:46:49 cshore: I remember the episode, but I forgot what the solution ended up being. Mar 03 00:47:21 But it is an interesting question, I will look up what I have done. Mar 03 00:47:53 Thanks. I'd like to add to trunk if it's generic enough Mar 03 00:48:14 And like I said, I tend to forget about it Mar 03 00:48:21 so haven't worked on it Mar 03 00:54:01 Today I was quite frustrated by my inability to reproduce the release candidate binaries. They seem to contain configurations that have never been checked into the repository... Mar 03 00:54:25 For example, where do the time server settings in rc4 come from? Mar 03 00:57:39 The package selection is also a mistery. It's certainly not the default selection after picking a platform and a configuration. Mar 03 01:01:52 epointsystem: https://dev.openwrt.org/browser/branches/backfire/package/base-files/files/etc/config/system?rev=20728 Mar 03 01:02:25 epointsystem: package selection has been always target + luci + config_all - ebtables Mar 03 01:05:42 BTW, while navigating around in LuCI I sometimes get uncommitted changes about rdate even though I don't touch anything. Very strange. Mar 03 01:07:23 hmm Mar 03 01:07:32 you sure? Mar 03 01:07:54 might need to investigate that Mar 03 01:08:03 Pretty sure. Mar 03 01:08:53 there are certain place where I let it create a uci section when the page is opened Mar 03 01:09:26 basically to prepopulate a section that might not be there from the get-go but where I do not want to allow add/remove either Mar 03 01:09:46 rdate might be one of those, dhcp on the interface admin page is the other Mar 03 01:10:07 I suppose an "addonly" attribute is in order... Mar 03 01:11:44 ah yes - http://luci.subsignal.org/trac/browser/luci/branches/luci-0.9/modules/admin-full/luasrc/model/cbi/admin_system/system.lua#L22 Mar 03 01:11:55 this is the thing you see Mar 03 01:12:46 Hmm, the revision log says that you touched quilt 7 months ago. Right now, if you check out the repository and try building it for the ar71xx architecture, it breaks at comiling quilt. I somehow doubt that this has been an issue for the past 7 months. Has anyone changed anything in the build system? Mar 03 01:12:46 I am talking about the backfire branch. Mar 03 01:13:27 Indeed. Mar 03 01:14:57 no Mar 03 01:15:17 I built backfire very recently in order to backport autotools stuff and I did not notice any issue with quilt Mar 03 01:15:26 whats the nature of the problem? Mar 03 01:15:39 non fitting patches, makefile foo, compilation errors? Mar 03 01:16:04 Compilation error, I believe, but let me do a clear-room experiment. Do you have the time to wait for it? Mar 03 01:16:23 I can't sleep so I am likely around for a while :) Mar 03 01:17:04 but can you paste the current error? Maybe I recall it Mar 03 01:17:09 Okay, so I start with svn co svn://svn.openwrt.org/openwrt/branches/backfire Mar 03 01:17:43 k Mar 03 01:17:46 Then change the feed sources to LuCI 0.9 Mar 03 01:18:04 And comment out Xwrt. Mar 03 01:19:07 ok Mar 03 01:19:20 Checked out revision 25840. Mar 03 01:20:40 src-svn luci http://svn.luci.subsignal.org/luci/tags/0.9.0/contrib/package Mar 03 01:21:02 ok Mar 03 01:21:47 ./scripts/feeds update Mar 03 01:21:47 ./scripts/feeds install -a Mar 03 01:24:23 right Mar 03 01:24:37 make menuconfig # Target System (Atheros AR71xx...) # Target Profile (Ubiquiti Products) Mar 03 01:25:39 thats difference one Mar 03 01:25:40 Many warnings about recursive dependencies, but let's ignore them for now. Mar 03 01:25:51 release images only do generic profiles Mar 03 01:26:03 just saying Mar 03 01:26:12 the recursive dependencies are known Mar 03 01:26:39 make V=99 Mar 03 01:27:08 ok Mar 03 01:28:18 The generic profile is a superset of the specialized profiles, isn't it? Mar 03 01:29:27 kind of Mar 03 01:29:44 basically profilÃes are just different package selections Mar 03 01:31:09 ping loswillios Mar 03 01:32:07 loswillios: your r25840 is missing a backslash in the makefile and the initscript seems to include binary garbage Mar 03 01:35:01 build #70 of etrax is complete: Failure [failed compile_3] Build details are at http://tksite.gotdns.org:8010/builders/etrax/builds/70 Mar 03 01:38:36 quilt building started... Mar 03 01:39:30 Right. But the generic profile has all the packages that specialized profiles have, doesn't it? Mar 03 01:39:42 in most cases yes Mar 03 01:40:59 Hmm. It seems like quilt finished building. Dentist effect... Mar 03 01:42:59 Strangely, in Ubiquiti products profile, there is no WiFi driver included; neither ath9k nor madwifi. The generic profile has both. Mar 03 01:46:25 jow * r25841 /trunk/rules.mk: [buildroot] add an abstract $(LIBRPC_DEPENDS) variable for feed packages to rely on Mar 03 01:48:37 jow * r25842 /packages/ (7 files in 7 dirs): [packages] use LIBRPC_DEPENDS (#8962) Mar 03 01:51:17 gcc 4.3.3 building Mar 03 02:16:43 build #77 of ar7 is complete: Failure [failed compile_7] Build details are at http://tksite.gotdns.org:8010/builders/ar7/builds/77 Mar 03 02:24:40 build #69 of ep93xx is complete: Failure [failed compile_4] Build details are at http://tksite.gotdns.org:8010/builders/ep93xx/builds/69 Mar 03 02:28:50 ^ missing kernel symbol(s) Mar 03 02:29:24 ah the rtc crap Mar 03 02:32:39 jow * r25843 /trunk/target/linux/generic/ (8 files): [kernel] add missing symbol Mar 03 02:36:27 and RTC_DRV_GENERIC (ppc44x/2.6.32.29, mpc52xx/2.6.30.10), RTC_DRV_PL030 (adm5120_router_le/2.6.32.29), IRQ_TIME_ACCOUNTING (rdc/2.6.32.29) Mar 03 02:40:18 and orbit host alignment (at91, avr32, x86) and gcc regrename (cris/powerpc) and ... Mar 03 02:40:49 yeah Mar 03 02:41:10 can't fix the toolchain issues but planned to look into orbit Mar 03 02:53:50 the etherpuppet tarball needs to be pushed to a mirror **** ENDING LOGGING AT Thu Mar 03 02:59:57 2011