**** BEGIN LOGGING AT Mon Dec 10 02:59:58 2012 Dec 10 06:56:53 <[florian]> swalker: thanks for python-crypo Dec 10 10:38:54 juhosg r34588 trunk/target/linux/ (6 files in 4 dirs) * ar71xx: ar934x_nfc: allow to control DMA data swap via platform data Dec 10 10:38:55 juhosg r34589 trunk/target/linux/ar71xx/modules.mk * ar71xx: load ar934x_nfc driver earlier Dec 10 12:12:55 is it just me, or are a whole bunch of packages and kernel modules automatically included when I choose "show broken packages" ? Dec 10 12:13:06 I'm not sure if I'm reading the depends list properly Dec 10 12:17:43 hmm, might be stuff related to mac80211, nvm. Dec 10 12:17:49 diffconfig doesn't show it up anymore Dec 10 13:44:31 Delboy, mesx: looking at your sx76x boards, can you guys maybe help me find what pull-up is missing for serial on my sx762? Dec 10 13:51:37 nbd r34590 trunk/package/mac80211/patches/300-pending_work.patch * mac80211: merge fix for carl9170 upstream regression Dec 10 13:51:42 nbd r34591 trunk/package/mac80211/ patches/852-brcmsmac-start-adding-support-for-core-rev-28.patch patches/851-brcmsmac-add-support-for-BCM43224.patch patches/860-brcmsmac-wflush-objmem.patch * mac80211: refresh patches Dec 10 13:51:45 nbd r34592 trunk/package/mac80211/patches/300-pending_work.patch * ath9k: improve signal strength / channel noise reporting accuracy Dec 10 13:51:48 nbd r34593 trunk/package/network/utils/iw/patches/001-nl80211_h_sync.patch * iw: sync with nl80211.h to fix setting antenna gain Dec 10 13:51:52 nbd r34594 trunk/package/mac80211/patches/523-mac80211_configure_antenna_gain.patch * mac80211: fix setting txpower (#12594) Dec 10 13:53:23 <[florian]> win 12 Dec 10 13:53:25 <[florian]> miss Dec 10 14:13:49 blogic r34595 trunk/target/linux/ (14 files in 10 dirs) * [ramips] add support for Hauppauge Broadway Dec 10 14:22:56 hi, Im doing "make kernel_menuconfig", enabling some module. it is correctly saved in target/linux/raspberrypi/config-3.1 (or config-default if I change). but it is not populationg to kernel's .config , thus its not building. I need advice where should I look yet ? Dec 10 14:23:19 what module? Dec 10 14:23:42 (all modules as kmod* are disabled) Dec 10 14:24:10 some usb net card driver Dec 10 14:24:24 you're not supposed to use make kernel_menuconfig to enable building kernel modules Dec 10 14:25:00 florian r34596 trunk/toolchain/musl/Config.in * toolchain/musl: fix typo on configuration symbol Dec 10 14:25:01 make a kmod-* package for the module if there isn't one already Dec 10 14:25:03 then select that Dec 10 14:25:05 florian r34597 trunk/toolchain/ (7 files in 2 dirs) * toolchain/musl: update to 0.9.8 Dec 10 14:25:07 florian r34598 trunk/toolchain/musl/Makefile * toolchain/musl: simplify musl install steps Dec 10 14:25:11 florian r34599 trunk/target/linux/generic/patches-3.6/811-pci_disable_usb_common_quirks.patch * kernel/generic: fix build failure on missing uhci symbol Dec 10 14:25:24 but I want to. it was always working ok. modules was building , .ko files was present in kernel's build directory. I know it was not installed into firmware. Dec 10 14:25:45 why do you want to do it the unsupported way? Dec 10 14:25:57 because it was always working okay ;) Dec 10 14:26:00 Im building this way modules for things which are not present as kmod packages. it was always ok. Dec 10 14:26:08 indeed :) Dec 10 14:26:28 some people prefer to spend medium effort 100 times isntead of a slightly bigger effort once Dec 10 14:26:30 so how about doing it the right way and submitting a patch for that Dec 10 14:26:40 makes them feel alive or so Dec 10 14:27:13 [florian]: that last fix might need porting to 3.7 Dec 10 14:27:21 <[florian]> nbd: oh right Dec 10 14:27:28 we're on 3.7 already? Dec 10 14:27:34 we have generic patches for it Dec 10 14:27:45 jow_laptop : actually I'v spend some time on something different - I build 'package' which installs all modules from kernel tree into firmware, instead using kmods. its way better for me. Dec 10 14:28:14 voyo: and now you need to spend even more time on fixing the unsupported hack :) Dec 10 14:28:45 jow_laptop : we are all hackers. thats our life ;) Dec 10 14:29:22 so, little hint where I need to dig to restore previous behaviour of owrt ? Dec 10 14:29:23 what a cop-out Dec 10 14:29:28 voyo: you won't get any help from us on keeping your unsupported hackery Dec 10 14:29:43 voyo: because that would be waste of our time Dec 10 14:29:56 voyo: i told you how you can make your kernel module stuff working with little effort Dec 10 14:30:03 voyo: to me it sounds liek you just forgot to run make target/linux/prepare Dec 10 14:30:06 voyo: so either do it this way, or stop asking ;) Dec 10 14:30:32 jow_laptop: don't encourage it Dec 10 14:30:46 you guys are awesome, i just made a make on trunk today and its at 3.6.9 ! Dec 10 14:30:53 shame. I was hoping maybe someone who did this is here on irc, and can quickly give me a hint Dec 10 14:31:14 jow_laptop : hm, tnx. I tried with clean/compile only. maybe thats it.. Dec 10 14:31:24 voyo: the hint is "don't do it" Dec 10 14:31:45 voyo: see if you change the kernel config to enable modules, things defined by kmod-* packages can override that Dec 10 14:31:59 so with some modules it'll appear to work, whereas with others it won't work because the kmod-* selection overrides it Dec 10 14:32:17 so as long as you try to work around the system instead of working with it, you'll run into these sorts of traps Dec 10 14:32:27 DonkeyHotei : Im not typical user, I have reasons. Dec 10 14:32:34 and while you feel like you may save some work in the short term, you'll end up spending more time debugging such issues Dec 10 14:33:04 and as somebody who has written most of the code related to kernel config handling, i will not change the code to support such hackery because it's not worth it Dec 10 14:33:24 voyo: your reasons need to be adapted to the way things work Dec 10 14:33:46 nbd yes, I realize that. but it is just better for me this way Dec 10 14:34:14 no, it really isn't Dec 10 14:34:19 'I have reasons' typically means 'I don't explain my reasons, because otherwise somebody might be able to prove me wrong and I don't want to change anything' Dec 10 14:34:38 if we all would adapt to the reality, then no one would invent the wheel ;) Dec 10 14:35:07 voyo: you're doing the exact opposite Dec 10 14:35:36 you're taking a wheel and deciding it's wrong to use Dec 10 14:36:29 Im thinking rather - I took the wheel, and improved it with spikes to use on my icy road ;) Dec 10 14:36:57 this is not an improvement Dec 10 14:37:00 nah, you're duct taping the wheel to your car instead of putting it on properly, and now you're complaining to us that it keeps falling off ;) Dec 10 14:37:43 and i'm telling you how to put it on properly, but you say you have your reasons for wanting to use duct tape instead Dec 10 14:37:44 you're taking existing steel spikes, removing them, and putting on new copper ones Dec 10 14:37:46 ;) Dec 10 14:38:34 steel does better on ice than copper does Dec 10 14:40:23 nbd Im not complaining! , Im not telling you are doing it wrong! no way. Im just doing it my way. Dec 10 14:40:42 but i'm telling you why your way leads to the issues that you're having Dec 10 14:40:46 Why can not you respect that? Dec 10 14:41:03 and this may be a bad time for pedantry, but it should have been duck tape, not duct tape, since it was originally made from cotton duckcloth and never had anything to do with heating ducts Dec 10 14:41:03 because he is right Dec 10 14:41:14 dirty hack lead to dirty problems Dec 10 14:41:19 *hacks Dec 10 14:42:37 that was no my question, I only asked what possibly has changed, I dont need sermon. I didnt asked to restore previous behavior from us, I can do everything myself. I only asked for small hint where should I loook :-/ Dec 10 14:42:53 indeed, it's much better to have clean problems than to have to wash the dirty ones Dec 10 14:43:33 actaally what is reason to use kmod instead enabling modules yourself ? Dec 10 14:44:02 or at least allowing powerusers to do this Dec 10 14:44:23 there are several reasons: Dec 10 14:44:51 first, hacking around in the kernel config creates useless delta of the local tree against svn (assuming the target is in svn) Dec 10 14:45:31 second, the build system already controls variables related to kmod-* building as to avoid building redundant stuff Dec 10 14:45:36 anything that isn't packaged is redundant Dec 10 14:46:05 third, i think it's useful to encourage devs to do the right thing instead of insisting to stick with the wrong thing Dec 10 14:46:16 so i won't add extra complexity to the core just for supporting the wrong thing Dec 10 14:47:24 kernel configs are target-specific, but kmod pkgs are part of the generic core and don't all have to be maintained separately for each target Dec 10 14:48:13 ok, I understand that all. makes sense. for building (and maintenancing) owrt. Dec 10 14:49:32 nbd: btw, there is still a pkg kmod-gpio-dev but the module itself seems to have been removed from the kernel tree in recent versions Dec 10 14:50:21 voyo: also note that you are asking for dev time for (helping) fixing problems not present in normal/intended openwrt use Dec 10 14:52:09 KanjiMonster : yes indeed, as I said above I was hopping that someone who has made this change is here online and can answer quickly for precise and stright question. I was not intended to fight with any one, or asking devs to do anything permament with owrt. Dec 10 14:54:11 voyo: this assumes that whoever made the change affecting you is even aware that this change affects you - and still requires time for looking into it Dec 10 14:54:14 no one in their right mind would make this change, so hoping such a person would be here is not going to be fruitful Dec 10 14:54:51 sorry I wasted your precious time. Dec 10 14:55:27 I will never ask for anything here again. once again very sorry. Dec 10 14:56:05 voyo: just expect to be lectured if you are "holding it wrong" Dec 10 14:56:19 oh, you're free to ask, just make sure you properly listen to the feedback that you get and accept it Dec 10 14:57:46 "what color is the sky?" "it's blue" "i don't care if you say it's blue, because it's yellow to me" is not going to fly Dec 10 15:02:19 arguing with you lead me to nowhere, its wasting of all our time, your and mine too. but once again - I asked very precisely and straight question, I was not expecting lecture or homily. Im not 15 years old I dont need advices how to live my life either. Shame you guys are unable to respect someones else oppinion and right to follow his path. Dec 10 15:04:35 florian r34600 trunk/target/linux/generic/patches-3.7/811-pci_disable_usb_common_quirks.patch * kernel/generic: port r34599 over 3.7 kernel Dec 10 15:04:43 florian r34601 trunk/target/linux/generic/patches-3.7/ 100-overlayfs.patch 863-gpiommc.patch 065-8139cp-fixes.patch 330-mips-add-crash-and-kdump-support.patch * kernel/generic: refresh 3.7 patches Dec 10 15:04:53 florian r34602 trunk/target/linux/generic/patches-3.7/067-mips_mt_fix_uidgid_strict_type_check.patch * generic/kernel: add a fix for a common MIPS build failure. Dec 10 15:06:05 usyually better if you dont know the answer or dont want to give it - is just to not say anything. Dec 10 15:06:11 voyo: your question was precisely answered Dec 10 15:06:30 if whatever you selected is covered by a kmod-* the deselecting of kmod-* will overrule your config changes Dec 10 15:06:40 hence the question about what module we're talking about exactly Dec 10 15:07:21 if you hold back the vital info you can't blame us for not precisely answering your question Dec 10 15:07:43 and that was answered before all the "sermon" Dec 10 15:08:17 so your approach will break whenever something you used to enable manually gets packaged as kmod Dec 10 15:08:40 but to find out whether this is the case we'd need to now the module in question Dec 10 15:10:13 jow_laptop : yes, only by you actually, by suggestion to do something (target prepare). and now , by precisely explain this dependence. Dec 10 15:10:41 no it was answered by nbd Dec 10 15:11:00 15:31 < nbd> voyo: see if you change the kernel config to enable modules, things defined by kmod-* packages can override that Dec 10 15:11:06 15:31 < nbd> so with some modules it'll appear to work, whereas with others it won't work because the kmod-* selection overrides it Dec 10 15:13:05 ok, then I didnt understand or it missed me in this mess Dec 10 15:14:00 tnx all guys for your help , I appreciate your time. Dec 10 15:14:54 so, if I want to add a kernel module built in, rather than as a module, do I still need to build a kmod package? or is that use of make kernel_menuconfig still ok? Dec 10 15:16:17 karlp: last time I checked kmod selection state does not override =y Dec 10 15:16:32 the reason I ask, and the reason people might use it, is that this isn't isn't listedn anywhere on the pages linked to here: http://wiki.openwrt.org/doc/start#building.openwrt Dec 10 15:17:41 karlp: there is some code for deteting that and build empty kmod ipks, so dependencies can always be fulfilled Dec 10 15:18:19 so what is the purpose of make kernel_menuconfig ? Dec 10 15:18:28 is it, "never run this, we just haven't deleted it yet" ? Dec 10 15:19:21 it is a developer tool intended to help with kernel porting Dec 10 15:19:47 not strictly a "user tool" since its side effects are not fully defined and its results are not intended to be usable as-is Dec 10 15:20:18 so, is there any docuemenation on "I want to add a kernel module, either built in or as a module, that is not currently selectable in the regular make menuconfig" Dec 10 15:20:44 that seems like it would be helpful to help people "do the right thing" Dec 10 15:26:06 karlp: something like http://wiki.openwrt.org/doc/devel/packages#creating.packages.for.kernel.modules ? Dec 10 15:28:17 that's good, if only there was something that mad eyou think, "I want to include this kernel module, I should go and look at the "how to build a package" page" Dec 10 15:28:38 because this: http://wiki.openwrt.org/doc/howto/build#kernel.configuration is the first thing you'd probably go to Dec 10 15:30:03 also, www.openwrt.org->Developement->Development documenation still takes you to kamikaze build docs, which also suggest using kernel_menuconfig Dec 10 15:34:35 make kernel_menuconfig is useful for making kernel config changes that aren't related to enabling modules Dec 10 15:35:10 but it can also clutter the config with irrellevant symbols Dec 10 15:43:17 nbd r34603 trunk/package/mac80211/patches/300-pending_work.patch * ath5k: fix skb leaks (should fix #12199) Dec 10 15:45:17 nbd r34604 branches/attitude_adjustment/package/mac80211/patches/300-pending_work.patch * ath5k: fix skb leaks (should fix #12199) Dec 10 15:46:56 juhosg r34605 trunk/package/mac80211/patches/301-pending-ath9k-ar9003-fix-OTP-register-offsets-for-AR9340.patch * mac80211: ath9k: fix OTP register offsets for AR9340 Dec 10 15:46:58 juhosg r34606 trunk/ (14 files in 2 dirs) * mac80211: ath9k: allow to load EEPROM data via firmware API Dec 10 16:40:44 hi blogic do you have some time today :) you were mentionig samo xtu bits were broken yesterday Dec 10 16:41:21 well Dec 10 16:41:28 do oyu get a clean pppd disconenct ? Dec 10 16:41:37 or does the unit leave showtime ? Dec 10 16:41:41 its really strange that ppp works from my pc, when i bridge nas0 Dec 10 16:41:51 ah ok Dec 10 16:41:54 so the driver is working Dec 10 16:41:58 its a pppd issue then Dec 10 16:42:32 yes /dsl status is ok, its synced all the time Dec 10 16:42:51 ok Dec 10 16:42:55 then its not the xtu bits Dec 10 16:43:11 start pppd on the unit with the same options you use on your laptop Dec 10 16:43:39 I tested this with three other Sx76x its the same Dec 10 16:43:44 nbd r34607 trunk/package/mac80211/ patches/300-pending_work.patch patches/820-b43-add-antenna-control.patch * b43: fix tx path skb leaks Dec 10 16:44:00 but I remeberd that I had one more, which worked correctly :/ Dec 10 16:44:12 I gave that one to my neigbhoor Dec 10 16:45:46 At that time I haven't had a any idea what was going on. Dec 10 16:45:56 ok Dec 10 16:47:38 cyrus r34608 trunk/package/network/ipv6/6distributed/Makefile * 6distributed: add status-call for luci-integration Dec 10 16:47:42 cyrus r34609 trunk/package/network/ ipv6/ipv6-support/Makefile ipv6/ipv6-support/files/support.sh * ipv6-support: Add support for NPT status tracking Dec 10 16:48:14 jow_laptop: ^ Dec 10 16:49:57 cyrusff: thanks Dec 10 16:50:58 watching at tcpdump when I changed echo_interval to 30... shows no Term Request... Dec 10 16:51:10 could it be that STP is to blame? Dec 10 16:53:19 Can ISP know if I'm using ppp on PC vs ppp on router? Dec 10 16:55:45 * tonikasch ciao Dec 10 17:18:20 AA rc1, what revision is that? 34505? Dec 10 17:18:31 is this the right way to check it out? svn co --revision=34505 svn://svn.openwrt.org/openwrt/trunk/ Dec 10 17:21:13 if you want AA, start with svn://svn.openwrt.org/openwrt/branches/attitude_adjustment/ Dec 10 17:21:41 and I'd suggest to use the latest revision in that branch, it should contain less bugs than the rc1 Dec 10 17:25:57 check Dec 10 18:30:00 nbd r34610 trunk/package/mac80211/ (7 files) Dec 10 18:30:00 mac80211: fix channel context related warnings when using wds ap mode (fixes #12585) Dec 10 18:50:13 nbd r34611 trunk/package/kernel/modules/netsupport.mk * kernel: include the xfrm_algo module on 3.6+ (fixes #12552) Dec 10 19:50:36 build #110 of pxcab is complete: Failure [failed shell_11] Build details are at http://buildbot.openwrt.org:8010/builders/pxcab/builds/110 Dec 10 19:57:20 build #109 of rb532 is complete: Failure [failed shell_11] Build details are at http://buildbot.openwrt.org:8010/builders/rb532/builds/109 Dec 10 20:04:17 build #109 of ppc44x is complete: Failure [failed shell_11] Build details are at http://buildbot.openwrt.org:8010/builders/ppc44x/builds/109 Dec 10 20:28:09 juhosg r34612 packages/multimedia/minidlna/files/minidlna.init * packages/minidlna: create log/db directories defined in the uci configuration Dec 10 20:28:11 juhosg r34613 packages/net/usbip/Makefile * packages: allow to build usbip on 3.6+ Dec 10 21:01:31 jow_laptop? Dec 10 21:05:49 florian r34614 trunk/package/kernel/modules/other.mk * package/kernel: package kmod-input-matrixkmap Dec 10 21:05:54 florian r34615 trunk/target/linux/ep93xx/profiles/01-simone.mk * [ep93xx] also list kmod-sound-core in the default packages Dec 10 21:05:58 florian r34616 trunk/package/kernel/modules/sound.mk * package/kernel: snd-soc-core depends on regmap Dec 10 21:06:10 florian r34617 trunk/target/linux/generic/patches-3.6/066-snd-soc-dmaengine-pcm.patch * generic: backport an upstream fix for building snd-soc-pcm as a module Dec 10 21:06:17 florian r34618 trunk/package/kernel/modules/sound.mk * package/kernel: enable CONFIG_SND_SOC_DMAENGINE_PCM in sound-soc-core Dec 10 21:06:19 florian r34619 trunk/target/linux/ep93xx/modules.mk * [ep93xx] keyboard driver depends on input-matrixkmap for 3.6+ Dec 10 21:06:25 florian r34620 trunk/target/linux/ generic/patches-3.7/067-input_keymap_module_license.patch generic/patches-3.6/067-input_keymap_module_license.patch * generic: backport a fix for the input keymap-matrix module Dec 10 21:06:30 florian r34621 trunk/target/linux/ep93xx/Makefile * [ep93xx] add sound feature Dec 10 21:06:48 florian r34622 trunk/target/linux/ (5 files in 2 dirs) * [ep93xx] add 3.6 kernel support Dec 10 21:06:57 florian r34623 trunk/target/linux/ep93xx/Makefile * [ep93xx] switch to 3.6 kernel Dec 10 22:22:52 is there a config option to build for all targets? Dec 10 22:23:19 CONFIG_ALL=y makes make build all the packages for specified target, right? Dec 10 22:28:20 if you want builds for all targets, you want to be looking at the bot output. That's going to take you an awfully long time Dec 10 22:30:31 ls Dec 10 22:30:43 ctrl-d Dec 10 22:31:00 lol Dec 10 22:31:38 lots of ls? Dec 10 22:31:47 Dec 10 22:38:12 say I'd like to build this http://downloads.openwrt.org/attitude_adjustment/12.09-rc1/ ,my own repository, perhaps with a few changes/default settings for the images Dec 10 22:38:39 you'd want to look at whatever build bot scripts there are, I'd think Dec 10 22:40:49 do you build them one at a time when you put them on site? Dec 10 22:41:05 I don't know how it's done Dec 10 22:49:20 <[florian]> threadlock: each target is built one by one Dec 10 22:56:41 [florian]: manually? Dec 10 22:56:51 sorry, I'm gonna read a bit more documentation Dec 10 22:57:05 I want/need to get into this in more depth either way Dec 10 22:57:38 good luck Dec 10 23:21:46 juhosg r34624 trunk/package/mac80211/patches/305-pending-ath9k-allow-to-load-EEPROM-content-via-firmware-API.patch * mac80211: fix possible NULL pointer dereference in ath9k Dec 11 00:00:59 build #116 of ramips is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/ramips/builds/116 Dec 11 01:02:50 build #113 of uml is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/uml/builds/113 Dec 11 01:30:12 build #114 of ppc40x is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/ppc40x/builds/114 Dec 11 01:36:18 build #100 of sibyte is complete: Failure [failed shell_11] Build details are at http://buildbot.openwrt.org:8010/builders/sibyte/builds/100 Dec 11 01:50:12 build #106 of ar7 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/ar7/builds/106 **** ENDING LOGGING AT Tue Dec 11 03:00:00 2012