**** BEGIN LOGGING AT Sat Apr 17 02:59:56 2010 Apr 17 08:47:00 build #42 of ixp4xx is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/ixp4xx/builds/42 Apr 17 08:49:02 file caled e2fsprogs-1.41.11.tar.gz is missing from http://downloads.openwrt.org/sources/ and sourceforge.net page isnt responding right now so compiling from that packages isnt posible : http://paste.debian.net/69345/ hope someone can add it to openwrt downloads source page :) Apr 17 08:49:25 -from Apr 17 08:59:52 build #41 of pxcab is complete: Failure [failed compile_3] Build details are at http://tksite.gotdns.org:8010/builders/pxcab/builds/41 Apr 17 09:24:21 build #42 of kirkwood is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/kirkwood/builds/42 Apr 17 09:29:29 juhosg * r20951 /trunk/target/linux/ppc40x/patches/004-magicbox.patch: ppc40x: fix MagicBox partition table Apr 17 09:29:31 juhosg * r20952 /trunk/target/linux/generic-2.6/patches-2.6.32/ (023-mips-fix-kexec.patch 651-swconfig-2.6.32-fix.patch): kernel: refresh 2.6.32 patches Apr 17 09:36:00 thepeople: does buildbot have an errors log file output or a brief (tail -100) stdio output? Apr 17 09:38:21 juhosg * r20953 /branches/backfire/target/linux/ppc40x/patches/004-magicbox.patch: backfire: ppc40x: fix MagicBox partition table (backport of r20951) Apr 17 09:43:17 juhosg * r20954 /branches/backfire/target/linux/ar71xx/patches-2.6.32/902-mips_clocksource_init_war.patch: backfire: ar71xx: nuke clocksource init patches, it is not required since 2.6.27 (backport of r20764) Apr 17 09:47:16 regarding that e2fsprogs-1.41.11.tar.gz file after adding it manula to backfire/dl/ dir everything else was fine to compile backfire current svn Apr 17 10:18:50 Is it possible to define custom rule ovverrides in the profiles? Like, I want an additional kernel patch for a profile. Apr 17 10:31:34 build #42 of orion is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/orion/builds/42 Apr 17 10:50:38 Also, is there a way to specify, that a profile wants some of the external packages? Apr 17 11:16:48 xl0: external = other package repository? Apr 17 11:20:03 joda_bot: Yes, the feeds thing. Apr 17 11:22:32 xl0: I don't think you can include a kernel patch just for single profile, but you could guard the changes through a kernel config option, and then create a kernel module that enables it, which your profile selects Apr 17 11:23:24 xl0: I do it for the rspro rtc stuff, see http://page.mi.fu-berlin.de/jgorski/openwrt/wip/trunk/0007-ar71xx-Enable-RTC-support-for-the-RouterStation-Pro.patch Apr 17 11:24:55 KanjiMonster: Thank you. Problem is, the patch is quite invasive, and in fact there a multiple such patches. But it's a vendor tree, so I'll probably just land them in the platform dir. ;( Apr 17 11:25:25 ah, I see Apr 17 11:28:20 Hmm, is downloads.sourceforge.net really dead? Apr 17 11:30:55 slashdot said "Our Geeknet sites upgrade activities are occurring on April 17th and 18th between 0500 and 1300 UTC." somewhen earlier, now its down, too Apr 17 11:31:08 Yeah, see. ;) Apr 17 11:35:31 regarding the package: I don't think anything is stopping you from including it, except that it won't be if the feed wasn't made available Apr 17 11:35:54 Well, you also need to "install" it, right? Apr 17 11:35:58 yeah Apr 17 11:36:04 KanjiMonster: can't you autodetect the rtc or the presence of a rs pro? Apr 17 11:36:08 And that's the annoying part. Apr 17 11:36:22 KanjiMonster: because this kmod solution doesn't strike me as a good idea Apr 17 11:37:10 rtz2: if I do then all other ar71xx targets get a warning in their kernel log that the system time couldn't be set because no rtc was found Apr 17 11:38:00 KanjiMonster: well, you could remove the warning Apr 17 11:38:15 KanjiMonster: and I don't see much of a problem with it Apr 17 11:38:34 KanjiMonster: where exactly does the warning come from? Apr 17 11:39:12 rtz2: the kernel is configured to automatically set the system time from the rtc, and if it doesn't find one, prints the warning (or if the time is invalid) Apr 17 11:40:09 KanjiMonster: which config symbol enables it? Apr 17 11:40:56 KanjiMonster: the automatic setting of the clock Apr 17 11:41:32 CONFIG_RTC_HCTOSYS_DEVICE="rtc0" and CONFIG_RTC_HCTOSYS=y in generic-2.6/config-* Apr 17 11:44:08 and it makes sense to have it if you have a rtc Apr 17 11:46:18 KanjiMonster: well, it shouldn't be to hard to turn it into a module Apr 17 11:47:10 KanjiMonster: the sys and class parts already are modules as far as I can see Apr 17 11:47:33 KanjiMonster: and even if it isn't possible, I don't see that much of a problem with the error message Apr 17 11:48:11 KanjiMonster: hmm, wait Apr 17 11:48:19 KanjiMonster: isn't it already one? Apr 17 11:49:40 you mean the rtc stuff? As I did it no, since its a platform device (connected through spi) Apr 17 11:50:41 KanjiMonster: well, that's no reason, why the driver can't be a module Apr 17 11:51:20 KanjiMonster: from what I can see, the whole rtc subsystem can be compiled as one Apr 17 11:51:22 I don't know much about the spi/kernel stuff; making it a platform device was the easiest Apr 17 11:52:10 KanjiMonster: sure Apr 17 11:53:05 KanjiMonster: but the drivers for platform devices still can be modules Apr 17 11:53:13 also automatically setting the system clock probably wouldn't work then anymore, or would it still work? Apr 17 11:54:17 KanjiMonster: i don't see, why not Apr 17 11:55:03 because its done by the kernel before loading (pre)init Apr 17 11:55:33 KanjiMonster: well, why can't it happen a bit later? Apr 17 11:57:28 then it must be done by some script as far as I see Apr 17 11:58:22 "The driver for this RTC device must be loaded before late_initcall functions run, so it must usually be statically linked." Apr 17 11:59:34 (for HCTOSYS) Apr 17 11:59:34 KanjiMonster: hmm, i'm not sure, if you can work around it Apr 17 12:00:00 KanjiMonster: but I also don't see such a big problem with the warning message Apr 17 12:01:11 yeah, but I thought it would be nice to not make the kernel bigger for all other devices that don't even have use for the whole rtc stuff ;) Apr 17 12:04:00 KanjiMonster: don't you still have to link it into the kernel with your solution? Apr 17 12:07:26 rtz2: only when selecting the rs pro profile Apr 17 12:09:35 rtz2: http://page.mi.fu-berlin.de/jgorski/openwrt/wip/trunk/0006-Add-support-for-the-RTC-on-the-Routerstation-Pro.patch is enough to make it work as soon as the driver is included (apart from some additional modifications to the spi master driver itself) Apr 17 12:24:35 hello guys :) Apr 17 12:39:11 hi stintel Apr 17 12:39:29 I messed up again Apr 17 12:39:49 I'm just gonna stop trying Apr 17 13:09:27 What happens if a package is present in more than one feed? How are the priorities set? Apr 17 13:14:30 <{Nico}> xl0: you can choose to install a package from a specific feed with the -p option Apr 17 13:16:02 {Nico}: Thanks. Is there a way to set default feed priorities? Apr 17 13:16:21 acinonyx * r20955 /packages/net/ (4 files in 3 dirs): Apr 17 13:16:21 [packages] mini_snmpd: Merge mini_snmpd_ipv6 with mini_snmpd Apr 17 13:16:21 - add ipv6 module dependency if IPV6 support is selected Apr 17 13:16:21 - update ipv6 patch to compile IPv6 support conditionally Apr 17 13:16:21 - cleanup Makefile Apr 17 13:16:21 - refresh patches Apr 17 13:17:21 <{Nico}> xl0: not yet Apr 17 15:32:57 build #42 of rb532 is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/rb532/builds/42 Apr 17 16:23:46 I'm getting a DHCP lease with a gateway on a different subnet, and openwrt refuses it. Apr 17 16:24:50 ex. IP 188.81.188.32; mask 255.255.0.0; gw 194.65.169.218 Apr 17 16:25:45 I think that's because route gives an error when trying to add this gateway on the different subnet, first I have to add a regular route to 194.65.169.218 255.255.255.255 on eth1 Apr 17 16:26:16 ubuntu accepts it Apr 17 16:26:34 udhcpc or route bug? Apr 17 16:50:02 nico * r20956 /packages/net/mini_snmpd/Makefile: [packages] mini_snmpd: build 2 variants, one for IPv4 and one for IPv6 Apr 17 17:12:11 nico * r20957 /branches/backfire/target/linux/brcm47xx/patches-2.6.32/610-pci_ide_fix.patch: [backfire] backport r20924 Apr 17 17:19:40 nico * r20958 /packages/multimedia/fswebcam/ (. Makefile): [packages] add fswebcam (closes: #7177) Apr 17 17:28:30 nico * r20959 /branches/backfire/toolchain/gcc/patches/3.4.6/701-pr16276-fix.patch: [backfire] backport r20900 Apr 17 17:45:57 acinonyx * r20960 /trunk/package/dropbear/files/dropbear.init: package] dropbear: Add bind 'Interface' option (#7149) Apr 17 17:47:43 I've seen its not a route bug - NOTE: The specified gateway must be reachable first. This usually means that you have to set up a static route to the gateway beforehand. Apr 17 17:48:31 I've no idea where is udhcpc doing the "route add default gw 194.65.169.218 eth1" Apr 17 17:49:12 I think we need to check first if that address is reachable, and if not doing first "route add -host 194.65.169.218 dev eth1" Apr 17 17:49:33 my question is: which file should I look at? Apr 17 17:52:52 nothing to do with udhcpc, it also happens with a static interface configuration, maybe this way I can find it... Apr 17 18:26:30 What sort of dev work is done in OpenWRT? Apr 17 18:27:07 Porting applications? Apr 17 18:31:51 pretty much any aspect of embedded hacking. platform porting, toolchain work, package porting, writing drivers, userland integration ... Apr 17 18:32:29 What would be suitable for beginners? Package porting? Apr 17 18:32:58 yes Apr 17 18:33:01 testing Apr 17 18:33:03 bug fixing Apr 17 18:33:06 work on the wiki Apr 17 18:36:14 How should I start? By deeply familiarizing oneself with the platform, then contribute to the wiki? Apr 17 18:36:22 yes Apr 17 18:49:55 Hmmm. strace isn't building in trunk for mipsel. Unless anyone knows of a fix, I'll start working on it. Apr 17 19:00:13 <{Nico}> noz: i have a fix, wait a sec Apr 17 19:00:37 k Apr 17 19:06:45 nico * r20961 /packages/utils/strace/Makefile: [packages] strace: only change CPPFLAGS when building with an external toolchain (see [19770]), fixes a build failure on mips/mipsel and probably others Apr 17 19:06:53 <{Nico}> noz: here you go :) Apr 17 19:13:05 build #40 of ramips is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/ramips/builds/40 Apr 17 19:13:58 {Nico}: Thanks. Would have taken me a while to find that. Apr 17 19:20:16 acinonyx * r20962 /packages/utils/ (pciutils/Makefile usbutils/Makefile): [packages] {pciutils,usbutils}: Fix updating of IDs when installing on destination other than root (#7077) Apr 17 19:27:53 build #40 of ps3 is complete: Failure [failed compile_3] Build details are at http://tksite.gotdns.org:8010/builders/ps3/builds/40 Apr 17 19:39:58 {Nico}: Want another one? ;-) Apr 17 19:40:05 make[2]: *** No rule to make target `package//host/compile', needed by `package/feeds/packages/avahi/compile'. Stop. Apr 17 19:40:31 I've looked at the Makefile for avahi, and nothing seems out of place. Apr 17 19:41:46 Do I need to do some sort of feeds update? Apr 17 20:01:55 build #41 of rdc is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/rdc/builds/41 Apr 17 20:23:37 acinonyx * r20963 /packages/net/lighttpd/Makefile: [packages] lighttpd: Fix missing IPv6 support (#7173) Apr 17 20:40:12 acinonyx * r20964 /packages/net/lighttpd/Makefile: [packages] lighttpd: Bump release number Apr 17 20:47:51 Acinonyx: did usbutils only get a partial fix in r20962? Apr 17 20:48:03 yes Apr 17 20:49:00 unfortunately, usbutils don't have an option to specify file so ID database is always installed on root /usr/share Apr 17 20:49:29 also, it refuses to display anything if database is missing Apr 17 20:52:32 build #38 of ubicom32 is complete: Failure [failed compile_4] Build details are at http://tksite.gotdns.org:8010/builders/ubicom32/builds/38 Apr 17 20:55:35 goto go. I am not feeling too well today... Apr 17 20:55:37 gnight Apr 17 22:01:39 ~. Apr 17 22:01:51 argl - schon wieder Apr 17 22:30:27 build #41 of au1000 is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/au1000/builds/41 Apr 17 22:36:08 cshore i just testet that patch to go back to flash boot, when the usb stick is unpluged and it works just fine :) (block-exroot) Apr 17 23:44:52 nico * r20965 /packages/net/strongswan4/ (11 files in 3 dirs): [packages] add strongswan4 (closes: #2538) Apr 17 23:58:12 nbd * r20966 /trunk/package/mac80211/ (9 files in 2 dirs): mac80211: update to wireless-testing 2010-04-16 Apr 17 23:58:19 nbd * r20967 /trunk/package/mac80211/patches/550-ath9k_bb_fix.patch: ath9k: improve reliability of the rx stuck fixup Apr 17 23:58:24 nbd * r20968 /trunk/package/mac80211/patches/ (4 files): ath9k: enable 3-stream and MCS16-23 support Apr 17 23:58:27 nbd * r20969 /trunk/package/mac80211/patches/530-minstrel_ht.patch: mac80211: enable 3-stream support in minstrel_ht Apr 17 23:58:34 nbd * r20970 /trunk/package/mac80211/patches/530-minstrel_ht.patch: mac80211: minstrel_ht - increase space between sampling attempts to allow for better aggregation Apr 17 23:58:38 nbd * r20971 /trunk/package/mac80211/Makefile: mac80211: use minstrel_ht by default for ath9k Apr 17 23:58:46 nbd * r20972 /trunk/package/mac80211/Makefile: mac80211: remove the unused PACKAGE_ATH9K_DEBUG option - PACKAGE_ATH_DEBUG turns on debugging for atheros drivers Apr 18 00:01:55 nbd * r20973 /branches/backfire/package/mac80211/patches/550-ath9k_bb_fix.patch: ath9k: backport the rx stuck bugfix change from r20967 Apr 18 00:01:57 nbd * r20974 /branches/backfire/package/mac80211/Config.in.ath9k: ath9k: backport the config option removal from r20972 Apr 18 00:56:34 nbd * r20975 /trunk/package/mac80211/patches/530-minstrel_ht.patch: mac80211: add ldpc and stbc support to minstrel_ht Apr 18 00:56:40 nbd * r20976 /trunk/package/mac80211/patches/ (3 files): ath9k: add stbc support Apr 18 00:57:20 nico * r20977 /packages/net/strongswan4/Makefile: [packages] strongswan4: rebuild when plugin selection is changed **** ENDING LOGGING AT Sun Apr 18 02:59:57 2010