**** BEGIN LOGGING AT Thu Mar 03 02:59:58 2011 Mar 03 03:10:48 build #76 of avr32 is complete: Failure [failed compile_7] Build details are at http://tksite.gotdns.org:8010/builders/avr32/builds/76 Mar 03 03:18:11 build #68 of rdc is complete: Failure [failed compile_4] Build details are at http://tksite.gotdns.org:8010/builders/rdc/builds/68 Mar 03 03:37:46 build #91 of ubicom32 is complete: Failure [failed compile_3] Build details are at http://tksite.gotdns.org:8010/builders/ubicom32/builds/91 Mar 03 04:04:41 build #69 of adm5120 is complete: Failure [failed compile_4] Build details are at http://tksite.gotdns.org:8010/builders/adm5120/builds/69 Mar 03 04:43:27 build #95 of atheros is complete: Failure [failed compile_4] Build details are at http://tksite.gotdns.org:8010/builders/atheros/builds/95 Mar 03 05:34:17 gmorning Mar 03 07:44:10 build #88 of ppc40x is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/ppc40x/builds/88 Mar 03 07:49:23 build #89 of orion is complete: Failure [failed shell_4 compile_1] Build details are at http://tksite.gotdns.org:8010/builders/orion/builds/89 Mar 03 10:01:50 [florian]: ping Mar 03 10:10:03 acoul * r25844 /trunk/toolchain/gcc/patches/4.5.2/995-fa526.patch: toolchain/gcc: fix r25709 (closes #8964) Mar 03 10:11:27 jow * r25845 /trunk/target/linux/rdc/config-2.6.32: [rdc] add missing kernel symbol Mar 03 10:16:49 jow * r25846 /trunk/target/linux/generic/ (8 files): [generic] add more missing kernel symbols Mar 03 10:19:54 'lo KanjiMonster Mar 03 10:20:05 hi EqUaTe Mar 03 10:22:26 how goes it? Mar 03 10:26:52 currently trying to figure out why sysupgrade breaks the rootfs when keeping settings on bcm63xx with 2.6.37 Mar 03 10:27:06 acoul * r25847 /trunk/tools/xz/Makefile: tools/xz: update to version 5.0.1 Mar 03 10:30:28 <[florian]> KanjiMonster: pong Mar 03 10:34:28 nbd * r25848 /trunk/target/linux/ar71xx/files/arch/mips/include/asm/mach-ar71xx/gpio.h: ar71xx: add missing gpio_set_debounce function Mar 03 10:34:32 [florian]: ah. something is breaking sysupgrade on 2.6.37 with bcm63xx (but only when keeping settings, so must be something with the fs); haven't been able to find out what yet. Mar 03 10:35:39 [florian]: "VFS: Busy inodes after unmount of jffs2. Self-destruct in 5 seconds. Have a nice day..." <- this is probably releated Mar 03 10:42:35 <[florian]> related to the bcm963xx-flash map driver no? Mar 03 10:45:22 unlikely Mar 03 12:18:59 getting this from latest trunk following a distclean, http://pastebin.com/dSpuJJUX Mar 03 12:19:26 ah yes, seen it on the buildbot as well Mar 03 12:19:37 nbd: ^ Mar 03 12:44:54 <__lore__> ping nbd Mar 03 12:54:21 ping nbd Mar 03 12:54:39 nbd the changeset 25600 breaks the compile of libiconv Mar 03 12:54:48 this line doenst work ... $(INSTALL_DATA) $(PKG_BUILD_DIR)/m4/* $(STAGING_DIR_HOST)/share/aclocal/ Mar 03 12:56:57 jow * r25849 /packages/libs/libiconv/Makefile: [packages] libiconv: fix typo in HostInstall section Mar 03 12:58:55 thanks jow Mar 03 13:15:57 ping loswillios Mar 03 13:18:05 build #72 of xburst is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/xburst/builds/72 Mar 03 13:37:13 jow * r25850 /packages/libs/libxslt/Makefile: [packages] libxslt: fix InstallDev after r25729 Mar 03 13:53:04 thepeople: ping Mar 03 14:08:21 hi guys, do you know how to compile one package without cache ? Mar 03 14:08:22 Im getting Mar 03 14:08:22 "Error: V8 doesn't like cache. " while compile something, and I dont want to change global ccache setting. any1 ? Mar 03 14:29:23 ping jow Mar 03 14:29:38 jow_laptop: this r25849 commit didnt fix the error Mar 03 14:30:02 i don see which m4 directory you wanna copy as there is no m4 dir Mar 03 14:30:10 error is install: cannot stat `/mnt/stuff/Openwrt/trunk_git/build_dir/host/libiconv/m4/*': No such file or directory Mar 03 14:34:45 jow * r25851 /packages/libs/libiconv/Makefile: [packages] libiconv: fix r25849 Mar 03 14:43:16 [florian]: ah, found it. It actually /was/ the flashmap driver, I didn't know that the totalLength is actually a reserved field and the rootfsLength get set to zero. Mar 03 14:45:14 [florian]: the field is actually used on some devices though (as the "image sequence" field, for finding out which image is newer with dual images) Mar 03 14:50:54 nbd * r25852 /trunk/package/base-files/Makefile: base-files: do not build or use libgcc on m68k and avr32 Mar 03 14:52:23 nbd, hi, I checked my previous builds for the wifi problem Mar 03 14:53:44 hi Mar 03 14:53:45 and? Mar 03 14:53:48 I can say that the problem doesn't exist at revision 24966. But it's exist at revision 25426. The problem is somewhere between these revisions. Mar 03 14:54:49 that's a wide range of changes Mar 03 14:54:56 start bisecting tornado :) Mar 03 14:55:03 please try some revisions inbetween (of just mac80211, you can leave the rest at current) Mar 03 14:55:15 jow * r25853 /packages/libs/taglib/ (Makefile patches/ patches/100-uclibc++-compat.patch): [packages] taglib: bump to v1.6.3, switch to cmake, add uclibc++ compat, throw away the libtool shit Mar 03 14:55:23 [florian]: hm, it seems flashing images through cfe on devices using the image sequence will pose a problem since cfe uses the same four bytes but as ascii, and writes this field when flashing the image (atoi -> ++ -> sprintf()), so I think we have to use a different offset for the totalLength Mar 03 14:55:47 if you're on current trunk, also please rebuild your toolchain Mar 03 14:55:56 there were some issues after the upgrade to the linaro gcc Mar 03 14:55:59 which are fixed now Mar 03 14:56:04 so make dirclean would be a good idead Mar 03 14:56:18 because one of those issues affects hostapd Mar 03 14:56:23 bbl Mar 03 14:58:22 nbd, I've updated my trunk 2 or 3 days days ago. and rebuilt it after dirclean Mar 03 14:59:00 EqUaTe, yes I think that's a good idea ;) Mar 03 14:59:24 tornado: the bug was fixed yesterday Mar 03 14:59:39 so you should do another dirclean Mar 03 15:00:02 nbd, wow that's great. I'll try it. thanks Mar 03 15:02:33 i have to say that git makes bisecting really easy... Mar 03 15:04:14 build #95 of at91 is complete: Success [build successful] Build details are at http://tksite.gotdns.org:8010/builders/at91/builds/95 Mar 03 15:06:05 ping OutBackDingo Mar 03 15:06:29 xMff: pong Mar 03 15:06:49 OutBackDingo: you originally committed libre/restund, did it ever work for you? Mar 03 15:07:02 it does not comple in backfire, it does not compile in trunk Mar 03 15:07:13 in fact I did never see it compile successfully Mar 03 15:07:15 xMff: odd it did Mar 03 15:07:34 i wouldnt commit comething that didnt compile and run Mar 03 15:07:40 ill look at it again Mar 03 15:07:46 i might have fubared something Mar 03 15:07:53 I assumed that, therefore I ask Mar 03 15:08:12 because its a quite obvious error, not something like subtle compile issues Mar 03 15:08:18 xMff: LOL that i fubared something or that i wouldnt commit something that didnt build/run :) Mar 03 15:08:38 I was referring to "i wouldnt commit comething that didnt compile and run" Mar 03 15:08:45 hehehhe Mar 03 15:09:07 ok ill give it another go and fix it up Mar 03 15:09:14 hm wait, a new error Mar 03 15:09:20 its not the missing re.mk anymore Mar 03 15:10:23 http://buildbot.openwrt.org:8010/broken_packages/at91/restund/compile.txt Mar 03 15:10:47 not possible to see whats going on cause of the stupid beaurification Mar 03 15:11:15 but I'd say its either missing -lre or a missing (-Wl,)-rpath-link=$(STAGING_DIR)/usr/lib Mar 03 15:14:02 xMff: ill run through it Mar 03 15:15:17 thank you Mar 03 15:30:59 xMff: args, thanks for notifying me Mar 03 15:32:48 I should switch to git for packages Mar 03 15:33:35 you mean git will magically solve pebkac ? ;) Mar 03 15:34:27 hehe Mar 03 15:34:36 heh Mar 03 15:34:40 I wonder why the file is blob anyway Mar 03 15:34:47 if it does then serious kudos to the git devs :P Mar 03 15:35:33 bbl Mar 03 15:42:05 build #76 of ar71xx is complete: Failure [failed compile_4] Build details are at http://tksite.gotdns.org:8010/builders/ar71xx/builds/76 Mar 03 15:42:12 xMff: does juhosg speak English? Mar 03 15:42:25 I sent him an e-mail a fair few days ago but haven't had a reply Mar 03 15:43:39 he tends to be very busy Mar 03 15:43:48 yes, he speaks english Mar 03 15:44:08 heh, ok Mar 03 15:44:23 I'll try Hungarian if I don't get a response in a week or two Mar 03 16:22:18 nbd * r25854 /trunk/target/linux/ar71xx/files/arch/mips/include/asm/mach-ar71xx/gpio.h: Revert "ar71xx: add missing gpio_set_debounce function" - wrong target Mar 03 16:28:30 <__lore__> ping nbd Mar 03 16:28:37 pong Mar 03 16:29:36 <__lore__> hi Mar 03 16:29:41 hi Mar 03 16:29:43 <__lore__> how are you? Mar 03 16:30:06 fine Mar 03 16:30:10 <__lore__> :) Mar 03 16:30:34 guys, do you know how to compile one package without cache ? Mar 03 16:30:34 Im getting "Error: V8 doesn't like cache. " while compile something, and I dont want to change global ccache setting. any1 ? I dont want to recompile everything... Mar 03 16:31:05 <__lore__> I carried out some measuraments disabling RX interrupt mitigation.. Mar 03 16:31:27 voyo: xMff already told you it's not possible Mar 03 16:31:47 <__lore__> do you remember the project I am working on? Mar 03 16:33:25 yes Mar 03 16:33:38 build #93 of brcm47xx is complete: Success [build successful] Build details are at http://tksite.gotdns.org:8010/builders/brcm47xx/builds/93 Mar 03 16:33:40 got_milk: ok tnx. hard to follow for responses if you dont know thats for you. however I dont believe in such things like "its not possible" ;) Mar 03 16:33:56 <__lore__> I obtained the RX delay decreases to 500 us Mar 03 16:34:04 <__lore__> from 1.5ms Mar 03 16:34:08 good Mar 03 16:34:59 <__lore__> I obtain no performance decrease Mar 03 16:35:17 voyo: make package//compile V=99 CONFIG_CCACHE= Mar 03 16:35:26 <__lore__> ~ 85Mbps with UDP iperf traffic Mar 03 16:35:36 ha! tnx nbd :) Mar 03 16:35:56 <__lore__> I am wondering why interrupt mitigation is enabled on embedded device Mar 03 16:36:32 __lore__: because on slower devices it does help for bringing down interrupt load when there is no NAPI support Mar 03 16:36:49 <__lore__> :) Mar 03 16:37:08 <__lore__> I am using a rocket M Mar 03 16:37:33 <__lore__> from my measurements the system works fine without RX interrupt mitigation Mar 03 16:39:08 <__lore__> is RX interrupt mitigation configurable? Mar 03 16:39:48 <__lore__> I mean, is it possible to specify how many interrupts aggregate? Mar 03 16:40:02 there are register settings for changing the timing Mar 03 16:40:16 see 520-ath9k_intr_mitigation_tweak.patch Mar 03 16:41:09 <__lore__> ok..tnx :) Mar 03 16:42:47 <__lore__> what are AR_RIMT_LAST and AR_RIMT_FIRST? Mar 03 16:43:08 <__lore__> 250 and 500 are us? Mar 03 16:48:55 build #74 of kirkwood is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/kirkwood/builds/74 Mar 03 16:50:11 <__lore__> upper and lower part of AR_RIMT register Mar 03 16:52:06 <__lore__> is it seem to me that in 520-ath9k_intr_mitigation_tweak.patch it's halfed a value.. Mar 03 16:52:24 <__lore__> what does it represent? Mar 03 17:19:58 build #19 of lantiq is complete: Exception [exception failed slave lost shell_13 compile_12] Build details are at http://tksite.gotdns.org:8010/builders/lantiq/builds/19 Mar 03 17:29:04 build #82 of rb532 is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/rb532/builds/82 Mar 03 17:44:34 build #82 of ps3 is complete: Failure [failed compile_3] Build details are at http://tksite.gotdns.org:8010/builders/ps3/builds/82 Mar 03 17:55:54 build #84 of pxcab is complete: Failure [failed shell_4 compile_3] Build details are at http://tksite.gotdns.org:8010/builders/pxcab/builds/84 Mar 03 17:58:37 nbd: you might want to close this ticket -> https://dev.openwrt.org/ticket/8969 Mar 03 18:06:54 jow * r25855 /packages/ipv6/send/patches/006-openssl-1.0.0-compat.patch: [packages] send: fix more OpenSSL 1.0.0 fallout Mar 03 19:08:54 build #84 of ramips is complete: Exception [exception shell failed slave lost shell_13 compile_12] Build details are at http://tksite.gotdns.org:8010/builders/ramips/builds/84 Mar 03 19:38:21 nbd, I updated my trunk and built it after dirclean. The problem still exists Mar 03 19:43:46 nbd, again when I run wifi at client link goes unstable and after some seconds will be done at all. I have to run wifi at AP to bring it back. Mar 03 19:46:31 nbd, it makes the wireless unusable. I have to use old revisions. Please check it out if you could. Thanks in advance Mar 03 19:48:59 build #89 of cobalt is complete: Success [build successful] Build details are at http://tksite.gotdns.org:8010/builders/cobalt/builds/89 Mar 03 23:41:28 build #72 of au1000 is complete: Exception [exception failed slave lost shell_13 compile_12] Build details are at http://tksite.gotdns.org:8010/builders/au1000/builds/72 Mar 03 23:41:30 build #76 of uml is complete: Exception [exception failed slave lost shell_13 compile_12] Build details are at http://tksite.gotdns.org:8010/builders/uml/builds/76 Mar 03 23:41:37 build #100 of brcm63xx is complete: Exception [exception failed slave lost shell_13 compile_12] Build details are at http://tksite.gotdns.org:8010/builders/brcm63xx/builds/100 Mar 03 23:41:39 build #95 of s3c24xx is complete: Exception [exception failed slave lost shell_13 compile_12] Build details are at http://tksite.gotdns.org:8010/builders/s3c24xx/builds/95 Mar 04 02:17:28 swalker * r25856 /packages/ipv6/ndisc/ (Makefile patches/130-wifcontinued.patch): Mar 04 02:17:28 [packages] ndisc: update to 1.0.1 Mar 04 02:17:28 * add missing librt dependencies Mar 04 02:18:43 swalker * r25857 /packages/net/lftp/Makefile: [packages] lftp: update to 4.1.3 Mar 04 02:19:58 swalker * r25858 /packages/net/nbd/Makefile: [packages] nbd: update to 2.9.20 Mar 04 02:33:01 swalker * r25859 /packages/net/ncftp/Makefile: Mar 04 02:33:01 [packages] ncftp: update to 3.2.5 Mar 04 02:33:01 * move to FTP submenu **** ENDING LOGGING AT Fri Mar 04 02:59:59 2011