**** BEGIN LOGGING AT Mon Mar 14 02:59:57 2011 Mar 14 03:39:17 cshore: Thanks. The r26155 fixed the issues. BTW, I am not sure if modules on these packages, i.e. iputils, krb5, net-tools, ppp, and snmp, are also mangling. Mar 14 05:28:31 nbd: there's a missing -htc for https://dev.openwrt.org/browser/trunk/package/mac80211/Makefile#L491's description Mar 14 07:35:41 blogic * r26157 /trunk/target/linux/lantiq/ (6 files in 2 dirs): Mar 14 07:35:41 [lantiq] Mar 14 07:35:41 * fixes arv4510 board support Mar 14 07:35:41 * fixes nor swizzle hack Mar 14 07:35:41 * fixes gpio .can_sleep Mar 14 07:35:41 * fixes compile warning inside pci driver Mar 14 07:36:23 blogic * r26158 /trunk/package/ (4 files in 4 dirs): Mar 14 07:36:23 [lantiq] Mar 14 07:36:23 * fixes .unlocked_ioctl functions Mar 14 07:38:13 xMff: gdbserver: can't load library 'libthread_db.so.1' Mar 14 08:52:01 build #97 of cobalt is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/cobalt/builds/97 Mar 14 09:42:28 russell--: crap Mar 14 10:21:57 jow_laptop: i found a forum comment suggesting PTHREADS_DEBUG_SUPPORT=y in toolchain/uClibc/config/mipsel might help, but i didn't find that it did. Mar 14 10:24:47 russell--: well for now you can just copy libthread_db.so.1 to your device Mar 14 10:24:58 russell--: it should be somewhere in build_dir/toolchain-* Mar 14 10:25:53 * russell-- is currently rebuilding just to confirm. i'll look when it finishes. Mar 14 10:28:35 ok Mar 14 10:36:33 gmorning Mar 14 11:00:51 build #91 of ramips is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/ramips/builds/91 Mar 14 11:17:18 <_trine> #j ##dd-wrt Mar 14 11:19:21 ? Mar 14 11:24:41 <_trine> blogic, is the project you are doing concerned with voip Mar 14 11:24:48 yes Mar 14 11:24:51 i have 2 Mar 14 11:24:57 1) make voip work fully on lantiq Mar 14 11:24:59 <_trine> lantiq or something Mar 14 11:25:05 2) make a small eval hw with voip Mar 14 11:25:27 <_trine> I would be interested in making a board up Mar 14 11:25:27 well Mar 14 11:25:36 i am working on a arm9 design Mar 14 11:26:02 it is a sam9260 it is lqfp so i can solder it Mar 14 11:26:11 once it works i plan to switch to sam9g20 Mar 14 11:26:25 <_trine> is there any info on what you are doing anywhere Mar 14 11:26:33 which is pretty much pin compatible, but bga and has 400mhz Mar 14 11:26:37 not yet Mar 14 11:26:46 i am making the schematic in eagle at the moment Mar 14 11:26:55 <_trine> are you OK with solder surface mount Mar 14 11:27:08 <_trine> soldering* Mar 14 11:27:12 yes Mar 14 11:27:16 easy ;) Mar 14 11:27:25 brb Mar 14 11:27:28 <_trine> it is wen you have the equipment Mar 14 11:27:51 i have a semi professional lab Mar 14 11:28:01 <_trine> very nice Mar 14 11:51:31 anyone here using a Samsung printer on OpenWRT? Mar 14 11:55:38 _trine: so basically once i advance with the schematic i will upload it and then get input Mar 14 12:40:16 build #97 of orion is complete: Failure [failed shell shell_4 compile_6] Build details are at http://tksite.gotdns.org:8010/builders/orion/builds/97 Mar 14 13:19:57 build #89 of rb532 is complete: Success [build successful] Build details are at http://tksite.gotdns.org:8010/builders/rb532/builds/89 Mar 14 13:38:52 build #76 of etrax is complete: Failure [failed compile_3] Build details are at http://tksite.gotdns.org:8010/builders/etrax/builds/76 Mar 14 13:49:40 kerneis * r26159 /packages/net/babel/ (Makefile patches/001-fix-local_notify_route.patch): [packages] babeld: patch minor bug until next release Mar 14 14:02:09 build #101 of brcm47xx is complete: Success [build successful] Build details are at http://tksite.gotdns.org:8010/builders/brcm47xx/builds/101 Mar 14 14:07:04 nbd * r26160 /trunk/package/mac80211/Makefile: mac80211: fix ath9k/ath9k-htc description (thx, swalker) Mar 14 15:18:18 build #103 of atheros is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/atheros/builds/103 Mar 14 15:55:42 build #72 of mpc52xx is complete: Failure [failed compile_4] Build details are at http://tksite.gotdns.org:8010/builders/mpc52xx/builds/72 Mar 14 16:05:48 ping xMff: this resolving but is inconsistent, sometimes it segfaults, sometimes not Mar 14 16:05:59 s/but/bug Mar 14 16:14:50 xMff: I have gdbserver installed and ping/wget segfaulting Mar 14 17:12:39 netprince: gdbserver :9000 /bin/ping example.org Mar 14 17:13:15 netprince: in your buildroot ./scripts/remote-gdb 192.168.1.1:9000 ./staging_dir/target-*/root-*/bin/ping Mar 14 17:13:31 netprince: then "c" enter Mar 14 17:13:44 netprince: once it crashes and drops to a prompt, "bt" enter Mar 14 18:09:53 jow_laptop: looks like I didn't rebuild the toolchain after adding CONFIG_GDB, cause it cant find gdb... argh Mar 14 18:17:34 jow_laptop: lol, well, i found build_dir/toolchain-mipsel_gcc-linaro_uClibc-0.9.32/uClibc-0.9.32/lib/libthread_db-0.9.32-rc2-git.so ... unfortunately, it is neary 7MB ;-) maybe i can load it on a thumbdrive, the wgt634u has usb afterall. Mar 14 18:18:05 russell--: strip it Mar 14 18:19:12 do you happen to know the path to the right strip program? Mar 14 18:19:48 probably staging_dir/toolchain-mipsel_gcc-linaro_uClibc-0.9.32/bin/mipsel-openwrt-linux-strip Mar 14 18:20:34 okay, thanks! Mar 14 18:22:06 ooh, that's better! only 35k now! Mar 14 18:25:48 okay, gdbserver runs now ... /me fishes the incantation out of irclogs ... Mar 14 18:31:56 so close! (/me needs to disable firewall, i think) Mar 14 18:41:15 build #82 of uml is complete: Failure [failed compile_10] Build details are at http://tksite.gotdns.org:8010/builders/uml/builds/82 Mar 14 18:47:51 thread_db stripped only 35k? yay, guess we should package it then Mar 14 18:47:53 * xMff looks Mar 14 18:59:59 build #100 of ubicom32 is complete: Failure [failed compile_3] Build details are at http://tksite.gotdns.org:8010/builders/ubicom32/builds/100 Mar 14 19:04:05 hmm. SIGCANCEL seems to be involved Mar 14 19:06:56 somewhere in pthread_cancel Mar 14 19:08:49 39 #ifdef SHARED Mar 14 19:08:49 40 pthread_cancel_init (); Mar 14 19:08:50 41 #endif Mar 14 19:09:19 that's where the unwind/dlopen stuff comes in, btw Mar 14 19:09:34 which library is that? Mar 14 19:12:30 hm Mar 14 19:12:45 after building the target gcc and the host gcc I cannot find a libthread_db anywhere Mar 14 19:13:25 erm s/gcc/gdb/ Mar 14 19:14:23 ah uclibc must be build with PTHREADS_DEBUG_SUPPORT Mar 14 19:15:05 yeah Mar 14 19:15:40 hm Mar 14 19:16:11 how do i ask gdb where i am? (/me rusty with gdb) Mar 14 19:17:23 i'm somewhere in pthread_cancel() anyway Mar 14 19:17:28 info line Mar 14 19:17:52 Line 91 of "libpthread/nptl/pthread_cancel.c" starts at address 0x2b365564 and ends at 0x2b36558c . Mar 14 19:17:55 -ish Mar 14 19:18:48 val = INTERNAL_SYSCALL (tkill, err, 2, pd->tid, SIGCANCEL); Mar 14 19:19:03 hm Mar 14 19:19:12 that should have no side effects Mar 14 19:19:53 another step or two and i get "Remote failure reply: E01" Mar 14 19:20:12 interesting Mar 14 19:20:23 is it killing the debuger? Mar 14 19:20:25 the strace seemed to indicate 32 was "unknown" by the receiver or something Mar 14 19:20:39 i can't talk to the gdbserver anymore Mar 14 19:20:46 I see Mar 14 19:20:50 without gdb, at this point, the terminal session freezes Mar 14 19:21:25 after the unknown signal ABORT gets called to kill the thread or something Mar 14 19:21:54 the weird thing is iftop hasn't changed in almost 2 years Mar 14 19:22:02 uclibc has changed Mar 14 19:22:03 a lot Mar 14 19:22:06 yeah Mar 14 19:22:06 yeah, but the uclibc threads implementation has Mar 14 19:22:18 so it sounds like uclibc does not properly set up the internal signal handler for SIGCANCEL Mar 14 19:22:24 which i assume is involved in handling thread cancellation Mar 14 19:22:48 and because it isn't set up, the default handler kicks in Mar 14 19:22:52 yeah Mar 14 19:22:52 and runs abort() Mar 14 19:23:42 i wonder where else this would manifest itself besides iftop? Mar 14 19:26:17 russell--: what is the iftop symptoms? Mar 14 19:26:48 freezes terminal session on quit Mar 14 19:27:49 i think i found something Mar 14 19:30:03 nbd * r26161 /trunk/package/base-files/Makefile: base-files: link libpthread against libc Mar 14 19:30:20 try svn up, make package/base-files/compile, then install the new libpthread .ipk Mar 14 19:30:49 * russell-- is a gitster, but will try Mar 14 19:31:38 svn up is shorter to type than 'update to latest from svn' ;) Mar 14 19:32:20 how often does the git tree sync? Mar 14 19:32:40 was 10 minutes or something a year ago (/me vaguely recalls) Mar 14 19:32:44 every 10 minutes or so, yes Mar 14 19:32:47 okay Mar 14 19:33:07 should be in sync now Mar 14 19:33:25 yes Mar 14 19:33:28 just pulled Mar 14 19:41:13 hmm. same thing. Mar 14 19:41:26 ok Mar 14 19:41:34 aborts in an indistinguisable way. Mar 14 20:17:20 nbd, ping Mar 14 20:22:03 tornado: pong Mar 14 20:22:27 Hi, Thanks again for solving the ath9k problem. Everything is great now. Just there's one problem that I told you before. Mar 14 20:22:47 About the delay in 4addr Mar 14 20:23:30 At first I should say that when I run ifup lan, the wlan0 enters disabled state and needs one wifi command after it. Now when it's connected after running wifi, I can simply ping the AP from client console, but it takes about 40 seconds or more to have ping or other traffics over 4addr. Mar 14 20:24:51 the ifup vs wifi issue is a known issue, it will be fixed once i make more progress with my rewrite of the network init stuff Mar 14 20:24:59 as for the 40 seconds - no idea where those come from Mar 14 20:25:21 nbd, I did some tests Mar 14 20:25:35 The problem exists most of the times after ifup lan. Mar 14 20:25:54 The only more thing that I can say is that when I run wifi immediately after ifup lan, the problem doesn't exist. (I can say before I get "destination host unreachable" or in other word when I'm getting "request timed out") Mar 14 20:26:23 But if I run wifi after getting an "destination host unreachable", I have to wait for about 40 seconds to have traffic over 4addr. Mar 14 20:26:23 maybe the mac address of br-lan changes when you run wifi Mar 14 20:26:31 have you tried clearing your arp cache after running wifi? Mar 14 20:26:49 no Mar 14 20:27:12 most of the time when i see something matching your description when bringing lan down/up it is related to ARP Mar 14 20:27:18 but br-lan mac address doesn't change. I know that Mar 14 20:27:47 ok Mar 14 20:27:59 ok I'll test it Mar 14 20:37:39 blogic, ping Mar 14 21:37:47 I have some common defconfig settings that I want to use for all architectures. Right now I'm maintaining 2 separate .configs. Is there a better way? Mar 14 21:38:10 ~/.openwrt/defconfig Mar 14 21:39:51 I thought ~/.openwrt/defconfig was an architecture specific config. What I really want is 1 file for the common options, then separate files for each arch with the arch-specific stuff Mar 14 21:40:16 defconfig is sourced and merged with whatever config you havei n your tree Mar 14 21:40:26 so it'd hold your generic stuff Mar 14 21:55:50 russell--: please try copying staging_dir/toolchain-*/lib/libpthread-*.so to your device and see if iftop still aborts Mar 14 21:58:06 xMff, so for the arch specific stuff... that goes in target/linux/[arch]/config-default? Mar 14 21:58:17 yeah Mar 14 21:58:24 if you really want to edit it Mar 14 21:58:37 but that's kernel config stuff Mar 14 21:58:41 right Mar 14 21:58:42 hrmm, actually that the kernel config... yeah Mar 14 21:59:02 Is there a place for the arch specific openwrt config? Mar 14 21:59:20 yes Mar 14 21:59:24 DEFAULT_PACKAGES in the target makefile Mar 14 21:59:48 it can select packages Mar 14 22:00:12 however at this point it might be useful if you tell us what kind of config changes you want to maintain Mar 14 22:00:15 maybe there is a better way to do it Mar 14 22:00:44 I want CONFIG_TARGET_ROOTFS_TGZ=y but only for one of the archs Mar 14 22:01:00 which arch? Mar 14 22:01:36 target/linux/fsP1020RDB (it's a new platform that I added) Mar 14 22:01:43 powerpc though Mar 14 22:01:57 FEATURES += tgz Mar 14 22:01:59 in the target makefile Mar 14 22:02:11 that's how other targets do this Mar 14 22:03:05 nice! thanks gents Mar 14 22:05:03 nbd: okay Mar 14 22:05:45 nbd, it looks like REATURES += tgz sets CONFIG_USES_TGZ not CONFIG_TARGET_ROOTFS_TGZ Mar 14 22:06:38 er: oops... it set CONFIG_TARGET_ROOTFS_TGZ in addition to CONFIG_USES_TGZ Mar 14 22:06:40 TARGET_ROOTFS_TARGZ defaults to y if USES_TARGZ is set Mar 14 22:06:49 yup... ok Mar 14 22:15:30 nbd: same-same Mar 14 22:16:17 ok, so i guess we can rule out the relinking stuff Mar 14 22:20:56 Hi all...anyone trying to yell at me about the package reorganization from the other night (I hope not) ? Mar 14 22:21:40 cshore: no. but looks like your yate package has a fan already Mar 14 22:22:28 cool...mazilo? He was asking about it the last night. Mar 14 22:22:37 cshore: check the tickets Mar 14 22:22:51 someone reported a bunch of bugs Mar 14 22:24:01 hmmm...I hope it's not from modules that are marked @BROKEN but they didn't realize it because the shebang is marked @BROKEN for now Mar 14 22:25:14 how do I set CONFIG_X86_VDI_IMAGES=y by default? Mar 14 22:26:10 nope, it's not, it's a good catch Mar 14 22:36:35 libpthread/nptl/unwind.c:120:1: warning: redeclaration of '__pthread_unwind' with different visibility (old visibility preserved) Mar 14 22:36:48 * russell-- wonders if that's relevant at all Mar 14 23:13:04 build #89 of ppc44x is complete: Success [build successful] Build details are at http://tksite.gotdns.org:8010/builders/ppc44x/builds/89 Mar 14 23:45:19 build #102 of brcm47xx is complete: Exception [exception failed slave lost shell_13 compile_12] Build details are at http://tksite.gotdns.org:8010/builders/brcm47xx/builds/102 Mar 14 23:45:20 build #83 of x86 is complete: Exception [exception failed slave lost shell_13 compile_12] Build details are at http://tksite.gotdns.org:8010/builders/x86/builds/83 Mar 14 23:45:22 build #75 of ep93xx is complete: Exception [exception failed slave lost shell_13 compile_12] Build details are at http://tksite.gotdns.org:8010/builders/ep93xx/builds/75 Mar 14 23:50:45 acinonyx * r26162 /trunk/package/hostap-driver/ (Makefile files/lib/wifi/hostap.sh): [package] hostap-driver: Add support for multiple interfaces Mar 15 00:16:59 If my ISP supports IPv6, will "Enable IPv6 support in packages" in the "Global build settings" through the 'make menuconfig' be enough to handle IPv6? If no, what other packages do I need to install? Mar 15 00:46:02 acinonyx * r26163 /trunk/package/hostap-driver/files/lib/wifi/hostap.sh: [package] hostap-driver: Fix unique device autodetection Mar 15 00:46:09 acinonyx * r26164 /trunk/package/madwifi/files/lib/wifi/madwifi.sh: [package] madwifi: Fix unique device autodetection Mar 15 00:46:21 acinonyx * r26165 /trunk/package/mac80211/files/lib/wifi/mac80211.sh: [package] mac80211: Fix unique device autodetection Mar 15 00:46:35 acinonyx * r26166 /trunk/package/base-files/files/etc/init.d/network: [package] base-files: Autodetect newly added wireless hardware Mar 15 01:39:04 nbd * r26167 /trunk/package/mac80211/patches/540-ath9k_ar9100_blockack_fix.patch: ath9k: enable an ar913x specific hardware fix for sending block-ack Mar 15 01:39:08 nbd * r26168 /trunk/package/mac80211/patches/541-ath9k_ar9100_rx_chainmask_fix.patch: ath9k: force the rx chainmask on ar913x to 0x7 - fixes calibration issues with 2-antenna devices such as the wrt160nl Mar 15 01:40:02 cshore * r26169 /packages/utils/coreutils/Makefile: Mar 15 01:40:02 [utils] coreutils: Stopped coreutils from building as long as no plugin is selected. Mar 15 01:40:02 with commit 26026 (https://dev.openwrt.org/changeset/26026) Mar 15 01:40:02 I see the coreutils package downloaded and build in my defconfig. Mar 15 01:40:02 With the following patch, this default selection really does no build, Mar 15 01:40:02 as long as no plugin is selected. Mar 15 01:40:03 Thanks to Thomas Langer for this patch. Mar 15 01:58:09 I just compiled and flashed my NetGear WGT634U with r26166 and LuCI main page shows a white blank page. Mar 15 02:40:10 nbd * r26170 /branches/backfire/scripts/config/conf.c: scripts/config: backport output file option from r25230 Mar 15 02:40:12 nbd * r26171 /branches/backfire/scripts/diffconfig.sh: backport scripts/diffconfig.sh from trunk Mar 15 02:40:15 nbd * r26172 /branches/backfire/package/mac80211/patches/ (2 files): ath9k: backport ar913x fixes from r26167, r26168 **** ENDING LOGGING AT Tue Mar 15 02:59:56 2011