**** BEGIN LOGGING AT Mon Feb 06 02:59:58 2012 Feb 06 04:45:05 build #120 of rb532 is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/rb532/builds/120 Feb 06 08:09:24 jow_laptop: pong Feb 06 09:18:51 <[florian]> [Fate]: ah so true, did you have a safe trip back? Feb 06 09:21:30 philipp64|laptop: alix2 kernel config questions: SCSI_MVSAS_TASKLET, SKGE_GENESIS, LEDS_NET5501, noticed while updating from r30004 to r30310 Feb 06 10:41:55 loswillios: whats the state of pulseaudio on backfire? I needs ntpl, right? Feb 06 10:51:52 jow * r30325 /packages/ (4 files in 4 dirs): [packages] re, rem, baresip, restund: use another approach to find the "sysroot", the current way still fails on Backfire - sigh Feb 06 10:53:00 jow * r30326 /branches/packages_10.03.2/ (4 files in 4 dirs): [packages_10.03.2] re, baresip, restund: merge r30325 Feb 06 11:01:36 jow * r30327 /branches/packages_10.03.2/libs/rem/ (. Makefile Makefile): [packages_10.03.2] add rem library from trunk, needed by baresip Feb 06 11:11:16 jow_laptop: exactly Feb 06 11:12:32 loswillios: so pulseaudio is completely broken for backfire? Feb 06 11:13:37 yes, that's why the @!UCLIBC_VERSION_0_9_30_1 stuff is added Feb 06 11:13:56 in theory it shouldn't be build for backfire Feb 06 11:13:57 in this case it might make sense to simply remove it Feb 06 11:14:01 yes Feb 06 11:14:20 ok, I put that in the queue, thanks Feb 06 11:14:37 thanks Feb 06 12:24:45 jow * r30328 /packages/net/baresip/patches/ (. 100-avcodec-compat.patch): [packages] baresip: fix build against libavcodec by removing broken compat code Feb 06 12:27:52 jow * r30329 /branches/packages_10.03.2/net/baresip/patches/: [packages_10.03.2] baresip: merge r30328 Feb 06 12:29:53 blogic: have you got a moment to take a look at: http://pastebin.com/ebBve9zF ref ADSL annex a. this is the result of kernel symbols and pakages with debug info turned on. Feb 06 12:33:10 1 sec Feb 06 12:33:53 try to rebuild the dsl driver Feb 06 12:34:02 that does really early in the init Feb 06 12:36:08 should have all build overnight. how do i ask for just the dsl to be rebuilt - make package somehting? Feb 06 12:37:17 http://www.youtube.com/watch?v=g2mZhWDnrpU&feature=related Feb 06 12:37:40 make package/ltq-dsl/clean && make V=99 Feb 06 12:39:41 :) like the footage Feb 06 12:40:15 felix from saturday Feb 06 12:41:51 quite noisy Feb 06 12:49:17 jow * r30330 /branches/packages_10.03.2/net/ntpd/Makefile: [packages_10.03.2] ntpd: remove libcap dependency, disable linux capabilities Feb 06 12:51:39 btsimonh|2: if you need to build only ltq-dsl package you can do: make package/ltq-dsl/{clean,compile} Feb 06 13:05:48 ok, going to try it on ADSL - will lose this connection. have installed atm-diagnostics and have atm-tools. bfore i go, what commands ar useful? Feb 06 13:06:54 ok Feb 06 13:06:55 wait Feb 06 13:07:04 leave it connected for 1-2 minutes Feb 06 13:07:12 it will say "showtime" in the logs Feb 06 13:07:21 with Daratrae XYZ etc Feb 06 13:07:31 once oyu see that do a ifup wan Feb 06 13:07:36 good luck Feb 06 13:16:22 jow * r30331 /branches/packages_10.03.2/utils/psmisc/ (Makefile patches/ patches/100-no-mountinfo.patch): [packages_10.03.2] psmisc: skip fuser mount into initialization due to missing posix_memalign() in Backfire Feb 06 13:17:16 btsimonh: what router are you online with ? Feb 06 13:18:46 http://pastebin.com/TGB8Hb3T - interestingly we have some tx data on nas0, but also a crashdump :( Feb 06 13:19:50 I'm on a BT Homehub V2A kindly hacked by PsiDoc - not that it need to be 'cos i'm with bt... p.s. my V3 is a type B, not Type A, so i can't play with that yet. Feb 06 13:20:13 so ... that crash look really weird Feb 06 13:20:17 never seen anything lik eit Feb 06 13:20:20 (type A is broadcom) Feb 06 13:20:23 yes Feb 06 13:21:00 [ 1435.868000] [DSL_BSP_Showtime 916]: Datarate US intl = 0, fast = 448000 <--- this however means its working and connected to the dslam Feb 06 13:21:53 try an image ith only atm + dsl in it Feb 06 13:22:01 and remove all the other cruft Feb 06 13:22:08 and see if it still blows up Feb 06 13:22:19 I can't find where either mei or dsl is configured. Could i have different IRQs (i guess not). Feb 06 13:22:36 Just remove the platform register commands from mach file? Feb 06 13:22:40 no Feb 06 13:22:45 the drivers from the image Feb 06 13:22:49 like usb/wifi/ ... Feb 06 13:23:23 [ 0.076000] Applying Patch for CPU1 IRQ Issue ?! Feb 06 13:24:58 http://pastebin.com/jw1UX6Uw - can remove this too.. not sure it made any difference to previous issues :) Feb 06 13:26:25 ugly Feb 06 13:26:35 ok, left in nand and ethernet only. Feb 06 13:27:32 hmmm Feb 06 13:27:39 is this danube or ar9 ? Feb 06 13:27:48 danube Feb 06 13:27:53 should work Feb 06 13:27:57 (came straight from SHC code; yes it's horrible, but I was clutching at straws). did you note the ath9k fix for eeprom byte swapping? that was a pain. Feb 06 13:27:58 tried it last week Feb 06 13:28:04 annex a? Feb 06 13:28:13 yes Feb 06 13:28:30 Can I turn off the compiler checks for unused code, with a pragma or something? Feb 06 13:28:43 what speed is your line supposed to be ? Feb 06 13:29:00 8mbit, and i get it all, quite close to exchange. Feb 06 13:33:43 the router hardware should handle the 20 mbit stuff... just we won't get it here for ages. Feb 06 13:36:51 remember the nand issue - 32 bit writes failing to do the same job on mine comapred to yours? i'm worried it's related... Feb 06 13:37:34 fits not Feb 06 13:37:44 dsl is a differnt matter really Feb 06 13:37:47 well should be Feb 06 13:51:52 ok about to try, back in a few mins Feb 06 13:58:38 blogic: exactly the same. Feb 06 13:59:55 hmm Feb 06 14:00:12 i will try tonight on a different board Feb 06 14:00:19 but yeah, migt be related Feb 06 14:02:59 I guess I should do my day job now :). Let me know if there is anything I can do to help. Feb 06 15:09:16 build #132 of ubicom32 is complete: Failure [failed compile_1] Build details are at http://buildbot.openwrt.org:8010/builders/ubicom32/builds/132 Feb 06 15:31:12 jow * r30332 /branches/packages_10.03.2/net/nginx/Makefile: [packages_10.03.2] nginx: disable posix_fadvice() on Backfire Feb 06 15:31:30 jow * r30333 /branches/packages_10.03.2/net/nginx/patches/101-feature_test_fix.patch: [packages_10.03.2] nginx: update patch Feb 06 15:34:55 jow * r30334 /packages/net/memcached/Makefile: [packages] memcached: update to v1.4.12 - fixes compilation on Backfire Feb 06 15:35:34 jow * r30335 /branches/packages_10.03.2/net/memcached/Makefile: [packages_10.03.2] memcached: merge r30334 Feb 06 15:48:35 jow * r30336 /branches/packages_10.03.2/net/lftp/Makefile: [packages_10.03.2] lftp: disable posix_fadvice on Backfire Feb 06 15:51:32 jow * r30337 /branches/packages_10.03.2/utils/lcd4linux/patches/ (6 files): [packages_10.03.2] lcd4linux: sync patches with trunk Feb 06 16:02:05 build #134 of at91 is complete: Failure [failed compile_4] Build details are at http://buildbot.openwrt.org:8010/builders/at91/builds/134 Feb 06 16:40:40 ping philipp64|laptop Feb 06 16:44:36 jow * r30338 /branches/packages_10.03.2/utils/lcd4linux/ (8 files in 2 dirs): [packages_10.03.2] lcd4linux: Backfire does not implement log2f(), workaround with logf(x)/logf(2) Feb 06 17:11:00 jow_laptop: pong Feb 06 17:45:31 philipp64|laptop: a user reported that he gets allocation errors on boot with most recent x86 / alix2 subtarget Feb 06 17:45:38 florian * r30339 /trunk/package/iptables/Makefile: [package] iptables: don't apply patches if building an external kernel Feb 06 17:45:42 florian * r30340 /trunk/package/mtd/src/mtd.c: mtd: define MTDREFRESH if not defined (external kernel) Feb 06 17:46:18 philipp64|laptop: http://pastebin.com/djNfNa3t Feb 06 17:46:34 its a 2D13 Feb 06 18:02:21 blogic: '--enable-linux-26' in makefile for dsl - are you sure there's no 3.1 patches i'm missing? Feb 06 18:03:36 yes Feb 06 18:03:46 but i will try tomorrow jut to be sure Feb 06 18:03:47 jow_laptop: I don't have a 2D13 any more... I lent it to Daniel for the work on sysupgrade-x86 he was supposedly going to do. Feb 06 18:04:29 anyway, I can't reproduce this.... I just booted an image on a 6F2 with tinyBIOS (factory) and another with coreboot+SeaBIOS. Feb 06 18:04:48 ok Feb 06 18:04:54 then it was maybe a miscompilation Feb 06 18:04:54 I see this: http://pastebin.com/stEPjUJ7 ... normalcy. is this user building from an entirely fresh checkout? Feb 06 18:05:35 I don't know Feb 06 18:05:38 if he had any .ko's left over from before an "svn update" that didn't get recompiled, the change in compiler options might cause incompatibilities... Feb 06 18:05:45 yeah Feb 06 18:05:53 I'd recommend he build from an entirely new checkout... Feb 06 18:06:06 I'd also suggest he apply the compiler options in 1829. Feb 06 18:07:07 blogic: there are some useful DSL patches that I submitted that haven't gone into the tree yet... Feb 06 18:07:18 they work with the new skbuff changes that went in 3.0... Feb 06 18:07:45 they might help with header reservation problems with using full-length headers. Feb 06 18:09:13 philipp64|laptop: can you give me a link to your DSL patches? Feb 06 18:10:39 1660 & 1662. Feb 06 19:53:19 jow * r30341 /branches/packages_10.03.2/libs/libgphoto2/ (3 files in 2 dirs): [packages_10.03.2] libgphoto2: merge r30323 Feb 06 19:53:41 jow * r30342 /branches/packages_10.03.2/multimedia/gphoto2/ (Makefile patches/002-disable-docs-translations.patch): [packages_10.03.2] gphoto2: merge r30324 Feb 06 20:00:56 jow * r30343 /branches/packages_10.03.2/libs/ftplib/: [packages_10.03.2] ftplib: merge r28996 Feb 06 20:21:28 jow_laptop: ask this alix2 user to jump on IRC some time. Feb 06 20:23:13 do we even know which modules were causing the allocation errors? Feb 06 20:23:34 he was on earlier Feb 06 20:23:42 his kernel is out of RAM Feb 06 20:23:45 missed him, then. Feb 06 20:25:04 who was it? Feb 06 20:25:22 russell--: LEDS_NET5501 shouldn't even be showing up. it's deprecated. Feb 06 20:26:45 see the last 30 lines of target/linux/x86/patches-3.2/005-net5501_platform.patch Feb 06 20:29:34 * russell-- just reports what he sees Feb 06 20:39:31 in general, how big is our stack on mips? Feb 06 20:42:18 you mean the buildenv? Feb 06 20:42:46 oops, nvm Feb 06 20:43:09 compiler warning about 'frame > 1024 bytes' Feb 06 20:44:25 russell--: can you build from a fresh checkout? Feb 06 20:44:39 warning: the frame size of 1496 bytes is larger than 1024 bytes - and another for 2100 bytes. Feb 06 20:50:16 my image just went from 11mbytes to 17mbytes in one build with no material changes :( can't fit it into the available nand space now. Feb 06 20:50:57 btsimonh|3: can you send me your art partition, I wanna se something about eeprom Feb 06 20:54:17 Delboy_: is anything coming through? Feb 06 20:54:34 nope Feb 06 20:54:44 it says it is cancelled Feb 06 20:54:50 one mo, i'll put it somewhere Feb 06 20:54:58 ok, thx Feb 06 20:56:24 try http://psychotherapist.redirectme.net/mtdbock2 Feb 06 20:57:04 it's format matches eeprom_def Feb 06 20:57:31 nope, the link is not working :( Feb 06 20:57:44 one mo Feb 06 20:58:11 oh, that's not good. Geos is panicking on network traffic... http://fpaste.org/6O56/ Feb 06 20:59:31 philipp64|laptop: rebuild with symbol table information Feb 06 21:00:00 try again. for some reason http was going to my openwrt hub! Feb 06 21:00:29 The object requested (/mtdbock2) could not be found on the server., can you upload it to rghost.net? Feb 06 21:00:42 http://psychotherapist.redirectme.net/mtdblock2 Feb 06 21:00:51 try the obvious spelling mistake. Feb 06 21:00:55 it is ok now :), thx Feb 06 21:03:46 168c:0027 is your wlan chip pci id right? Feb 06 21:04:57 yep, think so. but not what it is in the pci beforehand. the stuff at the start is 16 bit pci reg:32 bit value Feb 06 21:05:05 (until FFFF) Feb 06 21:05:24 Del, what's your proc? Feb 06 21:05:42 on sx763 or? Feb 06 21:05:53 danube? Feb 06 21:06:08 yes but 50702 Feb 06 21:06:16 not the denube-s one Feb 06 21:06:29 did you have any problems with DSL? what annex are you? Feb 06 21:07:01 nope, here in croatia we use both annex a and b, depends on a provider Feb 06 21:07:16 everyone i know who uses openwrt works fine Feb 06 21:07:29 are you running annex a or b personally? Feb 06 21:07:29 puth we use PPPoe Feb 06 21:07:48 i don't have adsl Feb 06 21:07:52 ahhhh.... Feb 06 21:15:30 how much of backfire is similar to trunk? Will I be able to backport my lantiq changes if i checkout a backfire? Feb 06 21:16:38 i think there are a lot of changes Feb 06 21:20:00 and if i do a fresh checkout of trunk to trunk2, is there a possibility that my DSL issue may go away? (i.e. could it be a build/config error) - is this worth the 3 hours? Feb 06 21:24:00 mhei * r30344 /packages/lang/php5/Makefile: [packages] php5: upgrade to 5.3.10, fixes CVE-2012-0830 Feb 06 22:08:27 Delboy_: do you use u-boot? What swicth do you have - adm? Feb 06 22:11:34 adm6996i Feb 06 22:12:29 yes i use u-boot from trunk Feb 06 22:12:47 so if I build a fresh gigaset buld, modify to allow it to reister ethernet without u-boot config, DSL should work :). bulding now.... Feb 06 22:14:27 if your dsl line is up on brhomehub, then something is wrong with pppd or your config Feb 06 22:15:33 no, i get a definate crash in ltq_atm :( - but will try a fresh trunk build of yours and see. Feb 06 22:32:40 jow_laptop: what's the symbol do do that? Feb 06 22:37:03 what drivers are needed on x86 hardware to be able to mount a USB (vfat) memory stick? Feb 06 22:37:53 kmod-usb-storage, kmod-fs-vfat, kmod-nls-cp*, kmod-nls-iso* Feb 06 22:39:49 ok, thanks. I left out the nls stuff. Feb 06 22:42:33 do I need all of them? hmm.... Feb 06 22:43:31 no Feb 06 22:43:44 only the ones for the codepage used by your vfat fs Feb 06 22:43:53 dmesg should tell once you attempt to mount it Feb 06 23:28:12 how do I setup a fresh ramdisk image for pppoa from command line? Feb 06 23:29:42 do you know vpi and vci? Feb 06 23:31:04 uci set network.wan.proto=pppoa; uci set network.wan.vci=x; uci set network.wan.vpi=y; uci set network.wan.username=...; uci set network.wan.password=...; uci commit network; ifup wan Feb 06 23:34:24 vci=38, vpi=0; on ifup wan, i get grep: /etc/config/wireless: no such file or directory - will this affect the ifup wan working? Feb 06 23:34:31 no Feb 06 23:34:50 thats just ifup, it tries to bringup related wifi ifaces Feb 06 23:35:04 long story :) Feb 06 23:35:33 check logread for actual info Feb 06 23:35:49 pppd should try to bridge the circuit to a pppoe-wan iface Feb 06 23:36:14 sorry, pppoa-wan Feb 06 23:36:23 have to wait for the wife to stop webboggle :( then i plug in and dmesg.... Feb 06 23:37:15 I need to go now Feb 06 23:37:24 good luck Feb 06 23:37:34 thanks for your help tonight/ Feb 06 23:37:40 ps Feb 06 23:37:53 whoops -wrong place to type that :) Feb 06 23:56:58 jow_laptop: mount: mounting /dev/sdb1 on /tmp/vfat failed: No such device Feb 07 00:28:29 philipp64|laptop: does the device actually exist? Feb 07 00:32:06 the directory entry does... and dmesg tells me there's a device found there.... Feb 07 00:34:20 [ 12.997874] sd 2:0:0:0: [sdb] 31326208 512-byte logical blocks: (16.0 GB/14.9 GiB) Feb 07 00:59:36 Ok, got the usb stick to mount... yay... and burnt the new BIOS to the board... Feb 07 00:59:55 still seeing general unhappiness, though: http://fpaste.org/mkrQ/ Feb 07 01:00:54 what .config symbols do I need to enable for debugging? Feb 07 02:00:04 russell--: still around? Feb 07 02:44:25 philipp64|laptop: yes **** ENDING LOGGING AT Tue Feb 07 02:59:58 2012