**** BEGIN LOGGING AT Wed Aug 27 02:59:58 2014 Aug 27 03:13:55 build #168 of x86_64 is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/x86_64/builds/168 Aug 27 03:36:41 http://wiki.openwrt.org/toh/ubiquiti/unifi#network.configuration.unifi.as.standalone.wireless.router looks to be because of a missing unifi (& other ubnt boards?) case in target/linux/ar71xx/base-files/etc/uci-defaults/02_network & falling through Aug 27 03:58:20 build #68 of brcm47xx.mips74k is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/brcm47xx.mips74k/builds/68 Aug 27 05:32:56 build #521 of octeon is complete: Failure [failed compile_4] Build details are at http://buildbot.openwrt.org:8010/builders/octeon/builds/521 Aug 27 05:35:58 build #255 of mvebu is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/mvebu/builds/255 Aug 27 07:14:03 cyrus r42306 packages/lang/luasoap * luasoap: moved to github Aug 27 08:50:10 nico r42307 trunk/package/kernel/linux/modules/spi.mk * kernel: revert [40784] Aug 27 08:53:58 i've problem with compiling kernel, I get: bc: error while loading shared libraries: libfl.so.2: cannot open shared object file: No such file or directory Aug 27 08:54:25 http://pastebin.com/AUUvtRY4 any idea? Aug 27 08:55:17 it seems that builbot can compile it just fine: http://buildbot.openwrt.org:8010/builders/gemini/builds/558 Aug 27 09:01:07 zajec: did you upgrade your system recently? Aug 27 09:01:57 you could try removing build_dir/host and running make again Aug 27 09:06:00 it was a clean git clone Aug 27 09:06:06 done only for this single build Aug 27 09:06:56 and there is build_dir/host/flex-2.5.39/.libs/libfl.so.2 Aug 27 09:07:08 ok, i've removed it anyway Aug 27 09:07:42 i've removed target subdir as well, leaving only a toolchain Aug 27 09:18:06 i'll add some fixes that might help Aug 27 09:18:54 same problem again Aug 27 09:20:33 nbd r42308 trunk/tools/flex/Makefile * tools/flex: disable building shared libfl.so Aug 27 09:20:43 nbd r42309 trunk/tools/Makefile * tools/bc: add missing dependency on bison Aug 27 09:21:01 try removing build_dir/host, staging_dir/host and then build again Aug 27 09:21:05 after updating Aug 27 09:21:05 ok Aug 27 09:21:59 i'll just wait for git sync :) Aug 27 09:22:53 it's in there now Aug 27 09:29:35 compilig Aug 27 09:49:56 nbd: works, thanks! Aug 27 09:52:58 nbd r42310 branches/barrier_breaker/tools/flex/Makefile * tools/flex: disable building shared libfl.so Aug 27 09:52:59 nbd r42311 branches/barrier_breaker/tools/Makefile * tools/bc: add missing dependency on bison Aug 27 10:10:00 https://dev.openwrt.org/ticket/17445 -> seems important Aug 27 10:10:29 n3ph_laptop: what's that? Aug 27 10:16:57 trying this atm... every 3.14.x kernel doesn't build atm on ar71xx Aug 27 10:29:28 n3ph_laptop: works for me ? Aug 27 10:29:41 Linux wrt0 3.14.17 #1 Mon Aug 25 16:00:34 CEST 2014 mips GNU/Linux Aug 27 10:29:49 machine : Ubiquiti RouterStation Pro Aug 27 10:30:15 works here Aug 27 10:30:24 ..too Aug 27 10:31:14 didn't work first time I tried, but I had been messing around as usual, later it did compile after make clean Aug 27 10:31:35 stintel: uhm - but you mean WITH this patch or without? Aug 27 10:32:01 ah but I think it is not relevant for my devices Aug 27 10:32:12 yeah - might be Aug 27 10:32:18 not using yaffs, missed the mikrotik part Aug 27 10:32:46 that seems to be the point - anyhow - defect is fixed with this patch applied Aug 27 10:33:26 I should probably get coffee first :) Aug 27 10:34:37 this should get merged... i am not familar with the processes yet - is there still sth. to do yet or will this be done by some maintainer theese days? Aug 27 10:39:14 n3ph_laptop1: stintel: only the mikrotik subtarget uses yaffs, so I assume only that one is affected Aug 27 10:39:26 KanjiMonster: correct Aug 27 11:13:45 hi. huawei b593, fairly typical broadcom 5358 but with huawei specific hacks in cfe resulting problems detecting flash partitions with openwrt. any tips on resolving that? some details here: http://pastebin.com/CyAkMgft Aug 27 11:14:19 i tried to force partition layout using mtdparts kernel parameter like ar71xx devices do but that seems to be ignored by bcm47xxsflash Aug 27 11:28:40 nico r42312 packages/lang/luasql * luasql: moved to github Aug 27 12:01:13 jr-: huh, two firmware images? that's a bit crazy Aug 27 12:10:03 wigyori r42313 trunk/target/linux/sunxi/ (112 files in 2 dirs) * sunxi: initial 3.14 patchset Aug 27 12:11:37 wigyori: thanks :) Aug 27 12:13:48 <_tano_> Hi everybody Aug 27 12:25:57 <_tano_> I'm working on a AR9344-based board. I would like to use the High Speed UART1 to read some data from GPIO (in particular GPS NMEA info). After doing some research I'm quite sure that AR9344 UART1 is the same serial interface that is mounted on AR9331. I'm guessing if I can use the ar933x_uart driver (provided in the kernel) to control the AR9344 UART1 serial interface Aug 27 12:26:12 <_tano_> Is there anybody that can confirm that? Aug 27 12:26:22 <_tano_> Or give me some tips? Aug 27 12:26:24 <_tano_> Thanks Aug 27 12:27:41 jr-: could you try to dump "TR0969" aka "auxfs" partition content? Aug 27 12:28:01 if the board file already registers the uart, then you should just have a second dev, otherwise, youll need to modify your board file to register it, but then it's just another tty Aug 27 12:29:11 Last night, I posted this article (https://forum.openwrt.org/viewtopic.php?id=52415) on OpenWRT Forum. So far, it has been read for 44x sans any response. I wonder if anyone here knows the answer to my inquiry in the post? Aug 27 12:29:41 wasn't this on the mailing list _really_ recently? Aug 27 12:30:01 look for the thread about adding separte debug info Aug 27 12:30:09 also, what do you need them unstripped for? Aug 27 12:30:14 make package/foo/compile STRIP=true Aug 27 12:30:20 karlp: I wouldn't know because I don't subscribe to mailing list? Is there an archive to the mailing list I can access to check? Aug 27 12:30:25 there you go, jow to the rescue :) Aug 27 12:30:41 karlp: for personal reason. Aug 27 12:30:51 jow_laptop: Thanks. Aug 27 12:31:03 note, STRIP=true does not mean "enable stripping" but it means "replace .../bin/strip" with ".../bin/true" Aug 27 12:31:14 so do not try STRIP=false, it will fail Aug 27 12:31:33 jow_laptop: Is there a way to include STRIP=true in the OpenWRT Makefile for the package? Aug 27 12:31:42 jow_laptop: OK. Aug 27 12:31:59 adding STRIP:=true to the makefile should work, yes Aug 27 12:32:12 jow_laptop: OK. Aug 27 12:34:50 Mazilo: I'd still love to hear of a reason to want unstripped on the target. Aug 27 12:35:12 probably due to unawareness of gdbserver Aug 27 12:35:15 karlp: To make debugging easier. Aug 27 12:36:07 jow_laptop: Thanks for the above explanation. I will update the post with your explanation for posterity purposes. Aug 27 12:38:10 Mazilo: right, but you can use the local unstripped ones with gdb too, Aug 27 12:38:52 karlp: Yes. But, I need to experiment doing that on the hardware Aug 27 12:39:15 local unstripped ones are available in staging_dir/target-*/root-*/ Aug 27 12:40:00 <_tano_> karlp: Thanks, I'll try Aug 27 12:40:22 KanjiMonster: Yes. And, it is also available in the $(PKG_INSTALL_DIR). Aug 27 12:41:27 added to http://wiki.openwrt.org/doc/devel/debugging#add.and.modify.compiler.debug.flags now as well. Aug 27 12:53:48 jowlaptop: I added "STRIP=true" in the OpenWRT Makefile for the package as shown here (http://paste.debian.net/117839), and the binary file still gets stripped. However, when I did a 'make package/foo/compile STRIP=true', it works and the binary file doesn't get stripped. I sure would prefer the former. This way, I don't have to manually compile the package with a 'STRIP=true'. Aug 27 13:00:28 When globally compile OpenWRT by issuing 'make', is there a way to speicify a 'STRIP=true' for a specific package only? Aug 27 15:16:30 zajec. yes there's two firmwares. one is for recovery and can be booted by holding wps+wlan+reset for x seconds at startup. which is "permanent" in sense that it changes nvram flag and device will only boot recovery fw until nvram is changed from recovery side. tr0609 is empty (0x00) Aug 27 15:17:08 nice research, thanks for info :) Aug 27 15:17:10 both cfe and main firmware have serial port disabled (not initialized) unless disable_console=0 is present in nvram. recovery fw enables serial always. Aug 27 15:17:14 could you provide nvram content? Aug 27 15:18:02 i can provide dump of whole flash if you want. 256MB secondary flash (connected via usb) doesn't contain anything interesting, just copies of trx files and lte modem firmware image Aug 27 15:18:53 just nvram will be fine :) Aug 27 15:19:38 just a second Aug 27 15:23:56 zajec. http://62.236.109.200/b593/ Aug 27 15:24:29 hi Aug 27 15:24:37 I don't quite understand something in http://wiki.openwrt.org/doc/devel/config-scripting#callbacks Aug 27 15:24:39 two zips, one with "auxfs" and "nvram" mtd partitions dumped from stock firmware. bigger one has all partitions except primary os and recovery os dumped using huawei "flashtool". Aug 27 15:24:47 "You can also alter option_cb from config_cb based on the section type. This allows you to process every single config section based on its type individually." Aug 27 15:25:01 what is meant by "alter option_cb from config_cb" exactly? Aug 27 15:25:49 zajec. bit more about hardware and how to enable serial console on my blog. http://blog.asiantuntijakaveri.fi/2014/08/whats-inside-huawei-b593u-12-lte-router.html http://blog.asiantuntijakaveri.fi/2014/08/serial-console-on-huawei-b593u.html Aug 27 15:31:04 oh, ok, defining option_cb *inside* config_cb Aug 27 15:31:06 makes sense Aug 27 15:40:20 jr-: when you write a firmware, where does it get written? Aug 27 15:40:35 at 0x40000 or 0xA40000 Aug 27 15:40:49 i'd expect 0x40000 Aug 27 15:40:57 it's bigger and it's called "Main image" Aug 27 15:41:27 but the "rootfs" partition has negative size... did you install some hacked firmware? Aug 27 15:41:39 ah no... wait Aug 27 15:42:03 "rootfs" partition is fine Aug 27 15:42:22 but why OpenWrt didn't detect JFFS2 "dead code" and why it didn't create "rootfs_data"? :| Aug 27 15:42:54 mtdsplit: no squashfs found in "bcm47xxsflash" Aug 27 15:43:01 i don't understand this part :| Aug 27 15:46:19 zajec. beginning of trx has different bit for "main image" and "subject image". probably depends on that where it gets written. so far i've only uploaded openwrt image to "main image" which then ends up corrupting flash triggering nvram reset which in turn disables serial port... :) Aug 27 15:47:21 zajec. i'm now tftp booting kernel and mounting root over nfs. i also removed mtd touching parts of init scripts to avoid recovering device after each boot. Aug 27 15:48:37 oh, so that are some of your hacks Aug 27 15:48:38 ok Aug 27 15:48:58 zajec. this device also requires either elf image over tftp or lzma compressed image on flash. except regular lzma even with those recently committed parameters fail. i have to use "lzma_4k" from asus gpl release to compress openwrt lzma loader and include that in trx image Aug 27 15:49:36 oh no... Aug 27 15:49:45 not another LZMA compression Aug 27 15:49:47 omg :| Aug 27 15:50:21 trx format is also different, cfe reads firmware version info etc. from between trx headers and beginning of lzma compressed data. headers are not before trx like most (all other?) devices Aug 27 15:51:29 oh, indeed Aug 27 15:51:37 0x00000004011c-0x0000001a14e8 : "linux" Aug 27 15:51:58 the beginning of trx data is usually at 0x1c offset, not 0x11c offset Aug 27 15:55:15 trx headers http://pastebin.com/8A6pWh4G Aug 27 17:14:25 luka r42314 trunk/target/linux/kirkwood/config-3.14 * kirkwood: refresh kernel config Aug 27 17:41:17 build #669 of ppc44x is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/ppc44x/builds/669 Aug 27 18:43:58 build #53 of ramips.rt3883 is complete: Failure [failed compile_4] Build details are at http://buildbot.openwrt.org:8010/builders/ramips.rt3883/builds/53 Aug 27 18:48:13 build #275 of imx6 is complete: Failure [failed compile_4] Build details are at http://buildbot.openwrt.org:8010/builders/imx6/builds/275 Aug 27 18:49:29 build #700 of ramips is complete: Failure [failed compile_4] Build details are at http://buildbot.openwrt.org:8010/builders/ramips/builds/700 Aug 27 18:50:54 build #638 of sibyte is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/sibyte/builds/638 Aug 27 19:46:25 build #650 of avr32 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/avr32/builds/650 Aug 27 23:01:39 anyone know blogic's https://dev.openwrt.org/changeset/42187 intentions? Aug 27 23:38:12 octeon also hasn't built since the 18th because of at least a missing PATA_OCTEON_CF symbol Aug 27 23:48:43 swalker: submit a patch? Aug 27 23:50:05 I could in a futile attempt I suppose, but what are maintainers for? and why aren't commits that break builds backed out? Aug 27 23:52:09 a patch can't hurt Aug 27 23:53:21 you're assuming it actually gets looked at which isn't exactly a given Aug 27 23:55:20 if it's not looked at, nothing is lost still Aug 28 02:11:17 build #621 of kirkwood is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/kirkwood/builds/621 Aug 28 02:40:38 build #715 of orion is complete: Failure [failed compile_8] Build details are at http://buildbot.openwrt.org:8010/builders/orion/builds/715 Aug 28 02:41:01 build #680 of uml is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/uml/builds/680 Aug 28 02:44:26 build #574 of iop32x is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/iop32x/builds/574 **** ENDING LOGGING AT Thu Aug 28 02:59:58 2014