**** BEGIN LOGGING AT Thu Aug 08 02:59:59 2013 Aug 08 07:28:51 blogic ping pls, this https://dev.openwrt.org/changeset/37703/ does not generate a factory image for dlink dir 505 ... Aug 08 08:15:24 dape_8708: ok, so you have a patch to make it a factory image ? Aug 08 08:15:41 of course not, i just reported te facts Aug 08 08:15:57 yes Aug 08 08:16:09 and i closed it as it is not a bug report but a feature request Aug 08 08:16:18 please use forum and irc for feature requests Aug 08 08:16:29 trac is filled with lots of crap already try to not make it worse Aug 08 08:16:37 you have been around here long enough to know better Aug 08 08:16:40 can i flash the sysupgrade image on the router? Aug 08 08:16:47 yes Aug 08 08:16:53 in uboot Aug 08 08:16:57 that i didnt know Aug 08 08:16:57 and then via sysupgrade Aug 08 08:17:09 wait, with serial acces? cause i dont have that Aug 08 08:17:18 i flashed it by booing a initramfs and then using that to flash the unit via sysupgrade Aug 08 08:17:25 yes with serial Aug 08 08:17:46 ok, so to recap: there is no way i can get openwrt on the router without opening the unit right now Aug 08 08:18:16 got it ^ Aug 08 08:18:32 ok Aug 08 08:18:47 man this is gonna hurt, i saw the serial pads are so tiny in the forum pictures Aug 08 08:19:01 be carefull i ripped mine of Aug 08 08:19:10 ;) Aug 08 08:19:29 i have a electronic wise dad, a couple of beers and his precision robotic eyes and hands will do (i hope) Aug 08 08:19:32 had to track the lanes and scratch the silkscreen of a via and solder the cable there Aug 08 08:19:51 or use clips to atach to it Aug 08 08:19:55 your honest advice is to wait or to go ahead and.. Aug 08 08:20:26 well Aug 08 08:20:28 no idea Aug 08 08:20:35 i wont take the blame if you frag the unit ;) Aug 08 08:20:42 you can try to flash it via webui Aug 08 08:20:45 i have not even tried Aug 08 08:21:05 ok, one more question: does this has the recovery webpage ? Aug 08 08:21:41 i think dlink router have a recovery webpage to flash firmware if booted with reset button, my dlink dir 615 has Aug 08 08:24:28 "the chose file is not a firmware file" meh.. let me try to recovery boot it Aug 08 08:54:26 blogic: I thought I fixed x86 preinit, but I just made it worse. :-D Aug 08 10:11:33 build #312 of rb532 is complete: Failure [failed shell_12] Build details are at http://buildbot.openwrt.org:8010/builders/rb532/builds/312 Aug 08 10:12:48 blogic, in target/linux/ar71xx/image/Makefile at line 353 in Image/Build/CameoHornet i think $$$$r and $(8) are null Aug 08 10:13:12 so maybe mkcameofw does not get the proper params Aug 08 10:13:43 build #312 of ppc44x is complete: Failure [failed shell_12] Build details are at http://buildbot.openwrt.org:8010/builders/ppc44x/builds/312 Aug 08 10:17:51 dape: ok Aug 08 10:18:16 if [ -e "/home/dani/OpenWrt/bin/ar71xx/openwrt-ar71xx-generic-dir-505-a1-squashfs-sysupgrade.bin" ]; then for r in ; do [ -n "$r" ] && dashr="-$r" || dashr=; [ -z "$r" ] && r="DEF"; mkcameofw -M HORNET -R "$r" -S "HORNET-RT-DIR505A1-3" -V -c -K 0xe0000 -I -k "/home/dani/OpenWrt/bin/ar71xx/openwrt-ar71xx-generic-dir-505-a1-squashfs-sysupgrade.bin" -o /home/dani/OpenWrt/bin/ar71xx/openwrt-ar71xx-generic-dir-505-a1-squashfs-factory$dashr.bin; done Aug 08 10:18:16 ; fi Aug 08 10:19:03 for r in ; Aug 08 10:19:09 that part seems wrong Aug 08 10:20:08 so $(9) .. Aug 08 11:07:02 build #326 of uml is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/uml/builds/326 Aug 08 13:12:55 build #296 of sibyte is complete: Failure [failed shell_12] Build details are at http://buildbot.openwrt.org:8010/builders/sibyte/builds/296 Aug 08 13:32:13 where can i get sources for 12.09.1 branch? Aug 08 13:35:12 nxis: https://dev.openwrt.org/wiki/GetSource ? Aug 08 13:35:35 there no 12.09.1 there Aug 08 13:35:40 is it the same as 12.09? Aug 08 13:36:19 nxis: commits are still being made on the 12.09 branch, yes: http://git.openwrt.org/?p=12.09/openwrt.git;a=shortlog Aug 08 13:39:01 understood. i have a very tiny experimental patch for ddns package =) i will check if it still actual for upcoming release Aug 08 13:50:49 Is it possible to fix problem with gcc-4.8.x with Linaro so that this option "Build/install java compiler and GNU classpath ?" will not cause gcc compilation crash as shown here (http://pastebin.com/tz2vDCe8) on line #3? The reason I asked this is I would like to compile jamvm package that requires classpath package. Without gcj from gcc, classpath won't compile. Aug 08 14:00:29 mazilo: i thought classpath could use a host javac Aug 08 14:00:37 not sure why it would need a cross gcj Aug 08 14:12:39 nxis: there are no separate branches for subreleases, just one for the main one (12.09), and all further releases are derived from that one Aug 08 14:35:08 build #345 of orion is complete: Failure [failed compile_8] Build details are at http://buildbot.openwrt.org:8010/builders/orion/builds/345 Aug 08 14:40:36 nbd: I thought so. But, the way it's packaged doesn't look like that way. Aug 08 14:44:09 mazilo: hello, could you please take a look at your freeswitch package. It seems that freeswitch-mod-spandsp needs liblzma as dependency too. :-) Aug 08 14:45:45 strasidlo: I looks like the commit (30c13cc7864386e18fa4bdea84f72708b174e118) I submitted on 07/27/2013 doesn't fix the issue. Aug 08 14:47:26 strasidlo: I changed the mod-spandsp line to "$(eval $(call BuildPlugin,spandsp,Span DSP,mod,,,,+libjpeg +@PACKAGE_xz:PACKAGE_xz:PACKAGE_liblzma)) # ~/conf" and that didn't help. Aug 08 14:47:58 strasidlo: I meant +@PACKAGE_xz:PACKAGE_liblzma Aug 08 14:48:25 i think including a lzma dependency is useless for most people Aug 08 14:48:32 better just force it to not use liblzma Aug 08 14:49:02 instead of keeping quirky build behavior that depends on whether xz is selected Aug 08 14:49:12 nbd: OK. I will try that through the configure line. Aug 08 14:49:19 nbd: Thanks for the pointer. Aug 08 15:00:28 nbd: Now, I just finished compiled gcc-4.8 with Linaro Enhancement to include the option "Build/install java compiler and GNU classpath ?" without a problem. The key here is the missing 301-missing-execinfo_h.patch file in toolchain/gcc/patches/4.8-Linaro directory. Aug 08 15:02:11 nbd: What I did was to copy the toolchain/gcc/patches/4.6.3/301-missing-execinfo_h.patch to toolchain/gcc/patches/4.8-linaro/301-missing-execinfo_h.patch and then 'make dirclean; make -j4'. Aug 08 15:09:45 nbd: But, classpath compilation still stumbled upon gcj as shown here (http://pastebin.com/2j1FSXPi) on line #284 and #285. :( Aug 08 15:11:27 you probably need to make a symlink from $target_cross-gcj to gcj Aug 08 15:18:56 nbd: Can you please add this link to gcc package so that it will automatically be created when building gcc with such an option? Aug 08 15:58:17 mazilo: not sure if that's a good idea, that might cause confusion with host tools Aug 08 15:58:45 either make the classpath package use the cross-gcj name Aug 08 15:58:54 or put the symlink somewhere in the package build dir Aug 08 15:58:59 and add that to PATH for the build Aug 08 16:04:34 nbd r37738 trunk/package/network/ (24 files in 2 dirs) * hostapd: update to latest version (20130707) Aug 08 16:15:54 build #306 of avr32 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/avr32/builds/306 Aug 08 16:25:04 nbd r37739 trunk/package/network/ services/hostapd/Makefile services/hostapd/files/wpa_supplicant-p2p.config Aug 08 16:25:04 hostapd: add a build variant for wpa_supplicant with p2p (aka Wi-Fi Direct) support Aug 08 16:26:36 no gdb for avr32? srs. Aug 08 16:27:17 well, avr32 is not a typical mainline-supported architecture Aug 08 16:27:21 so a lot of stuff is missing Aug 08 16:27:25 right Aug 08 16:27:25 btw. why are you using it? Aug 08 16:27:34 political reasons Aug 08 16:27:45 heh Aug 08 16:27:56 it's in a gateway we "have" to use. We want to try and push them on something better soon, but we have to deliver something on this Aug 08 16:28:24 udev didn't build either. was complaining about pipe2 and whatnot. Aug 08 16:28:41 and lighttpd is crashing Aug 08 16:29:03 <[florian]> avr32 just needs a bit of love :) Aug 08 16:29:28 yes, and there's stuff missing from gcc 4.8 for it, at least in openwrt, although it ought to have native support here Aug 08 16:29:30 there Aug 08 16:29:36 some stuff crashes GCC until I put -O0 Aug 08 16:29:45 so, good times! Aug 08 16:30:46 nbd: any advice? Aug 08 16:30:56 would specifics help? Aug 08 16:31:43 kwakkwak Aug 08 16:33:34 i don't have any advice on this Aug 08 16:33:38 aw Aug 08 16:33:53 * Chocks will make some bug reports in any case. I can always hope Aug 08 16:35:27 blogic: is this correct: https://dev.openwrt.org/browser/trunk/target/linux/lantiq/dts/danube.dtsi#L68? If yes, does ltq-vmmc do this job? Aug 08 16:36:39 i ask because i'm preparing to test voip with latest trunk rev Aug 08 16:37:08 lighttpd crash appears to be ssl-related Aug 08 17:17:11 KanjiMonster: the issues we were having with jffs2 seems to be solved with trunk r37737. We'll do deeper tests to confirm, but looks very promising. Aug 08 17:21:12 moo Aug 08 17:35:28 KanjiMonster: here are the bootlogs http://pastebin.com/ZtpmdUqr Aug 08 17:41:04 Pteridium: so this is >= r37737 or not? Aug 08 17:41:32 yes, r37737 Aug 08 17:41:48 with kernel 3.10.4 Aug 08 17:45:53 which gcc version are you using? Aug 08 17:46:45 moment.., Aug 08 17:47:15 gcc version 4.6.4 (OpenWrt/Linaro GCC 4.6-2012.12 Aug 08 17:49:18 not GCC 4.6-2013.05 ;) Aug 08 17:51:39 but if you want to test GCC 4.6-2013.05 i have no problem Aug 08 17:54:27 moo Aug 08 17:54:35 moo? Aug 08 17:54:39 exactly Aug 08 17:55:02 ? Aug 08 17:55:08 try "ip moo" from your desktop machine Aug 08 17:55:40 Chocks: did the apt-get maintainers get their hands on it? :) Aug 08 17:55:41 funny ;) Aug 08 17:55:46 (or aptitude..) Aug 08 17:55:56 Pteridium: yes, that would be good (or 4.8-linaro, that one should be good, too) Aug 08 17:56:03 I'm not sure which one was earlier Aug 08 17:57:53 KanjiMonster: i'll test these GCC versions. We'll send the feedback as soon as possible. Aug 08 17:59:15 but now i go to make a repo in github for the #seguridadwireless forum users Aug 08 20:34:08 nbd: I got your message. Thank you. Aug 08 20:45:23 nbd r37740 trunk/package/system/ubus/Makefile * ubus: update to latest version, fixes a bug in reconnect after ubusd dies Aug 08 20:45:26 nbd r37741 trunk/package/network/services/hostapd/patches/700-ubus_support.patch * hostapd: fix ubus reconnect handling Aug 08 20:45:29 nbd r37742 trunk/package/network/services/hostapd/Makefile * hostapd: fix typo in version number Aug 08 20:47:52 nbd r37743 branches/attitude_adjustment/package/ubus/Makefile * AA: ubus: sync with latest trunk version (as of r37740) Aug 08 22:08:51 build #263 of iop32x is complete: Failure [failed shell_12] Build details are at http://buildbot.openwrt.org:8010/builders/iop32x/builds/263 Aug 08 22:11:48 MartijnVdS: I have patched in another hash today (of just 5 numbers) but the router required me to enter 16 characters Aug 08 22:12:16 even return and backspace characters add a * to the password... Aug 08 22:12:43 doing a more tricky hack now... Aug 08 22:13:07 I've already managed to bypass the sha-1 hash with jtag ;) Aug 08 22:13:51 so now i'm trying to patch the hashing algorithm out from inside the lzma image... Aug 08 22:14:31 Pteridium, possibly interresting for you, I have found the secret firmware update website on my VGV7519, but it is protected with a PIN code as well Aug 08 22:14:49 again a SHA-1 hashing mechanism... Aug 08 22:16:00 The_Lizard: so, without the pin the firm can't be downloaded, i'm right? Aug 08 22:16:22 indeed... Aug 08 22:17:01 another victory, my patching in the lzma image worked...I'm now able to update MY flash images without a valid PIN code... Aug 08 22:17:08 you are making a wonderful job with this router. Congratulations! Aug 08 22:17:51 not entirely... my networking issue is gone when using BRN-boot bootloader Aug 08 22:17:52 amazing! you are great! Aug 08 22:18:13 so that means there is some evil in my u-boot loader Aug 08 22:18:31 but when using brn-boot, my storage isn't in place yet Aug 08 22:18:52 I can't seem to be able to configure my flashes correctly in the device tree... Aug 08 22:19:24 The storage can't alter my first flash contents, as it is hashed (CRC32) Aug 08 22:19:42 but you have done a hard job, and now there are some little issues Aug 08 22:20:16 so I tried to configure my image to use the second flash (the code 1 image actually) as storage... I have flashed in my squashfs root image, with the deadc0de marker, but it isn't working yet... Aug 08 22:20:31 So I guess there is a mess-up in my device tree setting... Aug 08 22:23:38 blogic and others have more knowledge about VR9 based routers; i'm still newbie and can't help you. At this development stage they are the only that can push your job forward. Aug 08 22:23:56 much more* Aug 08 22:27:47 KanjiMonster: with GCC 4.8-linaro still works fine http://pastebin.com/XkyDVVzb Another problem to the trashcan. Aug 08 22:29:55 The_Lizard: excuseme, but i was concentrated with an issue in jffs2 when enabling FPU emulation in the kernel. The board is based in BCM6368. Aug 08 22:30:07 I know, but now I have this firmware uploading in place, I hope I can test new firmware images more quickly than burning one each 15 minutes... Aug 08 22:30:17 using the XModem Aug 08 22:30:44 that is no problem, i'm going off to bed... good luck! Aug 08 22:30:55 good night Aug 08 22:33:35 The_Lizard: take in mind that if you can run openwrt in your board is a very good advance... and a hard job, of course Aug 08 22:34:13 I know, and I'm not done with it yet... Aug 08 22:34:29 i go to bed too, good night Aug 08 22:34:38 but it would be nice if there is an easy solution to flash the router for everybody... Aug 08 22:34:44 without having to open the device... Aug 08 22:36:23 the most of people here are experts, so if they have to open the router is not a problem Aug 08 22:38:30 For the arv4518 and arv7518 could be made something in this way, but if this is done a lot of people are mailing because they made a mistake Aug 08 22:39:24 good night The_Lizard Aug 08 23:07:07 build #266 of ep93xx is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/ep93xx/builds/266 **** ENDING LOGGING AT Fri Aug 09 02:59:59 2013