**** BEGIN LOGGING AT Wed Sep 10 02:59:58 2014 Sep 10 03:50:09 I'm seeing "ar2315-wdt ar2315-wdt.0: watchdog expired, rebooting system" on a D-Link DIR-300 A1, which has AR2317 as CPU and 16MB of RAM, anything I can do to stop this? Sep 10 03:56:16 build #651 of sibyte is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/sibyte/builds/651 Sep 10 04:54:25 build #663 of avr32 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/avr32/builds/663 Sep 10 05:43:01 build #634 of kirkwood is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/kirkwood/builds/634 Sep 10 05:45:23 build #587 of iop32x is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/iop32x/builds/587 Sep 10 06:39:44 build #754 of at91 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/at91/builds/754 Sep 10 07:11:17 build #82 of brcm47xx.mips74k is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/brcm47xx.mips74k/builds/82 Sep 10 07:20:40 hi, is it the case that lzma-loader will only be used by boards with u-boot that do not do lzma decompress? Sep 10 07:23:18 swalker: ping Sep 10 09:19:15 build #271 of mvebu is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/mvebu/builds/271 Sep 10 12:24:24 my god, AR231X to be removed :( Sep 10 12:26:19 it's been on and off for a while. Sep 10 12:26:31 * karlp hopes sergey can get a bit more time. Sep 10 12:27:18 Devastator: you mean from ath5k? well yes, it makes sense; nobody but us can test it, and the code can be added back if the board support finally lands upstream Sep 10 12:29:53 the target name will likely change before that, too (atheros is too generic, especially as it only targets a subset of atheros socs - so e.g. ath25 would be better suited, to be similar ath79) Sep 10 12:30:20 maybe then we'll get support for different machs too, instead of the same single device ;) Sep 10 12:31:04 or maybe even the current diehard ar2317 users in oipenwrt will get new hardware, and will simply die of starvation Sep 10 12:56:47 nbd r42456 trunk/package/system/procd/files/procd.sh * procd: allow adding empty strings to arrays Sep 10 12:56:52 nbd r42457 trunk/target/linux/ar71xx/files/drivers/net/ethernet/atheros/ag71xx/ag71xx_main.c Sep 10 12:56:52 ar71xx: ack completed tx descriptors only after the full frame has been completed Sep 10 12:57:35 nbd r42458 branches/barrier_breaker/package/system/procd/files/procd.sh * procd: allow adding empty strings to arrays Sep 10 13:02:21 nbd r42459 trunk/package/network/services/hostapd/patches/600-ubus_support.patch * hostapd: add ubus bindings for wps Sep 10 13:35:01 KanjiMonster well, I'm testing BB RC3 on a friend's router, it's a DIR-300 A1 which has AR2317 Sep 10 13:35:54 yesterday I installed serial header to discover why it is rebooting within minutes of uptime, and it seems related to watchdog expiring Sep 10 13:36:12 I don't know what it means but doesn't look good.. Sep 10 13:46:28 16meg is going to be sucky on ar2317.. Sep 10 13:46:45 I had some 16meg devices for a while, but even on AA, it was flaky. Sep 10 13:50:37 nbd r42460 trunk/target/linux/atheros/Makefile * atheros: enable the low_mem flag to better deal with 16M RAM devices Sep 10 13:51:22 nbd r42461 branches/barrier_breaker/target/linux/atheros/Makefile * atheros: enable the low_mem flag to better deal with 16M RAM devices Sep 10 13:51:26 karlp: maybe that will help :) Sep 10 13:52:25 isn't that enabled already? swalker showed me a changeset enabling it Sep 10 13:53:00 Devastator: ? Sep 10 13:54:03 https://dev.openwrt.org/changeset/41532 Sep 10 13:54:35 that's a different target Sep 10 13:54:50 wonderful undocumented knobs to twiddle :| Sep 10 13:54:57 what does low_mem actually _do_ ? Sep 10 13:55:12 it reduces the squashfs block size Sep 10 13:55:18 slightly more flash use, but much lower ram use Sep 10 13:55:51 nbd also, do you have any idea why I'm getting "ar2315-wdt ar2315-wdt.0: watchdog expired, rebooting system" on this router? Sep 10 13:56:18 if the system is under heavy memory pressure, a watchdog reboot is a likely outcome Sep 10 13:57:14 hum, so what you just pushed should fix this as well Sep 10 13:57:16 thanks! Sep 10 13:58:15 luka r42462 trunk/include/kernel-version.mk * kernel: remove unused versions Sep 10 14:01:40 so for atheros users with 32meg of ram, how much flash are we paying for this? Sep 10 14:01:49 5%? 25%? Sep 10 14:02:56 is that currently target wide? or is even an option on per platform? Sep 10 14:03:44 depends. sometimes it's one erase block more Sep 10 14:03:58 at the moment it's a per platform feature flag Sep 10 14:04:12 btw. it only adjusts the default value for the block size Sep 10 14:04:20 you can use make menuconfig to put in your own value if you want Sep 10 14:07:43 I saw that discussion yesterday I think Sep 10 14:07:54 about Huawei E970 Sep 10 14:11:29 I really hope AR231X doesn't die just now Sep 10 14:24:23 nbd can your latest commits be backported to BB? thanks Sep 10 14:26:53 already did, right after the commit Sep 10 14:29:54 sorry for that, I guess I'm blind today Sep 10 14:33:37 build #535 of octeon is complete: Failure [failed compile_4] Build details are at http://buildbot.openwrt.org:8010/builders/octeon/builds/535 Sep 10 14:48:05 20GB should be enough for a VM to build BB, right? Sep 10 15:01:06 nbd: you might know, is there a way to make procd reload inittab? like "init -q" used to do? Sep 10 15:01:50 karlp: blogic replied to your mail, iirc Sep 10 15:01:55 saying that this is not currently possible Sep 10 15:08:34 hrm, I wonder why I didn't see that at home last night. Sep 10 15:08:43 thanks for the pointer, I can see it in the mail archives :) Sep 10 17:50:22 where can I find LED definitions for atheros target? Sep 10 20:11:35 karlp do you have any pointers about where leds are defined for AR2317? Sep 10 20:12:21 wlan is lighting up sysled instead when enabled Sep 10 20:29:04 nbd: it seems there may be another package requiring PKG_USE_MIPS16: https://dev.openwrt.org/ticket/17710 Sep 10 20:29:11 nbd: do you think you could look at this? Sep 10 20:29:17 nbd: or should we wait for blogic? Sep 10 20:32:28 build #289 of imx6 is complete: Failure [failed compile_4] Build details are at http://buildbot.openwrt.org:8010/builders/imx6/builds/289 Sep 10 21:05:45 zajec: it's been broke for a month now Sep 10 21:41:08 luka r42463 trunk/ (376 files in 24 dirs) * kernel: update 3.14 to 3.14.18 Sep 10 23:05:24 Devastator: try something like this: https://github.com/remakeelectric/openwrt-bb/commit/c836b5c9a0a9222c5d0b100b06fdd93f369ed52c Sep 10 23:06:04 you can also get jiggy with quilt and edit the original board patches, but I'm not (in?)sane enough to keep those patches in my head. Sep 10 23:15:50 luka r42464 trunk/config/Config-kernel.in * config: enable cgroup freezer Sep 10 23:23:10 build #683 of ppc44x is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/ppc44x/builds/683 Sep 11 02:58:40 build #729 of orion is complete: Failure [failed compile_8] Build details are at http://buildbot.openwrt.org:8010/builders/orion/builds/729 **** ENDING LOGGING AT Thu Sep 11 02:59:58 2014