**** BEGIN LOGGING AT Wed Dec 11 02:59:59 2013 Dec 11 10:37:10 luka r39029 packages/utils/mtd-utils/Makefile * [packages] mtd-utils: add missing dependency Dec 11 10:56:59 luka r39030 packages/libs/ libsocketcan libsocketcan/Makefile * [packages] libsocketcan: add package Dec 11 10:57:00 luka r39031 packages/utils/ canutils canutils/Makefile * [packages] canutils: add package Dec 11 10:58:19 luka r39032 trunk/package/kernel/linux/modules/can.mk * kernel: add CAN modules Dec 11 10:58:21 luka r39033 trunk/target/linux/ imx6/patches-3.10/0061-can-flexcan-use-correct-clock-as-base-for-bit-rate-c.patch imx6/patches-3.10/0060-flexcan.patch imx6/patches-3.10/0062-ARM-i.MX6q-fix-the-wrong-parent-of-can_root-clock.patch * imx6: add flexcan support Dec 11 10:58:22 luka r39034 trunk/target/linux/ imx6/files-3.10/arch/arm/boot/dts/imx6dl-gw52xx.dts imx6/files-3.10/arch/arm/boot/dts/imx6q-gw54xx.dts imx6/files-3.10/arch/arm/boot/dts/imx6dl-gw53xx.dts imx6/profiles/120-gateworks.mk * imx6: add canbus support for Ventana boards Dec 11 11:25:20 jow_laptop: do you know how i could find out why "wifi" doesn't work anymore - when i start the router with the wifi disabled and enable it in the config - "wifi reload" doesnt bring up the interface - when i do a /etc/init.d/network restart - it works Dec 11 11:25:47 aterwards its possible the stop and reload the interface with wifi Dec 11 11:26:18 use ubus call network reload Dec 11 11:28:36 jow_laptop: why doesn't wifi work anymore? wifi down seems to work when the interface is up - should i also do a "ubus call network reload" instead? Dec 11 11:36:28 because the wifi implementation was changed Dec 11 11:40:51 jow_laptop: thanks for the info Dec 11 11:54:07 where does openwrt collect entropy by default? Dec 11 11:56:41 seems that brcm636x hwrng is not used Dec 11 11:57:20 jow_laptop: after i tried i a few times i now get this: ubus call network reload Dec 11 11:57:23 Command failed: Not found Dec 11 11:57:38 no idea what happens with marvell or ixp targets Dec 11 12:07:32 tripolar_: netifd died? Dec 11 12:07:55 jow_laptop: yes Dec 11 12:09:46 jow_laptop: Dec 11 12:09:46 Wed Dec 11 12:42:05 2013 user.emerg syslog: Instance network::instance1 s in a crash loop 6 crashes, 9 seconds since last crash Dec 11 12:09:50 Wed Dec 11 12:42:09 2013 auth.info sshd[3716]: Received disconnect from 78.142.159.226: 11: disconnected by user Dec 11 13:07:17 tripolar_: is there a core dump in /tmp ? Dec 11 13:14:47 jow_laptop: coredumps should be functional again for netifd ;) Dec 11 13:15:46 jow_laptop: yes some Dec 11 13:16:06 tripolar_: can you upload both the coredumps and your netifd builddir? Dec 11 13:16:35 the cores yes the builddir later - because i can'T reach the build pc now Dec 11 13:16:41 okay Dec 11 13:16:59 ping me when you got it Dec 11 13:17:05 jow_laptop i will Dec 11 13:24:13 blogic: http://sprunge.us/DZAa?diff this is the patch adding rt-n14u board Dec 11 13:24:58 everything seems to work except this shows up during boot. http://sprunge.us/QTSQ Dec 11 13:29:30 loblik: this is perfectly normal Dec 11 13:32:14 ok. then it looks fine to me. how do i submit? can i just send it to the mailing list? Dec 11 13:38:50 loblik: https://dev.openwrt.org/wiki/SubmittingPatches Dec 11 13:44:14 jow_laptop: these are the dumps Dec 11 13:44:16 https://anonfiles.com/file/a91bacef6bde7293fe3410d416008282 Dec 11 13:44:36 jow_laptop: i will send the netifd builddir later Dec 11 13:49:51 jow_laptop: did you see the link? Dec 11 13:50:58 yes Dec 11 15:19:28 I tried to build OpenWRT firmware for a DLink DIR600 rev.A1 to no avail. The compilation is successful to produce a lot of firmware files, but no 'openwrt-ar71xx-generic-dir-600-a1-squashfs-factory.bin' firmware file for a DLink DIR-600 rev.A1. I also posted this on the forum as shown here (https://forum.openwrt.org/viewtopic.php?id=47789). Can anyone help? Dec 11 15:20:28 mazilo, have you looked at your build logs? Dec 11 15:21:01 * russell-- recalls that happened to me once, when i had accidentally left debugging or something on and the image was too huge Dec 11 15:21:11 russell--: No. Perhaps, I should run a 'make BUILDLOG=1'? Dec 11 15:21:30 something like that Dec 11 15:21:38 russell--: OK. Dec 11 15:25:15 russell--: After I run a 'make BUILD_LOG=1', where should I look, i.e. logs/target/linux ? Dec 11 15:47:02 mazilo: did you find it? Dec 11 15:47:39 russell--: I think so. It looks like the firmware size is too big. Dec 11 15:47:50 yes, that sounds familiar Dec 11 16:25:04 russell--: I have removed/regenerated the .config file with 'make menuconfig'. In it, I added LuCI SSL, QoS, uPnP, IPv6 support, and the logs/target/linux/install.txt file still shows this: "Warning: /opt/openwrt-svn-trunk/bin/ar71xx/openwrt-ar71xx-generic-dir-600-a1-squashfs-sysupgrade.bin is too big (> 3538944 bytes)". :( Dec 11 16:26:06 russell--: A Fry's FR-RTR54 (a.k.a DLink DIR-600 rev.A1) has 4MB flash. Dec 11 16:26:32 russell--: And, this used to build without a problem. Dec 11 16:26:47 there have been various similar reports recently Dec 11 16:26:54 something probably got bloated again Dec 11 16:29:03 jow_laptop: Thank you. I hope it will get fixed soon. Dec 11 16:29:39 The snapshots (http://downloads.openwrt.org/snapshots/trunk/ar71xx/) show some of the firmware is 15MB in size! Dec 11 16:31:07 this can happen when the firmware is padded to make factory flash routines happy Dec 11 16:31:16 I believe this also affected BRCM63XX platform. Dec 11 16:31:48 don't know why it's the case for the alfa ap96 and the qca ap135 though Dec 11 16:31:59 heffer: Perhaps, we need to add an option to disable/enable the padding. Dec 11 16:32:19 its not about padding Dec 11 16:32:27 its about stuff inside the image becoming big Dec 11 16:32:35 for whatever reason Dec 11 16:33:50 people love to blame luci for example, but that had almost no changes in the past months Dec 11 16:33:56 so its likely kernel, userland etc. Dec 11 16:37:12 also 4MB is just tiny. if you want moar, get a device with moar flash! Dec 11 16:40:04 this looks like a nice device: http://www.gateworks.com/product/item/ventana-gw5410-network-processor Dec 11 16:40:06 mazilo: disabling the padding does not make sense, as it does not take away any usable flash space (because all the padding is marked as usuable flash), and it is actually needed for those devices enabled (else it will trigger annoying side effects) Dec 11 16:45:14 if we get a solution for this https://lists.openwrt.org/pipermail/openwrt-devel/2013-December/022842.html we basically get support for a dual band 3x3 device with 16MB flash and 128MB RAM available for 29US$ Dec 11 16:45:41 I wonder what speeds it will be able to route / nat, with qos enabled etc Dec 11 16:46:12 tharvey: do you have any numbers on that ? Dec 11 17:40:19 jow_laptop: here is the build folder of netifd - https://anonfiles.com/file/2840cf8e6257f162dd918f9237aaac22 Dec 11 17:56:12 tripolar: i think i figured out the bug Dec 11 17:56:44 missing an avl_delete call in the wireless device update Dec 11 18:15:51 nbd okay - will "wifi reload" work again or is this command gone forever? Dec 11 18:16:15 jow_laptop: I recompiled OpenWRT without any LuCI package and the problem remains the same "Warning: /opt/openwrt-svn-trunk/bin/ar71xx/openwrt-ar71xx-generic-dir-600-a1-squashfs-sysupgrade.bin is too big (> 3538944 bytes)". This confirms what you said ealier that people blame LuCI is unfounded. Dec 11 18:16:19 nbd: if you want i can test the patch Dec 11 18:23:03 tripolar: i was able to reproduce the bug Dec 11 18:23:10 tripolar: fixes will be committed in a minute Dec 11 18:24:13 nbd: nice, thanks. but i still have the same question will "wifi reload" work again or is "ubus call network reload" the new way to do the same? Dec 11 18:25:17 wifi reload calls 'ubus call network reload' Dec 11 18:26:05 nbd r39035 trunk/package/network/config/netifd/Makefile * netifd: fix crashes triggered by adding/removing wireless devices on reload Dec 11 18:27:15 nbd okay thanks Dec 11 19:50:31 nbd r39036 trunk/package/base-files/files/sbin/wifi * /sbin/wifi: reload network/wireless on wifi restart Dec 11 20:15:57 nbd: i don't use the latet patch but when i call ubus call netowrk reload - the wifi doesn't come up anymore without error - after i started and stopped it a few times Dec 11 20:16:12 i mean this patch - https://dev.openwrt.org/changeset/39036 Dec 11 20:17:16 is there some user-space app for reading PHY registers? Dec 11 20:17:44 it's possible with ioctl... right? so there should be one Dec 11 20:49:34 nbd even with the latest patch applied the wifi doesnt come up anymore after some up and downs Dec 11 21:23:31 * russell-- had a report (observed personally) that wifi on an ar71xx dual-band device stopped accepting connections from some devices (including mine), could see the network, but not connect. windows complained about low signal. it's possible that 2.4ghz just devices stopped working, maybe. wifi restart fixed it. Dec 11 21:28:45 yeah, that's what it looks like happened. about 2 days ago i stopped seeing hostapd traffic on the 2.4GHz radio Dec 11 21:28:53 in the syslogs Dec 11 21:43:46 tripolar: anything in the process list when that happens? what does 'wifi status' show? Dec 11 21:56:38 i'm trying to get sysupgrade to work on my new ramips board. sysupgrade says Dec 11 21:56:41 Sysupgrade is not yet supported on generic. Dec 11 21:57:11 and cat /tmp/sysinfo/board_name prints generic. what am i missing? Dec 11 22:08:22 loblik: look in target/linux/ramips/base-files/lib/ramips.sh ? Dec 11 22:24:33 i added board name to the file and now sysupgrade runs however it doesn't boot up then. but mtd write upgrade works. Dec 11 22:33:49 look at what sysupgrade is actually doing and whether it is appropriate for your board? Dec 11 23:09:15 it works. but i must manually run . /lib/ramips.sh ; ramips_board_detect before sysupgrade **** ENDING LOGGING AT Thu Dec 12 02:59:59 2013