**** BEGIN LOGGING AT Tue Dec 10 02:59:58 2013 Dec 10 11:43:31 nbd: looks like aa-mac80211.git isn't compatible with AA anymore Dec 10 11:43:42 install -m0644 ./files/netifd.sh /home/stijn/Development/OpenWrt/attitude_adjustment/build_dir/target-mips_r2_uClibc-0.9.33.2/hostapd-wpad-full/hostapd-20131120/ipkg-ar71xx/hostapd-common/lib/netifd/hostapd.sh Dec 10 11:43:46 install: cannot stat './files/netifd.sh': No such file or directory Dec 10 11:43:54 or I missed something :) Dec 10 12:21:30 jow r39026 packages/utils/wifitoggle/ Makefile files/wifitoggle.hotplug * wifitoggle: use ubus call network reload to apply wifi changes (#14587) Dec 10 13:01:28 blogic: thanks for the fix it works Dec 10 13:03:11 stintel: oh, forgot to remove that part Dec 10 13:05:13 stintel: fixed Dec 10 13:05:41 nbd: cheers Dec 10 13:05:42 ;) Dec 10 13:08:13 hmmm Dec 10 13:08:13 CONFLICT (add/add): Merge conflict in mac80211/patches/564-ath9k_p2p_ps_support.patch Dec 10 13:14:30 tripolar: its only intermediate Dec 10 13:14:40 i am working on the real fix just now Dec 10 13:17:18 okay - whats the problem? and how do you develop this? debugging on a router keyboard and screen somehow sucks Dec 10 13:17:36 keyboard/without keyboard Dec 10 13:19:21 ? Dec 10 13:19:32 what problem ? Dec 10 13:23:52 is anyone experienced with mt7620n based boards? is there some way to get wifi to work? Dec 10 13:25:05 loblik: its wip Dec 10 13:25:26 mt7620n boards should work without wifi Dec 10 13:25:31 whihc baord do you have ? Dec 10 13:30:02 it's this one http://wikidevi.com/wiki/ASUS_RT-N14U Dec 10 13:30:30 i wrote a dts file and everything seems working except wifi Dec 10 13:30:43 blogic> i am working on the real fix just now Dec 10 13:31:54 logread is missing the the logic for ubus reconnect Dec 10 13:32:30 so i am adding a generic reconnect to libubs to fix it and then remove the temporary fix i added last night Dec 10 13:32:57 loblik: it will take a few weeks to get wifi working Dec 10 13:33:06 blogic: how is it that it worked before? Dec 10 13:33:10 dont hold your breath, but we are working on it Dec 10 13:33:29 russell--: logd and logread get started within a few ms of eachother now Dec 10 13:33:46 it was just lucky when it worked, then? Dec 10 13:34:07 russell--: likely Dec 10 13:34:11 and when loading from slow flash the logd object is not registered on ubus when logread starts Dec 10 13:34:43 do logread enters its mainloop with no valid ubus/log context and never treis to reconnect Dec 10 13:35:05 logd collects syslog messages, and logread subscribes to them? Dec 10 13:35:17 klog and syslog Dec 10 13:35:35 logread subscribes to the events broadcast from logd Dec 10 13:35:52 and depending on the options passed puts the lines in a file, a socket or stdout Dec 10 13:36:10 and, what is ubus? is that used for other things as well? Dec 10 13:36:14 loblik: can you submit the dts file ? Dec 10 13:36:21 russell--: its like dbus Dec 10 13:36:30 an ipc framework for communication between processes Dec 10 13:39:37 blogic: sure. but i guess it needs some review. i'm not sure how define leds. Dec 10 13:40:43 which part is unclear ? Dec 10 13:41:15 sorry i gotta go now. i will be back in few hours. Dec 10 13:58:24 * russell-- just read a little about kdbus http://lwn.net/Articles/551969/ Dec 10 13:59:35 blogic could it be that the wifi is broken? it doesnt come up here with the latest trunk, created a few hours ago Dec 10 14:00:35 no idea Dec 10 14:02:09 im rebuilding the whole firmware maybe a build bug Dec 10 15:05:44 [florian], KanjiMonster: i think that i can make a diff for initial support for the BCM3380. The problem is that the usb is otg and ethernet driver surely is different than other bcm targets Dec 10 15:06:07 which are your opinions? Dec 10 15:07:32 sorry for my poor english :( Dec 10 15:10:33 Pteridium: talk to [florian], I think he also started adding support for it Dec 10 15:11:08 ok, thanks Dec 10 15:11:44 do you plan to push bcm63xx to kernel 3.12? Dec 10 15:12:12 Pteridium: also can you do one more test for me regarding the smp fs issue? You said jffs2-only also has the problem. Can you try disabling xattr-support for jffs2? this should be ok for non-overlay images Dec 10 15:12:23 I plan to push some updates to 3.10 first Dec 10 15:12:35 perfect Dec 10 15:12:54 also I don't plan to push bcm63xx to 3.12 (or anything) else until BB is branched Dec 10 15:13:58 ok, now understand: BB will be released soon as stable branch Dec 10 15:14:07 unlikely Dec 10 15:14:12 BB will take a while Dec 10 15:14:46 but the plan is to have 3.10 as the kernel for BB, so until BB is branched off, 3.10 stays the main kernel version Dec 10 15:15:22 of course: there are some things in BB that still need some debug Dec 10 15:17:52 last question: will be BB updated to kernel 3.10.23? Dec 10 15:26:26 KanjiMonster: compiling with the option you told Dec 10 15:31:56 next round will compile with kernel 3.10.23 Dec 10 15:35:07 Pteridium: BB will be updated in unregular intervals to the newest kernel releases Dec 10 15:35:25 ok Dec 10 15:36:17 so eventually it will be on 3.10.23 or newer Dec 10 15:37:05 newer? .23 is the last 3.10 version if i'm not wrong Dec 10 15:37:34 tripolar: I'm also seeing an issue with wifi not coming up after a reboot. It seems to work fine when I log in an bring it up manually. Dec 10 15:37:37 3.10 is a longterm kernel, there will be a 3.10.24. and a 3.10.25. etc pp Dec 10 15:37:55 yes, you are right Dec 10 15:45:13 Have the buildbots moved or are there some problems (regarding their ui/server)? Dec 10 15:50:04 sungami i was a pebkac - i used a wring configure option now it works Dec 10 16:26:59 jow r39027 trunk/package/network/services/hostapd/files/netifd.sh * hostapd: fix short_preamble option Dec 10 16:27:02 jow r39028 trunk/package/kernel/mac80211/files/lib/netifd/wireless/mac80211.sh * mac80211: add missing set_default to fix stray "radio0: sh: bad number" messages Dec 10 16:41:48 jow: can you close https://dev.openwrt.org/ticket/13258 ? the last post shows a working solution and i added an entry in the wiki regarding that Dec 10 16:45:20 plntyk: this solution looks wrong Dec 10 16:45:41 if a cross compiled package is fixed by updating headers on the host system then it is not properly cross compiling Dec 10 16:47:32 ah ok i see .. i have a patch with libftdi1 that uses cmake and manages to find correct headers Dec 10 17:49:57 <[florian]> Pteridium: yes the ethernet driver is much different Dec 10 17:50:14 <[florian]> Pteridium: and I wonder if we are not sort of forced to use the FPM for transfering packets Dec 10 17:51:20 so bcm3380 will not be supported in the near future Dec 10 17:52:10 correction: the serial por driver can work :-/ Dec 10 17:53:44 [florian]: and the PCIe i guess Dec 10 17:54:08 <[florian]> Pteridium: I have an experimental patch for it, which just supports serial, spi, hs-spi Dec 10 17:54:16 <[florian]> not sure about the GigE unimac interface Dec 10 17:54:33 <[florian]> on one hand I'd love to see it supported, on the other end, without docsis modem support, it is kind of useless Dec 10 17:55:13 do you want to try it with the cisco epc3825? Dec 10 17:55:25 ecos based Dec 10 17:55:36 <[florian]> is this the one you have? Dec 10 17:55:40 yes Dec 10 17:55:42 <[florian]> I am currently using a bcm93380 devboard Dec 10 17:55:58 same bootloader? Dec 10 17:56:02 <[florian]> yep Dec 10 17:56:09 <[florian]> pretty much all bcm33xx are ecos-based Dec 10 17:56:18 <[florian]> it's only recently that they somehow switched to linux Dec 10 17:56:31 yes, only some netgear have linux sources Dec 10 17:56:46 <[florian]> and even there, afair, the SoC might not entirely run Linux, just part of it Dec 10 17:57:00 sure? Dec 10 17:57:37 <[florian]> not entirely, I have seen contradictory things about that Dec 10 17:58:25 understood: cable modem part runs with something different, no? Dec 10 17:59:03 <[florian]> CM traditionnaly ran eCOS Dec 10 17:59:51 yes Dec 10 18:00:41 so are some parts ecos or something else based? Dec 10 18:01:21 <[florian]> on newer designs only such as bcm3383, bcm3384, not for bcm3380 apparently Dec 10 18:02:26 excuse me, what happens in newer SoCs? Dec 10 18:02:51 [florian]: thanks for you reviews :) Dec 10 18:02:55 <[florian]> zajec: you are welcome! Dec 10 18:03:07 <[florian]> Pteridium: they make the CM part only run eCos, and the rest runs linux Dec 10 18:03:37 strange... Dec 10 18:04:18 <[florian]> well, they realized that customers want Linux applications, not some crazy eCos OS to develop with :) Dec 10 18:04:30 <[florian]> and adding CM support to Linux was probably deemed too complicated Dec 10 18:04:36 like the livebox 2, that runs linux and then openrg Dec 10 18:05:07 crazy ecos :D Dec 10 18:05:27 what is "cm"? Dec 10 18:05:33 *CM Dec 10 18:05:39 cablemodem Dec 10 18:06:01 thx Dec 10 18:07:07 zajec: bcm33xx family Dec 10 19:12:53 nbd: after the latest updates my wifi card doesn't come up anymore, any idea how i could debug this? or has the command changed? wifi up only my second wifi card as deactivated Dec 10 19:14:49 nbd /etc/init.d/network restart brings it up - is there a better way to do this? Dec 10 19:16:52 okay it seems it is wifi reload now - i used "up" some time ago and it also worked Dec 10 20:05:50 nbd okay i was wrong - something is wrong with the wifi script - when the device is started with disabled '1' and then i turn it on wiht '0' and do a "wifi" the wifi device is down - if i do a /etc/init.d/network restart - it works Dec 10 20:05:54 see http://nopaste.info/d72e85a92c.html Dec 10 20:06:22 the device is started - means when i start the router Dec 10 20:48:56 This document describes the features of Ubicom's port of Das U-Boot. -> http://pastebin.com/a75x1kWa Dec 10 20:48:59 havefun :) Dec 10 21:42:05 KanjiMonster: same result with kernel 3.10.23 Dec 10 22:25:12 mtd: partition "rootfs" must either start or end on erase block boundary Dec 10 22:25:46 should i care about this? i'm not sure why this happens Dec 10 22:47:30 I just updated to SVN r39028 and built OpenWRT for a DIR-600 rev.A1. The compilation is OK, but it didn't produce a firmware for DIR-600 rev.A1 firmware even though I specifically selected the 'Target Profile (D-Link DIR-600 rev. A1)'. Does anyone here know why? **** ENDING LOGGING AT Wed Dec 11 02:59:59 2013