**** BEGIN LOGGING AT Thu Jan 30 02:59:58 2014
Jan 30 03:37:36 <builder_> build #441 of sibyte is complete: Failure [failed shell_12]  Build details are at http://buildbot.openwrt.org:8010/builders/sibyte/builds/441
Jan 30 04:44:07 <builder_> build #482 of uml is complete: Failure [failed compile_5]  Build details are at http://buildbot.openwrt.org:8010/builders/uml/builds/482
Jan 30 05:42:45 <builder_> build #451 of avr32 is complete: Failure [failed compile_5]  Build details are at http://buildbot.openwrt.org:8010/builders/avr32/builds/451
Jan 30 07:36:42 <OpenWrt-Commits> nbd r39424 trunk/include/kernel-defaults.mk * kernel: fix parallel build issues with header install
Jan 30 07:44:46 <Hopscotch> good morning
Jan 30 09:05:55 <OpenWrt-Commits> blogic r39425 trunk/package/ libs/libubox/Makefile system/ubus/Makefile system/ubox/Makefile system/procd/Makefile * libubox: ubox: procd: ubus: update to latest git head
Jan 30 09:10:24 <blogic> jow_laptop: ping
Jan 30 09:10:31 <blogic> i thought about the validation stuff
Jan 30 09:10:45 <blogic> i will add a service type to procd that does not run a daemon
Jan 30 09:11:05 <blogic> this allows us to hook certain uci options to shell scripts (fw3 and hostname, ....)
Jan 30 09:19:17 <builder_> build #334 of octeon is complete: Success [build successful]  Build details are at http://buildbot.openwrt.org:8010/builders/octeon/builds/334
Jan 30 10:39:09 <raorn> russell--: yep. i even managed to get usb serial, but don't have torx t8 screwdriver ;-)
Jan 30 10:39:34 <russell--> flatheads work too
Jan 30 10:39:47 <russell--> if you pick the right one
Jan 30 10:40:07 <russell--> (also, get a t-8, you won't be sorry)
Jan 30 10:58:00 <raorn> i have 50 pcs set ant t10 is the smallest torx...
Jan 30 11:13:07 <russell--> like i said, a small flathead will do in a pinch
Jan 30 11:13:42 <russell--> t-8's are pretty common though, so put it on your shopping list ;-)
Jan 30 11:14:19 <russell--> also a dremel cutting wheel will turn any screw into a flathead screw!
Jan 30 11:16:04 <karlp> anyone know if there's a way tomake uboot request files with nfsv3 instead of v2?
Jan 30 11:16:16 <karlp> turns out my fedora machine has dropped nfsv2 support,
Jan 30 11:16:40 <karlp> wireshark shows it mounting ok with v2, but then failing to get the files themselves with v2,
Jan 30 11:55:06 <plntyk> karlp: changing nfs bootargs to v3 does not work ?
Jan 30 11:56:37 <plntyk> no bootargs_nfs ?
Jan 30 11:58:38 <karlp> not bootinga kernel via nfs,
Jan 30 11:58:48 <karlp> this is the nfs command in uboot to flash an image from nfs
Jan 30 11:59:30 <karlp> I could try nfs kernel and nfs root file system, but haven't set tha tuðp before,
Jan 30 11:59:54 <karlp> does the owrt build root leave a root file system in place in the buildtree?
Jan 30 12:02:26 <plntyk> i think staging_dir/target-<arch...uclibc>/root-<arch>
Jan 30 12:04:42 <karlp> this needs a new kernel that can support this right?
Jan 30 12:06:00 <karlp> I'll have ot have damn tftp to get a new kernel, stupid fedora
Jan 30 12:07:22 <plntyk> sry dont know
Jan 30 13:24:29 <OpenWrt-Commits> nbd r39426 trunk/package/network/ services/hostapd/patches/340-reload_freq_change.patch services/hostapd/patches/440-dynamic_20_40_mhz.patch services/hostapd/patches/600-ubus_support.patch services/hostapd/patches/400-terminate_on_setup_failure.patch * hostapd: fix frequency selection for ap+sta
Jan 30 13:24:34 <OpenWrt-Commits> nbd r39427 trunk/package/kernel/ (56 files in 2 dirs) * mac80211: update to wireless-testing 2014-01-23
Jan 30 13:25:18 <plntyk> oh nice
Jan 30 15:34:54 <karlp> any ideas on how to fix, "Sysupgrade is not yet supported on unknown. Image check 'platform_check_image' failed."
Jan 30 15:35:01 <karlp> what needs to be fixed for this platform?
Jan 30 15:35:37 <karlp> http://wiki.dragino.com/index.php?title=Upgrade_Firmware_ms14 implies it should just be working?
Jan 30 15:58:57 <OpenWrt-Commits> nbd r39428 trunk/package/kernel/ (14 files in 2 dirs) * mac80211: refresh tarball with a newer backports version
Jan 30 16:00:10 <karlp> urh, looks like initial dragino2 patch didn't include _all_ of the upstream chances :|
Jan 30 16:02:28 <Devastator> karlp is this ralink?
Jan 30 16:05:05 <karlp> Devastator: no, ar9331,
Jan 30 16:05:58 <Devastator> hum, it probably is not add yet in image/upgrade/platform.c
Jan 30 16:06:01 <Devastator> something like that
Jan 30 16:06:17 <karlp> yeah, I found it missing in /lib/upgrade/platform.sh
Jan 30 16:06:35 <karlp> the intial patch for this platform got everything _else_ fromthe original authors, but missed including this,
Jan 30 16:06:46 <karlp> testing out a new build now |
Jan 30 16:20:14 <karlp> yup, that works.
Jan 30 17:10:59 <OpenWrt-Commits> juhosg r39429 trunk/target/linux/ ar71xx/base-files/lib/ar71xx.sh ar71xx/base-files/lib/upgrade/platform.sh * ar71xx: dragino2: add missing sysupgrade support.
Jan 30 17:11:01 <OpenWrt-Commits> juhosg r39430 trunk/package/kernel/linux/modules/virtual.mk * kernel/modules: add xen-privcmd package
Jan 30 19:27:58 <Pteridium> [florian]: can be used the bcm63268 ethernet driver with the bcm3380?
Jan 30 19:34:37 <[florian]> Pteridium: 63268's driver is the same as 6368, it's still a roboswitch
Jan 30 19:34:49 <Pteridium> ah
Jan 30 19:34:52 <Pteridium> thanks
Jan 30 19:36:08 <[florian]> Pteridium: 3380 is really special as it uses the "free pool manager" which is some sort of hardware dma allocator
Jan 30 19:36:27 <[florian]> the driver does not manipulate buffers directly but ids associated with these buffers
Jan 30 19:38:56 <Pteridium> i ask this because found this:
Jan 30 19:38:57 <Pteridium> http://pastebin.com/HqrPCM38
Jan 30 19:39:27 <Pteridium> seems that bcm63268 has both enet and gmac
Jan 30 19:40:02 <Pteridium> while bcm6818 has gmac only
Jan 30 19:52:50 <[florian]> Pteridium: yes there is a separate GMAC interface
Jan 30 19:53:00 <[florian]> Pteridium: but it is different from the 3380 one
Jan 30 19:53:31 <Pteridium> bad luck
Jan 30 19:53:47 <Pteridium> thanks for your time
Jan 30 19:53:49 <[florian]> and it is not bgmac either obviously ;)
Jan 30 19:54:06 <Pteridium> of course
Jan 30 19:57:09 <[florian]> and it looks like the gmac is actually feeding packets into the switch directly
Jan 30 19:57:19 <[florian]> not sure if there are any 63268 out there using the gmac
Jan 30 19:59:06 <Pteridium> so the 63268 based boards uses enet
Jan 30 20:48:15 <[florian]> Pteridium: the ones I have tested do, yes
Jan 30 20:49:22 <Pteridium> interesting, then why they added gmac?
Jan 30 20:49:41 <Pteridium> afaik 63268 is adsl/vdsl
Jan 30 20:49:48 <[florian]> not sure yes, maybe to offer more bandwidth and allow for a dedicated WAN port?
Jan 30 20:50:13 <Pteridium> ah, right
Jan 30 20:51:33 <Pteridium> so the 63168 is targeted for designs that doesn't need that features
Jan 30 20:51:44 <[florian]> yep
Jan 30 20:51:53 <[florian]> the 1 usually stands for low cost, low bill of material
Jan 30 20:51:58 <[florian]> and therefore low features too
Jan 30 20:52:08 <Pteridium> :D
Jan 30 20:54:27 <Pteridium> do you remember the problem with jffs2 and smp? i'm almost certain that it will not happen in boards with spi flash
Jan 30 20:55:33 <Pteridium> sadly in spain there aren't routers with bcm6361 or bcm6362
Jan 30 20:55:51 <Pteridium> Spain*
Jan 30 20:57:17 <[florian]> what flash do you have, parallel nor, right?
Jan 30 20:57:45 <[florian]> the boards I have here are mostly spi or nand
Jan 30 20:58:03 <[florian]> so it could very well be the paralle nor which is the key difference
Jan 30 20:59:32 <Pteridium> parallel nor, yes. http://wiki.openwrt.org/toh/comtrend/vr3025un
Jan 30 21:02:35 <[florian]> just to rule out some possible UART issues, this also happens from a telnet session right?
Jan 30 21:03:18 <Pteridium> yes, i've disabled uart driver but still happens the problem
Jan 30 21:03:24 <[florian]> ok, good
Jan 30 21:06:18 <Pteridium> later i'll send a mail to the mailing list with object debugging that shows something funny: http://pastebin.com/xKN987W4
Jan 30 21:06:56 <[florian]> humm
Jan 30 21:07:03 <[florian]> could be related for the jffs2 debug object warning
Jan 30 21:07:22 <Pteridium> sorry, but no idea
Jan 30 21:07:43 <[florian]> I wonder if the SPI flash is slow enough for not showing this problem
Jan 30 21:08:01 <[florian]> or if this is some sort of HW problem where the CPU is super speculative and P-NOR is fast enough to allow that
Jan 30 21:08:03 <Pteridium> with Ralink rt3052 it does not happen
Jan 30 21:08:36 <Pteridium> wait! i'm not as smart as you
Jan 30 21:10:46 <Pteridium> but with the lantiq danube with parallel nor flash happens something: http://pastebin.com/2FaMZXJm
Jan 30 21:11:38 <Pteridium> is there any way to slow the parallel flash?
Jan 30 21:12:27 <zinx> set the clock lower?
Jan 30 21:13:36 <Pteridium> more or less
Jan 30 21:15:31 <[florian]> I don't think so
Jan 30 21:16:27 <[florian]> you might be able to slightly change the timings for the chip-select 0
Jan 30 21:16:39 <[florian]> but they should already be the correct ones since the bootloader has configured those (hopefully)
Jan 30 21:17:16 <Pteridium> good point. cs.c, no?
Jan 30 21:18:01 <[florian]> yep
Jan 30 21:18:06 <[florian]> we do not touch it for cs0
Jan 30 21:18:18 <[florian]> we just read the base address configured to know the flash base address
Jan 30 21:19:41 <Pteridium> seems that parallel flash is attached to the mpi
Jan 30 21:24:53 <[florian]> it is yes
Jan 30 21:28:22 <Pteridium> thanks
Jan 30 22:13:51 <builder_> build #383 of iop32x is complete: Failure [failed shell_12]  Build details are at http://buildbot.openwrt.org:8010/builders/iop32x/builds/383
Jan 30 23:38:39 <builder_> build #65 of mvebu is complete: Failure [failed compile_5]  Build details are at http://buildbot.openwrt.org:8010/builders/mvebu/builds/65
Jan 31 01:22:20 * russell-- is seeing a failure to populate the hostname in recent alix2 builds
**** ENDING LOGGING AT Fri Jan 31 02:59:58 2014