**** BEGIN LOGGING AT Mon Sep 09 02:59:58 2013
Sep 09 03:23:12 <builder> build #249 of octeon is complete: Failure [failed compile_5]  Build details are at http://buildbot.openwrt.org:8010/builders/octeon/builds/249
Sep 09 03:29:55 <russell--> anybody know how to get a tp-link wdr3600 case off?
Sep 09 03:35:21 <russell--> ah, prying works
Sep 09 03:38:16 <Devastator> russell-- I don't know if I thanked you about your instructions on how to work with patches, but thank you :)
Sep 09 04:00:29 <russell--> Devastator: np
Sep 09 04:00:58 <russell--> branches are cheap (and useful) in git
Sep 09 04:01:10 <Devastator> russell-- have you ever worked with brcm47xx?
Sep 09 04:01:23 <russell--> yes
Sep 09 04:01:38 <russell--> but on pretty limited cross section of hardware
Sep 09 04:05:43 <Devastator> russell-- I was looking for something like this, but for brcm47xx, or brcm63xx is similar? http://wiki.openwrt.org/doc/devel/add.new.device
Sep 09 04:09:12 <russell--> but now you found it?
Sep 09 04:10:16 <Devastator> russell-- not really, unless I can use the same "step by step" of brcm63xx for brcm47xx
Sep 09 04:16:14 <russell--> probably pretty similar
Sep 09 04:17:25 <Devastator> thanks, I will see if I can ping hauke to get a confirmation
Sep 09 04:19:17 <russell--> or you could just try ... empiricism++
Sep 09 04:20:04 <Devastator> heheh
Sep 09 04:20:30 <Devastator> I could.. but I try to save time.. even if I'm losing already
Sep 09 04:23:18 <russell--> gah, so i put a serial console on the tp-link wdr3600, and it's asking me for a non-trivial passwd, wtf?
Sep 09 04:23:24 <russell--> (stock fw)
Sep 09 04:29:14 <Devastator> try blank :)
Sep 09 04:32:08 <russell--> nope, nor "admin" nor "password" nor the wireless password/pin printed on the bottom of the box
Sep 09 04:32:21 <russell--> nor "root"
Sep 09 04:34:13 <Devastator> admin/admin, admin/password, admin/1234
Sep 09 04:35:28 <russell--> i'm talking about the serial console, not the web admin interface
Sep 09 04:35:42 <russell--> and it complains about admin as a username, so i'm sticking with root
Sep 09 04:35:44 <quackquack1> admin 1234
Sep 09 04:35:52 <quackquack1> mostly in tp-link
Sep 09 04:36:13 <Devastator> he is using netgear
Sep 09 04:36:31 <russell--> TL-WDR3600 login: admin
Sep 09 04:36:32 <russell--> login: no valid shadow password
Sep 09 04:36:52 <russell--> this is a tp-link
Sep 09 04:37:08 <quackquack1> then admin 1234
Sep 09 04:37:11 <quackquack1> mostly
Sep 09 04:37:37 <russell--> admin at a loging generates "no valid shadow password"
Sep 09 04:39:40 <quackquack1> press t
Sep 09 04:39:44 <quackquack1> and then turn on
Sep 09 04:39:50 <quackquack1> and then admin 1234
Sep 09 04:47:25 <Devastator> russell-- try root and 5up
Sep 09 04:52:22 <russell--> quackquack1: what does 't' do?
Sep 09 04:52:40 <quackquack1> no idea
Sep 09 04:53:00 <quackquack1> http://wiki.openwrt.org/toh/tp-link/td-w8970
Sep 09 04:53:03 <quackquack1> thats why
Sep 09 04:53:06 <quackquack1> guess
Sep 09 04:53:49 <Devastator> according to what I've quickly searched, root/5up should work..
Sep 09 04:54:13 <russell--> quackquack1: this is substantially different hardware
Sep 09 04:54:32 <russell--> i can get to uboot with tpl during the one second before autobooting
Sep 09 04:56:20 <quackquack1> hmmmm
Sep 09 04:57:23 <Devastator> I need to get some sleep, good luck
Sep 09 05:38:41 <russell--> got an initramfs image working
Sep 09 05:38:55 <russell--> using tftpboot from uboot prompt
Sep 09 06:04:55 <russell--> grabbed copies of the flash, sysupgraded from initramfs
Sep 09 06:15:39 <quackquack1> which device?
Sep 09 08:02:59 <russell--> hmm. unsuccessful mounting or extracting files from the tp-link stock squashfs
Sep 09 08:03:40 <russell--> get an lzma error: "lzma uncompress failed with error code 0"
Sep 09 09:22:34 <builder> build #375 of orion is complete: Failure [failed compile_8]  Build details are at http://buildbot.openwrt.org:8010/builders/orion/builds/375
Sep 09 09:51:19 <stintel> nbd: not sure if you saw my paste about issues with Buffalo WZR-HP-G300NH and ath9k with AA r37838
Sep 09 09:51:29 <stintel> but today it seems the AP is in a state that no client can associate
Sep 09 09:52:29 <stintel> well, at least not my macbook air and also my Linux 3.10.10 workstation with ar9285
Sep 09 09:52:44 <nbd> i didn't see the paste
Sep 09 09:53:05 <stintel> ok. iw wlan0 station dump shows 1 station now, but with 4623470ms inactive time
Sep 09 09:53:32 <stintel> I'll paste dmesg soon
Sep 09 09:53:46 <nbd> hm, with that kind of inactive time it looks like there may be a hostapd issue
Sep 09 09:56:06 <stintel> http://bpaste.net/show/130765/
Sep 09 09:56:29 <stintel> the paste from last week looked similar with a lot failed to stop DMA
Sep 09 09:58:44 <stintel> nbd: is there any other output you can use ?
Sep 09 10:00:01 <nbd> go to /sys/kernel/debug/ieee80211/phy0/ath9k and show me xmit, recv, interrupt, misc, reset, queues
Sep 09 10:03:02 <stintel> nbd: http://pastebin.com/gMMKJ3Rr
Sep 09 10:03:49 <nbd> interesting
Sep 09 10:03:57 <nbd> not a hostapd issue then
Sep 09 10:04:03 <nbd> but something very messed up in the driver
Sep 09 10:04:40 <nbd> please try deleting the station using iw
Sep 09 10:04:44 <nbd> and then give me the queues output again
Sep 09 10:06:01 <stintel> http://pastebin.com/MXvmB4t4
Sep 09 10:06:48 <nbd> thx
Sep 09 10:06:55 <nbd> then it seems that there is either a leak or a counting error
Sep 09 10:07:02 <nbd> that definitely helps with the code review
Sep 09 10:07:02 <stintel> looks the same - iw wlan0 station del [mac] is what you
Sep 09 10:11:16 <stintel> ok, think I am going to reflash a previous AA build for now
Sep 09 10:17:29 <stintel> after running wifi everything is fine again, but I guess that expected :)
Sep 09 10:17:33 <stintel> that's*
Sep 09 10:19:03 <nbd> good to know
Sep 09 10:19:10 <nbd> then it can't be a counting error either
Sep 09 10:19:22 <nbd> then i have no idea what can trigger this
Sep 09 10:20:16 <nbd> is the queues output completely normal after you re-ran wifi?
Sep 09 11:21:21 <stintel> nbd: looks like it, all values 0
Sep 09 11:21:30 <nbd> ok
Sep 09 11:21:32 <stintel> except qnum ofcourse :)
Sep 09 11:21:32 <nbd> thx
Sep 09 11:21:50 <stintel> yw
Sep 09 12:39:20 <OpenWrt-Commits> nbd r37920 branches/attitude_adjustment/package/mac80211/patches/560-ath9k_tx_queueing_rework.patch
Sep 09 12:39:20 <OpenWrt-Commits> AA: ath9k: remove most of the tx queueing rework until it has been fully debugged
Sep 09 12:39:43 <nbd> stintel: this should get rid of the issues you were seeing
Sep 09 12:39:57 <nbd> stintel: btw. how easy was the issue to reproduce?
Sep 09 12:40:31 <nbd> i'd also like to see some test results for trunk (which has more fixes)
Sep 09 12:40:35 <nbd> if you have some time to run that
Sep 09 12:44:08 <stintel> nbd: it's the first time I saw this issue, but this image was only running since last wednesday
Sep 09 12:46:05 <stintel> I saw the "failed to stop ?X DMA" messages before though
Sep 09 12:48:53 <stintel> and would like to avoid running trunk on this device, it's an AP at work, other people use it too
Sep 09 12:50:16 <stintel> also, there is a lot of RF in the air here
Sep 09 12:50:47 <nbd> you could copy over the mac80211 patches from trunk to AA
Sep 09 12:50:55 <nbd> and leave the rest alone
Sep 09 12:51:55 <stintel> root@OpenWrt:~# iw wlan0 scan  |egrep -c '^BSS '
Sep 09 12:51:55 <stintel> 40
Sep 09 12:52:23 <stintel> just the patches dir? or the whole mac80211 dir ?
Sep 09 12:53:03 <nbd> just patches
Sep 09 12:53:07 <stintel> ok
Sep 09 12:53:40 <stintel> currently working on an IPTV project, so streaming quite some video on the AP, too
Sep 09 12:53:47 <stintel> s/on/over/
Sep 09 13:59:27 <builder> build #364 of ramips is complete: Failure [failed shell_7]  Build details are at http://buildbot.openwrt.org:8010/builders/ramips/builds/364
Sep 09 14:05:06 <builder> build #344 of rb532 is complete: Failure [failed shell_12]  Build details are at http://buildbot.openwrt.org:8010/builders/rb532/builds/344
Sep 09 14:15:04 <builder> build #344 of ppc44x is complete: Failure [failed shell_12]  Build details are at http://buildbot.openwrt.org:8010/builders/ppc44x/builds/344
Sep 09 14:15:57 <stintel> ok had a flood of "failed to stop ?X DMA" messages again, got disconnected, took about a minute before I could associate again
Sep 09 14:16:07 <stintel> I'll flash with mac80211/patches from trunk
Sep 09 14:36:00 <OpenWrt-Commits> jow r37921 trunk/package/ system/rpcd/files/rpcd.init system/rpcd/Makefile * rpcd: update to git head
Sep 09 15:02:43 <stintel> nbd: only up 7 minutes and saw this already (with mac80211 patches from trunk)
Sep 09 15:02:46 <stintel> Sep  9 16:59:27 OpenWrt kern.err kernel: [  293.130000] ath: phy0: DMA failed to stop in 10 ms AR_CR=0x00000024 AR_DIAG_SW=0x42000020 DMADBG_7=0x000286c0
Sep 09 15:02:49 <stintel> Sep  9 16:59:27 OpenWrt kern.err kernel: [  293.140000] ath: phy0: Could not stop RX, we could be confusing the DMA engine when we start RX up
Sep 09 15:03:01 <stintel> and SSH connection is choking every now and then
Sep 09 15:03:28 <nbd> is it significantly worse than the old AA build?
Sep 09 15:03:46 <nbd> if so, i'd like to know if that difference is reproducible
Sep 09 15:05:08 <stintel> I'll keep running this one for a while
Sep 09 15:05:38 <nbd> ok, thanks
Sep 09 15:07:20 <stintel> I guess this might be limited to some specific wireless chips. this Buffalo WZR-HP-G300NH is very problematic for me
Sep 09 15:07:51 <DonkeyHotei> nbd: thoughts? http://patchwork.openwrt.org/patch/3966/
Sep 09 15:07:58 <stintel> my rspro's at home with each an ar9285 and an ar9160 haven't produced that error in a long time afaict
Sep 09 16:05:21 <builder> build #325 of sibyte is complete: Failure [failed shell_12]  Build details are at http://buildbot.openwrt.org:8010/builders/sibyte/builds/325
Sep 09 16:26:31 <Devastator> luka hi, good morning, I've sent a refreshed version of my patch to the mailing list: http://patchwork.openwrt.org/patch/4037/
Sep 09 16:27:09 <Devastator> luka when you have the time, see if it's ok and if you can, push it :)
Sep 09 16:59:37 <Devastator> Hauke hi, can I follow the same described here for brcm47xx as well or this is brcm63xx specific? http://wiki.openwrt.org/doc/devel/add.new.device
Sep 09 17:01:06 <Hauke> Devastator: no it is different
Sep 09 17:01:57 <Hauke> OpenWrt will boot on your device without midifications, it could just be that you have to add some special header to the firmware image so the firmware flasher will accept it
Sep 09 17:02:48 <Hauke> Devastator: you could try network boot at first: http://wiki.openwrt.org/doc/hardware/soc/soc.broadcom.bcm47xx#network.boot
Sep 09 17:04:05 <Hauke> when it boots you should make it detect your router by adding it to target/linux/brcm47xx/patches-3.10/260-MIPS-BCM47XX-add-board-detection.patch and then adding a gpio configuration for this board in package/kernel/broadcom-diag/src/diag.c
Sep 09 17:04:15 <Devastator> Hauke device is with just cfe, no firmware loaded
Sep 09 17:04:39 <Devastator> thanks
Sep 09 17:05:02 <Hauke> Devastator: the try to flash bin/brcm47xx/openwrt-brcm47xx-squashfs.trx or use network boot
Sep 09 17:05:35 <Devastator> I will build a tg3 ethernet + proprietary wl initramfs image and see how it goes, what do you think?
Sep 09 17:06:42 <Hauke> that sounds good
Sep 09 17:08:39 <Devastator> and if you want, with your help, I can try to format a wiki entry similar to brcm63xx, to help people porting their routers, if they want to
Sep 09 17:09:38 <builder> build #376 of orion is complete: Failure [failed compile_8]  Build details are at http://buildbot.openwrt.org:8010/builders/orion/builds/376
Sep 09 17:09:49 <Hauke> Devastator: yes that would be nice
Sep 09 17:11:38 <builder> build #357 of uml is complete: Failure [failed compile_5]  Build details are at http://buildbot.openwrt.org:8010/builders/uml/builds/357
Sep 09 17:12:58 <Devastator> whatever I can do to help :)
Sep 09 17:13:35 <Devastator> I also have a bricked E3000, but it's hard to enable jtag
Sep 09 17:14:23 <quackquack> Devastator, what kind of flash is it?
Sep 09 17:14:43 <Devastator> serial flash, iirc
Sep 09 17:15:08 <quackquack> hmmmm
Sep 09 17:15:41 <Devastator> but it was hit by a lightening hehe
Sep 09 17:15:52 <quackquack> its tsop-48
Sep 09 17:16:07 <quackquack> i dont think you can use a buspirate
Sep 09 17:16:22 <quackquack> can you?
Sep 09 17:16:36 <quackquack> i guess you can....
Sep 09 17:16:46 <Devastator> is buspirate similar to jtag?
Sep 09 17:16:49 <quackquack> http://images3.wikia.nocookie.net/__cb20100803141012/infodepot/images/thumb/a/ac/Linksys_E3000_v1.0_FCCd.jpg/639px-Linksys_E3000_v1.0_FCCd.jpg
Sep 09 17:16:58 <quackquack> kind of.
Sep 09 17:17:21 <Devastator> jtag is not enabled by default
Sep 09 17:17:25 <quackquack> right next to flash?
Sep 09 17:17:38 <quackquack> there is soic-8 soic-16 footprint.
Sep 09 17:18:13 <quackquack> you can rip out the tsop flash and solder a soic-16 flash chip.
Sep 09 17:18:28 <quackquack> use the buspirate to reprogram the soic-16 flash chip
Sep 09 17:19:29 <Devastator> it's hard to find flashchips over here, even soic-16
Sep 09 17:19:38 <quackquack> soic?
Sep 09 17:19:43 <quackquack> dont tell me that!
Sep 09 17:19:49 <Devastator> yup
Sep 09 17:19:50 <quackquack> you are in antartica?
Sep 09 17:20:01 <Devastator> almost.. Brazil
Sep 09 17:20:05 <quackquack> lol
Sep 09 17:20:39 <quackquack> order it from ebay
Sep 09 17:21:16 <Devastator> doesn't worth it, switch chip has a hole in it, it's dead
Sep 09 17:21:30 <Devastator> but I can get a crazy serial output
Sep 09 17:21:36 <quackquack> oh
Sep 09 17:22:32 <Devastator> I thought about using it as a repeater, but I need jtag to erase and reflash
Sep 09 17:23:02 <quackquack> if you replace the tsop with soic flash then a buspirate is crazy enough.
Sep 09 17:25:41 <Devastator> I've bought a flashchip once with Farnell, it costed me about 20$
Sep 09 17:26:12 <quackquack> oh
Sep 09 17:26:57 <quackquack> i purchased from element14 which is farnell for india some 15 chips for 60 US$
Sep 09 17:27:22 <Devastator> whoa
Sep 09 17:27:39 <quackquack> soic 16 chips all of them of 4 MiB, 8 MiB and 16 MiB each 5 no
Sep 09 17:27:56 <MartijnVdS> can't you "rescue" one from some other device?
Sep 09 17:28:18 <quackquack> footprint must match isnt it MartijnVdS
Sep 09 17:28:56 <MartijnVdS> quackquack: yes, but flash chips are very common, and their form factors as well
Sep 09 17:29:01 <quackquack> wrt54g/l footprint and this chip looks identical MartijnVdS
Sep 09 17:29:23 <quackquack> there are 40, 48 and 56 tsop
Sep 09 17:29:29 <quackquack> so only i asked.
Sep 09 17:29:34 <quackquack> IIRC.
Sep 09 17:30:21 <quackquack> MartijnVdS, is right.
Sep 09 17:30:32 <quackquack> you can it on another device.
Sep 09 17:30:48 <quackquack> but MartijnVdS since he has an option for soic-8/16 thats why.
Sep 09 17:30:57 <MartijnVdS> ah yes.. that's harder to find
Sep 09 17:31:02 <quackquack> work is less with soic-8/16
Sep 09 17:31:24 <Devastator> soldering flashchips is not easy as well
Sep 09 17:31:48 <quackquack> true if your hands shake! then its a no no.
Sep 09 17:32:30 <Devastator> the paddings can lift up, then.. oh boy..
Sep 09 17:32:39 <OpenWrt-Commits> nbd r37922 trunk/package/kernel/mac80211/patches/ (7 files) * rt2x00: merge an rt2800 initialization order fix
Sep 09 17:33:33 <quackquack> "<Devastator> the paddings can lift up, then.. oh boy.." -> netgear dgn1000, serial pads are ripped....
Sep 09 17:33:45 <quackquack> so cant fix it using serial console.
Sep 09 17:41:19 <Devastator> quackquack yep, crappy build quality
Sep 09 17:42:35 * quackquack nods
Sep 09 17:42:44 <quackquack> bbiaw
Sep 09 18:29:10 <quackquack> can someone tell me how to port openwrt to wrt54g v7 the atheros chipset.
Sep 09 18:29:27 <quackquack> i did try to put the redboot but it didnt work.
Sep 09 18:30:12 <quackquack> i upgraded the ram 32 MiB and also flash soic-16 8 MiB
Sep 09 18:30:17 <quackquack> how to proceed?
Sep 09 18:33:05 <DonkeyHotei> i have a GLv1.1 and a GSv1.0 dropped on my doorstep i still don't know what to do with
Sep 09 18:33:31 <DonkeyHotei> consensus is their cpu is too slow to be useful
Sep 09 18:33:57 <quackquack> hmmm
Sep 09 18:34:03 <quackquack> i got them for free.
Sep 09 18:34:14 <quackquack> so beggers are not choosers.
Sep 09 18:34:38 <DonkeyHotei> but i don't know of a use case
Sep 09 18:35:10 <quackquack> hmmmm
Sep 09 18:35:52 <quackquack> -ChanServ- Channel #redboot is not registered.
Sep 09 18:35:53 <quackquack> -ChanServ- Channel ##redboot is not registered.
Sep 09 18:36:02 <quackquack> great hence no redboot support.
Sep 09 18:36:39 <DonkeyHotei> it's #eCos iirc
Sep 09 18:37:16 <quackquack> oh
Sep 09 18:37:21 <quackquack> #ecos
Sep 09 19:05:00 <builder> build #335 of avr32 is complete: Failure [failed compile_5]  Build details are at http://buildbot.openwrt.org:8010/builders/avr32/builds/335
Sep 09 20:02:49 <builder> build #365 of ramips is complete: Success [build successful]  Build details are at http://buildbot.openwrt.org:8010/builders/ramips/builds/365
Sep 09 21:00:11 <builder> build #284 of ep93xx is complete: Failure [failed compile_5]  Build details are at http://buildbot.openwrt.org:8010/builders/ep93xx/builds/284
Sep 09 21:59:25 <falstaff_> Hi, I try to boot OpenWRT on WNDR4300 with UBIFS. So far i've a running kernel which can read from nand, i flashed kernel and fs to the device
Sep 09 21:59:34 <falstaff_> now when I try to boot the device I get this error:
Sep 09 21:59:41 <falstaff_> [    0.250000] UBI error: ubi_init: UBI error: cannot initialize UBI, error -19
Sep 09 22:00:11 <falstaff_> My kernel command line: console=ttyS0,115200 ubi.mtd=8 rootdelay=5 root=ubi0:rootfs rootfstype=ubifs mtdparts=ar934x-nfc:256....
Sep 09 22:01:13 <falstaff_> I can see in the kernel output that mtd partitions don't get parsed.. When I run the same kernel using initramfs, I can see the MTD partitions... But later than ubi is initialized in this kernel
Sep 09 22:01:28 <falstaff_> => It looks as if the kernel initialize ubi too early
Sep 09 22:04:01 <falstaff_> -19 is ENODEV, so this supports my theory
Sep 09 22:08:07 <froek> where would i go about changing the boot sequence when the filesystem is looking to mount EXTROOT on a uuid-based device?
**** ENDING LOGGING AT Tue Sep 10 03:00:00 2013