**** BEGIN LOGGING AT Sun Nov 07 02:59:58 2010
Nov 07 03:30:18 <builder> build #18 of xburst is complete: Success [build successful]  Build details are at http://tksite.gotdns.org:8010/builders/xburst/builds/18
Nov 07 04:06:55 <builder> build #21 of ps3 is complete: Failure [failed compile_3]  Build details are at http://tksite.gotdns.org:8010/builders/ps3/builds/21
Nov 07 04:45:00 <russell_> hmm.  brcm47xx is failing for me at the moment
Nov 07 04:45:33 <russell_> /bin/tar: gcc-4.3.3/libstdc++-v3/testsuite/20_util/hash/operators: Cannot utime: No such file or directory
Nov 07 04:45:35 <russell_> wtf
Nov 07 04:47:28 <russell_> . /aux/src/openwrt/include/shell.sh; bzcat /src_archive/openwrt/dl/gcc-4.3.3.tar.bz2 | /bin/tar -C /aux/src/openwrt/build_dir/toolchain-mipsel_gcc-4.3.3+cs_uClibc-0.9.31/gcc-4.3.3/.. -xf -
Nov 07 05:25:34 <russell_> tar --version ==> tar (GNU tar) 1.24
Nov 07 05:33:03 <russell_> only seems to happen when doing bzcat foo.tar.bz2 | tar
Nov 07 05:33:10 <russell_> tar -xjf works fine
Nov 07 05:34:47 <russell_> reverting to 1.23.x worked
Nov 07 08:52:27 <builder> build #23 of pxcab is complete: Failure [failed compile_3]  Build details are at http://tksite.gotdns.org:8010/builders/pxcab/builds/23
Nov 07 09:51:31 <slonopotamus> mb__: does this thing actually charge? https://dev.openwrt.org/browser/trunk/target/linux/omap24xx/patches-2.6.36/900-n810-battery-management.patch
Nov 07 10:56:10 <loswillios> [florian]: ping
Nov 07 11:05:19 <loswillios> Wipster: you worked on ar7 networking too, right? I'm having trouble with an ADM6996LC on trunk (Speedport W701V)
Nov 07 11:06:17 <Wipster> loswillios, I'm trying too more learning atm, whats up
Nov 07 11:07:21 <loswillios> swconfig dev 0 help says failed to connect, I guess I don't have the switch driver installed. Looking for it atm
Nov 07 11:08:41 <loswillios> Wipster: do I need something like kmod-adm6996 or how is the switch supposed to be setup?
Nov 07 11:08:54 <loswillios> because I can't find anyhting
Nov 07 11:09:48 <Wipster> hmm the adm6996 phy driver is enabled in the kernel config
Nov 07 11:11:46 <loswillios> ah yes. does it expose something under eg /sys?
Nov 07 11:13:24 <loswillios> lots of /sys/devices/1:1f: driver -> ../../bus/mdio_bus/drivers/Infineon ADM6996
Nov 07 11:15:04 <loswillios> http://pastebin.ca/1984293
Nov 07 11:15:26 <loswillios> it shows RX/TX although it didn't even get an IP Address via DHCP
Nov 07 11:17:58 <Wipster> hmm ok lets see
Nov 07 11:18:32 <Wipster> tried setting lan to eth0.1?
Nov 07 11:20:57 <loswillios> yes
Nov 07 11:21:11 <loswillios> I'm not sure if the switch has vlan-support though
Nov 07 11:21:40 <loswillios> or the driver rather
Nov 07 11:23:38 <loswillios> I tried with 'config switch eth0' section and without
Nov 07 11:25:33 <Wipster> from the driver it looks like it does, what have you got set in config/network?
Nov 07 11:26:35 <[florian]> loswillios: pong
Nov 07 11:27:00 <[florian]> loswillios: there is switch driver for adm6996 which is swconfig aware
Nov 07 11:27:15 <loswillios> I tried those http://pastebin.ca/1984302
Nov 07 11:27:20 <[florian]> loswillios: the "switch" driver is just a simple PHY driver, which properly detects the phy id, that's all
Nov 07 11:27:31 <loswillios> ah, which package do I need?
Nov 07 11:27:44 <[florian]> there is not one for adm6996 yet
Nov 07 11:27:59 <[florian]> I mean, there is no proper switch driver for such switches
Nov 07 11:28:08 <[florian]> at least not on a platform different than brcm47xx
Nov 07 11:28:52 <loswillios> hm, but the switch worked with r12nnn
Nov 07 11:29:19 <[florian]> I do not think you were able to control the switch
Nov 07 11:29:23 <[florian]> back then
Nov 07 11:29:45 <loswillios> no, I only need to give it an ip-address. No fancy vlan stuff
Nov 07 11:29:55 <[florian]> well then, just configure br-lan and there you go
Nov 07 11:30:08 <[florian]> that works for me on the current trunk with a adm6996 device
Nov 07 11:30:33 <loswillios> config interface lan; option ifname   eth0; option proto    dhcp
Nov 07 11:30:50 <[florian]> and that does not work either?
Nov 07 11:30:53 <loswillios> nope
Nov 07 11:31:01 <[florian]> can you please show me a bootlog of your device?
Nov 07 11:31:06 <loswillios> http://pastebin.ca/1984293
Nov 07 11:32:20 <loswillios> there was some weird autodetection stuff in /e/c/network after I flashed trunk: http://pastebin.ca/1984307
Nov 07 11:32:36 <loswillios> (written from memory)
Nov 07 11:33:16 <[florian]> well I can see that you receive and emit packets on your interface
Nov 07 11:33:39 <loswillios> exactly, but no dhcp,ping etc
Nov 07 11:34:04 <[florian]> what if you configure the interface manullay and get a tcpdump running on the other side, do you see something coming in?
Nov 07 11:34:18 * loswillios tries
Nov 07 11:35:29 <[florian]> I am wondering if the option nat here is not causing some issues
Nov 07 11:37:19 <loswillios> I have it commented out
Nov 07 11:37:48 <loswillios> AFAICS there's nothing arriving from the router in tcpdump
Nov 07 11:38:35 <[florian]> ok, if you ping from the the speedport say, your router IP, do you see the tx packet counter increasing?
Nov 07 11:39:01 <[florian]> the link is seen "up" from the cpmac ethernet driver, so the packets should be able to be sent out
Nov 07 11:40:06 <loswillios> yes
Nov 07 11:45:41 <loswillios> I'm trying backfire now
Nov 07 11:47:35 <[florian]> it should be exactly the same
Nov 07 11:48:32 <Wipster> probably wont help but tried on a different ethernet port?
Nov 07 11:52:54 <loswillios> backfire-git won't boot: http://pastebin.ca/1984323 stopping at vlynq0: regs 0x08611800, irq 29, mem 0x04000000
Nov 07 11:53:49 <[florian]> have I backported the vlynq patch correctly to backfire :/
Nov 07 11:54:12 <loswillios> I don't know ;-) but I can help testing now
Nov 07 11:54:53 <[florian]> let me check
Nov 07 11:55:11 <loswillios> next line should be 'Found a VLYNQ device: 00000029'
Nov 07 11:55:35 <[florian]> the patch is there
Nov 07 12:09:36 <loswillios> [florian]: what else can I do regarding the network issue?
Nov 07 13:04:45 <CIA-76> jow * r23910 /branches/backfire/package/ (base-files/files/bin/login.sh busybox/files/telnet): [backfire] merge r22279 and r22282 (#6011)
Nov 07 13:16:25 <[florian]> loswillios: install tcpdump on the speedport and get me a capture of what you receive/emit
Nov 07 13:23:30 <KanjiMonster> blogic: gpioctl produces null pointer oopses when accessing invalid gpio numbers ;)
Nov 07 13:24:09 <jow_laptop> KanjiMonster: make a patch :)
Nov 07 13:25:08 <[florian]> KanjiMonster: gpioctl is that not useful when the device uses gpiolib
Nov 07 13:25:25 <KanjiMonster> jow_laptop: I'll do, probably after adding some more stuff to it. Looks like the bcm63xx devices have a range of "serial" gpios, for which I currently don't see a way of accessing
Nov 07 13:25:41 <KanjiMonster> [florian]: I am currently trying to find out which gpio belongs to which led
Nov 07 13:26:26 <KanjiMonster> and I'm still missing the wifi led after going through all 40 gpios
Nov 07 13:26:42 <[florian]> KanjiMonster: maybe it's directly controlled by the wifi card and therefore driver
Nov 07 13:27:40 <KanjiMonster> [florian]: it's a minipci wifi card, I wouldn't think that that works with those
Nov 07 13:27:54 <[florian]> oh it can, there is a pci signal for that if I recall right
Nov 07 13:28:06 <KanjiMonster> ah, I see
Nov 07 13:28:15 <KanjiMonster> then its a possibility
Nov 07 13:28:35 <[florian]> did you check the led polarity?
Nov 07 13:29:22 <KanjiMonster> my way of checking was read current value => inverse it
Nov 07 13:29:52 <[florian]> well that's not going to work in all cases, because by default gpiolib sets the gpio in output with a level at 0
Nov 07 13:30:16 <[florian]> erm not you will see if turning off
Nov 07 13:31:42 <KanjiMonster> btw, no wonder the dsl-2740b images didn't work, openwrt is missing the board definition for it ;)
Nov 07 13:32:24 <KanjiMonster> is there a naming convention for the dsl/internet leds?
Nov 07 13:32:42 <[florian]> I need to convert them to the linux leds class naming
Nov 07 13:33:06 <[florian]> "devicename:colour:function"
Nov 07 13:33:37 <[florian]> but usually that makes the userland scripts using them less generic
Nov 07 13:43:58 <KanjiMonster> [florian]: I meant in the board_info. The led(s) labeled with "internet" on the case are called "ppp" by broadcom, and I see "ppp"/"ppp-fail" as well as "internet:green"/"internet:red" in board_bcm963xx.c
Nov 07 13:44:11 <[florian]> oh ok
Nov 07 13:44:23 <[florian]> well, just follow the casing naming then
Nov 07 13:44:33 <[florian]> the broadcom led naming works when these are eval boards
Nov 07 13:49:33 <KanjiMonster> also, how do I get an active high led to not light up at boot time? (active_low isn't set, and default_state is also not set, so it should default to off)
Nov 07 13:50:18 <CIA-76> hauke * r23911 /trunk/target/linux/brcm47xx/ (18 files in 3 dirs):
Nov 07 13:50:18 <CIA-76> brcm47xx: Fix boot problem with wgt634u.
Nov 07 13:50:18 <CIA-76> Readd the workarround from the old version again which was removed in r22296 and refresh the patches.
Nov 07 13:50:18 <CIA-76> This should close #7874
Nov 07 13:50:18 <CIA-76> Thank you Russell Senior for testing.
Nov 07 13:53:56 <_Ernesto_> hello,
Nov 07 13:54:00 <_Ernesto_> can someone check this:
Nov 07 13:54:00 <_Ernesto_> https://dev.openwrt.org/ticket/8189
Nov 07 13:54:05 <_Ernesto_> thx
Nov 07 13:57:20 <CIA-76> nbd * r23912 /trunk/package/mac80211/patches/572-ath9k_xmit_queue_cleanup.patch: ath9k: fix a compile error with debugging disabled (#8189)
Nov 07 13:58:36 <CIA-76> nbd * r23913 /branches/backfire/package/mac80211/patches/572-ath9k_xmit_queue_cleanup.patch: ath9k: merge compile error fix from r23912
Nov 07 13:58:38 <_Ernesto_> i got the trunk before 10 minutes fresh
Nov 07 13:58:43 <_Ernesto_> thx
Nov 07 14:09:42 <loswillios> [florian]: ok
Nov 07 14:15:30 <loswillios> [florian]: http://pastebin.ca/1984448
Nov 07 14:18:53 <loswillios> and this is when pinging the router from another device: http://pastebin.ca/1984450
Nov 07 14:19:15 <loswillios> 1.2 is the speedport, 1.33 me
Nov 07 14:24:47 <das-Iro> there seems to be sth. wrong with http://tksite.gotdns.org:8010/builders/ps3/builds/21/steps/compile_3/logs/stdio
Nov 07 14:28:39 <das-Iro> some other platforms seem to have this too, powerpc-openwrt-linux-uclibc-gcc -> internal compiler error: Segmentation fault
Nov 07 14:49:45 <KanjiMonster> [florian]: Since I am renaming the image name for the dsl-2740b anyway to add a -C2, should I rename it to DSL-274XB also since it works (or should work) for both the DSL-2740B and 2741B?
Nov 07 14:52:06 <blogic> KanjiMonster: lol
Nov 07 14:54:27 <KanjiMonster> blogic: sorry, wasn't you code, its somewhere in the kernel, the gpiolib says you should make sure that the gpios exist, but it seems nothing actually does that ;)
Nov 07 15:09:17 <blogic> hehe
Nov 07 15:37:26 <KanjiMonster> [florian]: ping
Nov 07 15:45:01 <[florian]> KanjiMonster: pong
Nov 07 15:45:24 <[florian]> yes, please use DSL-274XB
Nov 07 15:46:31 <KanjiMonster> [florian]: I just noticed, r20652 probably broke half of the images for the AG226G and the AGV2+W, now the only difference between the normal and the -cfe images is the filename
Nov 07 15:47:23 <CIA-76> florian * r23914 /trunk/target/linux/uml/patches-2.6.32/007-uml_missing_stat_h.patch: [uml] fix build failure when targetting ramdisk
Nov 07 16:05:20 <CIA-76> hauke * r23915 /branches/backfire/target/linux/ (30 files in 17 dirs): kernel: update from kernel 2.6.32.16 to kernel 2.6.32.25 and refresh patches
Nov 07 16:13:22 <joda_bot1> I'd really lighttpd to be upgraded to 1.4.27 in the final release of backfire 10.03.1 - https://dev.openwrt.org/ticket/7755
Nov 07 16:13:40 <joda_bot1> Is there a change to promote this bug/patch update?
Nov 07 16:14:48 <xMff> did you test this patch?
Nov 07 16:14:51 <joda_bot1> That ticket would fix the following issue for me: http://redmine.lighttpd.net/issues/2125
Nov 07 16:15:15 <joda_bot1> xMff: no I did not, that might be an idea, but it would require me to setup a complete build environment
Nov 07 16:15:21 <KanjiMonster> Hauke: you just fixed https://dev.openwrt.org/ticket/7365 ;)
Nov 07 16:15:22 <joda_bot1> which I tried to avoid so far
Nov 07 16:15:35 <xMff> joda_bot1: the current version contains a fix for "issue 2125" already
Nov 07 16:16:10 <joda_bot1> oh cool
Nov 07 16:16:21 <joda_bot1> but the version is not upgraded to 1.4.27?
Nov 07 16:16:28 <xMff> https://dev.openwrt.org/browser/packages/net/lighttpd/patches/001-issue_2157.patch
Nov 07 16:16:30 <joda_bot1> beacuse I checked the version numbers of the packages
Nov 07 16:16:37 <joda_bot1> in the release candidates
Nov 07 16:16:59 <joda_bot1> not my problem I think?
Nov 07 16:17:00 <xMff> 1.4.26-1 -> 1.4.26-2
Nov 07 16:17:30 <xMff> oh right
Nov 07 16:17:33 <joda_bot1> My problem is that lighttpd selects the wrong certificate when using domain wildcards and SNI (multiple certificates on a single IP)
Nov 07 16:20:27 <Hauke> KanjiMonster: I have not noticed taht I just fixed a bug ;-) Thank you for showing it to me.
Nov 07 16:22:22 <KanjiMonster> Hauke: you're welcome :)
Nov 07 20:25:35 <[florian]> loswillios: ping?
Nov 07 20:25:46 <loswillios> pong
Nov 07 20:25:51 <[florian]> loswillios: when you said vlynq was freezing on your device with backfire, was it with a pre-built image?
Nov 07 20:26:31 <loswillios> no, backfire-git from today
Nov 07 20:26:54 <[florian]> but you do not see the lock on trunk right?
Nov 07 20:27:03 <loswillios> right
Nov 07 20:28:00 <[florian]> I am extremely surprised then, because both have the same patches
Nov 07 20:32:59 <loswillios> unfortunately I can't try anything for the next three weeks
Nov 07 20:33:25 <[florian]> no problems
Nov 07 20:38:22 <[florian]> Wipster: ping?
Nov 07 20:38:55 <philipp64|laptop> anyone have a chance to look at my patches?  they're pretty simple...
Nov 07 20:43:14 <CIA-76> florian * r23916 /trunk/target/linux/x86/net5501/target.mk:
Nov 07 20:43:14 <CIA-76> [x86] net5501: Add features to net5501 and remove HID support (unneeded)
Nov 07 20:43:14 <CIA-76> Since the net5501 doesn't have a VGA port, it doesn't need HID support either --
Nov 07 20:43:14 <CIA-76> all console access will be over a serial port.
Nov 07 20:43:14 <CIA-76> Also add FEATURES to it as well.
Nov 07 20:43:14 <CIA-76> Signed-of-by: Philip Prindeville <philip@redfish-solutions.com>
Nov 07 20:44:30 <Wipster> [florian], pong
Nov 07 20:45:20 <[florian]> Wipster: do you have any remaining work to do on the cpmac switch patches?
Nov 07 20:45:37 <[florian]> Wipster: and I if provide you with a ptach which adds support for 2.6.36 can you test it for me?
Nov 07 20:47:28 <Wipster> [florian], I am currently trying to track the transceiver deadlock. Yeh sure that would be really good
Nov 07 20:47:55 <[florian]> Wipster: are you able to reproduce it?
Nov 07 20:48:17 <[florian]> Wipster: I mean reliably deadlock it?
Nov 07 20:52:32 <Wipster> [florian], yeh without me doing anything it seems to lock up at the same point at boot but seeminly unrelated things are happening
Nov 07 20:54:16 <swalker> updated openwrt/upstream, https://home.comcast.net/~sdwalker/uscan/uscan.shtml
Nov 07 20:56:39 <[florian]> Wipster: ok, then let's put the priority on this issue since I cannot reproduce this
Nov 07 21:26:03 <Wipster> [florian], ok it seems to be ok I can telnet in
Nov 07 22:32:59 <CIA-76> florian * r23917 /trunk/target/linux/ar7/files/drivers/char/ar7_gpio.c: [ar7] avoid namespace clash on ar7_gpio_init
Nov 07 22:47:01 <CIA-76> florian * r23918 /trunk/target/linux/rdc/patches-2.6.32/ (6 files): [rdc] refresh patches, remove 015-r6040_fix_multicast.patch (upstreamed)
Nov 07 22:47:45 <philipp64|laptop> [florian]: thanks.
Nov 07 23:07:38 <Wipster> how does the uci-defaults work in the basefiles/etc like in atheros, I tried just adding to ar7 but it didn't look like it ran
Nov 07 23:11:31 <Wipster> *in the
Nov 07 23:28:59 <Wipster> doh
Nov 08 01:01:51 <philipp64|laptop> any chance of getting the patch "Make leds-5501 work" committed?
Nov 08 02:35:18 <philipp64|laptop> {Nico}: ping
Nov 08 02:35:32 <philipp64|laptop> Bartman007: ping
**** ENDING LOGGING AT Mon Nov 08 02:59:57 2010