**** BEGIN LOGGING AT Tue Mar 09 02:59:58 2010 Mar 09 03:50:40 build #26 of ep93xx is complete: Failure [failed compile_7] Build details are at http://tksite.gotdns.org:8010/builders/ep93xx/builds/26 Mar 09 09:32:16 build #23 of adm5120 is complete: Success [build successful] Build details are at http://tksite.gotdns.org:8010/builders/adm5120/builds/23 Mar 09 11:18:44 nico * r20080 /packages/net/arpwatch/Makefile: [packages] arpwatch: change PKG_BUILD_DIR to prevent configure for finding libpcap build dir Mar 09 11:51:57 tripolar * r20081 /packages/net/openssh/Makefile: Mar 09 11:51:58 update opennsh to 5.4p1 and move ssk-kegen out of openssh-client-utils and openssh-server Mar 09 11:51:58 and create a seperate package for it - to get rid of problems when both packages Mar 09 11:51:58 are installed at the same time Mar 09 12:03:50 lars * r20082 /trunk/target/linux/xburst/files-2.6.32/arch/mips/jz4740/irq.c: [xburst] jz4740: Properly setup intc irq handler. Mar 09 13:14:37 I read some mails from 2008 about patchwork being set up for openwrt. did that ever happen? Mar 09 13:45:37 build #23 of ixp4xx is complete: Success [build successful] Build details are at http://tksite.gotdns.org:8010/builders/ixp4xx/builds/23 Mar 09 14:25:43 Stintel: hehe, the silence says it all :P Mar 09 14:32:59 MarkV: ;-) Mar 09 14:35:43 nbd * r20083 /trunk/target/linux/generic-2.6/files/drivers/net/phy/ (ar8216.c ar8216.h): (log message trimmed) Mar 09 14:35:43 Several small fixes for ar8216 driver (patch by Jonas Gorski) Mar 09 14:35:43 * Create defines for some magic values/masks. Mar 09 14:35:43 * Change vlan_id to u16, to allow VIDs > 255. Mar 09 14:35:43 * Add a range check to set_pvid as it isn't a VID, but the index Mar 09 14:35:44 in the vlan table. Mar 09 14:35:44 * Set the max VID to 4094, since 4095 is a reserved value and Mar 09 14:36:59 nbd: ping Mar 09 14:38:03 pong Mar 09 14:38:47 nbd: there is another way to avoid recalculating the checksum Mar 09 14:39:13 how? Mar 09 14:39:18 nbd: one could include the squashfs image into the trx Mar 09 14:39:35 and then change the firstboot detection system Mar 09 14:39:58 change it how? Mar 09 14:40:03 instead of looking for 0xdeadcode, it could verify the jffs2 start magic is not there Mar 09 14:40:25 sorry, there are people here, can't write too fast Mar 09 14:40:47 well... does the web interface erase the flash at that point during reflash? Mar 09 14:40:56 or does it simply leave garbage there? Mar 09 14:41:04 garbage Mar 09 14:41:14 then it won't be reliable Mar 09 14:41:15 and we can't use it Mar 09 14:41:28 but it's rather unlikely this garbage is a jffs start header Mar 09 14:41:39 there are conditions where it'd be quite likely Mar 09 14:41:41 for instance: Mar 09 14:41:50 you have an openwrt image that is larger than the linksys firmware image Mar 09 14:41:59 you had openwrt on there before, switch to linksys fw, then switch back to openwrt Mar 09 14:42:03 => fail Mar 09 14:47:04 nbd: maybe I can trick it into erasing parts of the flash Mar 09 14:47:04 nbd: let me think about it Mar 09 14:47:39 how about just writing a small c program that fixes up the trx header before the jffs2 gets mounted read/write? Mar 09 14:49:38 nbd: this would also be a possbility Mar 09 14:49:45 let me think a little bit about this Mar 09 14:49:55 jow * r20084 /trunk/target/imagebuilder/Config.in: [imagebuilder] remove dependency on CONFIG_ALL Mar 09 14:51:15 build #23 of kirkwood is complete: Failure [failed compile_6 shell_19] Build details are at http://tksite.gotdns.org:8010/builders/kirkwood/builds/23 Mar 09 15:51:43 nbd * r20085 /trunk/scripts/metadata.pl: scripts/metadata.pl: fix handling of multiple conditional depends on the same package Mar 09 16:40:09 <[florian]> :win 11 Mar 09 16:40:13 <[florian]> grmnbl Mar 09 16:59:10 florian * r20086 /trunk/target/linux/ar7/files-2.6.30/include/asm-mips/ar7/ar7.h: [ar7] fix missing bits in ar7.h after r20037 Mar 09 16:59:16 florian * r20087 /trunk/target/linux/ar7/Makefile: [ar7] switch to 2.6.32 Mar 09 17:02:36 ping nbd Mar 09 17:04:06 ping {Nico} Mar 09 17:36:01 nbd * r20088 /trunk/package/hostapd/files/wpa_supplicant.sh: wpa_supplicant: fix bridging settings for wds Mar 09 17:36:06 nbd * r20089 /trunk/package/mac80211/files/lib/wifi/mac80211.sh: mac80211: add support for the noscan option for skipping the initial scan on hostapd startup Mar 09 17:36:13 nbd * r20090 /trunk/package/hostapd/ (12 files in 2 dirs): hostapd: upgrade to 20100309 Mar 09 18:02:39 [florian]: ping Mar 09 18:05:33 juhosg * r20091 /trunk/target/linux/ar71xx/base-files/etc/defconfig/ (5 files in 5 dirs): ar71xx: add missing protocol definitions to default network configurations Mar 09 18:05:37 juhosg * r20092 /trunk/target/linux/ar71xx/files/drivers/net/ag71xx/ (ag71xx.h ag71xx_main.c): ar71xx: ag71xx: fix MAC address setup Mar 09 18:05:40 juhosg * r20093 /trunk/target/linux/ar71xx/files/net/dsa/ar7240.c: ar71xx: fix a typo in the ar7240 dsa driver Mar 09 18:05:43 juhosg * r20094 /trunk/target/linux/ar71xx/base-files/sbin/wget2nand: ar71xx: improve wget2nand script Mar 09 18:05:47 juhosg * r20095 /trunk/target/linux/ar71xx/files/arch/mips/ar71xx/ (5 files): ar71xx: allow overriding of the lan port names on AP91 based boards Mar 09 18:05:53 juhosg * r20096 /trunk/target/linux/ar71xx/ (2 files in 2 dirs): ar71xx: override lan port names on the RB750 Mar 09 18:05:54 juhosg * r20097 /trunk/target/linux/ar71xx/base-files/etc/uci-defaults/rb750: ar71xx: add uci-defaults for the RB750 board Mar 09 18:05:58 juhosg * r20098 /trunk/target/linux/adm5120/files/arch/mips/adm5120/edimax/br-6104kp.c: Mar 09 18:05:58 adm5120: fix USB LED configuration for the BR-6104KP (closes #4187) Mar 09 18:05:58 * Thanks to Eugeny Zadevalov Mar 09 18:05:59 juhosg * r20099 /trunk/target/linux/ar71xx/base-files/etc/defconfig/pb44/ (. network): ar71xx: add default network configuration for the PB44 board Mar 09 18:43:03 nbd: ping Mar 09 18:43:29 Hauke1: ping Mar 09 18:43:42 rtz3: pong Mar 09 18:44:36 Hauke1: regarding this patch: https://dev.openwrt.org/changeset/20072/trunk/target/linux/brcm47xx Mar 09 18:44:46 rtz3: pong Mar 09 18:44:49 tripolar_: pong Mar 09 18:44:57 Hauke1: do you know, if this fixes the boot problem with the asus rt n16? Mar 09 18:45:32 rtz3: I do not know, I do not have such a device Mar 09 18:46:16 nico * r20100 /packages/net/ntpd/Makefile: [packages] ntp: disable snmp explicitely Mar 09 18:46:42 Hauke1: ok Mar 09 18:46:45 had to ask Mar 09 18:51:25 rtz3: the asus n16 seems to use a diverent cpu than most of the other chips Mar 09 18:56:06 nico * r20101 /packages/libs/libsdl/Makefile: [packages] SDL: use CONFIGURE_ARGS & CONFIGURE_VARS instead of Build/Configure/Default, install sdl-config in $(STAGING_DIR)/usr/bin/ and a relative symlink to it in $(STAGING_DIR)/host/bin/ Mar 09 19:11:55 nico * r20102 /packages/libs/libsdl_sound/Makefile: [packages] SDL_sound: use PKG_INSTALL, disable all external libs use explicitely, don't ship dev shlib symlink, bump release number Mar 09 19:14:23 what about this ticket: https://dev.openwrt.org/ticket/6824 ? Mar 09 19:18:37 nbd could u please apply my busybox patch? Mar 09 19:18:50 it updates busybox to 1.16.0 Mar 09 19:20:19 dogge10: -ENODETAIL Mar 09 19:23:58 jow_laptop need a serial log? Mar 09 19:26:53 dogge10: attach it to the ticket Mar 09 19:32:08 build #23 of iop32x is complete: Failure [failed shell_3] Build details are at http://tksite.gotdns.org:8010/builders/iop32x/builds/23 Mar 09 19:32:12 build #24 of mpc52xx is complete: Failure [failed svn compile_15] Build details are at http://tksite.gotdns.org:8010/builders/mpc52xx/builds/24 Mar 09 19:59:35 xMff that's it, i don't get more on the serial console: Mar 09 19:59:37 loading kernel from nand... OK Mar 09 19:59:37 setting up elf image... Mar 09 19:59:44 just hangs there :( Mar 09 20:00:17 and yes. i flashed the ramdisk elf image Mar 09 20:02:07 xMff any idea? Mar 09 20:21:49 cshore: ping Mar 09 20:43:12 xMff https://dev.openwrt.org/ticket/6812 Mar 09 21:05:14 xMff ping Mar 09 21:09:31 dogge10: #6824 sounded like there is an issue with preinit, the stuff you wrote looks like the kernel does not start at all, this is a big difference Mar 09 21:11:29 xMff https://dev.openwrt.org/ticket/6812 - also look here. that is my problem Mar 09 21:11:49 dogge10: yeah, seen that Mar 09 21:12:18 in ticket #6812 is the same on a 433ah, mine is 433uah Mar 09 21:14:59 xMff what could be wrong? Mar 09 21:15:22 no idea, did it ever work with 2.6.32 ? Mar 09 21:15:37 iirc, yes Mar 09 21:16:25 this narrows the time window to 6 weeks Mar 09 21:16:35 nbd * r20103 /trunk/package/hostapd/patches/350-wds_fix.patch: hostapd: fix setting the station vlan for wds stations on reauthentication Mar 09 21:16:39 nbd * r20104 /trunk/package/hostapd/ (Makefile files/hostapd.hotplug patches/360-wds_bridge.patch): hostapd: remove the hotplug script for adding wds station interfaces to a bridge - it suffers from race conditions. instead let hostapd add the interfaces to the bridge Mar 09 21:17:37 nbd https://dev.openwrt.org/ticket/6812 - you have any idea? Mar 09 21:17:49 nico * r20105 /packages/libs/ucl/Makefile: [packages] libucl: don't ship dev shlib symlink, bump release number Mar 09 21:17:49 i don't have this hardware Mar 09 21:17:53 no idea about this stuff Mar 09 21:17:53 dogge10: apart from bisecting I have no idea how to find out whats wrong, if the kernel does not start at all it is maybe a compression issue... lzma kernel vs. gz kernel Mar 09 21:18:31 k. trying with gziped kernel then Mar 09 21:19:08 maybe just works then Mar 09 21:23:19 nbd * r20106 /trunk/package/mac80211/files/lib/wifi/mac80211.sh: Mar 09 21:23:19 mac80211: fix interface setup for wds client Mar 09 21:23:19 sometimes an interface does not get added to the bridge properly Mar 09 21:23:19 this seems to be a race of interface setup vs wpa_supplicant Mar 09 21:23:19 fix this by configuring the interface only after wpa_supplicant has been started Mar 09 21:23:41 nico * r20107 /packages/utils/upx/Makefile: [packages] upx: define UCL_NO_ASM to fix build on x86 (ucl is built with --disable-asm), add missing dependencies, cleanup Mar 09 21:33:26 nbd * r20108 /trunk/package/hostapd/patches/350-wds_fix.patch: hostapd: restore wds sta state after the sta reassociates Mar 09 21:38:43 nico * r20109 /packages/mail/nullmailer/Makefile: [packages] nullmailer: remove CONFIGURE_ARGS with args already set by default, cleanup Mar 09 21:43:19 nbd * r20110 /trunk/target/linux/generic-2.6/files/drivers/net/phy/ar8216.h: ar8216: fix the mtu register definition Mar 09 21:52:42 nbd: ping Mar 09 22:07:23 rtz2: pong Mar 09 22:09:34 nbd: ok, the wrt160nl again Mar 09 22:09:50 nbd: there are 3 options I thought of Mar 09 22:10:14 nbd: 1. fix the trx on first boot the same way it's done on brcm47xx Mar 09 22:14:32 nbd: sorry, lost my internet connection Mar 09 22:15:03 nbd: anyway, 2. do the thing you suggested with a small program on first boot which fixes up the trx Mar 09 22:15:58 nbd: and 3. Mar 09 22:16:04 this is a bit more complex Mar 09 22:17:15 rtz2: do the thing I suggested with changing the detection logic to look for the jffs header not being present instead of 0xdeadcode being present Mar 09 22:18:10 nbd: it's possible to force the firmware to delete the first block of the jffs2 partition Mar 09 22:18:54 nbd: the trx would contain the first byte of the jffs2 header Mar 09 22:19:35 nbd: this byte doesn't change after the jffs2 is formated, so the checksum stays the same Mar 09 22:20:20 i'd prefer option 2 Mar 09 22:20:27 nbd: but the firmware will still erease the entry block, so there isn't the problem with old partitions Mar 09 22:20:42 the program only needs to rewrite one block Mar 09 22:20:48 and if that goes wrong, the boot loader will launch the tftp server Mar 09 22:21:25 if you look at older revs you can probably find a small c program that already does 90% of this Mar 09 22:21:31 and mtd also contains part of that Mar 09 22:21:42 so mtd could be changed to include a platform specific command for recalculating trx Mar 09 22:21:51 nbd: why wasn't this done so in the first place? Mar 09 22:21:56 reusing the code that was already added for broadcom Mar 09 22:22:03 you mean on broadcom? Mar 09 22:22:06 yes Mar 09 22:22:27 i think preinit was a bit different back then, making the kernel part a bit more practical Mar 09 22:22:30 but i'm not sure Mar 09 22:22:31 it was a long time ago Mar 09 22:22:44 when this extra stuff is done, we can probably switch broadcom over to it as well Mar 09 22:24:20 nbd: ok, I will see what I can do Mar 09 22:24:28 thanks Mar 09 22:24:45 nbd: out of interest, why not use the kernel solution like on broadcom? Mar 09 22:25:14 because it's ugly Mar 09 22:25:19 and i'd like to keep ugly things outside of the kernel Mar 09 22:25:20 ;) Mar 09 22:25:33 makes it easier to sync up with upstream at some point Mar 09 22:26:28 ok Mar 09 22:26:48 i'm studying physics, I always ask why :) Mar 09 22:33:08 ;) Mar 09 22:59:34 nbd: I would change mtd into a multicall binary, would this be ok? **** ENDING LOGGING AT Tue Mar 09 23:10:55 2010 **** BEGIN LOGGING AT Wed Mar 10 21:48:21 2010 Mar 10 21:51:20 hauke * r20125 /trunk/package/mac80211/Makefile: Mar 10 21:51:20 mac80211: fix build problems with kmod-mac80211=n Mar 10 21:51:20 Build compat-wireless also if kmod-mac80211=n, but do not build mac80211.ko. Mar 10 21:51:20 Fix problem in ticket #6813. Mar 10 22:24:54 nico * r20126 /trunk/target/linux/ar7/config-2.6.32: [target] ar7: platform has *no* PCI support, override CONFIG_PCI default **** BEGIN LOGGING AT Thu Mar 11 09:00:15 2010 Mar 11 09:11:46 build #33 of avr32 is complete: Failure [failed compile_4] Build details are at http://tksite.gotdns.org:8010/builders/avr32/builds/33 Mar 11 09:12:47 nico * r20133 /trunk/package/kernel/modules/sound.mk: [package] kernel: add missing CONFIG_SND_SOC_ALL_CODECS to kmod-sound-soc-core KCONFIG Mar 11 09:16:57 florian * r20134 /trunk/target/linux/rdc/files-2.6.30/ (arch/x86/mach-rdc321x/pci.c drivers/watchdog/rdc321x_wdt.c): Mar 11 09:16:57 [rdc] define panic_on_unrecovered_nmi in the southbridge driver and fix watchdog init message Mar 11 09:16:57 Patch from Bernhard Loos. Mar 11 09:19:35 nico * r20135 /trunk/target/linux/avr32/config-default: [target] avr32: add missing CONFIG_SND_ATMEL_SOC config symbol Mar 11 09:50:14 nico * r20136 /trunk/target/linux/ (59 files in 36 dirs): do *not* enable CONFIG_PCI in generic config, but on targets where it makes sense instead (partially revert [18370]) Mar 11 09:56:34 build #33 of brcm63xx is complete: Failure [failed compile_4] Build details are at http://tksite.gotdns.org:8010/builders/brcm63xx/builds/33 Mar 11 10:11:41 build #31 of brcm47xx is complete: Failure [failed compile_4] Build details are at http://tksite.gotdns.org:8010/builders/brcm47xx/builds/31 Mar 11 11:50:27 acoul * r20137 /trunk/toolchain/binutils/ (9 files in 2 dirs): add binutils-2.20.1 Mar 11 11:55:35 acoul * r20138 /trunk/target/linux/generic-2.6/config-2.6.34: generic-2.6: add 2.6.34 preliminary support Mar 11 11:57:23 acoul * r20139 /trunk/target/linux/ar71xx/config-2.6.34: ar71xx: add 2.6.34 preliminary support Mar 11 12:19:25 [florian]: ping Mar 11 12:24:20 acoul * r20140 /trunk/target/linux/generic-2.6/patches-2.6.34/ (105 files): generic-2.6: add 2.6.34 preliminary support (patches) Mar 11 12:25:52 acoul * r20141 /trunk/target/linux/ar71xx/patches-2.6.34/ (33 files): ar71xx: add 2.6.34 preliminary support (patches) Mar 11 12:26:19 <[florian]> rtz2: pong Mar 11 12:27:05 [florian]: I have a patch for rdc but it may break some things, so your would have to do some testing Mar 11 12:27:23 <[florian]> rtz2: allright, can you send it on the mailing-list, so that I can keep track of it? Mar 11 12:27:32 [florian]: can Mar 11 12:27:33 do Mar 11 12:27:49 [florian]: there are 2 main problems: Mar 11 12:28:32 1., I need the config and boot partitions from the bifferboard and the r8610, otherwise I can't find a string to identify them Mar 11 12:29:07 [florian]: 2. ar525w uses a checksum spanning the complete image file including the rootfs Mar 11 12:29:32 [florian]: this will lead to checksum errors on the second boot Mar 11 12:30:30 [florian]: I tried to solve it by only including the kernel in the image created by the ar525w tool and simply appending the rootfs to it Mar 11 12:30:54 [florian]: but this means, that the image length stored in the image header is too short Mar 11 12:31:44 [florian]: the bootloader flasher or the webinterface may or may not trunkate it to the lenght in the header during flashing Mar 11 12:31:51 you will have to try this Mar 11 12:32:20 but otherwise, I would have to fix up the checksum on first boot, which is pretty ugly Mar 11 12:32:20 <[florian]> rtz2: allright Mar 11 12:39:55 is juhosg online? Mar 11 12:43:51 swalker * r20142 /packages/libs/fftw3/Makefile: [packages] fftw3: correct SECTION variable Mar 11 13:08:44 acoul * r20143 /trunk/target/linux/adm5120/router_le/ (config-2.6.30 config-2.6.32): adm5120: add CONFIG_PCI=y on router_le too Mar 11 13:24:04 nbd * r20144 /packages/devel/binutils/Makefile: binutils: the shared variant of libbfd/libopcodes is being generated and used again, so it needs to be packaged as well Mar 11 13:38:24 is juhosg online? Mar 11 13:41:01 juhosg? Mar 11 13:54:08 when is juhosg online here? Mar 11 13:57:00 <[florian]> dogge10: are you doing to ask this every 10 minutes ? Mar 11 13:57:08 <[florian]> dogge10: if he does not answer he is not online Mar 11 13:57:14 <[florian]> dogge10: or not behind the keyboard Mar 11 15:30:40 nbd * r20145 /trunk/package/mac80211/patches/580-tx_status_optimization.patch: mac80211: improve performance a bit Mar 11 15:38:52 claudio * r20146 /trunk/package/kernel/modules/ (block.mk i2c.mk): [etrax] fix kernel modules dependences for i2c and ata-core Mar 11 16:30:52 build #24 of ixp4xx is complete: Failure [failed shell_3] Build details are at http://tksite.gotdns.org:8010/builders/ixp4xx/builds/24 Mar 11 16:37:13 nico * r20147 /trunk/package/kernel/modules/other.mk: [package] kernel: fix kmod-pcmcia-yenta on 2.4 Mar 11 16:50:32 build #30 of brcm_2_4 is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/brcm_2_4/builds/30 Mar 11 17:17:02 nico * r20148 /trunk/target/linux/ (5 files in 5 dirs): remove empty directories (again) Mar 11 18:48:55 nbd * r20149 /trunk/target/linux/ (13 files in 10 dirs): the mips -ffreestanding patch has been unnecessary for a long time now, remove it Mar 11 18:48:58 nbd * r20150 /trunk/target/linux/ (5 files in 5 dirs): ar71xx: fix oprofile support Mar 11 18:49:03 nbd * r20151 /trunk/target/linux/generic-2.6/ (5 files in 5 dirs): mips: allow the compiler to optimize memset, memcmp, memcpy for better performance and (in some instances) smaller code Mar 11 18:57:05 build #23 of xburst is complete: Failure [failed compile_4] Build details are at http://tksite.gotdns.org:8010/builders/xburst/builds/23 Mar 11 19:02:01 lars * r20152 /trunk/target/linux/xburst/image/Makefile: [xburst] Fix image generation for qi_lb60 subtarget Mar 11 19:20:50 sry 4 short wiki downtime. pls excuse the inconvience ;) Mar 11 19:21:03 inconvenience .. Mar 11 19:21:05 aregl Mar 11 19:21:08 o.O Mar 11 19:26:21 juhosg ? Mar 11 19:29:37 dogge10_: he is not on the channel Mar 11 19:29:50 :( Mar 11 19:30:04 dogge10_: I think you might have more luck by mail Mar 11 19:30:20 http://openwrt.pastebin.com/XJsW2Rbq - any idea about this? Mar 11 19:30:22 juhosg@openwrt.org Mar 11 19:30:52 ok. thanks Mar 11 19:36:20 jow_laptop: erm.. :p Mar 11 19:36:49 Kaloz: there's always hope Mar 11 19:37:28 jow_laptop: do you want me to transfer the spams to your inbox ?;) Mar 11 19:38:07 jow_laptop: pasting a full email address to a channel which has public logs isn't a smart decision ;) Mar 11 19:39:34 hehe Mar 11 19:39:55 brb Mar 11 19:40:40 ah you mean that, well sorry then Mar 11 19:41:12 this particular mail is all over the web anyway Mar 11 19:41:21 but won't happen again Mar 11 19:55:35 jow_laptop opened a ticket Mar 11 19:56:16 no I did not Mar 11 20:04:23 i did it :) Mar 11 22:08:10 nbd * r20153 /trunk/package/mac80211/patches/560-minstrel_ht.patch: minstrel_ht: make the ewma of the ampdu length more smooth by updating it on stats updates and reduce the frequency of aggregation checks Mar 12 00:35:38 nbd * r20154 /trunk/tools/sstrip/src/sstrip.c: sstrip: fix a section length corruption bug (patch from #6847) Mar 12 01:50:41 build #24 of pxcab is complete: Success [build successful] Build details are at http://tksite.gotdns.org:8010/builders/pxcab/builds/24 **** ENDING LOGGING AT Fri Mar 12 02:59:57 2010