**** BEGIN LOGGING AT Thu Jul 07 02:59:56 2011 Jul 07 03:13:18 cshore * r27502 /trunk/target/linux/ar7/base-files/etc/diag.sh: [ar7] base-files: diag.sh: added failsafe led definition Jul 07 03:50:28 cshore, ping Jul 07 03:50:35 pong Jul 07 03:50:59 package/base-files/files/lib/preinit/02_default_set_state Jul 07 03:51:07 yes? Jul 07 03:51:21 is /etc/functions.sh missing? Jul 07 03:51:29 idk just reading heh Jul 07 03:51:33 no, previously sourced Jul 07 03:51:40 kk :) Jul 07 03:51:42 oh wait. Jul 07 03:51:47 sorry, no, don't need it Jul 07 03:52:08 :) Jul 07 03:52:15 * RealOp bored Jul 07 04:24:02 cshore * r27503 /packages/net/openvpn/Makefile: [packages] openvpn: openvpn-easy-rsa doesn't actually depend on openvpn, it's just invented by openvpn and in that source package Jul 07 04:35:31 hi, anybody around? Jul 07 04:36:12 I have ported my board to OpenWR Jul 07 04:36:15 T Jul 07 04:36:42 I am using a kernelthat is based on 2.6.37 Jul 07 04:37:01 I've created a mega-patch to be applied on top of OpenWRT's generic patches Jul 07 04:37:57 the problem is the following: my patch won't apply on 2.6.37.6 (that is used for instance by the kirkwood port) Jul 07 04:38:37 and some OpenWRT patches won't apply on 2.6.37 (in particular, 821 and 950) Jul 07 04:45:43 cshore: ping Jul 07 04:45:48 [pmg Jul 07 04:45:50 pong Jul 07 04:46:47 just about done a build of my main router, then (tomorrow as I'm getting tired, but will be here for a little while yet), I will look at x86 sysupgrade Jul 07 04:50:58 I'm thinking the solution is to do something like ar71xx sysupgrade where there is an image that contains both the kernel and rootfs, writes the kernel to the boot (grub+kernel) partition, and then uses mtd to write the rootfs to the existing mtd (which exists via mtd2block). I think the ext4 rootfs case needs to be handled differently, because you can't use -j in that case. Jul 07 04:51:56 using dd won't work for the mtd case because mtd2block will die horribly Jul 07 04:58:32 well, what's the chance that someone will want to specify kernel params explicitly in their grub.conf that they'll want propagated across updates? Jul 07 04:58:46 and did you have a chance to look at the Perl eglibc patch? Jul 07 04:59:13 hmmm...you're thinking dd the kernel part? Jul 07 04:59:57 and no Jul 07 05:00:04 I'm thinking that we might want to save the grub config and restore it after the update. Jul 07 05:00:31 well if you only update the kernel the grub config doesn't change Jul 07 05:00:47 if you dd the kernel part on the other hand, it becomes an issue Jul 07 05:02:42 not sure if the filename of the kernel will be an issue or not...I think it's just vmlinuz or some such Jul 07 05:02:54 I don't know enough about how it's stored to say one way or another. Jul 07 05:03:17 I'll be looking at that Jul 07 05:11:44 oh yeah, what image have you been flashing (filename) for sysupgrade, up til now? Jul 07 05:12:52 openwrt-x86-geos-combined-squashfs.img Jul 07 05:13:47 btw: who owns the OCF stuff? in other words, who do I need to bug to get a code review? :-) Jul 07 05:15:11 no idea Jul 07 05:22:48 cshore * r27504 /packages/libs/avahi/Makefile: [packages] libs/avahi: Version bump (forgotten on previous update) Jul 07 05:37:51 is there a "list" of possible items to be put in "FEATURES" in the makefile? Jul 07 06:34:33 cshore: nbd says it's currently unowned, so I could wait for a long time to get OCF approved. Jul 07 07:08:15 hey nbd thanks for the help yesterday, today I was able to boot Jul 07 07:08:31 I think I made a preliminary port of OpenWRT for my board Jul 07 07:09:10 there is a small problem with the serial console anyway: my serial port is /dev/ttyAMA0 in the kernel and I pass it as command line option Jul 07 07:09:30 I modified /etc/inittab to open a terminal on that device Jul 07 07:09:59 but I still feel like some messages, for example the ones printed during userspace boot, are missing Jul 07 07:10:22 is there a way to set the default serial console when building an OpenWRT image? Jul 07 08:11:33 florian * r27505 /packages/libs/argp-standalone/Makefile: [package] argp-standalone: should be compile with fpic Jul 07 08:11:35 florian * r27506 /packages/libs/elfutils/ (Makefile patches/006-libdw_LIBS.patch): [package] elfutils: link with libargp Jul 07 08:58:29 cshore * r27507 /trunk/package/firewall/files/lib/core_redirect.sh: [package] firewall: fix wrong variable names for protocol command line parameter - were missed during r27500 Jul 07 09:03:40 cshore * r27508 /trunk/package/firewall/files/lib/core_redirect.sh: [package] firewall: also correct another variable missed in previous commit Jul 07 09:46:44 florian * r27509 /trunk/toolchain/eglibc/Makefile: [toolchain] remove reference to old eglibc versions. Jul 07 09:46:48 florian * r27510 /trunk/toolchain/eglibc/ (Config.in Config.version Makefile): [toolchain] eglibc: add support for 2.14 Jul 07 09:46:52 florian * r27511 /trunk/target/linux/malta/le/config-default: [malta] enable perf counter support Jul 07 09:46:54 florian * r27512 /trunk/target/linux/realview/config-default: [realview] enable support for perf events Jul 07 09:46:57 florian * r27513 /trunk/toolchain/uClibc/ (config-0.9.32/common patches-0.9.32/970-uclibc_fts.patch): [toolchain] uclibc: backport FTS support Jul 07 09:47:00 florian * r27514 /trunk/target/linux/generic/patches-2.6.39/101-perf_arm_uclibc.patch: [kernel] fix building of perf for arm on uclibc Jul 07 09:47:04 florian * r27515 /trunk/package/perf/ (. Makefile): [package] add perf Jul 07 10:27:51 jow * r27516 /packages/Xorg/lib/qt4/Makefile: [packages] qt4: disable on brcm-2.4, it fails due to missing tslib support Jul 07 10:28:38 kaloz * r27517 /trunk/target/linux/cns3xxx/patches-3.0/001-cns3xxx_v6k_cpu.patch: [cns3xxx]: select the right cpu Jul 07 10:30:17 kaloz * r27518 /trunk/target/linux/generic/ (config-3.0 patches-3.0/006-arm_kernel_xz_support.patch): [generic/3.0]: add support for and switch to XZ compressed kernels on ARM Jul 07 10:32:32 jow * r27519 /trunk/package/firewall/ (Makefile files/lib/core_redirect.sh): [package] firewall: make sure that -m mac is used with --mac-source, follow up to r27508 Jul 07 10:49:29 jow * r27520 /packages/libs/boost/Makefile: [packages] boost: disable on brcm-2.4, does not build Jul 07 11:15:00 jow * r27521 /packages/libs/avahi/Makefile: [packages] avahi: disable the dbus support on brcm-2.4, it does not build Jul 07 13:05:50 juhosg * r27522 /trunk/target/linux/ar71xx/generic/profiles/atheros.mk: ar71xx: add separate profiles for the AP121 and AP121-MINI boards Jul 07 13:05:52 juhosg * r27523 /trunk/target/linux/ar71xx/image/Makefile: ar71xx: build image for the AP121 boards with 4M of flash Jul 07 13:34:42 kaloz * r27524 /trunk/target/linux/generic/patches-3.0/999-drop_sublevel.patch: [generic/3.0]: 3.0 is at rc6 now Jul 07 13:55:52 jow_laptop: If you have sometimes to spare, can you please help me to resolve this issue (https://forum.openwrt.org/viewtopic.php?id=30584) where the 'HOST="$(GNU_TARGET_NAME)"' variable on an OpenWRT Makefile can't be passed to any source package Makefile which calls for a GNU configure tool. Jul 07 17:03:10 hi Jul 07 17:03:31 just checking if there is somebody that can comment on my "problem" Jul 07 17:03:40 my serial console is /dev/ttyAMA0 Jul 07 17:03:57 I have specified it as a command line option to the kernel so I get boot messages Jul 07 17:04:14 I also changed /etc/inittab so I get a login on that serial port Jul 07 17:04:27 but I still feel there is something missing Jul 07 17:04:43 is there a way to setup the default serial console in OpenWRT? Jul 07 17:05:42 <[florian]> you should provide a board specific /etc/inittab file like I suggested you Jul 07 17:05:47 <[florian]> see how it's done for ep93xx or realview Jul 07 17:06:54 yes I did that Jul 07 17:07:05 I mentioned it above Jul 07 17:07:27 but I'm not too concerned about the login - that is there Jul 07 17:07:34 it's more like the boot messages Jul 07 17:08:37 but maybe it's simply openwrt that is "terse" with messages at startup and shutdown Jul 07 17:08:40 <[florian]> there are no boot messages Jul 07 17:08:45 <[florian]> not more than - preinit - ... Jul 07 17:08:48 compared to a regular sysV Jul 07 17:08:49 <[florian]> and the modules loading Jul 07 17:08:50 oh ok Jul 07 17:08:58 okok Jul 07 17:09:00 <[florian]> the rest is logged in /var/log/ Jul 07 17:09:20 do you know if I can specify a default console when I compile the *kernel*? Jul 07 17:09:36 because for now I need to modify U-Boot's config to pass the correct option Jul 07 17:10:17 <[florian]> you can do that with CMDLINE_FORCE and CONFIG_CMDLINE in your kernel configuration Jul 07 17:11:03 so there is no "option" in the target/linux/{platform}/Makefile for that? Jul 07 17:11:17 <[florian]> that must be done in the kernel config for your platform Jul 07 17:11:25 <[florian]> so under target/linux/{platform}/config* Jul 07 17:12:42 I see... Jul 07 17:13:05 oh another, quite unrelated, thing: [lib]gphoto packages in OpenWRT are ANCIENT Jul 07 17:13:44 do you think they could be updated to the latest version? I could prepare the necessary modifications and test Jul 07 17:14:07 and then give the result to the maintainer (if there is one) Jul 07 17:14:24 <[florian]> sure, just send the patches on the mailing-list Jul 07 17:14:59 ok, openwrt-devel@lists.openwrt.org this is the mailing list address correct? Jul 07 17:18:49 I have added CONFIG_CMDLINE="console=ttyAMA0,115200 rootfstype=jffs2" to my config-default Jul 07 17:21:11 kaloz * r27525 /trunk/target/linux/cns3xxx/patches-3.0/049-cns3xxx_smp_support.patch: [cns3xxx]: make smp code more similar to upstream Jul 07 17:21:47 kaloz * r27526 /trunk/target/linux/cns3xxx/config-default: [cns3xxx]: enable smp support Jul 07 17:22:09 kaloz * r27527 /trunk/target/linux/generic/patches-3.0/006-arm_kernel_xz_support.patch: [cns3xxx]: refresh patches Jul 07 17:22:42 doh, bad commit message :) Jul 07 17:22:53 kaloz * r27528 /trunk/target/linux/cns3xxx/patches-3.0/ (001-cns3xxx_v6k_cpu.patch 100-laguna_support.patch): [cns3xxx]: refresh patches Jul 07 18:23:54 blogic * r27529 /trunk/package/pjsip/Config.in: pjsip-ltq-tap is for all lantiq targets Jul 07 18:24:01 blogic * r27530 /trunk/package/ltq-vmmc/patches/400-falcon.patch: fixes ltq-vmmc for none falcon targets Jul 07 18:24:05 blogic * r27531 /trunk/package/ltq-ifxos/ (5 files in 2 dirs): update ltq-ifxos to version 1.5.14 Jul 07 18:24:10 blogic * r27532 /trunk/target/linux/lantiq/patches-3.0/ (450-mach-arv45xx.patch 470-mach-gigasx76x.patch): remove CONFIG_MTD_PARTITIONS references from lantiq 3.0 patches Jul 07 18:24:15 blogic * r27533 /trunk/package/ltq-tapidemo/ (5 files in 2 dirs): update ltq-tapidemo to 5.1.0.53 Jul 07 18:24:20 blogic * r27534 /trunk/target/linux/lantiq/xway/profiles/ (001-lantiq.mk 003-gigaset.mk 003-netgear.mk 004-gigaset.mk): remove ar9 board from xway profiles Jul 07 18:24:25 blogic * r27535 /trunk/target/linux/lantiq/ (18 files in 4 dirs): rename xway profile to danube and remove ar9 specifics Jul 07 18:24:30 blogic * r27536 /trunk/target/linux/lantiq/ (10 files in 3 dirs): add ar9 specific profile Jul 07 18:24:35 blogic * r27537 /trunk/target/linux/lantiq/image/Makefile: fix lantiq image makefile to represent new split between ar9 and danube Jul 07 18:24:40 blogic * r27538 /trunk/ (7 files in 6 dirs): replace all references to xway with danube Jul 07 18:24:45 blogic * r27539 /trunk/package/ltq-vmmc/ (6 files in 2 dirs): update ltq-vmmc patches Jul 07 18:24:50 blogic * r27540 /trunk/package/ltq-vmmc/Config.in: fix ltq-vmmc SoC dependency Jul 07 18:24:55 blogic * r27541 /trunk/target/linux/lantiq/ (falcon/config-default falcon-stable/config-default): fix i2c-falcon kmod package selection Jul 07 18:25:00 blogic * r27542 /trunk/target/linux/lantiq/ (3 files in 2 dirs): fix up falcon kernel config Jul 07 18:25:05 blogic * r27543 /trunk/package/kernel/modules/netdevices.mk: kmod-dm9000 should build as a module Jul 07 20:00:07 blogic * r27544 /branches/backfire/ (144 files in 42 dirs): drop ifxmips Jul 07 20:00:11 blogic * r27545 /branches/backfire/target/linux/generic-2.6/patches-2.6.32/251-atm_header_changes.patch: add atm header patch needed by ltq-dsl Jul 07 20:00:23 blogic * r27546 /branches/backfire/target/linux/lantiq/ (87 files in 16 dirs): add new lantiq target Jul 07 20:00:35 blogic * r27547 /branches/backfire/package/ (58 files in 17 dirs): adds lantiq packages/drivers Jul 07 20:06:54 blogic * r27548 /branches/backfire/package/kernel/modules/other.mk: add 2 new kmod defines for at24/25 Jul 07 20:08:59 blogic: ping Jul 07 20:10:13 yes Jul 07 20:10:16 i am here Jul 07 20:10:40 luka12345|wiik: my patches will probably break your overlay tree on googlecode Jul 07 20:10:51 i see you had a busy night :) Jul 07 20:10:53 the danube boards are now danube and not xway anymore Jul 07 20:10:59 i am not done yet Jul 07 20:11:08 i'm looking at the code now Jul 07 20:11:17 it is mainly cleanup Jul 07 20:11:25 and a few updates to the tapi stuff Jul 07 20:11:27 i backed up my diff :) break all you want :) Jul 07 20:11:38 and merging stuff to backfire Jul 07 20:13:04 i have been playing some more with sx763 Jul 07 20:13:10 i'll update a wiki Jul 07 20:13:29 and mail you & ralph Jul 07 20:13:58 ok Jul 07 20:14:10 also i have discovered that adm6996m driver is not working :( Jul 07 20:14:15 i had not time yet to look at the switch issue Jul 07 20:14:20 the kernel driver ? Jul 07 20:14:22 ok Jul 07 20:14:25 what does it do worng ? Jul 07 20:14:31 i can do the driver (i think) Jul 07 20:14:42 well it's recognized Jul 07 20:14:49 its a I not a M Jul 07 20:15:05 but when i configure with swconfig it's not working Jul 07 20:15:09 i know Jul 07 20:15:20 that they are different Jul 07 20:15:29 but i have been looking at datasheets Jul 07 20:15:36 it's all the same to me :) Jul 07 20:15:46 finish your commiting Jul 07 20:15:54 we can talk later about this Jul 07 20:16:32 there is siemens source online for this so i guess i can do it... Jul 07 20:18:22 ok Jul 07 20:24:50 blogic * r27549 /trunk/target/linux/lantiq/ (6 files in 3 dirs): Jul 07 20:24:50 adds 2 ASC related patches to lantiq Jul 07 20:24:50 Signed-off-by: Thomas Langer Jul 07 20:24:54 blogic * r27550 /trunk/target/linux/lantiq/image/Makefile: fix generation of falcon images for profiles Jul 07 20:26:41 i think i am done Jul 07 20:26:48 * blogic gets a whisky Jul 07 20:26:49 brb Jul 07 20:54:49 blogic: personaly i'm more beer type, but whisky is also fine Jul 07 21:30:22 blogic: do you have those drivers for fxo port on reference board? I see that those 2 chips are the same as those on sx763 Jul 07 21:32:10 i have some but they are old from 2.4 amazon sources Jul 07 21:40:49 the litelink ? Jul 07 21:40:55 its not supported anymore Jul 07 21:41:19 afaik the tapi/vmmc sources dont support them anymore Jul 07 21:42:02 what spurces do you have exactly ? Jul 07 21:42:48 amazon_daa-0.1.1 thats for litelink Jul 07 21:44:07 and duslic-0.0.2 i think that's for peb3164 Jul 07 21:45:21 yes Jul 07 21:46:05 i will dig out the hdd that has that stuff on it tomorrow or so and see what we can do about it Jul 07 21:46:18 i have a version newer than 0.1.1 and 0.2 Jul 07 21:46:30 ah ok Jul 07 21:46:57 i think so at least Jul 07 21:46:59 :) Jul 07 21:48:08 but probably it's api is different, so it won't work with that danube-voip sip client :( Jul 07 21:48:24 indeed Jul 07 21:48:31 it might work with tapidemo Jul 07 21:56:04 Delboy: i am working on owsip just now Jul 07 21:56:17 not sure how much longer it will take Jul 07 21:56:20 it is a backgorund task Jul 07 22:06:09 how much ram does it use? because with all that adsl and tapi/vmmc and danube-voip only few megs are free Jul 07 22:13:52 alot Jul 07 22:13:53 ;) Jul 07 22:13:58 i am still optimizing Jul 07 22:14:05 something like 8MB Jul 07 22:21:24 not that good, looks like I'm gonna need to solder a bigger ram chip if i plan to use more stuff :) Jul 07 22:23:03 well Jul 07 22:23:06 not that easy Jul 07 22:23:13 the uboot has a ram specific init part Jul 07 22:23:20 changing the ram wont work Jul 07 22:23:43 kaloz * r27551 /trunk/toolchain/gcc/patches/linaro/200-ppc_include_config_linux_h.patch: [toolchain/gcc/linaro]: fix ppc toolchain build Jul 07 23:48:36 alesan: lots of packages are ancient & no, the [lib]gphoto packages don't have a maintainer (few packages do) Jul 07 23:49:47 swalker, I am pretty familiar with that package because I am the author of a program that uses it Jul 08 00:06:55 how do i make a host package? i always forget this syntax Jul 08 00:07:12 make package/host/mtd/{clean,compile}? Jul 08 00:07:27 i need to build host/mtd-20101124 Jul 08 00:07:53 and cut a patch, since u_long is undefined on OS X 10.7 for libfec.c Jul 08 00:08:02 without sys/types.h included Jul 08 00:08:58 oh its mtd-utils Jul 08 00:09:00 not mtd Jul 08 02:30:05 is there a comprehensive list of arguments for the FEATURES setting? **** ENDING LOGGING AT Fri Jul 08 02:59:58 2011