**** BEGIN LOGGING AT Sun Jun 19 02:59:57 2011 Jun 19 06:56:37 so, the buildslaves are still down :( Jun 19 06:56:41 it's been 9 days now Jun 19 10:14:16 ping johnyb Jun 19 10:14:19 ping jow_laptop Jun 19 10:14:53 tripolar: pong Jun 19 10:15:09 johnyb sorry i wanted to talk to jow .. Jun 19 10:15:18 ahh, never mind ;) Jun 19 10:17:03 ping nbd Jun 19 10:23:34 blogic: ping Jun 19 10:52:53 hi Jun 19 10:52:58 luka12346|wiik: pong Jun 19 10:53:05 hi Jun 19 10:53:08 was a misunderstnding with th uboot ;) Jun 19 10:53:14 i will do it tomorow in the evening Jun 19 10:53:14 np :) Jun 19 10:53:23 as for the mtd driver Jun 19 10:53:30 what is wrong with the exisiting one ? Jun 19 10:53:41 you mean mtd layout? Jun 19 10:53:58 no Jun 19 10:54:07 in your mail you have added a new nor driver Jun 19 10:54:32 the danube.c file Jun 19 10:54:43 that was just a reference Jun 19 10:54:51 how they did it in original source Jun 19 10:54:58 that is not important at all Jun 19 10:55:05 ok Jun 19 10:55:12 so the patch is ok? Jun 19 10:55:25 i had a quick broswe Jun 19 10:55:32 i will start merging the stuff tomorrow Jun 19 10:55:37 great :) Jun 19 10:55:44 i have one question Jun 19 10:56:02 i have added separate kernel and rootfs mtd partition Jun 19 10:56:07 yes Jun 19 10:56:20 but i see in mach-arv45xx.c that there is only linux Jun 19 10:56:25 yes Jun 19 10:56:25 what is the way to go? Jun 19 10:56:29 linux Jun 19 10:56:36 i will take care of that once i have uboot working Jun 19 10:56:43 dont worry about it Jun 19 10:56:43 ok Jun 19 10:56:48 basically we have a patch Jun 19 10:56:56 ok Jun 19 10:57:08 did that linux part in the mtd changed recently Jun 19 10:57:12 it looks at the start of the part for the uimage magic and works out the size of the kernel and dynamically allocates the rootfs/kernel partitoion Jun 19 10:57:20 thus making optimal use of the available space Jun 19 10:57:31 that is great Jun 19 10:57:53 and wwe will add the arteep.bin to the machfile in this case and also i will test it with ath5k Jun 19 10:58:11 so lets hope that tuesday/wednesday it is all working in trunk Jun 19 10:58:17 ok great Jun 19 10:58:20 i need to see how my workload looks monday morning Jun 19 10:58:25 np Jun 19 10:58:30 take it slow :) Jun 19 10:58:47 main thing is that we have working patch Jun 19 10:59:35 i'm going to try now luca's voip code Jun 19 10:59:53 what shall we do with usb? Jun 19 11:00:27 is the dwc driver obsolete? is there newer version? Jun 19 11:01:35 blogic can you look at changeset 27138 - jow made a patch to get base-files installed as last package - when an image is build, but this is not right, because if someone adds a package that modifies a file that base-files provide like groug/passwd, the changes will be lost... Jun 19 11:02:01 groug=group Jun 19 11:21:45 luka12346|wiik: the usb patch is weird Jun 19 11:21:50 you can use it for now as it is Jun 19 11:22:06 i will try to figure out why it is needed and add a sane/clean versio of it. Jun 19 11:22:24 as for voip, make sure to use the ltq_register_tapi() before loading the modules Jun 19 11:22:40 otherwise you will see something along the lines of "-1 firmware load failed" Jun 19 11:23:12 tripolar ... you leaft Jun 19 11:23:15 tripolar ... you left Jun 19 11:32:55 tripolar: hi Jun 19 11:32:58 let me have a look Jun 19 11:33:06 hi blogic Jun 19 11:33:28 hmmm Jun 19 11:33:33 i see what you mean Jun 19 11:33:33 its not right - base-files should be the first package that gets installed Jun 19 11:34:00 errrmmm Jun 19 11:34:07 i would say you are correct Jun 19 11:34:17 let me find out why it was done this way Jun 19 11:34:23 it will take a few hours Jun 19 11:34:43 its 13:33 here which is way too early ;) Jun 19 11:35:04 What pkg provide libcom_err.so.2? Jun 19 11:35:30 blkid need it. Jun 19 11:35:37 blogic: $(curdir)/builddirs-install:=. $(filter base-files,$(package-y)) $(filter-out base-files,$(sort $(package-y))) Jun 19 11:35:42 would be correct i think Jun 19 11:36:00 tripolar: yes Jun 19 11:36:26 or $(curdir)/builddirs-install:=. $(filter base-files,$(sort $(package-y))) $(filter-out base-files,$(sort $(package-y))) Jun 19 11:36:31 would be even better Jun 19 11:36:34 well Jun 19 11:36:41 packages should not overwrite eachothers files Jun 19 11:36:50 whyt is your reallife problem that you see ? Jun 19 11:36:55 *what Jun 19 11:37:00 or, lets me rephrase Jun 19 11:37:06 which package breaks due to this commit ? Jun 19 11:37:30 nbd: ping Jun 19 11:37:33 mirko: pong Jun 19 11:37:38 create a custom image with packages that add a group/user - this packages wont work, as base-files installed as last package will overwrite passwd/group Jun 19 11:37:47 https://dev.openwrt.org/ticket/9377 Jun 19 11:37:54 nbd: sth. like "DEPENDS:=+USE_EGLIBC:libbsd +USE_GLIBC:libbsd" seems not to work Jun 19 11:37:57 @blogic Jun 19 11:38:09 nbd: it just generates "$(curdir)/hotplug2/compile += $(if $(CONFIG_USE_GLIBC),$(curdir)/libbsd/compile)" Jun 19 11:38:14 nbd: is this intended? Jun 19 11:38:31 USE_EGLIBC is not taken into account here Jun 19 11:38:52 i'll take a look Jun 19 11:39:01 tripolar: package should use uci-defaults to add users/groups Jun 19 11:39:15 nbd: exmaple is the hotplug2-package Jun 19 11:39:17 nbd: thanks Jun 19 11:39:27 tripolar: nbd is right, use uci-daefaults Jun 19 11:39:46 tripolar: do you know how uci-defaults works ? Jun 19 11:40:02 there are several examples in /packages Jun 19 11:40:06 blogic: this will add a group even when the files passwd/group are overwritten? Jun 19 11:40:08 blogic: no Jun 19 11:40:20 tripolar: it will be run at first boot Jun 19 11:40:25 ahh okay Jun 19 11:41:26 nbd i tested the patch with a working image and the eth goes down when do the udp flood and remove ISTAT_RFO Jun 19 11:41:36 do/i do Jun 19 11:41:38 ok Jun 19 11:42:14 ok, then the code is probably ok as it is Jun 19 11:42:24 okay Jun 19 11:43:31 which package uses uci-defaults to add a user/group? Jun 19 11:45:36 hm, not sure Jun 19 11:45:46 i only vaguely remember seeing some that use it Jun 19 11:48:55 nbd: can take a look myself as well if you're busy.. just wanted to know whether it's known / intended.. Jun 19 11:51:47 nbd i found some packages that add a user/group for example tor-alpha and tor but they directly modify passwd/group Jun 19 11:53:00 also openssh (my package) - i would change it if i would know how to add an user with uci Jun 19 11:56:49 mirko: no, it's not intended Jun 19 11:57:10 tripolar: i think it may be better to just throw everything into the default passwd/group Jun 19 11:57:17 KanjiMonster: ping Jun 19 11:57:33 Hauke: pong Jun 19 11:57:38 since it's not many lines and the overhead and complexity of dealing with this directly is probably more than just putting it in the default file Jun 19 11:57:56 KanjiMonster: you had a patch to add support for serial flash somewhere? Jun 19 11:58:02 I lost the url Jun 19 11:58:10 nbd throw everything the default passwd group? Jun 19 11:58:16 the/in the Jun 19 11:59:12 nbd the problem i see is, that if someone adds tor to add it to the image, it wont work, as the passwd/group file is overwritten when the packages are installed, because base-files is installed as the last package Jun 19 11:59:13 Hauke: http://inuyasha.ath.cx/~jonas/0001-SSB-Add-Serial-Flash-support.patch - it probably doesn't compile, and still has some uglyness (I died during clean up when it became clear that adding brcma support to bcm47xx will take quite a while) Jun 19 11:59:42 eh, -died+stopped working Jun 19 12:01:10 Hauke: but I changed most of the code to match the kernel style more (or at least match ssb's) Jun 19 12:01:25 KanjiMonster: thank you I just want to hack that into my other patches to get my device boot Jun 19 12:02:18 Hauke: if you are using it with 3.0, remember to remove any #ifdefs for CONFIG_MTD_PARTITIONS ;) Jun 19 12:02:31 ok Jun 19 12:02:52 that's now always enabled, so there's config symbol for it anymore Jun 19 12:03:59 also you need to access the init_mtd_partitions from the bcm47xx-flashmap.c (or whatever it's called); for now it's easiest to throw a extern bla line in there Jun 19 12:07:15 KanjiMonster: then I will try to boot my device completly with taht patch Jun 19 12:07:22 nbd: ok, so are you looking into it or shall i? Jun 19 12:42:22 hey KanjiMonster Jun 19 12:42:28 you use ZNC yourself right? Jun 19 12:43:39 Olipro: yes Jun 19 12:43:47 #9577, #9578, more trac spam Jun 19 12:43:58 KanjiMonster: have you had any memleak issues with it lately? Jun 19 12:44:07 nope Jun 19 12:44:12 well, I don't think so Jun 19 12:44:24 I have, I think it's a module causing it Jun 19 12:44:39 possibly route_replies Jun 19 12:45:22 okay Jun 19 13:19:38 juhosg * r27221 /trunk/target/linux/ar71xx/files/arch/mips/ar71xx/devices.c: ar71xx: fix uart clock frequency for AR933X Jun 19 13:19:40 juhosg * r27222 /trunk/target/linux/ar71xx/files/ (2 files in 2 dirs): ar71xx: cleanup AR933X UART driver Jun 19 13:47:54 updated openwrt/upstream, https://home.comcast.net/~sdwalker/uscan/index.html Jun 19 14:39:05 swalker: any legend regarding the ticks and crosses? :) Jun 19 14:40:17 swalker: ah, enabling javascript helps :) Jun 19 14:40:21 mirko: hovering on rows show title attrs Jun 19 14:40:47 swalker: got it, thanks Jun 19 14:49:44 mirko: moo Jun 19 15:08:20 ipv6 is not working with 2.6.39,ar71xx target. Jun 19 15:35:45 gzanan: known problem with .38+, use http://patchwork.openwrt.org/patch/992/ Jun 19 16:24:33 Chocky: meh Jun 19 16:26:14 I updated that bug report. Just realized the patch refers to 2.14 as well. oh well Jun 19 16:45:08 Chocky: will look at it later Jun 19 16:46:49 \o/ Jun 19 17:11:21 swalker * r27223 /packages/utils/flashrom/ (Makefile patches/001-msr_dev_path.patch): [packages] flashrom: update source url, use PKG_INSTALL & MAKE_FLAGS, refresh patches Jun 19 17:18:15 swalker * r27224 /packages/utils/cpusage/Makefile: [packages] cpusage: update url Jun 19 17:26:34 swalker * r27225 /packages/libs/opencdk/Makefile: [packages] opencdk: update urls Jun 19 18:17:03 jow * r27226 /trunk/package/block-mount/ (Makefile files/block.sh files/fstab.init): [package] block-mount: remove explicit lock waits, solves some potential race conditions Jun 19 18:19:33 jow * r27227 /branches/backfire/package/block-mount/ (Makefile files/block.sh files/fstab.init): [backfire] merge r27226 Jun 19 19:46:33 <_trine> larsc, I have done some tests and crypto_hash compiles ok and it only fails when you try to compile kmod-crypto-mv-cesa at the same time Jun 19 20:09:16 florian * r27228 /trunk/target/linux/au1000/ (au1500/config-2.6.37 au1550/config-2.6.37 patches-2.6.37/): [au1000] remove 2.6.37 patches and config Jun 19 20:10:50 florian * r27229 /trunk/target/linux/ep93xx/ (8 files in 5 dirs): [ep93xx] remove pre 2.6.39 patches and config Jun 19 20:11:30 florian * r27230 /trunk/target/linux/cobalt/ (patches-2.6.32/ patches-2.6.35/): [cobalt] remove 2.6.35 and 2.6.32 support Jun 19 20:12:25 florian * r27231 /trunk/target/linux/ar7/files-2.6.32/: [ar7] cleanup empty directory Jun 19 20:14:35 florian * r27232 /trunk/target/linux/brcm63xx/ (26 files in 3 dirs): [brcm63xx] switch to 2.6.39 Jun 19 20:16:13 florian * r27233 /trunk/target/linux/brcm63xx/ (config-2.6.37 config-2.6.38 patches-2.6.37/ patches-2.6.38/): [brcm63xx] remove support for 2.6.37 and 2.6.38 Jun 19 20:18:49 florian * r27234 /trunk/target/linux/brcm63xx/patches-2.6.39/ (3 files): [brcm63xx] remove accidental bcm6345 related patches Jun 19 20:30:28 florian * r27235 /trunk/target/linux/brcm63xx/patches-2.6.39/ (17 files): [brcm63xx] fixup patches after r27234 Jun 19 20:47:50 florian * r27236 /trunk/target/linux/brcm63xx/ (4 files in 2 dirs): [brcm63xx] dev-spi.c checks for running CPU, not board code Jun 19 21:05:43 _trine: ok, i'll take a look Jun 19 21:08:24 _trine: could you paste a diff between your kernel config with and without mv-cesa? Jun 19 21:15:32 <_trine> larsc, would you mind if I did that tomorrow as its not far off my bedtime right now Jun 19 21:17:51 not at all Jun 19 21:18:34 <_trine> ok thanks,, I should be around in the morning Jun 19 22:50:46 jow * r27237 /trunk/target/linux/generic/ (3 files in 3 dirs): (log message trimmed) Jun 19 22:50:46 [PATCH] generic: Remove IPv6 depependency of bridge in 2.6.38+ Jun 19 22:50:46 Since 2.6.38 the bridge module has a dependency to IPv6 if IPv6 is Jun 19 22:50:46 enabled. Since the IPv6 module isn't exactly lightweight and bridge also Jun 19 22:50:46 only needs a single function from IPv6, it's rather easy to create a Jun 19 22:50:47 common "lib" module with a RCU pointer to the actual implementation, if Jun 19 22:50:48 the IPv6 module is loaded (although slightly hackish). Jun 19 22:54:10 jow * r27238 /packages/net/rsync/Makefile: Jun 19 22:54:10 [PATCH v2] rsyncd: Add description and conffiles entry Jun 19 22:54:10 Add description and conffiles entry for rsyncd to rsync's Makefile. Jun 19 22:54:10 Signed-off-by: Ian Leonard Jun 19 23:02:15 jow * r27239 /packages/utils/sane-backends/Makefile: Jun 19 23:02:15 [PATCH] No LaTeX documentation for sane-backend package Jun 19 23:02:15 If no LaTeX tools are present, building sane-backend package fails. Jun 19 23:02:15 This patch removes the building of the LaTeX documentation (it does not Jun 19 23:02:15 seem sensible/required for routers) Jun 19 23:02:16 Signed-off-by: Maarten Bezemer Jun 19 23:04:47 jow * r27240 /packages/net/ (nagios/ nagios-plugins/ nrpe/ nsca/): [packages] remove empty leftovers of nagios package move Jun 19 23:07:12 jow * r27241 /trunk/target/linux/x86/base-files.mk: [x86] Restrict Soekris specific base-files handling to Soekris profiles, based on patch by Philip Prindeville Jun 20 00:37:23 KanjiMonster: Did you have a time to see what happen to my 3.0 booting? Jun 20 00:39:17 http://pastebin.com/kR6DRm4X **** ENDING LOGGING AT Mon Jun 20 02:59:57 2011