**** BEGIN LOGGING AT Tue Jan 26 02:59:56 2010 Jan 26 03:00:09 im searching Jan 26 03:00:38 https://dev.openwrt.org/ticket/4963 Jan 26 03:00:44 florian closed it Jan 26 03:00:48 and sayed wont fix Jan 26 03:01:02 Please fix the user-space utilities to be able to make use of nfs4. Jan 26 03:01:56 nbd also a patch included Jan 26 03:01:58 :) Jan 26 03:03:00 nbd will u remove it? Jan 26 03:03:14 yes Jan 26 03:03:59 i'll remove nfs v3 and v4 Jan 26 03:04:07 this should not be in the kernel image by default Jan 26 03:04:07 why both? Jan 26 03:04:14 only a minority of users need this Jan 26 03:04:21 and they should be able to install kmod-* packages for it Jan 26 03:04:29 yeah thats okay Jan 26 03:04:43 but dont delete it completely please Jan 26 03:05:13 nfsdv3 is a module by the way Jan 26 03:05:32 no Jan 26 03:05:33 CONFIG_NFSD_V3=y Jan 26 03:05:48 also CONFIG_NFS_V3=y Jan 26 03:06:00 cshore: annnd it still doesn't work Jan 26 03:06:13 rtz2: did it say anything? Jan 26 03:06:44 nope Jan 26 03:06:55 rtz: do you have anything in addition to what I did? Jan 26 03:07:04 nbd upadting ocf would also be a nice idea .... https://dev.openwrt.org/ticket/6143 Jan 26 03:07:30 i use this for about 4 months now .. without any problems ... Jan 26 03:07:51 ah, seems i misread the kernel config Jan 26 03:08:32 ok, i think i won't change it for now until i've added support for changing the kconfig options dynamically Jan 26 03:08:54 ndb but removing nfsdv4 can be done with any damage Jan 26 03:08:59 rtz2: I can't see anything in 05_ether_mac_rdc that should be a problem Jan 26 03:09:01 it doesnt even work Jan 26 03:09:11 rtz2: it's basically cut 'n' paste Jan 26 03:09:24 with/without Jan 26 03:10:17 nbd so nfsdv4 will stay too? Jan 26 03:10:32 i'll talk to florian first Jan 26 03:10:40 and deal with the configurability when i have time Jan 26 03:10:55 okay Jan 26 03:11:05 so its really late have .. have a nice night Jan 26 03:11:07 bye Jan 26 03:11:13 bye Jan 26 03:11:19 cshore: forgot an s from gpio_buttons Jan 26 03:11:20 puchu: bye Jan 26 03:11:30 rtz2: I'm not the only one! Jan 26 03:35:40 cshore: still here? Jan 26 03:35:48 rtz: I'm here Jan 26 03:36:40 cshore: http://openwrt.pastebin.com/m68a4fa4 Jan 26 03:38:19 cshore: ahh, crap Jan 26 03:38:27 rtz? Jan 26 03:38:58 cshore: I know, what creates the insmod: can't insert 'input-core': File exists errors Jan 26 03:39:11 but I have no idea, why it failed to mount the rootfs Jan 26 03:39:18 in rw mode Jan 26 03:39:37 it's not behaving right at all Jan 26 03:40:01 during mount it should be erasing the filesystem until firstboot Jan 26 03:40:08 shouldn't be Jan 26 03:40:28 or did you sysupgrade...but no then it should be using /tmp/root Jan 26 03:41:00 cshore: no sysupgrade Jan 26 03:41:03 shouldn't again Jan 26 03:41:08 hmm.... Jan 26 03:41:36 let me build here and see if the patch that went in got screwed up somehow Jan 26 03:44:50 it looks to me like the mount stuff isn't working but I haven't touched that, nor does your stuff Jan 26 03:49:00 ok, building Jan 26 03:50:30 if have no idea what could go wrong there :/ Jan 26 03:58:01 cshore: it's in check_for_mtd Jan 26 03:58:10 root@(none):/lib/preinit# mount -o remount,rw /dev/root / Jan 26 03:58:11 mount: you must specify the filesystem type Jan 26 03:59:12 cshore: not sure, why this didn't turn up before Jan 26 03:59:26 rtz: it was there before...I just copied that Jan 26 04:00:32 rtz: It shouldn't be trying that mount at all Jan 26 04:01:08 well Jan 26 04:01:10 it does Jan 26 04:01:22 rtz: once it's got /tmp/root it *should* skip the rest of the mount stuff until firstboot Jan 26 04:01:29 so somethings borked Jan 26 04:01:41 two times actually Jan 26 04:03:06 I mean, it could be another mount command, but I don't see which one Jan 26 04:05:13 can you see anything wrong it 05_mount_skip or 40_mount_jffs2? Jan 26 04:05:25 or in the build dir, leftover files? Jan 26 04:06:01 rtz: I got this kind of stuff because target/root doesn't get cleared out, when I was doing devel Jan 26 04:07:04 cshore: I tried to check for leftover files, but didn't find anything Jan 26 04:07:49 it'll be another 10-15 minutes for my build Jan 26 04:08:13 to confirm it here or for it be just with rdc, or the build tree you've got Jan 26 04:08:39 rtz: what kernel are you using? Jan 26 04:10:52 I had no idea. Jan 26 04:11:02 cshore 30 Jan 26 04:11:06 2.6.30 Jan 26 04:11:20 rtz: ok same here...besides it shouldn't matter Jan 26 04:11:57 THIS would be a much better place to gripe about my NOR flash not being detected and to ask if there's any helpful pointers on where to begin... https://dev.openwrt.org/ticket/6552 Jan 26 04:12:40 cshore: this is another reason, why I would like to switch to a binary init Jan 26 04:12:50 the interactions are much easier to understand :/ Jan 26 04:13:16 rtz: it looks to me like it's executing everything twice Jan 26 04:14:17 MJEvans: 1. your bootlog is completly fucked up Jan 26 04:14:33 MJEvans: 2. try to riddle the mtd code if lots and lots of printks Jan 26 04:15:22 cshore: hmmm, yes Jan 26 04:15:28 build_dir/target-rdc/root-blah/lib/preinit.....what does it have? Jan 26 04:15:49 cshore: for the rdc stuff, that's because of backup files from the editor Jan 26 04:15:55 rtz: I know, and I know... I was hoping for advice on WHERE to plant printk's in relation to these lines... SSFDC read-only Flash Translation layer physmap platform flash device: 00800000 at 10000000 physmap-flash.0: Found 1 x16 devices at 0x0 in 16-bit bank Jan 26 04:15:57 but it's only those Jan 26 04:16:00 Amd/Fujitsu Extended Query Table at 0x0040 Jan 26 04:16:02 Unknown Amd/Fujitsu Extended Query version 0.0. Jan 26 04:16:10 MJEvans: everywhere ;) Jan 26 04:16:23 MJEvans: your patch probably didn't work because the flash is detected through CFI, not JEDEC Jan 26 04:16:40 cshore: http://openwrt.pastebin.com/m3c5a0241 Jan 26 04:17:13 nbd: See, THAT's helpful, knowing it's in CFI and not the obvious 'huge' list of chips. Jan 26 04:17:34 (which gives me a better place to go, add printks and other things.) Jan 26 04:18:12 Thank you, I'll do that next Jan 26 04:18:54 oohhh, the build is onto install stage Jan 26 04:21:01 rtz: I don't see anything obvious in the build_dir (other than what you already noticed) Jan 26 04:22:25 rtz: what does /etc/preinit look like? Jan 26 04:23:39 cshore: http://openwrt.pastebin.com/m73acba4d Jan 26 04:24:00 cshore: I also have a /etc/preinit~ but I don't see, how this could be a problem Jan 26 04:24:18 rtz: depend on how init calls it Jan 26 04:25:47 cshore: the kernel has a hardcoded call to /etc/preinit Jan 26 04:25:58 rtz: x86 works here Jan 26 04:26:01 also, it's the old preinit Jan 26 04:26:14 rtz: unlikely that's the problem then Jan 26 04:26:14 I would notice, if this got executed Jan 26 04:27:04 could you pastebin your gpio button files? Jan 26 04:27:47 cshore: http://openwrt.pastebin.com/m6b30f4ec Jan 26 04:29:54 rtz: hmph. I'm mystified....it works here on the x86 build I just did Jan 26 04:30:18 rtz: and I don't see anything the rdc specific stuff that should be causing double execution Jan 26 04:33:03 rtz: is this a clean build? Jan 26 04:34:41 no Jan 26 04:34:50 cshore: I will try one tomorrw Jan 26 04:34:53 o Jan 26 04:34:58 have to got to bed now Jan 26 04:35:00 good night Jan 26 04:35:04 rtz: good night Jan 26 04:36:08 * cshore opines that he really wishes he had one of every board; with a serial console Jan 26 04:40:01 http://pastebin.ca/1766128 "#" Jan 26 04:40:02 If I brick it... commenting out the unknown chip is what's going to do it. Jan 26 04:40:54 cshore: serial console IS useful; jtag would be more so. Jan 26 04:41:31 cshore: serial console IS useful; jtag would be more so. Jan 26 04:42:05 MJEvans: for what I do, I don't need the jtag so much Jan 26 04:42:29 MJEvans: I'd have to bork the boot loader to need that Jan 26 04:45:01 Can't you also use jtag to attach a debugger and step instruction by instruction? Jan 26 04:48:12 sure Jan 26 04:48:19 but to know where to start debugging, you need the serial console ;) Jan 26 04:49:12 With jtag the only brick you can't recover is a full hardware fry Jan 26 04:49:17 cshore: have an idea Jan 26 04:49:38 MJEvans: debugging shell scripts by stepping cpu instructions is kinda tiresome Jan 26 04:53:03 That does sound tiring Jan 26 04:53:18 rtz: yes? Jan 26 04:54:48 MJEvans: I think jtag is nice to have but not as essential as a serial console Jan 26 04:56:28 cshore: no, sorry Jan 26 04:56:36 but I really have to go now Jan 26 04:56:54 rtz: okay...I'll talk to you tomorrow Jan 26 04:58:39 MJEvans: I've only needed it two times: once when I set the wrong board id (which really isn't OpenWRT's fault) in the CFE config, and once when I flashed a image to big for the flash via CFE...and that board had no jtag header, so it was hosed Jan 26 05:02:33 Kaloz: did you get my email about setting up a trac account? Jan 26 06:25:34 various things are not finding openssl (during build)....I know this happens on occasion; what's the usual fix? Jan 26 06:31:54 hmm....ok it's the package/hostapd that's the problem, but it's rather 'interesting' to parse Jan 26 06:32:11 package/hostapd/Makefile Jan 26 06:32:24 for wpad-mini and wpasupplicant Jan 26 06:32:48 or for old hostapd Jan 26 06:33:45 it looks like the package definition is missing the openssl depends Jan 26 06:33:54 (conditional depends) Jan 26 07:34:42 hi Jan 26 07:34:45 ping cshore Jan 26 07:34:56 pong puchu Jan 26 07:35:00 it doesnt work Jan 26 07:35:13 i also found out that is seems there is no preinit.arch Jan 26 07:35:16 Frigga frizza git Jan 26 07:35:24 puchu: there's not supposed to be Jan 26 07:35:29 puchu: it's replaced Jan 26 07:35:53 puchu: target/linux/brcm47xx/base-files/lib/preinit Jan 26 07:36:04 hmm because i have this in preinit fs_failsafe_ifname= Jan 26 07:36:10 pi_ifname= Jan 26 07:36:17 both are empty in the image Jan 26 07:36:25 is this normal? Jan 26 07:36:39 /etc/preinit sets them empty as defaults Jan 26 07:36:48 they only apply to the preinit messages Jan 26 07:36:54 some archs set them Jan 26 07:37:00 hmm Jan 26 07:37:01 or you can configure them in menuconfig Jan 26 07:37:10 so no other error found? Jan 26 07:37:27 not by me...and this was a fresh build, right? Jan 26 07:37:32 yeah Jan 26 07:37:36 make dirclean Jan 26 07:37:37 make Jan 26 07:38:08 some managed to boot a image on brcm47xx? Jan 26 07:38:23 I don't know....I haven't heard Jan 26 07:39:06 so maybe other bugs are still there Jan 26 07:39:20 possibly Jan 26 07:39:49 but it works on x86 and brcm63xx Jan 26 07:40:01 so it's definitely arch-specific Jan 26 07:40:20 yeah but this also worked yesterday Jan 26 07:41:14 I really wish you had a serial console....but oh well...let me see if I can find anything Jan 26 07:41:26 cshore me too Jan 26 07:41:27 :) Jan 26 07:41:57 cshore but the 2 lines yesterday where for shure bugs Jan 26 07:42:06 yes Jan 26 07:42:22 I have a wgt634u with serial attached and will try to boot an image this evening Jan 26 07:42:36 loswillios when is evening? Jan 26 07:42:41 losillios: thanks! Jan 26 07:42:42 t+? hours? Jan 26 07:42:57 t=now Jan 26 07:43:21 loswillios: if you could turn off stderr suppression for the build that would help a lot Jan 26 07:43:39 menuconfig | Image Options | Preinit Jan 26 07:52:15 puchu: found another bug Jan 26 07:53:10 in 15_set_preinit_interface_brcm, in port_net_echo, it should be 'fi' after netmsg, not '}' Jan 26 07:57:55 http://openwrt.pastebin.com/m966e2ef Jan 26 07:58:33 puchu: it's currently 9am Jan 26 07:58:46 loswillios: okay Jan 26 07:59:20 cshore: maybe we should really wait for someone to test it with a device with a serial attached Jan 26 07:59:27 loswillios: where are u from? Jan 26 07:59:32 its 9am here too Jan 26 07:59:34 gmt+1 Jan 26 08:00:01 Tue Jan 26 00:00:00 PST 2010 Jan 26 08:00:05 germany Jan 26 08:00:14 puchu: it's up to you...I'm willing to keep hunting bugs, but it you who has to do the builds Jan 26 08:00:45 netmsg $pi_broadcast "$1" Jan 26 08:00:46 - } Jan 26 08:00:46 + fi Jan 26 08:00:58 puchu: and the bug I just found was a showstopper Jan 26 08:01:11 puchu: it wouldn't even load with that Jan 26 08:01:18 yeah, that one Jan 26 08:01:25 last build last try .. then i have to go ... maybe i can search for other bugs at work ... not maybe .. i can search for other bugs at work Jan 26 08:01:27 :) Jan 26 08:01:37 found antoher bug? Jan 26 08:01:59 just that one...I pastebinned the patched Jan 26 08:02:49 i should be enough when i do a make package/base-files/clean ; make ? Jan 26 08:03:29 i/it Jan 26 08:03:51 I discovered it with a silly trick; for file in *; do . $file; done (in my shell, while in the target...blah...preinit dir) Jan 26 08:04:12 puchu: you should probably make target/linux/clean too Jan 26 08:05:03 puchu: but if want to try without it; if you've got the time for a quick build; you could do that; it'd certainly be faster Jan 26 08:05:15 cshore i do another build in the evening .. maybe someone finds some other bugs today ... i think (but i dont hope so) that other not tested arches will have a few bugs too Jan 26 08:05:39 puchu: probably; I seem to be typo king Jan 26 08:06:03 cshore no you are not a typo king ... this is normal ... u are not a computer and mistakes/typos happen Jan 26 08:06:15 new init still rocks ... :) Jan 26 08:06:21 puchu: thanks :) Jan 26 08:06:30 with a serial console the few bugs would be history by now Jan 26 08:06:37 i bet Jan 26 08:07:02 in the meantime i ll try to fix some audacious bugs :) Jan 26 08:07:18 mediaplayer not related to openwrt :) Jan 26 08:07:19 puchu: is that what you work on at work? Jan 26 08:07:23 no Jan 26 08:07:27 also a hobby Jan 26 08:07:29 puchu: I've heard of it Jan 26 08:07:49 a xmms gtk2 rewrite Jan 26 08:07:56 puchu: used to be called something else ... yeah that's the one Jan 26 08:08:36 if u wanna test my mplayer plugin port ... i can give u the github adress Jan 26 08:09:23 the people at audacious dont wanna include it in their default image ... they mean playing videos with audacious is not a good idea .. beacuse its a audaio player ... but i loved this feature at xmms Jan 26 08:09:34 puchu: actually I use mpd, though I'm thinking of switching to rhythmbox and having my NAS be a media server instead of mpd server Jan 26 08:09:45 okay Jan 26 08:10:11 i dont like the client server thing when i comes to audio players Jan 26 08:10:30 small footprint, short startup time is all i want :) Jan 26 08:10:41 puchu: I did it when I had a separate media computer and wanted to control it remotely Jan 26 08:10:57 puchu: but I decided it was a waste to have the media computer Jan 26 08:11:12 i think audacious is also controllable by cmd line Jan 26 08:11:13 puchu: I only listen to music Jan 26 08:11:45 puchu: I like the music management vs music player thing Jan 26 08:11:58 no i can controll audacious with my cordless keyboard ... so i can sit at the couch and watch movies on my tv ... without any need to stand up to get to the next video :) Jan 26 08:12:39 puchu: heh....I don't watch videos these days...too busy mainlining OpenWRT Jan 26 08:12:45 :) Jan 26 08:13:06 i also dont ... but when i have time ... its a nice feature Jan 26 08:13:55 puchu: I get annoyed with movies and that because so much is a waste of time Jan 26 08:13:56 i just love this thing about linux ... u can "rape" it and do things that would take much more time on other os's Jan 26 08:14:18 cshore: yeah .. most movies nowadays are a waste of time Jan 26 08:14:22 thats true Jan 26 08:15:18 watching about 99,999% of all movies is just a dump way of burning time Jan 26 08:16:04 but music videos or stuff like this is nice ... and its more like ... i wanna program and involve ... so its a good way to do something and learn Jan 26 08:17:24 puchu: I understand....I haven't watched music videos lately, so I don't know if they're any good these days....never really got into them Jan 26 08:18:06 cshore yeah ... i also didnt watch them lately ... but as i said ... i just wanna program and i was missing this plugin in audacious Jan 26 08:18:25 puchu: an excuse...works for me :) Jan 26 08:18:35 :) Jan 26 08:18:49 puchu: besides some entertainment is good Jan 26 08:19:01 puchu: a little popcorn for the brain Jan 26 08:19:43 puchu: I probably ought to just turn off my brain once in a while and watch something meaningless but fun Jan 26 08:19:50 yeah a little bit ... but dont fill your had up with this stuff ... heard the dump movies and television show ... are one cause that people are getting dumper since 1999 Jan 26 08:20:25 dumb you mean? Jan 26 08:20:30 the overall iq has lowerd since 1999 Jan 26 08:20:38 puchu: heh Jan 26 08:21:25 people are getting sillier Jan 26 08:22:16 Since the twentieth century, IQ scores have increased at an average rate of around three IQ points per decade in most parts of the world. Jan 26 08:24:32 wrong text Jan 26 08:26:25 cshore: a read a german text from a docotor ... he means that the average iq is decreasing Jan 26 08:27:01 interesting Jan 26 08:27:52 and dumb media shows are one cause of it Jan 26 08:28:34 and when i watch tv and see things like big brother and other dumb things like simpsons or spong bob .... i think ... he could be right Jan 26 08:30:14 but this only happens in western civilisation ... Jan 26 08:32:37 okay but i'll leave now .. Jan 26 08:32:42 talk to you later Jan 26 08:33:04 puchu: ttyl Jan 26 08:33:04 Depressing is more like it Jan 26 08:33:07 Later Jan 26 09:57:31 juhosg * r19340 /trunk/target/linux/ar71xx/ (3 files in 3 dirs): ar71xx: fix 'section type conflict' build error Jan 26 09:57:33 juhosg * r19341 /trunk/target/linux/ramips/rt305x/config-2.6.30: ramips: sync kernel config Jan 26 09:57:36 juhosg * r19342 /trunk/target/linux/ramips/patches-2.6.30/ (102-rt288x-pci-driver-hook.patch 103-ethernet.patch): ramips: refresh patches Jan 26 09:57:39 juhosg * r19343 /trunk/target/linux/ramips/ (10 files in 3 dirs): ramips: add preliminary support for 2.6.32 Jan 26 10:45:29 jow * r19344 /trunk/target/linux/brcm-2.4/base-files/lib/preinit/15_set_preinit_interface_brcm: [brcm-2.4] [brcm47xx] fix syntax error in preinit (#6575) Jan 26 11:19:54 <{Nico}> /win 23 Jan 26 11:32:45 xMff: collectd-mod-iptables doesn't like the libiptc split Jan 26 11:41:36 cshore: hi Jan 26 11:41:45 did someone test the new patch? Jan 26 11:41:55 does init now for brcm47xx? Jan 26 11:42:03 does init now work for brcm47xx? Jan 26 12:19:06 not yet Jan 26 12:46:07 kaloz * r19345 /trunk/toolchain/gcc/patches/4.4.1+cs/000-codesourcery_2009q3_68.patch: cleanup patch Jan 26 13:50:48 swalker: will look later Jan 26 14:32:08 ping cshore Jan 26 14:41:28 juhosg * r19346 /trunk/package/base-files/Makefile: Jan 26 14:41:28 base-files: don't build '00_preinit.conf' if CONFIG_PREINITOPT is not selected Jan 26 14:41:28 Also, use the default pi_broadcast address if it is not specified Jan 26 14:41:30 juhosg * r19347 /trunk/package/kernel/modules/usb.mk: package: kmod-usb-core depends on kmod-nls-base on 2.6.32 Jan 26 15:04:34 heee, I have the source code for my infineon router Jan 26 15:04:58 not /particularly/ interesting, all told Jan 26 15:09:22 Anybody who had problems with ramdisk images? Jan 26 15:11:12 glp: dolphin did but he is not here Jan 26 15:13:01 blogic: can you remember on what platform? Jan 26 15:13:57 ramips Jan 26 15:15:08 I haev them on ar71xx Jan 26 15:15:17 nbd: ? Jan 26 15:15:18 :) Jan 26 18:03:25 cshore: http://www.pastebin.ca/1766816 Jan 26 18:04:13 latest trunk with 2.6.32.5 (don't bother about the wlan oops) Jan 26 18:13:09 maybe we need an ifconfig eth0 up before the switch is initialized Jan 26 18:41:28 kaloz * r19348 /trunk/toolchain/gcc/ (6 files in 2 dirs): add another gcc version based on 4.4.3 plus the 2009q3 codesourcery patch Jan 26 19:31:50 hi Jan 26 19:31:52 cshore: nice it works now Jan 26 19:31:56 on brcm47xx Jan 26 19:32:31 but the shadow creation is wrong ... i always end up when i loging with ssh with ... password expired .. please enter new password Jan 26 19:33:48 cshore: whats the best way to get something started befor 99_10_run_init Jan 26 19:33:58 id like to use my usb disk Jan 26 20:13:08 nbd * r19349 /trunk/toolchain/gcc/patches/4.4.3+cs/000-codesourcery_2009q3_68.patch: gcc: fix up broken chunks of the 4.4.3+cs patch and clean it up a bit more Jan 26 20:19:30 nbd * r19350 /trunk/toolchain/gcc/Config.version: add missing +cs version suffix to the 4.4.3 codesourcery toolchain Jan 26 20:25:04 puchu: http://www.pastebin.ca/1766816 Jan 26 20:29:28 loswillios i know it works for me too Jan 26 20:29:33 really nice Jan 26 20:34:39 loswillios but this was not the image you created yesterday Jan 26 20:35:24 ? Jan 26 20:36:00 loswillios ohh a oops Jan 26 20:36:05 just saw it Jan 26 20:48:31 loswillios: i think its this vfs error Jan 26 20:48:41 i dont know why this happens Jan 26 20:49:37 i tried to debug it ... but the router didnt come up with some printks in this kernel function Jan 26 20:49:50 it seems its called too often Jan 26 20:51:42 ping cshore Jan 26 21:11:42 loswillios: try to load the module in /etc/init.d/boot befor hotplug2 is loaded .. when i load the module there ... i have no oops ... dont excatly know why it works Jan 26 21:21:00 puchu: I'm not available for a couple of hours yet...got to get groceries and walk the dog (not in that order) Jan 26 21:21:19 okay Jan 26 21:38:13 <[Fate]> nbd: ping Jan 26 21:39:10 pong Jan 26 21:39:51 <[Fate]> one question about dependencies. the documentation basically says to use "bla" for dependencies within the same makefile, and "+bla" for dependencies in other packages' makefile. is that still correct? Jan 26 21:40:15 <[Fate]> because you could also read it this way: Jan 26 21:40:31 <[Fate]> "bla" means "bla" must have been selected beforehand, "+bla" will select "bla" Jan 26 21:41:59 correct Jan 26 21:42:26 <[Fate]> the first or the second? :) Jan 26 21:42:31 the second Jan 26 21:43:12 <[Fate]> ah ok. then I'd suggest a patch that I'd like to implement, but I wanted to talk about it first Jan 26 21:44:01 <[Fate]> right now, if "foo1" and "foo2" depend on "bar", then "foo1" and "foo2" will be visible only when "bar" is selected Jan 26 21:44:10 <[Fate]> except when "bar" is a menu (MENU=1) Jan 26 21:44:28 <[Fate]> then the menu is selectable only when "bar" is selected, but the user does not get confused by new options suddenly appearing Jan 26 21:44:47 <[Fate]> this (the grouping in a submenu), however, only works for seperate packages in some cases Jan 26 21:45:05 users are strange life forms, they get confused every time and they're scared by everything :D Jan 26 21:45:09 <[Fate]> namely when "foo1" would be named "bas" or the like, so that it immediately follows "bar" in alphasorting Jan 26 21:45:33 <[Fate]> that's why I'd like to implement a patch that does the following when generating /tmp/.config-package.in: Jan 26 21:46:19 <[Fate]> iterate over all pkgs in a category. if (and only if) a package x has a single dependency y, and y has MENU=1, move x behind y so that it will appear in the submenu reliably Jan 26 21:46:28 <[Fate]> I hope this was understandable :) Jan 26 21:55:00 <[Fate]> ok let me just implement it and we'll see Jan 26 22:42:03 puchu: no, it's current trunk. don't mind the oops, it's a bug in p54-driver. it's the roboswitch, see line 158 in the pastebin. does it work for you? Jan 26 22:42:46 currently running "make clean world" to be sure Jan 26 22:47:04 loswillios: canu repaste the pastebin Jan 26 22:47:11 my mouse is broken Jan 26 22:47:17 :( Jan 26 22:47:41 loswillios: but read my mailing ist mail ... Jan 26 22:47:54 http://www.pastebin.ca/1766816 Jan 26 22:48:28 that hotplug2 crpytodev error is still there when the module is loaded befor hotplug2 gets reinitialiced in init.d/boot Jan 26 22:48:37 and im booting from a usb disk Jan 26 22:48:47 when i load it befor .. everthing is nice Jan 26 22:48:51 its really weired Jan 26 22:51:48 loswillios: u mena this? Jan 26 22:51:51 http://nopaste.info/5d8e7b6c26.html Jan 26 22:51:54 mean Jan 26 22:52:10 yes Jan 26 22:52:31 but it works Jan 26 22:52:45 http://nopaste.info/5976c67076.html Jan 26 22:52:52 here is the complete dmesg output Jan 26 22:52:54 no error Jan 26 22:53:08 but i load the cryptodev modul befor hotplug2 Jan 26 22:53:17 is reloaded in init.d/boot Jan 26 22:53:45 when its loaded normally .. the oops is still there Jan 26 22:53:48 interesting Jan 26 22:53:54 i already use 2.6.32.6 btw Jan 26 22:57:51 loswillios the same i mentioned on the mailing list Jan 26 22:58:02 but i couldnt track it down Jan 26 22:58:22 weird.. I have no idea Jan 26 22:58:33 nbd ment when he has time he will look at it Jan 26 22:58:50 i allready use 2.6.32 .. a few days without an error Jan 26 22:59:11 except this load it befor hotplug2 ... which is really shitty Jan 26 23:00:04 but i think nearly no one will every see it .. because u have to modify some parts of the build_sys (openssl for example) to get the progs to use it Jan 26 23:00:20 and so no one will enable it :) Jan 26 23:00:26 expect of me :) Jan 26 23:00:37 where are the patches for that? Jan 26 23:00:47 which patches? Jan 26 23:00:52 for enabling it Jan 26 23:01:01 on my pc Jan 26 23:01:07 heh Jan 26 23:01:07 haven't been posted yet? Jan 26 23:01:07 its just ubsec_ssb Jan 26 23:01:22 and an patched openssl Jan 26 23:01:36 ubsec_ssb is allready in trunk Jan 26 23:01:39 for a long time Jan 26 23:02:08 else Jan 26 23:02:08 OPENSSL_OPTIONS:= shared no-ec no-err no-hw no-threads zlib-dynamic \ Jan 26 23:02:08 - no-engines no-sse2 Jan 26 23:02:08 + no-sse2 --with-cryptodev Jan 26 23:02:19 i just enable it by default Jan 26 23:02:39 i also requsted this some time ago Jan 26 23:02:55 nbd: like the remove nfsdv4 thing ;) Jan 26 23:02:57 nah, enabling it by default is a bad idea Jan 26 23:03:02 it should be made configurable Jan 26 23:03:05 i know .. but i know what im doing Jan 26 23:03:17 most of the time :P Jan 26 23:03:30 ifeq ($(BOARD),ixp4xx) Jan 26 23:03:30 OPENSSL_OPTIONS:= shared no-ec no-err no-hw no-threads zlib-dynamic \ Jan 26 23:03:30 no-sse2 --with-cryptodev Jan 26 23:03:30 else Jan 26 23:03:40 ixp4xx uses it by now Jan 26 23:03:41 no-engines is fine for 90% of the openssl users Jan 26 23:03:56