IRC Logs for #crux-devel Thursday, 2007-02-22

*** pitillo has joined #crux-devel00:17
*** jaeger has quit IRC01:07
*** treach has joined #crux-devel02:36
prologicoh god that's scary03:25
prologicwhy do I account for 41% of the contrib ports :/03:26
prologicdon't we have very many contrib maintainres ?03:26
treachtilman: got time to look at my first attempt ever at an official bugreport?03:41
prologichi treach wb03:44
prologicbeen sick ?03:45
treachyeah, the flu.03:46
prologicsorry to hear03:47
treachno problem, I survived. :)03:47
prologicI had chicken pox earlier this year03:48
prologicnot fun :)03:48
treachyikes. Aren't you a bit too old for that? :p03:48
prologicnever got it as a kid :/03:49
prologicmy gf gave it to me before xmas03:49
prologicand I came down with it on new year's day03:49
treachheh, some present, eh? :X03:49
prologicgreat love sharing :)03:49
*** treach has quit IRC04:23
*** treach has joined #crux-devel04:42
*** prologic has left #crux-devel04:45
*** prologic has joined #crux-devel04:45
treachtilman: I updated the bugreport, screenshots available at the same location, fvwm.with.{xaa,exa}.jpg04:47
tilmantreach: oh, list the version of xorg-xf86-video-ati, too :)04:49
tilmanit's okay otherwise04:49
treachnot yet..04:50
treachI meant I had added the ati-version. :)04:51
treachtilman: 1005905:01
*** Viper_ has joined #crux-devel05:13
*** Romster has quit IRC05:17
*** Romster has joined #crux-devel05:27
*** Viper_ has quit IRC05:44
*** aon has joined #crux-devel06:43
*** Viper_ has joined #crux-devel06:52
*** aon has joined #crux-devel07:00
*** Viper_ has quit IRC07:33
*** jaeger has joined #crux-devel07:47
*** j^2 has joined #crux-devel08:38
*** j^2 has left #crux-devel08:38
*** j^2 has joined #crux-devel08:39
j^2morning yall08:39
aonhi j^208:40
j^2feeling better aon?08:41
aonyeah, for a long time now08:42
aonmy server got somehow busted, running memcheck now08:42
aonno errors so far08:42
aonircing with gaim is horrible :)08:42
aonmorning jaeger08:43
Romster iputils fails to build but it built in a chroot don't know why...08:44
Romsterin 2.308:44
Romsteraon, eww :/08:44
Romsteryo jaeger08:44
aonRomster: yep :(08:45
jaegerthat was mentioned yesterday but I couldn't duplicate it08:45
Romsteraon, i'd rather use irssi than gaim :P08:46
Romsterbitchx is such a bitch :D08:46
Romsterjaeger, strange i can duplicate it.08:46
aonRomster: yeah, but i'm too lazy to compile it08:46
Romsteri rebuilt everything 3 times in core on that box same result08:46
Romsteraon, ah :/08:47
j^2morning jaeger08:47
Romsteri always have irssi and elinks as a backup.08:47
aonyes, well, i have irssi as primary08:47
aonand gaim as backup08:47
aonbut my irc box is fux0red now08:47
Romsterjaeger, anything i can try as this seems to be recreateable.08:47
Romsteraon, you reckon ya got bad mem08:48
jaegerfigure out why glibc-bugs.h doesn't exist08:49
aonRomster: _what_?08:49
Romsteraon, good one. I meant the irc box it crashed and corupted the hard disk from bad memory.08:51
aondunno, it booted up and crashed immediately08:52
aonit was actually running until the power chord was touched08:52
pitilloRomster, readed the error and here I got the same08:53
Romsterhrmm odd so you got that file...08:53
Romsterprt-get fsearch glibc-bugs.h, returns nothing..08:54
jaegerit should be in the iputils source08:54
Romsteroh hrmm08:54
jaegerperhaps it has something to do with non-default optimization flags08:55
jaegerthough that seems unlikely08:55
Romstertryed -O2 -mtune=pentium3 -pipe and -O2 -march=i686 -pipe, same result08:56
Romsterformer is my default for that box.08:56
*** cptn has joined #crux-devel08:57
*** ChanServ sets mode: +o cptn08:57
jaegerheyo, cptn :)08:57
Romsterfind / -name glibc-bugs.h dosn't return a thing so it has to be a local source path problem then.08:57
Romsterhi cptn08:57
jaegerI wouldn't expect to find it anywhere but in the source08:58
j^2morning cptn!08:59
jaegerhave a look at lines 13 and 17 of the Makefile08:59
pitillohi cptn08:59
cptnjaeger: your build log doesn't build any ipsec stuff08:59
cptnwhich is the part which fails for Romster08:59
jaegerperhaps because I don't have the /usr/src/linux symlink or use ipsec in the kernel09:00
jaegerhow lame09:00
Romsterhrmm seems to be ipsec related which i tink is kernel related09:02
Romsterthats running on that box i must of had that on in my desktop kernel.09:02
Romsterthe first line of the box is about libipsec09:03
Romsteryet on the chroot on my desktop it skips that part.09:03
jaeger <--- mind trying this?09:05
pitillohere built with image from cdrom and custom kernel without ipsec stuff. (created linux simlink in my case)09:05
Romsterjaeger, hmm i'll give it a try09:06
cptnjaeger: from that it's most certainly the /usr/src/linux symlink09:08
Romsteri added the symlink09:09
jaegerI hate stupid software09:09
cptnyep, that's the problem09:09
cptnit's conditionally compiled if it finds a particular header file09:09
pitilloat rc.shutdown line 30 command /usr/bin/sleep 5 -> /bin/sleep here (with sip test iso)09:11
Romstercan of whooparse on that patch heaps of conflicting types for 'dev_t'09:11
jaegerok, will look closer09:11
pitillo(do not know if it is fixed by now)09:11
jaegerRomster: try
jaegerpatch -i $SRC/$name-$version.patch Makefile09:14
Romsterthere is a later version i heard ya was saying about something needed to make the man pages..09:14
jaegeryeah, needs docbook/sgml stuff. I'll maybe build the man pages separately and package them, dunno. it's ugly but I'd like to be able to upgrade iputils09:15
Romsterb4 or after the ipg patch?09:15
jaegerdoesn't matter09:15
jaegerit works for me in my 2.3 vm (created the symlink)09:17
Romsterhmm can't download that patch odd09:19
cptnworked here09:19
cptnbut you can probably also use
Romsterprobbly gotta open port 8109:19
jaegerI suppose either would work09:20
Romsterok that worked09:21
Romsterthe one on jaeger's site09:21
Romsteralot of differ in signedness, but it compiled09:22
Romsterah your just skiping the check kernel check...09:25
jaegerthat was already skipped in the Pkgfile, anyway09:25
Romsteroh yeah in the sed09:26
Romsterok odd..09:27
Romstersed issue?09:27
jaegerdon't think so, that's coincidental09:28
Romster-e 's/^\(all:\) check-kernel/\1/' \        should of removed the kernel check but i guess it didn't09:30
Romstermaybe its a tab and not a space?09:30
cptnit removed the check for the directory09:31
cptnif that line wasn't there, the package would have failed to compile without the symlink09:31
cptnsince /usr/src/linux is deprecated, that check was disabled09:31
jaegerin other words, the check-kernel thing is not related to your problem09:31
cptnhowever the build will fail if some still have the /usr/src/linux symlink09:31
jaegerit builds both ways for me09:32
cptnwell, would fail before your patch09:32
jaegerit should work for both now, though09:32
cptnprobably it would have worked with a glibc from back then :-)09:32
jaeger(and does here)09:32
Romsternar that bit is ok i see it09:33
Romstermodules: check-kernel09:33
Romsterthink thats it :/ being called from another location09:33
jaegerI don't understand, could you elaborate?09:34
Romsteri might be dumbfounded but you seded out check-kernel from all:09:34
jaegerah, there's a modules: target in the makefile that builds check-kernel09:34
Romsterbut there is another call to check-kernel from modules09:34
jaegermodules shouldn't be called09:35
cptnbut make modules isn't called, is it?09:35
Romstershouldn't be was jsut a guess.09:36
RomsterSUBDIRS=libipsec setkey < there it is...09:41
Romstercalled from libipsec's Makefile09:41
tilmanlook who's visiting09:42
tilmanhey cptn :)09:42
cptntilman: :-)09:42
cptnRomster: didn't you understand what the problem is?09:42
Romsterpartly, i'm still learning this side of the code.09:43
Romsteri must admit this is a bit over the top for me, more guess work09:43
cptntilman: I figured I could drop in once in a while :-)09:43
Romsterbut trying to understand it.09:43
cptnwell, the original Makefile had a dependency on the 'subdirs' target09:44
tilmancptn: did you hear we're finally tying up 2.3? :D09:44
cptnRomster: if the symlink is there, the original Makefile would add libipsec to the list of subdirs09:45
cptntilman: I just saw some activities in the bug tracked09:45
cptnmissed the tuesday meeting unfortunately09:45
Romsterok i tested with and without, a symlink i think double checking that.09:46
Romstercptn, yeah your hardly about anymore :/09:47
Romsterok i'll be dammed without the /usr/src/linux symlink it builds...09:48
Romsterwhy didn't i check that earlier...09:48
Romsteryeash :/09:48
Romsterrebuilt core 3 times to make sure i hadn't had anything depending on anything else messed up.09:49
Romsterbootstraped another words.09:49
*** aon has quit IRC09:51
Romsterbut if its building libipsec it must hate this version of the kernel?09:52
jaegerRomster: if it finds that file, pfkeyv2.h, it adds those subdirs to the build09:52
cptnit hates the version of glibc probably09:52
cptnor rather expects a much older one09:52
*** aon has joined #crux-devel09:54
Romsterand the penny dropes, ka-ching09:58
Romsteraddes the subdirectorys yes that i can understand in the make file, but the not compiling bit didn't till i read that url jjpk pasted09:59
Romsternow it makes scense.09:59
aon 6:01PM  up 12 mins, 1 user, load averages: 0.06, 0.17, 0.1510:01
aon\o/, still hasn't crashed10:01
Romsterit should be seeing the kernel header files that glibc was linked with and that is the one bundeled in glibc kernel-headers-2.6.18.tar.bz210:01
tilmani guess we're good to go10:01
Romsteraon, so far so good :D10:01
Romsterso basicly my fault for doing that, linking to the kernel but not the headers that glibc was compiled with, so you shouldnt be patching a user fault :|10:02
*** j^2 has quit IRC10:03
cptnwell, it's not that easy10:03
cptnthe compilation is conditional and done purely when the kernel header exists10:04
cptnin /usr/src/linux that is10:04
cptnthe compilation itself includes a header which doesn't come with current glibcs anymore10:04
cptnso the compilation fails there10:04
jaegerhopefully a newer iputils will solve that problem but I've got to work around the documentation problem first10:05
Romsterwhich is incorrect ah so it should be pointing to usr/include10:05
cptnyeah, iputils git doesn't seem to have a libipsec subdirectory10:06
cptnyou should probably use glibc 2.1 with that iputils version10:06
cptnthat's just a guess BTW ;-)10:07
cptnbut yeah, don't add any /usr/src/linux symlink and you're all set10:07
*** j^2 has joined #crux-devel10:08
Romsterok i still don't get it...10:20
Romster-e 's|^KERNEL_INCLUDE=.*|KERNEL_INCLUDE=/usr/include|' \10:20
Romsteri added that.10:21
Romstershould point to the right kernel headers10:21
Romsterok i missed the libc is pointing to the right location so that other bit is redudent...10:21
Romsteri must be boring you lot with my blabering here :/ i'll leave it to you to fix, though that patch works i some how think bypassing isn't the right thing(tm) to do10:21
cptn/usr/include is not kernel headers10:21
cptnwell, the patch disables compilation of the libipsec subdirectory10:22
cptnwhich isn't even included anymore in the recent versions of iputils10:22
Romsterthats where the crux kernel-headers are pointing to in glibc ?10:22
cptnso it seems like it's the best thing to do, considering that some might hit a compilation error otherwise10:22
cptnalthough, not that /usr/include/linux contains the userspace headers10:24
cptnand packages relying on /usr/src/linux usually use all kinds of kernel headers10:24
cptnwhich is usually a bad idea10:24
Romstertilman, oh you are right :/ just looked the path is in the archive.10:25
Romstercptn, never knew about that problem but i learned about it now.10:30
cptntime to go home10:34
cptnhave a nice day :-)10:34
tilmanyou to :)10:35
tilmanugh, too10:35
cptnheh, thanks10:35
*** cptn has quit IRC10:35
Romsterhrmm let me know if this is a problem, my firewall is using iptables-imq thats linked to the kernel headers of but the rest of iptables is linked to the kernel headers, the new headers in /usr/include/linux don't work on the IMQ modules, thats a problem? incompatability? i'd be wise to recompile glibc with kernel-headers and the entire system again?10:36
clbUpdate from opt: 22 Feb 16:07 - smpeg: disabled gtv, removed manpage10:36
Romsterand same problem on the crux-2.2 though the kernel-headers on crux-2.2's glibc are much older, they don't match versions..10:37
Romsteractually thats a question i should be asking them, sorry about that.10:41
Romsteri should of read the faq seems its ok for that case10:44
RomsterThe iptables IMQ target shared libs are only built if your kernel tree has been patched to include the IMQ target10:44
Romsterso i guess that one is a excemption, but the other cases are not.10:45
*** Viper_ has joined #crux-devel11:17
clbUpdate from opt: 22 Feb 17:19 - mdadm: update to 2.6.1 || 22 Feb 17:18 - dropbear: update to 0.4911:38
*** sepen has quit IRC12:04
*** Viper_ has quit IRC12:20
*** treach_ has joined #crux-devel13:06
*** treach has quit IRC13:07
*** treach_ is now known as treach13:07
*** sepen has joined #crux-devel13:24
sepenwhat do you think about put my ati port on contrib?13:31
treachwhat a clown.13:57
*** sepen has quit IRC14:07
*** jaeger has quit IRC14:07
*** pitillo has quit IRC14:07
*** sepen has joined #crux-devel14:13
*** jaeger has joined #crux-devel14:13
*** pitillo has joined #crux-devel14:13
*** pitillo has quit IRC15:15
*** j^2 has quit IRC15:59
*** jaeger has quit IRC16:58
*** treach has quit IRC17:36
*** jaeger has joined #crux-devel18:43
*** _mavrick61 has quit IRC19:02
*** _mavrick61 has joined #crux-devel19:04
jaegerwonder where mark is these days19:54
prologicmar k?19:57
prologicI wonder where Anton is19:57
prologicwhere's the pkgutils-c disappeared to ?19:58
jaegerno idea19:59
*** Roomster has joined #crux-devel20:42
*** Romster has quit IRC21:01
*** jaeger has quit IRC22:19
*** Roomster is now known as Romster22:19
*** jaeger has joined #crux-devel22:23
*** Romster has quit IRC22:53
*** jaeger has quit IRC23:16
*** jaeger has joined #crux-devel23:23

Generated by 2.11.0 by Marius Gedminas - find it at!