T2 IRC Log: 2007-05-20

This is the log as captured by an IRC bot in the channel. The statements are those of the individual people and might not neccessarily reflect the policy and legal rules as set forth by the T2 SDE Project.

« prev | next »

--- Log opened Sun May 20 00:00:47 2007
01:29 -!- mtr_ [n=Michael@kobz-590cba5e.pool.einsundeins.de] has joined #t2
01:46 -!- mtr [n=Michael@kobz-590ca5f3.pool.einsundeins.de] has quit [Read error: 110 (Connection timed out)]
01:46 -!- mtr_ is now known as mtr
07:29 -!- misl [n=chatzill@84-104-172-187.cable.quicknet.nl] has joined #t2
07:48 -!- keinek [n=chatzill@201.250.17.144] has quit [Remote closed the connection]
08:45 -!- misl [n=chatzill@84-104-172-187.cable.quicknet.nl] has quit [Read error: 110 (Connection timed out)]
09:02 -!- misl [n=chatzill@84-104-172-187.cable.quicknet.nl] has joined #t2
09:56 -!- netist [n=chatzill@p548aa47e.dip0.t-ipconnect.de] has joined #t2
10:32 < sepp> http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30052
10:33 < rxr> moin
10:33 < sepp> moin rxr :)
10:33 < rxr> yes, I reported it here as well
10:34 < rxr> xorg-server in t2/trunk now needs very much memory to build, around 1G
10:34 < sepp> ohh, have not seen it ...
10:34 < rxr> that is swap is sufficient to get it build
10:34 < sepp> yesterday it ate my x86_64
10:36 < Stelz> moin rxr, sepp :)
10:36 < sepp> hi Stelz :)
10:37 < rxr> konqueror spinning 200 instances of econvert to rotate jpegs ate mine ...
10:37 < rxr> hi Stelz
10:38 [Users #t2]
10:38 [@ChanServ] [ Dallur] [ misl ] [ Ragnarin] [ Stelz ]
10:38 [ babygor ] [ idealm] [ mtr ] [ rxr ] [ TobiX ]
10:38 [ CIA-9 ] [ LMJ ] [ netist] [ sepp ] [ valentin]
10:38 -!- Irssi: #t2: Total of 15 nicks [1 ops, 0 halfops, 0 voices, 14 normal]
10:38 -!- idealm [n=ideal@c58-107-18-60.belrs2.nsw.optusnet.com.au] has quit [Remote closed the connection]
10:39 < sepp> well, with 2 jobs and 2gb ram + 1gb swap it was running for >1h compiling xf86ScanPci.c before it killed it - on my x2 ...
10:41 < CIA-9> rene * r24140 /trunk/package/multimedia/qdvdauthor/qdvdauthor.desc: * updated qdvdauthor (0.1.5 -> 1.0.0)
10:43 < CIA-9> rene * r24141 /trunk/package/x11/nao/ (. nao.cache nao.desc): * added nao (0.4.1) - A powerful, flexible, and utterly configurable file manager
10:44 < CIA-9> rene * r24142 /trunk/package/filesystem/sshfs/sshfs.desc: * updated sshfs (1.6 -> 1.8)
10:44 < CIA-9> rene * r24143 /trunk/package/database/opendbx/opendbx.desc: * updated opendbx (1.2.2 -> 1.2.3)
10:44 < CIA-9> rene * r24144 /trunk/package/develop/libtomcrypt/libtomcrypt.desc: * updated libtomcrypt (1.16 -> 1.17)
10:45 < CIA-9> rene * r24145 /trunk/package/fonts/dejavu-ttf-fonts/dejavu-ttf-fonts.desc: * updated dejavu-ttf-fonts (2.16 -> 2.17)
10:45 < CIA-9> rene * r24146 /trunk/package/network/tcpreplay/ (. tcpreplay.cache tcpreplay.desc): * added tcpreplay (3.0.1) - A tool which edits and replays captured network traffic back onto the wire
10:46 < CIA-9> rene * r24147 /trunk/package/network/silc-client/silc-client.desc: * updated silc-client (1.0.1 -> 1.0.4.1)
10:51 < rxr> Linux: German Linux Community Boycotting LinuxTag
10:51 < rxr> http://linux.slashdot.org/linux/07/05/19/2149231.shtml
10:52 < sepp> lol
10:53 < sepp> that is the right thing to do, instead of protesting against him directly
10:58 < CIA-9> rene * r24148 /trunk/package/archiver/alien/alien.desc: * updated alien (8.66 -> 8.67)
10:59 < CIA-9> rene * r24149 /trunk/package/develop/avrdude/avrdude.desc: * updated avrdude (5.3.1 -> 5.4)
11:04 < rxr> sepp: you had no success with the blender update ?
11:05 < sepp> running build-target ... dont have py2.5 installed locally
11:05 < sepp> i will know later
11:07 < sepp> was too lazy to mess around with python, i rather wait for build-target and install the new gcc-4.2 based build later
11:07 < rxr> yes sure
11:08 < rxr> just wanted to know if it is worth continuing takeing a look on the failed entry in my reference builder
11:08 < sepp> btw, 1-groff still fails on x86_84
11:08 < sepp> at least on my box
11:08 < rxr> oh, with what kind of error ?
11:08 < CIA-9> rene * r24150 /trunk/package/graphic/asymptote/asymptote.desc: * updated asymptote (1.28 -> 1.29)
11:08 < sepp> the strange c++ error
11:09 < sepp> in configure
11:09 < CIA-9> rene * r24151 /trunk/package/security/prelude-lml/prelude-lml.desc: * updated prelude-lml (0.9.9 -> 0.9.10)
11:10 < sepp> i can reproduce it in a moment ...
11:13 < CIA-9> rene * r24152 /trunk/package/network/hylafax/hylafax.desc: * updated hylafax (5.1.3 -> 5.1.4)
11:15 < sepp> http://rafb.net/p/jSrh0K21.html
11:17 < rxr> hm
11:18 < sepp> i moved it to stage 3 because i did not understand what it wrong ...
11:18 < sepp> to go on
11:18 < rxr> I'll check out locally
11:19 < rxr> maybe my rediff for the cross-g++-header patch is not yet perfect or so
11:19 < sepp> thanks :)
11:24 < rxr> sepp: best would be to throw it into Trac
11:24 < rxr> so I do not fotget it
11:24 < sepp> ;)
11:25 < rxr> forget even
11:25 < rxr> as I do two other things right onw: working on exact-image and waiting for a glibc-2.6 test build to finish .-)
11:34 < sepp> #196 ;)
11:41 -!- LMJ [n=serwou@laf31-2-82-224-107-105.fbx.proxad.net] has quit [Read error: 110 (Connection timed out)]
12:01 < CIA-9> rene * r24153 /trunk/package/develop/loki/ (. loki.cache loki.desc): * added loki (0.1.6) - A C++ library that demonstrates generic programming and design patterns
12:12 -!- write_erase [n=olivier@bon13-2-82-237-125-220.fbx.proxad.net] has joined #t2
12:12 < write_erase> hi
12:12 < sepp> hi write_erase
12:14 < write_erase> While (cross) compiling ircii, the make process first compile some helper tool 'mksiginc' which is used later in the Makefile.
12:14 -!- LMJ [n=serwou@laf31-2-82-224-107-105.fbx.proxad.net] has joined #t2
12:15 < write_erase> As the binary mksiginc is compiled for the target processor, it can't be executed, the makefile fails
12:15 < write_erase> What would be the proper way to solve that ?
12:18 < sepp> dont know, other irc client?
12:18 < sepp> :p
12:21 < rxr> write_erase: irssi should be fixed
12:21 < rxr> we do have similar patches for some x.org packages
12:22 < rxr> the helper that generate something need to be build with the $(HOSTCC)
12:22 < write_erase> I was about to replace $(CC) with 'gcc' .... $(HOSTCC) seems Better ... thx
12:23 < rxr> patches welcome
12:23 < rxr> dont' forget to add a [F] CROSS mark to the irssi package once it cross builds :-)
12:25 < write_erase> Ok
12:32 < rxr> damn - glibc-2.6 causes some regressions
12:32 < rxr> that will be fun to fix :-(
12:45 < CIA-8> rene * r422 /trunk/utility/Endianess.hh: * revisited endianess traits and auto-converter
12:46 -!- d-marc [n=d-marc@HSI-KBW-091-089-001-101.hsi2.kabelbw.de] has joined #t2
13:11 -!- sepp_ [n=sepp@p85.212.76.86.tisdip.tiscali.de] has joined #t2
13:17 -!- misl [n=chatzill@84-104-172-187.cable.quicknet.nl] has quit [Read error: 110 (Connection timed out)]
13:19 -!- sepp [n=sepp@p85.212.42.89.tisdip.tiscali.de] has quit [Read error: 60 (Operation timed out)]
13:34 < CIA-8> rene * r423 /trunk/utility/Endianess.hh: * fixed FreeBSD endianess support
13:38 < CIA-9> susan * r24154 /trunk/package/base/u-boot/u-boot.desc: * fixed u-boot Priority tag (superfluous -)
13:41 < write_erase> rxr mkmksig.pl generate getsignal.c , which is compiled and run > sig.inc containing a list "char *signals[] = {ZERO", "HUP", "INT", "QUIT", "ILL", "TRAP", "IOT", "BUS", "FPE", "KILL", "USR1""
13:42 < write_erase> I think I'll remove all that chain , and add a patch creating a static sig.inc ? is that OK ?
13:44 < rxr> I guess the helper is used to get the actual order or so of those signals for the actual CPU, the signal numbers are CPU & OS dependant
13:46 < write_erase> Yes... But no idea how to make some 'clean' patch
13:47 < rxr> generate a compatible list yourself, probably without using 'cc' but some tiny script or so
13:47 < rxr> try with a quick hack for now
13:48 < rxr> and if you got it running once we can still work out something that the upstream folks accept
13:52 < d-marc> hi
13:54 < write_erase> rxr, What about those kind of errors :
13:54 < write_erase> checking host system type... Invalid configuration `avr32-t2-linux-uclibc': machine `avr32-t2' not recognized
13:54 < write_erase> (glib)
14:02 < CIA-8> rene * r424 /trunk/gfx/EvasHelper.cc: * removed unused endias.h inclusion
14:02 < rxr> the config.{sub.guess} needs updating
14:03 < rxr> we have code that sould try to replace it, however it might not try to replace it in that case
14:05 < write_erase> you mean a patch for config.{sub.guess} so the avr32 is recognized ?
14:05 < rxr> yes you can just copy it from the ones crated by the T2 build
14:05 < rxr> T2 already has code to update the files automatically
14:05 < rxr> however it does not try very hard
14:06 < rxr> I locally have a tiny modification that replaces the files more often, but as I have not yet fully teseted it (whether to cause regressions) I did not yet commit it
14:07 < CIA-8> rene * r425 /trunk/utility/Makefile:
14:07 < CIA-8> * commented dl / pthread libs out for now, breaks exact-image on FreeBSD and if still
14:07 < CIA-8> needed should be just used for the utility module
14:08 < CIA-8> rene * r426 /trunk/utility/Endianess.hh: * adapted the endianess helpers for Apple OS X, sigh
14:18 -!- d-marc [n=d-marc@HSI-KBW-091-089-001-101.hsi2.kabelbw.de] has quit ["KVIrc 3.2.4 Anomalies http://www.kvirc.net/"]
14:53 -!- misl [n=chatzill@84-104-172-187.cable.quicknet.nl] has joined #t2
14:55 < rxr> http://www.heise.de/newsticker/meldung/89936
14:55 < rxr> ^- Oracle verwendete einen IBM System p6570 Server mit acht bei jeweils 4,7 GHz
14:55 < rxr> moin misl
15:27 < CIA-9> rene * r24155 /trunk/package/graphic/exact-image/exact-image.desc: * updated exact-image (0.3.6 -> 0.3.7)
15:41 < write_erase> rxr, glib and other gnome2 libraries are installed in opt/gnome2 ? Is that normal ?
15:42 < rxr> that is the t2 default, yes
15:42 < rxr> you can change that in scripts/config
15:43 < rxr> scripts/Config even
15:45 < write_erase> myavr-7.0-trunk-mediavr-avr32-ap7000-cross/usr/lib/libgmodule-2.0.so: No such file or ....
15:45 < write_erase> while compiling irssi
15:45 -!- write_erase [n=olivier@bon13-2-82-237-125-220.fbx.proxad.net] has left #t2 ["Leaving"]
15:46 -!- write_erase [n=olivier@bon13-2-82-237-125-220.fbx.proxad.net] has joined #t2
15:46 < write_erase> sorry
15:48 < write_erase> Is it possible to switch on the toolchain environement , so I can do the ./configure && make 'by hand' ?
15:49 < sepp_> in src.* - ./debug.sh
15:50 < write_erase> fine
16:03 -!- d-marc [n=d-marc@HSI-KBW-091-089-001-101.hsi2.kabelbw.de] has joined #t2
16:08 < rxr> hey d-marc
16:08 < d-marc> hello
16:08 < rxr> I'm not long in the office anymore, /me's hungry :-)
16:09 < d-marc> rxr: dont forget to look up the binray sizes when you find the time
16:09 < rxr> oh yes
16:12 * rxr booting the t-resc router testboard ...
16:14 < d-marc> thanks :)
16:16 < rxr> libstdc++.so.6.0.8 built against uclibc on x86 optimized for size -Os (at least I think the binry should be): 769768
16:17 < rxr> ^-> T-Resc from T2 SDE 7.0-trunk (2007/01/06)
16:18 < d-marc> hmm
16:31 < rxr> hmm === big ?
16:34 < d-marc> i think is matches what tommi had on his router
16:35 < d-marc> I guess we can save about 500 kb with uclibc++, right?
16:36 < d-marc> and it would be interesting to see what you can leave away from the gnu stdc++
16:36 < rxr> 896248 May 20 12:59 build/default-7.0-trunk-reference-x86-athlon-xp/usr/lib/libstdc++.so.6.0.9
16:36 < rxr> ^- gcc-4.2.0, glibc-2.6, x86-athlon-O1 or so ...
16:36 < d-marc> thats 100 kb more
16:37 < rxr> standard stdc++ is just fat alreay whethr to use a lightweight c library does not change that much
16:37 < d-marc> I mean we can just leave away some of the obsure stuff
16:37 < d-marc> locales and wchar_t base streams
16:38 < d-marc> only one locale needs to be present
16:39 < d-marc> type of locale I mean
16:39 < d-marc> for char
16:42 -!- misl [n=chatzill@84-104-172-187.cable.quicknet.nl] has quit [Read error: 110 (Connection timed out)]
16:46 < CIA-9> rene * r24156 /trunk/package/firmware/ (ipw2100-fw/ipw2100-fw.desc ipw2200-fw/ipw2200-fw.desc): * fixed firmware package License tags
16:51 -!- misl [n=chatzill@84-104-172-187.cable.quicknet.nl] has joined #t2
17:18 -!- keinek [n=chatzill@201.250.17.144] has joined #t2
17:25 < write_erase> rxr, could you help me install my build root and kernel on my ngw100 board ?
17:40 < CIA-9> sebastian * r24157 /trunk/package/graphic/poppler/poppler.desc:
17:40 < CIA-9> * changed poppler build priority (127.500 -> 127.750)
17:40 < CIA-9> right after gtk+, to really build the glib part (which needs gdk...)
18:03 < sepp_> $arch is 'x86-64' instead of 'x86_64' - weird
18:04 < sepp_> is that some type somewhere?
18:04 < sepp_> typo
18:12 < sepp_> hmm, no
18:28 < sepp_> o_O
18:28 < sepp_> -hook_add postmake 5 "cp -v `uname --machine`/libmpeg3.a $libdir/"
18:28 < sepp_> +hook_add postmake 5 "cp -v $arch_machine/libmpeg3.a $libdir/"
18:28 < sepp_> why would - fail and +work ?
18:30 < Ragnarin> uname during builds is a no-no ... breaks cross-building
18:30 < Ragnarin> ?
18:31 < sepp_> sure, but with debug.sh and `uname --machine` i still get x86_64
18:32 < sepp_> but with Build-Target it gets ''
18:33 < sepp_> btw, hi Ragnarin :)
18:34 < Ragnarin> hey sepp_ =) lol
18:34 < Ragnarin> the uname in build-target is a wrapper
18:35 < sepp_> i know, prints dont use me all the time ...
18:44 < rxr> re
18:45 < rxr> sepp_: I think our uname wrapper can not parse the long options
18:45 < rxr> can it ?
18:45 < rxr> oh - no should
18:46 < rxr> I added support for that
18:46 < sepp_> just wonder why it would work with debug.sh but not with Build-Target
18:46 < sepp_> `uname --machine`
18:48 < sepp_> though it should most probably be $arch_machine in the 1st place
18:53 < sepp_> and it was in libmpeg3.conf, so not my fault :p
18:54 < rxr> you said it is empty ?
18:54 < sepp_> only with Build-Target, not in debug.sh
18:54 < rxr> you could try using ' instead of " so that the content is not evaluated before adding to the hook
18:54 < rxr> but just using arch_machine is fine for me
18:59 < sepp_> ahh ' vs " :)
18:59 < rxr> did it make it work for you ?
18:59 < sepp_> yes
19:02 < sepp_> good to know
19:03 < rxr> btw valentin: I started using Exact:: as namsepace for new exact-{audio,image} code :-)
19:03 < rxr> that just looks cool :-)
19:03 -!- misl [n=chatzill@84-104-172-187.cable.quicknet.nl] has quit [Read error: 110 (Connection timed out)]
19:10 -!- misl [n=chatzill@84-104-172-187.cable.quicknet.nl] has joined #t2
19:10 -!- d-marc [n=d-marc@HSI-KBW-091-089-001-101.hsi2.kabelbw.de] has quit ["KVIrc 3.2.4 Anomalies http://www.kvirc.net/"]
19:14 < sepp_> bash segfaults running ./build.sh inside boost-jam
19:18 < rxr> not good
19:18 < sepp_> yup :/
19:19 < rxr> glibc-2.6 has lovely regressions:
19:19 < rxr> !> ///usr/include/sys/stat.h:508: warning: C99 inline functions are not supported; using GNU89
19:19 < rxr> !> ///usr/include/sys/stat.h:515: warning: C99 inline functions are not supported; using GNU89
19:19 < rxr> !> ///usr/include/sys/stat.h:523: warning: C99 inline functions are not supported; using GNU89
19:19 < rxr> !> In file included from utimecmp.c:34:
19:19 < rxr> Maybe we should start stripping -std=whatever from packages ...
19:20 < rxr> and just let the compiler accept whatever it's capable of compiling instead of random down-foring to some standard the system headers are not even compatible with
19:20 < rxr> and a new glibc maintainer shoudl be elected ...
19:20 < sepp_> hehe
19:25 -!- keinek [n=chatzill@201.250.17.144] has quit ["ChatZilla 0.9.78.1 [Firefox 2.0.0.3/2007031001]"]
20:03 < CIA-9> susan * r24158 /trunk/scripts/config.in: * fixed custom pkgsel rule parser to accept the new '=' (set to default) operator
20:20 -!- keinek [n=chatzill@201.250.17.144] has joined #t2
20:29 -!- Netsplit calvino.freenode.net <-> irc.freenode.net quits: write_erase
20:30 -!- Netsplit over, joins: write_erase
20:31 -!- Netsplit calvino.freenode.net <-> irc.freenode.net quits: write_erase
20:33 -!- Netsplit over, joins: write_erase
21:12 -!- d-marc [n=d-marc@HSI-KBW-091-089-001-101.hsi2.kabelbw.de] has joined #t2
22:07 -!- netist [n=chatzill@p548aa47e.dip0.t-ipconnect.de] has quit [Read error: 104 (Connection reset by peer)]
22:48 -!- d-marc [n=d-marc@HSI-KBW-091-089-001-101.hsi2.kabelbw.de] has quit ["KVIrc 3.2.4 Anomalies http://www.kvirc.net/"]
23:10 -!- misl [n=chatzill@84-104-172-187.cable.quicknet.nl] has quit [Read error: 110 (Connection timed out)]
--- Log closed Mon May 21 00:00:48 2007