T2 IRC Log: 2007-08-11

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 Sat Aug 11 00:00:18 2007
00:00 -!- yokoy_ [n=yokoy@gprs-pool-1-023.eplus-online.de] has joined #t2
00:03 -!- cyda [n=Cyda@208.115.42.5] has joined #t2
00:03 -!- yokoy [n=yokoy@gprs-pool-1-027.eplus-online.de] has quit [Remote closed the connection]
00:11 -!- mqueiros_ [n=mqueiros@c-217-70-66-47.bragatel.pt] has quit ["Ex-Chat"]
00:26 -!- kic- [n=rfoote@208.115.42.5] has quit [Read error: 110 (Connection timed out)]
00:42 < CIA-24> ragnar * r25454 /trunk/scripts/Build-Pkg: * Build-Pkg: ignore bogus empty lines when checking for file collisions between packages
01:37 -!- cyda [n=Cyda@208.115.42.5] has quit ["Leaving"]
01:41 -!- renderfull [n=nowhere@c-24-8-147-89.hsd1.co.comcast.net] has quit [Read error: 110 (Connection timed out)]
01:43 -!- Fatal [i=fatal@debian.as] has joined #t2
01:49 -!- kic- [n=rfoote@208.115.42.5] has joined #t2
03:03 -!- kic- [n=rfoote@208.115.42.5] has quit [Read error: 110 (Connection timed out)]
07:58 -!- pdinoto_ [n=pdinoto@host37.190-30-201.telecom.net.ar] has joined #t2
08:15 -!- pdinoto__ [n=pdinoto@host185.190-31-71.telecom.net.ar] has quit [Read error: 110 (Connection timed out)]
09:38 -!- tri_ [n=tri@p57ADAF38.dip0.t-ipconnect.de] has joined #t2
09:39 < tri_> moin
10:22 < LMJ> moin
10:43 < rxr> re
10:43 < rxr> Ragnarin: I fixed the shared file thing already
10:44 < rxr> Ragnarin:and your change contains permission changes
10:53 < CIA-24> rene * r25455 /trunk/scripts/Build-Pkg:
10:53 < CIA-24> * reverted r25454: ignore bogus empty lines when checking for file
10:53 < CIA-24> collisions between packages, as it was fixed for real before and
10:53 < CIA-24> also changed directory permissions in 3 places
10:55 < rxr> Linux: SCO Loses
10:58 -!- yokoy_ [n=yokoy@gprs-pool-1-023.eplus-online.de] has quit [Remote closed the connection]
11:05 < CIA-24> rene * r25456 /trunk/target/neo1973/ (7 files in 2 dirs): * added new neo1973 target - Mobile phone T2 platform for the Neo 1973 (aka. OpenMoko)
11:10 < CIA-24> susan * r25457 /trunk/package/filesystem/udev/udev.desc: * updated udev (113 -> 114)
11:10 < tri_> ah rene is your neo arrived :-)
11:10 < rxr> yes
11:10 < CIA-24> susan * r25458 /trunk/package/develop/uncrustify/uncrustify.desc: * updated uncrustify (0.35 -> 0.36)
11:11 < CIA-24> susan * r25459 /trunk/package/filesystem/ntfs-3g/ntfs-3g.desc: * updated ntfs-3g (1.710 -> 1.810)
11:12 < CIA-24> susan * r25460 /trunk/package/graphic/ufraw/ufraw.desc: * updated ufraw (0.12 -> 0.12.1)
11:12 < CIA-24> susan * r25461 /trunk/package/mail/getmail/getmail.desc: * updated getmail (4.7.5 -> 4.7.6)
11:12 < CIA-24> susan * r25462 /trunk/package/security/gsasl/gsasl.desc: * updated gsasl (0.2.18 -> 0.2.19)
11:13 < CIA-24> susan * r25463 /trunk/package/gnue/gnue-appserver/gnue-appserver.desc: * updated gnue-appserver (0.5.1 -> 0.5.2)
11:22 < CIA-24> rene * r25464 /trunk/scripts/Create-PkgList:
11:22 < CIA-24> * further shortened the internal argument list of Create-PkgList,
11:22 < CIA-24> as with our nearly 3000 packages the scripts sometimes reaches
11:22 < CIA-24> the usual Unix/Linux argument list limit, again ...
11:23 -!- yokoy [n=yokoy@e178102236.adsl.alicedsl.de] has joined #t2
11:49 < rxr> 2/n
11:49 [Users #t2]
11:49 [@ChanServ] [ Enqlave] [ mtr ] [ Ragnarin] [ tri_ ]
11:49 [ CIA-24 ] [ Fatal ] [ pdinoto ] [ rxr ] [ valentin]
11:49 [ Dallur ] [ LMJ ] [ pdinoto_] [ TobiX ] [ yokoy ]
11:49 -!- Irssi: #t2: Total of 15 nicks [1 ops, 0 halfops, 0 voices, 14 normal]
12:03 < CIA-24> rene * r25465 /trunk/scripts/Create-PkgList: * corrected scripts/Create-PkgList list fields after the path was changed earlier to avoid overlong argument lists
12:14 < rxr> ... configuration of 2.6.21.5-dist-moko11 finished!
12:14 -!- yokoy [n=yokoy@e178102236.adsl.alicedsl.de] has quit [Read error: 104 (Connection reset by peer)]
12:14 < rxr> Building the kernel with: make -j3 ARCH=arm CROSS_COMPILE=arm-t2-linux-gnueabi- KCC=arm-t2-linux-gnueabi-kcc vmlinux modules
12:15 < rxr> yuche
12:15 < rxr> heh - the noo / openmoko folks patch yaffs2 intot heir kernel and do not even use it ...
12:15 < rxr> freaky folks ...
12:18 -!- yokoy [n=yokoy@e178102236.adsl.alicedsl.de] has joined #t2
12:20 < rxr> 2== 12:15:28 =[1]=> Building base/linux26 [2.6.21.5-neo1973 8.0-trunk].
12:22 < CIA-24> rene * r25466 /trunk/target/neo1973/ (6 files in 3 dirs): * added neo1973 linux26 overlay (2.6.21.5-moko11)
12:24 -!- tri_ [n=tri@p57ADAF38.dip0.t-ipconnect.de] has left #t2 []
12:28 -!- mqueiros_ [n=mqueiros@c-217-70-65-242.bragatel.pt] has joined #t2
12:34 < rxr> == 12:22:55 =[1]=> Building base/linux26 [2.6.21.5-moko11 8.0-trunk].
12:34 < rxr> == 08/11/07 12:32:05 =[1]=> Finished building package linux26.
12:34 < rxr> :-)
12:37 -!- Dallur [n=earl@fire-out.ru.is] has quit [Read error: 110 (Connection timed out)]
12:38 < mqueiros_> hello! Got this "new" (for me) error message... while using Build-Target: "Hunk #1 FAILED at 11." and "1 out of 1 hunk FAILED -- saving rejects to file passwd.rej", can anyone give me a clue of what might be going wrong ?
12:42 < rxr> which package, which patcvh ?
12:42 < mqueiros_> I was using "sudo ./scripts/Build-Target -cfg tools603"
12:53 < rxr> which package did you got the error on ?
12:54 < mqueiros_> rxr: no package that I can see... I was building the target: http://pastebin.com/m332b98c9
13:00 < rxr> ah - that is in the overlay
13:00 < rxr> from target/.../rootfs processing
13:05 < rxr> http://www.t2-project.org/targets/neo1973/
13:06 < mqueiros_> rxr: "who created"...
13:07 < rxr> well - it is not yet really, there, so ...
13:08 < rxr> reload, better ?
13:08 < mqueiros_> ok... then it should be "to create"...
13:08 < mqueiros_> i think...
13:08 < rxr> yep, that is what I used now
13:09 < rxr> oh - still the functionality typo
13:09 < mqueiros_> yes... I was noticing that to...
13:10 < rxr> so - reload
13:10 < rxr> still something ?
13:12 < mqueiros_> no, it's ok...
13:16 < mqueiros_> Can you give me some hint... on where to look at to find where I got my error from ?
13:19 < rxr> at the end of your target the rootfs overlay is processed
13:19 < rxr> target/.../rootfs/...
13:19 < rxr> you have a patch in there trying to adapt a file
13:19 < rxr> etc/passwd apparently
13:19 < rxr> maybe you do not even need it due to copy from another target or so
13:19 < rxr> but at least it does not apply to your currentl build
13:20 < mqueiros_> ok, I will try to take a look at it... thanks!
13:30 < CIA-24> rene * r25467 /trunk/target/neo1973/pkgsel:
13:30 < CIA-24> * removed comment marks fro mthe neo1973/pkgsel, removed some stuff
13:30 < CIA-24> (apr*, subversion, etc.) addded xbitmaps)
13:33 -!- sepp [n=sepp@Z4bb0.z.pppool.de] has joined #t2
13:33 < sepp> moin :)
13:35 [Users #t2]
13:35 [@ChanServ] [ Fatal ] [ mtr ] [ Ragnarin] [ TobiX ]
13:35 [ CIA-24 ] [ LMJ ] [ pdinoto ] [ rxr ] [ valentin]
13:35 [ Enqlave ] [ mqueiros_] [ pdinoto_] [ sepp ] [ yokoy ]
13:35 -!- Irssi: #t2: Total of 15 nicks [1 ops, 0 halfops, 0 voices, 14 normal]
13:35 < rxr> moin sepp
13:35 < rxr> sepp: http://www.t2-project.org/targets/neo1973/
13:37 < sepp> :)
13:38 < sepp> cellphones annoy me :p
13:40 < sepp> all the morons broadcasting their private life into the bus or train ...
13:41 < sepp> but maybe you can run scummvm on the openmoko thing, would be at least some use
13:41 < sepp> :)
13:41 < rxr> yes, cellphones annoy me, too - crappy ui, dog slow, etc. etc.
13:42 < rxr> but that is why I got myself the Neo to finally get me a T2 phone that doesn't suck ...
13:44 < sepp> :)
13:45 < rxr> btw. I gradually migrated some www.t2 subsections to saner URLs
13:46 < rxr> but with permenent redirecty in the .htaccess
13:46 < rxr> like target/*.html target/#1
13:46 < rxr> and architectures/*.html architectures/#1/
13:46 < rxr> etc. pp.
13:53 < sepp> which package provides libintl, i thought that would be gettext
13:53 < sepp> but i don't have one
13:54 < rxr> glibc, uclibc, or libiconv, IIRC
13:54 < rxr> /var/adm/flists/dietlibc:dietlibc: usr/dietlibc/include/libintl.h
13:54 < rxr> /var/adm/flists/gettext:gettext: usr/lib64/preloadable_libintl.so
13:54 < rxr> /var/adm/flists/glibc:glibc: usr/include/libintl.h
13:55 < sepp> hmm
13:56 < rxr> 2http://www.heise.de/newsticker/meldung/94225
13:56 < rxr> http://www.heise.de/newsticker/meldung/94225
13:58 < sepp> hehe
13:59 < rxr> http://www.golem.de/0708/54061.html
13:59 < rxr> "Gericht: Unix gehört Novell (Update)
13:59 < rxr> Herbe Schlappe für SCO im Kampf gegen Linux"
14:11 < CIA-24> susan * r25468 /trunk/package/security/gsasl/gsasl.desc: * updated gsasl (0.2.19 -> 0.2.20)
14:17 [Users #t2]
14:17 [@ChanServ] [ Fatal ] [ mtr ] [ Ragnarin] [ TobiX ]
14:17 [ CIA-24 ] [ LMJ ] [ pdinoto ] [ rxr ] [ valentin]
14:17 [ Enqlave ] [ mqueiros_] [ pdinoto_] [ sepp ] [ yokoy ]
14:17 -!- Irssi: #t2: Total of 15 nicks [1 ops, 0 halfops, 0 voices, 14 normal]
14:38 -!- sepp [n=sepp@Z4bb0.z.pppool.de] has quit [Read error: 104 (Connection reset by peer)]
14:47 -!- kic- [n=rfoote@c-69-180-101-40.hsd1.fl.comcast.net] has joined #t2
14:48 < kic-> re
15:03 < kic-> SDECFGSET_ARM_ENDIANESS='le' is for little endian, so i would assume that SDECFGSET_ARM_ENDIANESS='be' is for big-endian?
15:28 -!- renderfulll [n=codey@c-24-8-147-89.hsd1.co.comcast.net] has joined #t2
15:28 -!- renderfulll [n=codey@c-24-8-147-89.hsd1.co.comcast.net] has quit [Remote closed the connection]
15:28 -!- renderfull [n=nowhere@c-24-8-147-89.hsd1.co.comcast.net] has joined #t2
15:42 -!- pdinoto__ [n=pdinoto@host167.190-137-8.telecom.net.ar] has joined #t2
15:59 -!- pdinoto_ [n=pdinoto@host37.190-30-201.telecom.net.ar] has quit [Read error: 110 (Connection timed out)]
16:22 < rxr> kic-: yes
16:22 < rxr> kic-: bash did not yet link for armbe yet, that's why I did not yet send a binry to you
16:22 < rxr> actually I did not yet found out what the error message during that link wants to indicate
16:23 < kic-> hrm
16:24 < kic-> i svn up'd
16:24 < kic-> and got the new neo target
16:24 < kic-> and change that config from 'le' to 'be'
16:24 < kic-> and did a build-target
16:24 < kic-> it had failed on binutils i believe..
16:24 < kic-> but after that..
16:24 < kic-> i did
16:25 < kic-> scripts/Build-Target -cfg arm-be -job 1-perl
16:25 < kic-> and it built successfully
16:25 < kic-> is that enough?
16:30 -!- sepp [n=sepp@Z4425.z.pppool.de] has joined #t2
16:33 < rxr> probably
16:33 < rxr> intersting that it linked for you
16:33 < rxr> what dose file build/.../usr/bin/perl say?
16:33 < rxr> MSB or LSB ?
16:34 < sepp> has someone a recent wine.tar.bz2 to share? :)
16:35 < rxr> sepp: binary one ?
16:35 < sepp> yes, x86 of course
16:36 < rxr> probably, just some seconds
16:36 < sepp> i downloaded some rpm crap from sf.net but that does not even support alsa, how stupid
16:42 < yokoy> moin
16:42 < sepp> moin yokoy
16:43 < yokoy> finaly, my computer parts arrived and I assembled them
16:44 < yokoy> the t2 image recognizes two cpus
16:44 < sepp> 2 diff ones? :p
16:44 < yokoy> but it seems the build prozess use only one cpu
16:44 < yokoy> two questions:
16:45 < yokoy> how can I indentify which cpu is used
16:45 < sepp> there is a Config option how many make jobs to use
16:45 < rxr> new: http://www.t2-project.org/hardware/portable/FIC/Neo1973/
16:45 < sepp> default is 2 if you have a dual core
16:45 < sepp> i think
16:45 < rxr> yes, default is detected at Config runtime based on sysfs/proc cpu entries
16:45 < yokoy> yes, i put at this option a "2"
16:46 < rxr> some packages do not build in parallel, and break then
16:46 < rxr> they are marked with NOPARALLEL
16:46 < rxr> t2 will print a status note for those packages that build in paralle, like this:
16:46 < rxr> == 14:45:22 =[1]=> Building base/pkgconfig [0.22 8.0-trunk].
16:46 < rxr> -> Parallel build using 2 jobs enabled.
16:46 < rxr> ...
16:46 < rxr> do you get such "Parallel" notes ?
16:47 < yokoy> yes
16:47 < rxr> then the CPUs should be utliized
16:47 < yokoy> i got them also on my singe cpu system
16:47 < rxr> ok - sure, if you enable parallel build you get the notes
16:47 < rxr> how do you determine CPU load ?
16:48 < rxr> hint: with the command line top you can hit the "1" key and you get a "per-cpu" load in the header lines
16:48 < yokoy> top
16:48 < sepp> or use htop
16:48 < rxr> (or atop)
16:48 < yokoy> ok, i have to build them first
16:48 < rxr> hm - damn - no wine on the reference builder as it was on a rebuild
16:49 < rxr> sepp: susan could tar the /var/adm/flist wine files for you on here Vaio ...
16:49 < rxr> should she ?
16:50 < sepp> no hurry, you could update to 0.9.43 and provide some tar the when it is done
16:50 < yokoy> binutils: 3:24
16:51 < yokoy> it is really faster than my old 32 bit single athlon
16:51 < yokoy> roundabout 3 times
16:51 < sepp> wait for 2 g++ jobs, then it will start to swap like mad :P
16:52 < sepp> with boost or something like that
16:52 < rxr> sepp: ok - notified susan about the update
16:52 < rxr> sepp: more RAM helps ...
16:52 < sepp> thanks
16:52 < sepp> yeah
16:52 < mqueiros_> speaking of speed... Lots of RAM helps ?
16:52 < yokoy> RAM - more money helps :-)
16:53 < mqueiros_> say... 2 Gigs will be way faster then 1 Gig ?
16:54 < rxr> mqueiros_: not necessarily
16:55 < rxr> but some packages require about 1GB of RAm per parallel compiler invocation to build without swapping
16:55 < sepp> with 2gb it still swaps bloatfox out
16:55 < rxr> including: boost, xorg-server, parts of KDE, OpenOffice.org, etc.
16:56 < mqueiros_> ok...
16:57 < rxr> but otherwise 1GB is fine, I have just that in my MacBook as well :-)
16:57 < mqueiros_> rxr: ok... I'v asked because I might buy a dedicated T2 machine...
16:57 < mqueiros_> and wanted to know that...
16:59 < rxr> 1GB per CPU if you compile normal software
16:59 < rxr> can be recommended these days
16:59 < rxr> if you only build tiny embedded stuff you can get away with less RAM ._)
17:16 < mqueiros_> yokoy: binutils 3m and 24s compiling on the new machine ?
17:21 -!- sepp_ [n=sepp@Z5541.z.pppool.de] has joined #t2
17:22 < yokoy> mqueiros_: yes, but I used a slow hd and ccache. I just restart the build without ccache and will give you in a few minutes the correct info...
17:22 < CIA-24> susan * r25469 /trunk/package/emulators/wine/wine.desc: * updated wine (0.9.42 -> 0.9.43)
17:24 < mqueiros_> yokoy: ok... My Pentium 4 3.00Ghz has done 4m25s compiling binutils.
17:27 < yokoy> mqueiros_: yes, I was shure that can't be the right time
17:28 < yokoy> binutils - 2:09 on an usb harddisk
17:29 < mqueiros_> that's faster... :)
17:30 < mqueiros_> yokoy: and you where building binutils on stage ??
17:31 < rxr> http://people.exactcode.de/~susan/wine-0.9.43.tar.bz2
17:31 < yokoy> mqueiros_: 1
17:31 < sepp_> rxr: thanks !
17:31 < yokoy> oh, sorry, 0
17:32 < yokoy> just restarted the build :-)
17:33 -!- Enqlave [i=stealth@unaffiliated/enqlave] has quit ["There is intelligent life out there..."]
17:33 < mqueiros_> yokoy: I was building on stage 2... will check later the time it takes...
17:33 < yokoy> what stage is the reference stage?
17:36 < mqueiros_> yokoy: I don't know...
17:36 < CIA-24> susan * r25470 /trunk/package/gpe/blueprobe/blueprobe.desc: * updated blueprobe (0.17 -> 0.18)
17:39 < CIA-24> susan * r25471 /trunk/package/gpe/bluez-pin/bluez-pin.desc: * updated download tag
17:40 < CIA-24> susan * r25472 /trunk/package/gpe/detect-stylus/detect-stylus.desc: * updated download tag
17:41 -!- Enqlave [i=stealth@unaffiliated/enqlave] has joined #t2
17:42 -!- sepp [n=sepp@Z4425.z.pppool.de] has quit [Read error: 110 (Connection timed out)]
17:49 < rxr> I made two changes to the package matrix:
17:49 < rxr> http://www.t2-project.org/packages/apache.html
17:49 < rxr> the 1) and 2) superscripts are new
17:49 < rxr> and comments or votes ?
17:52 [Users #t2]
17:52 [@ChanServ] [ Fatal] [ mqueiros_] [ pdinoto__ ] [ rxr ] [ valentin]
17:52 [ CIA-24 ] [ kic- ] [ mtr ] [ Ragnarin ] [ sepp_] [ yokoy ]
17:52 [ Enqlave ] [ LMJ ] [ pdinoto ] [ renderfull] [ TobiX]
17:52 -!- Irssi: #t2: Total of 17 nicks [1 ops, 0 halfops, 0 voices, 16 normal]
17:53 < CIA-24> susan * r25473 /trunk/package/gpe/gpe-aerial/ (gpe-aerial.cache gpe-aerial.desc no-iconlist.patch): * updated gpe-aerial (0.2.13 -> 0.3.0)
17:54 < sepp_> wine with sound!
17:54 < sepp_> :)
17:54 < rxr> heh
17:54 < rxr> what are you running in wine ?
17:55 < sepp_> some game from i found while searching the phone
17:55 < rxr> phone ?
17:56 < sepp_> sometimes i hide it from myself
17:56 < sepp_> and then i can not remember where
17:56 < sepp_> :P
17:56 < rxr> ah, searching your phone ...
17:56 < rxr> I though something in the www ...
17:58 < sepp_> http://www.quandaryland.com/jsp/dispArticle.jsp?index=399 ...
18:01 < rxr> :-)
18:02 < sepp_> old enough to work well
18:02 < rxr> lol
18:05 < CIA-24> susan * r25474 /trunk/package/gpe/libsoundgen/ (libsoundgen.cache libsoundgen.conf libsoundgen.desc): * updated libsoundgen (0.3 -> 0.6)
18:08 < CIA-24> susan * r25475 /trunk/package/gpe/libschedule/ (libschedule.cache libschedule.desc): * updated libschedule (0.16 -> 0.17)
18:26 < CIA-24> rene * r25476 /trunk/package/xorg/xorg-server/ (cross.patch xorg-server.desc):
18:26 < CIA-24> * added a litle gross hack to xorg-server to cross compile
18:26 < CIA-24> (though just forcing an uninportable sgml file check to no :-)
18:35 < CIA-24> rene * r25477 /trunk/package/textproc/libxslt/ (libxslt.conf libxslt.desc): * guided libxslt to cross compile (e.g. not pick up host python)
18:39 < CIA-24> rene * r25478 /trunk/package/xorg/ (2 files in 2 dirs): * marked xorg/font-misc-misc and xorg/font-cursor-misc to cross build
18:46 < kic-> root@localhost:/t2/t2-trunk/build/arm-be-8.0-trunk-neo1973-arm-gnueabi-cross/usr/bin# file perl
18:46 < kic-> perl: ELF 32-bit LSB executable, ARM, version 1 (SYSV), dynamically linked (uses shared libs), stripped
18:46 < kic-> oh
18:46 < kic-> its not static
18:51 < rxr> it is LSB
18:53 < rxr> you can select static builds in the Config
18:53 < rxr> somehwere at the expert options at the end
18:53 < rxr> and big-endian is not yet correctly selected ...
18:53 < kic-> hrm
18:53 < rxr> should be MSB executable, ARM, ...
18:54 < kic-> SDECFGSET_ARM_ENDIANESS='be'
18:54 < kic-> thats what i have in in.config
18:54 < kic-> *confg.in
18:54 < CIA-24> rene * r25479 /trunk/target/neo1973/pkgsel:
18:54 < CIA-24> * correctly select the matchbox-window-manager, keyboard and it's
18:54 < CIA-24> dependencies in the neo1973/pkgsel
18:55 < rxr> choice SDECFG_ARM_ENDIANESS el \
18:55 < rxr> eb "Build for big endian machines" \
18:55 < rxr> el "Build for little endian machines"
18:56 < rxr> kic-: I got finished the neo base, I take a look at my armebt-t2-linux / perl link issue
18:56 < kic-> k, thankee
19:09 < yokoy> 1-pkgconfig gives me an error: can not run testprogram while cross compiling
19:10 < yokoy> I remember I checked a box for testing, if possible
19:10 < yokoy> but do not remember I could find this box...
19:10 < yokoy> where I could find this box...
19:11 < yokoy> please, give me a hint
19:13 -!- Fatal [i=fatal@debian.as] has quit [Read error: 110 (Connection timed out)]
19:13 < rxr> yokoy: at the top
19:14 < rxr> 5th line in the config is cross compile
19:14 < kic-> ok, now it fails the build, after making those changes :D
19:16 < CIA-24> susan * r25480 /trunk/package/gpe/gpe-appmgr/ (compile.patch gpe-appmgr.cache gpe-appmgr.desc): * fixed gpe-appmgr to compile
19:16 < kic-> hrm
19:16 < kic-> cannot find -ldl
19:17 < kic-> wonder what provides dl
19:17 < yokoy> rxr: yes, as you advised - or how I understood - I crosscompile on a x686 the 64bit target
19:17 < yokoy> rxr: but there is an additional option for testing.
19:17 < yokoy> I do not find it anymore
19:17 < rxr> pkgconfig din trunk does indeed not cross compile
19:18 < rxr> I have to patch it, once again
19:18 < rxr> kic-: the C library, usually
19:18 < kic-> wee
19:18 < kic-> k
19:19 < rxr> ok - I did a fresh rebuild of my armeb build just to make sure it was not some random error due inconsisten build options while adding the armeb option, but still
19:19 < rxr> armeb-t2-linux-gnueabi-gcc -L/usr/local/lib -o miniperl \
19:19 < rxr> miniperlmain.o opmini.o libperl.a -lnsl -ldl -lm -lcrypt -lutil -lc -lm
19:19 < rxr> /home/rene/develop/t2-trunk/build/ixp425-8.0-trunk-embedded-arm-gnueabi-xscale-cross/usr/lib/libnsl.so: could not read symbols: File i
19:19 < rxr> n wrong format
19:19 < rxr> I have to review this, now
19:21 < kic-> hrm
19:23 -!- Fatal [i=fatal@debian.as] has joined #t2
19:25 < yokoy> is this an install or a live cd?
19:25 < CIA-24> susan * r25481 /trunk/package/gpe/gpe-beam/ (gpe-beam.cache gpe-beam.desc): * updated download tag and gpe-beam.cache file
19:25 < yokoy> http://nexus.tfh-berlin.de/t2/binary/stable/6.0.0/desktop-x86-64-athlon64/
19:26 < rxr> install
19:26 < yokoy> ah, this is smaller:
19:26 < yokoy> http://nexus.tfh-berlin.de/t2/binary/stable/6.0.1/minimal-x86-64-athlon64/
19:26 < rxr> yeah - install likewise, just minimal
19:27 < yokoy> just to go around cross building...
19:31 < rxr> kic-: I think we need to patc / guide gcc to correclty pass another, big-endian arm machine spec to the linker
19:33 < kic-> ah
19:33 < rxr> gcc passes armelf_linux_eabi, but it should be armelfb_linux_eabi
19:33 < rxr> debug-perl:[perl-5.8.8]# armeb-t2-linux-gnueabi-gcc -L/usr/local/lib -o miniperl miniperlmain.o opmini.o libperl.a -lnsl -ldl -lm -lcrypt -lutil -lc -lm -Wl,-marmelfb_linux_eabi
19:33 < rxr> debug-perl:[perl-5.8.8]# file miniperl
19:33 < rxr> miniperl: ELF 32-bit MSB executable, ARM, version 1 (SYSV), dynamically linked (uses shared libs), stripped
19:34 < rxr> I could hand you that resulting binary ...
19:34 < rxr> maybe it works :-)
19:34 < rxr> miniperl: ELF 32-bit MSB executable, ARM, version 1 (SYSV), dynamically linked (uses shared libs), stripped
19:34 < kic-> w00t
19:34 < kic-> oh
19:34 < kic-> those aren't static though?
19:34 < kic-> i could try em i guess
19:34 < kic-> not sure how they'll act
19:34 < kic-> not sure how anything will act actually
19:34 < kic-> so its better than nothin
19:35 < kic-> :D
19:35 < rxr> http://people.exactcode.de/~rene/miniperl
19:35 < rxr> oh - sorry
19:35 < rxr> miniperl: ELF 32-bit MSB executable, ARM, version 1 (SYSV), statically linked, stripped
19:35 < rxr> download, now
19:35 < rxr> 3054cbf3f6757cbc4fbcf016e610735a miniperl
19:35 < rxr> ^- the --static one
19:36 < kic-> w00t
19:36 < rxr> md5sum for the case you started downloading the first one :-)
19:36 < rxr> I have to hunt this machine spec passing in the gcc arm config now ...
19:36 < rxr> I'm really curious if that binary works :-)
19:36 < kic-> me too
19:36 < kic-> because if it does
19:36 < kic-> then i have to hire a perl coder by monday
19:37 < kic-> damn, i don't think i have one of the small clouseau's at the house here
19:37 < kic-> i think i'm gonna have to go into the office today
19:38 < rxr> :-(
19:38 < rxr> btw. how is the whether over there ? still the same hot sunshine ?
19:38 -!- yokoy [n=yokoy@e178102236.adsl.alicedsl.de] has quit [Read error: 104 (Connection reset by peer)]
19:39 < rxr> 3054cbf3f6757cbc4fbcf016e610735a miniperl
19:43 < kic-> if this works
19:43 -!- yokoy [n=yokoy@e178102236.adsl.alicedsl.de] has joined #t2
19:43 < kic-> i'll just use T2 to cross compile any other little binaries i need in the future for the small clouseau
19:43 < kic-> instead of snapgear
19:43 < kic-> that would be a huge help
19:44 < kic-> snapgear is such a pain
19:44 < kic-> rene, whats the different between miniperl binary
19:44 < kic-> and perl binary
19:44 < kic-> ?
19:44 < kic-> we're wanting to use perl for some of our update scripts that are currently written in bash
19:45 < rxr> miniperl does is even less full-featured than the the full cross-built perl
19:45 < kic-> would miniperl limit us to much? the bash scripts we have aren't advanced or anything
19:45 < rxr> when I get gcc fixed I can probably get the little more full featured one out of it
19:45 < kic-> neat
19:46 < rxr> well - I need to take a look mysself that the cross built perl binaries can do and can't
19:46 < kic-> and there'll be a patch or update that i can grab from svn?
19:46 < rxr> I think the C bindings is not available in both of them
19:46 < rxr> and I think the miniperl does not support most external modules or so
19:46 < kic-> i can send you a little clouseau to tinker with if you want an IXP425 arch to build for
19:46 < rxr> but I forgot the exact details
19:46 < kic-> :D
19:46 < rxr> but the cross built perl is enough to get all the other packages bootstapping when building a new system from scratch
19:47 < kic-> we got our new x86 hardware in also, FYI
19:47 < rxr> kic-: :-)
19:47 < kic-> which will be the clouseau 1000 (big clouseau)
19:47 < rxr> kic-: depending how urgent it is I can also take a tiny box with me the next time I'm over there .-)
19:47 < kic-> i put a t2 flash target on it last night
19:47 < kic-> and its handling about 4gb of continuous traffic at the moment
19:47 < kic-> left it running over the weekend
19:47 < kic-> and i'll check on it on monday
19:47 < rxr> but you can also send a tiny one if snapgear doesn't even yield a perl .-)
19:48 < kic-> but so far yesterday t2 + the hardware, was taking it like a champ
19:48 < rxr> :-)
19:48 < rxr> hope the latest code drop can boost performance further ...
19:48 < kic-> we haven't gotten to insmodding clouseau yet
19:48 < kic-> so far just testing t2 build with the hardware
19:49 < kic-> and just a couple of bridges
19:49 < kic-> monday if its still running with no kernel panics
19:49 < kic-> we'll insmod clouseau
19:49 < kic-> and tackle any issues from that
19:50 < rxr> kic-: sidenote, this channel is normally logged for web reference and help people googling for strange build errors :-) you know, right ?
19:50 < kic-> because he didn't understand that when he said strip the target down as small as possible, that you guys actually did that
19:50 < kic-> hah, he didn't have a vmstat command
19:51 < rxr> http://www.t2-project.org/irclogs/t2-2007-08-11.html
19:52 < CIA-24> susan * r25482 /trunk/package/gpe/libmimedir/ (. libmimedir.cache libmimedir.desc): * added libmimedir (0.4.2) - A gobject-based implementation of RFC 2425
19:52 < rxr> the irclogs are mentioned here: http://www.t2-project.org/contact/
19:52 < rxr> so non IRC people can find information asynchronously
19:54 < rxr> trimmed some lines
19:55 < kic-> thanke much
19:55 < kic-> i'll make sure cyda knows the channel is logged also
19:55 < kic-> heh
19:56 < rxr> that is what private messages are for :-)
19:56 < kic-> yep
19:56 < kic-> hrm
19:56 < kic-> wonder if our target will build for ARM
19:57 < kic-> after your gcc fix
19:57 < rxr> not the workstation stuff, but all the minimal stuff for the target: yes
19:57 < kic-> probably to many non marked packages in there though
19:57 < kic-> for the dev workstation
19:57 < kic-> yeah but if any of the workstation stuff fails, the build will halt i think
19:58 < kic-> atleast before stage 4
19:58 < rxr> if you select CROSS build then only the stuff that cross builds is left activated ...
19:58 < kic-> oh i see
19:58 < kic-> hrm
19:58 < rxr> if we have no spurius CROSS tags it probably will just continue to build :-)
19:58 < rxr> just a little less than for your current natvie build
19:58 < rxr> and wasting some CPU cycles on workstation packages that do not even end up in the deployment part
19:59 < kic-> hrm
19:59 < kic-> i can always create a seperate target directory sm-arm
19:59 < kic-> or soemthing
19:59 < kic-> and comment out everything that is not marked already with # BASE
20:00 < rxr> yes, or we introduce an additional code block to only disable the target-systems parts when the cross-build option is set
20:00 < rxr> err, disable the non-target-systems parts ...
20:01 < kic-> neat
20:02 < rxr> hm - the damn gcc has no code to default to big-endian ...
20:02 < rxr> damnmit
20:02 < rxr> patches rotting in gcc/bugzilla since the 2.95.3 days ...
20:02 < CIA-24> susan * r25483 /trunk/package/gpe/libeventdb/ (libeventdb.cache libeventdb.desc): * updated libeventdb (0.30 -> 0.90)
20:08 < kic-> heh, long time
20:10 < rxr> == 20:06:03 =[1]=> Building perl/perl [5.8.8 8.0-trunk].
20:10 < rxr> -> Preparing build in src.perl.ixp425.20070811.200603.5876.blackrider
20:10 < rxr> == 08/11/07 20:09:56 =[1]=> Finished building package perl.
20:11 < kic-> woot
20:12 < rxr> oh - damn, I need to hit the STATIC option :-)
20:12 < rxr> just one second
20:12 < kic-> lol
20:15 < kic-> hrm
20:16 < rxr> http://people.exactcode.de/~rene/perl
20:20 < CIA-24> rene * r25484 /trunk/architecture/arm/gcc-options.in:
20:20 < CIA-24> * added T2 specific -mbig-endian injection for big-endian ARM as GCC
20:20 < CIA-24> no code / support to automcailly default to big-endian for big-
20:20 < CIA-24> endian targets (yet)
20:20 < rxr> shoudl have big-endian _ARM_ targets (yet)
20:20 < rxr> anyway you get the idea
20:25 [Users #t2]
20:25 [@ChanServ] [ Fatal] [ mqueiros_] [ pdinoto__ ] [ rxr ] [ valentin]
20:25 [ CIA-24 ] [ kic- ] [ mtr ] [ Ragnarin ] [ sepp_] [ yokoy ]
20:25 [ Enqlave ] [ LMJ ] [ pdinoto ] [ renderfull] [ TobiX]
20:25 -!- Irssi: #t2: Total of 17 nicks [1 ops, 0 halfops, 0 voices, 16 normal]
20:28 < rxr> T2 package stats: 1026 packages out of the now 2913 do CROSS build
20:28 < rxr> 35.x %
20:30 < mqueiros_> Another newbie question: Is it possible to add a new package in T2 without having T2 installed as build machine ?
20:33 < rxr> add a package to be built along the target, or adding a new package to install in into your running systems ?
20:34 < mqueiros_> add a package to be built along the target... like if I wanted to add a package to T2... or update an existing one...
20:34 < mqueiros_> but using ubuntu... (as I'm now)...
20:34 < mqueiros_> don't know if made myself clear.
20:35 < CIA-24> susan * r25485 /trunk/package/gpe/libgpevtype/ (. compile.patch libgpevtype.cache libgpevtype.desc): * added libgpevtype (0.50) - A interface library to convert mime dir data
20:36 < rxr> yes of course
20:36 < rxr> you can update and add them to be build along with your target
20:36 < rxr> http://www.t2-project.org/developer/pkgtutorial.html
20:37 < mqueiros_> ok... I'v already took a look at the tutorial... just wanted to make sure... :)
20:37 < rxr> :-)
20:37 < rxr> don't forget to submit new packages to the t2 list or so
20:38 < rxr> even packages that are not fully finished, others can still polish them
20:38 < mqueiros_> ok, will do as long as I make it ! :)
20:38 < rxr> btw the OpenMoko distribution has all kinds of junk, but not even "file"
20:45 < kic-> hrm
20:45 < kic-> my target seems to require dietlibc
20:45 < CIA-24> susan * r25486 /trunk/package/gpe/libgpepimc/ (libgpepimc.cache libgpepimc.desc): * updated libgpepimc (0.8 -> 0.9)
20:45 < rxr> kic-: ?
20:45 < kic-> hrm
20:46 < rxr> why ?
20:46 < kic-> !> `bin-x86_64/diet-i' -> `/t2/t2-trunk/build/sm-arm-8.0-trunk-sm-arm-arm-gnueabi-xscale-cross/TOOLCHAIN/tools.cross/diet-bin/diet'
20:46 < kic-> !> `bin-x86_64/diet-dyn-i' -> `/t2/t2-trunk/build/sm-arm-8.0-trunk-sm-arm-arm-gnueabi-xscale-cross/TOOLCHAIN/tools.cross/diet-bin/diet ..
20:46 < kic-> !> Running make CC=armeb-t2-linux-gnueabi-gcc CPP=armeb-t2-linux-gnueabi-cpp CXX=armeb-t2-linux-gnueabi-g++ prefix=/usr/dietlibc CC_FO ..
20:46 < kic-> !> You should not call uname during the build!
20:46 < kic-> !> armeb-t2-linux-gnueabi-cc -I. -isystem include -pipe -nostdinc -Os -fomit-frame-pointer -fstrict-aliasing -mhard-float -Wall -W -Wc ..
20:46 < kic-> !> arm/start.S:0: sorry, unimplemented: -mfloat-abi=hard and VFP
20:46 < kic-> ack, two to many lines, sorry for long paste
20:46 < rxr> ah - just disable dietlibc for now
20:46 < CIA-24> susan * r25487 /trunk/package/gpe/libcontactsdb/ (libcontactsdb.cache libcontactsdb.desc): * updated download tag and libcontactsdb.cache
20:46 < kic-> k
20:46 < rxr> btw you did svn up go get the big-endian ARM fix ?
20:46 < rxr> and last but not least what kernel rums on the ARM box ?
20:46 < rxr> 2.6.x ?
20:46 < kic-> 2.6.19
20:47 < kic-> yeah, just svn'd up
20:47 < kic-> and re-ran build
20:47 < rxr> ah - ok, hopefully it supports the EABI binaries we build by default already
20:50 < kic-> heh
20:50 < kic-> hrm
20:51 < kic-> not sure
20:51 < kic-> that kernel .config i don't have
20:51 < kic-> sits on a snapgear machine that i haven't even touched yet
20:51 < kic-> i'll have to start eventually i suppose
20:52 < kic-> i need to put T2 on the 8 core machine i have now
20:53 < kic-> these builds would fly by
20:58 < mqueiros_> kic: that would be an dual processor, core quad ?
21:00 < kic-> yes
21:01 < kic-> i also have a quad processor, dual core machine
21:01 < kic-> but i don't have access to put T2 on it
21:01 < kic-> it is doing some other functions
21:01 < mqueiros_> nice... :)
21:02 < CIA-24> rene * r25488 /trunk/target/neo1973/config.in: * fixed neo1973 CPU optimization arm710t -> arm720t :-)
21:02 < kic-> erm
21:02 < kic-> danging
21:02 < kic-> same dietlibc error
21:02 < kic-> i disabled it
21:02 < kic-> # X dietlibc
21:02 < kic-> and also tried O dietlibc
21:02 < kic-> :D
21:03 < rxr> quick hack: touch build/.../var/adm/log/1-dietlibc.log
21:03 < rxr> :-)
21:03 < rxr> to so the build systems thinks it was built
21:04 < kic-> heh
21:04 < kic-> k
21:05 < kic-> woot, and perl is next
21:05 < kic-> wee
21:05 < kic-> lets see if it builds
21:05 < kic-> :D
21:06 < rxr> heh - the Neo phone just has a little more advanced ARM CPU than my good old Psion ...
21:06 < rxr> those had the 710T IIRC
21:07 < kic-> ugh, same perl error
21:07 < kic-> ahh well
21:07 < rxr> @ 36 MHz
21:07 < rxr> Neo 1973 is ARM 720T, nearly the same, @ 266 MHz ...
21:07 < rxr> kic-: the same bad error ?
21:07 < rxr> ld complaining about non matching format ?
21:08 < kic-> cannot find: -ldl
21:08 < rxr> cannot find ld ?
21:08 < rxr> dl
21:08 < rxr> strange
21:08 < rxr> which libc do you build ?
21:08 < kic-> collect2: ld returned 1 exit status
21:08 < kic-> uclibc
21:08 < rxr> that really should build
21:08 < rxr> was uclibc built already ?
21:08 < rxr> which target do you use ?
21:09 < kic-> our target
21:09 < kic-> uclibc was built already yes, and built fine
21:09 < rxr> hm
21:10 < kic-> ahh well, i will have to play tonight i guess, i have to go now and take care of the baby
21:10 < kic-> bbl
21:10 * kic- is away.
21:10 < rxr> cu
21:11 < rxr> though in some hours I'll be asleep
21:11 < rxr> 9pm here ...
21:14 < CIA-24> susan * r25489 /trunk/package/gpe/gpe-bootsplash/ (gpe-bootsplash.cache gpe-bootsplash.desc): * updated download tag and gpe-bootsplash.cache
21:18 < CIA-24> susan * r25490 /trunk/package/gpe/gpe-buttonbox/ (gpe-buttonbox.cache gpe-buttonbox.desc): * updated download tag and gpe-buttonbox.cache
21:22 < CIA-24> susan * r25491 /trunk/package/gpe/gpe-calculator/ (gpe-calculator.cache gpe-calculator.desc): * updated download tag and gpe-calculator.cache
21:24 -!- sepp_ [n=sepp@Z5541.z.pppool.de] has quit [Read error: 104 (Connection reset by peer)]
21:28 < kic-> rene
21:28 < kic-> we have our 10mb fiber line in the office now
21:28 < kic-> if i wanted to put up a t2 mirror on it
21:28 < kic-> do i need any scripts that you provide to stay up to date on packages
21:28 < kic-> ?
21:34 < rxr> it's best to use my binary stream from osiris.tfh-berlin.de
21:34 < rxr> I can also push it via rsync/ssh or ftp
21:36 < CIA-24> susan * r25492 /trunk/package/gpe/libtododb/ (. libtododb.cache libtododb.desc): * added libtododb (0.11) - API library
21:36 < CIA-24> susan * r25493 /trunk/package/gpe/libgpevtype/ (compile.patch libgpevtype.desc): * fixed download tag
21:38 < kic-> ok
21:38 < kic-> however you want me to do it is fine
21:38 < kic-> i just wanted to get it started downloading
21:38 < kic-> so that it can complete before the end of the weekend
21:38 < rxr> ssh account with rsync is best
21:39 < kic-> pl
21:39 < kic-> ok
21:39 < kic-> i just created you an account
21:39 < kic-> i'll msg you
21:44 < CIA-24> susan * r25494 /trunk/package/gpe/gpe-bluetooth/ (. gpe-bluetooth.cache gpe-bluetooth.desc): * added gpe-bluetooth (0.55) - A bluetooth connectivity tool
21:54 < CIA-24> susan * r25495 /trunk/package/gpe/gpe-clock/ (. gpe-clock.cache gpe-clock.desc): * added gpe-clock (0.25) - An alarm clock
21:56 < CIA-24> susan * r25496 /trunk/package/gpe/gpe-conf/ (gpe-conf.cache gpe-conf.desc): * updated gpe-conf (0.2.2 -> 0.2.6)
22:03 < rxr> ok - cu all folks
22:03 < rxr> have fun
22:03 < rxr> Ragnarin: note that I reverted your shared files "zero string" check, as I fixed it for real earlier and you commited permission changes along the lines ...
22:04 < rxr> Ragnarin: please let me know if you have duplicate shared files with trunk
22:08 < CIA-24> susan * r25497 /trunk/package/gpe/libxsettings/ (libxsettings.cache libxsettings.desc): * updated download tag and libxsettings.cache
22:31 -!- kic- [n=rfoote@c-69-180-101-40.hsd1.fl.comcast.net] has quit []
22:33 -!- mtr [n=michael@v29762.1blu.de] has quit [Remote closed the connection]
22:35 -!- mtr [n=michael@v29762.1blu.de] has joined #t2
23:26 < Ragnarin> rxr: okay, sorry about the permission changes
23:28 < Ragnarin> rxr: currently running the build for 8.0
--- Log closed Sun Aug 12 00:00:19 2007