--- Log opened Sat Apr 07 00:00:06 2007
00:04 -!- axion_ [n=axion@cpe-74-70-236-179.nycap.res.rr.com] has quit ["Lost terminal"]
00:12 -!- keinek [n=chatzill@201.250.10.4] has quit ["ChatZilla 0.9.78 [Firefox 2.0.0.3/2007031001]"]
00:57 -!- sepp [n=sepp@p213.54.180.29.tisdip.tiscali.de] has joined #t2
01:05 -!- sepp_ [n=sepp@p85.212.49.126.tisdip.tiscali.de] has quit [Read error: 60 (Operation timed out)]
01:05 -!- mtr_ [n=Michael@kobz-590cb064.pool.einsundeins.de] has joined #t2
01:14 -!- mtr [n=Michael@kobz-590cb09f.pool.einsundeins.de] has quit [Read error: 60 (Operation timed out)]
01:14 -!- mtr_ is now known as mtr
02:04 -!- mtr_ [n=Michael@kobz-590cb1ea.pool.einsundeins.de] has joined #t2
02:13 -!- mtr [n=Michael@kobz-590cb064.pool.einsundeins.de] has quit [Read error: 60 (Operation timed out)]
02:13 -!- mtr_ is now known as mtr
04:05 -!- mqueiros [n=mqueiros@c-217-70-65-169.bragatel.pt] has left #t2 ["Ex-Chat"]
05:03 -!- sparc-kly_ [n=mubex@adsl-64-237-252-124.prtc.net] has quit ["Leaving"]
05:18 -!- idealm [n=ideal@c58-107-18-60.belrs2.nsw.optusnet.com.au] has joined #t2
05:46 -!- emte_ [n=emte@d64-180-45-14.bchsia.telus.net] has joined #t2
05:48 -!- emte [n=emte@d64-180-45-14.bchsia.telus.net] has quit [Read error: 110 (Connection timed out)]
06:06 -!- idealm [n=ideal@c58-107-18-60.belrs2.nsw.optusnet.com.au] has quit [Remote closed the connection]
08:04 -!- idealm [n=ideal@c58-107-18-60.belrs2.nsw.optusnet.com.au] has joined #t2
09:15 -!- idealm [n=ideal@c58-107-18-60.belrs2.nsw.optusnet.com.au] has quit [Remote closed the connection]
09:24 -!- sepp_ [n=sepp@p213.54.184.236.tisdip.tiscali.de] has joined #t2
09:33 -!- sepp [n=sepp@p213.54.180.29.tisdip.tiscali.de] has quit [Read error: 60 (Operation timed out)]
09:43 < rxr> re
09:51 < CIA-18> rene * r23451 /trunk/package/filesystem/ntfsprogs/ntfsprogs.desc: * updated ntfsprogs Url tag
09:57 < CIA-18> rene * r23452 /trunk/package/network/cobbler/cobbler.desc: * updated cobbler (0.4.5 -> 0.4.6)
10:01 < CIA-18> rene * r23453 /trunk/package/gnome2/gnome-system-tools-backends/gnome-system-tools-backends.desc: * updated gnome-system-tools-backends (2.2.0 -> 2.2.1)
10:05 < CIA-18> rene * r23454 /trunk/package/filesystem/unison/unison.desc: * updated unison (2.13.16 -> 2.27.10)
10:07 -!- wbx [n=wbx@gprs-pool-1-020.eplus-online.de] has joined #t2
10:07 < CIA-18> rene * r23455 /trunk/package/network/nginx/nginx.desc: * updated nginx (0.5.16 -> 0.5.17)
10:15 -!- AStorm [n=astralst@host-81-190-179-124.gorzow.mm.pl] has quit ["Leaving"]
10:28 -!- AStorm [n=t2@host-81-190-179-124.gorzow.mm.pl] has joined #t2
10:38 -!- wbx_ [n=wbx@gprs-pool-1-023.eplus-online.de] has joined #t2
10:41 -!- wbx [n=wbx@gprs-pool-1-020.eplus-online.de] has quit [Remote closed the connection]
10:43 < AStorm> Hm, t2-resc is unsuitable for installing T2 system :/
10:43 < AStorm> No gcc there.
10:44 < rxr> yes, it is a rescue system
10:44 < AStorm> :/
10:44 < AStorm> The best rescue is the reinstall... never heard of it? :P
10:45 < AStorm> I've found static irssi and subversion, but static gcc... that would be hard :>
10:46 < rxr> just download a T2 minimal and unpack it in a fresh root and chroot() ?
10:47 < AStorm> That I'll do :>
10:47 < AStorm> But first, I have to update my t2 tree
10:47 < AStorm> and repartition the HD
10:47 < AStorm> (small, 4G root partition has to be enough)
10:47 < AStorm> (and one large for compiling)
10:49 < AStorm> Well, some network setup and irssi on the T2-Resc would be appreciated too :>
10:50 < AStorm> Doing these by hand is... funny.
10:50 < AStorm> irssi does compile against dietlibc.
10:50 < AStorm> Uhoh...
10:50 < AStorm> IO error...
10:50 < AStorm> Damn, I've probably messed up the size of the partition
10:51 < AStorm> stupid parted creates them in a weird way
10:53 < AStorm> reminder: don't use parted to create partitions
10:54 < AStorm> drat, IO error on the data partition...
10:55 < AStorm> Hmm... I better copy the maildir and etc backup from there to the encrypted one
10:56 < AStorm> That one is rock-solid
10:56 < AStorm> reminder 2: don't use ext4 just yet
10:57 < AStorm> Hmm... t2 doesn't support truecrypt yet - will have to fix that
10:57 < AStorm> It's a kernel module - are there any provisions for building these?
10:58 < rxr> yes, in T2 all third party kernel modules are build along, that is inside, the kernel build
10:58 < rxr> so all selected third pary kernel modules are available for the just build kernel
10:59 < rxr> just take a look at other packages shipping kernel modules, some wifi module or mol or so
10:59 < AStorm> To add one to the build I have to modify the target?
11:00 < rxr> to add truecrypt?
11:00 < rxr> you have to add it as package and if you want it in some target, at idd to the target's pkgsel
11:00 < AStorm> Ah, yes.
11:00 < rxr> if you use no target, but just generic it would be selected by default anyway
11:00 < AStorm> Heh, even better.
11:06 < CIA-18> rene * r23456 /trunk/package/base/linux26/linux26.desc: * updated linux26 (2.6.20.4 -> 2.6.20.5)
11:09 < AStorm> Niceness :-)
11:12 -!- wbx [n=wbx@gprs-pool-1-020.eplus-online.de] has joined #t2
11:12 < CIA-18> rene * r23457 /trunk/package/base/linux26/linux26.desc: * updated linux26 (2.6.20.5 -> 2.6.20.6)
11:15 < AStorm> Hm.
11:16 -!- wbx_ [n=wbx@gprs-pool-1-023.eplus-online.de] has quit [Remote closed the connection]
11:19 -!- sepp [n=sepp@p85.212.65.179.tisdip.tiscali.de] has joined #t2
11:28 -!- sepp_ [n=sepp@p213.54.184.236.tisdip.tiscali.de] has quit [Read error: 60 (Operation timed out)]
11:57 -!- wbx [n=wbx@gprs-pool-1-020.eplus-online.de] has quit ["Lost terminal"]
12:13 [Users #t2]
12:13 [@ChanServ] [ emte_] [ mtr ] [ sepp ] [ TobiX ]
12:13 [ AStorm ] [ fke ] [ R4gnar0k] [ Stelz] [ valentin]
12:13 [ CIA-18 ] [ LMJ ] [ rxr ] [ th ]
12:13 -!- Irssi: #t2: Total of 14 nicks [1 ops, 0 halfops, 0 voices, 13 normal]
13:06 -!- fke_ [n=fke@p3e9e4015.dip0.t-ipconnect.de] has joined #t2
13:14 -!- idealm [n=ideal@c58-107-18-60.belrs2.nsw.optusnet.com.au] has joined #t2
13:14 -!- sepp_ [n=sepp@p213.54.56.17.tisdip.tiscali.de] has joined #t2
13:14 -!- fke [n=fke@p3e9e31f9.dip0.t-ipconnect.de] has quit [Read error: 60 (Operation timed out)]
13:17 < AStorm> Hey guys... which bootloader will work on all these archs?
13:17 < rxr> which all these arches ?
13:17 < AStorm> x86(_64), sparc, ppc
13:17 < rxr> one on all, yet
13:18 < AStorm> :/
13:18 < rxr> grub2 maybe some time in the future
13:18 < AStorm> Damn.
13:18 < rxr> silo for sparc, yaboto for ppc
13:18 < AStorm> I need such a loader :>
13:18 < AStorm> For that smart pendrive
13:18 < rxr> well - then try grub2 and hope it already works on those ... :-)
13:18 < AStorm> or somehow to allow loading different ones
13:18 < rxr> well - you can embedded multiple
13:18 < AStorm> no, it doesn't, according to the TODO list :P
13:18 < rxr> well - you would need three version of grub2 anyway
13:19 < rxr> just add the three loaders
13:19 < AStorm> Mhm.
13:19 < AStorm> Good. That could work.
13:19 < rxr> as long as their requirements (sector location etc.) do not conflict it works
13:19 < AStorm> Now, on to hacking in that eprefix/libdir mod for T2
13:19 < AStorm> rxr: well, the M$ boot sector of the pendrive :P
13:20 < rxr> ppc and sparc use other mechanisms to load the loader from
13:20 < AStorm> :/
13:20 < AStorm> You mean OpenFirmware :>
13:20 < AStorm> I have to somehow make it work at least on x86(_64) and sparc
13:20 < rxr> yes, but it differs on sparc and powerpc
13:20 < AStorm> ppc is for later date
13:21 < AStorm> (just 2 machines, not worth it :P )
13:22 < rxr> valentin: even the HTML split files are named perfectly, in contrasat to the LaTeX random hash naming ...
13:22 -!- sepp [n=sepp@p85.212.65.179.tisdip.tiscali.de] has quit [Read error: 60 (Operation timed out)]
13:22 < AStorm> First, x86_64 with additional unionfs overlay
13:23 < AStorm> I mean, the basic environment will have a nice overlay added on my laptop :>
13:23 < AStorm> Hmm, where should I look for setting eprefix/libdir for all "makes"?
13:23 < AStorm> (where possible)
13:24 < AStorm> A'la multilib
13:24 < AStorm> (except w/o multilib - I'll use normal x86 instead of it)
13:27 < AStorm> autoconf would be easy, the other makefiles... would require patches
13:27 < AStorm> (maybe not cmake...)
13:27 < rxr> if you want to test it out for your target, target/astrorm/parse-config should do
13:28 < rxr> var_append extraconfopt ' ' "--with-whatever=xyz"
13:28 * rxr off for lunch
13:28 < AStorm> I know.
13:28 < AStorm> But the other makes?
13:28 < rxr> var_append makeopt ' ' "EPREFIX_FOO_BAR=xyz"
13:29 < AStorm> Good :>
13:29 < AStorm> libdir too
13:29 < rxr> and/or
13:29 < rxr> var_append makeinstopt ' ' "EPREFIX_FOO_BAR=xyz"
13:29 < rxr> cu then
13:29 < AStorm> Both.
13:35 -!- idealm [n=ideal@c58-107-18-60.belrs2.nsw.optusnet.com.au] has quit [Remote closed the connection]
13:38 < AStorm> drat, the "minimal desktop live" is too large :P
13:38 < AStorm> I better use some Gentoo stage2. Should be enough.
13:39 < AStorm> You could provide some minimal build environment too :-)
13:41 < AStorm> is the debug patch in already?
13:47 < AStorm> Hmm, I see it's not.
13:47 < AStorm> Will have to fix the cross for it and apply.
13:51 < AStorm> added some missing archprefixes
13:57 -!- sparc-kly [n=mubex@adsl-64-237-252-124.prtc.net] has joined #t2
13:59 < AStorm> Ok, will have to write SLiM .desc :>
13:59 < AStorm> http://slim.berlios.de
14:34 -!- emte__ [n=emte@d64-180-45-14.bchsia.telus.net] has joined #t2
14:35 -!- emte_ [n=emte@d64-180-45-14.bchsia.telus.net] has quit [Read error: 110 (Connection timed out)]
14:59 < AStorm> Uhoh
14:59 < AStorm> the current ./scripts/Config doesn't work in Gentoo stage2
14:59 < AStorm> I wonder why
15:00 < AStorm> it doesn't show any dialog :/
15:00 < AStorm> ncurses are available
15:01 < AStorm> the dialog doesn't appear, though the config is created
15:03 < AStorm> ah, no TERM variable :P
15:03 < AStorm> Ok, now it does work. Please document that behaviour
15:04 < AStorm> When will T2 contain gcc 4.2?
15:08 < AStorm> Hmm, maybe I'll just hack the .desc of gcc 4.1.x to support 4.2
15:08 < AStorm> I need these -march=generic
15:09 < sepp_> after its release?
15:09 < AStorm> It is released :-)
15:09 < AStorm> At least some -rc
15:10 < AStorm> Hmm, maybe not - almost there though.
15:10 < AStorm> You could prepare for it.
15:10 < AStorm> (6 regressions only)
15:10 < sepp_> get a faster computer?
15:10 < sepp_> :p
15:10 < AStorm> Even less.
15:11 < AStorm> sepp_: no, just -march=generic for those archs :>
15:11 < rxr> gcc-4.2 does miscompile most non x86 linux kernels
15:11 < AStorm> Because I can't set some other march
15:11 < AStorm> rxr: ?!?
15:11 < AStorm> Not 2.6.20
15:11 < sepp_> i mean as preparation
15:11 < AStorm> Whoops, not 2.6.20-git
15:11 < AStorm> sepp_: too :>
15:11 < rxr> but updating your working copy is just a matter of exchanging the versin tag
15:11 < rxr> AStorm: non x86
15:11 < rxr> AStorm: e.g. sparc, ppc, et al.
15:11 < AStorm> rxr: I do mean non x86
15:12 < AStorm> There are some kernel patches (for kernel stupidity hidden by older gcc :> )
15:12 < AStorm> they're in the git tree
15:12 < AStorm> will be in 2.6.21
15:12 < rxr> yes sure, I just wanted to point out why we do not have gcc-4.2-rc in the trunk ..
15:12 < AStorm> I know why :>
15:12 < AStorm> Just a gcc-4.2 desc :>
15:13 < AStorm> Not gcc normal
15:13 < AStorm> Hmm, or is installing 2 versions side-by-side not yet supported? (a major drawback for me...)
15:15 < AStorm> If so, I'll have to fix that somehow.
15:15 < AStorm> (I do need various cross-compilers)
15:18 < rxr> in the past, that is around the fork from ROCK Linux this was common practise
15:18 < rxr> in T2 we focused on getting the latest, greatest building everything in a stable way
15:25 < CIA-18> rene * r23458 /trunk/package/mail/fetchmail/fetchmail.desc: * updated fetchmail (6.3.7 -> 6.3.8)
16:04 < AStorm> rxr: good :-)
16:05 < AStorm> Though I do need multiple gcc versions for development :>
16:05 < AStorm> Except ABI compatibility will stop me, ehhh... bad anyway.
16:05 < AStorm> Would need full multiple systems, with libs
16:06 < AStorm> tags like: x86_64-generic-pc-linux-gnu-4.2 would be good
16:06 < AStorm> (with libdir and eprefix mod)
16:06 < AStorm> Maybe some true tags.
16:07 < AStorm> E.g. x86_64-generic-pc-linux-gnu-4.0 because abi compat is mostly unbroken on major 4.x
16:07 < AStorm> Similar system could be used for library versions.
16:08 < AStorm> So, the full tag would consist of $CHOST-$ABIVER-$LIBVER
16:09 < AStorm> On compiler, there would be no $LIBVER
16:09 < AStorm> That'd allow cross-compiling too, just add some symlinks or path manipulation
16:10 < AStorm> different compilers of the same ABI version could use the --program-suffix
16:10 < AStorm> :>
16:11 < AStorm> and some wrapper/symlink
16:11 < AStorm> Almost like Gentoo, but much better (supports cross builds much better :-) )
16:12 < AStorm> s/$CHOST/$CTARGET/
16:12 < AStorm> And place the build tools under:
16:13 < AStorm> $CHOST-$ABIVER/$CTARGET
16:13 < AStorm> libs like $CHOST-$ABIVER/$LIBNAME-$LIBVER (when needed)
16:13 < AStorm> ldconfig can take care of that
16:14 < AStorm> A special tag type [S] (or something) would mean that the lib headers and binaries have to be artificially versions
16:15 < AStorm> *versioned
16:15 < AStorm> I'll better write all of that down. :-)
16:16 < AStorm> To bind all of that for legacy purposes, you can use unionfs or bind mounts or even symlinks
16:17 < AStorm> Hmm, for the compilers, it has to be $CHOST-$ABIVER/$CTARGET-$ABIVER
16:17 < AStorm> the abi version has to be set by hand
16:21 < AStorm> E.g. the .desc file or parse-config
16:25 < AStorm> The hardest part, and the one I'll need your assistance, is getting the dependency checker to recognise these tags
16:39 < rxr> sorry, I just reformat the old LaTeX T2 Handbook sources into XML, Docbook
16:40 < rxr> guess I'll be a bit unresponsive over easter anyway
16:40 < rxr> if you have a specific question / problems always remember to drop me a note by mail ...
16:41 < AStorm> Not a problem at all :-)
16:41 < AStorm> (yet)
16:54 < AStorm> Hmm, I'm also missing an option to override CFLAGS for some packages
16:55 < AStorm> E.g. I want to build the whole system against dietlibc, but some performance-sensitive packages against glibc with -O2 instead of -Os
16:57 < AStorm> s/dietlibc/uclibc/
16:57 < AStorm> I wonder which is smaller :>
16:57 < rxr> you can set this in your target, though
16:58 < rxr> in the whole mighty parse-config
16:58 < AStorm> Hmm, just for some packages?
16:58 < rxr> sure
16:58 < rxr> switch $pkg in
16:58 < rxr> foo|bar) var_append GCC_WRAPPER_APPEND ' ' '-O2'
16:58 < AStorm> Good.
16:58 < rxr> ;;
16:58 < rxr> joo|yoo) var_append GCC_WRAPPER_APPEND ' ' '-Os'
16:58 < rxr> ...
16:58 < rxr> ;;
16:58 < rxr> esac
16:58 < AStorm> I'll use a table lookup instead of case.
16:59 < rxr> ok, s/switch/case/ :-)
16:59 < rxr> whatever
16:59 < AStorm> Anyway - is the uclibc build also static?
16:59 < rxr> just altering the c library is a bit harder
16:59 < rxr> dietlibc can be dynamic as well
16:59 < rxr> at least on x86 and arm ...
16:59 < AStorm> I know, on some archs.
16:59 < rxr> uclibc is optionally static or dynamic just as you wish
16:59 < AStorm> Mhm, driven by c flags?
17:00 < rxr> the static vs. dynamic ?
17:00 < AStorm> Yes.
17:00 < rxr> no, just cflags is not enough
17:00 < rxr> we have this SDECFG_STAITC or so
17:00 < rxr> that controls some more things, sometimes even per package
17:00 < AStorm> Let me search for that...
17:00 < rxr> as some need --enable-static and other special quirks
17:00 < AStorm> Ok, got that.
17:00 < rxr> you could reset this SDECFG_STATIC_something in the parse-config as well
17:00 < rxr> that should work
17:01 < rxr> (though noone asked for this yet and thus I guess that is untseted :-)
17:01 < AStorm> I wonder if the packages requiring dietlibc could be made to ue uclibc too.
17:01 < AStorm> *use
17:01 < rxr> btw. please be aweare that when you build ncurses against one libC you can not link programs needing curses with another libC and such
17:01 < AStorm> I know.
17:01 < AStorm> That's why I want to get rid of dietlibc as much as possible :-)
17:01 < rxr> that is why I have on the SoC page generic multilib to allow building a package's libraries multiple times for whatever multilib variants requested
17:02 < AStorm> Good, I need that too for different archs.
17:03 < AStorm> Hmm... what does "multilib support" offer except installing 32-bit libc?
17:03 < AStorm> I'd love to disable that (don't need it)
17:03 < AStorm> Can't even in cross build.
17:08 < rxr> multilib can be more than just a 32bit variant
17:08 < rxr> eg. no-fp on powerpc, et al.
17:08 < rxr> note that you need a functional 32bit compiler for the boot loader on all architectures
17:08 < AStorm> I know, but I want the stupid thing turned off... can't do so in the generic target :>
17:08 < AStorm> Guess I'll have to write one.
17:09 < rxr> should be possible to turn it off
17:09 < rxr> maybe some architectures forces it on, though ...
17:09 < AStorm> x86_64
17:09 < AStorm> It should be optional there.
17:11 < rxr> t2-trunk/architecture# grep MULTILIB */*
17:11 < rxr> mips64/config.in:SDECFGSET_MULTILIB=1
17:11 < rxr> powerpc64/config.in:SDECFGSET_MULTILIB=1
17:11 < rxr> sparc64/config.in:SDECFGSET_MULTILIB=1
17:11 < rxr> superh/config.in:SDECFGSET_MULTILIB=1
17:11 < rxr> superh/config.in:const SDECFG_MULTILIBLIST "m4,m4-nofpu"
17:11 < rxr> superh/config.in:SDECFGSET_MULTILIBLIST="m4,m4-nofpu"
17:11 < rxr> x86-64/config.in:SDECFGSET_MULTILIB=1
17:11 < rxr> though this is for dau users
17:12 < rxr> I think you should be able to overwrite it in your own target
17:12 < rxr> will all the stuff you mentioned here it's a good time to start to save all your code sniplets in a fres target anyway ...
17:12 < AStorm> Gah. I have to write one anyway, to get rid of dietlibc :-)
17:24 < AStorm> Hmm, I can't override the multilib from the target
17:24 < AStorm> :/
17:25 < AStorm> at least not from config.in
17:26 < rxr> hm
17:26 < rxr> hm
17:26 < rxr> should be
17:26 < rxr> what did you do ?
17:26 < rxr> SDECFGSET_... you must use
17:26 < AStorm> just set SDECFGSET_MULTILIB=0
17:26 < AStorm> I did.
17:26 < rxr> hm
17:26 < rxr> hm
17:26 < rxr> drop an note in the bugtracker and alter the architecture locally for now I would day ...
17:27 < rxr> say ...
17:27 < AStorm> Ok :-)
17:27 < rxr> do not forget the bugtracker step
17:27 < AStorm> Doing that right now...
17:27 < rxr> you even do not need to register, though that would be a nice bonus
17:30 < AStorm> Filed.
17:30 < AStorm> I will register when I get a real browser and be able to access my mail :>
17:36 < AStorm> It's ticket #185
17:36 < rxr> yep just took a look
17:37 < AStorm> Does your uclibc support NLS?
17:38 < AStorm> Uhoh
17:38 < rxr> should
17:38 < AStorm> Gentoo stage doesn't have curl, but has wget :-)
17:38 < AStorm> Hacking away again :-)
17:40 < AStorm> Too large hack for me. I'll install curl instead.
17:45 < AStorm> BRB
17:46 -!- AStorm [n=t2@host-81-190-179-124.gorzow.mm.pl] has quit ["leaving"]
17:47 -!- AStorm [n=t2@host-81-190-179-124.gorzow.mm.pl] has joined #t2
17:56 -!- fke [n=fke@p3E9E015D.dip0.t-ipconnect.de] has joined #t2
18:03 < AStorm> ok, upped gcc to 4.2.0-20070316
18:03 < AStorm> (prerelease, not snapshot)
18:05 < AStorm> Just a 14 MB diff against 4.1.2 :P
18:05 < AStorm> (of course, I'm using a tarball)
18:09 -!- fke_ [n=fke@p3e9e4015.dip0.t-ipconnect.de] has quit [Read error: 110 (Connection timed out)]
18:27 < CIA-18> rene * r23459 /trunk/package/lua/lpeg/ (hotfix.patch lpeg.desc): * updated lpeg (0.4 -> 0.5)
18:28 -!- sepp [n=sepp@p85.212.53.173.tisdip.tiscali.de] has joined #t2
18:31 -!- emte__ [n=emte@d64-180-45-14.bchsia.telus.net] has quit [Read error: 110 (Connection timed out)]
18:31 -!- emte_ [n=emte@d64-180-45-14.bchsia.telus.net] has joined #t2
18:37 -!- sepp_ [n=sepp@p213.54.56.17.tisdip.tiscali.de] has quit [Read error: 60 (Operation timed out)]
18:43 < AStorm> sepp: help please ;-) (too much AoE2)
18:43 < AStorm> !> * Code maturity level options
18:43 < AStorm> !> Due to previous errors, no 0-linux-header.log file!
18:44 < AStorm> You know, something is broken.
18:48 < AStorm> Hmm, the header build fails stupidly.
18:49 < AStorm> Shouldn't.
18:50 < AStorm> * Restart config...
18:50 < AStorm> Console input/output is redirected. Run 'make oldconfig' to update configuration
18:50 < AStorm> .
18:51 < AStorm> Somebody forgot about something?
18:51 < rxr> hm
18:52 < rxr> whats the exact error ?
18:52 < AStorm> Pasted above
18:52 < AStorm> linux-header asks for input...
18:52 < rxr> hm
18:53 < rxr> I moved something thre just recently
18:55 < AStorm> I can't sync with svn just right now (too dirty tree)
18:55 < rxr> hm - did not found a recent change though
18:56 < rxr> which exact line is that config run from ?=
18:56 < AStorm> ./scripts/Build-Target
18:56 < AStorm> and I shall check the xtrace
18:57 < AStorm> ====[main_lx_header:23 (last $?=0)> make -j4 ARCH=x86_64 CROSS_COMPILE= KCC=kcc
18:57 < AStorm> include/asm
18:57 < AStorm> (sorry for line-wrap)
18:59 < rxr> --- package/base/linux-header/linux-header.conf (revision 23440)
18:59 < rxr> +++ package/base/linux-header/linux-header.conf (working copy)
18:59 < rxr> @@ -20,12 +20,13 @@
18:59 < rxr>
18:59 < rxr> main_lx_header() {
18:59 < rxr> lx_patch
18:59 < rxr> - eval $MAKE $makeopt include/asm
18:59 < rxr>
18:59 < rxr> # we need to create an config since this generates some files
18:59 < rxr> yes '' |
18:59 < rxr> eval $MAKE $makeopt oldconfig > /dev/null
18:59 < rxr>
18:59 < rxr> + eval $MAKE $makeopt include/asm
18:59 < rxr> +
18:59 < rxr> # garuanteed newer than system to really install them
18:59 < rxr> find include -type f | xargs touch
18:59 < rxr>
18:59 < rxr> ^- does this work
19:00 < AStorm> Could use some paste :-)
19:00 < AStorm> Checking....
19:00 < rxr> just one line move
19:02 < AStorm> it seems it does work
19:02 < AStorm> yep, definitely does work
19:03 < AStorm> Downloading glibc:download/mirror/g/glibc-20061002.tar.bz2 ...
19:03 < AStorm> unrecognized!
19:03 < AStorm> INFO: download from mirror failed, trying original URL.
19:03 < AStorm> curl: (19) RETR response: 550
19:03 < AStorm> Something is broke with the server :>
19:03 < AStorm> or with mirror selection. How do I force it to run again?
19:05 < AStorm> Ah, now I know, just remove the Mirror-Cache file
19:05 < AStorm> :>
19:05 < AStorm> BTW, ever thought about updating the glibc?
19:05 < AStorm> It's half year old...
19:18 < AStorm> WTF?
19:18 < AStorm> !> /bin/pwd: couldn't find directory entry in `../../..' with matching i-node
19:18 < AStorm> !> /bin/pwd: couldn't find directory entry in `../../../..' with matching i-n ..
19:18 < AStorm> in glibc
19:18 < AStorm> on stage0
19:20 < AStorm> on configure
19:20 < AStorm> rxr: any idea?
19:20 < rxr> re
19:20 < AStorm> (except that t2-trunk dir is bind mounted)
19:20 < rxr> no idea about that nwe error
19:21 < rxr> we update the glibc on need, random updates usually break half of our architectures ...
19:21 < AStorm> >
19:21 < AStorm> :>
19:21 < rxr> and glibc is too much of a unwanted monster for comfortable debugging
19:21 < rxr> especially of dog slow CPU make runs ...
19:21 < AStorm> Who said about random? You do have the build farm :>
19:22 < AStorm> Check if a given version builds on all of these
19:22 < rxr> could be a bigger build farm ..
19:22 < CIA-18> rene * r23460 /trunk/package/base/linux-header/linux-header.conf:
19:22 < CIA-18> * fixed linux-header to not fail randomly, by making sure the .config
19:22 < CIA-18> exist before creating the include/asm link
19:23 < rxr> actually we have this snapshot because it was the time when the glibc 2.5.0 or what it is named was finalized but the usually Drepper moron did not wanted to spin a release tarball
19:23 < rxr> just keeping it secrent in Fedora Core as usual
19:23 < rxr> we pretty much touch glibc only if we have a need too
19:24 < rxr> but you can of course feel free to cross compile it to the major arches and propose an update, best with a list of new major features to know why we should update it :-)))
19:25 < CIA-18> susan * r23461 /trunk/package/mail/thunderbird/thunderbird.desc: * updated thunderbird (2.0b1 -> 2.0.0.0rc1)
19:26 < AStorm> rxr: crossing is easy...
19:26 < AStorm> checking if it runs is the other matter (right now at least)
19:26 < AStorm> I could check with sparc, x86 and x86_64
19:45 < AStorm> My C2D will love additional compiling... once I get the system up.
19:46 < AStorm> Ok, glibc ./configure hates bind mounts
19:47 < AStorm> uh.... why does the thing build the dietlibc even though I disabled it explicitly in the target?
19:47 < AStorm> (O = disable, yes?)
19:52 < rxr> grep for it in the config/.../package
19:52 < rxr> maybe it was forced on
19:52 < rxr> our initrd stuff relies on it by default
19:52 < rxr> when you do not want to give it a change to drop back on by default
19:52 < rxr> use not O, but -
19:53 < rxr> O is disable
19:53 < rxr> - is remove
19:53 < AStorm> Thanks :-)
19:53 < AStorm> Ay drat
19:54 < AStorm> I put X there :>
19:54 < AStorm> I have to just disable it per-package.
19:55 < AStorm> Because I do want it, just not build against it.
19:55 < rxr> the default setting is to build the stuff T2 embeds in the initrd by default with dietlibc
19:56 < rxr> unless you want your own initrd code that is a pretty good choice
19:56 < AStorm> Hmm, building with uclibc should work too, right?
19:56 < AStorm> (esp. statically)
19:56 -!- mtr [n=Michael@kobz-590cb1ea.pool.einsundeins.de] has quit [Read error: 60 (Operation timed out)]
19:56 < AStorm> The system will be static uclibc
19:56 < AStorm> So that should work anyway.
19:59 < AStorm> Anyway, I can build my own initramfs anyway
19:59 < AStorm> (whoops, duplicate words :P )
20:00 < AStorm> Where should I grep for the list of SDECFG_DIETLIBC_foo?
20:04 < rxr> config/.../config
20:04 < AStorm> :>
20:21 < rxr> so - cu
20:32 < AStorm> BTW - urxvt is smaller than nxterm
20:32 < AStorm> when built against dietlibc
21:23 < AStorm> Hmm, part of the sparc patch is already in that prerelease of gcc 4.2.0
21:24 < AStorm> Not part - the whole of the patch :>
21:28 < AStorm> Hmm, do I have to update all the "no-*" patches?
21:35 < sepp> no-* patches?
21:36 < AStorm> no-libstdc++
21:36 < AStorm> no-install-libiberty
21:36 < AStorm> libstdcpp-with-tag-cc
21:36 < AStorm> fixincl
21:37 < sepp> no-automated-way
21:37 < sepp> :)
21:37 < AStorm> :>
21:37 < AStorm> So I should do that
21:37 < AStorm> Maybe I'll be able to borrow some Gentoo ones :P
21:38 < AStorm> fixincl patch is a bit dumb :P
21:38 < sepp> they dont apply any longer?
21:38 < AStorm> libstdc++ doesn't with gcc 4.2.0
21:38 < AStorm> (the prerelease 20070316)
21:38 < sepp> rediff it
21:38 < AStorm> I'm in the process of it
21:39 < AStorm> fixincl is no longer needed AFAIK
21:39 < AStorm> They changed the build process to be less stupid
21:39 < AStorm> hmm
21:39 < AStorm> I'll have to change my C options after stage0 gcc
21:39 < AStorm> :>
21:39 < AStorm> after stage0 I mean
21:40 < AStorm> There should be a separate option for stage0
21:40 < AStorm> This gcc is uncontrolled :>
21:43 < AStorm> libstdc++ one is not necessary anymore
21:43 < AStorm> there's a configure option now
21:43 < AStorm> apply-protector.sh is not necessary either
21:43 < AStorm> stack protector is already in gcc
21:51 -!- emte_ [n=emte@d64-180-45-14.bchsia.telus.net] has quit [Read error: 60 (Operation timed out)]
21:57 < AStorm> Hmm, the no-libstdc++ is still "required"
21:58 < AStorm> I've provided a newer version
21:59 < AStorm> Hmm, is the libstdc++-tag patch really required?
22:00 < AStorm> New libtools shouldbe fixed... and libstdc++ isn't built in the first phase anywa
22:01 -!- emte_ [n=emte@d64-180-45-14.bchsia.telus.net] has joined #t2
22:01 < AStorm> I think the sparc niagara patch is also in
22:02 < AStorm> Yep, definitely, according to the changelog
22:07 < AStorm> the libiberty change should be applied using sed
22:07 < AStorm> Not as a patch that has to be fixed on each release
22:08 < AStorm> Instead of commenting, just cut the offender
22:14 < sepp> patches sent to the ml is welcome ;)
22:14 < sepp> are
22:14 < AStorm> I will send, just let me finish that gcc-4.2.0 prerelease port :>
22:15 < AStorm> I wonder if sed's .* is greedy or not
22:27 < AStorm> Hmm, it is
22:27 < AStorm> I have to use ranges :>
22:30 < AStorm> Like it?
22:30 < AStorm> sed -e "/install_to_libdir:/,/^$/ d" -i libiberty/Makefile.in
22:31 < AStorm> Instead of an ugly patch, some ranged removal
22:32 < AStorm> Wrong :>
22:32 < AStorm> It has to start at the _next_ line :>
22:45 < AStorm> Actually, I still may use that simple rule
22:45 < AStorm> then dump the rule back in at the end
22:46 -!- sepp_ [n=sepp@p85.212.43.175.tisdip.tiscali.de] has joined #t2
22:52 < AStorm> Oh, the gxx include dir patch is not necessary - there's a --with-gxx-include-dir option available
22:55 -!- sepp [n=sepp@p85.212.53.173.tisdip.tiscali.de] has quit [Read error: 60 (Operation timed out)]
23:24 -!- sepp [n=sepp@p213.54.7.213.tisdip.tiscali.de] has joined #t2
23:51 -!- sepp_ [n=sepp@p85.212.43.175.tisdip.tiscali.de] has quit [Read error: 110 (Connection timed out)]
--- Log closed Sun Apr 08 00:00:07 2007