--- Log opened Wed Feb 21 00:00:11 2007
00:07 < R4gnar0k> o_o
00:08 < R4gnar0k> rxr: the lx_version in linux 2.6.20.1 expands to 2.6.20.1 but the dir is still called 2.6.20 ... so how do I get the kernel dir in a clean way from my kernel module?
01:10 -!- emte [n=emte@d64-180-45-14.bchsia.telus.net] has quit [Read error: 60 (Operation timed out)]
01:15 -!- mtr_ [n=Michael@kobz-590cb027.pool.einsundeins.de] has joined #t2
01:20 -!- emte [n=emte@d64-180-45-14.bchsia.telus.net] has joined #t2
01:22 -!- idealm_ [n=ideal@CPE-203-45-99-98.nsw.bigpond.net.au] has joined #t2
01:27 -!- mtr [n=Michael@kobz-590cb5af.pool.einsundeins.de] has quit [Read error: 113 (No route to host)]
01:27 -!- mtr_ is now known as mtr
01:29 -!- idealm [n=ideal@CPE-203-45-99-98.nsw.bigpond.net.au] has quit [Read error: 60 (Operation timed out)]
01:40 -!- emte_ [n=emte@d64-180-45-14.bchsia.telus.net] has joined #t2
01:42 -!- emte [n=emte@d64-180-45-14.bchsia.telus.net] has quit [Read error: 110 (Connection timed out)]
02:25 -!- non2 [n=dc36734@213.144.157.75] has joined #t2
02:25 < non2> hi
02:30 < non2> anyone alive?
02:31 < R4gnar0k> a bit
02:32 < non2> may i ask somethin about eDesk which is based on t2?
02:33 < R4gnar0k> you can try tho I'm not familiar with eDesk in particular
02:34 < non2> maybe its a general question.. is it a default kde, what it has.. as i can build it on my own with the SDE.. how could i add patches to t2?
02:35 < non2> i mean t2 packages (which are mostly the original source packages as i understood)
02:35 < R4gnar0k> you copy a file called .patch into the package directory
02:35 < R4gnar0k> that file will get antomatically applied when the package is built
02:36 < non2> u mean diff files or a shellscript?
02:36 < R4gnar0k> no, a patch file. should apply with patch -p1
02:36 < R4gnar0k> sorry patch -p1 <
02:37 < non2> so diff-file
02:37 < R4gnar0k> yes
02:37 < non2> another question..
02:38 < non2> what are those characters X 01---5---- 000.000 base 00-dirtree 0000 / base/system DIETLIBC 0
02:38 < non2> X stands for install
02:38 < non2> mwhat is the rest?
02:38 < R4gnar0k> 01---5---- ... the stages where the package is built
02:39 < R4gnar0k> 000.000 build priority
02:39 < non2> so usually i dont have to change that, except i add new packages..?
02:39 < R4gnar0k> yea
02:40 < non2> cool
02:40 < R4gnar0k> or if the package now depends on another package
02:41 < non2> i m trying to compile a customized eDesk
02:42 < non2> are there any general hints for that?
02:43 < R4gnar0k> depends on what you want to customize
02:44 -!- sepp_ [n=sepp@p85.212.16.19.tisdip.tiscali.de] has joined #t2
02:44 < non2> i ll add some packages + remove some.. what i d like to add is also bootsplash and maybe i ll need some kernel modules (for mp3 player, but i m not sure now)..
02:45 < R4gnar0k> ah ok ... well that can mostly be done with the package selection
02:45 < R4gnar0k> and adding your own package for the kernel module ... there are several packages that build kernel mods
02:47 < non2> java wont be compiled right?
02:47 < R4gnar0k> hm?
02:48 < R4gnar0k> there's an option for the gcc package that disables the gcc java part ... and you'll also have to disable java/* in your pkgsel
02:49 < non2> i speak about sun java.. as i read compiling the jre needs some closed source stuff..
02:50 < R4gnar0k> that's not built by default tho
02:50 -!- sepp [n=sepp@p213.54.150.198.tisdip.tiscali.de] has quit [Read error: 145 (Connection timed out)]
02:50 < non2> so one can also package binaries into t2?
02:50 < non2> how is that done?
02:50 < R4gnar0k> yea
02:51 < R4gnar0k> check out the sun packages ... just basically have to either make a custmain script to copy stuff into the right locations, or invoke the binaries installer (like with the nvidia package)
02:52 < non2> ok cool
02:53 < non2> its hard to edit the packages list with a normal text editor.. do u kno somethin, where i can filter the lines?
02:54 < R4gnar0k> hm? nono don't edit the packages list itself ... use the pkgsel file
02:54 < R4gnar0k> the packages list will get overwritten each time you run Config
02:55 < non2> where should that pkgsel file be?
02:56 < R4gnar0k> config//pkgsel
02:56 < non2> i have no such a file
02:56 < R4gnar0k> you need to enable 'Custom package selection'
02:58 < non2> ok i have that file now but its empty..
02:58 < R4gnar0k> yea
02:58 < R4gnar0k> now you put lines of this format into it:
02:59 < R4gnar0k> X
02:59 < R4gnar0k> enables that package
02:59 < R4gnar0k> O
02:59 < R4gnar0k> makes it optional ... the package won't be built by Build-Target, but you can later Emerge-Pkg in
02:59 < R4gnar0k> it*
02:59 < non2> O gtk-engines for example?
02:59 < R4gnar0k> -
03:00 < R4gnar0k> removes the package completely
03:00 < R4gnar0k> yea
03:00 < non2> i dont see the difference between - and O
03:01 < R4gnar0k> as I said, you can emerge packages marked with O later ... packages with '-' will not exist for the build system any more, so you can't emerge them ... they also won't be included in dependency checks
03:02 < non2> i see
03:02 < R4gnar0k> so if you emerge something that needs gtk-engines to build, Emerge-Pkg will build gtk-engines before building that package IF you have it marked as O
03:03 < non2> is there any advantage of usin "-" ?
03:03 < R4gnar0k> for example, I have '- pam' in my pkgsel to build a non-pam system
03:04 < R4gnar0k> if I just had 'O' every emerge I did would build pam
03:04 < non2> i see..
03:05 < non2> do u kno if trunks would build now?
03:05 < R4gnar0k> atm? mostly
03:06 < non2> if i had done my package changes direct in that file "packages", would it also work ?
03:07 < non2> i wouldnt use the Config script afterwards..
03:07 < R4gnar0k> well it would be overwritten each time you run Config ...
03:07 < R4gnar0k> otherwise, it should
03:08 < non2> i hope it wont need more than 1 day to build it all :)
03:08 < R4gnar0k> good luck :)
03:08 < R4gnar0k> brb
03:09 < non2> ty for ur answers
03:09 < non2> advices
03:09 < R4gnar0k> yw :)
03:09 < non2> cu
03:10 -!- non2 [n=dc36734@213.144.157.75] has quit [Remote closed the connection]
03:22 -!- emte_ [n=emte@d64-180-45-14.bchsia.telus.net] has quit [Read error: 110 (Connection timed out)]
03:33 -!- emte [n=emte@d64-180-45-14.bchsia.telus.net] has joined #t2
04:04 -!- emte_ [n=emte@d64-180-45-14.bchsia.telus.net] has joined #t2
04:06 -!- emte [n=emte@d64-180-45-14.bchsia.telus.net] has quit [Read error: 110 (Connection timed out)]
04:26 -!- emte_ [n=emte@d64-180-45-14.bchsia.telus.net] has quit [Read error: 60 (Operation timed out)]
04:34 -!- emte_ [n=emte@d64-180-45-14.bchsia.telus.net] has joined #t2
05:08 -!- sepp_ [n=sepp@p85.212.16.19.tisdip.tiscali.de] has quit [Read error: 145 (Connection timed out)]
05:10 -!- sepp [n=sepp@p85.212.38.34.tisdip.tiscali.de] has joined #t2
06:04 -!- emte_ [n=emte@d64-180-45-14.bchsia.telus.net] has quit [Read error: 110 (Connection timed out)]
06:05 -!- emte_ [n=emte@d64-180-45-14.bchsia.telus.net] has joined #t2
06:39 -!- emte__ [n=emte@d64-180-45-14.bchsia.telus.net] has joined #t2
06:40 -!- emte_ [n=emte@d64-180-45-14.bchsia.telus.net] has quit [Read error: 110 (Connection timed out)]
07:15 -!- smarTUX [n=rvdalen@dsl-242-94-130.telkomadsl.co.za] has joined #t2
07:15 < smarTUX> hi everyone
07:23 -!- emte_ [n=emte@d64-180-45-14.bchsia.telus.net] has joined #t2
07:24 -!- emte__ [n=emte@d64-180-45-14.bchsia.telus.net] has quit [Read error: 110 (Connection timed out)]
08:19 < mtr> re
08:29 < rxr> re
08:29 < rxr> moin mtr and smarTUX
08:30 < mtr> hi rxr
08:32 < rxr> R4gnar0k: which module dire is not named correctly ?
08:32 < R4gnar0k> hm?
08:32 < R4gnar0k> d80211 doesn't find the kernel dir
08:32 < rxr> R4gnar0k: gor 2.6.20.1 I get: /lib/modules/2.6.20.1-dist/
08:33 < R4gnar0k> no
08:33 < R4gnar0k> the source dir
08:33 < rxr> R4gnar0k: maybe it must be supplied with some non standard makefile argument ?
08:33 < R4gnar0k> where it's built
08:33 < R4gnar0k> no no. look at the file for a second
08:34 < R4gnar0k> hook_add prepatch 1 "( tar xfj $pkg_d80211_tgz && cd d80211-${pkg_d80211_ver} && make KSRC=$builddir/linux-$ver && make KSRC=$builddir/linux-$ver patch_kernel )"
08:34 < R4gnar0k> $ver evaluates to 2.6.20.1 ... the dir is named 2.6.20
08:39 < rxr> heh
08:39 < rxr> KSRC=$kerneldir
08:39 < rxr> :-)
08:40 < rxr> $kerneldir and $moduledir are your friends
08:42 < R4gnar0k> oh duh lol
08:42 < R4gnar0k> thanks
08:43 < rxr> you'r welcome :-)
08:45 < R4gnar0k> hmm no
08:45 < R4gnar0k> that evaluates to /lib/modules/2.6.18-inside-sandbox/build in stage 1
08:45 < R4gnar0k> not good ;P
08:45 < rxr> kerneldir ?
08:45 < R4gnar0k> yea
08:45 < rxr> package/base/linux24/linux24.conf: kerneldir="$PWD"
08:46 < R4gnar0k> o_O
08:46 < rxr> this wireless stack is inside the new intel package in trunk?
08:46 < R4gnar0k> maybe cuz this is a pre-?
08:46 < rxr> I can I give it a try ?
08:46 < R4gnar0k> no extra package
08:46 < R4gnar0k> sure
08:46 < R4gnar0k> the package is names d80211
08:47 < R4gnar0k> named*
08:48 < CIA-21> rene * r22622 /trunk/package/mail/getmail/getmail.desc: * updated getmail (4.7.1 -> 4.7.2)
08:48 < CIA-21> rene * r22623 /trunk/package/filesystem/ntfs-3g/ntfs-3g.desc: * updated ntfs-3g (0.20070207-RC1 -> 1.0)
08:48 < CIA-21> rene * r22624 /trunk/package/x11/rxvt-unicode/rxvt-unicode.desc: * updated rxvt-unicode (8.1 -> 8.2)
08:48 < CIA-21> rene * r22625 /trunk/package/fonts/dejavu-ttf-fonts/dejavu-ttf-fonts.desc: * updated dejavu-ttf-fonts (2.14 -> 2.15)
08:49 < rxr> with 2.6.20 this d80211 did built ?
08:50 < rxr> what is was the new driver name for the 3945 ?
08:53 < CIA-21> susan * r22626 /trunk/package/network/dnrd/dnrd.desc: * updated dnrd (2.20.1 -> 2.20.2)
08:53 < CIA-21> susan * r22627 /trunk/package/develop/roundup/roundup.desc: * updated roundup (1.3.2 -> 1.3.3)
08:54 < CIA-21> susan * r22628 /trunk/package/dictionary/kanatest/kanatest.desc: * updated kanatest (0.4.0 -> 0.4.1)
08:54 < CIA-21> susan * r22629 /trunk/package/mail/dcc/dcc.desc: * updated dcc (1.3.45 -> 1.3.52)
08:55 < R4gnar0k> hmm iwlwifi
08:56 < R4gnar0k> it did build with 2.6.20 yea
08:57 < R4gnar0k> but the 3945 driver needs the d80211 framework
08:58 < CIA-21> rene * r22630 /trunk/package/e16/enlightenment16/enlightenment16.desc: * updated enlightenment16 (0.16.8.5 -> 0.16.8.6)
08:58 < rxr> R4gnar0k: ok I add them to the desktop target and give them a try
08:58 < R4gnar0k> ok
09:02 < rxr> I guess you vanish for some sleep now anyway?
09:03 < R4gnar0k> maybe lol
09:03 < R4gnar0k> unless you still need something?
09:06 < rxr> nope
09:06 < rxr> I have to do some work :-) I'll start the kernel rebuild when that is finished :-)
09:06 < rxr> just wanted to find out it is urgent :-)
09:09 < R4gnar0k> lol ok
09:13 < CIA-21> rene * r22631 /trunk/package/kde/dolphin/ (. dolphin.cache dolphin.desc): * added dolphin (0.8.2) - Dolphin is a file manager for KDE
09:17 -!- emte_ [n=emte@d64-180-45-14.bchsia.telus.net] has quit [Read error: 110 (Connection timed out)]
09:30 -!- nacho_macho [n=dc36734@213.144.157.75] has joined #t2
09:30 < nacho_macho> Hi
09:32 < nacho_macho> I cant compile trunks, the problem i have -->
09:32 < nacho_macho> At least one of the 3rd party kernel modules failed!
09:32 < nacho_macho> --- BUILD ERROR ---
09:32 < rxr> yes, this always happens after a major linux26 update
09:33 < nacho_macho> what would u recommend?
09:33 < rxr> you can either fix those third party module errors, ignore the error and continue the build, or use the stable 6.0 branch - where everything and his aount tilly builds fine :-)
09:33 < rxr> well, I use trunk for all development - I would look if I care about the modules that are printed out that failed are important to me and if not continue the build
09:33 < rxr> if they are I would go to the source and patch the third party kernel modules that are important to me ...
09:34 < nacho_macho> the problem is i cant find the problematic driver..
09:34 < nacho_macho> >>Due to previous errors, no 1-linux26.log file!
09:34 < rxr> they are printed out a bit above that message
09:35 < rxr> like Failed building ... d8211 ...
09:36 < nacho_macho> oo rt2500.ko
09:36 < nacho_macho> quite important i d say :(
09:37 < rxr> only if you have that NIC ...
09:39 < rxr> when you want to continue the build touch the log files in build/.../var/adm/logs/1-linux26.log
09:39 < rxr> and rerun Build-Target
09:51 < nacho_macho> i think i should use stable, as rt2500 is one of the most known wlan adapters, if that doesnt build other stuff might not work too..
09:52 < rxr> well the whole kernel builds of course
09:52 < rxr> this are just some non-kernel-intern add-ons
09:54 < nacho_macho> you kno its my first try, i m trying to compile eDesk with a few customizngs..
09:57 < nacho_macho> is there a binary iso of eDesk somewhere?
10:01 < rxr> yes 6.0.x
10:02 < rxr> http://osiris.tfh-berlin.de/mirror/t2/stable/6.0.0/
10:02 < nacho_macho> i need only the first cd for the installation right?
10:04 < rxr> nope both
10:05 < rxr> the files are currently alphabetically ordered ...
10:05 < rxr> x stuff is at the end ... :-(
10:05 < rxr> we always wanted to improve this ... :-)
10:07 < nacho_macho> hmm when i build the edesk target, do i get exactly the same?
10:10 < rxr> when you build in 6.0, yes
10:10 < rxr> when you build trunk you get a way updated flavour
10:11 < rxr> with some changes and fixes
10:11 < nacho_macho> dont those fixes go also into stable?
10:14 < rxr> yes, sure
10:15 < rxr> I meant changes such as more wlan drivers, deselected packages due seldome use
10:15 < rxr> soem additions due common request
10:15 < nacho_macho> i see
10:15 < rxr> and I think trunk also enfoces space optimization (-Os) and creation of a livecd
10:16 < nacho_macho> so there are 2 builds in the end? installcd+ livecd?
10:17 < rxr> nope, you select what you want
10:17 < rxr> however of course you can include an installer in the livecd that can copy the system onto HD
10:18 < nacho_macho> i asked as u said it enforces..
10:18 < nacho_macho> is there such an installer?
10:21 < rxr> nacho_macho: yes, normally the T2 ./scripts/Config allows all sorts of choices
10:21 < rxr> the trunk "ExactDesktop" more and more becomes a well-defined product, thus it enforces livecd as it becomes more and more dependant on beeing a livecd at the end
10:22 < rxr> however with just the generic target you can still select install or live - or do whatever you want in your own targ
10:22 < rxr> et
10:22 < rxr> e.g. I'm working right now on a custom target that does a whole lot custom postprocessing at the end ...
10:22 < rxr> not using either install or live
10:22 < rxr> T2 is all about the flexible System Development Environemt / Build System ...
10:35 < nacho_macho> i have an error in stable now :(
10:35 < nacho_macho> malloc: ../bash/subst.c:4135: assertion botched
10:35 < nacho_macho> realloc: start and end chunk sizes differ
10:35 < nacho_macho> Aborting...scripts/config.in: line 124: 19139 Abgebrochen
10:39 < nacho_macho> the packages file is empty!
10:40 < rxr> ouh
10:40 < rxr> interesting
10:40 < rxr> never saw that, what distribution are ou building on ?
10:41 < nacho_macho> debian/archlinux mix
10:41 < nacho_macho> debian with archlinux's package manager (pacman)
10:41 < rxr> appears your bash is buggy or in some way incompatible from the rest of the world
10:41 < rxr> well - if trunk worked otherwise maybe just stay with it
10:42 < nacho_macho> trunk started to build at least..
10:48 < rxr> trunk has a new bash plugin infrastructure
10:49 < rxr> instead of a custom C code plugin we embed the Lua language into the bash to allow more flexible and straight forward scripting
11:12 < rxr> yeah - go for trunk
11:12 < rxr> we can help fix current build regressions on the wa
11:12 < rxr> y
11:18 < nacho_macho> i m doin that..
11:30 -!- emte [n=emte@d64-180-45-14.bchsia.telus.net] has joined #t2
11:32 < CIA-21> rene * r22632 /trunk/package/network/d80211/ (d80211.conf d80211.desc pkg_linux26_pre.conf): * copyright and checksum patfh for d80211
11:34 < CIA-21> rene * r22633 /trunk/package/network/iwlwifi/ (4 files): * copyright and checksum patch for iwlwifi
11:49 < rxr> R4gnar0k: I think the error you got was because the $kerneldir was expanded while defining the hook, quoted in ' instead of " should wokr
11:49 < rxr> R4gnar0k: /me testing right now
11:51 < nacho_macho> would "O gnome2" work? i mean i dont want to install anything from gnome2 category
11:53 < rxr> O gnome2/* would work
11:53 < rxr> it would deselect all in the gnome2 repository
11:53 < rxr> including glib and gtk+
11:54 < rxr> if you need just those you would need to add a X glib, X gtk+ after the O gnome2/*
11:54 < nacho_macho> but if i add "X inkscape" it would install them anyway right?
11:55 < nacho_macho> so no need to add extra glib/gtk+ entries?
11:55 < rxr> when you want the Build-Target to build gtk+ for your inkscape you need to manually add them
11:55 < rxr> if you use Emerge-Pkg for an "in-system" installation Emerge would propose to instasll glib/gtk+ if you do not hinder it from doing so
11:56 < nacho_macho> with "in-system" u mean an installed system?
11:57 < rxr> yes
11:58 < rxr> after installation
11:58 < rxr> Emerge-Pkg builds the package into the running system
11:58 < rxr> Build-Target builds the package inside a fresh sandbox seperated from the host system
12:01 < nacho_macho> i think about usin a binary package management for the installed system.. it d be cool if the Package-Selection for Target had also a kind of dependency solver..
12:04 < rxr> your target will be deployed to "clients" or so ?
12:04 < rxr> T2 has automatic dependency tracking, the deps are tracked during the build and stored in the .cache files
12:05 < nacho_macho> i m my own "client" :)
12:05 < rxr> "your installer" can use those
12:05 -!- CIA-8 [i=cia@cia.navi.cx] has quit []
12:05 < nacho_macho> its just for fun
12:05 -!- CIA-21 [i=cia@cia.navi.cx] has quit []
12:07 -!- CIA-8 [i=cia@cia.navi.cx] has joined #t2
12:07 -!- CIA-6 [i=cia@cia.navi.cx] has joined #gsmp
12:08 < nacho_macho> i must say i like it
12:08 -!- ideal_ [n=ideal@203.45.99.98] has joined #t2
12:08 [Users #t2]
12:08 [@ChanServ] [ Demian] [ idealm_] [ nacho_macho] [ sepp ] [ Stelz ]
12:08 [ Capey ] [ emte ] [ LMJ ] [ R4gnar0k ] [ smarTUX ] [ valentin]
12:08 [ CIA-8 ] [ ideal_] [ mtr ] [ rxr ] [ sparc-kly]
12:08 -!- Irssi: #t2: Total of 17 nicks [1 ops, 0 halfops, 0 voices, 16 normal]
12:08 < rxr> nacho_macho: :-)
12:08 < rxr> hi ideal_
12:09 -!- idealm_ [n=ideal@CPE-203-45-99-98.nsw.bigpond.net.au] has quit [Read error: 110 (Connection timed out)]
12:09 < ideal_> Hi rxr
12:10 < CIA-8> rene * r22634 /trunk/package/base/linux24/linux24.conf:
12:10 < CIA-8> * define kerneldir and moduledir earlier during kernel builds,
12:10 < CIA-8> so extending packages also can use these variables earlier as well
12:10 < CIA-8> rene * r22635 /trunk/package/network/d80211/pkg_linux26_pre.conf:
12:10 < CIA-8> * fixed d80211 to not use hardcoded linux directory assumptions but
12:10 < CIA-8> the $kerneldir variable instead
12:18 -!- sparc-kly [n=mubex@adsl-72-50-101-246.prtc.net] has quit [Read error: 145 (Connection timed out)]
12:19 < CIA-8> sebastian * r22636 /trunk/package/graphic/pixie/ (pixie.conf pixie.desc): * updated pixie (1.7.6 -> 2.0.2)
12:22 < sepp> moin :)
12:22 < rxr> moin moin sepp
12:22 < rxr> ^- builds with x86-64
12:22 < rxr> ?
12:22 < sepp> i know :)
12:22 < sepp> yes, does
12:23 < sepp> hi rxr
12:23 < rxr> cool
12:23 < sepp> don't have the x86 standing around any longer ...
12:23 < sepp> was just collecting dust
12:24 < rxr> :-)
12:24 * rxr is happy either that I got rid of the 32bit MacBook ...
12:29 < nacho_macho> do u think 2GB will be enough for a ~500MB sized installer CD?
12:29 < sepp> for / ?
12:31 < sepp> only thing that does not work well on amd64 is the squeak vm ...
12:31 < nacho_macho> sepp: i m compiling a customized eDesk, it ll be about 500MB.. with all the sources and bin outputs, do i need more than 2GB hd space?
12:31 < rxr> 2GB would be nearly filled I estimate
12:32 < sepp> guess a little bit more space would be good
12:32 -!- ideal_ [n=ideal@203.45.99.98] has quit [Remote closed the connection]
12:40 < CIA-8> rene * r22637 /trunk/package/base/linux24/linux24.conf:
12:40 < CIA-8> * ok, the linux version is not valid as early as required with r22634,
12:40 < CIA-8> move kerneldir/moduledir to the earliest place where it is well defined
12:48 < CIA-8> rene * r22638 /trunk/package/base/linux24/linux24.conf:
12:48 < CIA-8> * ok, ok, d80211 needs the $kerneldir really early, split the
12:48 < CIA-8> definitions up, so the initially known $kerneldir is defined
12:48 < CIA-8> early on and the $moduledir just when the patching is done
13:37 < CIA-8> rene * r22639 /trunk/package/network/iwlwifi/ (iwlwifi.conf iwlwifi.desc postlinux.conf): (log message trimmed)
13:37 < CIA-8> * decustmained iwlwifi
13:37 < CIA-8> * fixed iwlwifi to actually build by moving the real drive
13:37 < CIA-8> source as the first [D] tag so it is unpacked durnig the
13:37 < CIA-8> kernel buidl and custom extract the ucode in the stand
13:37 < CIA-8> alone build
13:37 < CIA-8> * simplified the postlinux makeopt definition and derive
14:26 < nacho_macho> == 12:08:23 =[0]=> Building base/linux-header [2.6.19.1 7.0-trunk].
14:26 < nacho_macho> == 13:31:58 =[1]=> Building base/linux26 [2.6.20.1 7.0-trunk].
14:26 < nacho_macho> this takes really long.
14:32 < rxr> when you don't change something it builds as many modules possible
14:32 < rxr> and some third party stuff ...
14:34 < nacho_macho> how many days will it take :??
14:41 < sepp> depends on your hardware, unless you build on a 5 year old box less than a day
14:43 < rxr> I was speaking about just the linux kernel
14:43 < rxr> it is one of the longest things to build
14:43 < rxr> the few things that take longer are OpenOffice (8h++), gcc (with java), kde packages, ...
14:44 < sepp> i mean some "normal desktop" that fits onto 1 cd ...
14:48 < rxr> yep, I figured
14:48 < rxr> just wanted to clarify my statement :-)
14:48 [Users #t2]
14:48 [@ChanServ] [ Demian] [ mtr ] [ rxr ] [ Stelz ]
14:48 [ Capey ] [ emte ] [ nacho_macho] [ sepp ] [ valentin]
14:48 [ CIA-8 ] [ LMJ ] [ R4gnar0k ] [ smarTUX]
14:48 -!- Irssi: #t2: Total of 14 nicks [1 ops, 0 halfops, 0 voices, 13 normal]
15:23 < nacho_macho> rxr: do u have a list of non-working kernel modules? till now i have found 3 of them :(
15:28 -!- smarTUX [n=rvdalen@dsl-242-94-130.telkomadsl.co.za] has quit [" HydraIRC -> http://www.hydrairc.com <- Leading Edge IRC"]
15:28 < rxr> *> Failed building acx100.
15:28 < rxr> *> Failed building adm8211.
15:28 < rxr> *> Failed building at76c503a.
15:28 < rxr> *> Failed building madwifi.
15:28 < rxr> *> Failed building rt2500.
15:28 < nacho_macho> ok ty
15:35 -!- emte_ [n=emte@d64-180-45-14.bchsia.telus.net] has joined #t2
15:36 -!- emte [n=emte@d64-180-45-14.bchsia.telus.net] has quit [Read error: 110 (Connection timed out)]
15:42 -!- nacho_macho [n=dc36734@213.144.157.75] has quit [Remote closed the connection]
16:29 < Capey> hmm, what should define _GNU_SOURCE?
16:32 < rxr> hm - depends
16:32 < rxr> Capey: where are you stuck - what are you trying to build ?
16:32 < Capey> embutils
16:32 < Capey> its missing __USE_GNU which in turn needs _GNU_SOURCE to be enabled
16:33 < Capey> but i patched embutils to add -D_GNU_SOURCE when compiling
16:33 < Capey> seems to be quite "popular" case with uClibc
16:33 < rxr> embutils with wich c librares?
16:34 < rxr> I think embutils is not uclibc compatible
16:34 < Capey> hmm, i'll try this patch, if not, then i need to drop embutils package
16:35 < Capey> uClibc specific patches goes to package/base/uclibc/patches, right?
16:35 < rxr> yes
16:35 < rxr> maybe just use busybox for now then
16:36 < rxr> and only when v1 works try dietlibc/embutils later on ...
16:36 < Capey> whats the naming schema on this dir?
16:36 < Capey> for automatic file patching
16:42 < rxr> we really should unify that :-)
16:42 < rxr> just $pkg-something.patch
16:42 < Capey> yep
16:42 < rxr> I recently also fall into this uclibc trap
16:42 < rxr> just $pkg.patch is not enough there
16:42 < rxr> it needs some -text in the name
16:43 < rxr> as the code does $pkg-*.patch or so ...
16:52 -!- emte_ [n=emte@d64-180-45-14.bchsia.telus.net] has quit [Connection timed out]
16:54 -!- emte_ [n=emte@d64-180-45-14.bchsia.telus.net] has joined #t2
17:13 < CIA-8> rene * r22640 /trunk/target/share/livecd/rootfs/ (etc/inittab sbin/):
17:13 < CIA-8> * removed livecd/rootfs/sbin/init.d/system patch as well as etc/inittab
17:13 < CIA-8> as this is partially outdated, and sysv dependant and often not needed
17:13 < CIA-8> at all: let the target do this pre-configuration
17:13 -!- emte_ [n=emte@d64-180-45-14.bchsia.telus.net] has quit [Read error: 110 (Connection timed out)]
17:13 -!- emte_ [n=emte@d64-180-45-14.bchsia.telus.net] has joined #t2
17:21 -!- sepp_ [n=sepp@p85.212.48.195.tisdip.tiscali.de] has joined #t2
17:29 -!- sepp [n=sepp@p85.212.38.34.tisdip.tiscali.de] has quit [Read error: 145 (Connection timed out)]
17:54 < Capey> wow
17:55 < Capey> uClibc gets wrong configure options
17:55 < Capey> * uClibc Configuration
17:55 < Capey> *
17:55 < Capey> Target Architecture
17:55 < Capey> choice[1-23?]: 9
17:55 < Capey> which is i386
17:58 < rxr> .oO
17:58 < rxr> on my latest blackfin testbuild that works
17:58 < rxr> worked
17:58 < rxr> the .config based configuration of uclibc is a bit gragile
17:58 < rxr> fragile even
18:08 < Capey> ups, my bad... it does work
18:09 < Capey> its just outputs the its own config info many time and first is i386, but after that it includes the .config options defined by T2 and works
18:09 < rxr> puh
18:10 < Capey> but i still have a problem with it, it seems that it does not build libdl and openssl fails for that reason
18:11 < Capey> build or install that is
18:13 < rxr> hm
18:17 < LMJ> moin moin
18:22 < rxr> moin moin LMJ
18:22 < LMJ> how was the day today ?
18:23 < rxr> over busy as usual
18:24 < LMJ> hehe
18:25 < LMJ> I hope you will be able to handle it ;)
18:37 < rxr> handle it since years *g*
18:38 < LMJ> I updated some details on the Website layout since the email I sent to you
18:39 < CIA-8> rene * r22641 /trunk/ (package/base/mkinitrd/initrdinit.sh target/share/livecd/init):
18:39 < CIA-8> * more compact static module loading in the system and live
18:39 < CIA-8> initrd, also hiding errors as not all modules necessarily
18:39 < CIA-8> need to be available
18:44 -!- emte_ [n=emte@d64-180-45-14.bchsia.telus.net] has quit [Read error: 110 (Connection timed out)]
19:24 < R4gnar0k> rxr: so did you get it to work?
19:25 < rxr> R4gnar0k: yes
19:28 < R4gnar0k> hmmm ... that's exactly what I had and didn't work for me o_o
19:33 < rxr> yeah - kerneldir was undefined at that place
19:33 < rxr> I moved the definition up in linux-conf.in
19:34 < R4gnar0k> oh ok :p
19:35 < R4gnar0k> what was the custmain change about? o_o
19:36 -!- d-marc [n=d-marc@HSI-KBW-091-089-007-026.hsi2.kabelbw.de] has joined #t2
19:39 < rxr> custmain is dreprecated
19:39 < rxr> no new packages should use it
19:39 < rxr> it disables all hook and thus cripples target modifications
19:39 < rxr> the plan is to get rid of all custmain users and then remove the 5 lines implementing it
19:39 < rxr> custom can easily be replaced by:
19:39 < rxr> makeopt=
19:39 < rxr> makeinstopt=
19:39 < rxr> ^- disable make runs and co
19:40 < rxr> and then use a hook to inject the code
19:40 < R4gnar0k> ok ... o-o is that documented somewhere?
19:40 < rxr> well - custmain even was deprecated at ROCK Linux times ...
19:40 < rxr> no idea if we have that somewhere .-)
19:40 < rxr> I guess it is time for a
19:40 < R4gnar0k> :p lol
19:40 < R4gnar0k> maybe print a warning message when it's set :P
19:41 < rxr> [ "$custmain" ] && echo_warning "The use of custmain is deprecated, please use hooks instead."
19:41 < R4gnar0k> ok
19:41 < rxr> there are only 118 packages using it
19:42 < rxr> maybe we even can do a short sprint and remove it before 7.0
19:42 < rxr> hm - ok some packages do custmain="true" to build nothing
19:42 < rxr> that does not count :-)
19:42 < Stelz> w00h00
19:43 < Stelz> wesnoth 1.2.2 released :-)
19:43 < R4gnar0k> lol
19:47 < rxr> *> The use of custmain is deprecated, please use hooks instead.
19:47 < rxr> any typos ?
19:47 < R4gnar0k> not that I see
19:48 < CIA-8> rene * r22642 /trunk/scripts/functions.in: * added a warning note that the ust of custmain is deprecated
19:48 < R4gnar0k> uhm
19:48 < R4gnar0k> WARNING: Couldn't open directory /lib/modules/2.6.20.1-dist: No such file or directory
19:48 < R4gnar0k> FATAL: Could not open /lib/modules/2.6.20.1-dist/modules.dep.temp for writing: No such file or directory
19:48 < R4gnar0k> for the imlwifi module
19:49 < rxr> hm
19:49 < rxr> hm
19:49 < rxr> d80211 built fine?
19:49 < R4gnar0k> seems so
19:49 < rxr> this is just the postlinux built driver ?
19:49 < R4gnar0k> yea
19:49 < rxr> hm
19:49 < rxr> it built here
19:49 < rxr> does the dir exist?
19:49 < rxr> is this stage 1 ?
19:50 < R4gnar0k> yea
19:51 < R4gnar0k> it doesn't
19:54 < d-marc> hi!
19:55 < d-marc> does anybody remember how I find out how many bytes are available for reading on a file descriptor?
19:55 < rxr> hm
19:56 * rxr wonders if that was possible at all
19:56 < rxr> what do you want to do?
19:56 < rxr> just read what is there?
19:56 < rxr> maybe you want non blocking i/o ?
19:56 < d-marc> after select returns with data available, I want to know how many bytes I can read
19:57 < d-marc> yes, actually I want non-blocking I/O
19:57 < R4gnar0k> you can't afaik
19:58 < d-marc> Under win32 there is a ioctl
19:58 < R4gnar0k> just set the fd to nonblocking and read like 4096 bytes
19:58 < d-marc> yes, I know
19:58 < R4gnar0k> it will return less when there's less then 4096 bytes available
19:58 < d-marc> right
19:59 < d-marc> unfortunately someone implemented a availableBytes() function that I have to port to posix
19:59 < R4gnar0k> ouch >.<
19:59 < rxr> d-marc: let me take a look
19:59 < d-marc> thanks
20:01 < rxr> hm, ok, quick look did not reveal a match
20:01 < rxr> did you consult Google ?
20:05 < rxr> do you only work on files ?
20:05 < rxr> (in the area you imlement?
20:05 < rxr> you cou ld seek to the end and return the difference of the position ... :-)
20:07 < d-marc> not really
20:07 < rxr> google is not helpful ?
20:07 < d-marc> dont know
20:07 < d-marc> there is a ioctl for wsa sockets with FNIOREAD
20:12 < d-marc> some IODevices in other libraries have dataAvailable events that let you know how many bytes you can read without to block, but I dont know if they simple double buffer the data
20:12 < R4gnar0k> in unix libs? o_o
20:13 < d-marc> yes
20:13 < d-marc> FIONREAD exists for unix too I just found out
20:16 < d-marc> FIONREAD - Determines the number of bytes that are immediately available to be read on a file descriptor.
20:17 < R4gnar0k> o_o
20:17 < R4gnar0k> never heard of it
20:18 < d-marc> hehe
20:26 -!- sepp_ [n=sepp@p85.212.48.195.tisdip.tiscali.de] has quit [Read error: 145 (Connection timed out)]
20:26 < R4gnar0k> amazing lol :) well never stop to learn something new lol
20:26 < R4gnar0k> does it work?
20:27 < d-marc> dont know yet
20:27 < R4gnar0k> lol
20:28 < R4gnar0k> hm http://gcc.gnu.org/ml/java/1999-q3/msg00006.html
20:30 < rxr> X-/
20:30 < rxr> big fun (tm)
20:31 < R4gnar0k> lol
21:01 < CIA-8> rene * r22643 /trunk/package/graphic/pixie/pixie.cache: * updated pixie.cache
21:19 < R4gnar0k> rxr: so how to fix that error? o_o
21:22 < rxr> R4gnar0k: it is stage 1 ?
21:23 < rxr> the I guess just some $root missing somewhere :-)
21:23 < R4gnar0k> yea it is
21:25 < rxr> you wait me to find it, right?
21:25 < R4gnar0k> aparently lol
21:26 < R4gnar0k> if you don't have time I can try too ... I have a suspect
21:29 < CIA-8> rene * r22644 /trunk/target/share/livecd/init:
21:29 < CIA-8> * fixed live/init to translate all fat variants to vfat, not just fat32
21:29 < CIA-8> * improved it to wait 15s instead just 10s as we saw some USB devices
21:29 < CIA-8> to require slightly more time to settle on the USB bus
21:29 < CIA-8> * allow specifing the live image name on the kernel commandline via
21:29 < CIA-8> live= (defaults to the good old live.squash)
21:29 < CIA-8> * more precise diagnostics, with verbosity increasing over the time
21:30 < rxr> R4gnar0k: /me? time??
21:30 < rxr> I do not have a stage 1 build around this time anyway
21:30 < rxr> and I'm still hanging in the office anyway
21:30 < rxr> as I have a US customer with some toucle ticket in the mail loop right now ...
21:30 < rxr> need some food and sleep sometime soon
21:30 < R4gnar0k> ok :) I'll try
21:30 < rxr> s/toucle/trouble/ ...
21:31 < rxr> I'm positive you'll track that down quickly...
21:31 < R4gnar0k> :p lol
21:32 < R4gnar0k> jsut dunno if you'll like my fix *hrhr*
21:32 < R4gnar0k> just*
21:37 < rxr> usually :-)
21:38 < R4gnar0k> *grins*
21:54 < R4gnar0k> gah
21:54 < R4gnar0k> /export/alben2/t2-trunk/build/64-7.0-trunk-generic-x86-64-nocona//export/alben2/t2-trunk/build/64-7.0-trunk-generic-x86-64-nocona/lib/modules/2.6.20.1-dist/kernel/drivers/net/wireless
21:54 -!- d-marc [n=d-marc@HSI-KBW-091-089-007-026.hsi2.kabelbw.de] has quit ["KVIrc 3.2.4 Anomalies http://www.kvirc.net/"]
21:55 < R4gnar0k> seems moduledir sometimes has $root included
21:56 < rxr> maybe ever in cross builds ?
21:57 < R4gnar0k> this isn't a cross build tho
21:59 < rxr> stage 1 always is cross :-)
21:59 < rxr> the cross config option only disables all non stage1 parts ..
21:59 < rxr> ok - have to go hunt for food now ...
21:59 < CIA-8> rene * r22645 /trunk/package/x11/xfe/xfe.desc: * updated xfe (0.98.1 -> 0.98.2)
22:01 < R4gnar0k> have a good one
22:01 < CIA-8> rene * r22646 /trunk/package/base/sysstat/sysstat.desc: * updated sysstat (7.0.4 -> 7.1.1)
22:01 < R4gnar0k> I read that as 'stage 1 always is gross' first :p
22:12 < R4gnar0k> rxr: can I prevent a module from being loaded in initrd?
22:12 < R4gnar0k> the iwlwifi doesn't find the firmware there
22:46 -!- sparc-kly [n=mubex@66-50-123-218.prtc.net] has joined #t2
22:57 -!- sparc-kly [n=mubex@66-50-123-218.prtc.net] has quit ["Leaving"]
--- Log closed Thu Feb 22 00:00:14 2007