T2 IRC Log: 2008-12-10

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 Wed Dec 10 00:00:30 2008
01:10 -!- wtracy [n=wtracy@natusers1.stennerglenstudents.com] has joined #t2
02:25 -!- wtracy [n=wtracy@natusers1.stennerglenstudents.com] has quit [Remote closed the connection]
03:39 -!- wtracy [n=wtracy@natusers1.stennerglenstudents.com] has joined #t2
06:08 -!- wtracy [n=wtracy@natusers1.stennerglenstudents.com] has quit [Remote closed the connection]
06:28 -!- mpp [n=mpp@fulda025.server4you.de] has joined #t2
06:39 -!- wtracy [n=wtracy@natusers1.stennerglenstudents.com] has joined #t2
07:12 -!- wtracy [n=wtracy@natusers1.stennerglenstudents.com] has quit [Remote closed the connection]
07:24 -!- wtracy [n=wtracy@natusers1.stennerglenstudents.com] has joined #t2
07:33 -!- mpp [n=mpp@fulda025.server4you.de] has quit [Remote closed the connection]
07:34 -!- mpp [n=mpp@fulda025.server4you.de] has joined #t2
08:00 -!- wtracy [n=wtracy@natusers1.stennerglenstudents.com] has quit [Remote closed the connection]
08:26 < mtr> moin
08:44 < rxr> moin mtr
08:47 < mtr> hi rxr
08:55 < CIA-67> rene * r31486 /trunk/package/network/dbus/dbus.desc: * updated dbus (1.2.4 -> 1.2.8)
08:55 < CIA-67> rene * r31487 /trunk/package/network/dbus-glib/dbus-glib.desc: * updated dbus-glib (0.76 -> 0.78)
08:56 < CIA-67> rene * r31488 /trunk/package/base/rsyslog/rsyslog.desc: * updated rsyslog (3.20.0 -> 3.20.2)
09:40 < CIA-67> rene * r31490 /trunk/package/base/bonnie++/bonnie++.desc: * updated bonnie++ (1.03d -> 1.03e)
09:40 < CIA-67> rene * r31489 /trunk/package/network/aoetools/aoetools.desc: * updated aoetools (27 -> 29)
09:40 < CIA-67> rene * r31491 /trunk/package/develop/bazaar/bazaar.desc: * updated bazaar (1.9 -> 1.10)
09:40 < mpp> moinsen
10:30 < CIA-67> rene * r31492 /trunk/package/base/busybox/busybox.conf: * configure recent busybox for MMU-less blackfin
10:45 < mpp> hey rxr
10:46 < mpp> dl2 + dl3 get quite some load :-)
10:46 < rxr> now you know why our "not too bad" office line was oversaturated ...
10:46 < mpp> is it getting better over there ?
10:47 < mpp> i had to block searche enigne bots
10:47 < rxr> with the mirrors taking the load, yes .-)
10:47 < mpp> they killed the bw even on my server
10:47 < rxr> ah - yep
10:47 < mpp> esp google is insane inscpecting all bz2 packages and iso images
10:47 < mpp> pahhh.
10:49 < rxr> yeah - they should know better by now
10:49 < rxr> on the other hand they might like to know all your secrets :-)
10:50 < mpp> right
10:50 < mpp> ok i scheduled mirror of the dl every night so dl2 +dl3 should be behind the most 24h
10:51 < CIA-67> rene * r31493 /trunk/architecture/blackfin/package/linux26/ (5 files):
10:51 < CIA-67> * do not blindly use HEAD patches for the blackfin kernel,
10:51 < CIA-67> use the copy of the old, 2.6.20 files instead
10:52 < rxr> great
10:52 < rxr> we can start pushing new t2 releases then :-)
10:54 < rxr> can someone fix imagemagick and ghostscript to play together for me?
10:54 < rxr> their code hurts my eyes and I would have some eps to render to a bitmap :-(((
10:54 < rxr> guess it's time to implement an oen PostScript interpreter in exact-image :-(
10:55 < rxr> huh, !> cc1: error: unrecognized command line option "-mcpu=bf533-0.3"
11:07 -!- mpp [n=mpp@fulda025.server4you.de] has quit ["ChatZilla 0.9.84 [Firefox 3.0.4/2008111317]"]
11:22 < rxr> http://www.t2-project.org/zine/3/
11:24 -!- michael-i [n=michael-@141.41.40.103] has joined #t2
11:28 < CIA-67> rene * r31494 /trunk/target/embedded-minimal/ (. build.sh config.in pkgsel pkgsel.sed): * added truely minimal embedded example
11:29 < koan> rxr: a typo in the zine article "The CF card must be large enough to holt..."
11:29 < rxr> ah - yes thanks
11:29 < rxr> I even saw that in the email but forgot while cut'n pasting
11:30 < rxr> I wait with the commit for the case you find another one :-)
11:30 < koan> it's a nice walkthrough article
11:30 < koan> it's the only one I found :-)
11:30 < rxr> yeah - thanks to Iulian Demetrescu
11:30 < koan> oh wait
11:31 < koan> "Aditionally he is a system integrator..." at the end should be with two d's
11:32 < rxr> done
11:33 < koan> I have also such a Via mini-itx system, maybe I should try t2 on it
11:34 < koan> I'll have to think about a purpose for the machine, now it's collecting dust :-)
11:56 < michael-i> rxr, just synced up and started building. Thanks for your time in committing that!
11:58 < rxr> michael-i: the linux26 does not build anymore
11:58 < rxr> for blackfin
11:58 < rxr> ok - I just see that was the case before (old regression tester log)
11:58 < rxr> guess I did not particularly care about the blackfind kernel
11:59 < rxr> so when you build the new embedded-minimal for blackfin it should build except the linux26 package
11:59 < michael-i> I'm trying 486 for now, then moving on to blackfin. So far I've had to emerge-pkg gmp and mpfr because gcc fails without them
11:59 < rxr> I just update the blackfin toolchain as a new gcc / cpu option parsing is required for the mainline linux kernel
11:59 < rxr> ah - yesh - sorry forgot
11:59 < rxr> (I should delete them from my system :-)
12:00 < michael-i> Ok, good to know. Should I update those toolchains to the latest revision from blackfin.uclinux.org then?
12:00 < rxr> that is what I try right now
12:01 < rxr> maybe wait for my test to not waste time and test with 486 and/or the current version
12:01 < rxr> when I packaged the linux26 for blackfin overlay it built on that day - so even the current sources should be fixable ...
12:01 < rxr> when you do the same update right now you might just waste the time re-doing what I just did and wait to download and build ...
12:02 < CIA-67> rene * r31495 /trunk/target/embedded-minimal/ (pkgsel pkgsel.sed): * added gmp and mpfr (gcc dependencies) to the embedded-minimal target :-)
12:03 < michael-i> ok, I'll wait for now. 486 is compiling. I'll check the other target examples and try to add a libc option to the embedded-minimal target. Comparing a minimal glibc and a uclibc would be interesting.
12:14 -!- B-2 [i=stealth@sourcemage/guru/stealth] has joined #t2
12:16 -!- Stealth [i=stealth@sourcemage/guru/stealth] has quit [Read error: 113 (No route to host)]
12:16 -!- B-2 is now known as Stealth
12:19 < rxr> hi Stealth
12:19 [Users #t2]
12:19 [@ChanServ] [ Cyda ] [ LMJ ] [ rxr ] [ TobiX]
12:19 [ axionix_] [ dsoul] [ michael-i] [ Stealth]
12:19 [ CIA-67 ] [ koan ] [ mtr ] [ tfl0pz ]
12:19 -!- Irssi: #t2: Total of 13 nicks [1 ops, 0 halfops, 0 voices, 12 normal]
12:19 < Stealth> re
12:38 < mtr> what exactly was the optimisation problem with glibc-2.7?
12:38 < mtr> -Os is broken, right?
12:39 < mtr> the livecd target uses -Os, and build fails with:
12:39 < mtr> /work/T2/t2/src.glibc.tresc.20081208.215231.974.mwork/glibc-2.7/objdir/libc_pic.os: In function `faccessat':
12:39 < mtr> /work/T2/t2/src.glibc.tresc.20081208.215231.974.mwork/glibc-2.7/io/../sysdeps/unix/sysv/linux/faccessat.c:111: undefined reference to `fstatat64'
12:40 < mtr> is there a known workaround for this, other than using -O2?
12:47 < rxr> no this glibc stuff was too ugly for me
12:48 < rxr> but didn't I work around by exchanging Os with O2 in the glibc.conf ?
13:07 < michael-i> ar and strings are conflicting between binutils and busybox when embedded-minimal attempts to symlink the busybox programs. Would it be better/easier to disable binutils version or those in busybox?
13:09 < rxr> binutils should not have been installed in stage 1 of rhe embedded-minimal
13:09 < rxr> for the truely minimal thing no binutils and gcc should make it onto the target
13:09 < rxr> have you changed the target def ?
13:10 < rxr> oh - I got a typo in the pkgsel.sed .-(
13:10 < michael-i> yes, I added a uclibc/glibc selector and STATIC=0 and PKG_BUSYBOX_SYMLINKS=1 (http://pastebin.ca/1281567)
13:11 < michael-i> binutila....the Italian version? ;)
13:11 < CIA-67> rene * r31496 /trunk/target/embedded-minimal/pkgsel.sed: * fixed binutils typo in new embedded-minimal/pkgsel.sed
13:12 < rxr> yeah - as I did not got that error I guess it slipped in in soem last minute tweak of the file with vi ...
13:12 < rxr> binutila the nutela version ._9
13:12 < rxr> :-) even
13:12 < rxr> or how the nougat bread suggar thing is spelled :-)
13:13 < rxr> http://www.nutella.de/ even
13:13 < rxr> and just sugar sorry
13:13 < rxr> need lunch - cu
13:41 < michael-i> rxr, dropbear was missing zlib. I added it to pkgsel and now the image finishes compiling with uclibc (patch above). Packaging the image still fails because tar, readlink and rmdir are missing in /build/default-8.0-trunk-embedded-minimum-x86-i486-cross/embutils
13:42 < michael-i> total build directory size for uclibc 122MB. Now I'm trying using glibc
13:54 < rxr> you may not count the TOOLCHAIN dir
13:54 < rxr> to determine target image size use du of just [a-z]* in the build-dir
13:55 < rxr> the TOOLCHAIN will have hundreds of meg of cross compiler assembler linker and other support glue
13:55 < rxr> the uclibc target image will just be some single digit MB
13:55 < michael-i> I figured that but just took a quick measurement
13:56 < michael-i> I'll have another chance here soon hopefully with glibc (although I'm not sure my patch is working since uclibc is being buit again anyway)
13:56 < michael-i> ...and of course I have to for a presentation now
13:59 < michael-i> gotta go, be back later this evening
13:59 < michael-i> ciao
13:59 -!- michael-i [n=michael-@141.41.40.103] has quit []
14:44 < rxr> http://www.t2-project.org/zine/4/
15:24 < CIA-67> rene * r31497 /trunk/package/games/frozen-bubble/frozen-bubble.desc: * updated frozen-bubble (2.1.0 -> 2.2.0)
15:26 < koan> rxr: with which program did you create the graphs in the benchmark article?
15:31 < rxr> ploicus
15:31 < rxr> ploticus even
15:32 < rxr> http://www.t2-project.org/packages/ploticus.html
15:35 < koan> looks nice
16:53 < mtr> removed -Os with a preconf-hook for the livecd target, now glib32 fails with:
16:53 < mtr> !> pthread_rwlock_trywrlock.c:45: Error: bad register name `%dil'
16:53 < mtr> arghh :(
16:54 < rxr> hm
17:42 < mtr> known problem, already fixed in CVS:
17:42 < mtr> http://sourceware.org/cgi-bin/cvsweb.cgi/libc/nptl/sysdeps/unix/sysv/linux/i386/lowlevellock.h.diff?r1=1.38&r2=1.39&cvsroot=glibc
17:44 < rxr> patch import welcome
17:44 < rxr> btw. did I mention glibc is a stinky mess ?
17:49 < koan> several times :-)
18:11 -!- wtracy [n=wtracy@natusers1.stennerglenstudents.com] has joined #t2
19:59 -!- wtracy [n=wtracy@natusers1.stennerglenstudents.com] has quit [Remote closed the connection]
21:01 -!- haggis_pc [n=haggis__@93-97-152-107.zone5.bethere.co.uk] has joined #t2
21:29 < haggis_pc> any1 awake?
21:39 -!- d_oilen [n=iulian@89.18.29.87] has joined #t2
21:39 < d_oilen> hi
21:42 -!- d_oilen [n=iulian@89.18.29.87] has quit [Client Quit]
22:02 -!- haggis_pc [n=haggis__@93-97-152-107.zone5.bethere.co.uk] has quit ["Leaving"]
22:02 < koan> rxr: if I emerge lua-sqlite3, I get the following error:
22:02 < koan> /usr/lib64/gcc/powerpc64-t2-linux-gnu/4.3.2/../../../../powerpc64-t2-linux-gnu/bin/ld: cannot find -llualib
--- Log closed Thu Dec 11 00:00:36 2008