IRC log for #uclibc on 20100511

00:21.46*** join/#uclibc gnomon (~gnomon@CPE0022158a8221-CM000f9f776f96.cpe.net.cable.rogers.com)
01:20.59*** join/#uclibc dileX_ (~sd@p5B2ECFBB.dip.t-dialin.net)
01:54.15CIA-5503vda.linux 07master * rbfa1b2e8e8aa 10busybox/ (31 files in 10 dirs): randomtest fixes
02:32.22CIA-5503vda.linux 07master * rc0211e907ba0 10busybox/testsuite/ (busybox.tests cpio.tests mdev.tests mount.tests): randomtest fixes
02:37.36CIA-5503vda.linux 07master * r78d035f79645 10busybox/networking/ifplugd.c: ifplugd: fix typo in prev commit
03:56.57*** join/#uclibc dinar (~dinar@93.92.193.118)
04:13.47*** join/#uclibc tsukasa (~tsukasa@unaffiliated/tsukasa)
04:32.54*** join/#uclibc gnomon (~gnomon@CPE0022158a8221-CM000f9f776f96.cpe.net.cable.rogers.com)
06:00.30*** join/#uclibc sgh (~quassel@cpe.ge-0-2-0-950.faaqnqu1.customer.tele.dk)
06:01.13*** join/#uclibc dougmencken (~quassel@93.123.156.139)
06:16.22*** join/#uclibc fabled (~fabled@letku109.adsl.netsonic.fi)
06:34.27*** join/#uclibc tchan (~tchan@lunar-linux/developer/tchan)
06:52.39*** join/#uclibc sgh (~quassel@cpe.ge-0-2-0-950.faaqnqu1.customer.tele.dk)
06:59.27*** join/#uclibc zazenrasta (~zaz@adsl-074-244-082-076.sip.asm.bellsouth.net)
07:02.50CIA-5503Paul.Jones 07master * r18f72f97309b 10buildroot/ (CHANGES package/xvkbd/Config.in package/xvkbd/xvkbd.mk): xvkbd: bump version
07:02.51CIA-5503thomas.petazzoni 07master * rde73cdf478dd 10buildroot/package/hostapd/hostapd.mk: hostapd: use internal crypto library
07:02.52CIA-5503thomas.petazzoni 07master * r5b9384337a66 10buildroot/ (CHANGES package/less/less.mk): less: depends on curses
07:02.55CIA-5503thomas.petazzoni 07master * rc554cc2a2e62 10buildroot/ (CHANGES package/netplug/netplug-1.2.9-disable-werror.patch): netplug: disable -Werror to fix build failure
07:02.55CIA-5503thomas.petazzoni 07master * rc488481e9caf 10buildroot/ (CHANGES package/libnl/libnl-1.1-netlink-local-fix.patch): libnl: include limits.h to fix build failure
07:14.10CIA-5503thomas.petazzoni 07master * rfcdc9f891d28 10buildroot/package/php/Config.in: php: make sure either CLI or CGI is selected
07:14.10CIA-5503thomas.petazzoni 07master * r6bd1a449d61e 10buildroot/package/games/gnuchess/gnuchess-5.07-getline-conflicts.patch: gnuchess: add a patch to fix getline() conflict
07:14.11CIA-5503thomas.petazzoni 07master * re627989ad847 10buildroot/package/gawk/gawk.mk: gawk: remove CC override
07:14.18CIA-5503thomas.petazzoni 07master * rf5ff8cdc6218 10buildroot/package/hal/hal.mk: hal: add dependency on host-libxml-parser-perl
07:14.18CIA-5503thomas.petazzoni 07master * r8b47ede9e6d5 10buildroot/package/libgcrypt/libgcrypt.mk: libgcrypt: specify where gpg-error-config is
07:14.18CIA-5503thomas.petazzoni 07master * r6040190e9206 10buildroot/package/ (directfb/Config.in linux-fusion/Config.in): linux-fusion: add dependency on !BR2_KERNEL_none
07:14.19CIA-5503thomas.petazzoni 07master * raa3439e6a3fe 10buildroot/package/multimedia/vlc/Config.in: vlc: mark as broken
07:14.19CIA-5503thomas.petazzoni 07master * r844bbbb3fd10 10buildroot/package/sed/sed.mk: sed: bump to 4.2.1 to fix build failure
07:14.19CIA-5503thomas.petazzoni 07master * rbb39f18905e3 10buildroot/package/ltp-testsuite/Config.in: ltp-testsuite: mark as broken
07:14.20CIA-5503thomas.petazzoni 07master * rba9a902bf862 10buildroot/package/shared-mime-info/shared-mime-info.mk: shared-mime-info: missing dependency on host-intltool
07:14.20CIA-5503jacmet 07master * r77e2dfffae07 10buildroot/package/ (11 files in 11 dirs): Merge branch 'build-fixes' of git://git.busybox.net/~tpetazzoni/git/buildroot
07:15.08CIA-5503jacmet 07master * r49ce10cb3c7d 10buildroot/CHANGES: CHANGES: add updated packages from Thomas' build-fixes branch
07:20.40dougmenckenbuildroot guys, can you tell me which gcc are you using in buildroot? it's all about this bug : http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43911
07:30.32*** join/#uclibc matteo (~matteo@openwrt/developer/matteo)
07:36.59*** join/#uclibc sgh (~quassel@cpe.ge-0-2-0-950.faaqnqu1.customer.tele.dk)
07:48.05ncopablindvt: ping
08:45.55*** join/#uclibc Phurl (~mdupont@2001:0:53aa:64c:2c08:6fff:4d34:23db)
09:03.37*** join/#uclibc nataraj (~nataraj@122.165.223.135)
09:03.46natarajHi
09:10.14dougmenckenhello everyone and good luck
09:16.09*** join/#uclibc nataraj (~nataraj@122.165.223.135)
09:20.10*** join/#uclibc sgh (~quassel@cpe.ge-0-2-0-950.faaqnqu1.customer.tele.dk)
09:23.35*** join/#uclibc y_morin (~ymorin@ARennes-256-1-132-217.w90-32.abo.wanadoo.fr)
09:51.20natarajHi, how do i recompile qt package alone?
09:55.20nataraj"make qt clean" ?
10:02.55dougmenckennataraj: make clean only executes rule clean in the current build dir
10:03.07CIA-5503vda.linux 07master * rda1382410bbc 10busybox/testsuite/ (mdev.tests mount.tests tar.tests): randomtest fixes
10:04.14natarajdougmencken, yep, that what happened to me
10:07.13dougmenckennataraj: so cd to the correct buildir (for example, qt-build) and then 'make clean' then re-build, or better remove anything with rm -rf qt-build and unpack fresh tarball
10:10.51dougmenckenI just wonder (and wonder very much) why don't you like to build uclibc-binaries inside chroot? clean, simple, elegant
10:12.27*** join/#uclibc mirko (~daten@g225064113.adsl.alicedsl.de)
10:13.55natarajusing qemu?
10:14.17natarajor scratchbox?
10:15.09nataraji am on arm-linux
10:16.01natarajdougmencken, would you help me?
10:16.39dougmenckennataraj: your target is not you host, right? let me guess, you are using x86 as host
10:17.09natarajyep
10:17.26dougmenckenwell, then yes, qemu
10:18.33dougmenckendespite the fact, qemu is very and very raw, can't boot ppc livecd inside ppc
10:19.26natarajwhat could be wrong?
10:19.54dougmenckennataraj: on clean emulated system: nothing
10:20.18nataraji used to use qemu on debian-armel with a .qcow filesyetm
10:20.54dougmenckenbtw, even gcc dev asks for "how can I reproduce it via cross tooch"? I'm tired to tell them "you can't reproduce it on cross"
10:22.09dougmenckennataraj: once you got busybox and uclibc, try to look at my chroot srcipt: http://manulix.wikidot.com/files:chroot-here-sh
10:31.31*** join/#uclibc matteo` (~matteo@openwrt/developer/matteo)
10:45.21*** join/#uclibc sgh (~quassel@cpe.ge-0-2-0-950.faaqnqu1.customer.tele.dk)
11:11.05*** join/#uclibc sgh (~quassel@cpe.ge-0-2-0-950.faaqnqu1.customer.tele.dk)
11:22.19*** join/#uclibc sgh (~quassel@cpe.ge-0-2-0-950.faaqnqu1.customer.tele.dk)
11:34.40*** join/#uclibc likewise (~likewise@82-171-51-231.ip.telfort.nl)
11:35.45*** join/#uclibc likewise (~likewise@82-171-51-231.ip.telfort.nl)
11:40.52*** join/#uclibc bkuhn (~bkuhn@fsf/director/conservancy.president.bkuhn)
11:46.00*** join/#uclibc sgh (~quassel@cpe.ge-0-2-0-950.faaqnqu1.customer.tele.dk)
11:51.29*** join/#uclibc Phurl (~mdupont@2001:0:53aa:64c:2c08:6fff:4d34:23db)
11:54.07*** join/#uclibc y_morin_ (~ymorin@ARennes-256-1-117-210.w90-32.abo.wanadoo.fr)
11:58.10*** join/#uclibc sgh (~quassel@cpe.ge-0-2-0-950.faaqnqu1.customer.tele.dk)
12:01.09*** join/#uclibc gustavoz (~gustavoz@201.253.140.23)
12:27.40*** join/#uclibc calculus (~calculus@gentoo/user/calculus)
12:31.57*** join/#uclibc matteo` (~matteo@openwrt/developer/matteo)
12:32.06*** join/#uclibc mirko (~daten@zfd20.pia.fu-berlin.de)
12:39.51*** join/#uclibc pentanol (~pentanol@77-35-2-235.pppoe.primorye.net.ru)
12:41.12*** join/#uclibc pscheie (~pscheie@173-11-39-65-Minnesota.hfc.comcastbusiness.net)
12:50.22CIA-5503vda.linux 07master * r94043e8ad2d3 10busybox/ (editors/vi.c libbb/lineedit.c): lineedit: change how cmdedit_set_out_char works
12:54.44dougmenckencan you please help me with this bug : http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43911 ; it does really seem to be uclibc-only
12:55.13dougmenckenand soon or later, you will use 4.4.3 and above too
13:03.04gustavozdougmencken: builds for me with 4.4.4
13:03.52dougmenckengustavoz: natively? i.e. uclibc host, target, and build are the same?
13:04.07gustavozdougmencken: oh no, not natively :) cross
13:04.17dougmenckenah, not this case then
13:05.03dougmenckenanybody can try to bootstrap gcc 3.5.0 natively (yes, on x86, but on uclibc host)?
13:05.15dougmenckens/3./4./
13:05.24gustavozdougmencken: i could give it a shot but it'll take me a while to build on a 400mhz mpc8315
13:05.41dougmenckengustavoz: on my side, 1 bootstrap = 18-24 hours
13:08.56dougmenckenso it's okay; and be sure to include --enable-shared and try with --disable-checking (related to bug http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42347)
13:12.09dougmenckengcc is *very* important tool, without it, we are headless and eyeless, all of us
13:12.13*** join/#uclibc gnomon (~gnomon@CPE0022158a8221-CM000f9f776f96.cpe.net.cable.rogers.com)
13:25.39*** join/#uclibc sgh (~quassel@cpe.ge-0-2-0-950.faaqnqu1.customer.tele.dk)
13:53.51*** join/#uclibc matteo`_ (~matteo@93-32-56-237.ip32.fastwebnet.it)
14:08.56*** join/#uclibc matteo (~matteo@93-32-56-237.ip32.fastwebnet.it)
14:08.57*** join/#uclibc matteo (~matteo@openwrt/developer/matteo)
15:02.29*** join/#uclibc zazenrasta (~zaz@adsl-074-244-082-076.sip.asm.bellsouth.net)
15:20.56khemdougmencken: for the git query you posted
15:21.12khemyou can git log gcc/BASE-VER
15:21.28khemthat file is usually modified whenever the release happens
15:21.44khemso you can get the SHA from there for your good base
15:22.32dougmenckenkhem: from buildroot? okay, but... I already started bisecting, guessing bad and god by date
15:22.59khemdougmencken: I meant gcc git
15:23.11khemI was looking at your last comment on the bug
15:23.47dougmenckenah, that gcc git, okay, thanks
15:25.00dougmenckencat gcc/BASE-VER does really work ;)
15:26.24dougmenckenbut it's 4.3.0/4.4.0 and 4.5.0 only (i.e. 4.4.2 and 4.4.3 are undistinguishable)
15:29.11*** join/#uclibc mnt_real (~sinan@bas11-montrealak-1177756491.dsl.bell.ca)
15:35.08*** join/#uclibc mirko (~daten@p57B96C13.dip.t-dialin.net)
15:41.10*** join/#uclibc matteo` (~matteo@openwrt/developer/matteo)
16:09.03*** join/#uclibc mirko (~daten@p57B96C13.dip.t-dialin.net)
16:11.03*** join/#uclibc gnomon (~gnomon@CPE0022158a8221-CM000f9f776f96.cpe.net.cable.rogers.com)
16:43.07*** join/#uclibc mirko (~daten@p57B96C13.dip.t-dialin.net)
16:56.43*** join/#uclibc gnomon (~gnomon@CPE0022158a8221-CM000f9f776f96.cpe.net.cable.rogers.com)
17:00.33khemdougmencken: thats because 4.4 is a branch
17:00.41khemand 4.4.1 and 4.4.2 happens to be on that
17:19.56*** join/#uclibc likewhoa (~likewhoa@ool-45717734.dyn.optonline.net)
17:20.00*** join/#uclibc davidbpa (~davidbpa@200.129.46.176)
17:26.51*** join/#uclibc austinf (~austinf@74.92.231.110)
18:10.26*** join/#uclibc trem (~trem@mol92-1-81-57-136-23.fbx.proxad.net)
18:15.05*** join/#uclibc Xires (~Xires@66.190.79.122)
18:39.05*** join/#uclibc mirko (~daten@g225064113.adsl.alicedsl.de)
18:49.30blindvtncopa, pong
18:51.40khemblindvt: hey
18:51.49blindvtkhem, hi
18:51.56khemblindvt: dougmencken has tested the ppc patch
18:52.23khemhis results are here http://manulix.wikidot.com/local--files/start/out
18:52.41khemthe results are pretty good
18:57.49khemblindvt: and landley also booted his big native system
18:57.54khemwith nptl/ppc
18:58.31blindvtkhem, looks expected. Do we have explicit TID tests in there, btw? If not that should be added as sanity check
18:58.50khemblindvt: I wanted you to ask for that
18:58.59khemwhat should I test to make sure TIDs are ok
18:59.50*** join/#uclibc kos_tom (~thomas@humanoidz.org)
19:04.32blindvtkhem, there's a thread_id somewhere. top(1) or /proc, in case you have that. Dunno offhand
19:05.58khemblindvt: ps -ef shows in Parent Process ID
19:06.03khemand Process ID
19:06.07khemtop shows only PID
19:06.21blindvtkhem, but essentially it's the RESET_PID logic that is wrong. IIRC we need to distinct clone impls, one for pthread and one for c (and my last patch went through hoops to use only one, and that's wrong)
19:06.41blindvts/need to/need two/
19:07.36khemblindvt: there is only one for other implementations
19:07.41khemwhy do we need two
19:07.45khemwhy libc needs  clone
19:09.51ronghailis there a way to get mdev to modprobe modules for plugged-in hardware?
19:09.57ronghailor existing hardware
19:11.02blindvtronghail, mdev -s. see docs/mdev.txt
19:11.21ronghailit isn't calling modprobe
19:11.35blindvtkhem, i very much thought we need two
19:11.56ronghailit creates a bunch of devices, and if I modprobe foo the devices get created.  but it doesnt' seem to call modprobe on its own
19:12.07ronghailI have no mdev.conf though
19:12.10blindvtronghail, run modprobe manually to see if it works in principle. Check that you did everything that is listed in mdev.txt
19:12.20ronghailis mdev.conf the key to this?
19:12.59ronghailthe docs for the conf seem to say stuff in the conf will happen _after_ the device is created.  so seems like a chicken/egg problem
19:13.26blindvtronghail, you usually don't really need an mdev.conf
19:13.38ronghailyeah I have read that a few times actually.  still a little hazy on whether I need mdev.conf and why
19:14.06ronghailif I run modprobe manually everything is fine
19:14.29ronghailbut I was hoping mdev would magically see that I need to modprobe my scsi module
19:14.36blindvtronghail, my (busybox init) inittab for mdev looks like this, for your reference: http://repo.or.cz/w/buildroot.git/blob_plain/HEAD:/target/generic/target_busybox_skeleton/etc/inittab
19:14.59ronghailthanks, looking at it
19:15.07khemblindvt: actually the function should be in libc but it should come thru nptl or lt
19:15.27CIA-5503jacmet 07master * r6d1ab151cb4a 10buildroot/ (CHANGES Config.in docs/download.html docs/news.html): update for 2010.05-rc2
19:15.28khemblindvt: the reason is that it uses this getpid crap to cache pids
19:15.31khemin nptl
19:15.56ronghailI basically have all that sysinit stuff in my initramfs init up to before the mount -a
19:16.11blindvtkhem, i was under the impression that the nptl/lt variant needs to be distinct from the libc one, difference being that one resets the pid, as opposed to the other
19:16.24ronghailI mount proc and sys and dev and dev/pts, then I put /sbin/mdev in proc/.../hotplug and then I do mdev -s
19:16.33CIA-5503jacmet 07refs/tags/2010.05_rc2 * r712788423f8f 10buildroot/ (CHANGES Config.in docs/download.html docs/news.html): update for 2010.05-rc2
19:16.40khemblindvt: thats true but the function itself ends up in libc
19:16.55khemand we only choose one pthread implementation at a time
19:17.18khemso the one we have now for ppc nptl picks the one from nptl which is correct
19:19.14ronghailI will go confirm everything again to make sure
19:22.36blindvtkhem, so.. do both NO_THREADS and THREADS_NATIVE work as expected?
19:24.05khemblindvt: I havent tried NO_THREADS
19:24.08blindvtkhem, either way, i suggest you (or me, i don't care) push that thing out to master. We can fix it up later
19:24.11khembut it should work as usual
19:24.30khemblindvt: ok I can push it as I have few fixes on top
19:24.38khemand my tree is ready to go in
19:24.52blindvtkhem, i'd be glad if you could push it (my SOB should still be in there) i'm knee-deep into gcc ATM and want to finish that first
19:25.06khemyes its in there
19:25.14blindvtcool. Fore! :)
19:25.21kheminfact you are in the front line with first SOB
19:27.54CIA-5503raj.khem 07master * re3dec33ba03e 10uClibc/ (42 files in 19 dirs): powerpc: Add TLS and NPTL support
19:27.55CIA-5503raj.khem 07master * re4cf1aceb0e7 10uClibc/libpthread/nptl/sysdeps/powerpc/Makefile.arch: powerpc/Makefile.arch: Add objclean and headers_clean targets for NPTL.
19:29.51khemblindvt: http://pastebin.com/k6ndUu6j
19:30.00khemdo you approve of this patch
19:30.50blindvtkhem, sounds plausible from a quick look
19:31.08khemblindvt: ok
19:31.30khemblindvt: and landley also had errors compiling host getconf on ubuntu 8.04
19:32.08dougmenckenkhem: hoorah! time to remove external patch from build script :)
19:32.45blindvtkhem, did i send my config-parser addition and redoing all those /etc/* parsers with it to the list, btw? IIRC i did. Would be awesome if somebody could have a look at the locking in multithreaded apps using that patch
19:32.50khemblindvt: http://lists.uclibc.org/pipermail/uclibc/2010-May/044033.html
19:33.17khemblindvt: send me links to the patches I can queue them
19:33.18khemhere
19:33.58blindvtkhem, he's using out of date stuff and trolls about git. Not exactly news
19:35.12blindvti don't care about pre SUSv4 systems, which he apparently has. *shrug*
19:35.13khemblindvt: if you could reply to the mail would be nice
19:35.21khemok
19:36.39blindvtkhem, no. it's obvious and he should very well see what's going on. I'm not wasting my time to reply to rant-of-the-day-about-this-and-that-SCM-system mails.
19:38.37khemblindvt: heh ok
19:38.41khemunderstood
19:39.02blindvtdougmencken, could you please build and run the testsuite with no threads and show us the results?
19:40.19*** join/#uclibc gnomon (~gnomon@CPE0022158a8221-CM000f9f776f96.cpe.net.cable.rogers.com)
19:41.53dougmenckenblindvt: with no threads? okay, but not today; also, before I used LINUXTHREADS_OLD and saw no problem (never ran testsuite, just by experience)
19:43.00blindvtdougmencken, k, TIA. keeping testsuite logs would be very welcome.
19:43.40blindvtkhem, i suggest we should think about adding something like the .sum files to be able to spot regressions easily
19:44.38khemblindvt: hmmm git bisect is good enough isnt it
19:45.44blindvtkhem, for tracking down who caused breakage it is. For checking if you added a new regression on a port that may have had one or two missing testsuite cases it doesn't help
19:46.28khemI see
19:46.51khemblindvt: I think having something like a table of testresults like vapier did wud help
19:47.03khemblindvt: I am adding qemu targets to OE
19:47.10khemppc and SH are next in line
19:47.18khemit already has arm mips and x86
19:47.43khemand may be later add x86_64 and coldfire too
19:47.51khemthen we can automate the testing
19:47.52blindvtkhem, good thing. m68k afterwards? :)
19:48.34khemheh
19:49.34khemblindvt: may be sparc first
19:49.48khemaustinf: could be interested in that
19:53.41austinfoh yes
19:53.43austinf:)
20:17.03*** join/#uclibc gnomon (~gnomon@CPE0022158a8221-CM000f9f776f96.cpe.net.cable.rogers.com)
21:07.44CIA-5503raj.khem 07master * r008ebd962932 10uClibc/utils/Makefile.in: utils/Makefile.in: Define GETCONF_DIR for host builds of getconf.
21:39.35dj-deathkos_tom: around.
21:39.57dj-deathkos_tom: probably 1 day too late ...
22:19.17kos_tomdj-death: yeah. I'll be back on Thursday.
22:19.32dj-death'k
22:24.47kos_tomdj-death: another 8 build tests based on randpackageconfig, 8 failures.
22:25.27dj-death'k...
23:03.14*** join/#uclibc gnomon (~gnomon@CPE0022158a8221-CM000f9f776f96.cpe.net.cable.rogers.com)
23:04.22*** join/#uclibc zaz1R (~zaz@adsl-074-244-082-076.sip.asm.bellsouth.net)
23:49.31CIA-5503vda.linux 07master * ra2215b98f7d6 10busybox/ (editors/sed.c testsuite/sed.tests testsuite/testing.sh): sed: fix a command with multible trailing backslashes
23:49.38*** join/#uclibc zazenrasta (~zaz@adsl-074-244-082-076.sip.asm.bellsouth.net)

Generated by irclog2html.pl Modified by Tim Riker to work with infobot.