irclog2html for #uclibc on 20050322

00:38.35GyrosGeierhrm
01:20.01*** join/#uclibc SpanKY (~UserBah@pool-151-203-225-133.bos.east.verizon.net)
01:48.40*** part/#uclibc belracu (~getram@61-31-22-29.static.tfn.net.tw)
02:42.46*** join/#uclibc SecretSquirrel (~squirrel@adsl-209-30-10-25.dsl.rcsntx.swbell.net)
03:06.51*** join/#uclibc ShadowJK (jknutar@hylsy.saunalahti.fi)
03:30.32*** join/#uclibc null (~null___@g624-9034.itee.uq.edu.au)
03:30.44nullhello
03:31.14ashesbinutils-2.15.96 is out as of march 8. looks like 2.16 will be out soon
03:31.22nullany one know how to add search path for the includes ? when compiling with uclibc ?
03:31.26nullheya ashes
03:32.03SpanKYashes: it's a very snapshot-snapshot
03:32.12SpanKY2.16 is scheduled for april i believe
03:32.33nullany ideas ?
03:32.33*** join/#uclibc flatronf700B (~flatronf7@ns1.clipsalportal.com)
03:33.55SpanKYnull: -I
03:34.21nullahh will give dat a try thanks SpanKY
03:45.29ashes2.15 was botched. released too late and had too many incomplete untested features. 2.16 should be alright
03:57.39Gnat42hello
03:57.47Gnat42I am trying to add a package to buildroot
03:57.51Gnat42DirectFB to be exact
03:58.05Gnat42yesterday I learned not to use ldd as it is tied to the host compiler chain
03:58.14Gnat42today, using readelf -d
03:58.26Gnat42I'm noticing that some of the libraries created
03:58.32Gnat42reference the correct libc.so.0
03:58.39Gnat42and some reference libc.so.6
03:58.42Gnat42from the host.
03:59.18nullGnat42, well kool, i actually wanted to know the equivalent of ldd
03:59.31Gnat42glad I could be of help
03:59.44Gnat42the exact command is readelf -d path/to/so | grep NEEDED
03:59.58Gnat42to find the external referenced libraries
04:00.11Gnat42I learned it the *hard* way of submitting a bug report
04:00.17Gnat42and then looking like a dweeb
04:00.25Gnat42as it wasn't anywhere near accurate.
04:00.35nullwell sure every one goes thro dat stage
04:01.01Gnat42yes
04:01.06Gnat42doesn't make it fun though
04:01.07Gnat42;)
04:01.16Gnat42I'm confused about this one though
04:01.23Gnat42some of the files created are proper
04:01.25Gnat42others not
04:02.03nullGnat42, but what about binaries ? they should be able to show the libs linked to them using readelf ?
04:02.30Gnat42well I'm compiling DirectFB
04:02.39Gnat42it creates a number of .so shared libraries
04:02.51nullahh well i am compiling uclinux :-)
04:03.04Gnat42when I run readelf -d .so, some of them show that they need the hosts libc instead of the target
04:03.11Gnat42oddly enough though, some
04:03.13Gnat42don't
04:03.20Gnat42this is using the buildroot system.
04:03.25Gnat42where is uclinux?
04:03.56Gnat42ah
04:04.01Gnat42uclinux.org
04:04.03nullahh uclinux.org i think
04:04.04nullheh
04:04.11Gnat42I have an MMU
04:04.21nullgood
04:04.23Gnat42ah well
04:04.31Gnat42back to scratching my head
04:06.49nullme 2 :-) c ya buddy
04:16.54Gnat42if a package in buildroot uses libtool to deal with the compile & linking
04:17.15Gnat42does there need to be anything in particular passed to configure to make sure it uses the cross compilers
04:17.28Gnat42besides the usual --host --build --target ?
04:30.03*** join/#uclibc Alexander__ (~Alex@203.101.41.151)
04:35.05SpanKYno, libtool should do the right thing
04:37.10Gnat42any reasons you can think of that some of the libraries reference libc.so.6 ?
04:37.37Gnat42but most reference libc.so.0
04:42.32solaryeah miscompile failed to use the correct -I -L options
04:46.12Gnat42alright
04:46.29Gnat42I passed it a -I$(STAGING_DIR)
04:46.34Gnat42in the CFLAGS section
04:46.44Gnat42but I guess I need to add the -L option
04:49.40Gnat42thanks
05:26.21SpanKYandersee: see this before ?
05:26.22SpanKY../../include/bits/uClibc_page.h:24:1: warning: "PAGE_SHIFT" redefined
05:26.32SpanKY<PROTECTED>
05:26.32SpanKY<PROTECTED>
05:26.32SpanKY../../include/asm/page.h:8:1: warning: this is the location of the previous definition
05:26.37SpanKYi snipped some stuff there ...
05:34.09ashesis that util-linux?
05:34.24ashesim thinking PAGE_SIZE i think
05:34.28SpanKYthat's uclibc :p
05:34.41ashesah
06:10.09Gnat42hello once again
06:10.22Gnat42this should be my last question for a bit
06:10.31Gnat42I created a target/device/Via/epia-mii
06:10.33Gnat42directory
06:10.38Gnat42with the required files
06:10.45Gnat42so as to be able to build a kernel
06:10.50Gnat42there is only one patch
06:11.07Gnat42which is in the kernel-patches directory of epia-mii
06:11.14Gnat42unfortunately try as I might
06:11.19Gnat42it doesn't patch
06:11.42Gnat42the linux.mk file is patterned after the Soekris/net4521
06:11.47Gnat42and I see that it does
06:11.56Gnat42mkdir -p package/linux/kernel-patches
06:11.56Gnat42toolchain/patch-kernel.sh $(LINUX_DIR) package/linux/kernel-patches
06:12.14solarI get all happy when somebody says soekris
06:12.33Gnat42which seems a bit odd, as I don't see anywhere the patches copied to the kernel-patches dir
06:12.40Gnat42oh and again this is buildroot
06:12.46Gnat42if it wasn't already evident
06:12.52Gnat42you like soekris?
06:12.58Gnat42or just the pronounciation?
06:13.15solarcorrect I like the 45xx series
06:13.53Gnat42do you like it enough to know where I should look to know what is happening?
06:14.00Gnat42;)
06:14.23solarwhats not happening?
06:14.34Gnat42the patch isn't being applied to the kernel sources.
06:14.42Gnat42this isn't the soekris itself
06:14.46solarwhich patch?
06:14.49Gnat42but patterned pretty closely
06:14.54Gnat42viafb.diff
06:14.58solarwhat dir is the patch in
06:15.37Gnat42target/device/Via/epia-mii/kernel-patches
06:15.43Gnat42viafb_03.diff
06:15.45Gnat42to be exact
06:18.01solarthis is new
06:18.08Gnat42I'm adding it
06:18.14Gnat42haven't submitted the patch
06:18.23Gnat42as this last part is not *working* as it should
06:18.27Gnat42it compiles
06:18.30Gnat42and everything
06:18.31Gnat42is good
06:18.38Gnat42but doesn't enable my module.
06:19.18solarhey target/device/Via/epia-mii/kernel-patches/
06:19.23solarls target/device/Via/epia-mii/kernel-patches/
06:19.44Gnat42you want `ls target....` ?
06:20.12solarI want you to show me where the viafb.diff is
06:20.26Gnat42k
06:20.37Gnat42ls target/device/Via/epia-mii/kernel-patches/
06:20.45Gnat42= viafb_03.diff
06:20.53solarok thats wrong
06:21.03Gnat42wrong file name?
06:21.52solarmv viafb_03.diff 99_viafb_03.patch ; bzip2 99_viafb_03.patch
06:22.47solarthe patching is done in order by numeric and all patches are bzip2
06:23.03Gnat42oh
06:23.07solarnotice how they are in the Soekris/kernel-patches subdir
06:23.09Gnat42didn't know it mattered.
06:23.10Gnat42yes
06:23.16Gnat42didn't think to try it like that
06:23.19Gnat42lets see now...
06:23.23solaryou pretty much want those patches also
06:23.44solarmix and match erik patchsets. alot little goodies in there
06:24.01Gnat42oh?
06:24.02ashesthere's a new patch for shadow
06:24.13asheslastlog doesn't work for console
06:24.18solarbusybox shadow?
06:24.34Gnat42are these kernel patches or other patches?
06:24.34solarhey do you use java?
06:24.35asheshttp://cvs.pld.org.pl/shadow/libmisc/log.c?r1=1.7&r2=1.8
06:24.59ashesno, shadow-utils. it could be possible that busybox is affected
06:25.12Gnat42I don't use java
06:25.16Gnat42though I can
06:25.20Gnat42but don't need it
06:25.52asheslogins via tty are fine, but when logging in to a console lastlog doesn't work
06:26.06ashesits easy to check
06:26.20solarthanks ashes
06:35.26Gnat42couple more questions
06:35.36Gnat42lots of his patches are for kernel 2.4
06:35.41Gnat42I'm using 2.6.11.2
06:36.00Gnat42any known problems?
06:36.32Gnat42and q2 must I have a kernel name -erik (or whatever) after patching?
06:36.43Gnat42or can I leave the after patch name the same as the start?
06:37.40solarwhen people start talking about kernel 2.6 I lose whatever happiness that I had from thinking about soekris boxes being deployed
06:37.45Gnat42lol
06:37.47Gnat42why?
06:38.02solarincomming patch rate of the 2.6.x kernel
06:38.06Gnat42oh?
06:38.12Gnat42many patches
06:38.14solarit's more of a dev thing than production thing
06:38.21Gnat42I wasn't aware.
06:38.21kergothnot a fan of the new development model, i take it
06:38.42solarit's gear more towards development which is great
06:38.45Gnat42I figured that since it was in use in some of the *big* name distros it had stabalized
06:38.50solarI'm geared more towards security
06:39.04Gnat42not to start a distro war or anything
06:39.38Gnat42so it is a personal belief that ' so many patches so fast = not mature' type thing?
06:39.48kergothit has nothing to do with maturity.
06:40.47SpanKYsoooooooooooolar
06:40.54SpanKYanswer my Q in #-embedded
06:41.05Gnat42I'm a little at a loss as to why then
06:41.16Gnat42I thought 2.6 was stable enough
06:41.23Gnat42is there something I should look at to change my mind?
07:42.18Zta/home/sth/au/buildroot/build_arm/linux-2.4.28-pre4-erik/include/asm/rwsem.h:215: error: unknown register name `edx' in `asm'
07:43.12ZtaThen I get two assembler messages from standardinput that are a bad instruction:  `movl %cr0,r3'   and   `movl r3,%cr0'
07:43.32ZtaI get this error when compiling the linux package from buildroot.
07:43.54ZtaPerhaps I should try latest buildroot from svn
08:01.55*** part/#uclibc Gnat42 (~gnat@24.109.128.223)
08:13.30Zta/bin/sh: line 1: ./ntpdc-layout: cannot execute binary file
08:14.05ZtaI've been getting this error for the last three versions of buildroot, I've used =\
09:17.19*** join/#uclibc rgb (~rgb@Quebec-HSE-ppp230809.qc.sympatico.ca) [NETSPLIT VICTIM]
09:17.19*** join/#uclibc Tweedle-dee (~animeloe@animeloe.user) [NETSPLIT VICTIM]
09:17.19*** join/#uclibc hgb (~hgb@cm-80.111.5.236.chello.no) [NETSPLIT VICTIM]
09:17.19*** join/#uclibc telmich (~deadghost@telmich.gpm) [NETSPLIT VICTIM]
09:17.19*** join/#uclibc thomasez (~thomasez@castle.linpro.no) [NETSPLIT VICTIM]
09:18.02*** join/#uclibc mjn3-work (~mjn3@h-66-134-116-110.dnvtco56.covad.net) [NETSPLIT VICTIM]
09:18.02*** join/#uclibc Kaloz (kaloz@217.27.212.53) [NETSPLIT VICTIM]
09:19.03*** join/#uclibc CIA-1 (~CIA@flapjack.navi.cx) [NETSPLIT VICTIM]
09:21.02*** join/#uclibc kevc (~kev@kerouac.projectcolo.org.uk) [NETSPLIT VICTIM]
09:21.03*** join/#uclibc tahorg (~loic@zone4.gcu-squad.org) [NETSPLIT VICTIM]
09:22.42*** join/#uclibc Alexander__ (~Alex@203.101.41.151) [NETSPLIT VICTIM]
09:22.42*** join/#uclibc _drak0_ (~rob@ip68-98-148-202.dc.dc.cox.net) [NETSPLIT VICTIM]
09:28.26*** join/#uclibc mjn3 (~mjn3@c-67-164-173-93.client.comcast.net)
09:46.39ZtaMAN it's difficult to make buildroot compile the linux kernel!
09:59.06*** join/#uclibc SecretSquirrel (~squirrel@adsl-209-30-10-25.dsl.rcsntx.swbell.net)
10:00.35*** join/#uclibc mr_claus (random@p54995501.dip.t-dialin.net)
10:28.25ZtaHAH!
10:28.54ZtaI think I've fixed a bug in the package named "linux" that compiles the kernel.
10:29.16ZtaPrior to configuring the kernel, the linux.mk script does this:
10:29.16Zta$(SED) "s,^CROSS_COMPILE.*,CROSS_COMPILE=$(KERNEL_CROSS),g;" $(LINUXAU_DIR)/Makefile
10:29.27ZtaBut it forgot to do this:
10:29.28Zta$(SED) "s,^ARCH.*,ARCH=$(LINUXAU_KARCH),g;" $(LINUXAU_DIR)/Makefile
10:29.37ZtaNow it compiles =)
10:29.52ZtaI'm bad..
10:33.34*** join/#uclibc GyrosGeier (~geier@port-212-202-73-190.dynamic.qsc.de)
10:45.17ZtaIt is possible to store a uClibc.config and have buildroot copy that to .config and make oldconfig before compiling uClibc?  Similiar to how it's done with busybox and busybox.config?
10:46.01ZtaAh.. buildroot/toolchain/uClibc/uClibc.config
11:17.53*** join/#uclibc mr_claus_ (random@p549959D6.dip.t-dialin.net)
11:33.55*** join/#uclibc Alexander_ (~Alex@203.101.41.151)
12:21.47*** join/#uclibc SpanKY (~UserBah@pool-141-154-213-5.bos.east.verizon.net)
13:07.24*** join/#uclibc mr_claus_ (random@p549959D6.dip.t-dialin.net)
13:12.16*** join/#uclibc telmich (~deadghost@ns.schottelius.org) [NETSPLIT VICTIM]
13:12.18*** join/#uclibc hgb (~hgb@cm-80.111.5.236.chello.no) [NETSPLIT VICTIM]
13:12.31*** join/#uclibc Tweedle-dee (~animeloe@animeloe.user) [NETSPLIT VICTIM]
13:24.52*** join/#uclibc rgb (~rgb@Quebec-HSE-ppp230809.qc.sympatico.ca) [NETSPLIT VICTIM]
13:30.38*** join/#uclibc thomasez (~thomasez@castle.linpro.no) [NETSPLIT VICTIM]
13:47.58rfswhy is it that when you email someone to tell them their message was mangled badly by their mailer, do they assume that you want to help them with their problem...?
13:49.03rfswell, at least the re-posted message was not mangled this time ;)
15:23.57*** join/#uclibc bjbL (~bjbL@67.71.251.210)
17:07.30*** join/#uclibc sjhill (~sjhill@eth13.com-link.com)
18:04.44Sgt-Donanyop
18:15.16*** join/#uclibc mjn3-work (~mjn3@h-66-134-116-110.dnvtco56.covad.net)
18:39.27*** join/#uclibc andersee (~andersee@codepoet.org)
18:55.20*** join/#uclibc Gnat42 (~gnat@24.109.128.223)
18:56.05*** join/#uclibc Fidget001 (~rpage@adsl-214-114-38.gnv.bellsouth.net)
19:38.30Gnat42anyone have problems building iptables with binutils 2.15.94.0.2/gcc-3.4.3 using buildroot?
19:38.33Gnat42I get
19:38.46Gnat42extensions/libipt_DNAT.c:16: error: field `mr' has incomplete type
19:38.47Gnat42extensions/libipt_DNAT.c:232: error: invalid application of `sizeof' to incomplete type `ip_nat_multi_range_compat'
19:38.47Gnat42extensions/libipt_DNAT.c:233: error: invalid application of `sizeof' to incomplete type `ip_nat_multi_range_compat'
19:46.46mjn3-workwhat kernel headers?
19:52.41Sgt-Donanandersee are u here ?
19:53.54Sgt-Donanor maybe someone else could help me
19:54.31vapierno one can help you if you dont ask a question
19:55.08anderseeGnat42: I know what that is
19:55.13Gnat42oh?
19:55.28Gnat42kernel headers 2.6.11
19:55.29anderseeGnat42: Its a bug in the upstream linux-libc-headers package
19:55.34Gnat42okay
19:55.43Gnat42is it related to binutils?
19:55.50Gnat42or just the libc-headers?
19:55.52Sgt-Donani want to cross compile my m68k kernel on an x86, i have compiled a cross-m68k-gcc with crosstool at the end of the compilation it gives me error
19:56.11Sgt-Donancan i post the error ?
19:56.18Sgt-Donan~6 lines
19:56.23anderseelinux/include/linux/netfilter_ipv4/ip_nat.h is missing the definition for 'struct ip_nat_multi_range_compat'
19:56.39Gnat42so not related at all to binutils
19:56.43Gnat42do we have a patch?
19:56.50anderseeright.  The kernel headers needs a patch
19:58.30anderseeGnat42: http://codepoet.org/linux-libc-headers-2.6.11.0.patch
20:00.01Sgt-Donan(with the buildroot m68k toolchain i get the same error)
20:00.38vapierthe kernel fails to build ?
20:00.43Sgt-Donanyes
20:00.45vapierwhich version ?
20:00.55Sgt-Donan2.4.28
20:01.15Gnat42andersee: thanks, I copied it to toolchain/kernel-headers/linux-libc-headers-2.6.11.0.patch
20:01.26vapierhmm, i know mainline 2.6 for m68k does not work ... you need to get cvs work from the m68k website
20:01.28Gnat42will it apply automagically?
20:01.37vapieryou should try going to the m68k website and d/ling their 2.4 cvs tree
20:01.58solarhow do we feel about supporting __secure_getenv() ?
20:02.04Gnat42seems like it should
20:02.10Sgt-Donanof the kernel ?
20:02.25vapierSgt-Donan: what else would i be talking about :P
20:02.36Sgt-Donanyes but i could be the toolchain
20:02.41Sgt-Donanit
20:02.55Sgt-Donanok i try
20:03.00solarsolar designer keeps suggesting that upstream use __secure_getenv() on any program thats +s. We can't wrapper to simple getenv() cuz it would lower security
20:04.04Sgt-Donanvapier i have linux kernel binary for my board how does they compile it successfull
20:04.33Sgt-Donanon a 25Mhz 68040 it could take 1 day for compile kernel
20:05.03vapieri didnt say build it natively ...
20:05.10vapieri got a 33mhz 68040 :p
20:05.19CIA-103andersen * r10057 10buildroot/toolchain/kernel-headers/linux-libc-headers-2.6.11.0.patch:
20:05.19CIA-1Add missing definition for 'struct ip_nat_multi_range_compat' so
20:05.19CIA-1that iptables can compile
20:06.39Gnat42yeah
20:06.43Gnat42just testing the patch right now
20:06.50Gnat42supplied by andersee
20:06.54Gnat42oh
20:08.10Sgt-Donanvapier you have ?
20:08.23Sgt-Donanbut how do you get it work ?
20:08.29vapiermagic ?
20:08.41Sgt-Donanlol
20:09.19Sgt-Donani know that linux is magic but it doesn't make all itself
20:10.23vapiersure it does
20:10.31vapierthat's the magical part
20:11.05Sgt-Donanlol
20:11.08Sgt-Donansure
20:12.09vapieriiiiif you want you can pmsg me the error
20:12.18vapieri'll tell you if it's the fault of your toolchain or the kernel
20:12.58Sgt-Donanok
20:15.17vapierhmm ive seen that once before
20:15.22vapieri dont remember where though :)
20:15.59vapierbbl
20:16.22Sgt-Donank
20:42.24*** join/#uclibc d9rz (~user@82.152.95.50)
20:47.36*** join/#uclibc JockeHome (~JockeHome@84.217.0.212)
20:58.11*** join/#uclibc rgb (~rgb@Quebec-HSE-ppp230809.qc.sympatico.ca)
21:14.59*** part/#uclibc d9rz_ (~user@82.152.95.50)
22:17.12*** join/#uclibc kos_tom (~thomas@humanoidz.org)
22:50.43Sgt-Donanbn
23:07.53*** join/#uclibc jockeHome2 (~JockeHome@84.217.0.212)
23:09.25jockeHome2andersee, how do you feel about a little testing? see the path I just sent on uclibc list
23:10.16anderseejockeHome2: just saw the patch
23:10.39anderseeI do need to do some testing tonight, so I'll give it a try
23:10.50jockeHome2thanks
23:11.05jockeHome2fingers crossed :)
23:11.12andersee:)
23:11.21anderseeuhoh
23:11.25anderseebaby peter just woke up
23:11.32jockeHome2:(
23:12.22jockeHome2try using the silencer :)
23:13.35solarhey we can save 48 bytes per executable
23:13.37solarhttp://rafb.net/paste/results/AWt3ya87.txt
23:13.45Gnat42andersee: there is still a problem with the linux libc kernel headers 2.6.1
23:13.47solaron systems not needing java support
23:13.48Gnat422.6.11
23:14.53Gnat42the file kernel-headers.mk tries patching toolchain/patch-kernel.sh $(LINUX_HEADERS_DIR) toolchain/kernel-headers linux-libc-headers-$(LINUX_VERSION)-\*.patch
23:15.01Gnat42and you named the file .patch
23:15.14Gnat42don't know if you'd rather remove the last - before the \*.patch
23:15.21Gnat42or rename the headers patch
23:16.57jockeHome2solar, you up for a little testing?
23:19.52solarnew ldso stuff?
23:20.46jockeHome2yup
23:20.52jockeHome2see uclibc list
23:21.55jockeHome2possibly enable ldso debug first, early as well
23:27.12solarhey sorry about that
23:27.33solarnot sure I can do that kinda testing cuz it would take a full bootstrap
23:28.28JockeHomeWhy, you just need to rebuild ldso&libdl
23:29.11solardid you not say it broke the ABI and the existing bins would break?
23:29.58JockeHomeahh, that was the previous patch, but I modified it not to break anything yet
23:31.49solarI can test in about an hr
23:31.54solarwill you still be here?
23:32.36JockeHomemaybe, it is getting late.
23:34.07JockeHomeJust to be clear, the path to test is "[PATCH] Optimize relocation"
23:34.18JockeHomes/path/patch
23:35.28solarthat depends on SVN current or 0.9.27
23:35.53solarI can test 0.9.27 (mixed with gentoo patches pretty easy) but + cvs/svn takes a little more work.
23:37.05JockeHomeit is against current :(
23:37.36JockeHomewhats arches do you have?

Generated by irclog2html.pl by Jeff Waugh - find it at freshmeat.net! Modified by Tim Riker to work with blootbot logs, split per channel, etc.