IRC log for #uclibc on 20080929

00:12.25*** join/#uclibc tchan (n=tchan@lunar-linux/developer/tchan)
01:01.18*** join/#uclibc hiyuh (n=hiyuh@KD059133117089.ppp.dion.ne.jp)
05:15.02*** join/#uclibc doc|schaffen (n=doc@82.115.100.147)
06:43.52*** join/#uclibc vinilios (n=k88@adsl77-101.ath.forthnet.gr)
07:27.54*** topic/#uclibc by blindvt`_ -> discussion of uClibc and Busybox | uClibc 0.9.29 released 6 May 2007, 0.9.30-rc2 == trunk; PLEASE TEST! | busybox 1.12.1 28 September 2008 | for non uClibc/busybox such as buildroot and general setup issues try #elinux or #edev
08:16.53CIA-903aldot * r23548 10uClibc/libc/sysdeps/linux/ (alpha/bits/atomic.h mips/bits/atomic.h): - fix ISO C keywords
08:20.58*** join/#uclibc likewise (n=likewise@82-171-51-231.ip.telfort.nl)
08:51.32CIA-903carmelo * r23549 10uClibc/libc/misc/wchar/wchar.c: (log message trimmed)
08:51.32CIA-9i18n: Fix mbrtowc function to handle 0xc0 and 0xc1 sequence.
08:51.32CIA-9When the first byte of the multibyte sequence start with 0xc0
09:37.08*** join/#uclibc k88_ (n=k88@212.89.168.145)
09:57.12*** join/#uclibc k88_ (n=k88@adsl77-101.ath.forthnet.gr)
11:07.38blindvt`_hi
11:08.19fabohi blindvt`_
11:23.50*** join/#uclibc likewise (n=likewise@82-171-51-231.ip.telfort.nl)
11:36.50CIA-903aldot * r23550 10uClibc/libc/sysdeps/linux/i386/ (Makefile.arch posix_fadvise.c posix_fadvise64.S): - now passes all LTP tests
11:39.34blindvt`_on somewhat recent kernels, i may add. Although i even tested it on 2.6.11, i did not try something older
11:56.05CIA-903carmelo * r23551 10uClibc/ldso/ldso/ldso.c:
11:56.05CIA-9ldso: do not use hard-coded fd in _dl_dprintf. Use dl_debug_file consinstently.
11:56.05CIA-9Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
11:58.18CIA-903carmelo * r23552 10uClibc/ldso/libdl/libdl.c:
11:58.18CIA-9libdl: use stderr consistently.
11:58.18CIA-9Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
12:09.53CIA-903carmelo * r23553 10uClibc/ldso/ldso/sh/elfinterp.c:
12:09.53CIA-9ldso: allow undefined references to weak symbols
12:09.53CIA-9Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
12:41.22blindvt`_HcE, ping
12:41.22HcEblindvt`_: You sent me a contentless ping.  This is a contentless pong.  Please provide a bit of information about what you want and I will respond when I am around.
12:42.02blindvt`_HcE, i cannot reproduce http://bugs.busybox.net/view.php?id=4314
12:43.27HcEblindvt`_: I thought that was fixed?
12:43.41HcEblindvt`_: hmmm, check if glob resets errno
12:43.42blindvt`_HcE, is this your toolchain miscompiling something or a real issue somewhere?
12:44.06HcEblindvt`_: the user had not set the UCLIBC_HAS_GNU_GLOB, but used glob like it was set
12:44.10blindvt`_HcE, it's not really fixed for he turned off glob-susv3
12:51.57blindvt`_HcE, where did he use GNU_GLOB specific stuff? I don't see any
12:55.06HcEuClibc?
12:55.11HcEperhaps it is removed in trunk?
12:55.18HcEthis is 0.9.29 remember
12:57.18blindvt`_HcE, hm. Could be but that's not really satisfying
12:57.57HcEIn Big and tall, I can choose:
12:57.58HcEUCLIBC_HAS_GLOB=y
12:57.58HcEUCLIBC_HAS_GNU_GLOB=y
12:58.11blindvt`_HcE, # UCLIBC_HAS_GNU_GLOB is not set
12:59.46CIA-903carmelo * r23554 10uClibc/ldso/ldso/sh/elfinterp.c: Fix comment.
13:00.22blindvt`_HcE, would be awesome if you see it with 0.9.29 but not trunk and can pinpoint the fix to make sure that it's not some fallout of a broken^W^Whis toolchain but that it's really fixed
13:04.11CIA-903aldot * r23555 10uClibc/libc/sysdeps/linux/common/ssp.c: - honour UCLIBC_HAS_SYSLOG
13:19.36CIA-903carmelo * r23556 10uClibc/ldso/ldso/sh/elfinterp.c: sh_ldso: Fix typo: missing semi-colon.
13:19.55*** join/#uclibc hiyuh (n=hiyuh@KD125054017176.ppp-bb.dion.ne.jp)
13:20.11blindvt`_HcE, likewise for 4514
13:20.44blindvt`_HcE, mind if i assign all those atmel stuff to you?
13:31.35HcEblindvt`_: sure, go ahead, but I am a bit stuck for two weeks, I forgot we have training here these weeks :/
13:32.49blindvt`_HcE, aha, so we will ship without further atmel updates, IIUC?
13:36.22HcEonly bugfixes
13:36.37HcEI'll have time to go through the test framework
13:39.23blindvt`_k
13:50.05HcEshould I close the bug as well if a bug is fixed and verified by my co-worker?
13:52.02blindvt`_HcE, if it's fixed then yes, please close it and set the version in which it will be fixed accordingly
13:53.02blindvt`_HcE, i.e. 0.9.30 if it's fixed on trunk. TIA
13:56.15HcEblindvt`_: if it is not a uClibc bug, but toolchain bug?
13:56.42HcEhmm, perhaps I should have set it to something completely different :/
13:59.26HcEblindvt`_: why did I get the gethostbyname() stuff?
14:00.12HcEblindvt`_: and please only assigne AVR32 stuff to me, I have little knowledge of ARM
14:00.24blindvt`_HcE, if it's a toolchain bug then close it with "No fix required" or something like that. In these cases please state which toolchain version the user is supposed to use
14:01.07blindvt`_HcE, i only assigned that "0004514: hang when sending request to netlink socket on AVR32 NGW100 dev kit " to you
14:01.31blindvt`_HcE, i've updated the resolver bug, you just got a copy since you listen to -cvs
14:01.46blindvt`_HcE, ok, no ARMs for you :)
14:02.25blindvt`_HcE, oh and please *close* stuff. You set it to "resolved" which still makes it show up per default.
14:04.38HcEblindvt`_: I just closed it :)
14:28.26*** join/#uclibc wberrier (n=wberrier@12.191.193.40)
14:34.41CIA-903aldot * r23557 10uClibc/extra/Configs/Config.m68k: - do not pass bitwise-or to the assembler. Closes #3164
15:07.43HcEhmmm, uClibc makefile in BR2 seems a bit broken, does $(cross_compiler) really exist? I think it should be cross_compiler without the $()
15:19.10blindvt`_HcE, no, that's in fact ok. $(cross_compiler) refers to the binary sometimes later as opposed to cross_compiler (the target) which is handy for me
15:19.38blindvt`_HcE, the $(cross_compiler) is needed for toplevel parallel make to get the deps right.
15:20.05HcEhmm, okay
15:20.47blindvt`_HcE, i guess landley broke your uclibc, didn't he ;)
15:23.39blindvt`_~tell landley i'm looking forward to you adding ilnuxthreads.new (or nptl) support for m68k ;)
15:23.58HcEhihi
15:23.59blindvt`_~seen landley
15:24.00ibotlandley <n=landley@static-71-162-243-5.phlapa.fios.verizon.net> was last seen on IRC in channel #uclibc, 1d 17h 17m 4s ago, saying: 'Hello world!'.
15:25.04blindvt`_ibot, it's a pity that you don't keep or relay msgs. *sigh*
15:37.27HcEhmmm
15:37.37HcEdoes make headers in uClibc now require a working compiler?
15:40.27faboHcE: broken recently. I built a rootfs using buildroot+uclibc trunk friday, not today ...
15:40.59blindvt`_HcE, no it doesn't. later on pregen does
15:41.32fabomaybe commit r23534
15:41.43HcEaha
15:41.48HcEI grabbed latest snapshot
15:41.50HcEsilly me
15:43.15blindvt`_it's just once again svn lagging a bit behind *shrug*
15:43.25solarblindvt`_: was r23550 you? if so thanks
15:44.11HcE-install_dev: install_headers
15:44.11HcE+install_dev: install_headers all
15:44.14HcElooks fishy
15:45.22blindvt`_solar, yes that was me. I didn't test it on 2.4.x though, only 2.6.26.5 and 2.6.11-suse-something
15:46.17blindvt`_solar, sorry that it took me so long to provide an accurate impl..
15:46.30HcEyepp, the r23534 commit is not correct
15:47.24HcEit assumes you have a cross compiler, which is a bit tricky to make without headers
15:47.47blindvt`_solar, please holler if you still encounter problems
15:48.38solargetting near a rc2 ?
15:48.51solarthats a better time for me to test
15:49.33blindvt`_solar, i'll do an rc2 during this weekend or early next week
15:49.48HcEurk, something is wrong with the Linux headers as well
15:49.50HcEhmpf
15:51.02blindvt`_solar, i'm aiming for friday evening, but i'm getting into the early-november-birthday-mindfield starting thursday, so i'm not yet sure how friday evening will work out
15:51.20blindvt`_s/november/october/;#
15:51.28*** join/#uclibc pcgeil (n=steffen@p549E6730.dip.t-dialin.net)
15:53.43solarskip friday. Tuesdays and Wednesdays oddly enough are the best days of the week to release (middle of work week)
15:54.19solarnot just for me. But the same holds true for most vendor security updates as well
15:58.05*** join/#uclibc pcgeil (n=steffen@p549E6730.dip.t-dialin.net)
16:01.56HcEin SVN, is there a smooth way of reverting a previous version? Or should I just do the change myself and commit with a message with says "reverting r<num>"?
16:02.25blindvt`_HcE, followup landleys commit first and discuss reverting it, please
16:02.55HcEblindvt`_: on email you mean? For some reason, I do not have it :/
16:03.28blindvt`_HcE, while he's doing it in a inelegant way him doing that "all" doesn't really matter
16:04.35HcEmy brain got a parse error on that sentence
16:05.24blindvt`_HcE, that "all" dep is not elegant but does not impose any real damage. Fix your buildscript instead
16:05.39HcEblindvt`_: Buildroot in that matter then
16:07.36blindvt`_HcE, I've forwarded that commit message to you
16:07.53HcEthanks
16:08.13HcEI'll see if I can massage Buildroot to install headers in the right order
16:08.28HcEit should not need the uClibc headers before making the gcc-final
16:08.52HcEso it should be kernel headers, gcc bootstrap, binutils, uClibc and gcc-final
16:09.17blindvt`_HcE, let me explicitely point out that i've fixed this recently in my tree, fwiw. Perhaps that saves you some headache
16:10.21HcEsmooth
16:11.09blindvt`_that's what i mean with svn lagging behind, once again :P
16:11.20HcEhehe, well, did you commit it? ;)
16:11.58blindvt`_HcE, i'm pretty sure i did. mompls
16:12.18blindvt`_HcE, yes, i did.
16:14.16HcEhmmm
16:14.37HcEweird, I grabbed the snapshot from today
16:14.41HcEdid I hit the wall again?
16:27.49CIA-903egtvedt * r23558 10buildroot/package/qtopia4/qtopia4.mk:
16:27.49CIA-9qtopia4: bump version to 4.4.3
16:27.49CIA-9Signed-off-by: Hans-Christian Egtvedt <hans-christian.egtvedt@atmel.com>
16:40.37HcEI could not find something about the build order of the toolchain in your GIT repo or SVN for BR2
16:41.25HcEs/I/blindvt`_: I/
16:41.28HcEhihi
16:43.41blindvt`_HcE, it's the very last commit; fire up gitk or look at git log
16:44.25blindvt`_HcE, the solution is easier than you think
16:44.37HcEthanks
16:44.53HcEI did not really connect the commit subject to the change :/
16:44.57HcEsorry for making noise
16:45.30blindvt`_HcE, no reordering, no fancy stuff. Heck, _I_ should ask you all that stuff, it's you who's lucky enough to be payed for that stuff, after all :-|
16:46.23HcEhihi
16:46.32HcEI attacked the problem from the wrong angle
16:46.51blindvt`_yes, you did indeed. It's dead simple, as you can see
16:46.53HcEassume BR was fundamental wrong, instead it was the uclibc makefile which was wrong
16:47.03HcEbut it is not committed upstream
16:47.06HcEI'll do that
16:47.35blindvt`_fabo, that's not really constructive
16:49.20blindvt`_fabo, there is absolutely no point in jumping up and down WRT threading right now. zilch reason, nada, überhaupt kei nwieauchimmer gearteter Grund
16:50.48blindvt`_fabo, http://uclibc.org/lists/uclibc/2008-September/020191.html
16:52.02blindvt`_adds a filter to thrash everything containing "/(linuxthreads|nptl)" to /dev/null
16:57.33CIA-903aldot * r23559 10uClibc/libc/sysdeps/linux/m68k/sys/user.h: - add a copy of user_regs
17:03.20*** join/#uclibc bbradley (n=bbradley@78-105-167-16.zone3.bethere.co.uk)
17:06.16*** join/#uclibc Redhatter (n=vk4msl@2001:388:f000:0:0:0:0:279)
17:23.19*** join/#uclibc pcgeil (n=steffen@p549E6730.dip.t-dialin.net)
19:11.47*** join/#uclibc landley (n=landley@static-71-162-243-5.phlapa.fios.verizon.net)
19:46.15*** join/#uclibc likewise (n=likewise@82-171-51-231.ip.telfort.nl)
20:32.19*** join/#uclibc pcgeil (n=steffen@p549E6730.dip.t-dialin.net)
20:41.00faboblindvt`_: the point is related to testing linuxthreads as suggested by some people on the ml. the results is linuxthread is broken for me (at least) and the conclusion is : IMHO, it's a bit premature to deprecate one of linuxthread implementation.
20:41.17fabonothing more, nothing less
21:02.59landleyis happy to remove the _new_ implementation instead.
21:03.07landleyThey're both being deprecated by NPTL when it arrives.
21:03.16landleyPutting lots of effort into pthreads at this point seems kind of silly, that's all.
21:12.50*** join/#uclibc pcgeil_ (n=steffen@p549E5C8D.dip.t-dialin.net)
21:38.49*** join/#uclibc ivob (n=ivob@unaffiliated/tiny)

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