hardened/linux/uclibc contains "nls" in the use.mask However, it is not possible to emerge hddtemp without the nls use flag; this is even mentioned in all ebuilds there: # disabling nls breaks compiling use nls || myconf="--disable-nls ${myconf}" I unmasked the nls flag, and voila: hddtemp emerges and runs fine. So what would be a feasible fix here? - Patch hddtemp, so it compiles without nls? - Remove the nls use flag from the ebuild and never econf with --disable-nls? Reproducible: Always
1) Please post your `emerge --info' output in a comment. 2) Please attach the entire (failing) build log to this bug report.
Created attachment 338996 [details] build.log
emerge --info says: Portage 2.1.11.50 (hardened/linux/uclibc/x86, gcc-4.6.3, uclibc-0.9.33.2-r2, 3.5.7-gentoo i686) ================================================================= System uname: Linux-3.5.7-gentoo-i686-QEMU_Virtual_CPU_version_1.2.2-with-gentoo-2.1 KiB Mem: 2074136 total, 1129316 free KiB Swap: 0 total, 0 free Timestamp of tree: Fri, 15 Feb 2013 09:45:01 +0000 ld GNU ld (GNU Binutils) 2.22 ccache version 3.1.8 [disabled] app-shells/bash: 4.2_p37 dev-lang/python: 2.7.3-r2, 3.2.3 dev-util/ccache: 3.1.8 dev-util/cmake: 2.8.9 dev-util/pkgconfig: 0.28 sys-apps/baselayout: 2.1-r1 sys-apps/openrc: 0.11.8 sys-apps/sandbox: 2.5 sys-devel/autoconf: 2.13, 2.69 sys-devel/automake: 1.10.3, 1.11.6 sys-devel/binutils: 2.22-r1 sys-devel/gcc: 4.6.3 sys-devel/gcc-config: 1.7.3 sys-devel/libtool: 2.4-r1 sys-devel/make: 3.82-r4 sys-kernel/linux-headers: 3.6 (virtual/os-headers) sys-libs/uclibc: 0.9.33.2-r2 Repositories: gentoo ACCEPT_KEYWORDS="x86" ACCEPT_LICENSE="* -@EULA Oracle-BCLA-JavaSE" CBUILD="i686-gentoo-linux-uclibc" CFLAGS="-O2 -march=geode -pipe -fomit-frame-pointer -mmmx -m3dnow" CHOST="i686-gentoo-linux-uclibc" CONFIG_PROTECT="/etc /usr/share/gnupg/qualified.txt" CONFIG_PROTECT_MASK="/etc/ca-certificates.conf /etc/env.d /etc/fonts/fonts.conf /etc/gconf /etc/gentoo-release /etc/revdep-rebuild /etc/sandbox.d /etc/terminfo" CXXFLAGS="-O2 -march=geode -pipe -fomit-frame-pointer -mmmx -m3dnow" DISTDIR="/usr/portage/distfiles" FCFLAGS="-O2 -pipe" FEATURES="assume-digests binpkg-logs collision-protect compress-build-logs config-protect-if-modified distlocks ebuild-locks fixlafiles merge-sync news parallel-fetch protect-owned sandbox sfperms strict unknown-features-warn unmerge-logs unmerge-orphans userfetch xattr" FFLAGS="-O2 -pipe" GENTOO_MIRRORS="ftp://ftp.wh2.tu-dresden.de/pub/mirrors/gentoo" LDFLAGS="-Wl,-O1 -Wl,--as-needed -Wl,--hash-style=gnu" MAKEOPTS="-j3" PKGDIR="/usr/portage/packages" PORTAGE_CONFIGROOT="/" PORTAGE_RSYNC_OPTS="--recursive --links --safe-links --perms --times --compress --force --whole-file --delete --stats --human-readable --timeout=180 --exclude=/distfiles --exclude=/local --exclude=/packages" PORTAGE_TMPDIR="/var/tmp" PORTDIR="/usr/portage" PORTDIR_OVERLAY="" SYNC="rsync://rsync.europe.gentoo.org/gentoo-portage" USE="cli cracklib crypt cxx dri fortran hardened iconv ipv6 modules mudflap ncurses nptl openmp pax_kernel pcre pic readline session ssl uclibc unicode x86 xattr zlib" ALSA_PCM_PLUGINS="adpcm alaw asym copy dmix dshare dsnoop empty extplug file hooks iec958 ioplug ladspa lfloat linear meter mmap_emul mulaw multi null plug rate route share shm softvol" APACHE2_MODULES="authn_core authz_core socache_shmcb unixd actions alias auth_basic authn_alias authn_anon authn_dbm authn_default authn_file authz_dbm authz_default authz_groupfile authz_host authz_owner authz_user autoindex cache cgi cgid dav dav_fs dav_lock deflate dir disk_cache env expires ext_filter file_cache filter headers include info log_config logio mem_cache mime mime_magic negotiation rewrite setenvif speling status unique_id userdir usertrack vhost_alias" CALLIGRA_FEATURES="kexi words flow plan sheets stage tables krita karbon braindump" CAMERAS="ptp2" COLLECTD_PLUGINS="df interface irq load memory rrdtool swap syslog" CURL_SSL="gnutls" ELIBC="uclibc" GPSD_PROTOCOLS="ashtech aivdm earthmate evermore fv18 garmin garmintxt gpsclock itrax mtk3301 nmea ntrip navcom oceanserver oldstyle oncore rtcm104v2 rtcm104v3 sirf superstar2 timing tsip tripmate tnt ubx" INPUT_DEVICES="keyboard mouse" KERNEL="linux" LCD_DEVICES="bayrad cfontz cfontz633 glk hd44780 lb216 lcdm001 mtxorb ncurses text" LIBREOFFICE_EXTENSIONS="presenter-console presenter-minimizer" LINGUAS="de en_GB" LIRC_DEVICES="serial" PHP_TARGETS="php5-3" PYTHON_SINGLE_TARGET="python2_7" PYTHON_TARGETS="python2_7 python3_2" RUBY_TARGETS="ruby18 ruby19" SANE_BACKENDS="snapscan" USERLAND="GNU" XTABLES_ADDONS="quota2 psd pknock lscan length2 ipv4options ipset ipp2p iface geoip fuzzy condition tee tarpit sysrq steal rawnat logmark ipmark dhcpmac delude chaos account" Unset: CPPFLAGS, CTARGET, EMERGE_DEFAULT_OPTS, INSTALL_MASK, LANG, LC_ALL, PORTAGE_BUNZIP2_COMMAND, PORTAGE_COMPRESS, PORTAGE_COMPRESS_FLAGS, PORTAGE_RSYNC_EXTRA_OPTS, USE_PYTHON
I'll take a look into hddtemp source. If it is easy fix, I can patch it; otherwise I'll remove 'nls' USE flag.
I have decided to remove nls flag. Getting rid of iconv depended code wouldn't be very problematic, but anyway I don't think it's worth it. +*hddtemp-0.3_beta15-r23 (06 Jul 2013) +*hddtemp-0.3_beta15-r6 (06 Jul 2013) + + 06 Jul 2013; Amadeusz Żołnowski <aidecoe@gentoo.org> + hddtemp-0.3_beta15-r3.ebuild, hddtemp-0.3_beta15-r5.ebuild, + +hddtemp-0.3_beta15-r6.ebuild, hddtemp-0.3_beta15-r22.ebuild, + +hddtemp-0.3_beta15-r23.ebuild, files/hddgentoo.db, +files/hddtemp.service: + Added systemd unit file (by Pacho Ramos <pacho@gentoo.org>) + - fixes bug #474458. + + Removed 'nls' flag from IUSE, because it doesn't build if disabled + - fixes bug #457648. + + Added WDC WD3200BEKT-75A25T0 and Hitachi HTS542512K9SA00 to hddgentoo.db + - fixes bug #470970 and bug #473686. +
Um, it works fine on most profiles, so why is it now impossible for everyone to set USE=-nls, instead of masking the USE flag in the affected profiles? checking for iconv... yes checking how to link with libiconv... /usr/lib/libiconv.so It checks that nicely, but the linker command in src/Makefile doesn't pick that up. We can surely fix this: sed -i src/Makefile.am -e '/hddtemp_LDADD/s:$: @LIBICONV@:g' || die Does this help? You would also need to add virtual/libiconv to DEPEND. Also missing are gettext deps.
(In reply to Jeroen Roovers from comment #6) > Um, it works fine on most profiles, so why is it now impossible for everyone > to set USE=-nls, instead of masking the USE flag in the affected profiles? OK, I'll try to restore and fix it if having such an option is so important. :-)
(In reply to Amadeusz Żołnowski from comment #7) > (In reply to Jeroen Roovers from comment #6) > > Um, it works fine on most profiles, so why is it now impossible for everyone > > to set USE=-nls, instead of masking the USE flag in the affected profiles? > > OK, I'll try to restore and fix it if having such an option is so important. > :-) No, really, the ebuilds were broken already (missing deps, automagic deps) and now you've made it worse. Reverting to the old state would *improve* the ebuilds slightly. Your recent changes didn't improve anything.
I have restored nls flag. I hope I haven't messed up too much. :-) Next time when I am going to have a bit more time, I'll try to actually fix the problem. + 08 Jul 2013; Amadeusz Żołnowski <aidecoe@gentoo.org> + hddtemp-0.3_beta15-r3.ebuild, hddtemp-0.3_beta15-r5.ebuild, + hddtemp-0.3_beta15-r6.ebuild, hddtemp-0.3_beta15-r22.ebuild, + hddtemp-0.3_beta15-r23.ebuild: + Restore nls flag in IUSE. Rels bug #457648. +
I see you didn't add the sed script that would fix the case of having dev-libs/libconv and not sys-libs/uclibc[iconv]?
No, what's going on here is that uclibc is highly configurable and its locales support can be turned off. Since it is in bad shape, I did so on the stage3 images under /experimental and used the freebsd breakout library dev-lib/libiconv. This is an adequate solution with one problem: you need CFLAGS="-liconv". There are quite a few of these so here's my recommendation: 1) make sure the ebuilds build against glibc. do as little possible to make them work against uclibc unless there's a good reason (and there is eg glibc-ism which are just dumb). 2) for breakage because of breakout libraries (there linker doesn't know where to find some symbol because its not in libc), then add CFLAGS="-liconv" to the cflags. Do this on a per package basis by adding: app-admin/hddtemp hddtemp.conf to /etc/portage/package.env and then adding CFLAGS="-liconv" to /etc/portage/env/hddtemp.conf. Here libiconv was an example. Expect breackage with argp (from argp-standalone) and libintl (from gettext). 3) users will get annoyed, so if you want this in the ebuild, then append-cflags when USE=elibc_uclibc and DEPEND on uclibc? ( dev-lib/libiconv ) but what happens when locales is turned on in uclibc's config? boom! The configurability of uclibc is a two edged sword.
(In reply to Anthony Basile from comment #11) > then add CFLAGS="-liconv" to the cflags hddtemp already checks for this, but fails to include @LIBICONV@ when it matters. Comment #6 has the trivial patch to fix this bug report. I tested it. It works on some setups of uclibc +/- dev-libs/libiconv. ;-) But you want to actually not support this fully (because of the complications adding proper dependencies would introduce)? Then shouldn't this be assigned to embedded@ and simply set to WONTFIX? Finally, even in the glibc case, we don't DEPEND on gettext yet, and there may be an RDEPEND missing, too.
(In reply to Jeroen Roovers from comment #12) > (In reply to Anthony Basile from comment #11) > > then add CFLAGS="-liconv" to the cflags > > hddtemp already checks for this, but fails to include @LIBICONV@ when it > matters. Comment #6 has the trivial patch to fix this bug report. I tested > it. It works on some setups of uclibc +/- dev-libs/libiconv. ;-) > Okay, I'll look at hddtemp later and see. Looks like this is a particular case. > But you want to actually not support this fully (because of the > complications adding proper dependencies would introduce)? Then shouldn't > this be assigned to embedded@ and simply set to WONTFIX? I really need to sort out the locales issue in uclibc and eventually drop the breakout library. This would be the correct long term fix. However, my comments about other breakout libraries still applies, like argp. > > Finally, even in the glibc case, we don't DEPEND on gettext yet, and there > may be an RDEPEND missing, too. Probably correct.
for now, please use the workaround in comment #11
