Try to migrate to "True multilib" on my system after last portage update. All was fine except QT (dev-qt/qt-sql in particularly). It was failed on configure a 32-bit version. Reproducible: Always Steps to Reproduce: 1. emerge --update --deep --newuse --complete-graph --with-bdeps=y --quiet-build=y --pretend @world 2. See this particular output. (I'm sorry. Error was happen, when i already was in deep a @world update after set all needed ABI_X86 flags in /etc/portage/packages.use/*. So you may not able to reproduce this exactly). Calculating dependencies... done! [ebuild N ~] dev-qt/qtsql-4.8.6-r1 USE="debug exceptions oci8 postgres qt3support sqlite (-aqua) -freetds -mysql -odbc -pch" ABI_X86="32 (64) (-x32)" [ebuild N ~] dev-qt/qttest-4.8.6-r1 USE="debug exceptions (-aqua) -pch" ABI_X86="(64) -32 (-x32)" [ebuild N ~] dev-qt/qtxmlpatterns-4.8.6-r1 USE="debug (-aqua) -pch" ABI_X86="32 (64) (-x32)" [ebuild N ~] dev-qt/qt3support-4.8.6-r1 USE="accessibility debug exceptions (-aqua) -pch" ABI_X86="32 (64) (-x32)" [ebuild N ~] dev-qt/qtsvg-4.8.6-r1 USE="accessibility debug exceptions (-aqua) -pch" ABI_X86="(64) -32 (-x32)" [ebuild N ~] dev-qt/designer-4.8.6-r1 USE="debug exceptions qt3support (-aqua) -declarative -kde -pch -phonon -webkit" ABI_X86="(64) -32 (-x32)" [ebuild N ~] dev-qt/qtwebkit-4.8.6-r1 USE="debug exceptions gstreamer jit (-aqua) -icu -pch" ABI_X86="32 (64) (-x32)" [ebuild N ~] dev-qt/qtopengl-4.8.6-r1 USE="debug exceptions qt3support (-aqua) -egl -pch" ABI_X86="(64) -32 (-x32)" [ebuild N ~] dev-qt/qtmultimedia-4.8.6-r1 USE="alsa debug exceptions (-aqua) -pch" ABI_X86="(64) -32 (-x32)" [ebuild N ~] dev-qt/qtdeclarative-4.8.6-r1 USE="accessibility debug exceptions qt3support (-aqua) -pch -webkit" ABI_X86="(64) -32 (-x32)" [ebuild U ] sys-auth/polkit-kde-agent-0.99.1-r1 [0.99.1] 3. emerge --update --deep --newuse --complete-graph --with-bdeps=y --quiet-build=y @world 4. Failed on following error OCI auto-detection... () x86_64-pc-linux-gnu-g++ -c -pipe -march=native -O2 -pipe -O2 -Wall -W -I../../../../qt-everywhere-opensource-src-4.8.6/mkspecs/linux-g++ -I../../../../qt-everywhere-opensource-src-4.8.6/config.tests/unix/oci -I/usr/lib64/oracle/11.2.0.4/client/rdbms/public -I../../../../qt-everywhere-opensource-src-4.8.6/config.tests/unix/oci -I. -o oci.o ../../../../qt-everywhere-opensource-src-4.8.6/config.tests/unix/oci/oci.cpp x86_64-pc-linux-gnu-g++ -Wl,-O1 -Wl,--as-needed -Wl,-O1 -o oci oci.o -L/usr/lib64/oracle/11.2.0.4/client/lib32 -lclntsh /usr/lib/gcc/x86_64-pc-linux-gnu/4.8.3/../../../../x86_64-pc-linux-gnu/bin/ld: skipping incompatible /usr/lib64/oracle/11.2.0.4/client/lib32/libclntsh.so when searching for -lclntsh /usr/lib/gcc/x86_64-pc-linux-gnu/4.8.3/../../../../x86_64-pc-linux-gnu/bin/ld: cannot find -lclntsh collect2: error: ld returned 1 exit status Makefile:98: recipe for target 'oci' failed gmake: *** [oci] Error 1 OCI disabled. Oracle (OCI) support cannot be enabled due to functionality tests! Turn on verbose messaging (-v) to /var/tmp/portage/dev-qt/qtsql-4.8.6-r1/work/qt-everywhere-opensource-src-4.8.6/configure to see the final report. If you believe this message is in error you may use the continue switch (-continue) to /var/tmp/portage/dev-qt/qtsql-4.8.6-r1/work/qt-everywhere-opensource-src-4.8.6/configure to continue. * ERROR: dev-qt/qtsql-4.8.6-r1::gentoo failed (configure phase): * configure failed * * Call stack: * ebuild.sh, line 93: Called src_configure * environment, line 3181: Called multilib-minimal_src_configure * environment, line 2045: Called multilib_foreach_abi 'multilib-minimal_abi_src_configure' * environment, line 2252: Called multibuild_foreach_variant '_multilib_multibuild_wrapper' 'multilib-minimal_abi_src_configure' * environment, line 1976: Called _multibuild_run '_multilib_multibuild_wrapper' 'multilib-minimal_abi_src_configure' * environment, line 1974: Called _multilib_multibuild_wrapper 'multilib-minimal_abi_src_configure' * environment, line 389: Called multilib-minimal_abi_src_configure * environment, line 2039: Called multilib_src_configure * environment, line 2472: Called qt4_multilib_src_configure * environment, line 2958: Called die * The specific snippet of code: * "${S}"/configure "${conf[@]}" || die "configure failed"; * * If you need support, post the output of `emerge --info '=dev-qt/qtsql-4.8.6-r1::gentoo'`, * the complete build log and the output of `emerge -pqv '=dev-qt/qtsql-4.8.6-r1::gentoo'`. * The complete build log is located at '/var/tmp/portage/dev-qt/qtsql-4.8.6-r1/temp/build.log'. * The ebuild environment file is located at '/var/tmp/portage/dev-qt/qtsql-4.8.6-r1/temp/environment'. * Working directory: '/var/tmp/portage/dev-qt/qtsql-4.8.6-r1/work/qt-everywhere-opensource-src-4.8.6-abi_x86_32.x86' I suppose that problem in qt4-build-multilib.eclass file in qt4_get_mkspecs() function (at the end of file). For ABI_X86="32" it return "linux-g++" specification, that in QT source directory does not contain any flags to do 32-bit object code ("-m32"). I think it should be returned "linux-g++-32"! For 64-bit compilation it is returned correct "linux-g++-64" result and any build is finished fine. Please look on Makefile and oci.o in ./config.tests/unix/oci. ~ head -n 10 Makefile ############################################################################# # Makefile for building: oci # Generated by qmake (2.01a) (Qt 4.8.6) on: ?? ??? 30 15:58:28 2015 # Project: ../../../../qt-everywhere-opensource-src-4.8.6/config.tests/unix/oci/oci.pro # Template: app # Command: /var/tmp/portage/dev-qt/qtsql-4.8.6-r1/work/qt-everywhere-opensource-src-4.8.6-abi_x86_32.x86/bin/qmake -spec ../../../../qt-everywhere-opensource-src-4.8.6/mkspecs/linux-g++ CONFIG+= CONFIG-=debug_and_release LIBS*=\ -L/usr/lib64/oracle/11.2.0.4/client/lib32 LIBS+= INCLUDEPATH*=\ /usr/lib64/oracle/11.2.0.4/client/include QMAKE_CXXFLAGS*= QMAKE_CXXFLAGS+= QMAKE_CC=x86_64-pc-linux-gnu-gcc QMAKE_CXX=x86_64-pc-linux-gnu-g++ QMAKE_LINK=x86_64-pc-linux-gnu-g++ QMAKE_CFLAGS+=-march=native\ -O2\ -pipe QMAKE_CXXFLAGS+=-march=native\ -O2\ -pipe QMAKE_LFLAGS+=-Wl,-O1\ -Wl,--as-needed -o Makefile ../../../../qt-everywhere-opensource-src-4.8.6/config.tests/unix/oci/oci.pro ############################################################################# ####### Compiler, tools and options ~ less oci.o Header ELF: Magic: 7f 45 4c 46 02 01 01 00 00 00 00 00 00 00 00 00 Class: ELF64 Data: 2's complement, little endian Version: 1 (current) OS/ABI: UNIX - System V I set "stupid" workaround for this, and my @world update is finished correctly and all dependencies from QT works fine. But, please, thinks many times to set it for you. I suppose that more experienced developers should correct it in another way. diff -Naur /usr/portage/eclass/qt4-build-multilib.eclass.bak /usr/portage/eclass/qt4-build-multilib.eclass --- /usr/portage/eclass/qt4-build-multilib.eclass.bak 2015-03-23 12:31:10.000000000 +1000 +++ /usr/portage/eclass/qt4-build-multilib.eclass 2015-03-30 23:39:04.516952386 +1000 @@ -815,6 +815,13 @@ die "qt4-build-multilib.eclass: unsupported compiler '$(tc-getCXX)'" ;; esac + # Add -32 for explicit 32-bit flag +# ewarn "MULTIBUILD_VARIANT=${MULTIBUILD_VARIANT}" + if [[ "${MULTIBUILD_VARIANT}" == "abi_x86_32.x86" ]] + then + [[ -d ${S}/mkspecs/${spec}-32 ]] && spec+=-32 + fi + # Add -64 for 64-bit prefix profiles if use amd64-linux || use ia64-linux || use ppc64-linux || use x64-macos || My emerge --info ~ # emerge --info Portage 2.2.14 (python 2.7.9-final-0, default/linux/amd64/13.0, gcc-4.8.3, glibc-2.20-r2, 3.14.36-gentoo x86_64) ================================================================= System uname: Linux-3.14.36-gentoo-x86_64-Intel-R-_Core-TM-_i5-3550_CPU_@_3.30GHz-with-gentoo-2.2 KiB Mem: 8049380 total, 1832360 free KiB Swap: 1998844 total, 1971832 free Timestamp of tree: Sun, 29 Mar 2015 00:45:02 +0000 ld GNU ld (Gentoo 2.24 p1.4) 2.24 app-shells/bash: 4.2_p53 dev-java/java-config: 2.2.0 dev-lang/perl: 5.20.1-r4 dev-lang/python: 2.7.9-r1, 3.3.5-r1, 3.4.1 dev-util/cmake: 2.8.12.2-r1 dev-util/pkgconfig: 0.28-r1 sys-apps/baselayout: 2.2 sys-apps/openrc: 0.13.11 sys-apps/sandbox: 2.6-r1 sys-devel/autoconf: 2.69 sys-devel/automake: 1.11.6-r1, 1.12.6, 1.13.4 sys-devel/binutils: 2.24-r3 sys-devel/gcc: 4.8.3 sys-devel/gcc-config: 1.7.3 sys-devel/libtool: 2.4.6 sys-devel/make: 4.1-r1 sys-kernel/linux-headers: 3.18 (virtual/os-headers) sys-libs/glibc: 2.20-r2 Repositories: gentoo ACCEPT_KEYWORDS="amd64" ACCEPT_LICENSE="* -@EULA" CBUILD="x86_64-pc-linux-gnu" CFLAGS="-march=native -O2 -pipe" CHOST="x86_64-pc-linux-gnu" CONFIG_PROTECT="/etc /usr/share/config /usr/share/gnupg/qualified.txt" CONFIG_PROTECT_MASK="/etc/ca-certificates.conf /etc/dconf /etc/env.d /etc/fonts/fonts.conf /etc/gconf /etc/gentoo-release /etc/php/apache2-php5.5/ext-active/ /etc/php/cgi-php5.5/ext-active/ /etc/php/cli-php5.5/ext-active/ /etc/revdep-rebuild /etc/sandbox.d /etc/terminfo /etc/texmf/language.dat.d /etc/texmf/language.def.d /etc/texmf/updmap.d /etc/texmf/web2c" CXXFLAGS="-march=native -O2 -pipe" DISTDIR="/usr/portage/distfiles" EMERGE_DEFAULT_OPTS="--autounmask=n" FCFLAGS="-O2 -pipe" FEATURES="assume-digests binpkg-logs config-protect-if-modified distlocks ebuild-locks fixlafiles merge-sync news parallel-fetch preserve-libs protect-owned sandbox sfperms strict unknown-features-warn unmerge-logs unmerge-orphans userfetch userpriv usersandbox usersync" FFLAGS="-O2 -pipe" GENTOO_MIRRORS="http://mirror.yandex.ru/gentoo-distfiles/" LANG="ru_RU.UTF-8" LDFLAGS="-Wl,-O1 -Wl,--as-needed" MAKEOPTS="-j5" PKGDIR="/usr/portage/packages" PORTAGE_CONFIGROOT="/" PORTAGE_RSYNC_OPTS="--recursive --links --safe-links --perms --times --omit-dir-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="" USE="X acl amd64 avx berkdb bzip2 cli consolekit cracklib crypt cxx dbus dri fortran gdbm iconv ieee1394 ipv6 mmx mmxext modules multilib ncurses nls nptl openmp pam pcre readline session smp sse sse2 sse3 sse4_1 ssl ssse3 tcpd unicode vim-syntax zlib" ABI_X86="64" ALSA_CARDS="ali5451 als4000 atiixp atiixp-modem bt87x ca0106 cmipci emu10k1x ens1370 ens1371 es1938 es1968 fm801 hda-intel intel8x0 intel8x0m maestro3 trident usb-audio via82xx via82xx-modem ymfpci" 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 author" CAMERAS="ptp2" COLLECTD_PLUGINS="df interface irq load memory rrdtool swap syslog" CPU_FLAGS_X86="aes avx mmx mmxext popcnt sse sse2 sse3 sse4_1 sse4_2 ssse3" ELIBC="glibc" 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 ublox ubx" GRUB_PLATFORMS="pc" INPUT_DEVICES="keyboard mouse evdev" KERNEL="linux" LCD_DEVICES="bayrad cfontz cfontz633 glk hd44780 lb216 lcdm001 mtxorb ncurses text" LIBREOFFICE_EXTENSIONS="presenter-console presenter-minimizer" LINGUAS="en en_US ru ru_RU" OFFICE_IMPLEMENTATION="libreoffice" PHP_TARGETS="php5-5" PYTHON_SINGLE_TARGET="python2_7" PYTHON_TARGETS="python2_7 python3_3" QEMU_SOFTMMU_TARGETS="i386 x86_64" QEMU_USER_TARGETS="i386 x86_64" RUBY_TARGETS="ruby19 ruby20" USERLAND="GNU" VIDEO_CARDS="intel" 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, INSTALL_MASK, LC_ALL, PORTAGE_BUNZIP2_COMMAND, PORTAGE_COMPRESS, PORTAGE_COMPRESS_FLAGS, PORTAGE_RSYNC_EXTRA_OPTS, SYNC, USE_PYTHON
emerge --info please
(In reply to Davide Pesavento from comment #1) > emerge --info please I'm stupid. I meant attach the full build.log
Created attachment 400338 [details] build.log
I think i found the right way to correctly reproduce error. Sorry for above long explanations. My initial conditions: 1. Sync portage emerge --sync or emerge-webrsync -k In my case this is Portage 2.2.14 (python 2.7.9-final-0, default/linux/amd64/13.0, gcc-4.8.3, glibc-2.20-r2, 3.14.36-gentoo x86_64) ================================================================= System uname: Linux-3.14.36-gentoo-x86_64-Intel-R-_Core-TM-_i5-3230M_CPU_@_2.60GHz-with-gentoo-2.2 KiB Mem: 5672912 total, 4290148 free KiB Swap: 6291452 total, 6291452 free Timestamp of tree: Mon, 30 Mar 2015 00:45:01 +0000 2. Fully update system. I use the following emerge --update --newuse --deep --complete-graph --with-bdeps=y --pretend @world 3. Install dev-db/oracle-instantclient-basic with abi_x86_32 use-flag. Do the following: 1. USE="abi_x86_32 oci" ebuild /usr/portage/dev-qt/qtsql/qtsql-4.8.6-r1.ebuild configure See an error: OCI auto-detection... () x86_64-pc-linux-gnu-g++ -c -pipe -march=corei7 -O2 -pipe -O2 -Wall -W -I../../../../qt-everywhere-opensource-src-4.8.6/mkspecs/linux-g++ -I../../../../qt-everywhere-opensource-src-4.8.6/config.tests/unix/oci -I/usr/lib64/oracle/11.2.0.4/client/rdbms/public -I../../../../qt-everywhere-opensource-src-4.8.6/config.tests/unix/oci -I. -o oci.o ../../../../qt-everywhere-opensource-src-4.8.6/config.tests/unix/oci/oci.cpp x86_64-pc-linux-gnu-g++ -Wl,-O1 -Wl,--as-needed -Wl,-O1 -o oci oci.o -L/usr/lib64/oracle/11.2.0.4/client/lib32 -lclntsh /usr/lib/gcc/x86_64-pc-linux-gnu/4.8.3/../../../../x86_64-pc-linux-gnu/bin/ld: skipping incompatible /usr/lib64/oracle/11.2.0.4/client/lib32/libclntsh.so when searching for -lclntsh /usr/lib/gcc/x86_64-pc-linux-gnu/4.8.3/../../../../x86_64-pc-linux-gnu/bin/ld: cannot find -lclntsh collect2: error: ld returned 1 exit status Makefile:98: recipe for target 'oci' failed gmake: *** [oci] Error 1 OCI disabled. Oracle (OCI) support cannot be enabled due to functionality tests! Turn on verbose messaging (-v) to /var/tmp/portage/dev-qt/qtsql-4.8.6-r1/work/qt-everywhere-opensource-src-4.8.6/configure to see the final report. If you believe this message is in error you may use the continue switch (-continue) to /var/tmp/portage/dev-qt/qtsql-4.8.6-r1/work/qt-everywhere-opensource-src-4.8.6/configure to continue. * ERROR: dev-qt/qtsql-4.8.6-r1::gentoo failed (configure phase): * configure failed * * Call stack: * ebuild.sh, line 93: Called src_configure * environment, line 3178: Called multilib-minimal_src_configure * environment, line 2042: Called multilib_foreach_abi 'multilib-minimal_abi_src_configure' * environment, line 2249: Called multibuild_foreach_variant '_multilib_multibuild_wrapper' 'multilib-minimal_abi_src_configure' * environment, line 1973: Called _multibuild_run '_multilib_multibuild_wrapper' 'multilib-minimal_abi_src_configure' * environment, line 1971: Called _multilib_multibuild_wrapper 'multilib-minimal_abi_src_configure' * environment, line 386: Called multilib-minimal_abi_src_configure * environment, line 2036: Called multilib_src_configure * environment, line 2469: Called qt4_multilib_src_configure * environment, line 2955: Called die * The specific snippet of code: * "${S}"/configure "${conf[@]}" || die "configure failed"; * * If you need support, post the output of `emerge --info '=dev-qt/qtsql-4.8.6-r1::gentoo'`, * the complete build log and the output of `emerge -pqv '=dev-qt/qtsql-4.8.6-r1::gentoo'`. * The complete build log is located at '/var/tmp/portage/dev-qt/qtsql-4.8.6-r1/temp/build.log'. * The ebuild environment file is located at '/var/tmp/portage/dev-qt/qtsql-4.8.6-r1/temp/environment'. * Working directory: '/var/tmp/portage/dev-qt/qtsql-4.8.6-r1/work/qt-everywhere-opensource-src-4.8.6-abi_x86_32.x86' * S: '/var/tmp/portage/dev-qt/qtsql-4.8.6-r1/work/qt-everywhere-opensource-src-4.8.6' 2. Go to the working directory cd /var/tmp/portage/dev-qt/qtsql-4.8.6-r1/work/qt-everywhere-opensource-src-4.8.6-abi_x86_32.x86 3. Go to the configure tests subdirectory cd ./config.tests/unix/oci/ 4. Check oci.o file # LANG= objdump -f oci.o oci.o: file format elf64-x86-64 architecture: i386:x86-64, flags 0x00000011: HAS_RELOC, HAS_SYMS start address 0x0000000000000000 5. See 64-bit object code, that can't be linked with 32-bit oci library. ----------- More over! We can enter into other test's directory (for example sse2) cd ../sse2 and also see 64-bit code during configure stage for 32-bit mode # LANG= objdump -f sse2.o sse2.o: file format elf64-x86-64 architecture: i386:x86-64, flags 0x00000011: HAS_RELOC, HAS_SYMS start address 0x0000000000000000 But this test doesn't requery external libs, so its testing will be Ok and QT-lib continues to build. ----------- More over 2! I check these steps with another qt library (qtgui). Configuring is finished and OK!!!, but the same results inside working directory. 64-bit code in testing subdirectories during configure 32-bit mode. But i think this is trivial tests that doesn't require external libs, so nobody catch this. Build.log is attached. Thanks!
Yep, that's what I suspected. C(XX)FLAGS are not being passed to configure checks...
I think the reason in wrong specification that used to generate Makefile for tests. I attach one. You may see this line in oci/Makefile # Command: /var/tmp/portage/dev-qt/qtgui-4.8.6-r2/work/qt-everywhere-opensource-src-4.8.6-abi_x86_32.x86/bin/qmake -spec ../../../../qt-everywhere-opensource-src-4.8.6/mkspecs/linux-g++ It uses spec linux-g++ that does not contain flags for 32-bit compilation. But it SHOULD (or MUST?! :) ) use linux-g++-32.
Created attachment 400342 [details] Makefile from working directory of dev-qt/qtsql
Ups... In my Comment #6 i submitted line from qtgui, not qtsql. Sorry. That's right line # Command: /var/tmp/portage/dev-qt/qtsql-4.8.6-r1/work/qt-everywhere-opensource-src-4.8.6-abi_x86_32.x86/bin/qmake -spec ../../../../qt-everywhere-opensource-src-4.8.6/mkspecs/linux-g++
(In reply to Davide Pesavento from comment #5) > Yep, that's what I suspected. C(XX)FLAGS are not being passed to configure > checks... Actually, no. The issue is caused by the sedding of the various config.tests that we do in qt4-build-multilib_src_prepare()
I'm working on a fix
https://gitweb.gentoo.org/proj/qt.git/commit/?id=5e16bcd72951e73dbb89fdaa9b0dee793acdc4e8 The above commit should fix this issue. Please test. Since it's a pretty important behavior change and touches a delicate part of the eclass, any kind of testing (even if you're not affected by this particular bug) is _highly_ appreciated.
(In reply to Davide Pesavento from comment #11) > Since it's a pretty important behavior change and touches a delicate part of > the eclass, any kind of testing (even if you're not affected by this > particular bug) is _highly_ appreciated. I installed Qt4 from scratch (on a Qt5-only system) as well as konqueror-4.14.3. No problems here.
In cvs. 09 May 2015; Davide Pesavento <pesa@gentoo.org> qt4-build-multilib.eclass: Update from qt overlay: overhaul toolchain and *FLAGS handling for proper multilib support during the configure phase. Fixes bug #545106.
Tested on pure system (in VirtualBox) without any installed packages (except some base ones). There were no Qt and X. All builds with 32- and 64-bit ABIs were fine. Timestamp of repository gentoo: Mon, 11 May 2015 00:45:01 +0000
(In reply to George Tarasov from comment #14) > Tested on pure system (in VirtualBox) without any installed packages (except > some base ones). There were no Qt and X. All builds with 32- and 64-bit ABIs > were fine. > > Timestamp of repository gentoo: Mon, 11 May 2015 00:45:01 +0000 Thanks! This was with USE=oci8 enabled for qtsql, right?
(In reply to Davide Pesavento from comment #15) > > Thanks! This was with USE=oci8 enabled for qtsql, right? Yes, sure!
Great, thanks for testing.