I missed to compile gcc with support for fortran which blas-19980702 apparently needs. The thing is, I tried to install it and the build command failed. Even though, portage installed the sandbox, and the package was installed, even though no files was compiled nor installed. Reproducible: Always Steps to Reproduce: 1. Make sure to have gcc compiled _without_ support for fortran. 2. emerge -vp blas 3. emerge blas Actual Results: >>> md5 src_uri ;-) blas.tgz >>> Unpacking source... >>> Unpacking blas.tgz to /var/tmp/portage/blas-19980702/work >>> Source unpacked. g77 -O2 -march=pentium4 -pipe -fomit-frame-pointer -c -o caxpy.o caxpy.f make: g77: Command not found make: *** [caxpy.o] Error 127 >>> Install blas-19980702 into /var/tmp/portage/blas-19980702/image/ category app-sci /usr/lib/portage/bin/dolib.a: libblas.a does not exist man: prepallstrip: strip: >>> Completed installing into /var/tmp/portage/blas-19980702/image/ >>> Merging app-sci/blas-19980702 to / --- /usr/ --- /usr/lib/ * Caching service dependencies... >>> app-sci/blas-19980702 merged. Expected Results: The ebuild should warn or even die if gcc was compiled without support for fortan. Portage 2.0.50-r6 (default-x86-2004.0, gcc-3.3.3, glibc-2.3.3_pre20040420-r0, 2.6.5-gentoo-r1) ================================================================= System uname: 2.6.5-gentoo-r1 i686 Intel(R) Pentium(R) 4 CPU 3.00GHz Gentoo Base System version 1.4.10 ccache version 2.3 [enabled] Autoconf: sys-devel/autoconf-2.59-r3 Automake: sys-devel/automake-1.8.3 ACCEPT_KEYWORDS="x86 ~x86" AUTOCLEAN="yes" CFLAGS="-O2 -march=pentium4 -pipe -fomit-frame-pointer" CHOST="i686-pc-linux-gnu" COMPILER="gcc3" CONFIG_PROTECT="/etc /usr/X11R6/lib/X11/xkb /usr/kde/2/share/config /usr/kde/3.2/share/config /usr/kde/3/share/config /usr/lib/mozilla/defaults/pref /usr/share/config /usr/share/texmf/dvipdfm/config/ /usr/share/texmf/dvips/config/ /usr/share/texmf/tex/generic/config/ /usr/share/texmf/tex/platex/config/ /usr/share/texmf/xdvi/ /var/qmail/control" CONFIG_PROTECT_MASK="/etc/gconf /etc/terminfo /etc/env.d" CXXFLAGS="-O2 -march=pentium4 -pipe -fomit-frame-pointer" DISTDIR="/usr/portage/distfiles" FEATURES="autoaddcvs ccache sandbox" GENTOO_MIRRORS="http://ftp.du.se/pub/os/gentoo ftp://mirrors.sec.informatik.tu-darmstadt.de/gentoo/ http://ds.thn.htu.se/linux/gentoo http://mirror.pudas.net/gentoo" MAKEOPTS="-j3" PKGDIR="/usr/portage/packages" PORTAGE_TMPDIR="/var/tmp" PORTDIR="/usr/portage" PORTDIR_OVERLAY="/usr/local/portage" SYNC="rsync://rsync.gentoo.org/gentoo-portage" USE="X apm arts avi berkdb cdr crypt cups doc dvdr emacs encode esd foomaticdb gdbm gif gnome gpm gtk gtk2 imlib jpeg libg++ libwww mad mikmod motif mozilla mpeg mysql nas ncurses nls oggvorbis opengl oss pam pdflib perl png ppds python qt quicktime readline ruby samba sdl slang spell ssl svga tcltk tcpd tetex truetype x86 xml xml2 xmms xv zlib"
Forgot to add that the script does not die either if the make process fails or if the files it tries to install does not exist.
Hi Emil. Please take a look at app-sci/blas-* packages (blas-reference and blas-atlas). These are going to be the future of blas (and the same for lapack-*) in portage. They introduce an ability to mix and match different variants of blas via blas-config. Plus they should have this issue fixed already (but please check!). I am actually about to unmask blas-* packages and do a bit of explanation on the lists (the change is sizeable and will effect other devs), so if you could post your test results that would be awesome! See #30453 for more details (and please report to that bug as well). Closing this bug as a dup correspondingly.. George *** This bug has been marked as a duplicate of 30453 ***
Ok, I shall see what I can do. I re-compiled GCC with support for fortran, so should I recompile again without it, and try to install the other blas-packages instead?
Hm, that's quite involved, but I am not sure if there is an easier way to do the testing. May be try renaming g77 into something, then try emerging the package and then rename that back into g77? IIRC the blas build script tests for g77 presence, but I am not totally sure atm. I'll be revisiting blas_* packages soon, before unmasking them, but it would be nice to have many test reports in order to be sure that we do not screw things up. George.
