The SIMD extensions are hard disabled for amd64. Reproducible: Always Steps to Reproduce: Portage 2.0.51-r15 (default-linux/amd64/2004.3, gcc-3.4.3, glibc-2.3.4.20041102-r0, 2.6.10-gentoo-r6 x86_64) ================================================================= System uname: 2.6.10-gentoo-r6 x86_64 Mobile AMD Athlon 64 Processor 2800+ Gentoo Base System version 1.5.3 Python: dev-lang/python-2.3.4-r1 [2.3.4 (#1, Feb 7 2005, 09:47:43)] ccache version 2.3 [enabled] dev-lang/python: 2.3.4-r1 sys-devel/autoconf: 2.59-r6, 2.13 sys-devel/automake: 1.9.4, 1.8.5-r3, 1.4_p6, 1.6.3, 1.5, 1.7.9-r1 sys-devel/binutils: 2.15.92.0.2-r1 sys-devel/libtool: 1.5.10-r4 virtual/os-headers: 2.6.8.1-r2 ACCEPT_KEYWORDS="amd64" AUTOCLEAN="yes" CFLAGS="-O2" CHOST="x86_64-pc-linux-gnu" CONFIG_PROTECT="/etc /usr/kde/2/share/config /usr/kde/3.3/env /usr/kde/3.3/share/config /usr/kde/3.3/shutdown /usr/kde/3/share/config /usr/lib/X11/xkb /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" DISTDIR="/usr/portage/distfiles" FEATURES="autoaddcvs autoconfig ccache distlocks sandbox" GENTOO_MIRRORS="http://distfiles.gentoo.org http://distro.ibiblio.org/pub/Linux/distributions/gentoo" LANG="de_DE" 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="amd64 X aac aalib accessibility acpi alsa audiofile bash-completion berkdb bitmap-fonts blender-game bluetooth bonobo bzlib c++ cairo cddb cdparanoia cdr cgi cjk cross crypt cups curl directfb divx4linux doc dri dvd dvdr dvdread edl emacs emacs-w3 encode esd ethereal ex f77 fam fbcon fftw flac flash flatfile font-server fortran ftp gd gdbm gif ginac glx gmp gnome gphoto2 gpm gstreamer gtk gtk2 gtkhtml imagemagick imap imlib insecure-drivers ipv6 java javascript jp2 jpeg kde lcms leim libwww lirc lzw lzw-tiff mad mbox mikmod mime mng motif mozilla mpi mule multilib mysql ncurses nls offensive oggvorbis opengl oss pam pcntl pcre perl php png pnp ppds python qt readline recode samba scanner sdk sdl sharedmem simplexml slang ssl svg sysvipc tcltk tcpd tetex theora tiff tokenizer truetype truetype-fonts type1-fonts unicode usb userlocales v4l v4l2 vhosts videos wmf xfs xine xinerama xml xml2 xmms xosd xpm xprint xrandr xv xvid xvmc yv12 zlib video_cards_radeon linguas_en linguas_de" Unset: ASFLAGS, CBUILD, CTARGET, LC_ALL, LDFLAGS
Created attachment 52364 [details, diff] xmms-1.2.10-r13.ebuild.diff Please apply attached patch to xmms-1.2.10-r13.ebuild This works....
Hi Roland, please have a look at [xmms-workdir]/Input/mpg123/Makefile.am: if ARCH_X86 if USE_X86ASM if USE_SIMD EXTRA_SRC = decode_i386.c dct64_i386.c decode_i586.s \ decode_3dnow.s dct64_3dnow.s dct36_3dnow.s getcpuflags.s \ dct64_MMX.s decode_MMX.s else EXTRA_SRC = decode_i386.c dct64_i386.c decode_i586.s endif You see, even if you enable SIMD in that ebuild, it doesn't get used. Now have a look at dct64_3dnow.s (fex). You will see plenty of x86 SIMD assembler, but not one line of x86-64 SIMD assembler. If you overwrite the tests, you'd get an assembler error. I resolve this one as INVALID because I'm interested in rewriting x86 asm to x86-64 asm. Assigning to AMD64 team...
s/I resolve/I don't resolve/ :-/ Should learn how to write...
Comment on attachment 52364 [details, diff] xmms-1.2.10-r13.ebuild.diff see comment 2
Danny, still interested in rewriting x86 asm to x86-64 asm? Otherwise I'd suggest we close this one
This is not an entirely trivial process. I did it for the image shading routines in Eterm by porting http://cvs.sourceforge.net/viewcvs.py/enlightenment/eterm/Eterm/src/mmx_cmod.S?rev=1.7&view=auto to http://cvs.sourceforge.net/viewcvs.py/enlightenment/eterm/Eterm/src/sse2_cmod.c?rev=1.1&view=auto If you are interested may I suggest using this piece of code: http://cvs.sourceforge.net/viewcvs.py/enlightenment/e17/libs/evas/src/lib/include/evas_mmx.h?rev=1.5&view=auto Written by H. Dietz and R. Fisher and now part of E-17. It appears that the MMX/SSE code is now generic enough to be used on both x86 and x86-64 processors. I would have used that code had I known it was there prior to my porting Eterm's code. Just a thought.
looks like kugelfang has other things to do ;) closing for now, everybody feel free to reopen with a patch
