| View previous topic :: View next topic |
| Author |
Message |
Thaidog
Veteran
Joined: 19 May 2004
Posts: 1053
|
 Posted: Fri Apr 09, 2010 3:04 am Post subject: |
 |
|
| costel78 wrote: | @Thaidog
That was my presumptions also, so I already compiled my world set using:
| Code: |
export GCC=${OCC}
export GCXX=${OCXX}
export OCC="icc"
export OCXX="icpc"
export CFLAGS=${ICCCFLAGS}
export CXXFLAGS=${ICCCXXFLAGS}
[ -r "${ROOT}"/etc/portage/package.gcc ] || return 0
while read -a target; do
if [ "${target}" = "${CATEGORY}/${PN}" ]; then
export OCC=${GCC}
export OCXX=${GCXX}
export CFLAGS="$flags"
export CXXFLAGS="$CFLAGS"
if [ -r ${ROOT}/etc/portage/package.gcc-cflags ]; then
while read target flags; do
if [ "${target}" = "${CATEGORY}/${PN}" ]; then
export CFLAGS="$flags"
export CXXFLAGS="$CFLAGS"
break
fi
done < "${ROOT}"/etc/portage/package.gcc-cflags
fi
break
fi
done < "${ROOT}"/etc/portage/package.gcc
if [ "${GCC}" = "icc" ]; then
# found="no"
# while read -a target; do
# if [ "${target}" = "${CATEGORY}/${PN}" ]; then
# found="yes"
# break;
# fi
# done < "${ROOT}"/etc/portage/package.icc-plain
grep "${CATEGORY}/${PN}" "${ROOT}"/etc/portage/package.icc-plain
if [ $? -eq 0 ]; then
found="no";
else
# if [ "${found}" = "no" ]; then
export AR=xiar
export LD=xild
export LDFLAGS="--as-needed,--warn-once,--sort-common,--hash-style=gnu"
fi
fi
if [ -r ${ROOT}/etc/portage/package.icc-cflags ]; then
while read target flags; do
if [ "${target}" = "${CATEGORY}/${PN}" ]; then
export CFLAGS="$flags"
export CXXFLAGS="$CFLAGS"
break
fi
done < ${ROOT}/etc/portage/package.icc-cflags
fi
if [ "${OCC}" != "" ]; then
export CC_FOR_BUILD="${OCC}" #workaround gcc detection function in toolchain-funcs.eclass
fi
unset GCC
unset GCXX
|
The only problems were:
| Code: |
cat /etc/portage/package.icc-plain
net-misc/openssh
sys-cluster/openmpi
www-apache/mod_security
|
|
How is your bashrc file working out? I'd like to add this to the Google Group files if that is ok?
_________________
Registered Linux User: 437619
"I'm a big believer in technology over politics" - Linus Torvalds |
|
| Back to top |
|
 |
costel78
Guru
Joined: 20 Apr 2007
Posts: 407
|
| Posted: Fri Apr 09, 2010 1:58 pm Post subject: |
|
|
@Thaidog
Sorry for long absence.
Everything except kernel is working great. About kernel, it seem that, in my configuration, I hit a bug, so I reverted it to gcc. The bug it's posted on LinuxDNA Google Group.
There are an error in script, -Wl was missing
export LDFLAGS="-Wl,--as-needed,--warn-once,--sort-common,--hash-style=gnu"
As I said previously replace ld with xild cause only 3 packages to fail. So, I consider it a success. I started at least every major application, start and use every service, without problems. I somebody may suggest a more professional way to test, please let me know.
I finished yesterday in the morning to compile with ipo flag every package witch didn't fail to compile or break related package. I still have to test few application for stability, at least starting it, test it if it crash whet used it's basic functions . Where was possible, I used the test portage FEATURE. But compiling with ipo and testing it's a long try and error process... so I need few more hours to test.
I'll post here package.icc-cflags, make conf and package.gcc list, bashrc when I finish testing. Again, if there are a better test method, please let me know.
Thank you!
_________________
Sorry for my English. I'm still learning this language. |
|
| Back to top |
|
|
broch
Tux's lil' helper
Joined: 09 Jul 2005
Posts: 94
|
| Posted: Fri Apr 09, 2010 3:55 pm Post subject: |
|
|
Hello,
I am trying out dna patch for 32-bit systems
Applied patch
dna-2.6.33-intel32-1.patch
on 2.6.33.2
source /opt/intel/Compiler/11.1/069/bin/iccvars.sh ia32
run make:
make AR=xiar LD=xild
I am getting the following error:
CC [M] drivers/net/wireless/iwlwifi/iwl-calib.o
CC [M] drivers/net/wireless/iwlwifi/iwl-scan.o
drivers/net/wireless/iwlwifi/iwl-scan.c(532) (col. 2): internal error: 0_0
compilation aborted for drivers/net/wireless/iwlwifi/iwl-scan.c (code 4)
make[4]: *** [drivers/net/wireless/iwlwifi/iwl-scan.o] Error 4
make[3]: *** [drivers/net/wireless/iwlwifi] Error 2
make[2]: *** [drivers/net/wireless] Error 2
make[1]: *** [drivers/net] Error 2
make: *** [drivers] Error 2
Is there a way to fix this?
thank you. |
|
| Back to top |
|
|
Thaidog
Veteran
Joined: 19 May 2004
Posts: 1053
|
| Posted: Fri Apr 09, 2010 5:42 pm Post subject: |
|
|
| broch wrote: | Hello,
I am trying out dna patch for 32-bit systems
Applied patch
dna-2.6.33-intel32-1.patch
on 2.6.33.2
source /opt/intel/Compiler/11.1/069/bin/iccvars.sh ia32
run make:
make AR=xiar LD=xild
I am getting the following error:
CC [M] drivers/net/wireless/iwlwifi/iwl-calib.o
CC [M] drivers/net/wireless/iwlwifi/iwl-scan.o
drivers/net/wireless/iwlwifi/iwl-scan.c(532) (col. 2): internal error: 0_0
compilation aborted for drivers/net/wireless/iwlwifi/iwl-scan.c (code 4)
make[4]: *** [drivers/net/wireless/iwlwifi/iwl-scan.o] Error 4
make[3]: *** [drivers/net/wireless/iwlwifi] Error 2
make[2]: *** [drivers/net/wireless] Error 2
make[1]: *** [drivers/net] Error 2
make: *** [drivers] Error 2
Is there a way to fix this?
thank you. |
Give Chi an email on this:
chi at linuxdna dot com
_________________
Registered Linux User: 437619
"I'm a big believer in technology over politics" - Linus Torvalds |
|
| Back to top |
|
|
Thaidog
Veteran
Joined: 19 May 2004
Posts: 1053
|
| Posted: Fri Apr 09, 2010 5:47 pm Post subject: |
|
|
| costel78 wrote: | @Thaidog
Sorry for long absence.
Everything except kernel is working great. About kernel, it seem that, in my configuration, I hit a bug, so I reverted it to gcc. The bug it's posted on LinuxDNA Google Group.
There are an error in script, -Wl was missing
export LDFLAGS="-Wl,--as-needed,--warn-once,--sort-common,--hash-style=gnu"
As I said previously replace ld with xild cause only 3 packages to fail. So, I consider it a success. I started at least every major application, start and use every service, without problems. I somebody may suggest a more professional way to test, please let me know.
I finished yesterday in the morning to compile with ipo flag every package witch didn't fail to compile or break related package. I still have to test few application for stability, at least starting it, test it if it crash whet used it's basic functions . Where was possible, I used the test portage FEATURE. But compiling with ipo and testing it's a long try and error process... so I need few more hours to test.
I'll post here package.icc-cflags, make conf and package.gcc list, bashrc when I finish testing. Again, if there are a better test method, please let me know.
Thank you! |
Can you give me the thread for the kernel bug at the google group? Probably using the system is the best way to test beside benchmarks and other synthetic tests. I'll be sure to add your script to the uploads!
_________________
Registered Linux User: 437619
"I'm a big believer in technology over politics" - Linus Torvalds |
|
| Back to top |
|
|
costel78
Guru
Joined: 20 Apr 2007
Posts: 407
|
| Posted: Sat Apr 10, 2010 3:53 pm Post subject: |
|
|
I finished (for now) my system compiled as much as possible with icc using ip and ipo flags.
I have to say that stability it's not easy to archive, because the following thing may happen: today a package may compile and work fine and tomorrow, at next update, it may compile fine but fail to run stable. That's the worst case scenario.
Comparative with a full gcc system with emerge in background (--buildpkgonly) which imply only a short period of time for maintenance , a system compiled with icc require much of your time.
Does it worth ? The answer depend by person. My answer is Yes. Not for system speed you get (may be impressive in some places), but for experience you got during the process. That's the Gentoo way 
I want to thanks everyone from this forum who helped me during process, especially Thaidog, darkbasic and linuxrouter (from LinuxDNA Google Group) and entire LinuxDNA team that makes a tremendous job.
make.conf:
| Code: | # These settings were set by the catalyst build script that automatically
# built this stage.
# Please consult /etc/make.conf.example for a more detailed example.
CHOST="x86_64-pc-linux-gnu"
CBUILD="x86_64-pc-linux-gnu"
ACCEPT_LICENSE="*"
PORTAGE_ECLASS_WARNING_ENABLE="0"
ACCEPT_KEYWORDS="~amd64 ~x86 amd64 x86"
AUTOCLEAN="yes"
LANG="ro_RO.UTF-8"
LC_ALL="ro_RO.UTF-8"
LINGUAS="ro en"
WANT_MP="true"
INPUT_DEVICES="evdev keyboard mouse"
VIDEO_CARDS="nv nvidia"
ALSA_CARDS="hda-intel"
LIRC_DEVICES="hauppauge"
COLLISION_IGNORE="/usr /lib /lib64 /bin /sbin /var"
SYNC="rsync://rsync.europe.gentoo.org"
GENTOO_MIRRORS="http://localhost http://mirrors.xservers.ro/gentoo/ http://ftp.roedu.net/pub/mirrors/gentoo.org/ http://mirrors.evolva.ro/gentoo/ http://distfiles.gentoo.org/ ${GENTOO_MIRRORS}"
FETCHCOMMAND="/usr/bin/wget --bind-address=192.168.2.5 --timeout=4 -t 2 --passive-ftp -O \"\${DISTDIR}/\${FILE}\" \"\${URI}\""
RESUMECOMMAND="/usr/bin/wget -c --bind-address=192.168.2.5 --timeout=4 -t 2 --passive-ftp -O \"\${DISTDIR}/\${FILE}\" \"\${URI}\""
PORTAGE_RSYNC_EXTRA_OPTS="--progress --delete-before --human-readable"
DISTDIR="/usr/portage/distfiles"
GRAPHITE="-floop-interchange -floop-strip-mine -floop-block"
CFLAGS="-O2 -pipe -w -march=native -mtune=native ${GRAPHITE}"
CXXFLAGS="${CFLAGS}"
LDFLAGS="-Wl,-O1,--sort-common,--warn-once,--hash-style=gnu"
MAKEOPTS="-j8 --load-average=11"
APACHE2_MODULES="authn_alias cern_meta charset_lite dumpio log_forensic proxy_ftp version actions alias auth_basic auth_digest authn_anon authn_dbd authn_dbm authn_default authn_file authz_dbm authz_default authz_groupfile authz_host authz_owner authz_user autoindex cache dav dav_fs dav_lock dbd deflate dir disk_cache env expires ext_filter file_cache filter headers ident imagemap include info log_config logio mem_cache mime mime_magic negotiation proxy proxy_ajp proxy_balancer proxy_connect proxy_http rewrite setenvif so speling status unique_id userdir usertrack vhost_alias asis substitute"
APACHE2_MPMS="worker"
W_USE="-R -akonadi -berkdb -bluetooth -branding -handbook -ipv6 -ldap -xattr"
N_USE="bindist caps curl custom-cflags eds ffmpeg fontconfig geoip gmp gnome gstreamer icu imagemagick jpeg jpeg2k kde lame lm_sensors matroska mpi mysql nls oss perl php png postgres quicktime samba slang smp startup-notification subversion symlink tcl threads tk v4l v4l2 vhosts"
L_USE="system-sqlite glitz clamdtop cleartype custom-cpuopts custom-optimization faac faad fuse git gmp inotify ithreads logrotate optimization rar replytolist resolvconf ssh tools vdpau"
USE="${W_USE} ${N_USE} ${L_USE}"
CLEAN_DELAY="3"
FEATURES="assume-digests candy distlocks fixpackages nodoc -parallel-fetch prelink strict -test unmerge-orphans -sandbox"
# SEARCH_DIRS_MASK=""
OVR="/usr/portage/local"
PORTDIR_OVERLAY="${OVR}/added ${OVR}/gnome ${OVR}/gentoo-php-rewrite"
RUBY_TARGETS="ruby18 jruby"
EMERGE_DEFAULT_OPTS="--jobs=4 --load-average=7.9 --keep-going --with-bdeps=y --complete-graph"
PORTAGE_COMPRESS="bzip2"
PORTAGE_COMPRESS_FLAGS="-9"
RSYNC_TIMEOUT=500
ICCCFLAGS="-O2 -xSSE4.2 -gcc -w -ip -ipo"
ICCCXXFLAGS="${ICCCFLAGS}"
|
Relevant files from /etc/portage:
bashrc - the new part (comparative to wiki) is replace of ld with xild. Packages that do not support xild are in package.icc-plain
| Code: | export GCC=${OCC}
export GCXX=${OCXX}
export OCC="icc"
export OCXX="icpc"
export CFLAGS=${ICCCFLAGS}
export CXXFLAGS=${ICCCXXFLAGS}
[ -r ${ROOT}/etc/portage/package.gcc ] || return 0
while read -a target; do
if [ "${target}" = "${CATEGORY}/${PN}" ]; then
export OCC=${GCC}
export OCXX=${GCXX}
export CFLAGS="$flags"
export CXXFLAGS="$CFLAGS"
if [ -r ${ROOT}/etc/portage/package.gcc-cflags ]; then
while read target flags; do
if [ "${target}" = "${CATEGORY}/${PN}" ]; then
export CFLAGS="$flags"
export CXXFLAGS="$CFLAGS"
break
fi
done < ${ROOT}/etc/portage/package.gcc-cflags
fi
break
fi
done < ${ROOT}/etc/portage/package.gcc
if [ "${GCC}" = "icc" ]; then
found="no"
while read -a target; do
if [ "${target}" = "${CATEGORY}/${PN}" ]; then
found="yes"
break;
fi
done < ${ROOT}/etc/portage/package.icc-plain
if [ "${found}" = "no" ]; then
export AR=xiar
export LD=xild
export LDFLAGS=" -Wl,-O,--sort-common,--warn-once,--hash-style=gnu"
fi
fi
if [ -r ${ROOT}/etc/portage/package.icc-cflags ]; then
while read target flags; do
if [ "${target}" = "${CATEGORY}/${PN}" ]; then
export CFLAGS="$flags"
export CXXFLAGS="$CFLAGS"
break
fi
done < ${ROOT}/etc/portage/package.icc-cflags
fi
if [ "${OCC}" != "" ]; then
export CC_FOR_BUILD="${OCC}" #workaround gcc detection function in toolchain-funcs.eclass
fi
unset GCC
unset GCXX
|
package.gcc - packages that don't compile with icc, or, by compile them with icc break stuff
| Code: | # Basic, important packages
app-shells/bash
sys-apps/sed
sys-devel/binutils
sys-devel/gcc
sys-libs/glibc
sys-libs/libstdc++-v3
sys-libs/ncurses
sys-libs/zlib
sys-process/procps
# It's compile, but the result during runtime is cathastophic
app-arch/xz-utils
dev-libs/openssl
sys-fs/ntfs3g
# Qt - it compile and works fine with icc, but kdelibs can't be compiled, even with gcc, so switched them to gcc
x11-libs/qt-core
x11-libs/qt-dbus
x11-libs/qt-gui
x11-libs/qt-opengl
x11-libs/qt-qt3support
x11-libs/qt-script
x11-libs/qt-sql
x11-libs/qt-svg
x11-libs/qt-test
x11-libs/qt-webkit
x11-libs/qt-xmlpatterns
# Various packages
app-antivirus/clamav
app-arch/gzip
app-doc/doxygen
app-emulation/virtualbox-modules
app-emulation/virtualbox-ose
app-misc/lirc
app-office/openoffice
app-text/aspell
dev-cpp/glibmm
dev-cpp/gtkmm
dev-db/cdb
dev-lang/perl
dev-lang/php
dev-libs/boost
dev-libs/check
dev-libs/elfutils
dev-libs/klibc
dev-libs/libatomic_ops
dev-libs/libffi
dev-libs/liboil
dev-libs/libpcre
dev-libs/nss
dev-libs/ppl
dev-libs/STLport
dev-util/kdevelop
dev-util/subversion
gnome-extra/wp_tray
kde-base/gwenview
kde-base/juk
kde-base/kcheckpass
kde-base/kdeartwork-kscreensaver
kde-base/kdebase-kioslaves
kde-base/kdelibs
kde-base/kdepimlibs
kde-base/kdm
kde-base/klinkstatus
kde-base/kopete
kde-base/krfb
kde-base/krosspython
kde-base/krunner
kde-base/ksysguard
kde-base/kwin
kde-base/libkcompactdisc
kde-base/libkdcraw
kde-base/okular
kde-misc/krusader
mail-client/mailx
media-gfx/graphviz
media-gfx/inkscape
media-libs/babl
media-libs/gd
media-libs/gegl
media-libs/id3lib
media-libs/libcanberra
media-libs/libmp4v2
media-libs/libv4l
media-libs/openjpeg
media-libs/taglib
media-libs/xine-lib
media-plugins/gst-plugins-ffmpeg
media-plugins/gst-plugins-taglib
media-sound/sox
media-video/avidemux
media-video/mkvtoolnix
net-analyzer/nmap
net-analyzer/ntop
net-firewall/iptables
net-fs/samba
net-ftp/lftp
net-libs/libnice
net-libs/xulrunner
net-misc/ntp
net-p2p/ktorrent
perl-core/version
sys-apps/busybox
sys-apps/coreutils
sys-apps/memtest86+
sys-apps/parted
sys-apps/preload
sys-apps/util-linux
sys-apps/v86d
sys-boot/lilo
sys-boot/syslinux
sys-devel/m4
sys-devel/prelink
sys-fs/btrfs-progs
sys-fs/mdadm
sys-libs/gpm
www-client/chromium
www-plugins/moonlight
x11-apps/x11perf
x11-apps/xauth
x11-apps/xhost
x11-apps/xinit
x11-base/xorg-server
x11-drivers/nvidia-drivers
x11-drivers/xf86-video-nv
x11-libs/cairo
x11-libs/libXvMC
x11-misc/makedepend
x11-wm/compiz
x11-wm/metacity
|
package.icc-cflags - accept flags per package. Most are here because they don't support or do not run well when compiling with ipo flag
| Code: |
app-admin/syslog-ng -O2 -xSSE4.2 -gcc -w -ip
app-arch/cabextract -O2 -xSSE4.2 -gcc -w -ip
app-arch/cpio -O2 -xSSE4.2 -gcc -w
app-arch/tar -O2 -xSSE4.2 -gcc -w -ip
app-cdr/brasero -O2 -xSSE4.2 -gcc -w -ip
app-crypt/pinentry -O2 -xSSE4.2 -gcc -w -ip
app-editors/gedit -O2 -xSSE4.2 -gcc -w -ip
app-misc/strigi -O2 -xSSE4.2 -gcc -w -ip
app-office/scribus -O2 -xSSE4.2 -gcc -w -ip
app-text/rarian -O2 -xSSE4.2 -gcc -w -ip
dev-db/postgresql-base -O2 -xSSE4.2 -gcc -w -ip
dev-db/sqlite -O3 -xSSE4.2 -gcc -w -ip -ipo
dev-lang/python -O3 -xSSE4.2 -gcc -w -ip -ipo
dev-lang/yasm -O2 -xSSE4.2 -gcc -w -ip
dev-libs/apr -O2 -xSSE4.2 -gcc -w -ip
dev-libs/apr-util -O2 -xSSE4.2 -gcc -w -ip
dev-libs/dbus-glib -O2 -xSSE4.2 -gcc -w -ip
dev-libs/fribidi -O2 -xSSE4.2 -gcc -w -ip
dev-libs/glib -O2 -xSSE4.2 -gcc -w -ip
dev-libs/libassuan -O2 -xSSE4.2 -gcc -w -ip
dev-libs/libgamin -O2 -xSSE4.2 -gcc -w -ip
dev-libs/libgcrypt -O2 -xSSE4.2 -gcc -w -ip
dev-libs/libgweather -O2 -xSSE4.2 -gcc -w -ip
dev-libs/libunique -O2 -xSSE4.2 -gcc -w -ip
dev-python/compizconfig-python -O2 -xSSE4.2 -gcc -w -ip
dev-python/dbus-python -O2 -xSSE4.2 -gcc -w -ip
dev-python/gconf-python -O2 -xSSE4.2 -gcc -w -ip
dev-python/gnome-applets-python -O2 -xSSE4.2 -gcc -w -ip
dev-python/gnome-vfs-python -O2 -xSSE4.2 -gcc -w -ip
dev-python/gst-python -O2 -xSSE4.2 -gcc -w -ip
dev-python/libbonobo-python -O2 -xSSE4.2 -gcc -w -ip
dev-python/libgnomecanvas-python -O2 -xSSE4.2 -gcc -w -ip
dev-python/libgnome-python -O2 -xSSE4.2 -gcc -w -ip
dev-python/numpy -O2 -xSSE4.2 -gcc -w -ip
dev-python/pygobject -O2 -xSSE4.2 -gcc -w -ip
dev-python/pygtk -O2 -xSSE4.2 -gcc -w -ip
dev-python/pygtksourceview -O2 -xSSE4.2 -gcc -w -ip
dev-python/pyorbit -O2 -xSSE4.2 -gcc -w -ip
dev-python/PyQt4 -O2 -xSSE4.2 -gcc -w -ip
dev-python/sexy-python -O2 -xSSE4.2 -gcc -w -ip
dev-util/cmake -O2 -xSSE4.2 -gcc -w -ip
dev-util/cvs -O2 -xSSE4.2 -gcc -w -ip
dev-util/gperf -O2 -xSSE4.2 -gcc -w -ip
dev-util/kbuild -O2 -xSSE4.2 -gcc -w -ip
games-emulation/dosbox -O2 -xSSE4.2 -gcc -w -ip
gnome-base/gnome-keyring -O2 -xSSE4.2 -gcc -w -ip
gnome-base/gnome-menus -O2 -xSSE4.2 -gcc -w -ip
gnome-base/gnome-session -O2 -xSSE4.2 -gcc -w -ip
gnome-base/gnome-vfs -O2 -xSSE4.2 -gcc -w -ip
gnome-base/gvfs -O2 -xSSE4.2 -gcc -w -ip
gnome-base/libglade -O2 -xSSE4.2 -gcc -w -ip
gnome-base/libgnome-keyring -O2 -xSSE4.2 -gcc -w -ip
gnome-base/libgnome -O2 -xSSE4.2 -gcc -w -ip
gnome-base/libgnomeui -O2 -xSSE4.2 -gcc -w -ip
gnome-base/orbit -O2 -xSSE4.2 -gcc -w -ip
gnome-extra/gnome-media -O2 -xSSE4.2 -gcc -w -ip
gnome-extra/gucharmap -O2 -xSSE4.2 -gcc -w -ip
gnome-extra/libgsf -O2 -xSSE4.2 -gcc -w -ip
kde-base/ksystemlog -O2 -xSSE4.2 -gcc -w -ip
kde-base/pykde4 -O2 -xSSE4.2 -gcc -w -ip
mail-mta/postfix -O2 -xSSE4.2 -gcc -w -ip
media-gfx/gimp -O2 -xSSE4.2 -gcc -w -ip
media-gfx/imagemagick -O2 -xSSE4.2 -gcc -w -ip
media-libs/aften -O2 -xSSE4.2 -gcc -w -ip
media-libs/alsa-lib -O2 -xSSE4.2 -gcc -w -ip
media-libs/gstreamer -O2 -xSSE4.2 -gcc -w -ip
media-libs/libdvdnav -O2 -xSSE4.2 -gcc -w -ip
media-libs/libdvdread -O2 -xSSE4.2 -gcc -w -ip
media-libs/libgphoto2 -O2 -xSSE4.2 -gcc -w -ip
media-libs/libtheora -O2 -xSSE4.2 -gcc -w -ip
media-libs/mesa -O2 -xSSE4.2 -gcc -w -ip
media-libs/schroedinger -O2 -xSSE4.2 -gcc -w -ip
media-libs/speex -O2 -xSSE4.2 -gcc -w -ip
media-libs/swfdec -O2 -xSSE4.2 -gcc -w -ip
media-libs/x264 -O2 -xSSE4.2 -gcc -w -ip
media-plugins/gst-plugins-a52dec -O2 -xSSE4.2 -gcc -w -ip
media-plugins/gst-plugins-alsa -O2 -xSSE4.2 -gcc -w -ip
media-plugins/gst-plugins-cdparanoia -O2 -xSSE4.2 -gcc -w -ip
media-plugins/gst-plugins-dvdread -O2 -xSSE4.2 -gcc -w -ip
media-plugins/gst-plugins-faac -O2 -xSSE4.2 -gcc -w -ip
media-plugins/gst-plugins-flac -O2 -xSSE4.2 -gcc -w -ip
media-plugins/gst-plugins-gconf -O2 -xSSE4.2 -gcc -w -ip
media-plugins/gst-plugins-gio -O2 -xSSE4.2 -gcc -w -ip
media-plugins/gst-plugins-gnomevfs -O2 -xSSE4.2 -gcc -w -ip
media-plugins/gst-plugins-lame -O2 -xSSE4.2 -gcc -w -ip
media-plugins/gst-plugins-mad -O2 -xSSE4.2 -gcc -w -ip
media-plugins/gst-plugins-mpeg2dec -O2 -xSSE4.2 -gcc -w -ip
media-plugins/gst-plugins-ogg -O2 -xSSE4.2 -gcc -w -ip
media-plugins/gst-plugins-oss -O2 -xSSE4.2 -gcc -w -ip
media-plugins/gst-plugins-pango -O2 -xSSE4.2 -gcc -w -ip
media-plugins/gst-plugins-resindvd -O2 -xSSE4.2 -gcc -w -ip
media-plugins/gst-plugins-theora -O2 -xSSE4.2 -gcc -w -ip
media-plugins/gst-plugins-v4l2 -O2 -xSSE4.2 -gcc -w -ip
media-plugins/gst-plugins-v4l -O2 -xSSE4.2 -gcc -w -ip
media-plugins/gst-plugins-vorbis -O2 -xSSE4.2 -gcc -w -ip
media-plugins/gst-plugins-x -O2 -xSSE4.2 -gcc -w -ip
media-plugins/gst-plugins-xvideo -O2 -xSSE4.2 -gcc -w -ip
media-plugins/live -O2 -xSSE4.2 -gcc -w -ip
media-sound/audacious -O2 -xSSE4.2 -gcc -w -ip
media-sound/audacity -O2 -xSSE4.2 -gcc -w -ip
media-sound/rosegarden -O2 -xSSE4.2 -gcc -w -ip
media-sound/wavpack -O2 -xSSE4.2 -gcc -w -ip
media-tv/xawtv -O2 -xSSE4.2 -gcc -w -ip
media-video/dirac -O2 -xSSE4.2 -gcc -w -ip
media-video/gnome-mplayer -O2 -xSSE4.2 -gcc -w -ip
media-video/mplayer -O2 -xSSE4.2 -gcc -w -ip
media-video/transcode -O2 -xSSE4.2 -gcc -w -ip
net-analyzer/ipfm -O2 -xSSE4.2 -gcc -w -ip
net-analyzer/iptraf -O2 -xSSE4.2 -gcc -w -ip
net-dns/bind-tools -O2 -xSSE4.2 -gcc -w -ip
net-firewall/ebtables -O2 -xSSE4.2 -gcc -w -ip
net-ftp/filezilla -O2 -xSSE4.2 -gcc -w -ip
net-ftp/gftp -O2 -xSSE4.2 -gcc -w -ip
net-ftp/pure-ftpd -O2 -xSSE4.2 -gcc -w -ip
net-libs/farsight2 -O2 -xSSE4.2 -gcc -w -ip
net-libs/libssh2 -O2 -xSSE4.2 -gcc -w -ip
net-mail/fetchmail -O2 -xSSE4.2 -gcc -w -ip
net-misc/aria2 -O2 -xSSE4.2 -gcc -w -ip
net-misc/dhcp -O2 -xSSE4.2 -gcc -w -ip
net-misc/ipv6calc -O2 -xSSE4.2 -gcc -w -ip
net-misc/openssh -O2 -xSSE4.2 -gcc -w -ip
net-misc/wget -O2 -xSSE4.2 -gcc -w -ip
net-p2p/transmission -O2 -xSSE4.2 -gcc -w -ip
net-print/cups -O2 -xSSE4.2 -gcc -w -ip
net-proxy/squid -O2 -xSSE4.2 -gcc -w -ip
sys-apps/dbus -O2 -xSSE4.2 -gcc -w -ip
sys-apps/devicekit-power -O2 -xSSE4.2 -gcc -w -ip
sys-apps/diffutils -O2 -xSSE4.2 -gcc -w -ip
sys-apps/grep -O2 -xSSE4.2 -gcc -w -ip
sys-apps/groff -O2 -xSSE4.2 -gcc -w -ip
sys-apps/module-init-tools -O2 -xSSE4.2 -gcc -w -ip
sys-apps/sandbox -O2 -xSSE4.2 -gcc -w -ip
sys-apps/sysvinit -O2 -xSSE4.2 -gcc -w -ip
sys-apps/texinfo -O2 -xSSE4.2 -gcc -w -ip
sys-auth/polkit -O2 -xSSE4.2 -gcc -w -ip
sys-cluster/openmpi -O2 -xSSE4.2 -gcc -w -ip
sys-devel/bc -O2 -xSSE4.2 -gcc -w -ip
sys-devel/bison -O2 -xSSE4.2 -gcc -w -ip
sys-devel/flex -O2 -xSSE4.2 -gcc -w -ip
sys-devel/gdb -O2 -xSSE4.2 -gcc -w
sys-devel/libtool -O2 -xSSE4.2 -gcc -w -ip
sys-fs/multipath-tools -O2 -xSSE4.2 -gcc -w -ip
sys-fs/reiserfsprogs -O2 -xSSE4.2 -gcc -w -ip
sys-fs/udev -O2 -xSSE4.2 -gcc -w -ip
sys-libs/db -O2 -xSSE4.2 -gcc -w -ip
sys-libs/readline -O2 -xSSE4.2 -gcc -w -ip
sys-libs/tdb -O2 -xSSE4.2 -gcc -w -ip
www-servers/apache -O2 -xSSE4.2 -gcc -w -ip
x11-libs/gtk+ -O2 -xSSE4.2 -gcc -w -ip
x11-libs/gtksourceview -O2 -xSSE4.2 -gcc -w -ip
x11-libs/libfontenc -O2 -xSSE4.2 -gcc -w -ip
x11-libs/libwnck -O2 -xSSE4.2 -gcc -w -ip
x11-libs/libX11 -O2 -xSSE4.2 -gcc -w -ip
x11-libs/libXaw -O2 -xSSE4.2 -gcc -w -ip
x11-libs/libXfontcache -O2 -xSSE4.2 -gcc -w -ip
x11-libs/libXfont -O2 -xSSE4.2 -gcc -w -ip
x11-libs/libXmu -O2 -xSSE4.2 -gcc -w -ip
x11-libs/libXpm -O2 -xSSE4.2 -gcc -w -ip
x11-libs/libXt -O2 -xSSE4.2 -gcc -w -ip
x11-libs/pango -O2 -xSSE4.2 -gcc -w -ip
x11-libs/pixman -O2 -xSSE4.2 -gcc -w -ip
x11-libs/vte -O2 -xSSE4.2 -gcc -w -ip
x11-proto/fontcacheproto -O2 -xSSE4.2 -gcc -w -ip
x11-proto/fontsproto -O2 -xSSE4.2 -gcc -w -ip
x11-proto/xproto -O2 -xSSE4.2 -gcc -w -ip
|
I almost forgot: package.icc-plain - does not support xild
| Code: | net-misc/openssh
sys-cluster/openmpi
www-apache/mod_security
|
_________________
Sorry for my English. I'm still learning this language. |
|
| Back to top |
|
|
Thaidog
Veteran
Joined: 19 May 2004
Posts: 1053
|
| Posted: Sun Apr 11, 2010 2:04 am Post subject: |
|
|
Wow great stuff costel! We're glad your enjoying using our patches and learning about icc! 
_________________
Registered Linux User: 437619
"I'm a big believer in technology over politics" - Linus Torvalds |
|
| Back to top |
|
|
linuxrouter
n00b
Joined: 03 Apr 2010
Posts: 3
|
| Posted: Tue Apr 13, 2010 2:51 am Post subject: |
|
|
| tranquilcool wrote: |
doesn't work for me. driver installs but crashes. |
My systems are running 64-bit currently and nvidia driver works okay with this setup. I hope to have some time soon so I can do a 32-bit install and see if I can reproduce the crash with the 32-bit nvidia driver. |
|
| Back to top |
|
|
darkbasic
Tux's lil' helper
Joined: 06 Sep 2006
Posts: 133
|
| Posted: Tue Apr 13, 2010 10:29 am Post subject: |
|
|
Wow, nice work 
_________________
Computers are like air conditioners:
they stop working properly when you open Windows...
Coltiva Linux, Windows si pianta da solo.
http://www.linuxsystems.it/ |
|
| Back to top |
|
|
broch
Tux's lil' helper
Joined: 09 Jul 2005
Posts: 94
|
| Posted: Tue Apr 13, 2010 2:03 pm Post subject: |
|
|
| Thaidog wrote: | | broch wrote: | Hello,
I am trying out dna patch for 32-bit systems
Applied patch
dna-2.6.33-intel32-1.patch
on 2.6.33.2
source /opt/intel/Compiler/11.1/069/bin/iccvars.sh ia32
run make:
make AR=xiar LD=xild
I am getting the following error:
CC [M] drivers/net/wireless/iwlwifi/iwl-calib.o
CC [M] drivers/net/wireless/iwlwifi/iwl-scan.o
drivers/net/wireless/iwlwifi/iwl-scan.c(532) (col. 2): internal error: 0_0
compilation aborted for drivers/net/wireless/iwlwifi/iwl-scan.c (code 4)
make[4]: *** [drivers/net/wireless/iwlwifi/iwl-scan.o] Error 4
make[3]: *** [drivers/net/wireless/iwlwifi] Error 2
make[2]: *** [drivers/net/wireless] Error 2
make[1]: *** [drivers/net] Error 2
make: *** [drivers] Error 2
Is there a way to fix this?
thank you. |
Give Chi an email on this:
chi at linuxdna dot com |
I emailed him, but I have no idea if he received it.
I tried to build iwl into kernel but the error is exactly the same:
| Quote: | CC drivers/net/wireless/iwlwifi/iwl-scan.o
drivers/net/wireless/iwlwifi/iwl-scan.c(532) (col. 2): internal error: 0_0
compilation aborted for drivers/net/wireless/iwlwifi/iwl-scan.c (code 4)
make[4]: *** [drivers/net/wireless/iwlwifi/iwl-scan.o] Error 4
make[3]: *** [drivers/net/wireless/iwlwifi] Error 2
make[2]: *** [drivers/net/wireless] Error 2
make[1]: *** [drivers/net] Error 2
make: *** [drivers] Error 2 |
without wireless kernel builds fine, I can also install nvidia driver 195.36.15 (32-bit) without any modifications. So for me aside of wireless icc kernel works.
With nvidia drivers installation for icc kernel I cheated:
installed for running kernel compiled with gcc, then still in gcc kernel I have run "offline" installation of nvidia driver for icc kernel, then booted icc kernel and all worked.
In other words: nvidia driver module was not compiled for icc kernel, it was only installed with "offline" installation command. It installed just fine and does not crash. |
|
| Back to top |
|
|
Thaidog
Veteran
Joined: 19 May 2004
Posts: 1053
|
| Posted: Mon Apr 19, 2010 12:41 am Post subject: |
|
|
We have a new 2.6.34 patch out:
wget http://www.linuxdna.com/dna-2.6.34-rc4-intel64-1.patch
In this patch linuxrouter added linking directly to the icc libs for increased performance 
_________________
Registered Linux User: 437619
"I'm a big believer in technology over politics" - Linus Torvalds |
|
| Back to top |
|
|
costel78
Guru
Joined: 20 Apr 2007
Posts: 407
|
| Posted: Tue Apr 20, 2010 5:53 am Post subject: |
|
|
So AR=xiar LD=xild are not more required for kernel and we have a patch for 2.6.34-rc.
That's great news! I'll test the new patch when I'll return home.
Thank you!
_________________
Sorry for my English. I'm still learning this language. |
|
| Back to top |
|
|
broch
Tux's lil' helper
Joined: 09 Jul 2005
Posts: 94
|
| Posted: Tue Apr 20, 2010 1:01 pm Post subject: |
|
|
Any chance to see also a nw 32-bit patch?
Thanks |
|
| Back to top |
|
|
mattst88
Developer
Joined: 28 Oct 2004
Posts: 422
|
| Posted: Tue Apr 20, 2010 11:38 pm Post subject: |
|
|
Ostensibly, the purpose of these patches is for performance, but I only see two tests where any significant change in performance in achieved.
http://linuxdna.com/benchmarks.pdf (By the way, use a Monospace Font, that document is awful)
| Code: | Context switching - times in microseconds - smaller is better
-------------------------------------------------------------------------
Host OS 2p/0K 2p/16K 2p/64K 8p/16K 8p/64K 16p/16K 16p/64K
ctxsw ctxsw ctxsw ctxsw ctxsw ctxsw ctxsw
--------- ------------- ------ ------ ------ ------ ------ ------- -------
atom-gcc Linux 2.6.33- 15.0 16.6 13.5 17.9 24.2 22.9 27.0
atom-icc Linux 2.6.33- 4.0300 7.2000 4.3400 9.0700 14.2 12.3 17.8
*Local* Communication latencies in microseconds - smaller is better
---------------------------------------------------------------------
Host OS 2p/0K Pipe AF UDP RPC/ TCP RPC/ TCP
ctxsw UNIX UDP TCP conn
--------- ------------- ----- ----- ---- ----- ----- ----- ----- ----
atom-gcc Linux 2.6.33- 15.0 34.6 34.2 71.1 90.9 158.
atom-icc Linux 2.6.33- 4.030 11.5 24.1 49.5 73.2 216. |
Wouldn't a more worthwhile goal be to figure out what ICC does differently from gcc in these two cases? I imagine that whatever it is, it would also be useful for AMD CPUs as well.
_________________
My Wiki page |
|
| Back to top |
|
|
Thaidog
Veteran
Joined: 19 May 2004
Posts: 1053
|
| Posted: Wed Apr 21, 2010 5:42 am Post subject: |
|
|
| broch wrote: |
Any chance to see also a nw 32-bit patch?
Thanks |
EDIT:
http://www.chihoang.de/uploads/media/dna-2.6.34-rc4-intel32-1.patch
Chi may have something working on 32 bit:
chi at linuxdna dot com
_________________
Registered Linux User: 437619
"I'm a big believer in technology over politics" - Linus Torvalds
Last edited by Thaidog on Wed Apr 21, 2010 5:59 am; edited 1 time in total |
|
| Back to top |
|
|
Thaidog
Veteran
Joined: 19 May 2004
Posts: 1053
|
| Posted: Wed Apr 21, 2010 5:51 am Post subject: |
|
|
| mattst88 wrote: | Ostensibly, the purpose of these patches is for performance, but I only see two tests where any significant change in performance in achieved.
http://linuxdna.com/benchmarks.pdf (By the way, use a Monospace Font, that document is awful)
| Code: | Context switching - times in microseconds - smaller is better
-------------------------------------------------------------------------
Host OS 2p/0K 2p/16K 2p/64K 8p/16K 8p/64K 16p/16K 16p/64K
ctxsw ctxsw ctxsw ctxsw ctxsw ctxsw ctxsw
--------- ------------- ------ ------ ------ ------ ------ ------- -------
atom-gcc Linux 2.6.33- 15.0 16.6 13.5 17.9 24.2 22.9 27.0
atom-icc Linux 2.6.33- 4.0300 7.2000 4.3400 9.0700 14.2 12.3 17.8
*Local* Communication latencies in microseconds - smaller is better
---------------------------------------------------------------------
Host OS 2p/0K Pipe AF UDP RPC/ TCP RPC/ TCP
ctxsw UNIX UDP TCP conn
--------- ------------- ----- ----- ---- ----- ----- ----- ----- ----
atom-gcc Linux 2.6.33- 15.0 34.6 34.2 71.1 90.9 158.
atom-icc Linux 2.6.33- 4.030 11.5 24.1 49.5 73.2 216. |
Wouldn't a more worthwhile goal be to figure out what ICC does differently from gcc in these two cases? I imagine that whatever it is, it would also be useful for AMD CPUs as well. |
The ultimate goal of the kernel patches is to get support working for both IPO and PGO cflags... in which case you would have something that GCC can NOT do at all. The patches we put out now are faster but as you notice not completely faster or in some cases not noticeably faster. IPO and PGO have been employed in earlier patches (2.6.4 -.9) but support has not been continued due to the fact we are a small and unpaid opensource project. My main goal is to bring attention to the project in any way I can to get more and more support so we can get these things going. Meanwhile the baseline patches we do make are for those who need every advantage... be it for underpowered atom cpus or big supercommputers. Once PGO and IPO are available we will have the fastest patches in any senerio available for linux... mark my words! Mean while patches freely exsist for AMD cpus to benifit just as much as intel cpus from our project. But I don't go in to detail since AMD has not contributed anything to this project while Intel has contributed much.
(FYI I've been on vacation with no computer for the last week... sorry for the delay in response)
_________________
Registered Linux User: 437619
"I'm a big believer in technology over politics" - Linus Torvalds |
|
| Back to top |
|
|
broch
Tux's lil' helper
Joined: 09 Jul 2005
Posts: 94
|
| Posted: Wed Apr 21, 2010 1:22 pm Post subject: |
|
|
Thank you very much Thaidog for 32-bit patch. Unfortunately, this patch will not apply on 2.6.34-rc5. I did dry-run patching:
| Quote: | [/usr/src/linux-2.6.34-rc5]$ patch -p1 --dry-run < ~/dna-2.6.34-rc4-intel32-1.patch
can't find file to patch at input line 3
Perhaps you used the wrong -p or --strip option?
The text leading up to this was:
--------------------------
|--- /usr/src/latest/Makefile 2010-04-02 23:36:12.000000000 +0200
|+++ /usr/src/linux/Makefile 2010-04-18 15:01:59.000000000 +0200
--------------------------
File to patch:
|
after correction, another, similar error pops up
unless i am doing something wrong?
Thanks |
|
| Back to top |
|
|
mothersh1p
n00b
Joined: 21 Apr 2010
Posts: 15
|
| Posted: Wed Apr 21, 2010 10:31 pm Post subject: |
|
|
| broch wrote: | Thank you very much Thaidog for 32-bit patch. Unfortunately, this patch will not apply on 2.6.34-rc5. I did dry-run patching:
| Quote: | [/usr/src/linux-2.6.34-rc5]$ patch -p1 --dry-run < ~/dna-2.6.34-rc4-intel32-1.patch
can't find file to patch at input line 3
Perhaps you used the wrong -p or --strip option?
The text leading up to this was:
--------------------------
|--- /usr/src/latest/Makefile 2010-04-02 23:36:12.000000000 +0200
|+++ /usr/src/linux/Makefile 2010-04-18 15:01:59.000000000 +0200
--------------------------
File to patch:
|
after correction, another, similar error pops up
unless i am doing something wrong?
Thanks |
Hi! I'm Chi. I'm sorry, I didn't receive your e-mail. I'm sorry about your problem with the patch, too. The default behaviour of the -p flag doesn't take count of a leading slash /usr/src/linux-2.6.34-rc5. The right comand would be of course:
patch -p4 -t --dry-run < ~/dna-2.6.34-rc4-intel32-1.patch.
This will strip the smallest prefix containing num leading slashes from each file name found in the patch file.
patch -p1 ... would give you usr/src/latest/Makefile, usr/src/latest/arch/x86/boot/compressed/Makefile etc.
Congratulation to Tyler for his new occupation! |
|
| Back to top |
|
|
Thaidog
Veteran
Joined: 19 May 2004
Posts: 1053
|
| Posted: Thu Apr 22, 2010 2:32 am Post subject: |
|
|
Thanks Chi
_________________
Registered Linux User: 437619
"I'm a big believer in technology over politics" - Linus Torvalds |
|
| Back to top |
|
|
broch
Tux's lil' helper
Joined: 09 Jul 2005
Posts: 94
|
| Posted: Thu Apr 22, 2010 3:44 am Post subject: |
|
|
Thanks guys!
All is working now |
|
| Back to top |
|
|
tranquilcool
Veteran
Joined: 25 Mar 2005
Posts: 1246
|
| Posted: Fri Apr 23, 2010 10:43 pm Post subject: |
|
|
thanks. works fine for me with nouveau. x32
nvidia is no show. last time i tried nvidia it crashed.
_________________
this is a strange strange world. |
|
| Back to top |
|
|
Thaidog
Veteran
Joined: 19 May 2004
Posts: 1053
|
| Posted: Wed May 05, 2010 2:55 am Post subject: |
|
|
More news - Chi has a new 32bit patch out:
wget http://www.linuxdna.com/dna-2.6.34-rc5-intel32-4.patch
Also Luyi found that using -ipo1 instead of -ipo to compile packages like MySQL and PHP works! And we should use -no-gcc in most of the packages. (A lot of strange problems like segment fault disappeared. But we need to change -KPIC to -fPIC in the configure script.)
_________________
Registered Linux User: 437619
"I'm a big believer in technology over politics" - Linus Torvalds |
|
| Back to top |
|
|
mothersh1p
n00b
Joined: 21 Apr 2010
Posts: 15
|
| Posted: Wed May 05, 2010 8:31 am Post subject: |
|
|
Greetings! My name is Chi Hoang. I'm an experienced webdeveloper.
I'm using an Amilo Xa 3530 Laptop. With a Turion 64 Ultra X2 2,2 Ghz CPU and 4 Gb Ram
with a hybrid graphic-card-system based on the Radeon 2400 and 3200. My OS
is OpenSuse 11.2. Thanks to the switcheroo-patch both graphic-cards
are supported by Linux and running! My webcam is running, too.
Touchpad is running even better than in Windows (Scrolling supported).
I use Linux mainly with php and mysql. Sometimes I do a little bash-
script. And I like to compile! I like do Typo3, too. I want 65€/hour. I'm looking forward to your post! |
|
| Back to top |
|
|
broch
Tux's lil' helper
Joined: 09 Jul 2005
Posts: 94
|
| Posted: Wed May 05, 2010 1:02 pm Post subject: |
|
|
| Thaidog wrote: | More news - Chi has a new 32bit patch out:
wget http://www.linuxdna.com/dna-2.6.34-rc5-intel32-4.patch
Also Luyi found that using -ipo1 instead of -ipo to compile packages like MySQL and PHP works! And we should use -no-gcc in most of the packages. (A lot of strange problems like segment fault disappeared. But we need to change -KPIC to -fPIC in the configure script.) |
great,
thank you both
note:
dna patch applied on 2.6.34-rc6-git4
booted successfully, works with nvidia driver 195.36.24
stable for at least 16min  |
|
| Back to top |
|
|
tranquilcool
Veteran
Joined: 25 Mar 2005
Posts: 1246
|
| Posted: Wed May 05, 2010 11:28 pm Post subject: |
|
|
| broch wrote: | | Thaidog wrote: | More news - Chi has a new 32bit patch out:
wget http://www.linuxdna.com/dna-2.6.34-rc5-intel32-4.patch
Also Luyi found that using -ipo1 instead of -ipo to compile packages like MySQL and PHP works! And we should use -no-gcc in most of the packages. (A lot of strange problems like segment fault disappeared. But we need to change -KPIC to -fPIC in the configure script.) |
great,
thank you both
note:
dna patch applied on 2.6.34-rc6-git4
booted successfully, works with nvidia driver 195.36.24
stable for at least 16min |
how did you get nvidia-drivers to work.
can't make it work here.
_________________
this is a strange strange world. |
|
| Back to top |
|
|
|
Unsupported Software
