View previous topic :: View next topic |
Author |
Message |
Jojobinha_2009 Tux's lil' helper
Joined: 27 Mar 2021 Posts: 77 Location: Brazil
|
Posted: Mon Apr 12, 2021 12:47 am Post subject: Setting -march to native = slow merge times. |
|
|
This is kind of related to my previous LibreOffice post.
I have noticed that if I use in my make.conf, I get overall longer merge times, in fact much longer for certain packages.
My machine stuck compiling PyQt5 for almost 5 hours using that parameter.
However, once I do what the wiki says and put I get much faster merge times, In fact I think this was one of the factors that made my system able to build LibreOffice in a whopping 48 minutes and Firefox in 33 minutes.
What does the "Native" mean here and why does it seem to influence compile times so much? _________________ Intel Core i5-9400F / 24GB DDR4 2666MHz / GeForce GTX 1060 3GB
Powered by Gentoo for x86_64
======================================================
Seize the day, and remember to have fun! |
|
Back to top |
|
|
mustafasalih1993 n00b
Joined: 09 Feb 2021 Posts: 38
|
Posted: Mon Apr 12, 2021 1:52 am Post subject: |
|
|
Hello Jojobinha_2009
the gcc man page tells the following about -march=native:
"-march=native causes the compiler to auto-detect the architecture of the build computer. At present, this feature is only supported on GNU/Linux, and not all architectures are recognized. If the auto-detect is unsuccessful the option has no effect."
it's basically tells the compiler to auto-detect your hardware this is useful if you don't know your hardware |
|
Back to top |
|
|
dmpogo Advocate
Joined: 02 Sep 2004 Posts: 3267 Location: Canada
|
Posted: Mon Apr 12, 2021 2:51 am Post subject: |
|
|
mustafasalih1993 wrote: | Hello Jojobinha_2009
the gcc man page tells the following about -march=native:
"-march=native causes the compiler to auto-detect the architecture of the build computer. At present, this feature is only supported on GNU/Linux, and not all architectures are recognized. If the auto-detect is unsuccessful the option has no effect."
it's basically tells the compiler to auto-detect your hardware this is useful if you don't know your hardware |
I wonder how I can check the result of autodetection ? |
|
Back to top |
|
|
mustafasalih1993 n00b
Joined: 09 Feb 2021 Posts: 38
|
Posted: Mon Apr 12, 2021 3:05 am Post subject: |
|
|
dmpogo wrote: |
I wonder how I can check the result of autodetection ?
|
Hello dmpogo
you can the see the result with:
Code: |
gcc -march=native -E -v - </dev/null 2>&1 | grep cc1 | grep march
|
|
|
Back to top |
|
|
Ionen Developer
Joined: 06 Dec 2018 Posts: 2719
|
Posted: Mon Apr 12, 2021 4:07 am Post subject: |
|
|
Reminder that setting =specific has a different meaning than =native
Like my cheap minipc haswell is missing some features, app-misc/resolve-march-native gives me: Code: | -march=haswell -mabm -mno-avx -mno-avx2 -mno-bmi -mno-bmi2 -mno-f16c -mno-fma -mno-xsave -mno-xsaveopt --param=l1-cache-line-size=64 --param=l1-cache-size=32 --param=l2-cache-size=2048 | So -march=haswell wouldn't be the same as -march=native, I don't even have avx.
On my skylake, =native additionally enables two flags: Code: | -march=skylake -mabm -mrtm --param=l1-cache-line-size=64 --param=l1-cache-size=32 --param=l2-cache-size=12288 |
GCC might also change how this is handled in new versions, add support for more things, etc.. and =native stays up to date (not that it can be helped if cross-compiling / distcc).
Edit: For me native vs exact flags doesn't seem to make any (noticeable) differences in compile times, likely some isolated hardware-specific issue if it does. |
|
Back to top |
|
|
kukibl Apprentice
Joined: 10 Jun 2008 Posts: 237
|
Posted: Mon Apr 12, 2021 6:07 am Post subject: |
|
|
@Jojobinha_2009
Is this subjective thing or you have some exact merge time differences (maybe to query it with genlop)? |
|
Back to top |
|
|
Jojobinha_2009 Tux's lil' helper
Joined: 27 Mar 2021 Posts: 77 Location: Brazil
|
Posted: Mon Apr 12, 2021 10:33 pm Post subject: |
|
|
kukibl wrote: | @Jojobinha_2009
Is this subjective thing or you have some exact merge time differences (maybe to query it with genlop)? |
Its subjective, however I can feel a noticeable slow down in large merges if I use native.
As I said, last time I used this, my machine got stuck merging PyQt5 for over 4 hours. I could hear the CPU cooler chugging along at high speed... And no end on sight.
Whereas once I used march=skylake, I didn't even notice that PyQt5 was compiled... It just flashed through. _________________ Intel Core i5-9400F / 24GB DDR4 2666MHz / GeForce GTX 1060 3GB
Powered by Gentoo for x86_64
======================================================
Seize the day, and remember to have fun! |
|
Back to top |
|
|
mike155 Advocate
Joined: 17 Sep 2010 Posts: 4438 Location: Frankfurt, Germany
|
Posted: Tue Apr 13, 2021 12:40 am Post subject: |
|
|
@Jojobinha_2009: please measure and post emerge compile times for different '-march' variants on your machine.
Below are the results on my Ivybridge machine for 'time emerge --oneshot --nodeps PyQt5':
Code: | CFLAGS="-O2 -pipe" : 612 secs
CFLAGS="-O2 -pipe -march=native" : 609 secs
CFLAGS="-O2 -pipe -march=skylake" : 605 secs
CFLAGS="-O2 -pipe -march=ivybridge" : 613 secs
CFLAGS="-O2 -pipe -march=znver2" : 608 secs |
I don't see any difference. |
|
Back to top |
|
|
NeddySeagoon Administrator
Joined: 05 Jul 2003 Posts: 54236 Location: 56N 3W
|
Posted: Tue Apr 13, 2021 8:38 am Post subject: |
|
|
Jojobinha_2009,
Are you sure its the -march setting?
Show us before and after the changes you make.
I would expect some CFLAGS to change the build time but -march is not one of them. _________________ Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail. |
|
Back to top |
|
|
Jojobinha_2009 Tux's lil' helper
Joined: 27 Mar 2021 Posts: 77 Location: Brazil
|
Posted: Tue Apr 13, 2021 11:26 pm Post subject: |
|
|
NeddySeagoon wrote: | Jojobinha_2009,
Are you sure its the -march setting?
Show us before and after the changes you make.
I would expect some CFLAGS to change the build time but -march is not one of them. |
Sorry I can't show the output of the command before I make the change (because I've already rebuilt my entire install using -march=skylake).
But here's the result after:
Code: | emerge --info
Portage 3.0.18 (python 3.8.9-final-0, default/linux/amd64/17.1/desktop/plasma/systemd, gcc-10.3.0, glibc-2.33, 5.11.13-gentoo-x86_64 x86_64)
=================================================================
System uname: Linux-5.11.13-gentoo-x86_64-x86_64-Intel-R-_Core-TM-_i5-9400F_CPU_@_2.90GHz-with-glibc2.2.5
KiB Mem: 24443880 total, 20767204 free
KiB Swap: 25165820 total, 25165820 free
Timestamp of repository gentoo: Tue, 13 Apr 2021 21:30:01 +0000
Head commit of repository gentoo: 30dc5d18c70ec8859ea68dff03c14e3324c4800e
sh bash 5.1_p4
ld GNU ld (Gentoo 2.35.2 p1) 2.35.2
app-shells/bash: 5.1_p4::gentoo
dev-lang/perl: 5.32.1::gentoo
dev-lang/python: 2.7.18_p8::gentoo, 3.8.9::gentoo, 3.9.4::gentoo
dev-lang/rust: 1.51.0::gentoo
dev-util/cmake: 3.20.1::gentoo
sys-apps/baselayout: 2.7-r1::gentoo
sys-apps/sandbox: 2.23::gentoo
sys-devel/autoconf: 2.13-r1::gentoo, 2.69-r5::gentoo
sys-devel/automake: 1.16.3-r1::gentoo
sys-devel/binutils: 2.35.2::gentoo
sys-devel/gcc: 10.3.0::gentoo
sys-devel/gcc-config: 2.4::gentoo
sys-devel/libtool: 2.4.6-r6::gentoo
sys-devel/make: 4.3::gentoo
sys-kernel/linux-headers: 5.11::gentoo (virtual/os-headers)
sys-libs/glibc: 2.33::gentoo
Repositories:
gentoo
location: /var/db/repos/gentoo
sync-type: rsync
sync-uri: rsync://rsync.gentoo.org/gentoo-portage
priority: -1000
sync-rsync-verify-max-age: 24
sync-rsync-extra-opts:
sync-rsync-verify-metamanifest: yes
sync-rsync-verify-jobs: 1
ACCEPT_KEYWORDS="amd64 ~amd64"
ACCEPT_LICENSE="*"
CBUILD="x86_64-pc-linux-gnu"
CFLAGS="-march=skylake -O2 -pipe"
CHOST="x86_64-pc-linux-gnu"
CONFIG_PROTECT="/etc /usr/lib64/libreoffice/program/sofficerc /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/revdep-rebuild /etc/sandbox.d /etc/terminfo"
CXXFLAGS="-march=skylake -O2 -pipe"
DISTDIR="/var/cache/distfiles"
EMERGE_DEFAULT_OPTS="--jobs=6 --load-average=6"
ENV_UNSET="CARGO_HOME DBUS_SESSION_BUS_ADDRESS DISPLAY GOBIN GOPATH PERL5LIB PERL5OPT PERLPREFIX PERL_CORE PERL_MB_OPT PERL_MM_OPT XAUTHORITY XDG_CACHE_HOME XDG_CONFIG_HOME XDG_DATA_HOME XDG_RUNTIME_DIR"
FCFLAGS="-march=skylake -O2 -pipe"
FEATURES="assume-digests binpkg-docompress binpkg-dostrip binpkg-logs config-protect-if-modified distlocks ebuild-locks fixlafiles ipc-sandbox merge-sync multilib-strict network-sandbox news parallel-fetch pid-sandbox preserve-libs protect-owned qa-unresolved-soname-deps sandbox sfperms strict unknown-features-warn unmerge-logs unmerge-orphans userfetch userpriv usersandbox usersync xattr"
FFLAGS="-march=skylake -O2 -pipe"
GENTOO_MIRRORS="http://gentoo.gossamerhost.com rsync://gentoo.gossamerhost.com/gentoo-distfiles/ ftp://mirrors.tera-byte.com/pub/gentoo http://gentoo.mirrors.tera-byte.com/ rsync://mirrors.tera-byte.com/gentoo"
LANG="pt_BR.utf8"
LDFLAGS="-Wl,-O1 -Wl,--as-needed"
MAKEOPTS="-j7"
PKGDIR="/var/cache/binpkgs"
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 --exclude=/.git"
PORTAGE_TMPDIR="/var/tmp"
USE="L10N=pt_BR X a52 aac acl acpi activities alsa amd64 berkdb branding bzip2 cairo cdda cdr cli crypt cups dbus declarative dri dts dvd dvdr emboss encode exif flac fortran gdbm gif gpm gui iconv icu imlib ipv6 jpeg jumbo-build kde kipi kwallet lcms libglvnd libnotify libtirpc lto mad mng mp3 mp4 mpeg multilib ncurses nls nptl ogg opengl openmp pam pango pcre pdf phonon plasma png policykit ppds pulseaudio qml qt5 readline sdl seccomp semantic-desktop spell split-usr ssl startup-notification svg systemd tcpd tiff truetype udev udisks unicode upower usb vorbis widgets wxwidgets x264 xattr xcb xml xv xvid zlib" ABI_X86="64" ADA_TARGET="gnat_2018" 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="karbon sheets words" COLLECTD_PLUGINS="df interface irq load memory rrdtool swap syslog" CPU_FLAGS_X86="mmx mmxext sse sse2" ELIBC="glibc" GPSD_PROTOCOLS="ashtech aivdm earthmate evermore fv18 garmin garmintxt gpsclock greis isync itrax mtk3301 nmea ntrip navcom oceanserver oldstyle oncore rtcm104v2 rtcm104v3 sirf skytraq superstar2 timing tsip tripmate tnt ublox ubx" GRUB_PLATFORMS="efi-64" INPUT_DEVICES="libinput" KERNEL="linux" LCD_DEVICES="bayrad cfontz cfontz633 glk hd44780 lb216 lcdm001 mtxorb ncurses text" LIBREOFFICE_EXTENSIONS="presenter-console presenter-minimizer" LUA_SINGLE_TARGET="lua5-1" LUA_TARGETS="lua5-1" OFFICE_IMPLEMENTATION="libreoffice" PHP_TARGETS="php7-3 php7-4" POSTGRES_TARGETS="postgres10 postgres11" PYTHON_SINGLE_TARGET="python3_8" PYTHON_TARGETS="python3_8" RUBY_TARGETS="ruby26" USERLAND="GNU" VIDEO_CARDS="nvidia" XTABLES_ADDONS="quota2 psd pknock lscan length2 ipv4options ipset ipp2p iface geoip fuzzy condition tee tarpit sysrq proto steal rawnat logmark ipmark dhcpmac delude chaos account"
Unset: CC, CPPFLAGS, CTARGET, CXX, INSTALL_MASK, LC_ALL, LINGUAS, PORTAGE_BINHOST, PORTAGE_BUNZIP2_COMMAND, PORTAGE_COMPRESS, PORTAGE_COMPRESS_FLAGS, PORTAGE_RSYNC_EXTRA_OPTS, RUSTFLAGS |
_________________ Intel Core i5-9400F / 24GB DDR4 2666MHz / GeForce GTX 1060 3GB
Powered by Gentoo for x86_64
======================================================
Seize the day, and remember to have fun! |
|
Back to top |
|
|
Tony0945 Watchman
Joined: 25 Jul 2006 Posts: 5127 Location: Illinois, USA
|
Posted: Wed Apr 14, 2021 12:22 pm Post subject: |
|
|
Phebom II X6
Code: | Casti ~ # gcc -march=native -E -v - </dev/null 2>&1 | grep cc1 | grep march
/usr/libexec/gcc/x86_64-pc-linux-gnu/9.3.0/cc1 -E -quiet -v - -march=amdfam10 -mmmx -m3dnow -msse -msse2 -msse3 -mno-ssse3 -msse4a -mcx16 -msahf -mno-movbe -mno-aes -mno-sha -mno-pclmul -mpopcnt -mabm -mno-lwp -mno-fma -mno-fma4 -mno-xop -mno-bmi -mno-sgx -mno-bmi2 -mno-pconfig -mno-wbnoinvd -mno-tbm -mno-avx -mno-avx2 -mno-sse4.2 -mno-sse4.1 -mlzcnt -mno-rtm -mno-hle -mno-rdrnd -mno-f16c -mno-fsgsbase -mno-rdseed -mprfchw -mno-adx -mfxsr -mno-xsave -mno-xsaveopt -mno-avx512f -mno-avx512er -mno-avx512cd -mno-avx512pf -mno-prefetchwt1 -mno-clflushopt -mno-xsavec -mno-xsaves -mno-avx512dq -mno-avx512bw -mno-avx512vl -mno-avx512ifma -mno-avx512vbmi -mno-avx5124fmaps -mno-avx5124vnniw -mno-clwb -mno-mwaitx -mno-clzero -mno-pku -mno-rdpid -mno-gfni -mno-shstk -mno-avx512vbmi2 -mno-avx512vnni -mno-vaes -mno-vpclmulqdq -mno-avx512bitalg -mno-movdiri -mno-movdir64b -mno-waitpkg -mno-cldemote -mno-ptwrite --param l1-cache-size=64 --param l1-cache-line-size=64 --param l2-cache-size=512 -mtune=amdfam10
Casti ~ # gcc -march=amdfam10 -E -v - </dev/null 2>&1 | grep cc1 | grep march
/usr/libexec/gcc/x86_64-pc-linux-gnu/9.3.0/cc1 -E -quiet -v - -march=amdfam10 |
|
|
Back to top |
|
|
Tony0945 Watchman
Joined: 25 Jul 2006 Posts: 5127 Location: Illinois, USA
|
Posted: Wed Apr 14, 2021 12:23 pm Post subject: |
|
|
Phenom II X6
Code: | Casti ~ # gcc -march=native -E -v - </dev/null 2>&1 | grep cc1 | grep march
/usr/libexec/gcc/x86_64-pc-linux-gnu/9.3.0/cc1 -E -quiet -v - -march=amdfam10 -mmmx -m3dnow -msse -msse2 -msse3 -mno-ssse3 -msse4a -mcx16 -msahf -mno-movbe -mno-aes -mno-sha -mno-pclmul -mpopcnt -mabm -mno-lwp -mno-fma -mno-fma4 -mno-xop -mno-bmi -mno-sgx -mno-bmi2 -mno-pconfig -mno-wbnoinvd -mno-tbm -mno-avx -mno-avx2 -mno-sse4.2 -mno-sse4.1 -mlzcnt -mno-rtm -mno-hle -mno-rdrnd -mno-f16c -mno-fsgsbase -mno-rdseed -mprfchw -mno-adx -mfxsr -mno-xsave -mno-xsaveopt -mno-avx512f -mno-avx512er -mno-avx512cd -mno-avx512pf -mno-prefetchwt1 -mno-clflushopt -mno-xsavec -mno-xsaves -mno-avx512dq -mno-avx512bw -mno-avx512vl -mno-avx512ifma -mno-avx512vbmi -mno-avx5124fmaps -mno-avx5124vnniw -mno-clwb -mno-mwaitx -mno-clzero -mno-pku -mno-rdpid -mno-gfni -mno-shstk -mno-avx512vbmi2 -mno-avx512vnni -mno-vaes -mno-vpclmulqdq -mno-avx512bitalg -mno-movdiri -mno-movdir64b -mno-waitpkg -mno-cldemote -mno-ptwrite --param l1-cache-size=64 --param l1-cache-line-size=64 --param l2-cache-size=512 -mtune=amdfam10
Casti ~ # gcc -march=amdfam10 -E -v - </dev/null 2>&1 | grep cc1 | grep march
/usr/libexec/gcc/x86_64-pc-linux-gnu/9.3.0/cc1 -E -quiet -v - -march=amdfam10 |
Same with gcc 10.2.0 |
|
Back to top |
|
|
|