| View previous topic :: View next topic |
| Author |
Message |
SteveBallmersChair
Tux's lil' helper
Joined: 12 Jul 2006
Posts: 84
Location: Being thrown around in Redmond
|
 Posted: Tue Feb 27, 2007 5:24 am Post subject: |
 |
|
| Bloodsurfer wrote: | OK, I have to participate here too... 
Running gentoo on my new Core2Duo E6600 with 2,4 GHz and two gigs of DDR2-667 Corsair RAM.
To be honest, I am not pleased with that build times (second time was faster because it didn't have to download the sources). We know that the openoffice ebuild only uses one core to compile, it could be so much faster if it did use all both cores...
Tried WANT_DISTCC="true", but that didn't help, it only used one core... That sucks.
There was one with a notebook and a compiletime of around two hours - how did you manage that? Some trick with niceness or such a thing? I didn't try something like that yet... |
The latest OOo 2.1.x compile time I've gotten on my machine is about 4 hours 20 minutes. (Machine is an X2 4200+, 2x1GB DDR-400.) Turning off ccache helps by about 20 minutes, and moving /var/tmp from a 74 GB 10,000 rpm Raptor to a 45 GB md RAID 0 stripe across the first 15 GB of 3 250GB SATA hard drives on a PCIe x4 controller also helped, particularly in the unpacking and file-shuffling portions. But the compile *is* still single-threaded, and that stinks. At least it's faster than my 2.20 GHz P4-M notebook with 1GB DDR-266 and a 5400 RPM HDD. That took a whopping 10 hours and 52 minutes to compile- and that was OOo 2.0.4, not 2.1. When 2.1 came out, it got the binary version as waiting 11 hours for a compile is nuts, especially when the laptop fan sounds like something taken off the back of a 1U rackmount server.
_________________
Unix is user friendly- it is just picky who its friends are. |
|
| Back to top |
|
 |
yamakawa
Guru
Joined: 28 Jul 2003
Posts: 340
|
| Posted: Wed Apr 25, 2007 10:10 am Post subject: |
|
|
sub-2 hours record (WITHOUT distcc) has been just set! 
(I assume nowhereman's record of 1 hour, 35 minutes and 50 seconds is set by using distcc from the fact he uses "-j8" parameter.)
| Code: | # genlop -t openoffice
* app-office/openoffice
Wed Apr 25 11:01:26 2007 >>> app-office/openoffice-2.2.0
merge time: 2 hours, 51 minutes and 32 seconds.
Wed Apr 25 15:03:09 2007 >>> app-office/openoffice-2.2.0
merge time: 2 hours, 50 minutes and 28 seconds.
Wed Apr 25 18:30:59 2007 >>> app-office/openoffice-2.2.0
merge time: 1 hour, 50 minutes and 36 seconds. |
The first two emerges are made by not specifying multi-threads, which is close to natros's record set by C2D 6400 with 2MB L2 cache. The last one uses WANT_MP set as "true." All files are prefetched by issuing "emerge -f openoffice."
As a fresh install just a few days ago, the system is quite simple otherwise I set "~amd64" arch.
One thing to note is that I separated PORTAGE_TMPDIR="/var/tmp" location and /usr partition. /var sits in a physically different HDD. /usr included under /(root) locates itself in another HDD (WD Raptor X).
| Code: | # emerge --info
Portage 2.1.2.4 (default-linux/amd64/2006.1/desktop, gcc-4.1.2, glibc-2.5-r1, 2.
6.17-gentoo-r7 x86_64)
=================================================================
Gentoo Base System release 1.12.10
Timestamp of tree: Tue, 24 Apr 2007 22:00:10 +0000
dev-lang/python: 2.4.4
dev-python/pycrypto: 2.0.1-r5
sys-apps/sandbox: 1.2.18.1
sys-devel/binutils: 2.17
sys-devel/gcc-config: 1.3.16
sys-devel/libtool: 1.5.23b
virtual/os-headers: 2.6.20-r2
LANG="ja_JP.eucJP"
MAKEOPTS="-j3"
PKGDIR="/usr/portage/packages"
PORTAGE_RSYNC_OPTS="--recursive --links --safe-links --perms --times --compress
--force --whole-file --delete --delete-after --stats --timeout=180 --exclude=/di
stfiles --exclude=/local --exclude=/packages --filter=H_**/files/digest-*"
PORTAGE_TMPDIR="/var/tmp"
PORTDIR="/usr/portage"
SYNC="rsync://rsync.gentoo.org/gentoo-portage"
USE="X acpi alsa amd64 arts avi berkdb bitmap-fonts cairo cdr cjk cli cracklib c
rypt cups dbus dlloader dri dvd dvdr eds emboss encode esd fam firefox fortran g
dbm gif gnome gpm gtk gtk2 hal iconv ipv6 isdnlog jpeg kde libg++ mad midi mikmo
d mp3 mpeg multilibs ncurses nls nptl nptlonly ogg opengl oss pam pcmcia pcre pd
flib perl png ppds pppd python qt3 qt4 quicktime readline reflection sdl session
spell spl ssl tcpd truetype truetype-fonts type1-fonts udev unicode vorbis xml
xorg xv zlib" ALSA_CARDS="ali5451 als4000 atiixp atiixp-modem bt87x ca0106 cmipc
i emu10k1x ens1370 ens1371 es1938 es1968 fm801 hda-intel intel8x0 intel8x0m maes
tro3 trident usb-audio via82xx via82xx-modem ymfpci" ALSA_PCM_PLUGINS="adpcm ala
w asym copy dmix dshare dsnoop empty extplug file hooks iec958 ioplug ladspa lfl
oat linear meter mulaw multi null plug rate route share shm softvol" ELIBC="glib
c" INPUT_DEVICES="keyboard mouse evdev" KERNEL="linux" LCD_DEVICES="bayrad cfont
z cfontz633 glk hd44780 lb216 lcdm001 mtxorb ncurses text" LINGUAS="ja" USERLAND
="GNU" VIDEO_CARDS="nvidia nv vesa"
Unset: CTARGET, EMERGE_DEFAULT_OPTS, INSTALL_MASK, LC_ALL, LDFLAGS, PORTAGE_COM
PRESS, PORTAGE_COMPRESS_FLAGS, PORTAGE_RSYNC_EXTRA_OPTS, PORTDIR_OVERLAY
# equery s openoffice
[ Searching for packages matching openoffice... ]
* size of app-office/openoffice-2.2.0
Total files : 3010
Total size : 245973.48 KiB
# emerge -vp openoffice
These are the packages that would be merged, in order:
Calculating dependencies... done!
[ebuild R ] app-office/openoffice-2.2.0 USE="cairo cups dbus eds firefox gn
ome gtk kde pam -binfilter -branding -debug -gstreamer -java -ldap -mono -odk -s
eamonkey -sound -webdav" LINGUAS="ja -af -ar -as_IN -be_BY -bg -bn -br -bs -ca -
cs -cy -da -de -dz -el -en -en_GB -en_US -en_ZA -eo -es -et -fa -fi -fr -ga -gl
-gu_IN -he -hi_IN -hr -hu -it -km -ko -ku -lt -lv -mk -ml_IN -mr_IN -nb -ne -nl
-nn -nr -ns -or_IN -pa_IN -pl -pt -pt_BR -ru -rw -sh_YU -sk -sl -sr_CS -ss -st -
sv -sw_TZ -ta_IN -te_IN -tg -th -ti_ER -tn -tr -ts -uk -ur_IN -ve -vi -xh -zh_CN
-zh_TW -zu" 0 kB
Total: 1 package (1 reinstall), Size of downloads: 0 kB
|
You can see moderately overclocked CPUs (2.16GHz -> near 3.0GHz) make some difference. This OC is not the maxmum limit yet. I will try it up later soon. 
| Code: | # cat /proc/cpuinfo
processor : 0
vendor_id : GenuineIntel
cpu family : 6
model : 15
model name : Intel(R) Core(TM)2 CPU 6420 @ 2.13GHz
stepping : 6
cpu MHz : 2928.006
cache size : 4096 KB
physical id : 0
siblings : 2
core id : 0
cpu cores : 2
fpu : yes
fpu_exception : yes
cpuid level : 10
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov
pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm syscall nx lm constant_tsc
clflush size : 64
cache_alignment : 64
address sizes : 36 bits physical, 48 bits virtual
power management:
processor : 1
vendor_id : GenuineIntel
cpu family : 6
model : 15
model name : Intel(R) Core(TM)2 CPU 6420 @ 2.13GHz
stepping : 6
cpu MHz : 2928.006
cache size : 4096 KB
physical id : 0
siblings : 2
core id : 1
cpu cores : 2
fpu : yes
fpu_exception : yes
cpuid level : 10
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov
pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm syscall nx lm constant_tsc
pni monitor ds_cpl vmx est tm2 cx16 xtpr lahf_lm
bogomips : 5855.99
clflush size : 64
cache_alignment : 64
address sizes : 36 bits physical, 48 bits virtual
power management: |
Thanks for the tips by neysx and Bloodsurfer (multi-threads), and SteveBallmersChair (separating partitions).
One more tip to add:
unless adding "vba" USE flag, WANT_MP emerge would fail. It does not happen when disabling WANT_MP. |
|
| Back to top |
|
|
yamakawa
Guru
Joined: 28 Jul 2003
Posts: 340
|
| Posted: Thu Apr 26, 2007 2:39 am Post subject: |
|
|
The record goes up to | Code: | Thu Apr 26 11:21:30 2007 >>> app-office/openoffice-2.2.0
merge time: 1 hour, 24 minutes and 30 seconds. |
with OCed 3.2GHz C2D.
Sub-1 hour record is quite possible with near 4GHz dual CPU system. Anyone to try? |
|
| Back to top |
|
|
hermanng
Apprentice
Joined: 28 Jan 2004
Posts: 173
Location: Nussberg,Germany,Earth
|
| Posted: Thu Jun 14, 2007 1:35 pm Post subject: |
|
|
The record goes up even more ... | Code: | tornado ~ # genlop -t openoffice
* app-office/openoffice
Thu Jun 14 09:06:22 2007 >>> app-office/openoffice-2.2.1
merge time: 1 hour, 2 minutes and 43 seconds. |
on a (dual boot) MacPro (2x Xeon 5150 (2.66 GHz), 4GB). I have to admit, it was done with "-java". ccache was off and the emerge was done with WANT_MP and MAKE_OPTS=-j9. /var/tmp/portage sits on a different disk( 250GB and 400 GB SATA drives ). Could possibly be enhanced some more, the machine was not saturated with that - load was between 5 and 13 all the time.
Someone with 8 GB memory could possibly use a 4GB tmpfs to do the compile on and make it *real* fast. 
EDIT: forgot to say: distcc was not used - really don't need it on this machine 
... and 1 more trick : I used | Code: | | WANT_MP="true" emerge openoffice > /dev/null |
to suppress the screen output
_________________
Hermann |
|
| Back to top |
|
|
phsdv
Guru
Joined: 13 Mar 2005
Posts: 372
Location: Europe
|
| Posted: Tue Dec 25, 2007 9:15 pm Post subject: |
|
|
First to be below 1 hour?!? and then by 12 minutes....
| Code: | $ genlop -t openoffice
* app-office/openoffice
Sun Dec 23 17:20:35 2007 >>> app-office/openoffice-2.3.1
merge time: 47 minutes and 58 seconds.
|
system: | Code: | 2.6.23-gentoo-r5 x86_64 Intel(R) Core(TM)2 Quad CPU Q6600 @ 2.40GHz
ACCEPT_KEYWORDS="amd64"
CBUILD="x86_64-pc-linux-gnu"
CFLAGS="-O2 -march=nocona -pipe -mno-tls-direct-seg-refs"
CHOST="x86_64-pc-linux-gnu" |
| Code: | MAKEOPTS="-j5" # 4+1 for quad core
WANT_MP="true" |
| Code: | $ mount
none on /var/tmp/portage type tmpfs (rw,size=5000M,nr_inodes=1M) |
I have 'only' 4G of mem, so some swap was used, but never more than 1G (i did not have more installed...)
Use flags: | Code: | | app-office/openoffice-2.3.1 USE="cairo cups dbus firefox gstreamer gtk pam -binfilter -debug -eds -gnome -java -kde -ldap -mono -odk -seamonkey -sound -webdav -xulrunner" |
|
|
| Back to top |
|
|
agboon
n00b
Joined: 26 Dec 2007
Posts: 1
|
| Posted: Wed Dec 26, 2007 11:15 pm Post subject: |
|
|
Sorry to grab your new record so soon phsdv... 
I tried it and got it compiled even quicker:
| Code: |
# genlop -t openoffice
* app-office/openoffice
Wed Dec 26 23:46:20 2007 >>> app-office/openoffice-2.3.1
merge time: 44 minutes and 6 seconds.
|
System:
| Code: |
# uname -a
Linux einstein 2.6.23-gentoo-r3 #1 SMP Thu Dec 20 23:25:09 CET 2007 x86_64 Intel(R) Core(TM)2 Quad CPU Q6600 @ 2.40GHz GenuineIntel GNU/Linux
|
Intel Core2Quad Q6600 2.40GHz@3.15GHz
with 4GB Kingston HyperX PC8500 @ 900MHz
I also used
| Code: |
MAKEOPTS="-j5"
WANT_MP="true"
|
Use flags:
| Code: |
[ebuild R ] app-office/openoffice-2.3.1 USE="cairo cups dbus firefox gnome gtk java pam -binfilter -debug -eds -gstreamer -kde -ldap -mono -odk -seamonkey -sound -webdav -xulrunner"
|
Looks like an openoffice emerge is a nice multi-core Gentoo-benchmark |
|
| Back to top |
|
|
sLumpia
Tux's lil' helper
Joined: 10 Jul 2006
Posts: 114
Location: troller land
|
| Posted: Thu Dec 27, 2007 6:45 am Post subject: |
|
|
| agboon wrote: |
Looks like an openoffice emerge is a nice multi-core Gentoo-benchmark |
|
|
| Back to top |
|
|
phsdv
Guru
Joined: 13 Mar 2005
Posts: 372
Location: Europe
|
| Posted: Thu Dec 27, 2007 9:16 am Post subject: |
|
|
| agboon wrote: | | Sorry to grab your new record so soon phsdv... |
Was to be expected, congratulations! |
|
| Back to top |
|
|
posey
n00b
Joined: 15 Dec 2007
Posts: 16
|
| Posted: Wed Feb 20, 2008 11:48 pm Post subject: |
|
|
I realize this isn't record breaking, but I'm feeling really good about my system right now, especially comparing with other people's times. Athlon 64 with 1gb ddr ram. Gnome was running at the time, but I wasn't using the box. Maybe I should also mention I bootstrapped the box?
| Code: | genlop -t openoffice
* app-office/openoffice
Wed Feb 20 12:23:51 2008 >>> app-office/openoffice-2.3.1
merge time: 3 hours, 45 minutes and 52 seconds.
|
| Code: | cat /proc/cpuinfo
processor : 0
vendor_id : AuthenticAMD
cpu family : 15
model : 4
model name : AMD Athlon(tm) 64 Processor 3200+
stepping : 8
cpu MHz : 2000.000
cache size : 1024 KB
fpu : yes
fpu_exception : yes
cpuid level : 1
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 syscall nx mmxext lm 3dnowext 3dnow rep_good
bogomips : 4009.75
TLB size : 1024 4K pages
clflush size : 64
cache_alignment : 64
address sizes : 40 bits physical, 48 bits virtual
power management: ts fid vid ttp
|
| Code: | emerge --info
Portage 2.1.4.4 (default-linux/amd64/2007.0, gcc-4.1.2, glibc-2.6.1-r0, 2.6.23-gentoo-r8 x86_64)
=================================================================
System uname: 2.6.23-gentoo-r8 x86_64 AMD Athlon(tm) 64 Processor 3200+
Timestamp of tree: Tue, 19 Feb 2008 23:46:01 +0000
app-shells/bash: 3.2_p17-r1
dev-java/java-config: 1.3.7, 2.1.4
dev-lang/python: 2.4.4-r6
dev-python/pycrypto: 2.0.1-r6
sys-apps/baselayout: 1.12.10-r5
sys-apps/sandbox: 1.2.18.1-r2
sys-devel/autoconf: 2.13, 2.61-r1
sys-devel/automake: 1.5, 1.7.9-r1, 1.8.5-r3, 1.9.6-r2, 1.10
sys-devel/binutils: 2.18-r1
sys-devel/gcc-config: 1.4.0-r4
sys-devel/libtool: 1.5.26
virtual/os-headers: 2.6.23-r3
ACCEPT_KEYWORDS="amd64"
CBUILD="x86_64-pc-linux-gnu"
CFLAGS="-march=k8 -O2 -pipe"
CHOST="x86_64-pc-linux-gnu"
CONFIG_PROTECT="/etc /usr/share/X11/xkb"
CONFIG_PROTECT_MASK="/etc/env.d /etc/env.d/java/ /etc/fonts/fonts.conf /etc/gconf /etc/terminfo /etc/udev/rules.d"
CXXFLAGS="-march=k8 -O2 -pipe"
DISTDIR="/usr/portage/distfiles"
FEATURES="distlocks metadata-transfer sandbox sfperms strict unmerge-orphans userfetch"
GENTOO_MIRRORS="ftp://mirror.iawnet.sandia.gov/pub/gentoo/ "
MAKEOPTS="-j2"
PKGDIR="/usr/portage/packages"
PORTAGE_RSYNC_OPTS="--recursive --links --safe-links --perms --times --compress --force --whole-file --delete --stats --timeout=180 --exclude=/distfiles --exclude=/local --exclude=/packages"
PORTAGE_TMPDIR="/var/tmp"
PORTDIR="/usr/portage"
SYNC="rsync://rsync.namerica.gentoo.org/gentoo-portage"
USE="X aac acl alsa amd64 avahi berkdb bitmap-fonts cairo cdr cli cracklib crypt css cups dbus dri dvd dvdr dvdread encode firefox flac fortran gdbm gif gnome gpm gstreamer gtk hal iconv isdnlog java jpeg krb4 mad midi mmx mp3 mudflap ncurses nls nptl nptlonly nsplugin nvidia ogg opengl openmp pam pcre pdf perl png pppd python quicktime readline reflection samba session spell spl sse sse2 ssl svg tcpd tiff truetype truetype-fonts type1-fonts unicode vorbis xml xorg xv xvmc zlib" 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" ALSA_PCM_PLUGINS="adpcm alaw asym copy dmix dshare dsnoop empty extplug file hooks iec958 ioplug ladspa lfloat linear meter mulaw multi null plug rate route share shm softvol" APACHE2_MODULES="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 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" ELIBC="glibc" INPUT_DEVICES="keyboard mouse" KERNEL="linux" LCD_DEVICES="bayrad cfontz cfontz633 glk hd44780 lb216 lcdm001 mtxorb ncurses text" USERLAND="GNU" VIDEO_CARDS="nvidia"
Unset: CPPFLAGS, CTARGET, EMERGE_DEFAULT_OPTS, INSTALL_MASK, LANG, LC_ALL, LDFLAGS, LINGUAS, PORTAGE_COMPRESS, PORTAGE_COMPRESS_FLAGS, PORTAGE_RSYNC_EXTRA_OPTS, PORTDIR_OVERLAY
|
|
|
| Back to top |
|
|
speeddemon
Apprentice
Joined: 27 Sep 2003
Posts: 162
|
| Posted: Fri Aug 29, 2008 12:01 am Post subject: |
|
|
Just noticed it updated, checked the time, figured I'd post it here. Never had noticed the bit about WANT_MP="true"
| Code: | ELHCshee ryan # genlop -t openoffice
* app-office/openoffice
Sat Mar 29 02:42:21 2008 >>> app-office/openoffice-2.3.1-r1
merge time: 2 hours, 39 minutes and 15 seconds.
Thu Aug 28 19:44:23 2008 >>> app-office/openoffice-2.4.1
merge time: 38 minutes and 57 seconds.
|
Anybody beat 39 minutes?
_________________
Cats are deadly animals. If you stick your nose up their crotch and snort their piss, THEY CAN KILL YOU!!! |
|
| Back to top |
|
|
srunni
Guru
Joined: 26 Dec 2007
Posts: 365
|
| Posted: Mon Sep 01, 2008 7:06 am Post subject: |
|
|
| speeddemon wrote: | Just noticed it updated, checked the time, figured I'd post it here. Never had noticed the bit about WANT_MP="true"
| Code: | ELHCshee ryan # genlop -t openoffice
* app-office/openoffice
Sat Mar 29 02:42:21 2008 >>> app-office/openoffice-2.3.1-r1
merge time: 2 hours, 39 minutes and 15 seconds.
Thu Aug 28 19:44:23 2008 >>> app-office/openoffice-2.4.1
merge time: 38 minutes and 57 seconds.
|
Anybody beat 39 minutes? |
What are the specs on that box?
_________________
| Patrick Ewing wrote: | | if it walks like a duck and talks like a duck, it’s a duck, right? So if this duck is not giving you the noise that you want, you’ve got to just punch that duck until it returns what you expect. |
|
|
| Back to top |
|
|
sal
n00b
Joined: 01 Sep 2008
Posts: 2
|
| Posted: Mon Sep 01, 2008 8:15 pm Post subject: |
|
|
| thanks for the valuable information you guys... i'm trying to absorb things all at once... really glad to have found this forum |
|
| Back to top |
|
|
speeddemon
Apprentice
Joined: 27 Sep 2003
Posts: 162
|
| Posted: Tue Sep 02, 2008 9:44 pm Post subject: |
|
|
| srunni wrote: | | speeddemon wrote: | Just noticed it updated, checked the time, figured I'd post it here. Never had noticed the bit about WANT_MP="true"
| Code: | ELHCshee ryan # genlop -t openoffice
* app-office/openoffice
Sat Mar 29 02:42:21 2008 >>> app-office/openoffice-2.3.1-r1
merge time: 2 hours, 39 minutes and 15 seconds.
Thu Aug 28 19:44:23 2008 >>> app-office/openoffice-2.4.1
merge time: 38 minutes and 57 seconds.
|
Anybody beat 39 minutes? |
What are the specs on that box? |
8G DDR2667 (fb-dimms) (quad channel, 4 sticks)
2 x Xeon 5405 quad cores (80 watts, 45 nm)
Tyan Tempest i5400XT mobo
Could be faster if it supported the higher FSB speed and had faster memory. Not really worried about it though. Its not as much of a concern as people think in practical usage. People seem to ignore the cache, 12mb shared cache. For what I use it for its great, love the mobo.
_________________
Cats are deadly animals. If you stick your nose up their crotch and snort their piss, THEY CAN KILL YOU!!! |
|
| Back to top |
|
|
srunni
Guru
Joined: 26 Dec 2007
Posts: 365
|
| Posted: Tue Sep 02, 2008 10:27 pm Post subject: |
|
|
| speeddemon wrote: | | srunni wrote: | | speeddemon wrote: | Just noticed it updated, checked the time, figured I'd post it here. Never had noticed the bit about WANT_MP="true"
| Code: | ELHCshee ryan # genlop -t openoffice
* app-office/openoffice
Sat Mar 29 02:42:21 2008 >>> app-office/openoffice-2.3.1-r1
merge time: 2 hours, 39 minutes and 15 seconds.
Thu Aug 28 19:44:23 2008 >>> app-office/openoffice-2.4.1
merge time: 38 minutes and 57 seconds.
|
Anybody beat 39 minutes? |
What are the specs on that box? |
8G DDR2667 (fb-dimms) (quad channel, 4 sticks)
2 x Xeon 5405 quad cores (80 watts, 45 nm)
Tyan Tempest i5400XT mobo
Could be faster if it supported the higher FSB speed and had faster memory. Not really worried about it though. Its not as much of a concern as people think in practical usage. People seem to ignore the cache, 12mb shared cache. For what I use it for its great, love the mobo. |
Are you using 4 GB of the RAM for tmpfs? Is that how you got it to build so fast? Or is it just the CPUs?
_________________
| Patrick Ewing wrote: | | if it walks like a duck and talks like a duck, it’s a duck, right? So if this duck is not giving you the noise that you want, you’ve got to just punch that duck until it returns what you expect. |
|
|
| Back to top |
|
|
speeddemon
Apprentice
Joined: 27 Sep 2003
Posts: 162
|
| Posted: Tue Sep 02, 2008 10:46 pm Post subject: |
|
|
| srunni wrote: | | speeddemon wrote: |
8G DDR2667 (fb-dimms) (quad channel, 4 sticks)
2 x Xeon 5405 quad cores (80 watts, 45 nm)
Tyan Tempest i5400XT mobo
Could be faster if it supported the higher FSB speed and had faster memory. Not really worried about it though. Its not as much of a concern as people think in practical usage. People seem to ignore the cache, 12mb shared cache. For what I use it for its great, love the mobo. |
Are you using 4 GB of the RAM for tmpfs? Is that how you got it to build so fast? Or is it just the CPUs? |
Just the CPUs. Only tmpfs at the moment is /var/run/jack (jack audio connection kit). I have a few smaller disks in RAID 0 for the build, temp, and other non-important directories (140ish MB/s). I haven't really played with it too much, no real point (its a server; nothing I do is going to be noticeable). Few programs take over 5 minutes to build. Run -j9 in make.conf. I could enable hyperthreading too.
_________________
Cats are deadly animals. If you stick your nose up their crotch and snort their piss, THEY CAN KILL YOU!!! |
|
| Back to top |
|
|
srunni
Guru
Joined: 26 Dec 2007
Posts: 365
|
| Posted: Wed Sep 03, 2008 2:02 am Post subject: |
|
|
| speeddemon wrote: | | srunni wrote: | | speeddemon wrote: |
8G DDR2667 (fb-dimms) (quad channel, 4 sticks)
2 x Xeon 5405 quad cores (80 watts, 45 nm)
Tyan Tempest i5400XT mobo
Could be faster if it supported the higher FSB speed and had faster memory. Not really worried about it though. Its not as much of a concern as people think in practical usage. People seem to ignore the cache, 12mb shared cache. For what I use it for its great, love the mobo. |
Are you using 4 GB of the RAM for tmpfs? Is that how you got it to build so fast? Or is it just the CPUs? |
Just the CPUs. Only tmpfs at the moment is /var/run/jack (jack audio connection kit). I have a few smaller disks in RAID 0 for the build, temp, and other non-important directories (140ish MB/s). I haven't really played with it too much, no real point (its a server; nothing I do is going to be noticeable). Few programs take over 5 minutes to build. Run -j9 in make.conf. I could enable hyperthreading too. |
You should try building it with /var/tmp/portage on tmpfs. I bet it'd drop the build time by another few minutes.
Also, what is the clock speed on those Xeons?
_________________
| Patrick Ewing wrote: | | if it walks like a duck and talks like a duck, it’s a duck, right? So if this duck is not giving you the noise that you want, you’ve got to just punch that duck until it returns what you expect. |
|
|
| Back to top |
|
|
speeddemon
Apprentice
Joined: 27 Sep 2003
Posts: 162
|
| Posted: Wed Sep 03, 2008 7:50 pm Post subject: |
|
|
| srunni wrote: | You should try building it with /var/tmp/portage on tmpfs. I bet it'd drop the build time by another few minutes.
Also, what is the clock speed on those Xeons? |
Tried it, only about 30 second improvement. I think I remembered having tried it before and it didn't help much.
Clock speed of 2.0 GHz.
_________________
Cats are deadly animals. If you stick your nose up their crotch and snort their piss, THEY CAN KILL YOU!!! |
|
| Back to top |
|
|
juantxorena
Apprentice
Joined: 19 Mar 2006
Posts: 176
Location: The Shire
|
| Posted: Thu Sep 04, 2008 1:06 pm Post subject: |
|
|
Are you sure that these benchmarks worth something? I don't know what things have an impact on the compile time, but I don't see this genlop thing as a good benchmark. This is what I have since I installed genlop:
| Code: |
Mon Oct 23 22:17:13 2006 >>> app-office/openoffice-2.0.4
merge time: 5 hours, 51 minutes and 17 seconds.
Fri Nov 17 17:49:02 2006 >>> app-office/openoffice-2.0.4
merge time: 5 hours, 49 minutes and 22 seconds.
Wed Dec 6 21:20:49 2006 >>> app-office/openoffice-2.0.4
merge time: 5 hours, 37 minutes and 28 seconds.
Fri Dec 8 22:40:58 2006 >>> app-office/openoffice-2.0.4
merge time: 5 hours, 14 minutes and 46 seconds.
Sun Dec 24 16:33:04 2006 >>> app-office/openoffice-2.0.4
merge time: 5 hours, 55 minutes and 40 seconds.
Fri Dec 29 11:41:02 2006 >>> app-office/openoffice-2.0.4
merge time: 2 hours and 31 seconds.
Tue Jan 9 14:33:20 2007 >>> app-office/openoffice-2.0.4
merge time: 4 hours, 11 minutes and 27 seconds.
Wed Jan 10 20:53:36 2007 >>> app-office/openoffice-2.0.4
merge time: 5 hours, 40 minutes and 32 seconds.
Sat Apr 7 17:52:05 2007 >>> app-office/openoffice-2.1.0-r1
merge time: 5 hours, 45 minutes and 51 seconds.
Tue May 1 22:38:54 2007 >>> app-office/openoffice-2.2.0
merge time: 5 hours, 47 minutes and 49 seconds.
Sat Jun 16 23:48:10 2007 >>> app-office/openoffice-2.2.1
merge time: 6 hours, 39 minutes and 28 seconds.
Sun Aug 12 13:27:38 2007 >>> app-office/openoffice-2.2.1
merge time: 3 hours, 16 minutes and 51 seconds.
Fri Sep 21 16:24:41 2007 >>> app-office/openoffice-2.3.0
merge time: 5 hours, 28 minutes and 47 seconds.
Tue Nov 20 17:48:43 2007 >>> app-office/openoffice-2.3.0
merge time: 4 hours, 23 minutes and 14 seconds.
Wed Dec 5 18:59:49 2007 >>> app-office/openoffice-2.3.1
merge time: 5 hours, 1 minute and 18 seconds.
Thu Mar 6 15:30:47 2008 >>> app-office/openoffice-2.3.1-r1
merge time: 5 hours, 56 minutes and 11 seconds.
Sat Apr 19 12:27:11 2008 >>> app-office/openoffice-2.3.1-r1
merge time: 1 hour, 57 minutes and 11 seconds.
Sun Apr 20 23:32:31 2008 >>> app-office/openoffice-2.4.0
merge time: 6 hours, 4 minutes and 40 seconds.
Fri Apr 25 14:35:48 2008 >>> app-office/openoffice-2.4.0
merge time: 6 hours, 18 minutes and 46 seconds.
Tue Jul 8 10:54:29 2008 >>> app-office/openoffice-2.4.1
merge time: 1 hour, 39 minutes and 3 seconds.
Sat Jul 19 18:45:21 2008 >>> app-office/openoffice-2.4.1
merge time: 4 hours, 39 minutes and 37 seconds.
Tue Aug 19 11:44:13 2008 >>> app-office/openoffice-2.4.1
merge time: 3 hours, 35 minutes and 53 seconds.
|
In these 2 years my computer has been the same, AMD 64 3500+ with 1Gb @ 400 MHz, the same HD with the same partitions and the same filesystem, the same ccache size (2Gb), always on (I erase it from time to time, though). Sometimes I compile it while doing stuff, others there are only some programs running on the background (with KDE, like ktorrent or amule), and a few times without starting the X server, compiling it from a terminal (these times are not the ones with the shortest time). And nearly 5 hours of difference between compiles are too much (3 hours between 2 compilations of the last version).
So, what's happening? Maybe there are too many things that affect the compilation time that isn't a good benchmark?
_________________
I cannot write English very well. Please, correct any mistake so that I can improve. |
|
| Back to top |
|
|
speeddemon
Apprentice
Joined: 27 Sep 2003
Posts: 162
|
| Posted: Thu Sep 04, 2008 4:00 pm Post subject: |
|
|
Well for one there are a bunch of different versions there. You also never know what else you were doing that could have had an effect too. Nobody ever said it was a perfect benchmark. Its just interesting to see how fast modern hardware can compile a massive program like openoffice.
Some of us from the "way back when" gentoo era remember it taking 12+ hours just to get X installed. Full systems would take a couple days. Stuff like openoffice was one of those where you came back the next morning to see if it was done.
_________________
Cats are deadly animals. If you stick your nose up their crotch and snort their piss, THEY CAN KILL YOU!!! |
|
| Back to top |
|
|
arnvidr
Guru
Joined: 19 Aug 2004
Posts: 441
Location: Oslo, Norway
|
| Posted: Fri Sep 05, 2008 5:58 am Post subject: |
|
|
| speeddemon wrote: | | Stuff like openoffice was one of those where you came back the next morning to see if it was done. |
It wasn't. You had to let it run while you were at work/school too.
_________________
|
|
| Back to top |
|
|
xmaes
n00b
Joined: 13 Aug 2008
Posts: 69
Location: Madrid
|
| Posted: Sat Sep 06, 2008 9:08 am Post subject: |
|
|
Thank you for the tip went from
| Quote: | Wed Aug 27 19:08:23 2008 >>> app-office/openoffice-2.4.1
merge time: 3 hours, 41 minutes and 1 second. |
to
| Quote: | Thu Sep 4 20:01:20 2008 >>> app-office/openoffice-2.4.1
merge time: 52 minutes and 22 seconds. |
on a Q6600 with WANT_MP="true"
Xavier |
|
| Back to top |
|
|
ericxx2005
Apprentice
Joined: 05 Apr 2005
Posts: 231
Location: 10 square miles surrounded by reality
|
| Posted: Sat Sep 06, 2008 9:41 pm Post subject: |
|
|
I'm going to give this a shot now. I've pretty much been using openoffice-bin since I realized that every openoffice update was taking 8+ hours on my Athlon XP 1700+. I would expect to be able to compile it in 2-3 hours with my new system, GCC, and WANT_SMP="true".
_________________
AMD 1090T | 2 GB | AMD 890GX | Gentoo x86_64
Core i7-920 | 6 GB | AMD 5850 | Windows 7 Pro |
|
| Back to top |
|
|
srunni
Guru
Joined: 26 Dec 2007
Posts: 365
|
| Posted: Sat Sep 06, 2008 9:49 pm Post subject: |
|
|
It says on the Wiki that "you really shouldn't do this". Does anyone understand why? I don't understand what they mean by "if the package is prone to break otherwise". Does this mean you shouldn't use it on unstable ebuilds?
_________________
| Patrick Ewing wrote: | | if it walks like a duck and talks like a duck, it’s a duck, right? So if this duck is not giving you the noise that you want, you’ve got to just punch that duck until it returns what you expect. |
|
|
| Back to top |
|
|
platojones
Veteran
Joined: 23 Oct 2002
Posts: 1425
Location: Just over the horizon
|
| Posted: Sun Sep 07, 2008 4:00 am Post subject: |
|
|
Mainstream CPU+Motherboard makes all the difference...intriguing..YES!
First bought a Q9550 Core 2 Quad for use in my 18 month old i957x Asus P5W64 WS Pro...it was certified to work with the new Yorkies, so I figured I would save a few denari and hold off on a motherboard purchase...So I compiled OO as a benchmark. Compared to my older E6700 Conroe, this is what I got:
| Code: |
Thu Aug 14 10:10:08 2008 >>> app-office/openoffice-2.4.1
merge time: 2 hours, 19 minutes and 49 seconds.
Tue Sep 2 00:58:25 2008 >>> app-office/openoffice-2.4.1
merge time: 1 hour, 48 minutes and 24 seconds.
|
Well, I didn't think that was as much bang as I was expecting...read about the new P45 chipset and went ahead and splurged thinking a 3.4 stable OC was better than a 3.1 stable OC...Gee, I was right...but I didn't think I would be this right:
| Code: |
Tue Sep 2 00:58:25 2008 >>> app-office/openoffice-2.4.1
merge time: 1 hour, 48 minutes and 24 seconds.
Sat Sep 6 23:48:31 2008 >>> app-office/openoffice-2.4.1
merge time: 37 minutes and 55 seconds.
|
This is a technology snapshot of about 1 year's difference...amazing what a difference a year makes in computers. What is really sad is that in another year...that will probably be about 15 minutes emerge time...btw, the original emerge with my 18 month old E6700 (OC to 3.1) and original i975x MB was this:
| Code: |
Thu Aug 14 10:10:08 2008 >>> app-office/openoffice-2.4.1
merge time: 2 hours, 19 minutes and 49 seconds.
|
Please note...this is not with WANT_MP="true" set! |
|
| Back to top |
|
|
ericxx2005
Apprentice
Joined: 05 Apr 2005
Posts: 231
Location: 10 square miles surrounded by reality
|
| Posted: Sun Sep 07, 2008 2:50 pm Post subject: |
|
|
| Quote: | genlop -t openoffice
* app-office/openoffice
Sun Sep 7 00:41:59 2008 >>> app-office/openoffice-2.4.1
merge time: 1 hour, 37 minutes and 48 seconds |
Wow, even less then I thought. I think I'll stay with the regular version from now on, no need to go "-bin"
_________________
AMD 1090T | 2 GB | AMD 890GX | Gentoo x86_64
Core i7-920 | 6 GB | AMD 5850 | Windows 7 Pro |
|
| Back to top |
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
|
Gentoo Chat
