View previous topic :: View next topic |
Author |
Message |
kristian aka unsolo n00b
Joined: 05 Jun 2005 Posts: 4
|
Posted: Sun Jun 05, 2005 9:31 pm Post subject: Do you have a 64bit cpu maybe you dont know it |
|
|
Intel 8xx and 6xx 775 series have x86_64 aka 64bit or amd64 or em64t or i32e
All intel XEONS with 800MHZ FSB is also x86_64
To take advantage of this you should
Use the amd64 livecd's and amd64 guide when installing
example /etc/make.conf
Code: |
CHOST="x86_64-pc-linux-gnu"
#CFLAGS="-O3 -march=nocona -pipe -ffast-math -mfpmath=sse,387"
CFLAGS="-O2 -march=nocona -pipe "
USE"-3DNow" |
you should also activate hyperthreading in your kernels as the x86_64 version is a lot better than the old one
also ACCEPT_KEYWORDS="~amd64" in case you wonder.
join #gentoo-amd64 at freenode for further help
I have been running gentoo on x86_64 intel xeons since september 2004 and i can ashure you a lot of things work |
|
Back to top |
|
|
adaptr Watchman
Joined: 06 Oct 2002 Posts: 6730 Location: Rotterdam, Netherlands
|
Posted: Sun Jun 05, 2005 10:55 pm Post subject: |
|
|
Quote: | Intel 8xx and 6xx 775 series have x86_64 aka 64bit |
Are you sure about this?
Intel Pentium-4s that support emt64 instructions have 64bit memory support - they are not 64bit processors.
That's what the emt stands for - Extended Memory Technology.
So far, only Intel Itaniums and AMD64 cpu's in 64bit mode are true 64bit CPU's
And, of course, all DEC/Compaq Alpha's and MIPS processors - but then, they have been 64bit for years.
And while it is true that for the x86 platform the move from 32 to 64bit memory will make the biggest impact - moving from 32bit to 64bit instructions would break the x86 compatibility - that's not what you say; they are not 64bit processors.
Quote: | also ACCEPT_KEYWORDS="~amd64" in case you wonder. |
I don't, actually - I would much prefer to go with the stable amd64 branch, seeing as you say later on:
Quote: | i can ashure you a lot of things work |
I would like to see "everything" work before committing myself. _________________ >>> emerge (3 of 7) mcse/70-293 to /
Essential tools: gentoolkit eix profuse screen |
|
Back to top |
|
|
kristian aka unsolo n00b
Joined: 05 Jun 2005 Posts: 4
|
Posted: Mon Jun 06, 2005 6:33 am Post subject: |
|
|
Ok i can agree with the fact that its not a 100% 64 bit as they support x86
However the x86_64 is adopted from amd64 processors so i not only guess when i say they can run without multilib with only lib(64) which is quite hard unless the entire execution process is in 64bit or emulates 64 bit. the latteral would be insanely slower than x86 which it is not.
And regarding which processors i am 99% shure its according to intels own pages. |
|
Back to top |
|
|
projectle Tux's lil' helper
Joined: 30 Jan 2005 Posts: 90
|
Posted: Mon Jun 06, 2005 6:34 am Post subject: Intel's 64-bit processors all blow monkey turds |
|
|
All of Intel's 64-bit processors blow monkey turds.
Take a look at performance comparisons of the Intel 64-bit processors (or their dual core for that matter) vs the Athlon64 or the Opteron in the single or dual core varieties.
Intel had it for a long time, although like Microsoft, they took too much of the market, felt comfortable and in turn let their quality and products slip until someone else could come along and a** rape them from behind. |
|
Back to top |
|
|
adaptr Watchman
Joined: 06 Oct 2002 Posts: 6730 Location: Rotterdam, Netherlands
|
Posted: Mon Jun 06, 2005 3:12 pm Post subject: |
|
|
That's... not how I would have put it
But it's out now, so let's run with it. _________________ >>> emerge (3 of 7) mcse/70-293 to /
Essential tools: gentoolkit eix profuse screen |
|
Back to top |
|
|
kristian aka unsolo n00b
Joined: 05 Jun 2005 Posts: 4
|
Posted: Mon Jun 06, 2005 6:44 pm Post subject: |
|
|
It still pumps some more in x86_64 mode than in x86 mode
Some people might have a reason for haiving this
so why not just let them have these cpus and happy that they run gentoo ?
And please refer to linux not windows benchmarks on perfomance comparison also make shure its only on x86_64
not some ia32 running on x86_64
lets say HPL for instance
High performance linpack benchmark used for top500.org |
|
Back to top |
|
|
projectle Tux's lil' helper
Joined: 30 Jan 2005 Posts: 90
|
Posted: Mon Jun 06, 2005 7:41 pm Post subject: |
|
|
I said nothing about Microsoft benchmarks. I merely said that Intel got too big for their own good.
If you are saying that Intel's D or EM64T line products are even in the same league as the AMD Products, then your logic and thought processes are 95% craptastic, 5% craptacular. |
|
Back to top |
|
|
adaptr Watchman
Joined: 06 Oct 2002 Posts: 6730 Location: Rotterdam, Netherlands
|
Posted: Tue Jun 07, 2005 8:53 am Post subject: |
|
|
Then I'm sure you wouldn't mind telling us what actual facts you base this entertaining yet vulgar assessment on, would you ?
Cite references and source material, please. _________________ >>> emerge (3 of 7) mcse/70-293 to /
Essential tools: gentoolkit eix profuse screen |
|
Back to top |
|
|
markkuk Guru
Joined: 29 Nov 2002 Posts: 446
|
|
Back to top |
|
|
kristian aka unsolo n00b
Joined: 05 Jun 2005 Posts: 4
|
Posted: Tue Jun 07, 2005 11:12 pm Post subject: |
|
|
your ignorance was so great that i didnt even have to answer my self |
|
Back to top |
|
|
augury l33t
Joined: 22 May 2004 Posts: 722 Location: philadelphia
|
Posted: Wed Jun 29, 2005 4:56 am Post subject: |
|
|
lshw -C cpu
intel 640
Code: |
*-cpu
product: Intel(R) Pentium(R) 4 CPU 3.20GHz
vendor: Intel Corp.
physical id: 1
bus info: cpu@0
version: 15.4.3
serial: 0000-0F43-0000-0000-0000-0000
size: 18EHz
width: 64 bits
capabilities: fpu fpu_exception wp 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 pbe x86-64 pni monitor ds_cpl est cid cx16 xtpr
configuration: id=0
*-logicalcpu:0
description: Logical CPU
physical id: 0.1
width: 64 bits
capabilities: logical
*-logicalcpu:1
description: Logical CPU
physical id: 0.2
width: 64 bits
capabilities: logical
|
intel 3.2ghz p4-c
Code: |
*-cpu
description: CPU
product: Intel(R) Pentium(R) 4 CPU 3.20GHz
vendor: Intel Corp.
physical id: 4
bus info: cpu@0
version: 15.2.9
slot: CPU 1
size: 3200MHz
capacity: 3600MHz
width: 32 bits
clock: 200MHz
capabilities: fpu fpu_exception wp 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 pbe cid xtpr cpufreq
configuration: id=0
*-logicalcpu:0
description: Logical CPU
physical id: 0.1
width: 32 bits
capabilities: logical
*-logicalcpu:1
description: Logical CPU
physical id: 0.2
width: 32 bits
capabilities: logical
|
im wondering if -m64 and associated flags are needed or if nocona takes care of all of this. |
|
Back to top |
|
|
bollucks l33t
Joined: 27 Oct 2004 Posts: 606
|
Posted: Wed Jun 29, 2005 7:24 am Post subject: |
|
|
You need the "lm" flag which means long mode ie 64 bit. |
|
Back to top |
|
|
julot n00b
Joined: 27 Aug 2003 Posts: 55 Location: Mexicus
|
Posted: Wed Jun 29, 2005 10:08 pm Post subject: Yes, x86 Indeed. |
|
|
The 6XX series of P4 and the newcomer celeron, 326 and so on, will support x86_64.
But you are wrong of using 3dnow, even sse2, sse1, because are 32 bits based instructions.
In this moment GCC are not optimised for the complete architecture and thus is not desirable to mix the 32 and 64 bits instructions, (let the longhorn programmers do that kind of rubbish, The AMD X_64 set already have sse2 and sse 64 bits counterparts so it is NOT needed to include:
So my Make.conf for the opterons is:
Quote: |
CFLAGS="-march=k8 -O2 -pipe -fomit-frame-pointers"
#CFLAGS="-march=k8 -O2 -pipe"
CHOST="x86_64-pc-linux-gnu"
CXXFLAGS="${CFLAGS}"
MAKEOPTS="-j5"
FEATURES="ccache"
CCACHE_SIZE="2G"
USE="threads doc javascript java imap nptl apache2"
|
And it is pretty stable, so the intel counterpart must be:
Quote: |
CFLAGS="-march=nocona -O2 -pipe -fomit-frame-pointer"
CHOST="x86_64-pc-linux-gnu"
CXXFLAGS="${CFLAGS}"
MAKEOPTS="-j5"
FEATURES="ccache"
CCACHE_SIZE="2G"
USE="threads doc javascript java imap nptl apache2"
|
For a server and for games/workstation consider this:
Quote: |
#Bloddy conservative
CFLAGS="-march=nocona -O2 -pipe"
#Using architecture specific flags
-O2 -march=nocona -funroll-loops -frerun-cse-after-loop
#Break my gentoo stable avoiding any 32 bits flags
-O3 -march=nocona -funroll-loops -frerun-cse-after-loop
|
Compare the p4 with the following of my opterons:
Quote: |
lshw -C cpu
*-cpu:0
description: CPU
product: AMD Opteron(tm) Processor 244
vendor: Advanced Micro Devices [AMD]
physical id: 4
bus info: cpu@0
version: AMD Opteron(tm)
slot: CPU0-Socket 940
size: 1800MHz
capacity: 3GHz
width: 64 bits
clock: 200MHz
capabilities: fpu fpu_exception wp vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 pni syscall nx mmxext x86-64 3dnowext 3dnow
*-cpu:1
description: CPU
product: AMD Opteron(tm) Processor 244
vendor: Advanced Micro Devices [AMD]
physical id: 5
bus info: cpu@1
version: AMD Opteron(tm)
slot: CPU1-Socket 940
size: 1800MHz
capacity: 3GHz
width: 64 bits
clock: 200MHz
capabilities: fpu fpu_exception wp vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 pni syscall nx mmxext x86-64 3dnowext 3dnow
|
So I believe that in a pure x64 binaries, it is compatible to AMD counterparts, but as I said so, EVADE sse, sse2, sse3, and 3dNow and 3dnowext, use it only if you use 32 bits compatibility mode of the processor.
Cheers. _________________ "Sine ira et studio" Tacitus. (c.56-c.177 AD).
(With neither anger nor partiality). |
|
Back to top |
|
|
|