View previous topic :: View next topic |
Author |
Message |
gauntalus Tux's lil' helper
Joined: 23 Dec 2004 Posts: 125 Location: University of Michigan
|
Posted: Tue Aug 01, 2006 3:44 am Post subject: Make flags for Intel Core 2 [SOLVED] |
|
|
I'm actually booted from my liveCD right now, I was wondering if someone could suggest some good make flags for an Intel Core 2 machine. The actual make is a Dell e1505, but all I'm really concerned with are settings I should select to get the most out of the core 2.
Suggestions? _________________ Mike Cook Online
-- Brutal Chess
Last edited by gauntalus on Tue Aug 01, 2006 7:06 am; edited 1 time in total |
|
Back to top |
|
|
thorpe l33t
Joined: 09 May 2005 Posts: 618 Location: Sydney, Australia.
|
|
Back to top |
|
|
gauntalus Tux's lil' helper
Joined: 23 Dec 2004 Posts: 125 Location: University of Michigan
|
|
Back to top |
|
|
stupidkid Apprentice
Joined: 17 Apr 2006 Posts: 247 Location: 127.0.0.1
|
Posted: Tue Aug 01, 2006 6:14 pm Post subject: |
|
|
Are you sure that's the right one? Core 2 Duo is Conroe. Core Duo (for laptops) is actually Yonah. |
|
Back to top |
|
|
vonr Guru
Joined: 23 Mar 2006 Posts: 300
|
Posted: Wed Aug 02, 2006 11:14 am Post subject: |
|
|
According to this post, Core 2 Duo is actually -march=nocona. However, the E1505 has a 'regular' Core Duo I think. |
|
Back to top |
|
|
slycordinator Advocate
Joined: 31 Jan 2004 Posts: 3065 Location: Korea
|
Posted: Thu Aug 03, 2006 7:44 am Post subject: |
|
|
vonr wrote: | However, the E1505 has a 'regular' Core Duo I think. |
Correct.
From dell's info page on the E1505:
Quote: | Dual-Core Performance
Do more with optional dual-core power. The new Intel® CoreTM Duo mobile processors leverage two execution cores on a single chip to help increase your PC's ability to maintain maximum performance while running multiple applications such as background virus scans. No more waiting, no more hiccups - that's the dual-core advantage. |
|
|
Back to top |
|
|
gauntalus Tux's lil' helper
Joined: 23 Dec 2004 Posts: 125 Location: University of Michigan
|
Posted: Fri Aug 04, 2006 2:30 am Post subject: |
|
|
i actually ended up using -march=prescott. Should I have used something different? _________________ Mike Cook Online
-- Brutal Chess |
|
Back to top |
|
|
stupidkid Apprentice
Joined: 17 Apr 2006 Posts: 247 Location: 127.0.0.1
|
Posted: Fri Aug 04, 2006 2:40 am Post subject: |
|
|
No, that's right. _________________ How many people can read hex if you and dead people can read hex? |
|
Back to top |
|
|
gauntalus Tux's lil' helper
Joined: 23 Dec 2004 Posts: 125 Location: University of Michigan
|
Posted: Fri Aug 04, 2006 2:45 am Post subject: |
|
|
And I had a lot of trouble finding which "Processor Family" I was supposed to choose in menuconfig. I chose pentium-m, again, is that right or should I have chosen pentium 4? _________________ Mike Cook Online
-- Brutal Chess |
|
Back to top |
|
|
mamac l33t
Joined: 29 Feb 2004 Posts: 890
|
Posted: Sat Aug 05, 2006 8:29 am Post subject: |
|
|
Hi,
What about MAKEOPTS, did you use "-j2" or "-j3" ? _________________ Powered by Gentoo Linux since 2003 |
|
Back to top |
|
|
slycordinator Advocate
Joined: 31 Jan 2004 Posts: 3065 Location: Korea
|
Posted: Sat Aug 05, 2006 9:17 pm Post subject: |
|
|
Since it's dual-core you'd use MAKEOPTS="-j3"
The "general rule" is to have it be MAKEOPTS="jX+1" where X is the number of processors. But with dual-core there are really 2 processors in one.
Though you'd obviously have to have support for SMP in your kernel and on new kernels specify the subsetting specifically for dual-core. And you'd want to make sure that when you do "cat /proc/cpuinfo" that it displays 2 CPUs that the computer sees. |
|
Back to top |
|
|
Dio-Sama n00b
Joined: 18 Mar 2005 Posts: 19
|
Posted: Sun Aug 06, 2006 11:46 am Post subject: |
|
|
Hi,
sorry for posting in a thread that's already been tagged solved, but since my question is quite similar to this one i didn't feel it's right opening a new thread just for this. so correct me if i'm wrong, but if i want to have my core 2 duo running in 32bit, i would have to put -march=prescott, and for 64bit it should be -march=nocona, would that be right? (and if they are the correct switches, i don't understand why, since prescott and nocona are netburst cores, completely different from the core/core2 architecture. so why should i be telling gcc to optimise for a processor i don't run the system on? this is a *little* confusing...)
furthermore, i read up on the icc 9.1 optimization guide from intel, but i found nothing especially for core 2 duo systems. can anybody help me with this?
i just can't decide if i should use icc or gcc for my system.
thx,
dio. _________________ "Death is but perchance to dream." |
|
Back to top |
|
|
Gregoire Apprentice
Joined: 15 Apr 2006 Posts: 292
|
Posted: Wed Apr 18, 2007 9:32 pm Post subject: |
|
|
From http://gcc.gnu.org/gcc-4.3/changes.html :
Quote: |
IA-32/x86-64
* Tuning for Intel Core 2 processors is available via -mtune=core2 and -march=core2.
* Tuning for AMD Geode processors is available via -mtune=geode and -march=geode.
* Code generation of block move (memcpy) and block set (memset) was rewritten. GCC can now pick the best algorithm (loop, unrolled loop, instruction with rep prefix or a library call) based on the size of the block being copied and the CPU being optimized for. A new option -minline-stringops-dynamically has been added. With this option string operations of unknown size are expanded such that small blocks are copied by in-line code, while for large blocks a library call is used. This results in faster code than -minline-all-stringops when the library implementation is capable of using cache hierarchy hints. The heuristic choosing the particular algorithm can be overwritten via -mstringop-strategy. Newly also memset of values different from 0 is inlined.
* GCC no longer places the cld instruction before string operations. Both i386 and x86-64 ABI documents mandate the direction flag to be clear at the entry of a function. It is now invalid to set the flag in asm statement without reseting it afterward.
* Support for SSSE3 built-in functions and code generation are available via -mssse3.
|
But of course gcc-4.3 is still at stage 1, so I won't use it right now |
|
Back to top |
|
|
julakali n00b
Joined: 27 Mar 2006 Posts: 27
|
Posted: Thu Apr 19, 2007 7:28 pm Post subject: |
|
|
So... can i run x86 32bit with march=nocona , too?
I haven't read anything like "If you use nocona, you need a 64bit distribution..."
edit: hmm seems not... can somebody tell me the truth about that thingy, anyway? |
|
Back to top |
|
|
darf n00b
Joined: 08 Jan 2007 Posts: 27
|
Posted: Fri Apr 20, 2007 5:43 pm Post subject: |
|
|
Yes, I use nocona on 32 bit works great. |
|
Back to top |
|
|
|