[gentoo-user] kernel config for eee w/Atom N270 CPU

[gentoo-user] kernel config for eee w/Atom N270 CPU

[Maxim Wexler]

Hi group,

Been tracking down other's kern config for 900A w/ N270 Atom cpu and
notice, so far, everyone goes for CONFIG_SMP=y. Why, particularly when
most commenters stress keeping the kernel as slim as possible?

In the <help> window we have 'If you have a system with only one CPU,
like most personal computers, say N.' As far as I know the 900A has
only one CPU, unless there is some sort of virtual one I haven't heard
of yet.

Maxim

Re: [gentoo-user] kernel config for eee w/Atom N270 CPU

[Paul Hartman]

On Thu, Jun 11, 2009 at 1:14 PM, Maxim Wexler<maxim.wexler@gmail.com> wrote:
> Hi group,
>
> Been tracking down other's kern config for 900A w/ N270 Atom cpu and
> notice, so far, everyone goes for CONFIG_SMP=y. Why, particularly when
> most commenters stress keeping the kernel as slim as possible?
>
> In the <help> window we have 'If you have a system with only one CPU,
> like most personal computers, say N.' As far as I know the 900A has
> only one CPU, unless there is some sort of virtual one I haven't heard
> of yet.

Does the CPU have multiple cores? If so, that's probably why.

Re: [gentoo-user] kernel config for eee w/Atom N270 CPU

[Maxim Wexler]

On 6/11/09, Paul Hartman <paul.hartman+gentoo@gmail.com> wrote:
> On Thu, Jun 11, 2009 at 1:14 PM, Maxim Wexler<maxim.wexler@gmail.com> wrote:
>> Hi group,
>>
>> Been tracking down other's kern config for 900A w/ N270 Atom cpu and
>> notice, so far, everyone goes for CONFIG_SMP=y. Why, particularly when
>> most commenters stress keeping the kernel as slim as possible?
>>
>> In the <help> window we have 'If you have a system with only one CPU,
>> like most personal computers, say N.' As far as I know the 900A has
>> only one CPU, unless there is some sort of virtual one I haven't heard
>> of yet.
>
> Does the CPU have multiple cores? If so, that's probably why.
>

hmm, when I %cat /proc/cpuinfo I get processor :0 at the top of a list
of stuff then, just below that, processor :1 at the top of an
identical list.

Are those two cores? I think CONFIG_MCORE2=y is recommended for the
Atom. Does MCORE2 mean 'two cores' or 'core no.2'? The <help>
window(under Symmetric multi-processor support) says nothing about
'cores'  Are 'cores' and 'cpus' the same thing in this context?

Well, this has taken me all morning and half the afternoon and I
haven't even started 'Power management...' options yet. I'll just say
'yes' to SMP and carry on. Hopefully have the definitive answer  by
midnight and I can make the darn kernel.

mw

mw

Re: [gentoo-user] kernel config for eee w/Atom N270 CPU

[Paul Hartman]

On Thu, Jun 11, 2009 at 2:56 PM, Maxim Wexler<maxim.wexler@gmail.com> wrote:
> On 6/11/09, Paul Hartman <paul.hartman+gentoo@gmail.com> wrote:
>> On Thu, Jun 11, 2009 at 1:14 PM, Maxim Wexler<maxim.wexler@gmail.com> wrote:
>>> Hi group,
>>>
>>> Been tracking down other's kern config for 900A w/ N270 Atom cpu and
>>> notice, so far, everyone goes for CONFIG_SMP=y. Why, particularly when
>>> most commenters stress keeping the kernel as slim as possible?
>>>
>>> In the <help> window we have 'If you have a system with only one CPU,
>>> like most personal computers, say N.' As far as I know the 900A has
>>> only one CPU, unless there is some sort of virtual one I haven't heard
>>> of yet.
>>
>> Does the CPU have multiple cores? If so, that's probably why.
>>
>
> hmm, when I %cat /proc/cpuinfo I get processor :0 at the top of a list
> of stuff then, just below that, processor :1 at the top of an
> identical list.
>
> Are those two cores? I think CONFIG_MCORE2=y is recommended for the
> Atom. Does MCORE2 mean 'two cores' or 'core no.2'? The <help>
> window(under Symmetric multi-processor support) says nothing about
> 'cores'  Are 'cores' and 'cpus' the same thing in this context?
>
> Well, this has taken me all morning and half the afternoon and I
> haven't even started 'Power management...' options yet. I'll just say
> 'yes' to SMP and carry on. Hopefully have the definitive answer  by
> midnight and I can make the darn kernel.

That is correct, SMP is required to use the multi-core option.
Multi-Cores and SMP are not the same thing, but are "basically" the
same thing, so the same code handles both. It appears you do have a
dual-core CPU, so you should use SMP and the multi-core option in your
kernel config. Otherwise you'll only be using half of your processor's
ability.

Re: [gentoo-user] kernel config for eee w/Atom N270 CPU

[Florian Philipp]

[-- Attachment #1: Type: text/plain, Size: 1691 bytes --]

Maxim Wexler schrieb:
> On 6/11/09, Paul Hartman <paul.hartman+gentoo@gmail.com> wrote:
>> On Thu, Jun 11, 2009 at 1:14 PM, Maxim Wexler<maxim.wexler@gmail.com> wrote:
>>> Hi group,
>>>
>>> Been tracking down other's kern config for 900A w/ N270 Atom cpu and
>>> notice, so far, everyone goes for CONFIG_SMP=y. Why, particularly when
>>> most commenters stress keeping the kernel as slim as possible?
>>>
>>> In the <help> window we have 'If you have a system with only one CPU,
>>> like most personal computers, say N.' As far as I know the 900A has
>>> only one CPU, unless there is some sort of virtual one I haven't heard
>>> of yet.
>> Does the CPU have multiple cores? If so, that's probably why.
>>
> 
> hmm, when I %cat /proc/cpuinfo I get processor :0 at the top of a list
> of stuff then, just below that, processor :1 at the top of an
> identical list.
> 
> Are those two cores? I think CONFIG_MCORE2=y is recommended for the
> Atom. Does MCORE2 mean 'two cores' or 'core no.2'? The <help>
> window(under Symmetric multi-processor support) says nothing about
> 'cores'  Are 'cores' and 'cpus' the same thing in this context?
> 
> Well, this has taken me all morning and half the afternoon and I
> haven't even started 'Power management...' options yet. I'll just say
> 'yes' to SMP and carry on. Hopefully have the definitive answer  by
> midnight and I can make the darn kernel.
> 
> mw
> 
> mw
> 

The N270 (as well as any other 2xx Atom) has one core. However, it
supports HyperThreading which produces that second entry in /proc/cpuinfo.

For this, you need SMP as well as SMT (an option which you will only see
when SMP is enabled).


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 261 bytes --]

Re: [gentoo-user] kernel config for eee w/Atom N270 CPU

[Kelly Hirai]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

the N270 is a single core with hyperthreading, which will apear as 2
cpus (with the same core id) in dmesg.

k.


Paul Hartman wrote:
> On Thu, Jun 11, 2009 at 1:14 PM, Maxim Wexler<maxim.wexler@gmail.com> wrote:
>> Hi group,
>>
>> Been tracking down other's kern config for 900A w/ N270 Atom cpu and
>> notice, so far, everyone goes for CONFIG_SMP=y. Why, particularly when
>> most commenters stress keeping the kernel as slim as possible?
>>
>> In the <help> window we have 'If you have a system with only one CPU,
>> like most personal computers, say N.' As far as I know the 900A has
>> only one CPU, unless there is some sort of virtual one I haven't heard
>> of yet.
> 
> Does the CPU have multiple cores? If so, that's probably why.
> 

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.11 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iEYEARECAAYFAkoxY/IACgkQaI4xTzulgR/UogCcDhBPtPP53Jsmm0zz+B0L0g7l
LzYAoMtXWYfA60RpdFAOQHSJv9D5N1fV
=/RR8
-----END PGP SIGNATURE-----

Re: [gentoo-user] kernel config for eee w/Atom N270 CPU

[Paul Hartman]

On Thu, Jun 11, 2009 at 3:07 PM, Kelly Hirai<kelly@met.fsu.edu> wrote:
> the N270 is a single core with hyperthreading, which will apear as 2
> cpus (with the same core id) in dmesg.

Ah, I forgot about hyperthreading masquerading as multiple CPUs. In
that case, Maxim can safely disable SMP if he wants to. I don't know
if the theoretical speed gains of disabling SMP outweigh the
theoretical speed gains of enabling hyperthreading. I think it'll
probably be about the same either way.

Re: [gentoo-user] kernel config for eee w/Atom N270 CPU

[Florian Philipp]

[-- Attachment #1: Type: text/plain, Size: 1474 bytes --]

Paul Hartman schrieb:
> On Thu, Jun 11, 2009 at 3:07 PM, Kelly Hirai<kelly@met.fsu.edu> wrote:
>> the N270 is a single core with hyperthreading, which will apear as 2
>> cpus (with the same core id) in dmesg.
> 
> Ah, I forgot about hyperthreading masquerading as multiple CPUs. In
> that case, Maxim can safely disable SMP if he wants to. I don't know
> if the theoretical speed gains of disabling SMP outweigh the
> theoretical speed gains of enabling hyperthreading. I think it'll
> probably be about the same either way.
> 

Well, I don't know about real workloads but once I did a little
benchmark: One versus two instances of `dd < /dev/zero | md5sum`.

Two instances had a 30% higher throughput than one. I haven't tried it
with disabled SMP but I really can't imagine that the extra scheduling
would cost nearly enough to compensate for this.

From a technical point of view, I think HT makes more sense for an Atom
than for a Nehalem: The Atom has only one pipeline, no out-of-order
execution and probably a less effective branch prediction. HT might
compensate this.

However, I'm wondering if it wouldn't have been better to implement
out-of-order execution instead of HT (like VIA Nano, for example). Maybe
HT doesn't need as many transistors as out-of-order execution?

In the end, unless there is some hard evidence against the use of HT,
I'd say: They've spend their transistor budget on HT, now we should use
what we've got.


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 261 bytes --]