Problem with 2.5.38 recognizing all CPUs ?

Problem with 2.5.38 recognizing all CPUs ?

[Mihai Burcea]

Hi, 

I have the following problem: I have a PE6600 Xeon machine, 8 CPUs with
the HyperThreading tech. Running a Linux RedHat 7.3,  2.4.18-4smp kernel
on it, makes it recognize that there are 8 CPUs in the system. 
After I have installed 2.5.38, it only notices there are 4 CPUs in the
system. 

Now the 2.5.x kernel may indeed report 4 CPUs because it's really really
smart and knows it all about the HT technology, but it should still at
least give me access to all 8 of the CPUs, I'd reckon. However, I doubt
that this is the case.

I would like to know whether anyone else has confronted with a similar
problem, or any feedback at all.

Thank you
Mihai Burcea

Re: Problem with 2.5.38 recognizing all CPUs ?

[Dave Jones]

On Tue, Oct 01, 2002 at 02:18:49PM -0400, Mihai Burcea wrote:
 > 
 > Hi, 
 > 
 > I have the following problem: I have a PE6600 Xeon machine, 8 CPUs with
 > the HyperThreading tech. Running a Linux RedHat 7.3,  2.4.18-4smp kernel
 > on it, makes it recognize that there are 8 CPUs in the system. 
 > After I have installed 2.5.38, it only notices there are 4 CPUs in the
 > system. 
 > 
 > Now the 2.5.x kernel may indeed report 4 CPUs because it's really really
 > smart and knows it all about the HT technology, but it should still at
 > least give me access to all 8 of the CPUs, I'd reckon. However, I doubt
 > that this is the case.
 > 
 > I would like to know whether anyone else has confronted with a similar
 > problem, or any feedback at all.

Indeed, the behaviour should be no different (as long as your
2.5 kernel was also built with CONFIG_SMP=y)
Can you post a dmesg output from both kernels ?

        Dave

-- 
| Dave Jones.        http://www.codemonkey.org.uk

Re: Problem with 2.5.38 recognizing all CPUs ?

[Mihai Burcea]

[-- Attachment #1: Type: TEXT/PLAIN, Size: 2017 bytes --]

Hi, 

I apologize for posting the dmesg's so late. I couldn't get it running
till now.
I have attached the dmesg's from 2.4.18 and from 2.5.38...one will
probably notice that in the beginning of the 2.5.38 one, there is a
warning saying "Sibling not found" for each of the 4 main CPUs...

It may also be of relevance that when I had recompiled the 2.4.18 kernel,
the default CPU type in menuconfig was Pentium III / Celeron / ..., while
in 2.5.38, it was Pentium II / ...
I think the first time I had tried to compile the 2.5.38 kernel I left it
with the default setting (i.e., Pentium II). 
The currently running kernel, the one I have attached the dmesg for, is
compiled with CPU type Pentium III / ... (as in the 2.4 kernel).

Thank you
mihai

On Tue, 1 Oct 2002, Dave Jones wrote:

> On Tue, Oct 01, 2002 at 02:18:49PM -0400, Mihai Burcea wrote:
>  > 
>  > Hi, 
>  > 
>  > I have the following problem: I have a PE6600 Xeon machine, 8 CPUs with
>  > the HyperThreading tech. Running a Linux RedHat 7.3,  2.4.18-4smp kernel
>  > on it, makes it recognize that there are 8 CPUs in the system. 
>  > After I have installed 2.5.38, it only notices there are 4 CPUs in the
>  > system. 
>  > 
>  > Now the 2.5.x kernel may indeed report 4 CPUs because it's really really
>  > smart and knows it all about the HT technology, but it should still at
>  > least give me access to all 8 of the CPUs, I'd reckon. However, I doubt
>  > that this is the case.
>  > 
>  > I would like to know whether anyone else has confronted with a similar
>  > problem, or any feedback at all.
> 
> Indeed, the behaviour should be no different (as long as your
> 2.5 kernel was also built with CONFIG_SMP=y)
> Can you post a dmesg output from both kernels ?
> 
>         Dave
> 
> -- 
> | Dave Jones.        http://www.codemonkey.org.uk
> -
> To unsubscribe from this list: send the line "unsubscribe linux-smp" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 

[-- Attachment #2: Type: APPLICATION/octet-stream, Size: 15264 bytes --]

[-- Attachment #3: Type: APPLICATION/octet-stream, Size: 15270 bytes --]