Xen devel - Mar 2010 - [PATCH]x86: Increase the default NR_CPUS to 128.

x86: Increase the default NR_CPUS to 128.

We have newer systems which have more than 64 CPUs, and users
often complain some cpus can''t be waken up when play with Xen.
Certainly, MAX_PHYS_CPUS option also can support more CPUs, but
it is still inconvenient for them, so change the default value to 128. 

Signed-off-by: Xiantao Zhang <xiantao.zhang@intel.com>

diff -r 1afa8d7bb9ad xen/include/asm-x86/config.h
--- a/xen/include/asm-x86/config.h	Mon Mar 08 21:34:43 2010 +0000
+++ b/xen/include/asm-x86/config.h	Tue Mar 09 09:41:37 2010 +0800
@@ -49,7 +49,7 @@
 #ifdef MAX_PHYS_CPUS
 #define NR_CPUS MAX_PHYS_CPUS
 #else
-#define NR_CPUS 64
+#define NR_CPUS 128
 #endif
 
 #ifdef __i386__

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

>>> "Zhang, Xiantao" <xiantao.zhang@intel.com> 09.03.10
02:51 >>>
>x86: Increase the default NR_CPUS to 128.
While I realize my reply is a little late, I''d still like to raise the
question
about the tools side not coping with the new default - I would think
that any default setting ought to imply that (within the scope of a
project) the whole software stack should still work. At the very
minimum CPU affinity changes (default settings on NUMA as well as
manual ones through xm) do not presently work for CPUs beyond 64.

Jan


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

On 10/03/2010 13:29, "Jan Beulich" <JBeulich@novell.com> wrote:
>>>> "Zhang, Xiantao" <xiantao.zhang@intel.com>
09.03.10 02:51 >>>
>> x86: Increase the default NR_CPUS to 128.
> 
> While I realize my reply is a little late, I''d still like to raise
the
> question
> about the tools side not coping with the new default - I would think
> that any default setting ought to imply that (within the scope of a
> project) the whole software stack should still work. At the very
> minimum CPU affinity changes (default settings on NUMA as well as
> manual ones through xm) do not presently work for CPUs beyond 64.
It would certainly be nice, but at least we can boot all CPUs and run VMs
across them, minus the finer-grained control. Is this halfway default
solution for these bigger machines worse than not making an effort at all?
At the end of the day packagers can make up their own minds.

 -- Keir



_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

>>> Keir Fraser <keir.fraser@eu.citrix.com> 10.03.10 15:21
>>>
>It would certainly be nice, but at least we can boot all CPUs and run VMs
>across them, minus the finer-grained control. Is this halfway default
>solution for these bigger machines worse than not making an effort at all?
Not sure - with NUMA also defaulting to on, guest creation may fail
because of the failure to enforce the affinity. Not nice, especially
since (I think; a general problem of the tools) the error indication is
rather hard to interpret by someone not knowing the source code.
>At the end of the day packagers can make up their own minds.
Certainly.

Jan


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

On 10/03/2010 14:38, "Jan Beulich" <JBeulich@novell.com> wrote:
>>>> Keir Fraser <keir.fraser@eu.citrix.com> 10.03.10 15:21
>>>
>> It would certainly be nice, but at least we can boot all CPUs and run
VMs
>> across them, minus the finer-grained control. Is this halfway default
>> solution for these bigger machines worse than not making an effort at
all?
> 
> Not sure - with NUMA also defaulting to on, guest creation may fail
> because of the failure to enforce the affinity. Not nice, especially
> since (I think; a general problem of the tools) the error indication is
> rather hard to interpret by someone not knowing the source code.
Hm, I didn''t really think of that...

 -- Keir



_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel