Re: [PATCH] mm/vmalloc: don't use vmalloc_end

[Mike Travis <travis@sgi.com>]

Christoph Lameter wrote:
> On Wed, 9 Dec 2009, Mike Travis wrote:
> 
>>> Tony: Can you confirm that the new percpu stuff works on IA64? (Or is
>>> there nobody left to care?)
>> Christoph,  I have access to a 640p system for a couple more weeks if
>> there's anything you'd like me to check out.
> 
> Boot with 2.6.32 and see if the per cpu allocator works. Check if there
> are any changes to memory consumption. Create a few thousand virtual
> ethernet devices and see if the system keels over.

Any advice on how to go about the above would be helpful... ;-)

> 
> It may also be good to run some scheduler test. Compare AIM9 of latest
> SLES with 2.6.32. Concurrent page fault test? Then a performance test with
> lots of concurrency but the usual stuff wont work since HPC apps usually
> pin.

I'm doing some aim7/9 comparisons right now between SPARSE and DISCONTIG
memory configs using sles11 + 2.6.32.  Which other benchmarks would you
recommend for the other tests?

> 
> Run latencytest (available in the lldiag package) from
> kernel.org/pub/linux/kernel/people/christoph/lldiag and see how the
> disturbances by the OS are changed.

I'll put that on the list.

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>