tlb shootdown

[Kostik Belousov]

On Thu, Jun 02, 2011 at 05:24:47PM +0300, Andriy Gapon wrote:
> on 02/06/2011 15:02 Ivan Voras said the following:
> > On 01/06/2011 13:11, Andriy Gapon wrote:
> >>
> >> Anyone knows of a benchmark/test that can measure/demonstrate difference in tlb
> >> shootdown performance (or its lack)?
> > 
> > The "tlb" utility from lmbench may help you.
> 
> Just because it's named tlb and I asked for tlb something? :)
> 
> tlb - TLB size and latency benchmark
> ...
> tlb tries to determine the size, in pages, of  the  TLB.   The  largest
> amount of memory it will examine is len bytes.
> ...
> Once  the  TLB  boundary is located tlb reports the TLB miss latency as
> the TLB latency for twice as many pages as the TLB can hold.
> 
> I am not this will tell anything about TLB _shootdown_ performance of SMP systems.
> Perhaps I wasn't specific enough in my original question.

Apparently, lmbench does not install a single multithreaded binary.
So, despite my expectation, it is not useful for what you have asked.

I think, the working scenario would be to have multithreaded process,
each thread of which does mmap, then accesses the mmaped pages, then
unmaps the region. You would count the number of accesses per time unit,
the more the better.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
Url : http://lists.freebsd.org/pipermail/freebsd-performance/attachments/20110602/8ed980c3/attachment.pgp