DragonFly BSD
DragonFly users List (threaded) for 2006-09
[Date Prev][Date Next]  [Thread Prev][Thread Next]  [Date Index][Thread Index]


From: "Miguel Figueiredo Mascarenhas Sousa Filipe" <miguel.filipe@xxxxxxxxx>
Date: Sun, 24 Sep 2006 23:01:50 +0100

Hi all,

On 9/22/06, Joerg Sonnenberger <joerg@xxxxxxxxxxxxxxxxx> wrote:
 On Fri, Sep 22, 2006 at 09:38:31AM -0700, Matthew Dillon wrote:
>     As part of my parallel buildworld test I record the contention
>     count delta for each -j 8 build.  I typically get numbers in the
>     800-900 million range.  x 100 ns pertick == about 100 seconds of wasted
>     of cpu (for both cpus together) for a buildworld that takes 1450 seconds,
>     and dividing by the two cpus, that probably represents a real-world
>     loss in build time of between 4% and 7% over what it could hav achieved.
>     That is a very rough estimate, I haven't actually measured what a 'tick'
>     is... it could be less then 100ns but I don't think it is more.

Hm. On my build machine: around 3e6/s during a parallel build...


I'm not shure I understood correctly the meaning of those number, and
their impact on actual smp speedup (or lack thereof)..

How does that correlate to speedup on a parallel build ?
for instance..  build -j8:
- in UP mode > 1450s
- in SMP mode >  825s (1450/2 + 100)?

sorry, I don't have a smp system to test out...

best regards

-- Miguel Sousa Filipe

Miguel Sousa Filipe

[Date Prev][Date Next]  [Thread Prev][Thread Next]  [Date Index][Thread Index]