On Mon, 28 Aug 2006 Ara.T.Howard / noaa.gov wrote:

> with those mods my results are largely the same: mutex shows no leak at all
> and sync show a gradual trend up.  with respect to sync, however, the trend
> is from a small amount of initial memory and it may well be the case that it
> eventually gets cleared so i am not asserting a leak in sync at this point -
> i'm doing a long run to see which is the case.

after a few hours i've finally seen the sync code cycle down after growing
slowly for a looooong time, so it does not appear to have a leak either.

except from output:

---
iter: 449
threads loaded in 23.4420.
memory_usage : pct_mem: 44.0%, vsz: 453944b, rss: 452456b
n_threads_in_object_space : 421
threads killed.
memory_usage : pct_mem: 22.3%, vsz: 230740b, rss: 229380b
n_threads_in_object_space : 1
---
iter: 450
threads loaded in 23.2819.
memory_usage : pct_mem: 44.0%, vsz: 453776b, rss: 452372b
n_threads_in_object_space : 421
threads killed.
memory_usage : pct_mem: 22.3%, vsz: 230740b, rss: 229380b
n_threads_in_object_space : 1
---
iter: 451
threads loaded in 23.4827.
memory_usage : pct_mem: 44.0%, vsz: 453776b, rss: 452416b
n_threads_in_object_space : 421
threads killed.
memory_usage : pct_mem: 8.0%, vsz: 83740b, rss: 82380b
n_threads_in_object_space : 1
---
iter: 452
threads loaded in 23.3958.
memory_usage : pct_mem: 44.1%, vsz: 454336b, rss: 452932b
n_threads_in_object_space : 421
threads killed.
memory_usage : pct_mem: 8.5%, vsz: 89336b, rss: 87976b
n_threads_in_object_space : 1
---
iter: 453
threads loaded in 23.4515.
memory_usage : pct_mem: 44.0%, vsz: 454300b, rss: 452856b
n_threads_in_object_space : 421
threads killed.
memory_usage : pct_mem: 9.4%, vsz: 98472b, rss: 97116b
n_threads_in_object_space : 1


cheers.


-a
-- 
to foster inner awareness, introspection, and reasoning is more efficient than
meditation and prayer.
- h.h. the 14th dalai lama