Add another kmem test to check for lock contention in the slab
allocator. I have serious contention issues here and I needed
a way to easily measure how much the following batch of changes
will improve things. Currently things are quite bad when the
allocator is highly contended, and interestingly it seems to
get worse in a non-linear fashion... I'm not sure why yet.
I'll figure it out tomorrow.
kmem:kmem_lock Pass
kmem_lock: time (sec) slabs objs
kmem_lock: tot/max/calc tot/max/calc
kmem_lock: 0.
061000000 75/60/64 2400/1894/2048
kmem_lock: 0.
157000000 134/125/128 4288/3974/4096
kmem_lock: 0.
471000000 263/249/256 8416/7962/8192
kmem_lock: 2.
526000000 518/499/512 16576/15957/16384
kmem_lock: 14.
393000000 990/978/1024 31680/31270/32768
git-svn-id: https://outreach.scidac.gov/svn/spl/trunk@134
7e1ea52c-4ff2-0310-8f11-
9dd32ca42a1c