I've placed a benchmark program at

ftp://ftp.netbsd.org/pub/NetBSD/misc/gmcgarry/lockprof.tar.gz

to test the performance of different userlevel locking schemes.
It compares RAS locks and CPU locks (inlined and not).  All CPUs
(except SH5) should now support CPU locks so the benchmark should
compile OK for all machines.  A recent -current should be OK.
RAS kernel support for your CPU isn't required.

I'd appreciate receiving the timings on different CPUs.  I guess
the results are somewhat academic since there is marginally any
difference between the locking schemes.  But numbers are cool.

The assembler routines will be the ones used in the pthread library
if it is decided that RAS locks should be used by default on
uniprocessor machines.  I guess the jury is still out on that one.

	-- Gregory McGarry <g.mcgarry@ieee.org>