hi-
    using the posted dhry.c and my sun-4/370 running NetBSD/sparc
i got the following results:

CASE 1:
    - single user
    - cache disabled (i.e. i patched cpu.c to not call cache_enable()):

	# time dhry ; time dhry_static
	total time was 203.671630
	null time was 1.260815
	page faults (0), swaps (0), context switches (2066)
	number of dhrystones per sec (4940.45)
	205.0u 1.6s 3:28.75 98.9% 0+0k 0+0io 0pf+0w
	total time was 200.842977
	null time was 1.249775
	page faults (0), swaps (0), context switches (2244)
	number of dhrystones per sec (5010.19)
	201.8u 1.3s 3:25.60 98.7% 0+0k 1+0io 0pf+0w

CASE 2:
    - single user
    - cache enabled
	# time dhry ; time dhry_static
	total time was 27.349044
	null time was 0.121112
	page faults (0), swaps (0), context switches (273)
	number of dhrystones per sec (36727)
	27.5u 0.1s 0:27.64 99.8% 0+0k 3+0io 0pf+0w
	total time was 26.972541
	null time was 0.125035
	page faults (0), swaps (0), context switches (271)
	number of dhrystones per sec (37247.4)
	27.0u 0.0s 0:27.26 99.5% 0+0k 6+0io 0pf+0w

CASE 3:
    - multiuser (only me logged in)
    - cache enabled
	# time dhry ; time dhry_static
	total time was 41.548356
	null time was 0.121107
	page faults (0), swaps (0), context switches (465)
	number of dhrystones per sec (24138.7)
	41.8u 0.1s 0:42.14 99.5% 0+0k 0+0io 0pf+0w
	total time was 27.24508
	null time was 0.121124
	page faults (0), swaps (0), context switches (304)
	number of dhrystones per sec (37170)
	27.1u 0.0s 0:27.44 99.1% 0+0k 1+0io 0pf+0w


the machine i used has 32MB ram and a cache like this:
cpu0 at mainbus0: SUN-4/300 series (L64812 or ACT8847 FPU)
cpu0: cache chip bug; trap page uncached
cpu0: 131072 byte write-back, 16 bytes/line, sw flush cache enabled

chuck