Re: Spinlocks, yet again: analysis and proposed patches

Tom, et al.,

Updated, with full recompiles between everything and the new
modification:

N, runtime:
Tip:                    1 31s   2 37s   4 86s   8 159s
no-cmpb:                1 32s   2 43s   4 83s   8 168s
spin:                   1 32s   2 51s   4 84s   8 160s
spin+mod:               1 32s   2 51s   4 89s   8 158s
spin+no-cmpb:           1 32s   2 51s   4 87s   8 163s
spin+mod+no-cmpb:       1 32s   2 50s   4 86s   8 161s

Unfortunately, the results don't seem to be terribly consistent between
runs anyway:

Run 2:
Tip:                    1 32s   2 43s   4 87s   8 160s
no-cmpb:                1 31s   2 47s   4 83s   8 167s
spin:                   1 32s   2 52s   4 88s   8 154s
spin+no-cmpb:           1 32s   2 51s   4 102s  8 166s
spin+mod:               1 32s   2 53s   4 85s   8 154s
spin+mod+no-cmpb:       1 32s   2 51s   4 91s   8 161s

Hope it helps,

	Stephen