performance review

[jamiealquiza]

This is an in-memory cache; every cycle counts.

Also see: #20, #17

[jamiealquiza]

PromoteEvict / sll.Remove has a > 500x speedup w/ 25K nodes here: 49dd2b1

[jamiealquiza]

^ Had bugs, rolled back. Pending reintroduction in #7.

[jamiealquiza]

Exchanged sll's mutex-guarded counter with an atomic CAS here 817289f

[jamiealquiza]

Switched the removeFromScores search algo to an unrolled linear search here 0dec799

[jamiealquiza]

Sll can pre-allocate an internal nodeScoreList with specific size (e.g. the capacity of a cache implementing sll). This improved initial insert times for 100k items by ~50%. a9e9743

[jamiealquiza]

Sll refactor reduces instructions per op a8ebb45

[jamiealquiza]

Removed deferred mutex unlocks 52c1f3b. Get perf seems to have improved; needs comparative testing.

[jamiealquiza]

#22 and #11 were closed.

[jamiealquiza]

1,500x improvement in PromoteEvict 4309170...5aa26f2

[jamiealquiza]

Tail eviction speed up c72347b

[jamiealquiza]

Sll node removal avoids mem copies ca2287d

[jamiealquiza]

Faster FNV-1 32-bit 72f8824