math/bits: faster TrailingZeroes8

For sizes > 8, the existing code is faster.

benchmark                     old ns/op     new ns/op     delta
BenchmarkTrailingZeros8-8     1.95          1.29          -33.85%

Measured on 2.3 GHz Intel Core i7 running macOS 10.12.3.

Change-Id: I6f3a33ec633a2c544ec29693c141f2f99335c745
Reviewed-on: https://go-review.googlesource.com/37581
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

src/math/bits/bits.go

@@ -34,7 +34,7 @@ func LeadingZeros64(x uint64) int { return 64 - blen(uint64(x)) }
 func TrailingZeros(x uint) int { return ntz(x) }
 
 // TrailingZeros8 returns the number of trailing zero bits in x; the result is 8 for x == 0.
-func TrailingZeros8(x uint8) int { return ntz8(x) }
+func TrailingZeros8(x uint8) int { return int(ntz8tab[x]) }
 
 // TrailingZeros16 returns the number of trailing zero bits in x; the result is 16 for x == 0.
 func TrailingZeros16(x uint16) int { return ntz16(x) }

src/math/bits/bits_impl.go

@@ -23,14 +23,6 @@ var deBruijn32tab = [32]byte{
 	31, 27, 13, 23, 21, 19, 16, 7, 26, 12, 18, 6, 11, 5, 10, 9,
 }
 
-func ntz8(x uint8) (n int) {
-	if x == 0 {
-		return 8
-	}
-	// see comment in ntz64
-	return int(deBruijn32tab[uint32(x&-x)*deBruijn32>>(32-5)])
-}
-
 func ntz16(x uint16) (n int) {
 	if x == 0 {
 		return 16