Implement cache index randomization for large allocations.

Extract szad size quantization into {extent,run}_quantize(), and .
quantize szad run sizes to the union of valid small region run sizes and
large run sizes.

Refactor iteration in arena_run_first_fit() to use
run_quantize{,_first,_next(), and add support for padded large runs.

For large allocations that have no specified alignment constraints,
compute a pseudo-random offset from the beginning of the first backing
page that is a multiple of the cache line size.  Under typical
configurations with 4-KiB pages and 64-byte cache lines this results in
a uniform distribution among 64 page boundary offsets.

Add the --disable-cache-oblivious option, primarily intended for
performance testing.

This resolves #13.

include/jemalloc/internal/prng.h

@@ -26,22 +26,22 @@
  *   const uint32_t a, c : See above discussion.
  */
 #define	prng32(r, lg_range, state, a, c) do {				\
-	assert(lg_range > 0);						\
-	assert(lg_range <= 32);						\
+	assert((lg_range) > 0);						\
+	assert((lg_range) <= 32);					\
 									\
 	r = (state * (a)) + (c);					\
 	state = r;							\
-	r >>= (32 - lg_range);						\
+	r >>= (32 - (lg_range));					\
 } while (false)
 
 /* Same as prng32(), but 64 bits of pseudo-randomness, using uint64_t. */
 #define	prng64(r, lg_range, state, a, c) do {				\
-	assert(lg_range > 0);						\
-	assert(lg_range <= 64);						\
+	assert((lg_range) > 0);						\
+	assert((lg_range) <= 64);					\
 									\
 	r = (state * (a)) + (c);					\
 	state = r;							\
-	r >>= (64 - lg_range);						\
+	r >>= (64 - (lg_range));					\
 } while (false)
 
 #endif /* JEMALLOC_H_TYPES */