Do I understand the tree_hash() code correctly?

Question

    memcpy(ints, hashes, (2 * cnt - count) * HASH_SIZE);

    for (i = 2 * cnt - count, j = 2 * cnt - count; j < cnt; i += 2, ++j) {
      cn_fast_hash(hashes[i], 64, ints + j * HASH_SIZE);
    }

Isn't there a buffer overrun? i is going over of size of hashes[].

For example - count=53, cnt=32. At j=26, i=52 buffer overrun.

Answer 1

Isn't there a buffer overrun? i is going over of size of hashes[].

No, it wont and isn't.

At j=26, i=52 buffer overrun.

No. When j=26, i is 41, so there's no buffer overrun.