mirrors/core-lightning
1
0
Fork 0
mirror of https://github.com/ElementsProject/lightning.git synced 2025-01-18 21:35:11 +01:00
Code Issues Projects Releases Wiki Activity

common/bolt12_merkle.c: fix comments.

Browse Source 
I spent an hour thinking this code had a bug (see test vector fix);
we *do* overallocate the tree, but that's deliberate: we fill with NULLs
and ignore on recursion.

The Merkle recurse comment had an out-by-one, and the NULL-pad
technique used was uncommented.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
This commit is contained in:
Rusty Russell 2021-07-29 15:59:53 +09:30
parent 207f1def93
commit c448170d5f
1 changed files with 10 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

13
common/bolt12_merkle.c
View File

@ -144,12 +144,15 @@ static const struct sha256 *merkle_recurse(const struct sha256 **base,
SUPERVERBOSE("Merkle recurse [%zu - %zu] and [%zu - %zu]\n",
arr - base, arr + len / 2 - 1 - base,
arr + len / 2 - base, arr + len - base);
arr + len / 2 - base, arr + len - 1 - base);
left = merkle_recurse(base, arr, len / 2);
right = merkle_recurse(base, arr + len / 2, len / 2);
/* left is never NULL if right is not NULL */
if (!right)
if (!right) {
SUPERVERBOSE("[%zu - %zu] is NULL!\n",
arr + len / 2 - base, arr + len - base);
return left;
}
return merkle_pair(base, left, right);
}
@ -162,7 +165,11 @@ void merkle_tlv(const struct tlv_field *fields, struct sha256 *merkle)
SUPERVERBOSE("nonce tag:");
h_lnall_ctx(&lnall_ctx, fields);
/* NULL-pad to next power of 2 */
/* We build an oversized power-of-2 symmentic tree, but with
* NULL nodes at the end. When we recurse, we pass through
* NULL. This is less efficient than calculating the
* power-of-2 split as we recurse, but simpler. */
arr = tal_arrz(NULL, struct sha256 *,
1ULL << (ilog64(tal_count(fields)) + 1));