ecmult: Document and test ng=NULL in ecmult #1834

1. 
   real-or-random commented at 4:49 pm on March 3, 2026: contributor
2. real-or-random force-pushed on Mar 3, 2026
3. real-or-random marked this as ready for review on Mar 3, 2026
4. real-or-random added the label tweak/refactor on Mar 3, 2026
5. real-or-random added the label assurance on Mar 3, 2026
6. theStack approved
7. 
   theStack commented at 6:26 pm on March 10, 2026: contributor

   ACK 87160614f6fff769226ba2c781e325b46c7c55ec

   I was curious on whether there would be a speedup noticeable in _pubkey_tweak_mul with NULL used instead of secp256k1_scalar_zero (second commit), but it seems there isn’t on my machine, according to a hacked together benchmark https://github.com/theStack/secp256k1/commit/b9d523b31c733758035658d5debdbbd98c2d505e. Still good to document and test for this.

8. ecmult: Document and test ng=NULL in ecmult 7e68c0c88b
9. eckey: Call ecmult with NULL instead of zero scalar 3a403639dc
10. in src/tests.c:5491 in 87160614f6

    5488@@ -5481,23 +5489,24 @@ static int test_ecmult_accumulate_cb(secp256k1_scalar* sc, secp256k1_ge* pt, siz
    5489 
    5490 static void test_ecmult_accumulate(secp256k1_sha256* acc, const secp256k1_scalar* x, secp256k1_scratch* scratch) {
    5491     /* Compute x*G in 6 different ways, serialize it uncompressed, and feed it into acc. */
    

    ---

    ---

    

    real-or-random commented at 6:45 am on March 11, 2026:

    0    /* Compute x*G in 7 different ways, serialize it uncompressed, and feed it into acc. */
    

    ---

    

    real-or-random commented at 10:10 am on March 11, 2026:

    fixed
11. real-or-random force-pushed on Mar 11, 2026
12. theStack approved
13. 
    theStack commented at 1:26 pm on March 11, 2026: contributor

    re-ACK 3a403639dc07e39aa6dc48fbcecfe3cb77f09770
14. real-or-random merged this on Mar 11, 2026
15. real-or-random closed this on Mar 11, 2026

    ---

Contributors
real-or-random theStack

Labels
assurance tweak/refactor