Add TODOs [skip ci]

Author: Raimo33

src/field_10x26_impl.h

@@ -261,6 +261,9 @@ static int secp256k1_fe_impl_normalizes_to_zero_var(const secp256k1_fe *r) {
 
 SECP256K1_INLINE static void secp256k1_fe_impl_set_int(secp256k1_fe *r, int a) {
     r->n[0] = a;
+
+    /* TODO: parallelize, SSE2 (32bit cpu only) */
+
     r->n[1] = r->n[2] = r->n[3] = r->n[4] = r->n[5] = r->n[6] = r->n[7] = r->n[8] = r->n[9] = 0;
 }
 

src/field_5x52_impl.h

@@ -224,6 +224,9 @@ SECP256K1_INLINE static void secp256k1_fe_impl_set_int(secp256k1_fe *r, int a) {
 
 SECP256K1_INLINE static int secp256k1_fe_impl_is_zero(const secp256k1_fe *a) {
     const uint64_t *t = a->n;
+
+    /* TODO: parallelize. mm_testz */
+
     return (t[0] | t[1] | t[2] | t[3] | t[4]) == 0;
 }
 

src/hash_impl.h

@@ -38,12 +38,12 @@ static void secp256k1_sha256_initialize(secp256k1_sha256 *hash) {
         0x6a09e667, 0xbb67ae85, 0x3c6ef372, 0xa54ff53a,
         0x510e527f, 0x9b05688c, 0x1f83d9ab, 0x5be0cd19
     );
-    _mm256_storeu_si256((__m256i*)hash->s, vec);
+    _mm256_storeu_si256((__m256i *)hash->s, vec);
 #elif defined(__SSE2__)
     const __m128i vec1 = _mm_setr_epi32(0x6a09e667, 0xbb67ae85, 0x3c6ef372, 0xa54ff53a); /* TODO: precompute */
     const __m128i vec2 = _mm_setr_epi32(0x510e527f, 0x9b05688c, 0x1f83d9ab, 0x5be0cd19); /* TODO: precompute */
-    _mm_storeu_si128((__m128i*)hash->s, vec1);
-    _mm_storeu_si128((__m128i*)(hash->s + 4), vec2);
+    _mm_storeu_si128((__m128i *)hash->s, vec1);
+    _mm_storeu_si128((__m128i *)(hash->s + 4), vec2);
 #else
     hash->s[0] = 0x6a09e667ul;
     hash->s[1] = 0xbb67ae85ul;

src/modinv32_impl.h

@@ -80,7 +80,7 @@ static void secp256k1_modinv32_normalize_30(secp256k1_modinv32_signed30 *r, int3
     VERIFY_CHECK(secp256k1_modinv32_mul_cmp_30(r, 9, &modinfo->modulus, 1) < 0); /* r < modulus */
 #endif
 
-    /* TODO: parallelize */
+    /* TODO: parallelize, SSE2 (32bit cpu only) */
 
     /* In a first step, add the modulus if the input is negative, and then negate if requested.
      * This brings r from range (-2*modulus,modulus) to range (-modulus,modulus). As all input