P-256/384/521.
New functions:
- sp_ecc_make_key_256_nb
- sp_ecc_make_key_384_nb
- sp_ecc_make_key_521_nb
- sp_ecc_secret_gen_256_nb
- sp_ecc_secret_gen_384_nb
- sp_ecc_secret_gen_521_nb
This commit also tweaks the testing for ECDSA testing in test.c. Instead of
testing with one of P-256/384/521, we test with all that are available.
GCM: make borrow constant time.
AES-GCM decrypt: compare at end and constant time.
Random: array_add touchs all elements every time.
RSA-OAEP: look for padding byte in constant time (look at every byte in
array).
SP
- reduce conditional use (make them bit ops)
- Fix point adds to not use double when adding infinity to infinity
- Implement signed div as __divi3 is not constant time.
- Move check sof input variables to API.
EC point add is a double if affine ordinates are the same.
Fix AVX2 implementation of mod inv to do conversion from 26 bits in
32-bit registers to 64 bits in 64-bit registers. Also handle negative
value.
RSA OAEP: loop down to last byte. Must have a separator.
Div for x86_64: fix to have all 1s when high word equals divisor.
Change Karatsuba implementations for x86_64.
Fix ECC code to better handle corner cases.
Add 'lower' versions of functions wehn an input is known to be less than m.
Add mont_add/dbl/tpl/sub for P384.
Change ECC point add to be cache-attack resistant.
Change mod_exp to be cache-attack resistant.
Need to right shift number down when hash is more than 521 bits.
Previously handled at a byte level and now at bit level.
Always return err from sp_*_ecc_mulmod_add_only_*().
When ECC add and double points are public and only have SP
implementation, check that the point ordinates are the right size.
When exponent bit length is a multiple of the window size and the top
word has only window bits in it, then n is shifted down by an undefined
value (size of a word). The n value is not used after this.
Check for this condition and don't attempt to shift n.
Improve the generated instructions for multiplying 32 bits by 32 bits
and squaring 32 bits with 64 bit results.
Minimize loads in montgomery reduction, mul and sqr.
Pull conditional subtract into montgomery reduction assembly code.
Added unrolled square and multiply
Define WOLFSSL_SP_LARGE_CODE to use the unrolled versions.
Faster small code and fast code.
Allow fixed 4096-bit FFDHE parameters in benchmark.
Convert [u]int[32|64|128]*_t types to sp_[u]int[32|64|128].
Add a div for when top bits are all 1
WOLFSSL_SP_FAST_LARGE_CODE added to make mul_add function faster on
non-embedded platforms.
Change mod_exp window sizes for same performance but less memory.
P256 with c32 now 9 words instead of 10.
Fix position of sp_1024_norm_18 now that div requires it:
./configure --disable-shared --enable-sakke --disable-eccsi
--enable-sp
Fix missing '{' in sp_mulmod_table_1024:
./configure --disable-shared --enable-sakke --enable-eccsi
--enable-smallstack --enable-sp
Do quick bit length check before loading the MP integers into fixed size
arrays.
Changed ECC to use SP key check function if SP enabled and not only with
SP Math.
