- 21 3月, 2022 2 次提交
- 20 3月, 2022 5 次提交
- 17 3月, 2022 1 次提交
-
-
由 PMheart 提交于
-
- 11 3月, 2022 2 次提交
- 10 3月, 2022 4 次提交
- 09 3月, 2022 7 次提交
- 08 3月, 2022 1 次提交
-
-
由 PMheart 提交于
-
- 06 3月, 2022 16 次提交
-
-
由 vit9696 提交于
-
由 vit9696 提交于
-
由 vit9696 提交于
-
由 vit9696 提交于
-
由 vit9696 提交于
-
由 vit9696 提交于
This optimisation once again gives us nothing on realistic values due to roughly uniform distribution.
-
由 vit9696 提交于
BigNumCalculateMontParams previously did not check BigNumMod for memory allocation failure anyway, which was thus harder to diagnose due to the Montgomery Inverse of N still returning correctly.
-
由 vit9696 提交于
Currently BigNumMod does not allocate memory for A words with leading zeroes. For us words are 8 bytes. When doing RSA computations with proper keys 8 zero bytes are nearly impossible and 16 zero bytes are completely impossible. Avoiding this optimisation does not change library behaviour for nearly all inputs but provides simpler guarantees about memory usage and simplifies testing.
-
由 vit9696 提交于
-
由 vit9696 提交于
-
由 vit9696 提交于
-
由 vit9696 提交于
-
由 PMheart 提交于
-
由 PMheart 提交于
-
由 PMheart 提交于
-
由 PMheart 提交于
-
- 03 3月, 2022 2 次提交
-
-
由 Savva Mitrofanov 提交于