SSWU parameters

[yelhousni]

This PR adds parameters for SSWU:

• On G1: BLS12-377, BLS24-315, BW6-761 and BW6-633 with a 2-isogeny
• On G2: BW6-633 with a 5-isogeny

Other curves, on G2:

• BN254 with a 193-isogeny (too big we might want to stick to generic SW)
• BLS24-315 with a 5-isogeny (Fp4 WIP)
• BLS12-377 with a 23-isogeny 85752112189-isogeny on G2 (too big we might want to stick to generic SW)
• BW6-761 with a 37-isogeny (too big? we might want to stick to generic SW)

on G1:

• BN254 with a 59-isogeny (too big we might want to stick to generic SW)

This will be addressed in a different PR. Also see #155 .

[Tabaie]

A 39-isogeny may still be good, as I remember the 11-isogeny in BLS12-381 took a very small portion of runtime. Should we generify the SW implementation from BLS12-381 to use for BLS12-377?

[yelhousni]

A 39-isogeny may still be good, as I remember the 11-isogeny in BLS12-381 took a very small portion of runtime. Should we generify the SW implementation from BLS12-381 to use for BLS12-377?

Yes we might give the 37-isogeny SSWU a try and yes we should generify th SW map for the remaining cases. We can take the implementation from any curve as the code is the same (b=0).

[yelhousni]

Last commit adds parameters of a 37-isogeny for the G2 bw6-761 case. I think a degree 37 is still fine. G1 and G2 for BW6 are both on Fp so comparing G1 SSWU with a 2-isogeny to G2 SSWU with a 37-isogeny is ~35 us difference.

[yelhousni]

and I was mistaken there is a suitable 23-isogeny on G2 for BLS12-377 (commit fadf1ae). So we should get rid of the slower SW map and keep only SSWU (#155).

[yelhousni]

#158 (fixed)