Possibility to use more than one curve at once about micro-ecc HOT 10 CLOSED

I'll probably need this at some point in the near future too, my temporary plan was just to have a script that batch renames things so that there are different symbols for each curve, seemed like an easy enough compromise but haven't tried it yet :)

from micro-ecc.

I am working on it. The eventual plan is to have a single #define that controls the size/speed tradeoff, and #defines to determine which curves are supported.

from micro-ecc.

Sounds good! I'm looking forward your solution! Thank your very much!

from micro-ecc.

master branch now supports runtime curve selection. I still need to add fast mult/square asm for AVR, and also implement optimization level 4 for both platforms.

from micro-ecc.

We integrated the new version in december. Very good. Thank you for that! Everything is working well. Today I noticed that our communication is about 250ms slower now. We have a long time graph of the duration in communication and after my checkin in december we have that 250ms jump in the graph. I set uECC_OPTIMIZATION_LEVEL=4 now but no change. Are there any new options I have to use? What is the new VLI thing?

from micro-ecc.

Which platform are you using (and which curves)? Unfortunately I've been pretty busy for a while, so haven't had time to work on improving the optimizations.

from micro-ecc.

It's an ARM Cortex M0. The tests running with secp192r1. Unfortunately I have no measurements for the secp256r1 but we also use it.

from micro-ecc.

Oh, ok. Adding the size parameter probably hurts Cortex M0 more because it has so few useful registers. I imagine that optimization level 4 will help, when I implement it. I am also looking into some improved multiplication/squaring implementations for Cortex M0.

from micro-ecc.

That would be nice! Please keep me updated. Thank you very much for that nice library.

from micro-ecc.

Should we discuss this in a new issue? I integrated your newest checkin. See the graphic for performance. It is not much better. We use uECC_OPTIMIZATION_LEVEL=4 is this right or should I change it?

from micro-ecc.