Shane
Shane
What am I supposed to do with the tmp inv param? I was trying to avoid allocating it internally? (add it as an optional last param?)
We now make EC keys produce a keypair, and now this is doing the opposite? What type of key is this actually for?
@paulidale I think I prefer the iterator approach.. See linux kernel list.h for example.. Have you seen the xor variant (would not make things easy to debug :)).
> I did mention the weight of WPACKET and my concerns were dismissed out of hand. Hence this PR. Can't have it both ways. Either it's heavy or it isn't....
> I've added some fuzzing. If the tests are happy, I'll leave it. If not I'll move it to a new PR. Can you do a quick test of this...
If it was added I would expect that WPACKET might use the lightweight layer. (Having any type of error in the new code would not be a good thing though...
@t8m this does not seem like an acceptable strategy. Putting a hold on this so it is discussed within OTC. i.e. This needs to have a CI solution.
I don't have a problem with your PR, but I do think it would be quite embarrassing if OpenSSL shipped another release that crashed with a similar issue. If there...
> I'm giving SDE a go in CI, but no promises yet (I'm also new to the tool). It's not the fastest so we'll see how it goes. Thanks for...
Not sure if this is helpful https://ci.chromium.org/ui/p/boringssl/builders/ci/linux32_sde/b8798585425292411169/steps