Filippo Valsorda

The default (single) is good for panics, but pretty useless for SIGQUIT since the signal handler is not the reason the program timed out. I think I picked system by...

Thank you!

It might be worth making these always static (ending in `` like `TEXT p256SqrInternal(SB),NOSPLIT,$0`) so that they don't conflict with symbols with the same name in other packages or modules.

I still support this, but in the interest of being aware of expected breakage, Synology apparently ships Linux 3.10 without getrandom() in recent (current?) firmwares. The "good" news is that...

> And then we also have 6,026 other active Linux users NOT using Synology running kernels older than 3.17. > How many of those are RHEL 7 and derivatives (which...

I suggest a slightly different semantic: the `HostKeyCallback` error (let's call it `UnknownKeyError`) would be returned for any unknown key, and have a field `KnownTypes []string`. The caller then filters...

Panics are recoverable, throw is an internal name for fatal errors. Think of it as a call to `exit(1)`. If we made crypto/rand panic that risks encouraging applications to wrap...

> Should we keep the /dev/urandom fallback for 3.17+ ? This is tempting, but I think making decisions based on kernel version is opening a can of worms. I think...

I was thinking about that but I have no intuition as to whether the cost of calling getrandom (to check if it's available) on `init()` for every Linux program is...

A follow-up question from the CL: should we have a `randcrash=0` GODEBUG to turn the throw off? It makes it a little less obvious/universal that it's ok to ignore the...