Gregory Maxwell

Then you're no longer testing the production libraries and small compiler differences change the constant time behaviour. The tests are compiled with special instrumentation and don't use the libraries in...

I'm a little dubious about this running in make check in bitcoin core. You're going to get reports of errors which will be hard to address. e.g. there are ones...

Unless the failure is extremely rare and specialized, documenting a failure thrown out by make check will just perpetually stop users from reporting other vaguely similar looking issues because they...

As it stands now the test fails on platforms these distributions targets! So the result would be what I pointed out above w/ debian-- suddenly the distribution refuses to upgrade...

When it fails will someone fix the failures? If not, there is only downsides from running it.

> I think this should be fixed I'm very interested in doing something about it! But I didn't know it. It's not documented anywhere (except that you hinted at it...

The reason for the rest behaviour is so that tests can get back to an exactly deterministic state, e.g. so that we can have a bunch of test cases that...

Your typical modern embedded device doesn't have much ram but they have a fair amount of memory mapped flash, static const data ends up in flash. Some devices only care...

The second commit is an alternative to #789.

Can someone with access please kick travis to clear the spurious failure?