Brooks Davis

I ended up working around this in our test infrastructure (where I suspect a hardware or pmap bug due to the symptom being a full page of 0s in a...

https://gist.github.com/brooksdavis/bf0977c69ebbac587e6e

It's not deterministic. Some number of tests run, but I've had failures occur between 30min and 3 hours into the test run. To be honest, I suspect a qemu bug...

I think either I or @staceyson could get it built for mips. I don't currently have a build environment, but do have a machine that I could set one up...

@staceyson: as a hack to see if there are any other issues, I might try replacing the exception in `datetime::timestamp::operator-` with returning 0 since I don't care all that much...

On another note, I've done some testing on our FPGA platform and thus far haven't hit this issue, but have managed to wedge the hardware before completing any runs (I've...

With @staceyson's package I'm still seeing seeing these exceptions. I've made very sure that I've installed that one and not another one, but something isn't working. The SHA256 of the...

I'm no longer able to trigger this so hopefully it's all fixed where ever the bug was.

I'm seeing this again with the 0.13,3 package on FreeBSD. ``` date/format_string_test:C_test -> kyua: E: Cannot subtract 1476919548082338us from 1476919547556756us as it would result in a negative datetime::delta, which are...

After further reflection, the core assumption this exception is asserting is false. `gettimeofday()` is not guaranteed to be monotonic. If you want something that is, `clock_gettime(CLOCK_MONOTONIC, &ts);` would be more...