RFR (S) 8217315: Proper units should print more significant digits

[Aleksey Shipilev]

On 1/17/19 4:08 PM, Thomas Stüfe wrote:
> Looks good to me too. 

Thanks!

> A better solution may be printing as float, see e.g. "print_human_readable_size"
> in metaspaceCommon.cpp . But that is outside the scope of your change.

Yes.

> As for the test, note that you could include all sizes up to the last one (50G) in 32bit too.

I could, but I would like not to, and follow what other tests in the same file do with "G". "G" is
very close to SIZE_MAX on 32/31 bit platforms, it seems risky to expose it too much.

-Aleksey