FYI: Make for faster windows build

[Tor Lillqvist]

> I have no idea what difference that'd make, I think there's Win32-specific
> code instead of fork(), I don't know if there's any Win32-specific stat()
> replacement or whether that'd make a noticeable difference.

It is quite possible that avoiding the Cygwin stat() and instead using
Win32 API to get *only* the information actually needed in each case
would give a measurable gain. I can imagine that especially
constructing the Unix-style mode bits from the Windows ACL can require
in the worst case quite a lot of code for Cygwin to do. Or not,
actually implementing a replacement and then doing some benchmarking
would be needed...

I once started hacking on that, but got distracted, and as far as I
recall, the use of stat() in make was somewhat convoluted and it was
not obvious in which cases just the st_mtime was needed, and when an
st_dev and st_ino was also needed.

--tml