problem using gcc-core for compile qemu

[Eliot Moss]

On 1/7/2021 11:18 AM, Marco Atzeri via Cygwin wrote:

 > It is not that is impossible, but will likely need a big effort.
 >
 > As there are Windows binaries, I do not think it is worth

Right - I had noticed on the web some mention of people maybe working on
getting qemu going under Cygwin.  There is no reason why that cannot be done
in principle, but since qemu can boot operating systems and such, it requires
proper mapping a many OS calls, etc.

Naturally I have no idea what the OP's overall goal is.  As a computer science
researcher, I have often found that for setups like this, a VirtualBox virtual
machine is better because I can get a well-controlled "pure Linux"
environment, that I can also package up and share with students and
colleagues.  Also, compared with Cygwin, I typically get higher speed.  This
is because there is overhead inherent in modeling Posix on top of Windows via
a library.  A VM runs actual Linux.  Don't get me wrong - I love Cygwin and
use it for much of my work!  I like having Windows apps and tools available
while living is a very Unix-like command-line world.  But it's not the only
thing out there.

Note that a modern alternative to VirtualBox (or vmware) is WSL, the Windows
Subsystem for Linux, which (it turns out) is also a VM.  (Note: Unless things
have changed recently, WSL version 2 (which may offer performance advantages
over WSL version 1) is not compatible with VirtualBox.  They use the hardware
virtualization in "opposite" ways.  WSL 1 can be kept compatible with
VirtualBox, but its emulation of Linux on top of Windows suffers from similar
performance issues to those of Cygwin, especially around the fork() call.)

Anyway, maybe this will contribute to the OP's overall need, maybe not ...

Eliot Moss