Re: NetBSD/vax compiler bounty?

[maya%NetBSD.org@localhost]

On Thu, Apr 16, 2020 at 09:22:52AM -0700, Jason Thorpe wrote:
> 
> > On Apr 16, 2020, at 9:18 AM, emanuel stiebler <emu%e-bbes.com@localhost> wrote:
> > 
> > Any idea, how long it actually takes, to compile gcc on a vax?
> 
> I don't think it's possible anymore due to memory consumption by the compiler.  That said, GCC doesn't self-host on a lot of targets (mainly the microcontroller ones), so I don't think this is a big deal.  We should focus on making a modern cross-vax-gcc work correctly.  As far as the test suite goes, perhaps a way to remotely execute the generated binaries (on a real machine or in a simulator) that need to run to ensure the compiler generated the correct code is the way to go... maybe the GCC test suite already supports that, even.
> 

An educated guess for the peak RAM usage of a native build would be
512MB RAM, this only matters if you can't add swap because you hit a
virtual address space limit.

I don't think a native build of GCC is necessary to test out GCC
patches. The marker of success I planned on using is a completed build
of NetBSD src booting and no visible regressions.