tech-kern archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: Regarding the ULTRIX and OSF1 compats



Le 09/03/2019 à 18:58, Jason Thorpe a écrit :
On Mar 9, 2019, at 8:22 AM, Maxime Villard <max%m00nbsd.net@localhost> wrote:
New thread, CC'ed to port-pmax@ and port-alpha@, in the continuity of:

	https://mail-index.netbsd.org/tech-kern/2019/03/09/msg024754.html

Basically, there were talks about retiring COMPAT_ULTRIX and COMPAT_OSF1,
because they are of questionable utility, in addition to being clear dead
wood; as I said, limited use case, no maintenance done in these areas,
inability to test changes, and the list goes on.

So, does anyone have anything to say about that?

I'm completely in favor of removing COMPAT_OSF1 ... it was useful back in
the early days of the Alpha port because the native toolchain was shaky (and
there was the whole ECOFF -> ELF transition).  But it was never good enough
to run many real applications because it doesn't implement any of the Mach
traps.  To fix it for real, we'd have to implement a ton of infrastructure
(including MiG), and I don't think anyone wants that horror show.

Alright

COMPAT_ULTRIX is a much thinner shim on top of of COMPAT_43, similar to
COMPAT_SUNOS, which we keep around for other reasons, and so I think the
wood is a little less obviously-dead.  Its biggest wart is having to exec
ECOFF binaries.  I don't know if anyone is actively using it at this point,
though.

Well, as far as I can tell no one has been maintaining it for quite some time
now, and it has now reached the point where people make blind changes without
being able to test; and the code is deprecating all along. Typically like
this:

	https://mail-index.netbsd.org/source-changes/2017/12/03/msg090179.html

It seems that, slowly, it will reach the final stage: being completely broken,
like SVR4. When the feature in question has a questionable utility, I'm even
more in favor of cutting the wood right away...


Home | Main Index | Thread Index | Old Index