Re: kern/52598 (/proc/self/exe misbehaves if it's on a null mount)

[Martin Husemann <martin%duskware.de@localhost>]

On Tue, Jan 14, 2020 at 05:31:52PM -0500, Christos Zoulas wrote:
> Well, you need a kernel bump because sizeof(struct proc) changes...
> This would probably be put on 8.2...

This would prevent pullup according to current releng policies.

We can make exceptions and document them carefully in case that are
considered extremely important, as the main idea behind the rule is
to prevent binaries compiled for newer netbsd-8 to not work on (say)
NetBSD 8.1.

This change is userland visible, and kmem users may break. It is not clear
to me if the what exactly the amount of breakage would be and if the
benefit of the change would be higher.

I have not followed the thread closely from the beginning, is this about
ld.elf_so's ORIGIN handling? I thought we don't rely on that in pkgsrc
rust anyway (past bootstrap).

Martin