On Fri, 2024-03-08 08:42:52 +0100, Jan-Benedict Glaw <jbglaw%lug-owl.de@localhost> wrote: > On Fri, 2024-03-08 12:28:54 +0900, Izumi Tsutsui <tsutsui%ceres.dti.ne.jp@localhost> wrote: > > jbglaw@ wrote: [...] > > NetBSD 6.1.5 (I'm not sure when I prepared this image) + open-simh > > from pkgsrc-2023Q4 partially worked: > > > > --- > > % simh-vax780 netbsd-boot780 > > > > VAX 11/780 simulator Open SIMH V4.1-0 Current simh git commit id: b23cde9c Build: Unsupported=include+lib [...] > > NetBSD 6.1.5 (GENERIC) > > VAX 11/780 > > total memory = 127 MB > > avail memory = 120 MB > > That's at least a very helpful data point! This could get me started > with some bisecting. (At least if I manage to get it build. My host > Docker images are always very recent, that might produce issues with > building the tools, -fcommon changes, ...) With that information, I started to `git bisect`. That's painfully slow and reminds me of worse days where it wasn't that easy (compared to today) to build NetBSD VAX. Compiler issues, ramdisk size issues, general code issues, ... I'm testing the vax780 simulator as well as the vax8600, both fail to boot the NetBSD ISOs these days (and showed the same behavior for the ISO images I've already fed in.) So I hope it's a common issue. Right now, I know that it broke somewhere between 4029805dcf59ae611e79df9f7e6a7966d62513a7 Wed May 10 08:02:37 2017 +0000 and 1c13c1a096d2a19333b76fdd585d1d7bcd4ad8bc Wed May 29 10:09:01 2019 +0000. As the issues are usually easy to fix, I'll just slowly build the stuff in between, that'll probably take a week or two. That way, I don't need to actually spend a lot of time. Too bad I don't have access to real hardware. I'd *love* to explore whether or not that will boot a current ISO image... MfG, JBG --
Attachment:
signature.asc
Description: PGP signature