Bug#886294: transition: nodejs

cc-ing s390x team to ask (re)building nodejs-8.9.3~dfsg-11 on zemlinsky.

2018-01-21 18:46 GMT+01:00 Jérémy Lal <kapouer@melix.org>:

2018-01-21 17:45 GMT+01:00 Jérémy Lal <kapouer@melix.org>:
2018-01-21 16:25 GMT+01:00 Emilio Pozuelo Monfort <pochu@debian.org>:
On 09/01/18 19:02, Jérémy Lal wrote:
> It should be all right on armxx.
> However what i believed to be a memory exhaustion bug might be
> some problem with v8 and some mips/s390x processors (which is bad news,
> but i'm not sure it's the case yet).

I'm not sure what the problem is on mips(el),

For mips(el) it was an issue with the number of jobs used during tests,
leading to memory exhaustion. Fixed in nodejs-8.9.3~dfsg-9 by setting
that number of jobs to "one" when SCHROOT_USER is set.

I was too optimistic here, judging from the build logs and failures it seems there
too there are failures on specific processors.

So that problem is #888021 and has been fixed thanks to James Cowgill in version

nodejs-8.9.3~dfsg-10.

but on s390x you're getting
illegal instructions on zemlinsky, which is a Z10 mainframe. Looks like newer
node possibly bumped the baseline, or just accidentally introduced instructions
not supported by our baseline.

Starting investigations about that. Hopefully it's a change that could have been
backward-compatible.

nodejs/v8 somewhat officially support s390x down to z196.

I removed the added march=z196 flag and uploaded it into nodejs-8.9.3~dfsg-11.

It would be wonderful to build it on zemlinsky to see what happens.

Jérémy