Re: rc2 testing: groff uses bogus putenv test

"G. Branden Robinson" <[email protected]> Wed, 4 Feb 2026 22:06:36 -0600

Newsgroups	gmane.comp.printing.groff.general
Message-ID	<20260205040636.6xdymiikopue2n7a@illithid>

Hi Ingo,

Thanks for the build report.

At 2026-02-05T04:09:25+0100, Ingo Schwarze wrote:
> i just tested rc2 for the first time.  As far as my testing in rc1 went
> previously (i never completed testing rc1), i see no further
> regressions in rc2 beyond what i already reported for rc1.
> 
> It appears parallel building now works (at least it did when i
> tested it once, which is admittedly not conclusive evidence because
> for parallel build errors, intermittent failures are typical).
> For now, i will do test builds in serial mode anyway because
> that makes diff(1)ing build logs easier.  Thoroughly testing
> parallel builds is not a priority at this point.
> 
> Here is the next crash on OpenBSD, present in both rc1 and rc2.
> 
> Groff uses the gnulib putenv module.  That module tests putenv(3)
> in a bogus way, which causes my OpenBSD build to crash:
[...]
> With the workaround appended below, i now managed for the first time
> to get groff to build on OpenBSD.  I do not think it is building
> correctly just yet, but at least the build now finishes without
> crashing.

I wonder why Bruno Haible has so much less difficulty building on
OpenBSD than you do.

Do you not build stock groff, but only attempt a build after applying
your raft of patches to make the formatter work more like mandoc(1)?

Regards,
Branden

signature.asc (application/pgp-signature, 833 B)