Re: local builds break "devel/gobject-introspection" et al.

["John D. Baker" <jdbaker%mylinuxisp.com@localhost>]

On Tue, 2 Aug 2016, John D. Baker wrote:

> Update the host OS from later TNF snapshots.  Go back and rebuild each
> of the above packages.  This should succeed.

For what it's worth, my i386 local-disk test system, which has only been
running TNF snapshots, spontaneously failed building "gstreamer1" in the
manner reported in pkg/51266.  See the PR for gdb backtraces of the
python2.7.core.  The "Gst-1.0.core" also present has no matching binary.
It was a JIT-generated tool in a temporary directory that was removed
as soon as it terminated.

The test system now fails to build all previously-mentioned problem
packages.

Fortunately, I'd just finished building my usual complement of packages
(except libreoffice, which is blocked by pkg/51221).  I attempted it
anyway and got the failure noted in the PR.  I know it sounds ridiculous,
but I wonder if that somehow triggered the problem with the other
packages.

I never had any of these troubles before the update to GCC 5.4.  I'm
pretty sure I didn't have trouble with the brief period that -current
used GCC 5.3 and I definitely didn't when it was still running GCC 4.8.5.


Meanwhile, my amd64 test system, running semi-diskless with a TNF snapshot
build, continues building packages without problems so far.

-- 
|/"\ John D. Baker, KN5UKS               NetBSD     Darwin/MacOS X
|\ / jdbaker[snail]mylinuxisp[flyspeck]com    OpenBSD            FreeBSD
| X  No HTML/proprietary data in email.   BSD just sits there and works!
|/ \ GPGkeyID:  D703 4A7E 479F 63F8 D3F4  BD99 9572 8F23 E4AD 1645