[pkg-gnupg-maint] Bug#926984: Bug#926984: gnupg2 FTBFS with gcc-9: dirmngr/dns.h:1058:24: error: lvalue required as unary '&' operand

[Helmut Grohne]

Control: tags -1 - moreinfo + bullseye sid

Hi Daniel,

On Mon, Apr 15, 2019 at 06:24:14PM -0400, Daniel Kahn Gillmor wrote:
> but i don't understand how the change works, or how it affects the build
> process that it says is failing.  The change just seems to be saying
> "don't bother ignoring the -Woverride-init warning" on gcc 9 or later.
> 
> How does that effect the lvalue of a unary & command?

To be honest, all I did here was google the message and when the first
hit was a patch from a respected community member, I gave up trying to
understand the problem at hand.

I think the bug itself is clear: You just build gnupg2 with gcc-9 (both
in Debian) and it fails. I'm also tagging it bullseye now to avoid
making buster rc-buggy when this bug becomes rc.

> Also, the chnage from Khem Raj (in cc here) says "Upstream-Status:
> pending", but i don't see it in the upstream bug tracker or revision
> control here: https://dev.gnupg.org/ and i haven't seen any messages
> from Khem on the upstream GnuPG mailing lists.

I'm very happy that you took matters in your own hands and reached out
to Khem here. That's going to be more productive. Thank you.

We still have plenty of time here. Afaik, archive-wide rebuilds with
gcc-9 haven't been performed yet. This is an early notice courtesy of
architecture bootstrapping efforts.

If you happen to find a patch that you prefer, I'd like to receive a
notice such that I can cherry pick it into rebootstrap.

Helmut