Re: Vim: Caught deadly signal ABRT

[Ottavio Caruso <ottavio2006-usenet2012%yahoo.com@localhost>]

On Sat, 13 Jul 2019 at 11:28, Ottavio Caruso
<ottavio2006-usenet2012%yahoo.com@localhost> wrote:
>
> On Sat, 13 Jul 2019 at 10:14, <maya%netbsd.org@localhost> wrote:
> >
> > So, upstream knows that vim crashes with FORTIFY_SOURCE, and tries to
> > force-disable it. They think it's not a bug.
> >
>
> This is what I found in work/editors/vim/work/.work.log:
>
> gcc/bin/gcc -m64 --version -std=gnu99 -D_FORTIFY_SOURCE=2
> -fstack-protector
> -I/home/oc/pkgsrc/work/editors/vim/work/.buildlink/include
> <Ottavio> cc -c -O2 -D_FORTIFY_SOURCE=2 -I/usr/include
> -I/opt/pkg/include/ncurses -I/usr/include -I/opt/pkg/include/ncurses
> conftest.c
>
> There must be some patch somewhere forcing D_FORTIFY_SOURCE=2 and
> overriding D_FORTIFY_SOURCE=1.
>

Out of inexperience, I forgot that we have:

PKGSRC_USE_FORTIFY?= strong

as default, unless declared otherwise. Therefore I assume just declaring:

PKGSRC_USE_FORTIFY?= weak # (or even removing '?' altogether)

should be enough for vim to produce a binary that doesn't crash. I
could go and re-build, but I'd like to know is this approach is
'kosher'.

I'm also curious to know if the same thing happen on a NetBSD box.



-- 
Ottavio Caruso