Re: Segfault 2.15.23 Span_bar_stub_engraver

[David Kastrup]

Jay Anderson <address@hidden> writes:

> On Wed, Jan 4, 2012 at 1:41 AM, David Kastrup <address@hidden> wrote:
>> Can you use git bisect for identifying the commit where things go wrong
>> for you?
>
> Git bisect results:
> 20670d51f8d97fd390210dd239b3b2427f071e7c is the first bad commit
> commit 20670d51f8d97fd390210dd239b3b2427f071e7c
> Author: Mike Solomon <address@hidden>
> Date:   Fri Sep 30 08:16:07 2011 +0200
>
> It looks like this was reverted and a revised version was committed
> the next day: 4f49b000d6e257724e311b406e2346b8388c1f0e. I've verified
> that the commit right before this doesn't cause a segfault and this
> one does. The only other information I have that's relevant is that
> I've only seen this segfault happen on my 64-bit OS and not 32-bit.

Grob::get_vertical_axis_group is not protected against the case where g
has an axis group interface but no Y_AXIs parent.

The second if in Grob::vertical_less has a test that can obviously not
be true.

Either don't look like 64bit problems.  Perhaps something needs a
live-check?

-- 
David Kastrup