bug#34394: 27.0.50; Emacs segfaults with SLY, company and C-g

[Eli Zaretskii]

> From: João Távora <joaotavora@gmail.com>
> Date: Mon, 18 Feb 2019 20:42:16 +0000
> Cc: 34394-done@debbugs.gnu.org
> 
> Thanks a lot.  I've only now read your complete explanation and
> it's very interesting. You make it sound slightly easier than what
> I though it was. Perhaps this part of Emacs isn't particularly hard,
> or maybe everything is easy in hindsight.

Once you gave me a recipe for reproducing the bug, the main problem
was how to catch the situation where we use the wrong jmp_buf.  That
took a few attempts, but eventually succeeded, and the solution from
there was almost obvious, because we use the various unwind_protect
methods all over the place.