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

[Eli Zaretskii]

> From: João Távora <joaotavora@gmail.com>
> Cc: 34394@debbugs.gnu.org,  schwab@linux-m68k.org
> Date: Sat, 09 Feb 2019 14:13:52 +0000
> 
> So I disabled SIGINT for gdb and let it go to Emacs for keyboard-quit.
> I reproduced with -nw, here's the backtrace.

Is it possible that somehow sit_for is called recursively in this
scenario?  If you put a breakpoint in sit_for, does it break when bt
shows another instance of sit_for up the call-stack?

We need to establish who clobbers the jump_buf being used to implement
quitting by C-g.