bug#44031: 28.0.50; Segfault with poly-gfm and atomic-chrome

[Lars Ingebrigtsen <larsi@gnus.org>]

toon@iotcl.com writes:

> All goes well up to this point, until I want to pass the buffer back to
> Firefox. When I press C-c C-c, bound to
> `atomic-chrome-close-current-buffer', Emacs crashes.
>
> The backtrace is attached below, but I must say I have no idea what
> exactly causes this crash.

(I'm going through old bug reports that unfortunately weren't resolved
at the time.)

> Thread 1 "emacs" received signal SIGSEGV, Segmentation fault.
> 0x000055555576f55f in find_newline (start=33, start_byte=33, end=1, end_byte=1, count=-1, counted=0x7fffffffd320, bytepos=0x7fffffffd358, allow_quit=true) at search.c:838
> 838			       && FETCH_BYTE (start_byte - 1) == '\n')
> (gdb) bt full
> #0  0x000055555576f55f in find_newline (start=33, start_byte=33, end=1, end_byte=1, count=-1, counted=0x7fffffffd320, bytepos=0x7fffffffd358, allow_quit=true) at search.c:838
>         lim1 = 1
>         next_change = 1
>         result = 0
>         tem = 93824994639943
>         ceiling_byte = 1
>         newline_cache = 0x555557a13bd0
>         cache_buffer = 0x55555e16cdf0
> #1  0x000055555576f9e0 in scan_newline_from_point (count=0, charpos=0x7fffffffd350, bytepos=0x7fffffffd358) at search.c:969

I vaguely remember there being some fixes in this area somewhat recently
(but I may well be misremembering).  Do you still see these crashes with
the current Emacs trunk?

-- 
(domestic pets only, the antidote for overdose, milk.)
   bloggy blog: http://lars.ingebrigtsen.no