bug#13743: 24.2.93; Segmentation fault when trying to [s]teal a file opened elsewhere

[Eli Zaretskii <eliz@gnu.org>]

> Date: Tue, 19 Feb 2013 04:52:59 +0400
> From: Dmitry Gutov <dgutov@yandex.ru>
> CC: 13743@debbugs.gnu.org
> 
> On 18.02.2013 20:11, Eli Zaretskii wrote:
> > Can you recompile without optimizations, and show the backtrace from
> > that build?  Thanks.
> 
> Here you go:
> 
> Program received signal SIGSEGV, Segmentation fault.
> 0x0000000000653481 in add_properties (plist=32025878, i=0x0, 
> object=31976325)
>      at textprop.c:378
> 378	      for (tail2 = i->plist; CONSP (tail2); tail2 = Fcdr (XCDR (tail2)))
> (gdb) bt full
> #0  0x0000000000653481 in add_properties (plist=32025878, i=0x0,
>      object=31976325) at textprop.c:378
>          tail1 = 32025878
>          tail2 = 31873526
>          sym1 = 13125730
>          val1 = 12886562
>          changed = 0
>          found = 0
>          gcpro1 = {
>            next = 0x7fffffff6180,
>            var = 0x57dc19 <modify_region+481>,
>            nvars = 4295010134
>          }
>          gcpro2 = {
>            next = 0x7fffffff6180,
>            var = 0x57dc19 <modify_region+481>,
>            nvars = 4295010134
>          }
>          gcpro3 = {
>            next = 0x7fffffff6180,
>            var = 0x57dc19 <modify_region+481>,
>            nvars = 4295010134
>          }
> ---Type <return> to continue, or q <return> to quit---
> #1  0x0000000000655963 in Fadd_text_properties (start=4, end=3068,
>      properties=32025878, object=31976325) at textprop.c:1212
>          i = 0x0
>          unchanged = 0x7fffffffa568
>          s = 1
>          len = 558
>          modified = 19
>          gcpro1 = {
>            next = 0xa31e4d <pure+1297709>,
>            var = 0xa31e29 <pure+1297673>,
>            nvars = 579820584989
>          }
> #2  0x00000000006559e0 in Fput_text_property (start=4, end=3068,
>      property=13125730, value=12886562, object=12886562) at textprop.c:1229
> No locals.

Thanks.  This shows quite a different story.  But the backtrace alone
is not enough to figure out what goes wrong in this scenario and why,
at least not for me (and I don't see anyone else jumping in to dig
into this problem).  And since reproducing this involves tricky
non-default setup and an external package, I wonder if you could
provide a recipe starting with "emacs -Q" or with a minimal .emacs
init file, and show every command you type to reproduce the crash?

> By the way, is there an easier way to compile without optimizations than 
> assigning CFLAGS before running configure?

That's the canonical way, AFAIK.