GNU bug report logs - #78444
30.1; Crash in GC (vector_marked_p)

Previous Next

Full log

Message #68 received at 78444 <at> debbugs.gnu.org (full text, mbox):

From: Pip Cet <pipcet <at> protonmail.com>
To: George P <georgepanagopo <at> gmail.com>
Cc: Eli Zaretskii <eliz <at> gnu.org>, acorallo <at> gnu.org, 78444 <at> debbugs.gnu.org
Subject: Re: bug#78444: 30.1; Crash in GC (vector_marked_p)
Date: Sat, 31 May 2025 08:18:38 +0000

"George P" <georgepanagopo <at> gmail.com> writes:

>  Have you customized anything to 256 MB, precisely or approximately?
>  What are your values for gcmh-high-cons-threshold and
>  gcmh-low-cons-threshold, assuming that was in use during the crash?
>
> Not that I could find, unfortunately. My gcmh-low-cons-threshold is at 800000 and gcmh-high-cons-threshold is at 134217728.

Thanks for checking!

> (gdb) p (struct Lisp_String *)0x000000001cfbfe40
> $19 = (struct Lisp_String *) 0x1cfbfe40

Oh, sorry, "p *(struct Lisp_String *)0x000000001cfbfe40".  I'm surprised
0xffdf6e1 is a valid pointer, and 0xffdf6e5 would be a Lisp_Object
pointing to a two-element vector, which might be the lost constants
vector.  So maybe someone flipped a bit, or incorrectly treated the
object as a fixnum and silently decremented it?

Eli's idea of adding -fno-tree-sra to the build options for the next
build is a good one.

Pip

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.