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

Previous Next

Full log

View this message in rfc822 format

From: Pip Cet <pipcet <at> protonmail.com>
To: George P <georgepanagopo <at> gmail.com>
Cc: Eli Zaretskii <eliz <at> gnu.org>, 78444 <at> debbugs.gnu.org
Subject: bug#78444: 30.1; Crash in GC (vector_marked_p)
Date: Mon, 19 May 2025 15:44:00 +0000

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

>  Can you please run x/64gx 0x3aa1ac0 so we can be sure
>  of this? 
>
> Sure:
>
> (gdb) x/64gx 0x3aa1ac0
> 0x3aa1ac0:      0x00000000098f1d0d      0x0000000000000030
> 0x3aa1ad0:      0x00000000098f1d65      0x0000000000000030
> 0x3aa1ae0:      0x00000000098f1dbd      0x0000000000000030
> 0x3aa1af0:      0x00000000098f1e15      0x0000000000000030
> 0x3aa1b00:      0x00000000098f1e6d      0x0000000000000030
> 0x3aa1b10:      0x00000000098f1ec5      0x0000000000000030
> 0x3aa1b20:      0x00000000098f1f1d      0x0000000000000030
> 0x3aa1b30:      0x00000000098f1f75      0x0000000000000030
> 0x3aa1b40:      0x00000000098f1fcd      0x0000000000000030
> 0x3aa1b50:      0x00000000098f2025      0x0000000000000030
> 0x3aa1b60:      0x00000000098f207d      0x0000000000000030
> 0x3aa1b70:      0x00000000098f20d5      0x0000000000000030
> 0x3aa1b80:      0x00000000098f212d      0x0000000000000030
> 0x3aa1b90:      0x00000000098f2185      0x0000000000000030
> 0x3aa1ba0:      0x00000000098f21dd      0x0000000000000030
> 0x3aa1bb0:      0x00000000098f2235      0x0000000000000030
> 0x3aa1bc0:      0x00000000098f228d      0x0000000000000030
> 0x3aa1bd0:      0x00000000098f22e5      0x0000000000000030
> 0x3aa1be0:      0x00000000098f233d      0x0000000000000030
> 0x3aa1bf0:      0x00000000098f2395      0x0000000000000030
> 0x3aa1c00:      0x00000000098f23ed      0x0000000000000030
> 0x3aa1c10:      0x00000000098f2445      0x0000000000000030
> 0x3aa1c20:      0x00000000098f249d      0x0000000000000030
> 0x3aa1c30:      0x00000000098f24f5      0x0000000000000030
> 0x3aa1c40:      0x00000000098f254d      0x0000000000000030
> 0x3aa1c50:      0x00000000098f25a5      0x0000000000000030
> 0x3aa1c60:      0x0000000000000007      0x0000000000000007
> 0x3aa1c70:      0x0000000000000007      0x0000000000000007
> 0x3aa1c80:      0x0000000000000007      0x0000000000000007
> 0x3aa1c90:      0x0000000000000007      0x0000000000000007
> 0x3aa1ca0:      0x0000000000000007      0x0000000000000007
> 0x3aa1cb0:      0x0000000000000007      0x0000000000000007
> (gdb) 

That's a native comp unit's lambda_gc_guard_h, which is quite curious.
Presumably 0x98e7985 is the native comp unit and 0x8f680f4 is its file
name, so could you please run

    p *(char **)0x8f68108

to retrieve it, as well as

    x/32gx 0x9e7980

to confirm it is (or was) a native comp unit?

Going back through the last_marked array, it seems we're looking at the
'function-history plist property of a symbol at 0x15554df1a3a0, but I'm
not sure which of the strings we mark after that is its name.  Best to
print all of them:

    p *(char **)0x8ae5588
    p *(char **)0x15554ec0ff98
    p *(char **)0x15554ec0ff78

The last vector or pseudovector we marked before that was 0x38294cd, so
I think we should look at

    x/32gx 0x38294c8

too.

> Keep them coming! Are you still suspecting X?

Currently, it seems more likely to involve the nativecomp code, but I've
stared at it for a while and I don't see how it can resurrect comp units
once they become unreachable.

Thanks again!

Pip

Previous Next

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