NetBSD Problem Report #59479

From he@smistad.uninett.no  Mon Jun 23 07:38:15 2025
Return-Path: <he@smistad.uninett.no>
Received: from mail.netbsd.org (mail.netbsd.org [199.233.217.200])
	(using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits)
	 key-exchange X25519 server-signature RSA-PSS (2048 bits) server-digest SHA256
	 client-signature RSA-PSS (2048 bits) client-digest SHA256)
	(Client CN "mail.NetBSD.org", Issuer "mail.NetBSD.org CA" (not verified))
	by mollari.NetBSD.org (Postfix) with ESMTPS id 288EA1A9239
	for <gnats-bugs@gnats.NetBSD.org>; Mon, 23 Jun 2025 07:38:15 +0000 (UTC)
Message-Id: <20250623073810.CE5BE43F55F@smistad.uninett.no>
Date: Mon, 23 Jun 2025 09:38:10 +0200 (CEST)
From: he@NetBSD.org
Reply-To: he@NetBSD.org
To: gnats-bugs@NetBSD.org
Subject: gcc crash building pcre2 on m68k -current
X-Send-Pr-Version: 3.95

>Number:         59479
>Category:       toolchain
>Synopsis:       gcc crash building pcre2 on m68k -current
>Confidential:   no
>Severity:       serious
>Priority:       low
>Responsible:    toolchain-manager
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Mon Jun 23 07:40:00 +0000 2025
>Last-Modified:  Mon Jun 23 19:05:00 +0000 2025
>Originator:     Havard Eidnes
>Release:        NetBSD 10.99.14 of June 16 2025
>Organization:
 I try...
>Environment:
System: NetBSD virt68k.urc.uninett.no 10.99.14 NetBSD 10.99.14 (GENERIC) #1: Mon Jun 16 15:52:20 CEST 2025  he@mt.urc.uninett.no:/u/build/HEAD/obj/virt68k/sys/arch/virt68k/compile/GENERIC virt68k
Architecture: m68k
Machine: virt68k
>Description:
	While building pkgsrc/devel/pcre2, gcc decided not to collaborate:

===> Building for pcre2-10.45
/usr/bin/make  all-am
  CC       src/libpcre2_8_la-pcre2_match.lo
during GIMPLE pass: cddce
src/pcre2_match.c: In function 'match':
src/pcre2_match.c:8072:1: internal compiler error: Segmentation fault
 8072 | }
      | ^
no stack trace because unwind library not available
Please submit a full bug report, with preprocessed source (by using -freport-bug).
See <http://www.NetBSD.org/support/send-pr.html> for instructions.
*** Error code 1

>How-To-Repeat:
	cd /usr/pkgsrc/devel/pcre2
	make
>Fix:
	That would be nice...

>Audit-Trail:
From: Martin Husemann <martin@duskware.de>
To: gnats-bugs@netbsd.org
Cc: 
Subject: Re: toolchain/59479: gcc crash building pcre2 on m68k -current
Date: Mon, 23 Jun 2025 09:54:29 +0200

 On Mon, Jun 23, 2025 at 07:40:00AM +0000, he@NetBSD.org wrote:
 > src/pcre2_match.c:8072:1: internal compiler error: Segmentation fault
 >  8072 | }
 >       | ^

 Can you try unlimiting the build and check if you run out of swap space
 here?

 Martin

From: Havard Eidnes <he@NetBSD.org>
To: gnats-bugs@netbsd.org, gnats-admin@NetBSD.org
Cc: toolchain-manager@netbsd.org, netbsd-bugs@netbsd.org
Subject: Re: toolchain/59479: gcc crash building pcre2 on m68k -current
Date: Mon, 23 Jun 2025 16:15:19 +0200 (CEST)

 >  > src/pcre2_match.c:8072:1: internal compiler error: Segmentation fault
 >  >  8072 | }
 >  >       | ^
 >
 >  Can you try unlimiting the build and check if you run out of swap space
 >  here?

 I did "unlimit datasize" and re-tried the build.

 It crashed the same way.

 I had a "top" window up at the same time, and the "SIZE" coloumn
 maxed out at around 71MB for the cc1 process.

 virt68k: {14} limit
 cputime      unlimited
 filesize     unlimited
 datasize     65536 kbytes
 stacksize    2048 kbytes
 coredumpsize unlimited
 memoryuse    122220 kbytes
 vmemoryuse   unlimited
 descriptors  128 
 memorylocked 40740 kbytes
 maxproc      160 
 maxthread    2048 
 sbsize       unlimited
 virt68k: {15} 

 is what I tried with.

 Note, I beleive the dynamically allocated memory is done by mmap,
 so is not accounted for in the "datasize" limit(?)

 Regards,

 - Havard

From: matthew green <mrg@eterna23.net>
To: Havard Eidnes <he@NetBSD.org>
Cc: toolchain-manager@netbsd.org, netbsd-bugs@netbsd.org,
    gnats-bugs@netbsd.org, gnats-admin@NetBSD.org
Subject: re: toolchain/59479: gcc crash building pcre2 on m68k -current
Date: Tue, 24 Jun 2025 05:03:06 +1000

 i wonder if we should turn NOGCCISL on for all the older platforms,
 it might reduce the GCC costs slightly and help this general issue.


 .mrg.

Home	PR Database Search