NetBSD Problem Report #58737

From www@netbsd.org  Thu Oct 10 15:59:49 2024
Return-Path: <www@netbsd.org>
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 DA1BD1A923B
	for <gnats-bugs@gnats.NetBSD.org>; Thu, 10 Oct 2024 15:59:48 +0000 (UTC)
Message-Id: <20241010155947.555B21A923D@mollari.NetBSD.org>
Date: Thu, 10 Oct 2024 15:59:47 +0000 (UTC)
From: andrew.cagney@gmail.com
Reply-To: andrew.cagney@gmail.com
To: gnats-bugs@NetBSD.org
Subject: building against gcc (GCC) 14.2.1 20240912 (Red Hat 14.2.1-3) + glibc-2.39-22.fc40.x86_64
X-Send-Pr-Version: www-1.0

>Number:         58737
>Category:       toolchain
>Synopsis:       building against gcc (GCC) 14.2.1 20240912 (Red Hat 14.2.1-3) + glibc-2.39-22.fc40.x86_64
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    toolchain-manager
>State:          closed
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Thu Oct 10 16:00:00 +0000 2024
>Closed-Date:    Sun Dec 15 23:10:49 +0000 2024
>Last-Modified:  Sun Dec 15 23:20:01 +0000 2024
>Originator:     Andrew Cagney
>Release:        10.x
>Organization:
>Environment:
Linux bernard 6.10.12-200.fc40.x86_64 #1 SMP PREEMPT_DYNAMIC Mon Sep 30 21:38:25 UTC 2024 x86_64 GNU/Linux

>Description:
  Machine:   -m evbarm64
  Objdir:    -O /home/netbsd/10/evbarm64
  Toolsdir:  -T /home/netbsd/10/tools
  X11:       -x -X /home/netbsd/10/xsrc
  Jobs:      -j 17

(I wonder if this should include the host compiler and libc, hmm Alpine; but I digress)

dependall ===> tools/date
#    create  date/date.d
/home/netbsd/10/evbarm64/tools/host-mkdep/host-mkdep -f date.d.tmp  --    -Wno-error -Wno-literal-suffix -Wno-expansion-to-defined -Wno-implicit-fallthrough  -D_LARGEFILE_SOURCE -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -I/home/netbsd/
10/evbarm64/tools/compat -I/home/netbsd/10/evbarm64/tools/compat/include -I/home/netbsd/10/src/tools/compat -DHAVE_NBTOOL_CONFIG_H=1 -D_FILE_OFFSET_BITS=64 -I/home/netbsd/10/src/tools/date/../../bin/date -Wno-format-nonliteral    /home/ne
tbsd/10/src/tools/date/../../bin/date/date.c &&  mv -f date.d.tmp date.d
#    create  date/.depend
rm -f .depend
/home/netbsd/10/evbarm64/tools/host-mkdep/host-mkdep -s .lo\ .ln\ .d -d -f .depend date.d
#   compile  date/date.lo
cc -fcommon -Wno-error -Wno-literal-suffix -Wno-expansion-to-defined -Wno-implicit-fallthrough  -D_LARGEFILE_SOURCE -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -I/home/netbsd/10/evbarm64/tools/compat -I/home/netbsd/10/evbarm64/tools/comp
at/include -I/home/netbsd/10/src/tools/compat -DHAVE_NBTOOL_CONFIG_H=1 -D_FILE_OFFSET_BITS=64 -I/home/netbsd/10/src/tools/date/../../bin/date -c -o date.lo.o -Wno-format-nonliteral   /home/netbsd/10/src/tools/date/../../bin/date/date.c
cc1: warning: command-line option '-Wno-literal-suffix' is valid for C++/ObjC++ but not for C
/home/netbsd/10/src/tools/date/../../bin/date/date.c: In function 'setthetime':
/home/netbsd/10/src/tools/date/../../bin/date/date.c:218:21: error: implicit declaration of function 'strptime'; did you mean 'strftime'? [-Wimplicit-function-declaration]
  218 |                 t = strptime(p, fmt, lt);
      |                     ^~~~~~~~
      |                     strftime
/home/netbsd/10/src/tools/date/../../bin/date/date.c:218:19: error: assignment to 'const char *' from 'int' makes pointer from integer without a cast [-Wint-conversion]
  218 |                 t = strptime(p, fmt, lt);
      |                   ^

*** Failed target:  date.lo
*** Failed command: cc -fcommon -Wno-error -Wno-literal-suffix -Wno-expansion-to-defined -Wno-implicit-fallthrough -D_LARGEFILE_SOURCE -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -I/home/netbsd/10/evbarm64/tools/compat -I/home/netbsd/10/
evbarm64/tools/compat/include -I/home/netbsd/10/src/tools/compat -DHAVE_NBTOOL_CONFIG_H=1 -D_FILE_OFFSET_BITS=64 -I/home/netbsd/10/src/tools/date/../../bin/date -c -o date.lo.o -Wno-format-nonliteral /home/netbsd/10/src/tools/date/../../b
in/date/date.c
*** Error code 1

>How-To-Repeat:

>Fix:
I'll try hacking the headers

>Release-Note:

>Audit-Trail:
From: Martin Husemann <martin@duskware.de>
To: gnats-bugs@netbsd.org
Cc: 
Subject: Re: toolchain/58737: building against gcc (GCC) 14.2.1 20240912 (Red
 Hat 14.2.1-3) + glibc-2.39-22.fc40.x86_64
Date: Thu, 10 Oct 2024 18:09:55 +0200

 This has been fixed in HEAD, needs pullups.

 Martin

From: Martin Husemann <martin@duskware.de>
To: gnats-bugs@netbsd.org
Cc: 
Subject: Re: toolchain/58737: building against gcc (GCC) 14.2.1 20240912 (Red
 Hat 14.2.1-3) + glibc-2.39-22.fc40.x86_64
Date: Thu, 10 Oct 2024 18:20:23 +0200

 There were a few include cleanup commits, like:

 https://mail-index.netbsd.org/source-changes/2024/09/16/msg153356.html
 https://mail-index.netbsd.org/source-changes/2024/09/16/msg153357.html
 https://mail-index.netbsd.org/source-changes/2024/09/16/msg153358.html

 and a few changes to date + sed:

 https://mail-index.netbsd.org/source-changes/2024/09/17/msg153360.html
 https://mail-index.netbsd.org/source-changes/2024/09/17/msg153361.html
 https://mail-index.netbsd.org/source-changes/2024/09/17/msg153362.html
 https://mail-index.netbsd.org/source-changes/2024/09/17/msg153363.html

 but I haven't tested or checked if that was all.

 Martin

From: Andrew Cagney <andrew.cagney@gmail.com>
To: gnats-bugs@netbsd.org
Cc: toolchain-manager@netbsd.org, gnats-admin@netbsd.org, 
	netbsd-bugs@netbsd.org
Subject: Re: toolchain/58737: building against gcc (GCC) 14.2.1 20240912 (Red
 Hat 14.2.1-3) + glibc-2.39-22.fc40.x86_64
Date: Thu, 10 Oct 2024 12:26:24 -0400

 Which commit?  Ah, thanks, never mind!  Testing.

 On Thu, 10 Oct 2024 at 12:15, Martin Husemann via gnats
 <gnats-admin@netbsd.org> wrote:
 >
 > The following reply was made to PR toolchain/58737; it has been noted by GNATS.
 >
 > From: Martin Husemann <martin@duskware.de>
 > To: gnats-bugs@netbsd.org
 > Cc:
 > Subject: Re: toolchain/58737: building against gcc (GCC) 14.2.1 20240912 (Red
 >  Hat 14.2.1-3) + glibc-2.39-22.fc40.x86_64
 > Date: Thu, 10 Oct 2024 18:09:55 +0200
 >
 >  This has been fixed in HEAD, needs pullups.
 >
 >  Martin
 >

State-Changed-From-To: open->needs-pullups
State-Changed-By: maya@NetBSD.org
State-Changed-When: Wed, 11 Dec 2024 03:27:41 +0000
State-Changed-Why:
Update bug status to match martin's latest reply


State-Changed-From-To: needs-pullups->feedback
State-Changed-By: gutteridge@NetBSD.org
State-Changed-When: Wed, 11 Dec 2024 16:26:01 +0000
State-Changed-Why:
Pullup states are tracked in PR lib/58674. There's a separate question for the submitter now.

From: "David H. Gutteridge" <david@gutteridge.ca>
To: gnats-bugs@netbsd.org
Cc: 
Subject: Re: toolchain/58737 (building against gcc (GCC) 14.2.1 20240912 (Red
 Hat 14.2.1-3) + glibc-2.39-22.fc40.x86_64)
Date: Wed, 11 Dec 2024 11:22:17 -0500

 This is effectively a duplicate of PR lib/58674. The changes and
 discussion were tracked there. Everything referenced in this ticket has
 either been pulled up to the netbsd-10 branch or pullup tickets have
 been submitted but not yet processed (as they will probably go in after
 branching 10.1).

 The only outstanding question (not referenced so far in this PR, but on
 tech-toolchain) is this:
 https://mail-index.netbsd.org/tech-toolchain/2024/10/15/msg004547.html

 It's unclear from that commentary at what stage the failure happens
 (i.e. in a tools build or after that point).

 Dave

From: Andrew Cagney <andrew.cagney@gmail.com>
To: gnats-bugs@netbsd.org
Cc: toolchain-manager@netbsd.org, gnats-admin@netbsd.org, 
	netbsd-bugs@netbsd.org
Subject: Re: toolchain/58737 (building against gcc (GCC) 14.2.1 20240912 (Red
 Hat 14.2.1-3) + glibc-2.39-22.fc40.x86_64)
Date: Wed, 11 Dec 2024 11:46:43 -0500

 The tools built cross compiler:

    /home/netbsd/10/tools/bin/aarch64--netbsd-c++

 compiled using the native compiler:

   cc (GCC) 14.2.1 20240912 + glibc-2.39

 (on fc40), barfed trying to compile:

 /home/netbsd/10/src/external/apache2/llvm/librt/libLLVMAMDGPUCodeGen/../../lib/../dist/llvm/lib/Target/AMDGPU/SIRegisterInfo.cpp:14:
 /home/netbsd/10/evbarm64/destdir.evbarm/usr/include/inttypes.h:71:1:
 internal compiler error: Segmentation fault
    71 | uintmax_t strtoumax_l(const char * __restrict,
       | ^~~~~~~~~

 On Wed, 11 Dec 2024 at 11:35, David H. Gutteridge via gnats
 <gnats-admin@netbsd.org> wrote:
 >
 > The following reply was made to PR toolchain/58737; it has been noted by GNATS.
 >
 > From: "David H. Gutteridge" <david@gutteridge.ca>
 > To: gnats-bugs@netbsd.org
 > Cc:
 > Subject: Re: toolchain/58737 (building against gcc (GCC) 14.2.1 20240912 (Red
 >  Hat 14.2.1-3) + glibc-2.39-22.fc40.x86_64)
 > Date: Wed, 11 Dec 2024 11:22:17 -0500
 >
 >  This is effectively a duplicate of PR lib/58674. The changes and
 >  discussion were tracked there. Everything referenced in this ticket has
 >  either been pulled up to the netbsd-10 branch or pullup tickets have
 >  been submitted but not yet processed (as they will probably go in after
 >  branching 10.1).
 >
 >  The only outstanding question (not referenced so far in this PR, but on
 >  tech-toolchain) is this:
 >  https://mail-index.netbsd.org/tech-toolchain/2024/10/15/msg004547.html
 >
 >  It's unclear from that commentary at what stage the failure happens
 >  (i.e. in a tools build or after that point).
 >
 >  Dave
 >

From: "David H. Gutteridge" <david@gutteridge.ca>
To: gnats-bugs@netbsd.org
Cc: 
Subject: Re: toolchain/58737 (building against gcc (GCC) 14.2.1 20240912
 (Red Hat 14.2.1-3) + glibc-2.39-22.fc40.x86_64)
Date: Wed, 11 Dec 2024 21:39:24 -0500

 Sorry, where I wrote

 > It's unclear from that commentary at what stage the failure happens
 > (i.e. in a tools build or after that point).

 I meant the reproduction steps weren't documented. If I run
 build.sh -U -m evbarm64 -j 8 tools
 for example, I can't reproduce the issue (on Fedora 40). That component
 isn't built in that context.

 So you could be referring to a non-standard tools target build, or a
 later (distribution) build stage. I'm guessing the latter. Still a tool
 issue but not a "tools build" issue. It helps to provide specific
 details to make it easy for someone else to look into, and for
 documentation's sake.

 Dave

From: "David H. Gutteridge" <david@gutteridge.ca>
To: gnats-bugs@netbsd.org
Cc: 
Subject: Re: toolchain/58737 (building against gcc (GCC) 14.2.1 20240912 (Red
 Hat 14.2.1-3) + glibc-2.39-22.fc40.x86_64)
Date: Thu, 12 Dec 2024 12:42:03 -0500

 On 2024-12-11 21:39, David H. Gutteridge wrote:
 > Sorry, where I wrote
 > 
 >> It's unclear from that commentary at what stage the failure happens
 >> (i.e. in a tools build or after that point).
 > 
 > I meant the reproduction steps weren't documented. If I run
 > build.sh -U -m evbarm64 -j 8 tools
 > for example, I can't reproduce the issue (on Fedora 40). That component
 > isn't built in that context.
 > 
 > So you could be referring to a non-standard tools target build, or a
 > later (distribution) build stage. I'm guessing the latter. Still a tool
 > issue but not a "tools build" issue. It helps to provide specific
 > details to make it easy for someone else to look into, and for
 > documentation's sake.

 FWIW, I'm not able to reproduce this on Fedora 40 with
 ./build.sh -U -m evbarm64 -j 8 release
 either. So either steps aren't clear, or this is perhaps a local issue.

 Dave

From: Andrew Cagney <andrew.cagney@gmail.com>
To: gnats-bugs@netbsd.org
Cc: toolchain-manager@netbsd.org, gnats-admin@netbsd.org, 
	netbsd-bugs@netbsd.org
Subject: Re: toolchain/58737 (building against gcc (GCC) 14.2.1 20240912 (Red
 Hat 14.2.1-3) + glibc-2.39-22.fc40.x86_64)
Date: Fri, 13 Dec 2024 09:02:12 -0500

 >  FWIW, I'm not able to reproduce this on Fedora 40 with

 Ditto, since reporting the problem, many things have changed so,
 presumably one of them made it work:

      build.sh command:    ./build.sh -m evbarm64 -u -U -x -X
 /home/netbsd/10/xsrc -O /home/netbsd/10/evbarm64 -T
 /home/netbsd/10/tools release install-image live-image
      build.sh started:    Thu Dec 12 13:25:24 EST 2024
      NetBSD version:      10.0_STABLE
      MACHINE:             evbarm
      MACHINE_ARCH:        aarch64
      Build platform:      Linux 6.11.11-200.fc40.x86_64 x86_64
      HOST_SH:             /usr/bin/sh
      No $TOOLDIR/bin/nbmake, needs building.
      Bootstrapping nbmake
      MAKECONF file:       /etc/mk.conf (File not found)
      TOOLDIR path:        /home/netbsd/10/tools
      DESTDIR path:        /home/netbsd/10/evbarm64/destdir.evbarm
      RELEASEDIR path:     /home/netbsd/10/evbarm64/releasedir
      Created /home/netbsd/10/tools/bin/nbmake
      Updated makewrapper: /home/netbsd/10/tools/bin/nbmake-evbarm64
      Successful make release
      Successful make install-image
      Successful make live-image
      build.sh ended:      Thu Dec 12 23:45:32 EST 2024

From: "David H. Gutteridge" <david@gutteridge.ca>
To: gnats-bugs@netbsd.org
Cc: 
Subject: Re: toolchain/58737 (building against gcc (GCC) 14.2.1 20240912 (Red
 Hat 14.2.1-3) + glibc-2.39-22.fc40.x86_64)
Date: Sun, 15 Dec 2024 18:03:29 -0500

 I'd thought there were partial LLVM component builds even without
 building X sets, but I see that's not the case. MKLLVMRT is only set
 with X sets (by default). I tried to replicate including X builds, and
 had no issues. I think this can be closed (as can PR lib/58674, now).

 Dave

State-Changed-From-To: feedback->closed
State-Changed-By: gutteridge@NetBSD.org
State-Changed-When: Sun, 15 Dec 2024 23:10:49 +0000
State-Changed-Why:
Most issues resolved. No one can reproduce the last issue discussed
anymore. Thanks for the PR (and references on tech-toolchain).

From: Andrew Cagney <andrew.cagney@gmail.com>
To: gnats-bugs@netbsd.org
Cc: toolchain-manager@netbsd.org, netbsd-bugs@netbsd.org, 
	gnats-admin@netbsd.org, gutteridge@netbsd.org
Subject: Re: toolchain/58737 (building against gcc (GCC) 14.2.1 20240912 (Red
 Hat 14.2.1-3) + glibc-2.39-22.fc40.x86_64)
Date: Sun, 15 Dec 2024 18:18:45 -0500

 thanks, I guess I can go back to checking one of my other bugs to see
 if it's fixed (just need to remember which :-^}

 On Sun, 15 Dec 2024 at 18:10, <gutteridge@netbsd.org> wrote:
 >
 > Synopsis: building against gcc (GCC) 14.2.1 20240912 (Red Hat 14.2.1-3) + glibc-2.39-22.fc40.x86_64
 >
 > State-Changed-From-To: feedback->closed
 > State-Changed-By: gutteridge@NetBSD.org
 > State-Changed-When: Sun, 15 Dec 2024 23:10:49 +0000
 > State-Changed-Why:
 > Most issues resolved. No one can reproduce the last issue discussed
 > anymore. Thanks for the PR (and references on tech-toolchain).
 >
 >

>Unformatted:

Home	PR Database Search