NetBSD Problem Report #57425

From www@netbsd.org  Sat May 20 19:22:45 2023
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))
	(Client CN "mail.NetBSD.org", Issuer "mail.NetBSD.org CA" (not verified))
	by mollari.NetBSD.org (Postfix) with ESMTPS id EF5CB1A9241
	for <gnats-bugs@gnats.NetBSD.org>; Sat, 20 May 2023 19:22:44 +0000 (UTC)
Message-Id: <20230520192243.191B61A9242@mollari.NetBSD.org>
Date: Sat, 20 May 2023 19:22:43 +0000 (UTC)
From: brandon@burn.net
Reply-To: brandon@burn.net
To: gnats-bugs@NetBSD.org
Subject: gettext-tools build errors out on >= 2023Q1 for sparc
X-Send-Pr-Version: www-1.0

>Number:         57425
>Category:       pkg
>Synopsis:       gettext-tools build errors out on >= 2023Q1 for sparc
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    pkg-manager
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Sat May 20 19:25:00 +0000 2023
>Last-Modified:  Wed Jun 28 17:55:01 +0000 2023
>Originator:     Brandon Applegate
>Release:        9.3 / 9.0_2023Q1
>Organization:
>Environment:
NetBSD qemu.internal.burn.net 9.3 NetBSD 9.3 (GENERIC.MP) #0: Thu Aug  4 15:30:37 UTC 2022  mkrepro@mkrepro.NetBSD.org:/usr/src/sys/arch/sparc/compile/GENERIC.MP sparc
>Description:
Trying to build gettext-tools on sparc on 9.0_2023Q1 or newer (I just tried a cvs pull the other day - same issue) results in it erroring out.  It builds fine on 9.0_2022Q4 however.

I have more information in this mailing list post I made:

https://mail-index.netbsd.org/pkgsrc-users/2023/05/12/msg037280.html

If you read that message - you will see 2 links of other folks (however running macOS / fink / brew) which seem to be hitting the same issue.

I was actually trying to build bash - of which gettext-tools was pulled in as a depend.  I believe gettext-tools is a depend on many things - so I would imagine this will cause many packages to fail to build.

The file causing this seems to be ftello.c.  Here is a snippet/beginning of the error:

gettext-tools/work/.buildlink/include -O2 -D_FORTIFY_SOURCE=2 -c ftello.c  -fPIC -DPIC -o .libs/
ftello.o
ftello.c: In function 'rpl_ftello':
ftello.c:122:12: error: 'FILE {aka struct __sFILE}' has no member named '_flag'; did you mean '_
flags'?
   if (fp_->_flag & _IOWRT)
            ^~~~~
            _flags
ftello.c:122:20: error: '_IOWRT' undeclared (first use in this function); did you mean '_IOFBF'?
   if (fp_->_flag & _IOWRT)
                    ^~~~~~
                    _IOFBF
ftello.c:122:20: note: each undeclared identifier is reported only once for each function it app
ears in
ftello.c:133:21: error: 'FILE {aka struct __sFILE}' has no member named '_flag'; did you mean '_
flags'?
           if ((fp_->_flag & _IONBF) == 0 && fp_->_base != NULL)
                     ^~~~~
                     _flags
>How-To-Repeat:
Attempt build of gettext-tools on sparc on pkgsrc 9.0_2023Q1 or later.
>Fix:

>Audit-Trail:
From: Joerg Sonnenberger <joerg@bec.de>
To: gnats-bugs@netbsd.org
Cc: pkg-manager@netbsd.org, gnats-admin@netbsd.org, pkgsrc-bugs@netbsd.org
Subject: Re: pkg/57425: gettext-tools build errors out on >= 2023Q1 for sparc
Date: Mon, 22 May 2023 00:52:45 +0200

 On Sat, May 20, 2023 at 07:25:00PM +0000, brandon@burn.net wrote:
 > >How-To-Repeat:
 > Attempt build of gettext-tools on sparc on pkgsrc 9.0_2023Q1 or later.

 Please check config.log for why it thinks that fseeko/ftello are not
 usable.

 Joerg

From: Brandon Applegate <brandon@burn.net>
To: gnats-bugs@netbsd.org
Cc: 
Subject: Re: pkg/57425: gettext-tools build errors out on >= 2023Q1 for sparc
Date: Sun, 21 May 2023 20:37:51 -0400

 --Apple-Mail=_FBD687A3-7134-4F24-857E-AEBCD90D7629
 Content-Transfer-Encoding: quoted-printable
 Content-Type: text/plain;
 	charset=utf-8

 Hmm, assuming I have the correct config.log - doesn=E2=80=99t look like =
 there is anything for ftello/fseek:

 qemu# pwd
 /usr/pkgsrc/devel/gettext-tools/work/gettext-0.21.1
 qemu# egrep -i 'ftell|fseek=E2=80=99 config.log
 qemu#

 I did find some more here:

 qemu# pwd
 /usr/pkgsrc/devel/gettext-tools/work/gettext-0.21.1/gettext-tools

 configure:34719: checking whether ftello is declared
 configure:34719: gcc -mcpu=3Dv9 -c -O2 -D_FORTIFY_SOURCE=3D2 =
 -I/usr/pkg/include -I/usr/include  -I/usr/pkg/include -I/usr/include =
 conftest.c >&5
 configure:34719: $? =3D 0
 configure:34719: result: yes

 configure:34834: checking for ftello
 configure:34852: gcc -mcpu=3Dv9 -o conftest -O2 -D_FORTIFY_SOURCE=3D2 =
 -I/usr/pkg/include -I/usr/include -I/usr/pkg/include -I/usr/include =
 -L/usr/pkg/lib -Wl,-R/usr/pkg/lib -L/usr/lib -Wl,-R/usr/lib conftest.c =
 -lintl >&5
 configure:34852: $? =3D 0
 configure:34862: result: yes
 configure:34875: checking whether ftello works
 configure:34955: gcc -mcpu=3Dv9 -o conftest -O2 -D_FORTIFY_SOURCE=3D2 =
 -I/usr/pkg/include -I/usr/include -I/usr/pkg/include -I/usr/include =
 -L/usr/pkg/lib -Wl,-R/usr/pkg/lib -L/usr/lib -Wl,-R/usr/lib conftest.c =
 -lintl >&5
 configure:34955: $? =3D 0
 configure:34955: ./conftest
 [1]   Illegal instruction (core dumped) ./conftest${ac_exeext}
 configure:34955: $? =3D 132
 configure: program exited with status 132
 configure: failed program was:

 I notice it=E2=80=99s doing =E2=80=9Cgcc -mcpu=3Dv9=E2=80=9D.  So it=E2=80=
 =99s trying to compile for 64 bit / >=3D Ultra it seems...

 On my other image that=E2=80=99s 2022Q4 snapshot - it=E2=80=99s not =
 doing the v9 cpu stuff:

 configure:27846: checking whether ftello is declared
 configure:27846: gcc -c -O2 -D_FORTIFY_SOURCE=3D2 -I/usr/pkg/include =
 -I/usr/include -I/usr/pkg/include -I/usr/include conftest.c >&5
 configure:27846: $? =3D 0
 configure:27846: result: yes
 configure:27870: checking for ftello
 configure:27887: gcc -o conftest -O2 -D_FORTIFY_SOURCE=3D2 =
 -I/usr/pkg/include -I/usr/include -I/usr/pkg/include -I/usr/include =
 -L/usr/pkg/lib -Wl,-R/usr/pkg/lib -L/usr/lib -Wl,-R/usr/lib conftest.c =
 -lintl >&5
 configure:27887: $? =3D 0
 configure:27896: result: yes
 configure:27909: checking whether ftello works
 configure:27987: gcc -o conftest -O2 -D_FORTIFY_SOURCE=3D2 =
 -I/usr/pkg/include -I/usr/include -I/usr/pkg/include -I/usr/include =
 -L/usr/pkg/lib -Wl,-R/usr/pkg/lib -L/usr/lib -Wl,-R/usr/lib conftest.c =
 -lintl >&5
 configure:27987: $? =3D 0
 configure:27987: ./conftest
 configure:27987: $? =3D 0
 configure:27998: result: yes

 Going down that rabbit hole:

 qemu$ # 2022Q4
 qemu$ pwd
 /usr/pkgsrc/devel/gettext-tools
 qemu$ grep -r mcpu=3Dv9 * | wc -l
        5

 qemu$ # CVS pull a few days ago
 qemu$ pwd
 /usr/pkgsrc/devel/gettext-tools
 qemu$ grep -r mcpu=3Dv9 * | wc -l
     2614

 I think the 5 hits on 20222Q4 are simply from gcc and related binaries =
 matching this string in themselves.

 Both of these environments still have everything built within the tree =
 (i.e. no =E2=80=9Cmake clean=E2=80=9D has been ran).

 Please let me know if there is something else you need me to provide (or =
 if I=E2=80=99ve provided the wrong information).  Thanks.







 --Apple-Mail=_FBD687A3-7134-4F24-857E-AEBCD90D7629
 Content-Transfer-Encoding: 7bit
 Content-Disposition: attachment;
 	filename=signature.asc
 Content-Type: application/pgp-signature;
 	name=signature.asc
 Content-Description: Message signed with OpenPGP

 -----BEGIN PGP SIGNATURE-----

 iQIzBAEBCgAdFiEEBkHShaNvUzpz5SVBSSBTPMYWcDoFAmRquV8ACgkQSSBTPMYW
 cDqRyw//Ud0SUGOHhQZap0Hdey62v9DUqc9+Sa9rp1g48LSNrK0vf6jayHNGAuOu
 7v7NulGSn54KB/v5ZzVc29uKsn9e2HtIO+MO7kegdhdjZrfMYcB0NrixAJpNwKWr
 LMgYO27CsxkEwl6luEPZbpE+5iN7PwVSta0k2T/mBDEVSvNoUhz3xhL0dljTr0gX
 eKZpxAk++tFTCBsraLPHf/3bARU/kWfGg/v6NsBbyWzp5RoKahrHxdZA+0p1xx0Z
 gAoQy2aRYYgbrirDXv7OH8OerW2TBTcR6MCCyo3gcEOSIlIK2kjF+VZ6pgGqcbc/
 dtLA/eaasKsDCsrhMqG3f1l1yoOtD3/5wwW50/h6m5KAFR9Qg+QlShwD7ZBwpvQ6
 G350TDqaOV47RQKo+4Gb81DHIuUpORceo0acBp70KnujV+qcxmQxDJwNt6EvHArY
 ixIQMz0OkiaerBlPv/qKk40ZPGcQSsi8lVuMa8lhuexHbh9Rg078aUNYykVVFq5m
 eabrSir9YFVSBvGAZAZC0u/qBuI/QZboKspsFLVVEfUMxbfk8M8G7YlP+tmL5DHZ
 Q7lLj8pLu8jJ/HJ/jdeGRIupCUKNdR7FR5GACw69wi+aNvwScE01zVD9W+7jLdsA
 8ANaAwmIdTs03X2WY61hemv3zEDWK5RMacAm2ZPI29/8rvbuoUk=
 =CyfL
 -----END PGP SIGNATURE-----

 --Apple-Mail=_FBD687A3-7134-4F24-857E-AEBCD90D7629--

From: Brandon Applegate <brandon@burn.net>
To: gnats-bugs@netbsd.org
Cc: 
Subject: Re: pkg/57425: gettext-tools build errors out on >= 2023Q1 for
 sparc
Date: Wed, 24 May 2023 10:44:20 -0400 (EDT)

 Bit of an update on this, see this thread I started on the bug-gettext ML:

 https://lists.gnu.org/archive/html/bug-gettext/2023-05/msg00042.html

 Seems like this is manifesting on QEMU due to some emultated CPU functions 
 (the lack thereof).  I assume this still builds on real hardware as 
 evidenced by the presence of pre-built packages out on the mirrors.

 One of the main devs. posted a patch in that thread, hopefully that makes 
 it into the official source tree as well.

From: Brandon Applegate <brandon@burn.net>
To: "gnats-bugs@netbsd.org" <gnats-bugs@NetBSD.org>
Cc: 
Subject: Re: pkg/57425
Date: Wed, 28 Jun 2023 13:52:32 -0400

 gettext 0.22 was released ~ 1 week ago.

 https://savannah.gnu.org/news/?id=3D10378

 I believe the fix for this issue is in that release.

 Could we get this bumped in pkgsrc ?  I just grabbed the 2023Q2 snapshot =
 to check and it=E2=80=99s still using 0.21.1.

 Thanks.=

Home	PR Database Search