NetBSD Problem Report #45773

From www@NetBSD.org  Tue Jan  3 04:32:37 2012
Return-Path: <www@NetBSD.org>
Received: from mail.netbsd.org (mail.netbsd.org [204.152.190.11])
	by www.NetBSD.org (Postfix) with ESMTP id 29D5263D675
	for <gnats-bugs@gnats.NetBSD.org>; Tue,  3 Jan 2012 04:32:37 +0000 (UTC)
Message-Id: <20120103043236.5107363BC35@www.NetBSD.org>
Date: Tue,  3 Jan 2012 04:32:36 +0000 (UTC)
From: pda@freeshell.org
Reply-To: pda@freeshell.org
To: gnats-bugs@NetBSD.org
Subject: Can't build ast-ksh from pkgsrc 11q-3
X-Send-Pr-Version: www-1.0

>Number:         45773
>Category:       pkg
>Synopsis:       Can't build ast-ksh from pkgsrc 11q-3
>Confidential:   no
>Severity:       critical
>Priority:       medium
>Responsible:    hpux-pkg-people
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Tue Jan 03 04:35:01 +0000 2012
>Closed-Date:    
>Last-Modified:  Mon Jul 23 05:45:01 +0000 2012
>Originator:     Paul Ackersviller
>Release:        Just pkgsrc 11q-3
>Organization:
>Environment:
HP-UX tdcndv01 B.11.31 U ia64 1579589996 unlimited-user license

>Description:
I'm mystified by this, I've experienced various different problems, including run-away processes filling up the system's process table. While I can't repeat that now, it's behaving just as strange in a different way. The build appears to go to sleep, as if waiting for input, very soon... static-ast-ksh does the same thing. Following is the tail end of what I see:

+ wrapper_result=0
+ eval /var/tmp/pda/pkgsrc/shells/static-ast-ksh/work/.hp/cc -o /var/tmp/pda/pkgsrc/shells/static-ast-ksh/work/arch/hp.ia64/bin/release release.c +DD64 -I/var/tmp/pda/pkgsrc/shells/static-ast-ksh/work/.buildlink/include -L/var/tmp/pda/pkgsrc/shells/static-ast-ksh/work/.buildlink/lib
+ /var/tmp/pda/pkgsrc/shells/static-ast-ksh/work/.hp/cc -o /var/tmp/pda/pkgsrc/shells/static-ast-ksh/work/arch/hp.ia64/bin/release release.c +DD64 -I/var/tmp/pda/pkgsrc/shells/static-ast-ksh/work/.buildlink/include -L/var/tmp/pda/pkgsrc/shells/static-ast-ksh/work/.buildlink/lib
+ . /var/tmp/pda/pkgsrc/shells/static-ast-ksh/work/.wrapper/tmp/empty
+ [ 0 -eq 0 ]
+ exit 0

>How-To-Repeat:
bmake -C shells/ast-ksh
bmake -C shells/static-ast-ksh
>Fix:

>Release-Note:

>Audit-Trail:

Responsible-Changed-From-To: pkg-manager->hpux-pkg-people
Responsible-Changed-By: obache@NetBSD.org
Responsible-Changed-When: Thu, 26 Apr 2012 13:53:24 +0000
Responsible-Changed-Why:
problem on HP-UX


State-Changed-From-To: open->feedback
State-Changed-By: dholland@NetBSD.org
State-Changed-When: Sun, 17 Jun 2012 19:24:45 +0000
State-Changed-Why:
Similar problems were seen on non-hpux and we think the newer version in
pkgsrc HEAD doesn't do this. Probably worth giving it a try, in any case.


From: Paul Ackersviller <pda@freeshell.org>
To: gnats-bugs@NetBSD.org
Cc: hpux-pkg-people@NetBSD.org, pkgsrc-bugs@NetBSD.org, gnats-admin@NetBSD.org,
        dholland@NetBSD.org
Subject: Re: pkg/45773 (Can't build ast-ksh from pkgsrc 11q-3)
Date: Sun, 8 Jul 2012 20:48:12 +0000

 On Sun, Jun 17, 2012 at 07:24:46PM +0000, dholland@NetBSD.org wrote:
 > Similar problems were seen on non-hpux and we think the newer version in
 > pkgsrc HEAD doesn't do this. Probably worth giving it a try, in any case.

 OK, I've tried the latest and and does look better... but I can't quite
 get all the way through a build.  These errors look to be more like some
 kind of standards-confomance disagreement between ksh and HP than
 anything in pkgsrc, but opinions are of course welcome.

 Here's a snippet:

 "/pkgsrc/shells/ast-ksh/work/src/lib/libast/include/ast_getopt.h", line 22: warning #2161-D: unrecognized #pragma
   #pragma prototyped
           ^

 "/usr/include/../include/wchar.h", line 121: error #2020: identifier "__va_list" is undefined
   extern int vfwprintf __((FILE *__restrict, const wchar_t *__restrict, __va_list));
                                                                         ^

 "/usr/include/../include/wchar.h", line 122: error #2020: identifier "__va_list" is undefined
   extern int vwprintf __((const wchar_t *__restrict, __va_list));
                                                      ^

 "/usr/include/../include/wchar.h", line 123: error #2020: identifier "__va_list" is undefined
   extern int vswprintf __((wchar_t *__restrict, size_t, const wchar_t *__restrict, __va_list));
                                                                                    ^

 "/usr/include/../include/wchar.h", line 257: error #2020: identifier "__va_list" is undefined
   extern int vwscanf __((const wchar_t * __restrict, __va_list));
                                                      ^

 "/usr/include/../include/wchar.h", line 258: error #2020: identifier "__va_list" is undefined
   extern int vfwscanf __((FILE * __restrict, const wchar_t * __restrict, __va_list));
                                                                          ^

 "/usr/include/../include/wchar.h", line 259: error #2020: identifier "__va_list" is undefined
   extern int vswscanf __((const wchar_t * __restrict, const wchar_t * __restrict, __va_list));
                                                                                   ^

From: David Holland <dholland-pbugs@netbsd.org>
To: gnats-bugs@netbsd.org
Cc: 
Subject: Re: pkg/45773 (Can't build ast-ksh from pkgsrc 11q-3)
Date: Mon, 23 Jul 2012 05:29:56 +0000

 not sent to gnats (if you reply to your own posting you often have to
 explicitly change the To: to gnats-bugs)

    ------

 From: Paul Ackersviller <pda@freeshell.org>
 To: hpux-pkg-people@NetBSD.org, gnats-admin@NetBSD.org, pkgsrc-bugs@NetBSD.org
 Subject: Re: pkg/45773 (Can't build ast-ksh from pkgsrc 11q-3)
 Date: Mon, 23 Jul 2012 03:10:26 +0000

 On Sun, Jul 08, 2012 at 08:50:04PM +0000, Paul Ackersviller wrote:
 > The following reply was made to PR pkg/45773; it has been noted by GNATS.
 > 
 > From: Paul Ackersviller <pda@freeshell.org>
 > 
 >  On Sun, Jun 17, 2012 at 07:24:46PM +0000, dholland@NetBSD.org wrote:
 >  > Similar problems were seen on non-hpux and we think the newer version in
 >  > pkgsrc HEAD doesn't do this. Probably worth giving it a try, in any case.

 I did eventually get this built, mostly thanks to a silly -D__va_list=va_list,
 then removing a bunch of libraries that I don't have or get.  This was very
 painful due to the opaque, unusual build.  Does anyone know a better way of
 affecting it, other than hacking in a script in place of the cc soft link?

 For the record, the linker flags I need to remove are -lfsg -ljobs -lsecdb -li
 ...I suppose the above can be closed, unless any of the above has to do with
 pkgsrc rather than upstream.

 P.S.  I thought perhaps static-ast-ksh might be a little better in
 respect to the above, but had completely different problems.  One is
 that the Makefile adds in -static, but that's only a gcc flag.  With
 HP's compiler, it should be -minshared instead.

State-Changed-From-To: feedback->open
State-Changed-By: dholland@NetBSD.org
State-Changed-When: Mon, 23 Jul 2012 05:33:49 +0000
State-Changed-Why:
feedback received
(sorry I apparently didn't notice your mail on july 9)


From: David Holland <dholland-pbugs@netbsd.org>
To: gnats-bugs@netbsd.org
Cc: 
Subject: Re: pkg/45773 (Can't build ast-ksh from pkgsrc 11q-3)
Date: Mon, 23 Jul 2012 05:41:07 +0000

  >  I did eventually get this built, mostly thanks to a silly
  >  -D__va_list=va_list,
  >  then removing a bunch of libraries that I don't have or get.  This was very
  >  painful due to the opaque, unusual build.  Does anyone know a better way of
  >  affecting it, other than hacking in a script in place of the cc soft link?

 You can probably set

 _WRAP_EXTRA_ARGS.CC=-D__va_list=va_list

 Next best is to patch the build system so you can feed in compiler
 flags in the environment, and then add what you need to BUILD_ENV.

  >  For the record, the linker flags I need to remove are -lfsg -ljobs -lsecdb -li
  >  ...I suppose the above can be closed, unless any of the above has to do with
  >  pkgsrc rather than upstream.

 and for these,

 BUILDLINK_TRANSFORM+=rm:-lfsg
 BUILDLINK_TRANSFORM+=rm:-ljobs
 BUILDLINK_TRANSFORM+=rm:-lsecdb

 should do it.

 pkgsrc shouldn't be causing any of this but it should be able to help
 work around it.

  >  P.S.  I thought perhaps static-ast-ksh might be a little better in
  >  respect to the above, but had completely different problems.  One is
  >  that the Makefile adds in -static, but that's only a gcc flag.  With
  >  HP's compiler, it should be -minshared instead.

 The cc wrapper should probably be taking care of this; if it's not, it
 should be taught to. However, I don't know how.

 At the moment static-ast-ksh doesn't build even on NetBSD so it may
 not be worth trying to fight with it.

 -- 
 David A. Holland
 dholland@netbsd.org

>Unformatted:

NetBSD Home
NetBSD PR Database Search

(Contact us) $NetBSD: query-full-pr,v 1.39 2013/11/01 18:47:49 spz Exp $
$NetBSD: gnats_config.sh,v 1.8 2006/05/07 09:23:38 tsutsui Exp $
Copyright © 1994-2007 The NetBSD Foundation, Inc. ALL RIGHTS RESERVED.