NetBSD Problem Report #44577
From cheusov@tut.by Tue Feb 15 14:32:50 2011
Return-Path: <cheusov@tut.by>
Received: from mail.netbsd.org (mail.netbsd.org [204.152.190.11])
by www.NetBSD.org (Postfix) with ESMTP id C327163B100
for <gnats-bugs@gnats.netbsd.org>; Tue, 15 Feb 2011 14:32:50 +0000 (UTC)
Message-Id: <87vd0l74tc.fsf@asrock.chizhovka.net>
Date: Tue, 15 Feb 2011 16:25:35 +0200
From: cheusov@tut.by
To: gnats-bugs@gnats.NetBSD.org
Subject: audio/pulseaudio fails to build with modular XOrg
X-Send-Pr-Version: 3.95
>Number: 44577
>Category: pkg
>Synopsis: audio/pulseaudio fails to build with modular XOrg
>Confidential: no
>Severity: serious
>Priority: medium
>Responsible: pkg-manager
>State: closed
>Class: sw-bug
>Submitter-Id: net
>Arrival-Date: Tue Feb 15 14:35:00 +0000 2011
>Closed-Date: Thu Nov 17 13:04:57 +0000 2011
>Last-Modified: Thu Nov 17 13:05:02 +0000 2011
>Originator: Aleksey Cheusov
>Release: NetBSD 5.1
>Organization:
>Environment:
System: NetBSD asrock.chizhovka.net 5.1 NetBSD 5.1 (GENERIC) #0: Mon Jan 31 23:53:52 EET 2011 cheusov@asrock.chizhovka.net:/srv/obj/sys/arch/amd64/compile/GENERIC amd64
Architecture: x86_64
Machine: amd64
>Description:
Full log is here
http://mova.org/~cheusov/pub/pkgsrc-distbb/NetBSD/current/logs/20110213.2304/audio_pulseaudio_pulseaudio-0.9.21nb7/stage_Package.html
Short log is below
=> Automatic manual page handling
=> Generating post-install file lists
pkg_create: can't stat `/tmp/obj-pkgsrc/audio/pulseaudio/work/.destdir/usr/pkg/lib/pulse-0.9.21/modules/libraop.a'
pkg_create: can't stat `/tmp/obj-pkgsrc/audio/pulseaudio/work/.destdir/usr/pkg/lib/pulse-0.9.21/modules/libraop.so'
pkg_create: can't stat `/tmp/obj-pkgsrc/audio/pulseaudio/work/.destdir/usr/pkg/lib/pulse-0.9.21/modules/module-raop-sink.a'
pkg_create: can't stat `/tmp/obj-pkgsrc/audio/pulseaudio/work/.destdir/usr/pkg/lib/pulse-0.9.21/modules/module-raop-sink.so'
=> Checking file-check results for pulseaudio-0.9.21nb7
ERROR: ************************************************************
ERROR: The following files are in the PLIST but not in /usr/pkg:
ERROR: /tmp/obj-pkgsrc/audio/pulseaudio/work/.destdir/usr/pkg/lib/pulse-0.9.21/modules/libraop.a
ERROR: /tmp/obj-pkgsrc/audio/pulseaudio/work/.destdir/usr/pkg/lib/pulse-0.9.21/modules/libraop.so
ERROR: /tmp/obj-pkgsrc/audio/pulseaudio/work/.destdir/usr/pkg/lib/pulse-0.9.21/modules/module-raop-sink.a
ERROR: /tmp/obj-pkgsrc/audio/pulseaudio/work/.destdir/usr/pkg/lib/pulse-0.9.21/modules/module-raop-sink.so
*** Error code 1
>How-To-Repeat:
X11_TYPE=modular
>Fix:
Unknown
>Release-Note:
>Audit-Trail:
From: "OBATA Akio" <obache@netbsd.org>
To: gnats-bugs@netbsd.org
Cc:
Subject: Re: pkg/44577: audio/pulseaudio fails to build with modular XOrg
Date: Sun, 01 May 2011 19:07:32 +0900
On Tue, 15 Feb 2011 23:35:00 +0900, <cheusov@tut.by> wrote:
> ERROR: ************************************************************
> ERROR: The following files are in the PLIST but not in /usr/pkg:
> ERROR: /tmp/obj-pkgsrc/audio/pulseaudio/work/.destdir/usr/pkg/lib/pulse-0.9.21/modules/libraop.a
> ERROR: /tmp/obj-pkgsrc/audio/pulseaudio/work/.destdir/usr/pkg/lib/pulse-0.9.21/modules/libraop.so
> ERROR: /tmp/obj-pkgsrc/audio/pulseaudio/work/.destdir/usr/pkg/lib/pulse-0.9.21/modules/module-raop-sink.a
> ERROR: /tmp/obj-pkgsrc/audio/pulseaudio/work/.destdir/usr/pkg/lib/pulse-0.9.21/modules/module-raop-sink.so
> *** Error code 1
>
>> How-To-Repeat:
> X11_TYPE=modular
It is not came from X11_TYPE, but failed to detect openssl.
http://mova.org/~cheusov/pub/pkgsrc-distbb/NetBSD/current/logs/20110213.2304/audio_pulseaudio_pulseaudio-0.9.21nb7/stage_configure.html
checking for OPENSSL... no
Please check why it is happened for you.
--
OBATA Akio / obache@NetBSD.org
State-Changed-From-To: open->feedback
State-Changed-By: obache@NetBSD.org
State-Changed-When: Mon, 02 May 2011 11:42:39 +0000
State-Changed-Why:
requested feedback.
From: Aleksey Cheusov <cheusov@tut.by>
To: gnats-bugs@NetBSD.org
Cc:
Subject: Re: pkg/44577: audio/pulseaudio fails to build with modular XOrg
Date: Sun, 08 May 2011 11:29:33 +0300
> On Tue, 15 Feb 2011 23:35:00 +0900, <cheusov@tut.by> wrote:
>
>> ERROR: ************************************************************
>> ERROR: The following files are in the PLIST but not in /usr/pkg:
>> ERROR: /tmp/obj-pkgsrc/audio/pulseaudio/work/.destdir/usr/pkg/lib/pulse-0.9.21/modules/libraop.a
>> ERROR: /tmp/obj-pkgsrc/audio/pulseaudio/work/.destdir/usr/pkg/lib/pulse-0.9.21/modules/libraop.so
>> ERROR: /tmp/obj-pkgsrc/audio/pulseaudio/work/.destdir/usr/pkg/lib/pulse-0.9.21/modules/module-raop-sink.a
>> ERROR: /tmp/obj-pkgsrc/audio/pulseaudio/work/.destdir/usr/pkg/lib/pulse-0.9.21/modules/module-raop-sink.so
>> *** Error code 1
>>
>>> How-To-Repeat:
>> X11_TYPE=modular
>
> It is not came from X11_TYPE, but failed to detect openssl.
>
> http://mova.org/~cheusov/pub/pkgsrc-distbb/NetBSD/current/logs/20110213.2304/audio_pulseaudio_pulseaudio-0.9.21nb7/stage_configure.html
>
> checking for OPENSSL... no
>
> Please check why it is happened for you.
# make patch DEPENDS_TARGET=do-bin-install
# /tmp/obj-pkgsrc/audio/pulseaudio/work/.tools/bin/pkg-config --exists --print-errors " openssl > 0.9 "
Requested 'openssl > 0.9' but version of OpenSSL is
# echo $?
1
# cat /tmp/obj-pkgsrc/audio/pulseaudio/work/.tools/bin/pkg-config
#!/bin/sh
wrapperlog="${TOOLS_WRAPPER_LOG-/tmp/obj-pkgsrc/audio/pulseaudio/work/.work.log}"
echo "[*] "/tmp/obj-pkgsrc/audio/pulseaudio/work/.tools/bin/pkg-config" $@" >> $wrapperlog
echo "<.> PKG_CONFIG_LIBDIR=/tmp/obj-pkgsrc/audio/pulseaudio/work/.buildlink/lib/pkgconfig:/tmp/obj-pkgsrc/audio/pulseaudio/work/.buildlink/share/pkgconfig PKG_CONFIG_PATH= /usr/pkg/bin/pkg-config "$@"" >> $wrapperlog
PKG_CONFIG_LIBDIR=/tmp/obj-pkgsrc/audio/pulseaudio/work/.buildlink/lib/pkgconfig:/tmp/obj-pkgsrc/audio/pulseaudio/work/.buildlink/share/pkgconfig PKG_CONFIG_PATH= /usr/pkg/bin/pkg-config "$@"
# cat /tmp/obj-pkgsrc/audio/pulseaudio/work/.buildlink/lib/pkgconfig/openssl.pc
Name: OpenSSL
Description: Secure Sockets Layer and cryptography libraries and tools
Version:
Libs: -L/usr/lib -lssl -lcrypto
Cflags: -I/usr/include
#
--
Best regards, Aleksey Cheusov.
From: "OBATA Akio" <obache@netbsd.org>
To: gnats-bugs@netbsd.org
Cc:
Subject: Re: pkg/44577: audio/pulseaudio fails to build with modular XOrg
Date: Sun, 08 May 2011 18:02:55 +0900
On Sun, 08 May 2011 17:30:04 +0900, Aleksey Cheusov <cheusov@tut.by> wrote:
> # cat /tmp/obj-pkgsrc/audio/pulseaudio/work/.buildlink/lib/pkgconfig/openssl.pc
> Name: OpenSSL
> Description: Secure Sockets Layer and cryptography libraries and tools
> Version:
> Libs: -L/usr/lib -lssl -lcrypto
> Cflags: -I/usr/include
I don't know why, but it is because failed to detect version of your builtin OpenSSL
(it's ${BUILTIN_VERSION.openssl} in pkgsrc/security/openssl/builtin.mk).
--
OBATA Akio / obache@NetBSD.org
From: John Marino <netbsd@marino.st>
To: gnats-bugs@NetBSD.org
Cc:
Subject: Re: pkg/44577 audio/pulseaudio openssl detection failure
Date: Fri, 28 Oct 2011 23:08:11 +0200
I think there may be something to this PR.
One of the packages I was building inside tinderbox-dragonfly pulled in
audio/pulseaudio, but stop with very similar error messages as seen at
the top of the PR.
Tinderbox is chrooted. I don't think a standard DragonFly installation
runs into this issue. It appears to me that pkgsrc mk scripts are
detecting that openssl is on the system just fine, but the pulseaudio
configure script isn't able to.
It seems to mainly rely on pkg-config. There aren't any openssl.pc
files for the /usr/bin/openssl executable, so I am I not sure how that
this is detected ok on a standard system.
I was only able to solve the problem by using PREFER_PKGSRC option to
tell it that pkgsrc's openssl needed to get pulled in.
In a related topic, pkg/43420 questions why so many dependencies are
required. In this case, openSSL is only needed to support Airtunes. I
would think that this should be a non-suggested option rather than a
mandatory dependency as the author of pkg/43420 proposes.
In any case, I wanted to provide feedback that Aleksey is not the only
one to see this problem in some form and basically confirm that it is a
real issue, at least in some configurations.
John
From: Matthew Mondor <mm_lists@pulsar-zone.net>
To: gnats-bugs@NetBSD.org
Cc:
Subject: Re: pkg/44577 audio/pulseaudio openssl detection failure
Date: Fri, 28 Oct 2011 17:29:03 -0400
On Fri, 28 Oct 2011 21:10:10 +0000 (UTC)
John Marino <netbsd@marino.st> wrote:
> It seems to mainly rely on pkg-config. There aren't any openssl.pc
> files for the /usr/bin/openssl executable, so I am I not sure how that
> this is detected ok on a standard system.
I also quickly looked for a -config script for openssl but didn't find
one either.
> In a related topic, pkg/43420 questions why so many dependencies are
> required. In this case, openSSL is only needed to support Airtunes. I
> would think that this should be a non-suggested option rather than a
> mandatory dependency as the author of pkg/43420 proposes.
I agree that it should then be an option (although that doesn't solve
the openssl buildlink issue).
--
Matt
From: Joerg Sonnenberger <joerg@britannica.bec.de>
To: gnats-bugs@NetBSD.org
Cc:
Subject: Re: pkg/44577 audio/pulseaudio openssl detection failure
Date: Fri, 28 Oct 2011 23:50:21 -0700
On Fri, Oct 28, 2011 at 09:10:10PM +0000, John Marino wrote:
> It seems to mainly rely on pkg-config. There aren't any openssl.pc
> files for the /usr/bin/openssl executable, so I am I not sure how that
> this is detected ok on a standard system.
It should be creating a fake openssl.pc file under .buildlink.
> In any case, I wanted to provide feedback that Aleksey is not the only
> one to see this problem in some form and basically confirm that it is a
> real issue, at least in some configurations.
I can't reproduce it.
Joerg
From: John Marino <netbsd@marino.st>
To: gnats-bugs@NetBSD.org
Cc:
Subject: Re: pkg/44577 audio/pulseaudio and openssl
Date: Tue, 08 Nov 2011 13:56:06 +0100
I tried, but I couldn't reproduce this either on a standard pkgsrc setup.
The environment that I can see it is within the tinderbox chroot jail.
I've tracked the issue to security/openssl/builtin.mk line 35 where it
tries to define BUILTIN_VERSION.openssl.
The result of that awk statement is blank within the chroot jail. The
fake openssl.pc file is getting created, but it looks something like:
Name: OpenSSL
Description: Secure Sockets Layer and cryptography libraries and tools
Version:
Libs: -L/usr/lib -lssl -lcrypto
Cflags: -I/usr/include
I think the makefile can determine the file exists, but it can't read
it. There are a number of null mounts within the setup, so their
configuration might be the culprit in my case. Looking back through the
history of this PR, that's what happened to Aleksey as well, but likely
for a different reason.
Maybe the builtin.mk can be beefed up a bit such that if
BUILIN_VERSION.openssl evaluates to blank then IS_BUILTIN.openssl gets
set to no?
Maybe that's treating the symptoms, but at least it would be more
robust. I'm not married to the idea, it's just a though.
John
From: John Marino <marino@netbsd.org>
To: gnats-bugs@NetBSD.org
Cc:
Subject: Re: pkg/44577 openssl builtin detection, getting closer
Date: Mon, 14 Nov 2011 01:32:36 +0100
I know which line of security/openssl/builtin.mk is causing openssl not
to be detected by the fake openssl.pc.
.if !defined(BUILTIN_PKG.openssl) && \
!empty(IS_BUILTIN.openssl:M[yY][eE][sS]) && \
empty(H_OPENSSLV:M__nonexistent__)
(line 32) is not evaluating to true because for some reason
BUILT_PKG.openssl is already defined in the two cases reported in this
PR. If I comment out that clause to change the line to:
if !empty(IS_BUILTIN.openssl:M[yY][eE][sS]) && \
empty(H_OPENSSLV:M__nonexistent__)
then the system openssl is detected as expected.
I can't figure out where BUILT_PKG.openssl is previously defined.
Is "!defined(BUILTIN_PKG.openssl)" really required?
Answering these two questions will lead to the solution, I believe.
John
From: David Holland <dholland-pbugs@netbsd.org>
To: gnats-bugs@NetBSD.org
Cc:
Subject: Re: pkg/44577 openssl builtin detection, getting closer
Date: Mon, 14 Nov 2011 01:48:04 +0000
On Mon, Nov 14, 2011 at 12:30:05AM +0000, John Marino wrote:
> I know which line of security/openssl/builtin.mk is causing openssl not
> to be detected by the fake openssl.pc.
>
> .if !defined(BUILTIN_PKG.openssl) && \
> !empty(IS_BUILTIN.openssl:M[yY][eE][sS]) && \
> empty(H_OPENSSLV:M__nonexistent__)
>
> (line 32) is not evaluating to true because for some reason
> BUILT_PKG.openssl is already defined in the two cases reported in this
> PR. If I comment out that clause to change the line to:
>
> if !empty(IS_BUILTIN.openssl:M[yY][eE][sS]) && \
> empty(H_OPENSSLV:M__nonexistent__)
>
> then the system openssl is detected as expected.
>
> I can't figure out where BUILT_PKG.openssl is previously defined.
> Is "!defined(BUILTIN_PKG.openssl)" really required?
>
> Answering these two questions will lead to the solution, I believe.
My first guess would be that the builtin.mk file is being included
twice for some reason.
You can generally track this kind of problem down by inserting
.if defined(IS_BUILTIN.openssl)
.warning "foo 1"
.endif
blocks strategically.
--
David A. Holland
dholland@netbsd.org
From: John Marino <marino@netbsd.org>
To: gnats-bugs@NetBSD.org
Cc:
Subject: Re: pkg/44577 openssl builtin detection, getting closer
Date: Mon, 14 Nov 2011 07:45:06 +0100
I need to clarify that this is happening with EVERY package that sucks
in the openssl buildlink3 (in the environments the issue is seen of
course). It's only noticeable on those packages where openssl is
detected via pkgconfig. The same thing is occurring on security/heimdal
for example, but it doesn't use the faulty openssl.pc file.
The "included twice" theory is a pretty good one, but I still have the
questions:
1) Why not use C header still defines to ignore re-included makefiles?
2) What's the problem with redefining BUILTIN_PKG.openssl? That implies
it's sacred, but the rest of the files can be redefined (albeit
erroneously) e.g. BUILTIN_VERSION.openssl.
3) I'm not sure how your trick is going to help. I'm able to detect the
case that it's redefined already.
I'm testing with "> bmake -V '${BUILTIN_VERSION.openssl}'" after the ">
bmake wrapper" phase. There's no echoes with that, so I'm not seeing
how warnings are going to help?
Looking closer, it seems that half the makefile is blocked off:
###
### The section below only applies if we are not including this file
### solely to determine whether a built-in implementation exists.
###
CHECK_BUILTIN.openssl?= no
.if !empty(CHECK_BUILTIN.openssl:M[nN][oO])
That would imply the openssl file would get created on the first pass
I don't see anywhere else CHECK_BUILTIN.openssl is defined.
From what I can tell, there's no problem with just removing ".if
!defined(BUILTIN_PKG.openssl) && \". Is the original author of this
builtin.mk file around, can we ask him?
John
From: David Holland <dholland-pbugs@netbsd.org>
To: gnats-bugs@NetBSD.org
Cc:
Subject: Re: pkg/44577 openssl builtin detection, getting closer
Date: Mon, 14 Nov 2011 18:04:06 +0000
On Mon, Nov 14, 2011 at 06:45:01AM +0000, John Marino wrote:
> The "included twice" theory is a pretty good one, but I still have the
> questions:
> 1) Why not use C header still defines to ignore re-included makefiles?
> 2) What's the problem with redefining BUILTIN_PKG.openssl? That implies
> it's sacred, but the rest of the files can be redefined (albeit
> erroneously) e.g. BUILTIN_VERSION.openssl.
> 3) I'm not sure how your trick is going to help. I'm able to detect the
> case that it's redefined already.
The question is why it's being included twice and why it's not getting
the right answers the first time. The fact that it isn't means
something's wrong; and if it *were* getting the right answers the
first time it wouldn't need to reset/change the variable.
(Plus, having the value change partway through evaluation of the
makefiles is probably worse than having it be wrong.)
All or almost all builtin.mk files are structured the same way, btw.
--
David A. Holland
dholland@netbsd.org
From: John Marino <marino@netbsd.org>
To: gnats-bugs@NetBSD.org
Cc:
Subject: Re: pkg/44577 openssl builtin detection, getting closer
Date: Mon, 14 Nov 2011 23:17:30 +0100
David Holland:
> The question is why it's being included twice and why it's not getting
> the right answers the first time. The fact that it isn't means
> something's wrong; and if it *were* getting the right answers the
> first time it wouldn't need to reset/change the variable.
>
> (Plus, having the value change partway through evaluation of the
> makefiles is probably worse than having it be wrong.)
>
> All or almost all builtin.mk files are structured the same way, btw.
I confirmed that security/openssl/builtin.mk is not getting included
twice. Besides grep not unearthing any additions includes of this file,
using your ".if" trick, I determine that it's getting run once, and the
first time it's run, BUILTIN_VERSION.openssl is already defined.
I don't know where else to look to see if this variable is getting
defined. AFAICT, it's limited to the files in security/openssl directory.
John
From: John Marino <marino@netbsd.org>
To: gnats-bugs@NetBSD.org
Cc:
Subject: Re: pkg/44577 openssl builtin detection, getting closer
Date: Tue, 15 Nov 2011 08:46:13 +0100
On 11/14/2011 11:17 PM, John Marino wrote:
>
> I confirmed that security/openssl/builtin.mk is not getting included
> twice. Besides grep not unearthing any additions includes of this
> file, using your ".if" trick, I determine that it's getting run once,
> and the first time it's run, BUILTIN_VERSION.openssl is already defined.
>
> I don't know where else to look to see if this variable is getting
> defined. AFAICT, it's limited to the files in security/openssl
> directory.
>
> John
>
By the way, the value of the definition of BUILTIN_PKG.openssl is
"openssl-1.0.0e" which is what it would be if builtin.mk was getting
included twice. Despite that damning evidence, I don't think it is
getting included twice from all the .if/.endif checks I put in to test
the theory.
Is there any kind of variable cache in pkgsrc? Anything that could
explain how this variable is already defined?
John
From: John Marino <marino@netbsd.org>
To: gnats-bugs@NetBSD.org
Cc:
Subject: Re: pkg/44577 openssl builtin detection, getting closer
Date: Wed, 16 Nov 2011 07:53:18 +0100
On 11/15/2011 8:46 AM, John Marino wrote:
>
> By the way, the value of the definition of BUILTIN_PKG.openssl is
> "openssl-1.0.0e" which is what it would be if builtin.mk was getting
> included twice. Despite that damning evidence, I don't think it is
> getting included twice from all the .if/.endif checks I put in to test
> the theory.
>
> Is there any kind of variable cache in pkgsrc? Anything that could
> explain how this variable is already defined?
> John
>
To support my assertion, I replaced security/openssl/builtin.mk with a
dummy file. Later, I replaced
security/openssl/buildlink3.mk with a dummy file. In both cases,
executing "bmake -V '${BUILTIN_PKG.openssl}'
from the directory security/heimdal afterwards returned "openssl-1.0.0e".
I think this proves BUILTIN_PKG.openssl has the possibility of being
defined outside of the openssl builtin logic
in both NetBSD and DragonFly. Can I have some assistance on figuring
out how?
John
From: "OBATA Akio" <obache@netbsd.org>
To: gnats-bugs@netbsd.org
Cc:
Subject: Re: pkg/44577 openssl builtin detection, getting closer
Date: Wed, 16 Nov 2011 19:19:00 +0900
On Wed, 16 Nov 2011 15:50:05 +0900, John Marino <marino@netbsd.org> wrote:
> To support my assertion, I replaced security/openssl/builtin.mk with a
> dummy file. Later, I replaced
> security/openssl/buildlink3.mk with a dummy file. In both cases,
> executing "bmake -V '${BUILTIN_PKG.openssl}'
> from the directory security/heimdal afterwards returned "openssl-1.0.0e".
> I think this proves BUILTIN_PKG.openssl has the possibility of being
> defined outside of the openssl builtin logic
> in both NetBSD and DragonFly. Can I have some assistance on figuring
> out how?
I feel you did not "make clean" after those changes.
See ${WRKDIR}/.*_makevars.mk (and pkgsrc/mk/bsd.makevars.mk).
--
OBATA Akio / obache@NetBSD.org
From: John Marino <netbsd@marino.st>
To: gnats-bugs@NetBSD.org
Cc:
Subject: Re: pkg/44577 openssl builtin detection, getting closer
Date: Wed, 16 Nov 2011 14:19:09 +0100
BINGO!!!
I know exactly what's happening and anybody can reproduce this.
This works:
======================
A) bmake clean
B) bmake wrapper
C) check ${WRKDIR}/lib/pkgconfig/openssl.pc (it's good)
This doesn't work
======================
A) bmake clean
B) bmake patch
C) bmake wrapper
D) check ${WRKDIR}/lib/pkgconfig/openssl.pc (BAD, version undefined)
The variable BUILTIN_PKG.openssl is cached, but BUILTIN_VERSION.openssl
is not.
I would love to hear somebody spin this as a feature and not a
straight-up bug. Executing individual make targets is a legitimate user
action.
Now that the anomaly is understood well, can we discuss how to fix this now?
John
From: "OBATA Akio" <obache@netbsd.org>
To: gnats-bugs@netbsd.org
Cc:
Subject: Re: pkg/44577 openssl builtin detection, getting closer
Date: Wed, 16 Nov 2011 23:03:19 +0900
On Wed, 16 Nov 2011 22:20:04 +0900, John Marino <netbsd@marino.st> wrote:
> The variable BUILTIN_PKG.openssl is cached, but BUILTIN_VERSION.openssl
> is not.
> I would love to hear somebody spin this as a feature and not a
> straight-up bug. Executing individual make targets is a legitimate user
> action.
> Now that the anomaly is understood well, can we discuss how to fix this now?
BUILTIN_VERSION.openssl must be added to MAKEVARS.
How about following patch?
Index: security/openssl/builtin.mk
===================================================================
RCS file: /cvsroot/pkgsrc/security/openssl/builtin.mk,v
retrieving revision 1.30
diff -u -r1.30 builtin.mk
--- security/openssl/builtin.mk 4 Oct 2011 14:15:35 -0000 1.30
+++ security/openssl/builtin.mk 16 Nov 2011 14:02:16 -0000
@@ -104,6 +104,7 @@
BUILTIN_PKG.openssl= openssl-${BUILTIN_VERSION.openssl}
.endif
MAKEVARS+= BUILTIN_PKG.openssl
+MAKEVARS+= BUILTIN_VERSION.openssl
.if !defined(BUILTIN_OPENSSL_HAS_THREADS) && \
!empty(IS_BUILTIN.openssl:M[yY][eE][sS]) && \
--
OBATA Akio / obache@NetBSD.org
From: John Marino <netbsd@marino.st>
To: gnats-bugs@NetBSD.org
Cc:
Subject: Re: pkg/44577 openssl builtin detection, getting closer
Date: Thu, 17 Nov 2011 08:36:33 +0100
OBATA Akio :
> BUILTIN_VERSION.openssl must be added to MAKEVARS.
> How about following patch?
Yes, that patch seems to work fine.
Would you please commit it? I'm on travel for the next few days.
I think this PR could also be closed upon the commit.
Thanks,
John
State-Changed-From-To: feedback->closed
State-Changed-By: obache@NetBSD.org
State-Changed-When: Thu, 17 Nov 2011 13:04:57 +0000
State-Changed-Why:
committed the fix.
From: "OBATA Akio" <obache@netbsd.org>
To: gnats-bugs@gnats.NetBSD.org
Cc:
Subject: PR/44577 CVS commit: pkgsrc/security/openssl
Date: Thu, 17 Nov 2011 13:03:20 +0000
Module Name: pkgsrc
Committed By: obache
Date: Thu Nov 17 13:03:19 UTC 2011
Modified Files:
pkgsrc/security/openssl: builtin.mk
Log Message:
Add BUILTIN_VERSION.openssl to MAKEVARS for later use.
fixes PR pkg/44577.
To generate a diff of this commit:
cvs rdiff -u -r1.30 -r1.31 pkgsrc/security/openssl/builtin.mk
Please note that diffs are not public domain; they are subject to the
copyright notices on the relevant files.
>Unformatted:
(Contact us)
$NetBSD: query-full-pr,v 1.51 2026/08/10 02:28:17 riastradh Exp $
$NetBSD: gnats_config.sh,v 1.10 2026/05/13 22:00:09 riastradh Exp $
Copyright © 1994-2026
The NetBSD Foundation, Inc. ALL RIGHTS RESERVED.