NetBSD Problem Report #41966
From kre@munnari.OZ.AU Tue Sep 1 04:44:51 2009
Return-Path: <kre@munnari.OZ.AU>
Received: from mail.netbsd.org (mail.netbsd.org [204.152.190.11])
by www.NetBSD.org (Postfix) with ESMTP id 5F9E563BAC2
for <gnats-bugs@gnats.NetBSD.org>; Tue, 1 Sep 2009 04:44:51 +0000 (UTC)
Message-Id: <200909010444.n814iiPw027495@jade.coe.psu.ac.th>
Date: Tue, 1 Sep 2009 11:44:44 +0700 (ICT)
From: kre@munnari.OZ.AU
To: gnats-bugs@gnats.NetBSD.org
Subject: x11/qt4-qdbus demands absurd quantities of (v)ram to compile (NetBSD 4/0/i386)
X-Send-Pr-Version: 3.95
>Number: 41966
>Category: pkg
>Synopsis: x11/qt4-qdbus demands absurd quantities of (v)ram to compile (NetBSD 4/0/i386)
>Confidential: no
>Severity: serious
>Priority: medium
>Responsible: adam
>State: closed
>Class: sw-bug
>Submitter-Id: net
>Arrival-Date: Tue Sep 01 04:45:00 +0000 2009
>Closed-Date: Mon May 27 12:39:58 +0000 2019
>Last-Modified: Mon May 27 12:39:58 +0000 2019
>Originator: Robert Elz
>Release: NetBSD 4.0 / i386 pkgsrc -current
>Organization:
Prince of Songkla University
>Environment:
System: NetBSD jade.coe.psu.ac.th 4.0_STABLE NetBSD 4.0_STABLE (JADE-1.696-20080517) #9: Fri May 23 18:55:13 ICT 2008 kre@jade.coe.psu.ac.th:/usr/obj/4/kernels/JADE i386
Architecture: i386
Machine: i386
>Description:
One of the files (generators/mac/pbuilder_pbx.cpp) in x11/qt4-qdbus
has started deciding it needs > 250MB data size to successfully
compile - in qt4-qdbus-4.5.2nb1. It did compile successfully for
me in the same environment, without this kind of problem (ie: no
special ulimit settings or anything) when it was qt4-qdbus-4.5.2
(I last compiled that on July 10)
The only change between 4.5.2 and 4.5.2nb1 was the revbump to
account for the jpeg lib major version change, and I cannot see
how that could be relevant to anything, so I am guessing that some
other change in the environment has caused this.
But I have no idea what - qt4-qdbus does not appear to install
a pkgsrc version of gcc, and the standard NetBSD 4.0 release version
cannot possibly have altered - the source files beig compiled
haven't altered, all that's reallyt left is some .h file installed
by some other package.
Of course, it is probably a gcc fault that it is consuming this
much ram (and hence a toolchain bug) but for all I know that
might already have been fixed - but something in the environment
of this particular package seems to have altgered to trigger the
problem.
>How-To-Repeat:
I use pkg_comp with libkver and NetBSD 4.0 release sets (except x*)
to similate a fairly pure NetBSD 4.0 release environment. Pkgsrc
modular xorg is enabled for X stuff (but that is not likely relevant
here).
With that environment I see ...
c++ -c -o pbuilder_pbx.o -O2 -I/usr/pkg/include -I/usr/pkg/qt4/include -I/usr/include -I/usr/pkg/include/freetype2 -O2 -I/usr/pkg/include -I/usr/pkg/qt4/include -I/usr/include -I/usr/pkg/include/freetype2 -DQ_TRAILING_DIR_SEP -DQMAKE_OPENSOURCE_EDITION -I/usr/pkg/include -I/usr/pkg/qt4/include -I/usr/include -I/usr/pkg/include/freetype2 -I. -Igenerators -Igenerators/unix -Igenerators/win32 -Igenerators/mac -I/pkg_comp/obj/pkgsrc/x11/qt4-qdbus/4x/qt-x11-opensource-src-4.5.2/include -I/pkg_comp/obj/pkgsrc/x11/qt4-qdbus/4x/qt-x11-opensource-src-4.5.2/include/QtCore -I/pkg_comp/obj/pkgsrc/x11/qt4-qdbus/4x/qt-x11-opensource-src-4.5.2/src/corelib/global -I/pkg_comp/obj/pkgsrc/x11/qt4-qdbus/4x/qt-x11-opensource-src-4.5.2/src/script -DQT_NO_PCRE -DQT_BUILD_QMAKE -DQT_BOOTSTRAPPED -DQT_NO_TEXTCODEC -DQT_NO_UNICODETABLES -DQT_NO_COMPONENT -DQT_NO_STL -DQT_NO_COMPRESS -I/pkg_comp/obj/pkgsrc/x11/qt4-qdbus/4x/qt-x11-opensource-src-4.5.2/mkspecs/netbsd-g++ -DHAVE_QCONFIG_CPP -DQT_NO_THRE!
AD -DQT_NO_QOBJECT -DQT_NO_GEOM_VARIANT generators/mac/pbuilder_pbx.cpp
cc1plus: out of memory allocating 15825784 bytes after a total of 252073488 bytes
make: *** [pbuilder_pbx.o] Error 1
*** Error code 2
Stop.
make: stopped in /usr/pkgsrc/x11/qt4-qdbus
*** Error code 1
Stop.
>Fix:
???
>Release-Note:
>Audit-Trail:
Responsible-Changed-From-To: pkg-manager->adam
Responsible-Changed-By: wiz@NetBSD.org
Responsible-Changed-When: Tue, 01 Sep 2009 09:24:00 +0000
Responsible-Changed-Why:
Over to maintainer.
From: Robert Elz <kre@munnari.OZ.AU>
To: gnats-bugs@NetBSD.org
Cc:
Subject: Re: pkg/41966: x11/qt4-qdbus demands absurd quantities of (v)ram to compile (NetBSD 4/0/i386)
Date: Mon, 05 Oct 2009 20:20:10 +0700
Since this has now become considerably more important with the
import of kde4 which needs it (previously nothing that mattered
to anything important seemed to care about this), I thought I'd
see if I could work out how to make it build with some other
compiler than the "gcc 4.1.2 20061021 prerelease" that is the
standard compiler in NetBSD 4.0
To do that, I copied (kind of) some stuff from qt3-tools which
wants a different compiler for sparc - left out the test for
sparc (which clearly isn't relevant to me, I'm on i386) and as
I'm not distributing to anything, left out conditional tests
completely (a genuine fix would probably want to test and only
do this where needed), but by adding the following 3 lines ...
.include "../../mk/bsd.prefs.mk"
GCC_REQD+= 3.4
USE_PKGSRC_GCC= yes
this package got past its demand for huge memory, and built fine,
or seemed to (aside from a myriad of warnings, everything completed OK).
I picked gcc34 because it seemed to be the most recent gcc in pkgsrc,
other than gcc44 which is very new - much newer than qt4-qdbus so I
assumed it was unlikely to be required.
Unfortunately, when I then attempted to use the qt4-qdbus package that
just finished building to build sysutils/strigi (one of the packages that
had been waiting on qt4-qdbus) I get this ...
Linking CXX executable strigiclient
/usr/pkg/qt4/lib/libQtDBus.so: undefined reference to `_ZdlPv@GLIBCXX_3.4'
/usr/pkg/qt4/lib/libQtDBus.so: undefined reference to `_ZTVN10__cxxabiv121__vmi_class_type_infoE@CXXABI_1.3'
/usr/pkg/qt4/lib/libQtDBus.so: undefined reference to `_ZdaPv@GLIBCXX_3.4'
/usr/pkg/qt4/lib/libQtDBus.so: undefined reference to `_Znaj@GLIBCXX_3.4'
/usr/pkg/qt4/lib/libQtDBus.so: undefined reference to `_Znwj@GLIBCXX_3.4'
/usr/pkg/qt4/lib/libQtDBus.so: undefined reference to `_ZTVN10__cxxabiv120__si_class_type_infoE@CXXABI_1.3'
/usr/pkg/qt4/lib/libQtDBus.so: undefined reference to `_ZTVN10__cxxabiv117__class_type_infoE@CXXABI_1.3'
which suggests (to me anyway) that something isn't quite right about the
qt4-qdbus package I managed to build... (Everything else which doesn't
request a specific compiler, which includes qt4-libs, would be using the
4.1.2 compiler that's NetBSD 4.0's standard cc).
I switched to
GCC_REQD+= 4.4
(leaving the other two lines I added as they were in the earlier test).
This one also managed to produce a binary package for qt4-dbus without
notable incident (certainly no memory demands so absurd as to be laughable).
Unfortunately, building sysutils/strigi failed with this one, in seemingly
the exact same way (including the "GLIBCXX_3.4" strings in the undefined
symbol names - even though, as best I can tell, nothing ever had gcc3.4
installed when it was being built in this version - I did verify, just
by the occasional "ps" while the compile was happening, that gcc34 was being
used in the first case above, and gcc44 in the second - I also verified
that the gcc44 built version of qt4-qdbus was actually what was installed
when strigi was being compiled).
I am appending all this (apparently useless) information to the original PR
in the hope that it will supply a hint to someone who actually has a clue
(which is not I) and lead to a solution to actually building this stuff on
NetBSD 4.0.
Note, none of these issues seem remotely Qt related - the person with a
clue probably needs to have gcc &/or toolchain clues, along with how all
that is made to work in pkgsrc, the Qt stuff just appears to be a trigger,
especially, as noted in the original PR, qt4-qdbus (same version) did
actually compile with the NetBSD 4.0 compiler previously - it needed
recompiling because of the jpeg update inspired revbump, and just for
that - and failed at that time. Something else altered between July 10
(my last successful compile) and Sept 1 (date of the PR, I think I ctually
detected this failing a little earlier than that.)
kre
ps: it isn't just sysutils/strigi, textproc/sporano generates the same
messages. Most likely anything using qt4-qdbus would.
Also, I note that both gcc34 and gcc44 packages include a message indicating
how they're to be used in pkgsrc in their MESSAGE - and that I didn't do
what those told me (partly as they seem to require more pkgsrc knowledge
than I have - qt4-qdbus has no pre-configure for example...) - and because
qt3-tools (which does have the GCC_REQD+= 3.4 - at least for sparcs - did
none of that extra stuff anywhere I would see, and I assume it works (no sparc
here to test on).
From: Joerg Sonnenberger <joerg@britannica.bec.de>
To: gnats-bugs@NetBSD.org
Cc:
Subject: Re: pkg/41966: x11/qt4-qdbus demands absurd quantities of (v)ram
to compile (NetBSD 4/0/i386)
Date: Mon, 5 Oct 2009 15:43:52 +0200
On Mon, Oct 05, 2009 at 01:25:02PM +0000, Robert Elz wrote:
> Linking CXX executable strigiclient
> /usr/pkg/qt4/lib/libQtDBus.so: undefined reference to `_ZdlPv@GLIBCXX_3.4'
> /usr/pkg/qt4/lib/libQtDBus.so: undefined reference to `_ZTVN10__cxxabiv121__vmi_class_type_infoE@CXXABI_1.3'
> /usr/pkg/qt4/lib/libQtDBus.so: undefined reference to `_ZdaPv@GLIBCXX_3.4'
> /usr/pkg/qt4/lib/libQtDBus.so: undefined reference to `_Znaj@GLIBCXX_3.4'
> /usr/pkg/qt4/lib/libQtDBus.so: undefined reference to `_Znwj@GLIBCXX_3.4'
> /usr/pkg/qt4/lib/libQtDBus.so: undefined reference to `_ZTVN10__cxxabiv120__si_class_type_infoE@CXXABI_1.3'
> /usr/pkg/qt4/lib/libQtDBus.so: undefined reference to `_ZTVN10__cxxabiv117__class_type_infoE@CXXABI_1.3'
Wild guess: This is the canonical GCC issue of not adding rpath entries
to find the newer version of libstdc++.
Joerg
From: Robert Elz <kre@munnari.OZ.AU>
To: gnats-bugs@NetBSD.org
Cc:
Subject: Re: pkg/41966: x11/qt4-qdbus demands absurd quantities of (v)ram to compile (NetBSD 4/0/i386)
Date: Mon, 05 Oct 2009 21:49:51 +0700
Date: Mon, 5 Oct 2009 13:45:02 +0000 (UTC)
From: Joerg Sonnenberger <joerg@britannica.bec.de>
Message-ID: <20091005134502.11A9963B8B6@www.NetBSD.org>
| Wild guess: This is the canonical GCC issue of not adding rpath entries
| to find the newer version of libstdc++.
I'm willing to believe anything about this one --- any suggestions
on what to do to verify the guess, and if correct, overcome it?
kre
From: Joerg Sonnenberger <joerg@britannica.bec.de>
To: gnats-bugs@NetBSD.org
Cc:
Subject: Re: pkg/41966: x11/qt4-qdbus demands absurd quantities of (v)ram
to compile (NetBSD 4/0/i386)
Date: Mon, 5 Oct 2009 17:00:04 +0200
On Mon, Oct 05, 2009 at 02:55:02PM +0000, Robert Elz wrote:
> The following reply was made to PR pkg/41966; it has been noted by GNATS.
>
> From: Robert Elz <kre@munnari.OZ.AU>
> To: gnats-bugs@NetBSD.org
> Cc:
> Subject: Re: pkg/41966: x11/qt4-qdbus demands absurd quantities of (v)ram to compile (NetBSD 4/0/i386)
> Date: Mon, 05 Oct 2009 21:49:51 +0700
>
> Date: Mon, 5 Oct 2009 13:45:02 +0000 (UTC)
> From: Joerg Sonnenberger <joerg@britannica.bec.de>
> Message-ID: <20091005134502.11A9963B8B6@www.NetBSD.org>
>
> | Wild guess: This is the canonical GCC issue of not adding rpath entries
> | to find the newer version of libstdc++.
>
> I'm willing to believe anything about this one --- any suggestions
> on what to do to verify the guess, and if correct, overcome it?
ldd the library and check which libstdc++ it is using.
Joerg
From: Robert Elz <kre@munnari.OZ.AU>
To: gnats-bugs@NetBSD.org
Cc:
Subject: Re: pkg/41966: x11/qt4-qdbus demands absurd quantities of (v)ram to compile (NetBSD 4/0/i386)
Date: Mon, 05 Oct 2009 23:03:52 +0700
Date: Mon, 5 Oct 2009 15:05:03 +0000 (UTC)
From: Joerg Sonnenberger <joerg@britannica.bec.de>
Message-ID: <20091005150503.B0EB263B8B6@www.NetBSD.org>
| ldd the library and check which libstdc++ it is using.
It certainly looks odd ...
libQtDBus.so.4.5.2:
-lz.1 => /usr/lib/libz.so.1
-lstdc++.6 => /usr/lib/libstdc++.so.6
-lm.0 => /usr/lib/libm387.so.0
-lm.0 => /usr/lib/libm.so.0
-lgcc_s.1 => /usr/lib/libgcc_s.so.1
-lpthread.0 => /usr/lib/libpthread.so.0
-lQtCore.4 => /usr/pkg/qt4/lib/libQtCore.so.4
-lQtXml.4 => /usr/pkg/qt4/lib/libQtXml.so.4
-lstdc++.6 => /usr/pkg/gcc44/lib/libstdc++.so.6
It has two different libstdc++.so.6's which cannot be good
(though as they both have the same major number, they should be
ABI equivalent, right?)
This was from the gcc44 version of qt4-dbus, obviously, I'm
guessing the gcc34 version would have been pretty much the same.
kre
From: Joerg Sonnenberger <joerg@britannica.bec.de>
To: gnats-bugs@NetBSD.org
Cc:
Subject: Re: pkg/41966: x11/qt4-qdbus demands absurd quantities of (v)ram
to compile (NetBSD 4/0/i386)
Date: Mon, 5 Oct 2009 18:18:43 +0200
On Mon, Oct 05, 2009 at 04:05:03PM +0000, Robert Elz wrote:
> libQtDBus.so.4.5.2:
> -lz.1 => /usr/lib/libz.so.1
> -lstdc++.6 => /usr/lib/libstdc++.so.6
> -lm.0 => /usr/lib/libm387.so.0
> -lm.0 => /usr/lib/libm.so.0
> -lgcc_s.1 => /usr/lib/libgcc_s.so.1
> -lpthread.0 => /usr/lib/libpthread.so.0
> -lQtCore.4 => /usr/pkg/qt4/lib/libQtCore.so.4
> -lQtXml.4 => /usr/pkg/qt4/lib/libQtXml.so.4
> -lstdc++.6 => /usr/pkg/gcc44/lib/libstdc++.so.6
objdump -p | grep RPATH will tell you the rpath, most likely libQtDBus
doesn't contain /usr/pkg/gcc44/lib for some reason. This is exactly the
kind of issue I meant.
> It has two different libstdc++.so.6's which cannot be good
> (though as they both have the same major number, they should be
> ABI equivalent, right?)
One is newer and contains more symbols...
Joerg
From: Robert Elz <kre@munnari.OZ.AU>
To: gnats-bugs@NetBSD.org
Cc:
Subject: Re: pkg/41966: x11/qt4-qdbus demands absurd quantities of (v)ram to compile (NetBSD 4/0/i386)
Date: Mon, 05 Oct 2009 23:39:10 +0700
Date: Mon, 5 Oct 2009 16:20:03 +0000 (UTC)
From: Joerg Sonnenberger <joerg@britannica.bec.de>
Message-ID: <20091005162003.EC2E763B8B6@www.NetBSD.org>
| objdump -p | grep RPATH will tell you the rpath,
RPATH /usr/pkg/gcc44/lib:/usr/pkg/qt4/lib:/usr/pkg/gcc44/lib/gcc/i386-unknown-netbsdelf4.0/4.4.1:/usr/pkg/gcc44/lib/gcc/i386-unknown-netbsdelf4.0/4.4.1/:/usr/pkg/lib
| most likely libQtDBus doesn't contain /usr/pkg/gcc44/lib
except it looks like it does.
I also tried the gcc34 version, and aside from having gcc34 where the
above (and the ldd) had gcc44, they're essentially the same
(the long components with netbsdelf in them also have the corresponding
version number change).
kre
From: Joerg Sonnenberger <joerg@britannica.bec.de>
To: gnats-bugs@NetBSD.org
Cc:
Subject: Re: pkg/41966: x11/qt4-qdbus demands absurd quantities of (v)ram
to compile (NetBSD 4/0/i386)
Date: Mon, 5 Oct 2009 18:48:30 +0200
On Mon, Oct 05, 2009 at 04:40:03PM +0000, Robert Elz wrote:
> The following reply was made to PR pkg/41966; it has been noted by GNATS.
>
> From: Robert Elz <kre@munnari.OZ.AU>
> To: gnats-bugs@NetBSD.org
> Cc:
> Subject: Re: pkg/41966: x11/qt4-qdbus demands absurd quantities of (v)ram to compile (NetBSD 4/0/i386)
> Date: Mon, 05 Oct 2009 23:39:10 +0700
>
> Date: Mon, 5 Oct 2009 16:20:03 +0000 (UTC)
> From: Joerg Sonnenberger <joerg@britannica.bec.de>
> Message-ID: <20091005162003.EC2E763B8B6@www.NetBSD.org>
>
> | objdump -p | grep RPATH will tell you the rpath,
>
>
> RPATH /usr/pkg/gcc44/lib:/usr/pkg/qt4/lib:/usr/pkg/gcc44/lib/gcc/i386-unknown-netbsdelf4.0/4.4.1:/usr/pkg/gcc44/lib/gcc/i386-unknown-netbsdelf4.0/4.4.1/:/usr/pkg/lib
And the dependencies? That's the NEEDED field.
Joerg
From: Robert Elz <kre@munnari.OZ.AU>
To: gnats-bugs@NetBSD.org
Cc:
Subject: Re: pkg/41966: x11/qt4-qdbus demands absurd quantities of (v)ram to compile (NetBSD 4/0/i386)
Date: Mon, 5 Oct 2009 23:52:38 +0700 (ICT)
NEEDED libQtXml.so.4
NEEDED libQtCore.so.4
NEEDED libz.so.1
NEEDED libstdc++.so.6
NEEDED libm.so.0
NEEDED libpthread.so.0
(this was from the gcc34 version, but I expect the gcc44 is the same,
it has been for everything else).
kre
From: Robert Elz <kre@munnari.OZ.AU>
To: gnats-bugs@NetBSD.org
Cc:
Subject: Re: pkg/41966: x11/qt4-qdbus demands absurd quantities of (v)ram to compile (NetBSD 4/0/i386)
Date: Tue, 13 Oct 2009 16:36:57 +0700
This virus (cc1plus out of memory) has now spread from qt4-qdbus
and into qt4-mysql qt4-pgsql qt4-sqlite3 and qt4-tiff
The file the compiler blows up on is the same (name anyway) in each
case (generators/mac/pbuilder_pbx.cpp)
Before these packages were upgraded to version 4.5.3 they all compiled
correctly (or at least seemed to) - I last compiled qt4-sqlite3 when
it moved from version 4.5.2nb1 to 4.5.2nb2 (compiled Oct 1), the others
I last compiled very late August (but qt4-qdbus failed at 4.5.2nb1
about the same time the others were successfully compiling...)
Does anyone have any ideas of a rational way to figure out what is
making the compiler decide it needs 252073488 + 15825784 bytes of
(malloc'd I assume) data space ?
That or why the NetBSD 4.0 compiler can't handle this, but apparently
both gcc34 and gcc44 (from pkgsrc) can compile it (though I get the link
errors reported earlier in the PR, that's most probably just some kind
of finger trouble though, which could be worked out if doing that was
really the right way to handle this.)
It is also worth noting that NetBSD -current's gcc version is only
one minor rev increased over that in NetBSD 4.0 (as best I can work
out).
kre
From: Joerg Sonnenberger <joerg@britannica.bec.de>
To: gnats-bugs@NetBSD.org
Cc:
Subject: Re: pkg/41966: x11/qt4-qdbus demands absurd quantities of (v)ram
to compile (NetBSD 4/0/i386)
Date: Tue, 13 Oct 2009 11:47:47 +0200
On Tue, Oct 13, 2009 at 09:40:04AM +0000, Robert Elz wrote:
> Does anyone have any ideas of a rational way to figure out what is
> making the compiler decide it needs 252073488 + 15825784 bytes of
> (malloc'd I assume) data space ?
Optimiser + big file, I guess. It is not that surprising.
Does UNLIMIT_RESOURCES = datasize work for you?
Joerg
From: Robert Elz <kre@munnari.OZ.AU>
To: gnats-bugs@NetBSD.org
Cc:
Subject: Re: pkg/41966: x11/qt4-qdbus demands absurd quantities of (v)ram to compile (NetBSD 4/0/i386)
Date: Tue, 13 Oct 2009 17:42:53 +0700
Date: Tue, 13 Oct 2009 09:50:04 +0000 (UTC)
From: Joerg Sonnenberger <joerg@britannica.bec.de>
Message-ID: <20091013095004.1E3BF63B877@www.NetBSD.org>
| Optimiser + big file, I guess. It is not that surprising.
The last time I worked on an optimiser (long ago now) it would
just abandon an optimisation if it ran out of resources (ie: ram
usually, but cpu time occasionally) rather than abort a compilation.
| Does UNLIMIT_RESOURCES = datasize work for you?
Yes, thanks, that does work (for qt4-qdbus, and a build of qt4-mysql
has proceeded past the faile that was causing problems). I assume
the rest of them will now be OK too.
I added that line into x11/qt4-libs/Makefile.common which is included
by all the others (it also contains a fair slab of system dependent
code already, so if you wanted to include this just for NetBSD 4 that
file would be a reaosnable place to do it I think).
I can't test whether this version of the qt4-qdbus package works
(without link errors) because of the problem reported by Hasso and
wiz (on the pkgsrc-changes list) in the past hour or two (or at least
I assume that's why textproc/soprano failed to get just about anywhere
with its compile when I tried it just a few minutes ago.)
kre
From: Robert Elz <kre@munnari.OZ.AU>
To: gnats-bugs@NetBSD.org
Cc:
Subject: Re: pkg/41966: x11/qt4-qdbus demands absurd quantities of (v)ram to compile (NetBSD 4/0/i386)
Date: Tue, 13 Oct 2009 18:53:58 +0700
Date: Tue, 13 Oct 2009 10:45:02 +0000 (UTC)
From: Robert Elz <kre@munnari.OZ.AU>
Message-ID: <20091013104502.881FD63B877@www.NetBSD.org>
| I assume the rest of them will now be OK too.
Yes, all the x11/qt4-* packages that had earlier failed, now compile
successfully with the UNLIMIT_RESOURCES=datasize line included in
x11/qt4-libs/Makefile.common
kre
From: coypu@sdf.org
To: gnats-bugs@netbsd.org
Cc:
Subject: Re: pkg/41966: x11/qt4-qdbus demands absurd quantities of (v)ram to
compile (NetBSD 4/0/i386)
Date: Mon, 27 May 2019 02:33:40 +0000
It sounds like we worked around it, but GCC could be less ridiculous
about memory usage. That's kinda out of scope for a package bug, though.
From: Robert Elz <kre@munnari.OZ.AU>
To: gnats-bugs@netbsd.org
Cc:
Subject: Re: pkg/41966: x11/qt4-qdbus demands absurd quantities of (v)ram to compile (NetBSD 4/0/i386)
Date: Mon, 27 May 2019 19:08:11 +0700
Date: Mon, 27 May 2019 02:35:01 +0000 (UTC)
From: coypu@sdf.org
Message-ID: <20190527023501.E409B7A181@mollari.NetBSD.org>
| It sounds like we worked around it, but GCC could be less ridiculous
| about memory usage. That's kinda out of scope for a package bug, though.
pkg/* PR's from me that are older than (say) about 2017 (certainly 2013)
can all be closed without further thought - I find it hard to imagine that
any of that ancient stuff is still relevant, and I actually cared about
almost none of it back then - I was just doing pkg builds (in my particular
environment) because I could, and they took very little of my time to make
happen (time back then being kind of a scarce resource).
But regardless of how stupid GCC might be, if it stops a pkg building,
that's a pkgsrc problem (might also be a gcc problem, but that is a
different issue).
kre
State-Changed-From-To: open->closed
State-Changed-By: bsiegert@NetBSD.org
State-Changed-When: Mon, 27 May 2019 12:39:58 +0000
State-Changed-Why:
Submitter says this can be closed.
>Unformatted:
(Contact us)
$NetBSD: query-full-pr,v 1.39 2013/11/01 18:47:49 spz Exp $
$NetBSD: gnats_config.sh,v 1.9 2014/08/02 14:16:04 spz Exp $
Copyright © 1994-2007
The NetBSD Foundation, Inc. ALL RIGHTS RESERVED.