Subject: bin/8200: ar(1) segfaults while building stock ImageMagick 4.2.8
To: None <gnats-bugs@gnats.netbsd.org>
From: Matthias Buelow <mkb@altair.mayn.de>
List: netbsd-bugs
Date: 08/12/1999 16:58:57
>Number:         8200
>Category:       bin
>Synopsis:       ar(1) segfaults while building stock ImageMagick 4.2.8
>Confidential:   no
>Severity:       non-critical
>Priority:       medium
>Responsible:    bin-bug-people (Utility Bug People)
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Thu Aug 12 16:20:01 1999
>Last-Modified:
>Originator:     Matthias Buelow
>Organization:
>Release:        1.4
>Environment:
	
System: NetBSD altair.mayn.de 1.4 NetBSD 1.4 (ALTAIR) #21: Thu Aug 12 23:10:39 CEST 1999 root@altair.mayn.de:/usr/src/sys/arch/i386/compile/ALTAIR i386


>Description:
On building a stock ImageMagick 4.2.8 distribution on NetBSD 1.4/i386
(not via pkgsrc, I actually didn't even look if it's in there) the
archiver ar(1) is invoked to create libMagick.a.
It never completes this, since it is killed via SIGSEGV before:

ar cru .libs/libMagick.a  PreRvIcccm.o animate.o annotate.o avs.o blob.o bmp.o cmyk.o colors.o compress.o dcm.o decorate.o delegates.o display.o dps.o draw.o effects.o enhance.o ept.o error.o fax.o fits.o fpx.o gems.o gif.o gradation.o gray.o hdf.o histogram.o html.o icc.o icon.o image.o iptc.o jbig.o jpeg.o label.o logo.o magick.o map.o matte.o memory.o miff.o monitor.o mono.o montage.o mtv.o null.o pcd.o pcl.o pcx.o pdf.o pict.o pix.o plasma.o png.o pnm.o preview.o ps.o ps2.o ps3.o psd.o pwp.o quantize.o rgb.o rla.o rle.o segment.o sct.o sfw.o sgi.o shear.o signature.o stegano.o sun.o tga.o tiff.o tile.o tim.o transform.o ttf.o txt.o uil.o utility.o uyvy.o vicar.o vid.o viff.o widget.o x.o xbm.o xc.o xpm.o xwd.o xwindows.o yuv.o zoom.o 
Segmentation fault - core dumped
*** Error code 139

Stop.

I wanted to compile ar with debugging infos but found out that with the
Gnu binutils tree, compiling a single tool is a little difficult; whatever,
I didn't have the time to dive into it, a quick attempt at configuring
and compiling it failed.  I would be thankful if some kind soul gave me
a hint on how to compile only ar(1) with sufficient debugging info for
identifying the bug in the source without having to compile the entire
binutils subtree.


>How-To-Repeat:
Obtain ImageMagick-4.2.8, untar and build.
>Fix:
None known to me.
>Audit-Trail:
>Unformatted: