Bugzilla Bug 69467

Trying to launch some programs, e.g. gvim or firefox, triggers the
 "assert(width>0 && height>0);" on line 51 in wrlib/raster.c, resulting in a automatic restart of windowmaker. While for example xmms, Eterm or xterm launches fine..

Reproducible: Always
Steps to Reproduce:





Portage 2.0.51-r2 (default-linux/amd64/gcc34-2004.2, gcc-3.4.2,
glibc-2.3.4.20041021-r0, 2.6.8-gentoo-r4 x86_64)
=================================================================
System uname: 2.6.8-gentoo-r4 x86_64 AMD Athlon(tm) 64 Processor 3000+
Gentoo Base System version 1.6.4
ccache version 2.3 [enabled]
Autoconf: sys-devel/autoconf-2.59-r5
Automake: sys-devel/automake-1.8.5-r1
Binutils: sys-devel/binutils-2.15.92.0.2-r1
Headers:  sys-kernel/linux26-headers-2.6.8.1-r1
Libtools: sys-devel/libtool-1.5.2-r5
ACCEPT_KEYWORDS="amd64 ~amd64"
AUTOCLEAN="yes"
CFLAGS="-O2 -pipe -fomit-frame-pointer -ftracer -fprefetch-loop-arrays
-ffast-math -fweb -frename-registers -mtune=athlon64 -march=athlon64"
CHOST="x86_64-pc-linux-gnu"
COMPILER=""
CONFIG_PROTECT="/etc /usr/X11R6/lib/X11/xkb /usr/kde/2/share/config
/usr/kde/3.2/share/config /usr/kde/3.3/env /usr/kde/3.3/share/config
/usr/kde/3.3/shutdown /usr/kde/3/share/config /usr/lib/mozilla/defaults/pref
/usr/share/config /usr/share/texmf/dvipdfm/config/
/usr/share/texmf/dvips/config/ /usr/share/texmf/tex/generic/config/
/usr/share/texmf/tex/platex/config/ /usr/share/texmf/xdvi/ /var/qmail/control"
CONFIG_PROTECT_MASK="/etc/gconf /etc/terminfo /etc/env.d"
CXXFLAGS="-O2 -pipe -fomit-frame-pointer -ftracer -fprefetch-loop-arrays
-ffast-math -fweb -frename-registers -mtune=athlon64 -march=athlon64"
DISTDIR="/usr/portage/distfiles"
FEATURES="autoaddcvs ccache distlocks"
GENTOO_MIRRORS="http://ftp.du.se/pub/os/gentoo http://mirror.pudas.net/gentoo
http://ftp.linux.ee/pub/gentoo/distfiles/ http://mirror.gentoo.no/"
MAKEOPTS="-j2"
PKGDIR="/usr/portage/packages"
PORTAGE_TMPDIR="/var/tmp"
PORTDIR="/usr/portage"
PORTDIR_OVERLAY="/usr/local/portage"
SYNC="rsync://rsync.europe.gentoo.org/gentoo-portage"
USE="X alsa amd64 apm arts avi berkdb bitmap-fonts cdr crypt encode esd f77 fam
flac foomaticdb gdbm gif gnome gpm gstreamer gtk gtk2 imagemagick imlib ipv6
java jpeg junit kde libg++ libwww mad mikmod motif mozilla mpeg mysql ncurses
nls nptl odbc offensive oggvorbis opengl oss pam pdflib perl png python qt
quicktime readline ruby samba sdl slang spell sqlite ssl tcltk tcpd tetex
threads tiff truetype unicode xml2 xmms xv zlib

Hello, another confirmation here. I even get it when starting windowmaker.
Suppose I start my X server and then launch wmaker -display :0.0 from console:

helios@psychotic /root $ wmaker -display :0.0
wmaker warning: error loading image file "/usr/share/WindowMaker/Icons/clip.tiff"
wmaker warning: could not load default icon "defaultAppIcon.tiff":the image format in the file is not supported and can't be loaded
wmaker warning: error loading image file "/usr/share/WindowMaker/Icons/GNUstep.tiff"
wmaker warning: could not load logo image for panels: the image format in the file is not supported and can't be loaded
wmaker warning: error loading image file "/usr/share/WindowMaker/Icons/GNUstepGlow.tiff"
wmaker warning: error loading image file "/usr/share/WindowMaker/Icons/Terminal.tiff"
wmaker: raster.c:51: RCreateImage: Assertion `width>0 && height>0' failed.


After that I get a "Fatal error, Window Maker recieved signal 6 ...." error screen in which I  have the option to restart window maker. When I do this windowmaker starts but the icons won't appear.
When I launch an application like galeon I get:

wmaker: raster.c:51: RCreateImage: Assertion `width>0 && height>0' failed.
wmaker warning: Window Maker exited due to a crash (signal 6) and will be restarted.

And no galeon.

My emerge info:
ortage 2.0.51-r3 (default-amd64-2004.2, gcc-3.3.4, glibc-2.3.4.20041102-r0, 2.6.8-gentoo-r10 x86_64)
=================================================================
System uname: 2.6.8-gentoo-r10 x86_64 AMD Athlon(tm) 64 Processor 3000+
Gentoo Base System version 1.6.6
Autoconf: sys-devel/autoconf-2.59-r5
Automake: sys-devel/automake-1.8.5-r1
Binutils: sys-devel/binutils-2.15.92.0.2-r1
Headers:  sys-kernel/linux26-headers-2.6.8.1-r1
Libtools: sys-devel/libtool-1.5.2-r7
ACCEPT_KEYWORDS="amd64 ~amd64"
AUTOCLEAN="yes"
CFLAGS="-O2 -pipe -fPIC"
CHOST="x86_64-pc-linux-gnu"
COMPILER=""
CONFIG_PROTECT="/etc /usr/X11R6/lib/X11/xkb /usr/kde/2/share/config /usr/kde/3.2/share/config /usr/kde/3.3/env /usr/kde/3.3/share/config /usr/kde/3.3/shutdown /usr/kde/3/share/config /usr/lib/mozilla/defaults/pref /usr/share/config /usr/share/texmf/dvipdfm/config/ /usr/share/texmf/dvips/config/ /usr/share/texmf/tex/generic/config/ /usr/share/texmf/tex/platex/config/ /usr/share/texmf/xdvi/ /var/qmail/control"
CONFIG_PROTECT_MASK="/etc/gconf /etc/terminfo /etc/env.d"
CXXFLAGS="-O2 -pipe -fPIC"
DISTDIR="/usr/portage/distfiles"
FEATURES="autoaddcvs ccache distlocks"
GENTOO_MIRRORS="http://ftp.gentoo.skynet.be/pub/gentoo/"
MAKEOPTS="-j2"
PKGDIR="/usr/portage/packages"
PORTAGE_TMPDIR="/var/tmp"
PORTDIR="/usr/portage"
PORTDIR_OVERLAY=""
SYNC="rsync://rsync.gentoo.org/gentoo-portage"
USE="X amd64 apm arts avi berkdb bitmap-fonts cdr crypt encode esd f77 foomaticdb fortran gdbm gif gnome gpm gtk gtk2 imlib java jpeg kde libg++ libwww mikmod motif mozilla mpeg ncurses nls nogcj oggvorbis opengl oss pam pdflib perl png postgres python qt quicktime readline sdl slang spell ssl tcpd tetex truetype xml2 xmms xv zlib"


And this still appears in 0.91.0-r1

this works for me in 0.91.0-r1

Created an attachment (id=44768) [details]
Patch to wrlib/raster.c

This problem happens on my amd64 box too. The problem appears to be when the
assertion in RGetSubImage is called with height and with one less than 0. The
patch is an ugly ugly ugly hack, but it seems to work ok.

Is this patch going to be added to the ebuild soon or could somebody please
give some pointers on how/where to apply it, that file doesn't seem to be on my
system. At the moment I'm stuck on KDE and hating every minute of it.

Created an attachment (id=45511) [details]
patch to wrlib/raster.c

Aaah nuts, I submitted the patch *after* I'd already done it and deleted it.
This is what the patch *should* have been. I've been running
windowmaker-0.91.0-r1 with this hack for a month or so - it's been rock solid -
so hopefully the failed assert() is harmless.

Well it runs alright with the hack, kde popups seem to spawn off the screen,
zinf doesn't change as song progresses, no icons on active minimised
aplications. Other than that it runs alright.

Is this an AMD64 error only?  Can someone using an x86 install verify?

Added gnustep@ as a CC.  Seems like they should be involved in this.

I can confirm that x86 is not affected. For amd64 the latest unstable build
still does not work.

I'm working on the compile warnings on this first.  There are a fairly large
amount of "cast from pointer to integer of different size" warnings.  Let's see
if this fixes it.  If not, I'll go from there.

Debian guys fixed the problem.
Here is the link
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=280533

Maybe this will help.

Debian fix worked for me.  Fix in CVS.  Other arches should test this as well. 
Marked ~ for all arches.

BTW, let me know if this build works for you so I can close this bug.

works for me with x86 and amd64 (windowmaker-0.91-r2). thanx!

Works beautifully here, thank you.

Marked FIXED.

I think this bug should be reopened.
I have frequent crashes of mozilla and gaim.

They segfault and there is an output.
No running windows found.

As I know the resolution lies not in raster.c.
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=280533

[quote]
I think, the offending part is:

+        new_rimage = RCreateImage(data[best_pos-2], data[best_pos-1], True);

and/or wherever else "best_pos - sth" occures (because of best_pos
apparently being 0 in some cases[*]). I'm not sure, why actually this
part was modified this very way, so I introduced only small workaround, ie.:
     if ( best_pos < 2 )  best_pos = 2;
before best_pos is used for the first time. It seems to fork fine now.
[/quote]

Can anyone confirm that there is still a problem?  I am having no problems with
the patch.

applications which did not run before are now running smoothly.

nevertheless my thunderbird won't start with same error as described above: "No running windows found"

I can't tell if the problem is related to the windowmaker fix (I emerged 1-2 updates to x.org in the meanwhile as well)

The "No running windows found" is not an error.  That's just something that
prints out when thunderbird determines that there is not already an instance of
the application running.  I doubt that a Windowmaker bug would keep a single
application from running.  It's possible though but we would need more
information to track it down.