Based partially on a patch by wbk to fix Gentoo bug #410093 [1]. From the
bug report:
After fixing the /proc/meminfo issue, another bug will emerge on systems
where total system RAM in bytes exceeds INT_MAX. The correct top three
processes will be identified, but their load bars will have the wrong
value. This is caused by a value overflow due to storing total system
memory in an "int" type variable. We will end up thinking ((Total RAM)
modulo (INT_MAX)) is our total system RAM, so our percentages will be
inflated when drawing the load bar. This fix will require a bit more care
to follow the flow of data and ensure this value isn't being cast to "int"
along the way. Function return types will need to be changed.
[1] https://bugs.gentoo.org/show_bug.cgi?id=410093
It was resulting in compilation errors:
profzoom@cayke:~/src/dockapps/dockapps/wmget$ xsltproc --nonet
wmget.refentry.xml
I/O error : Attempt to load network entity http://docbook.sourceforge.net/
release/xsl/1.62/manpages/docbook.xsl
warning: failed to load external entity "http://docbook.sourceforge.net/
release/xsl/1.62/manpages/docbook.xsl"
error
xsltParseStylesheetFile : cannot parse http://docbook.sourceforge.net/
release/xsl/1.62/manpages/docbook.xsl
compilation error: file wmget.refentry.xml line 19 element refentry
xsltParseStylesheetProcess : document is not a stylesheet
We therefore remove the xml source file and in the future, we will edit the
troff file directly.
We also add the version number to the manpage via autoconf.
Fixes the following warning:
In file included from src/wmtv.c:47:0:
/usr/include/X11/extensions/xf86dga.h:9:2: warning: #warning "xf86dga.h is
obsolete and may be removed in the future." [-Wcpp]
#warning "xf86dga.h is obsolete and may be removed in the future."
^
/usr/include/X11/extensions/xf86dga.h:10:2: warning: #warning "include
<X11/extensions/Xxf86dga.h> instead." [-Wcpp]
#warning "include <X11/extensions/Xxf86dga.h> instead."
^
Need to divide by 16 or 16000 depending on whether using kHz or MHz.
Patch by Nicolas Boullis <nboullis@debian.org> for Debian package version
0.6.5-14 or 15. (Not mentioned in debian/changelog.)
(I couldn't find the patches separately in, e.g., a bug report.)
Patches by Nicolas Boullis <Boullis.Nicolas@libertysurf.fr>.
Corresponding entries from debian/changelog:
0.6.5-8
=======
* Drop privileges to access files, and then regain privileges.
(Closes: #120825)
* A new user's configuration file is automatically created when needed.
* Cleaned up a few possible buffer overflows.
0.6.5-9
=======
* The program does not need any more to be setuid root.
(Closes: #120832)
0.6.5-10
========
* Trying to change channel while the TV is off does nothing, and does
not segfault anymore.
* Corrected strange behavior when the mouse was moved out of a pressed
button.
0.6.5-11
========
* Corrected a bug introduced in version 0.6.5-9, causing the program to
segfault while reverting from fullscreen back to normal mode.
(Closes: #128409)
0.6.5-12
========
* Catch SIGCHLD signals to remove zombies. (Closes: #129435)
0.6.5-13
========
* Added missing #include in wmtv.c.
* Automatically restarts the TV when the external application finishes.
* In fullscreen mode, the left and right arrow keys now behave as
specified in the README file and in the manpage. (Closes: #133301)
* Added the ability to pass some parameters to the external
application.
Patch by Nicolas Boullis <Boullis.Nicolas@libertysurf.fr>. From [1]:
From: Nicolas Boullis <Boullis.Nicolas@libertysurf.fr>
Subject: wmtv: shortens user's configuration file
Date: Sun, 18 Nov 2001 02:38:00 +0100
Hello !
When wmtv writes it configuration back to the user's configuration file
(after the user tried to fine-tune the channels) it forgets the last
channel. And then... one can't start it any more without correcting the
configuration file first.
The patch included makes wmtv write all the channels back.
[1] https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=120035
Patch by Nicolas Boullis <Boullis.Nicolas@libertysurf.fr>. From [1]:
From: Nicolas Boullis <Boullis.Nicolas@libertysurf.fr>
Subject: wmtv: does not start without user's ~/.wmtvrc
Date: Sun, 18 Nov 2001 00:17:26 +0100
Hi !
The current version of wmtv does not start if the user does not have a
~/.wmtvrc configuration file, while one could expect it to work with the
system wide configuration file /etc/wmtvrc.
This patch allows wmtv to start with the system wide configuration file
alone.
[1] https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=120024
Patch by Nicolas Boullis <Boullis.Nicolas@libertysurf.fr>. From [1]:
From: Nicolas Boullis <Boullis.Nicolas@libertysurf.fr>
Subject: wmtv: some french channels are not available
Date: Fri, 16 Nov 2001 01:33:13 +0100
wmtv does not currently support a few channels which are used in France.
These channels are called K01-K10, KB-KQ and H01-H19.
The patch included defines a new "freqnorm" called "secam-france", and
the new channels.
[1] https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=119767
Patch by Nicolas Boullis <Boullis.Nicolas@libertysurf.fr>. From [1]:
From: Nicolas Boullis <Boullis.Nicolas@libertysurf.fr>
To: Debian Bug Tracking System <submit@bugs.debian.org>
Subject: wmtv: dangerous suid root
Date: Thu, 08 Nov 2001 20:07:52 +0100
Hi !
I think there is a huge security hole with wmtv and, when wmtv is installed,
anyone can easily get a root account. Here is what I have in my terminal:
(everytime I launch wmtv, I double-clicked in the tv subwindow to call the
external program)
----------------------------------------------------------------------
Tintin:~> wmtv -e whoami
root
Tintin:~> cat > crack_root.sh
#!/bin/sh
cp /bin/sh /tmp
chmod u+s /tmp/sh
Tintin:~> chmod +x crack_root.sh
Tintin:~> wmtv -e ~/crack_root.sh
Tintin:~> ll /tmp/sh
-rwsr-xr-x 1 root users 407356 Nov 8 19:25 /tmp/sh*
----------------------------------------------------------------------
I tried to make wmtv non-suid root, and... sometimes it works (despite an
error message), sometimes it does not...
----------------------------------------------------------------------
Tintin:~> ll /usr/bin/X11/wmtv
-rwxr-xr-x 1 root root 62588 Jul 31 01:55 /usr/bin/X11/wmtv*
Tintin:~> wmtv
ioctl VIDIOCSFBUF: Operation not permitted
Tintin:~> wmtv
ioctl VIDIOCSFBUF: Operation not permitted
wmtv: no physical frame buffer access
----------------------------------------------------------------------
Hence, I guess you should either correct wmtv so that it always work without
being suid root, or make wmtv lose its privileges before it runs an external
program.
[1] https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=118778
Patch by Yann Vernier <yann@donkey.dyndns.org>. From [1]:
From: Malcolm Parsons <malcolm@ivywell.screaming.net>
Subject: wmtv: incorrectly calculates bytes per line
Date: Mon, 09 Apr 2001 21:15:52 +0100
wmtv does not put the tv display in its window on my second head.
According to bttv, wmtv is telling bttv that the display is:
Display at ea800000 is 800 by 600, bytedepth 2, bpl 1600
If I use xawtv, it correctly sets the bpl value:
Display at ea800000 is 800 by 600, bytedepth 2, bpl 1664
wmtv is probably incorrectly assuming width * bytedepth = bpl.
From: Yann Vernier <yann@donkey.dyndns.org>
Subject: wmtv: incorrectly calculates bytes per line
Date: Sun, 15 Jul 2001 14:21:56 +0200
Found the problem, at least this fixes it for me at 1600x1200.
[1] https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=93439
Patch by Yann Vernier <yann@algonet.se>. From [1]:
wmtv failed to read my .wmtvrc (hand written) correctly, I tracked it
down to incorrect memory management (sizeof() instead of strlen()).
The code was so bad that I chose to rewrite it to a simpler version,
patch below.
Other bugs fixed include some of the memory leaks (which are quite
numerous) and truncating the .wmtvrc when writing to it.
Cosmetic changes include keeping comments in .wmtvrc, C++-style
comments changed to C style, and numbering channels from 1.
[1] https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=105325
This fixes the package-contains-timestamped-gzip warning given by Lintian
for the Debian package. (This warning is really unnecessary, as its purpose
it to check for reproducible builds and the fonts are not compressed at
build time, but I see no harm in removing these timestamps.)
