[Bug 1337827] Re: Bash crashes with a segmentation fault on "trap 'kill 0' SIGTERM EXIT"

2014-09-29 Thread Harry Willis
So, after some browsing, it apears that trap recursion was introduced intentionally in 4.3-rc2. There is commented code in bash.c that makes it look like trap recursion was originally intended to be a configurable build option, but was made default behaviour instead. One fairly easy fix would be t

[Bug 1337827] Re: Bash crashes with a segmentation fault on "trap 'kill 0' SIGTERM EXIT"

2014-09-29 Thread Harry Willis
It's the infinite loop of trapped SIGTERMs that raises this exception, and can be performed in any way that deliberately infinitely loops a signal trap. $ bash -c 'trap "kill 0" EXIT SIGTERM' Segmentation fault (core dumped) $ bash -c 'trap "kill $$" EXIT SIGTERM' Segmentation fault (core dumped)

[Bug 1374207] Re: CVE-2014-7169 fix not effective on trusty

2014-09-25 Thread Harry Willis
As per comment #5 on bug #1373781, the executables appear not to have been updated to their patched versions (forgot to recompile, I guess?) before the packages were generated. Appears to be the case all bash and bash-static packages of the 4.3-7ubuntu1.2 ilk. -- You received this bug notificati

[Bug 1373781] Re: bash incomplete fix for CVE-2014-6271

2014-09-25 Thread Harry Willis
Re the above: the patch was *not* correctly applied in trusty package bash_4.3-7ubuntu1.2. lucid package bash_4.3-7ubuntu1.2 appears to have been upgraded fine, and handles the test case correctly. harry@mars:~$ md5sum Downloads/bash_4.3-7ubuntu1.1_amd64/bin/bash Downloads/bash_4.3-7ubuntu1.2_am

[Bug 1373781] Re: bash incomplete fix for CVE-2014-6271

2014-09-25 Thread Harry Willis
Marc, I've just upgraded to 4.3.7-ubuntu1.2 in trusty (https://launchpad.net/ubuntu/+source/bash/4.3-7ubuntu1.2) which I assume was supposed to protect against the test case provided for CVE-2014-7169. It doesn't appear to have done so. Confirmed that the upgrade was successfully applied. harry@ma

[Bug 986841] Re: Debug log created by Adobe Reader browser plugin (nppdf.so)

2012-06-21 Thread Harry Willis
alphapapa, I've just had a look into this. The acroread package (version 9.5.1) does intend to install a copy of the plugin to /usr/lib/firefox/plugins (symlinked to /usr/lib/firefox-addons/plugins) - this is triggered by a postinst trigger to /opt/Adobe/Reader9/install_browser_plugin. However, thi

[Bug 986841] Re: Debug log created by Adobe Reader browser plugin (nppdf.so)

2012-05-21 Thread Harry Willis
It should be noted that in Firefox (tested in 12.0), the necessary nppdf plugin is located in /usr/lib/firefox-addons/plugins: $ locate nppdf /home/harry/C:\nppdf32Log\debuglog.txt /opt/Adobe/Reader9/Browser/intellinux/nppdf.so /usr/lib/firefox-addons/plugins/nppdf.so /usr/lib/mozilla/plugins/nppd