|
|
|
Credit:
The information has been provided by iDefense Labs Security Advisories.
The original article can be found at: http://labs.idefense.com/intelligence/vulnerabilities/display.php?id=596
|
|
Vulnerable Systems:
* ImageMagick version 6.3.4
Immune Systems:
* ImageMagick version 6.3.5-9
The first vulnerability exists in the ReadDCMImage() function. Since the return value of ReadBlobByte() is not properly checked, it can enter an infinite loop.
The second vulnerability exists in the ReadXCFImage() function. Since the return value of ReadBlobMSBLong() is not properly checked, it can enter an infinite loop.
Analysis:
Exploitation of these vulnerabilities allows an attacker to consume excessive CPU resource on the system using the ImageMagick library to process images.
One way of exploiting these vulnerabilities is to persuade a targeted user to open a malicious image file with a program that utilizes the ImageMagick library.
As the tools that are part of ImageMagick are sometimes used as helper tools by other applications, this user may be the same as the web server user. This scenario is somewhat more severe than the previously described attack vector since the image processing can occur automatically.
Workaround:
Exposure to some of these vulnerabilities can be mitigated by moving or deleting the related module files. The file locations may vary between distributions. The globbing expressions listed below correspond to a Red Hat Linux system.
/usr/lib/ImageMagick-*/modules*/coders/dcm.*
/usr/lib/ImageMagick-*/modules*/coders/xcf.*
Vendor response:
The ImageMagick maintainers have addressed these vulnerabilities with the release of version 6.3.5-9. More information is available from the following URL. http://studio.imagemagick.org/pipermail/magick-announce/2007-September/000037.html
CVE Information:
CVE-2007-4985
Disclosure timeline:
09/04/2007 - Initial vendor notification
09/05/2007 - Initial vendor response
09/19/2007 - Coordinated public disclosure
|
|
|
|