[ Return to Bugs & Features | Roadmap 2.0 | Post Text | Post File ]
|Status:||5 - New|
|Priority:||2 - Low, e.g. a documentation error or undocumented side-effect|
|Scope:||3 - Applies to all machines and operating systems|
|Summary:||potentially unsafe behaviour for images|
Trouble Report Files:
[ Post File ]
Trouble Report Comments:
[ Post Text ]
22:11 Jul 12, 2010
|This isn't an overly critical patch to the image test functions (which are spread as thinly and over as many files as last night's pumpkin soup.....), but, as far as I'm aware, there's no standard to suggests that memcmp() *has* to stop when it reaches a null character, or even when it meets the first differing character (IIRC, it's allowed to compare all n characters if the implementation is done that way, or could potentially compare garbage data to be equal and continue). Whilst this is done sensibly in most implementations, I believe that it's a safer option to keep these as strncmp() and let the compiler deal with it's own implementation (in most cases, especially with gcc, AFAIK strncmp is turned into a memcmp anyway, so it's not going to create much of a difference in most systems). |
Beyond this, gifImage::test() had only a half-test for a GIF file, jpegImage had no size check and could thus segfault on an overly small file with a user-read header, xpmFileImage::test() also only had a half-test for an XPM image as well as no size checking, so on.
Tested and working on my i386 ubuntu; not that I'd expect otherwise given the small fixes this patch makes.
It certainly makes the code far more logical though.....
[ Return to Bugs & Features | Post Text | Post File ]