Date: Sun, 10 Feb 2013 13:10:13 +0400 From: Solar Designer <solar@...nwall.com> To: john-dev@...ts.openwall.com Subject: remaining DMG issues magnum, Dhiru, all - Here are a few I noticed, most important first: 1. We use a non-split OpenCL kernel. At high iteration counts like 250k it's bound to cause AMD ASIC hangs. Perhaps this is already on your to-do for fixing? 2. I just got both the dmg and dmg-opencl formats to segfault when feeding them with truncated input. Specifically, I took the first quoted string from the tests in dmg_fmt_plug.c. The one that ends in ...60156. I put it in a file on its own. 3. There are warnings on unused read() return value in dmg2john.c. We need proper checks of the return value (or better yet, also loops similar to read_loop() in popa3d, although when reading regular files partial reads are unlikely). A non-code issue: 4. We need to have DMG test cases on this wiki page: http://openwall.info/wiki/john/sample-non-hashes (including for both 1000 and non-1000 iteration count files). ... and we should probably use alphabetical order for the different formats listed there. Thanks, Alexander
Powered by blists - more mailing lists
Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.