MD5 algorithm considered toxic (and harmful)

["Kristian Erik Hermansen" <kristian.hermansen () gmail com>]

I know of many commercial security products which still utilize MD5 to
prove integrity of the data they distribute to customers.  This should
no longer be considered appropriate.  Now that tools are readily
available to exploit newer MD5 collision research, I think it is safe
to say that the public should retire its usage for good.


Read the most recent research regarding chosen-prefix collisions:
http://www.win.tue.nl/hashclash/EC07v2.0.pdf


A concrete example for your perusal:
khermans@khermans-laptop:/tmp$ wget
http://www.win.tue.nl/hashclash/SoftIntCodeSign/HelloWorld-colliding.exe
--04:36:32--  http://www.win.tue.nl/hashclash/SoftIntCodeSign/HelloWorld-colliding.exe
           => `HelloWorld-colliding.exe'
Resolving www.win.tue.nl... 131.155.70.190
Connecting to www.win.tue.nl|131.155.70.190|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 41,792 (41K) [application/octet-stream]

100%[====================================>] 41,792       109.16K/s

04:36:33 (108.92 KB/s) - `HelloWorld-colliding.exe' saved [41792/41792]

khermans@khermans-laptop:/tmp$ wget
http://www.win.tue.nl/hashclash/SoftIntCodeSign/GoodbyeWorld-colliding.exe
--04:36:37--  http://www.win.tue.nl/hashclash/SoftIntCodeSign/GoodbyeWorld-colliding.exe
           => `GoodbyeWorld-colliding.exe'
Resolving www.win.tue.nl... 131.155.70.190
Connecting to www.win.tue.nl|131.155.70.190|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 41,792 (41K) [application/octet-stream]

100%[====================================>] 41,792       127.20K/s

04:36:38 (126.82 KB/s) - `GoodbyeWorld-colliding.exe' saved [41792/41792]

khermans@khermans-laptop:/tmp$ ls -lsha *.exe
44K -rw-r--r-- 1 khermans khermans 41K 2007-11-23 01:08
GoodbyeWorld-colliding.exe
44K -rw-r--r-- 1 khermans khermans 41K 2007-11-23 01:08 HelloWorld-colliding.exe
khermans@khermans-laptop:/tmp$ strings HelloWorld-colliding.exe | tail
SetFilePointer
MultiByteToWideChar
LCMapStringA
LCMapStringW
GetStringTypeA
GetStringTypeW
SetStdHandle
CloseHandle
KERNEL32.dll
Hello World ;-)
khermans@khermans-laptop:/tmp$ strings GoodbyeWorld-colliding.exe | tail
SetFilePointer
MultiByteToWideChar
LCMapStringA
LCMapStringW
GetStringTypeA
GetStringTypeW
SetStdHandle
CloseHandle
KERNEL32.dll
Goodbye World :-(
khermans@khermans-laptop:/tmp$ md5sum HelloWorld-colliding.exe | awk
'{print $1}' | tee hw
18fcc4334f44fed60718e7dacd82dddf
khermans@khermans-laptop:/tmp$ md5sum GoodbyeWorld-colliding.exe | awk
'{print $1}' | tee gw
18fcc4334f44fed60718e7dacd82dddf
khermans@khermans-laptop:/tmp$ cmp hw gw
khermans@khermans-laptop:/tmp$ echo $?
0


There you have it.  Surely a GPL'd tool implementing this attack style
will be available shortly.  And since Chinese researchers have been
attacking SHA-1 lately, should SHA-256 be considered the proper
replacement?  I am unsure :-(
-- 
Kristian Erik Hermansen
"I have no special talent. I am only passionately curious."

_______________________________________________
Full-Disclosure - We believe in it.
Charter: http://lists.grok.org.uk/full-disclosure-charter.html
Hosted and sponsored by Secunia - http://secunia.com/