Description: A vulnerability with an unknown impact has been reported in LibTomCrypt. The vulnerability is caused due to an unspecified error in the signature scheme used for ECC (Elliptic Curve Cryptography). The vulnerability has been reported in version 1.02. Prior versions may also be affected. Solution: The vulnerability has been fixed in the CVS repository.
eh, sorry
vapier please advise.
we'll have to wait for an upstream release since tom doesnt provide a public cvs server for now i'll just package.mask it
jaervosz is not available to watch the bug so I'll check regularly upstream for the new versions.
The new version should be comign tommorow (June 9th)
1.03 has been released http://libtomcrypt.org/download.html vapier please bump
1.03 now in portage
thanks Vapier arches please mark stable
i took care of that too
Sorry, I couldn't know because of the lag
Without more info I vote 1/2 NO
updated section from secunia: Description: A vulnerability has been reported in LibTomCrypt, which can be exploited by malicious people to create valid signatures without the private key. The vulnerability is caused due to a mathematical error in the implementation of the El Gamal signature algorithm. This can be exploited to create valid signatures and sign arbitrary messages without the private key via an application using the ECC (Elliptic Curve Cryptosystem) signature scheme routines. The vulnerability has been reported in version 1.02. Prior versions may also be affected. ________ CAN-2005-1600 Description: A "mathematical flaw" in the implementation of the El Gamal signature algorithm for LibTomCrypt 1.0 to 1.0.2 allows attackers to generate valid signatures without having the private key. ________ http://www.securityfocus.org/archive/1/397649 LibTomCrypt version <=1.02 contained weak signature scheme used with ECC keys, allowing trivial signature forgeries. Description: ------------ During recent cryptographic review by the author, a mathematical flaw was found within the implementation of the El Gamal signature algorithm used in LibTomCrypt versions <=1.02 An attacker can create a valid random signature by selecting a random value for a, and then computing (a^-1)C (where the inverse is modulo the order of the curve), essentially allowing an attacker to sign arbitrary messages without the private key.
I tend to vote NO. libtomcrypt is not used in any portage package...
please vote
I tend to vote NO too.
Anybody thinking yes, please reopen.