On Mon, Sep 17, 2007 at 12:48:59AM -0400, Eric St-Laurent wrote: > Hi, > > > I've seen some posts that say the loop-aes is faster and more secure > than dm-crypt. Is there any truths or facts behind this? As to secuity, dm-crypt with essiv or lrw is at least as secure as loop-aes, but probably more so. As to speed, I seem to remember that loop-aes contains an AES implementation that was denied entry into the kernel because of some licensing issues (or ego issues....). It may be a bit faster, but I doubt it matters. > I prefer the dm-based approach but I want to know the trade-offs > involved. > > I presume the speed difference is mostly because a more optimized aes > implementation? > > The "more secure" claim seemed based on older dm-crypt code. Does the > situation has changed? I would say it has, and dramatically so. One major problem with loop-aes is that it is not in the kernel. I used if for some time and there were constant problems with newer kernel version and no real visible advantage. My impression is that loop-AES is the ego-boosting project of its maintainer, who thinks he can do it better than everybody else, and btw the kernel maintainers have no clue about security. These claims have no merit today IMO. Arno -- Arno Wagner, Dipl. Inform., CISSP --- CSG, ETH Zurich, wagner@xxxxxxxxxxxxxx GnuPG: ID: 1E25338F FP: 0C30 5782 9D93 F785 E79C 0296 797F 6B50 1E25 338F ---- Cuddly UI's are the manifestation of wishful thinking. -- Dylan Evans If it's in the news, don't worry about it. The very definition of "news" is "something that hardly ever happens." -- Bruce Schneier --------------------------------------------------------------------- dm-crypt mailing list - http://www.saout.de/misc/dm-crypt/ To unsubscribe, e-mail: dm-crypt-unsubscribe@xxxxxxxx For additional commands, e-mail: dm-crypt-help@xxxxxxxx
