On 11/09/2020 18:30, Ard Biesheuvel wrote: > (cc Milan and dm-devel) > > On Fri, 11 Sep 2020 at 19:24, Van Leeuwen, Pascal > <pvanleeuwen@xxxxxxxxxx> wrote: >> >>> -----Original Message----- >>> From: linux-crypto-owner@xxxxxxxxxxxxxxx <linux-crypto-owner@xxxxxxxxxxxxxxx> On Behalf Of Ard Biesheuvel >>> Sent: Friday, September 11, 2020 4:11 PM >>> To: linux-crypto@xxxxxxxxxxxxxxx >>> Cc: herbert@xxxxxxxxxxxxxxxxxxx; ebiggers@xxxxxxxxxx; Ard Biesheuvel <ardb@xxxxxxxxxx> >>> Subject: [PATCH] crypto: mark unused ciphers as obsolete >>> >>> <<< External Email >>> >>> We have a few interesting pieces in our cipher museum, which are never >>> used internally, and were only ever provided as generic C implementations. >>> >>> Unfortunately, we cannot simply remove this code, as we cannot be sure >>> that it is not being used via the AF_ALG socket API, however unlikely. >>> So let's mark the Anubis, Khazad, SEED and TEA algorithms as obsolete, >>> >> Wouldn't the IKE deamon be able to utilize these algorithms through the XFRM API? >> I'm by no means an expert on the subject, but it looks like the cipher template is >> provided there directly via XFRM, so it does not need to live in the kernel source. >> And I know for a fact that SEED is being used for IPsec (and TLS) in Korea. >> > > I have been staring at net/xfrm/xfrm_algo.c, and as far as I can tell, > algorithms have to be mentioned there in order to be usable. None of > the ciphers that this patch touches are listed there or anywhere else > in the kernel. > >> The point being, there are more users to consider beyond "internal" (meaning hard >> coded in the kernel source in this context?) and AF_ALG. >> > > That is a good point, actually, since dm-crypt could be affected here > as well, hence the CCs. > > Milan (or others): are you aware of any of these ciphers being used > for dm-crypt? Cryptsetup/dm-crypt can use them (talking about Seed, Khazad, Anubis, TEA), but I think there is no real use of these. (IOW these are used only if someone deliberately uses them - manually specifying on format.) For dm-crypt. there should be no big harm if these are marked obsolete. Milan
