Re: Complete list of AF_ALG algorithms used by iwd
James Prestwood <[email protected]> Mon, 4 May 2026 05:58:39 -0700
| Newsgroups | dev.linux.lists.iwd,org.kernel.vger.linux-crypto,org.kernel.vger.linux-kernel,org.kernel.vger.linux-wireless |
|---|---|
| Message-ID | <[email protected]> |
Hi Eric, On 5/3/26 9:50 PM, Eric Biggers wrote: > With AF_ALG being on its way out due to its frequent vulnerabilities, > iwd will need to be fixed to follow the standard practice of using > userspace crypto code (for example, libcrypto) instead. In the mean > time, AF_ALG is at least going to need to be hardened by removing > unnecessary functionality, for example unnecessary algorithms. > > Based on the iwd and libell source code, it appears iwd is relying on > the following algorithm names from AF_ALG. Please confirm that this > list is correct and that these are the ones we need to keep for iwd: > > cmac(aes) > hmac(md5) > hmac(sha1) > hmac(sha224) > hmac(sha256) > hmac(sha384) > hmac(sha512) > md4 > md5 > sha1 > sha224 > sha256 > sha384 > sha512 > ecb(aes) > cbc(aes) > ctr(aes) > ecb(des) > cbc(des) > cbc(des3_ede) This looks correct from IWD's standpoint, but ELL does support additional checksums: - sha3-224 - sha3-256 - sha3-384 - sha3-512 - hmac(sha3-224) - hmac(sha3-256) - hmac(sha3-384) - hmac(sha3-512) Marcel may be able to comment on what projects are using these, if any. > > Also, does iwd always hold CAP_NET_ADMIN? We're also considering adding > a privilege check to AF_ALG. (It's TBD whether it would be > unconditional or controlled by a sysctl.) Yes we require CAP_NET_ADMIN: https://git.kernel.org/pub/scm/network/wireless/iwd.git/tree/src/iwd.service.in#n15 Thanks, James > > - Eric