On Wed, Sep 22, 2010 at 3:58 PM, Steve French <smfrench@xxxxxxxxx> wrote: >> Yep, that's correct. Note that the secType is a per-server parameter >> right now. >> >> We could mix sectypes on a socket. I sent a patchset that did that this >> past spring, but Steve basically NAK'ed it. > > At least for the case of signing, and for CIFS especially (where all > users of the socket, use the 1st users credentials in signing) - it > makes it less confusing. If we want to have different sectypes to > the same user for different users we could still do it over different > sockets (and with Ben's patch, different interfaces). > > -- > Thanks, > > Steve > OK. I will not touch that logic. But I took out the function match_security and I am able to authenticate and sign in any order over the same smb connection using session key created and stored by the auth mech of the very first smb session on that smb connection. mount -t cifs //9.3.123.456/winshare2 /mnt/smb_e -o sec=ntlmi,user=root,pass=<passwd> mount -t cifs //9.3.123456/winshare2 /mnt/smb_f -o sec=ntlmsspi,user=root,pass=<passwd> or in reverse order. -- To unsubscribe from this list: send the line "unsubscribe linux-cifs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
