There was some discussion a while back about the proper method for doing kernel module checking. (see: https://lists.fedorahosted.org/pipermail/scap-security-guide/2012-August/001...)
The OVAL checks for disabling kernel modules are currently checking for `install [module] /bin/true`.
I'm sure there is a reason for doing this as opposed to `install [module] /bin/false`. Just a shot in the dark: we want the install to fail and return as if a failure is expected? Would it make more sense to run /bin/false, as the actual install is failing to install?
Additionally, it seems the checks are using a mixture of `install [module] /bin/true` and `alias [module] off`. Should these be made uniform, or is there a reason for the variation in method?
Any and all insight is greatly appreciated.
Thanks, --Mike
On 12/14/12 6:45 PM, Mike Palmiotto wrote:
There was some discussion a while back about the proper method for doing kernel module checking. (see: https://lists.fedorahosted.org/pipermail/scap-security-guide/2012-August/001...)
The OVAL checks for disabling kernel modules are currently checking for `install [module] /bin/true`.
I'm sure there is a reason for doing this as opposed to `install [module] /bin/false`. Just a shot in the dark: we want the install to fail and return as if a failure is expected? Would it make more sense to run /bin/false, as the actual install is failing to install?
Additionally, it seems the checks are using a mixture of `install [module] /bin/true` and `alias [module] off`. Should these be made uniform, or is there a reason for the variation in method?
Any and all insight is greatly appreciated.
Did this get lost in the pre-Christmas shuffle? I can't find any responses to this =/
I'd wager existing code is mixed simply because there was no standardized approach and we needed to "just get it done" between multiple coders. Standardizing on /bin/false seems ideal to me. Anyone have strong opinions on this?
And Mike was that you volunteering to submit patches for this?... ;)
Yes -- I think it got lost in the shuffle.
Standardizing on language/method here is desirable, and I'm afraid I don't recall any original motivations for particular choices (or if these were intentional).
It's possible that we chose /bin/true in order to quiet down some boot scripts. Using /bin/false certainly seems more desirable, assuming there are no undesirable side effects. Testing and patches welcome!
On 01/13/2013 11:17 PM, Shawn Wells wrote:
On 12/14/12 6:45 PM, Mike Palmiotto wrote:
There was some discussion a while back about the proper method for doing kernel module checking. (see: https://lists.fedorahosted.org/pipermail/scap-security-guide/2012-August/001...)
The OVAL checks for disabling kernel modules are currently checking for `install [module] /bin/true`.
I'm sure there is a reason for doing this as opposed to `install [module] /bin/false`. Just a shot in the dark: we want the install to fail and return as if a failure is expected? Would it make more sense to run /bin/false, as the actual install is failing to install?
Additionally, it seems the checks are using a mixture of `install [module] /bin/true` and `alias [module] off`. Should these be made uniform, or is there a reason for the variation in method?
Any and all insight is greatly appreciated.
Did this get lost in the pre-Christmas shuffle? I can't find any responses to this =/
I'd wager existing code is mixed simply because there was no standardized approach and we needed to "just get it done" between multiple coders. Standardizing on /bin/false seems ideal to me. Anyone have strong opinions on this?
And Mike was that you volunteering to submit patches for this?... ;) _______________________________________________ scap-security-guide mailing list scap-security-guide@lists.fedorahosted.org https://lists.fedorahosted.org/mailman/listinfo/scap-security-guide
scap-security-guide@lists.fedorahosted.org