Re: [PATCH v22 12/23] LSM: Specify which LSM to display

[Casey Schaufler]

On 11/7/2020 2:05 PM, John Johansen wrote:
> On 11/7/20 1:15 AM, Greg KH wrote:
>> On Fri, Nov 06, 2020 at 04:20:43PM -0800, Casey Schaufler wrote:
>>> On 11/5/2020 1:22 AM, Greg KH wrote:
>>>> On Wed, Nov 04, 2020 at 03:41:03PM -0800, Casey Schaufler wrote:
>>>>> Create a new entry "display" in the procfs attr directory for
>>>>> controlling which LSM security information is displayed for a
>>>>> process. A process can only read or write its own display value.
>>>>>
>>>>> The name of an active LSM that supplies hooks for
>>>>> human readable data may be written to "display" to set the
>>>>> value. The name of the LSM currently in use can be read from
>>>>> "display". At this point there can only be one LSM capable
>>>>> of display active. A helper function lsm_task_display() is
>>>>> provided to get the display slot for a task_struct.
>>>>>
>>>>> Setting the "display" requires that all security modules using
>>>>> setprocattr hooks allow the action. Each security module is
>>>>> responsible for defining its policy.
>>>>>
>>>>> AppArmor hook provided by John Johansen <john.johansen@xxxxxxxxxxxxx>
>>>>> SELinux hook provided by Stephen Smalley <sds@xxxxxxxxxxxxx>
>>>>>
>>>>> Reviewed-by: Kees Cook <keescook@xxxxxxxxxxxx>
>>>>> Acked-by: Stephen Smalley <sds@xxxxxxxxxxxxx>
>>>>> Acked-by: Paul Moore <paul@xxxxxxxxxxxxxx>
>>>>> Signed-off-by: Casey Schaufler <casey@xxxxxxxxxxxxxxxx>
>>>>> Cc: linux-api@xxxxxxxxxxxxxxx
>>>>> ---
>>>>>  fs/proc/base.c                       |   1 +
>>>>>  include/linux/lsm_hooks.h            |  17 +++
>>>>>  security/apparmor/include/apparmor.h |   3 +-
>>>>>  security/apparmor/lsm.c              |  32 +++++
>>>>>  security/security.c                  | 169 ++++++++++++++++++++++++---
>>>>>  security/selinux/hooks.c             |  11 ++
>>>>>  security/selinux/include/classmap.h  |   2 +-
>>>>>  security/smack/smack_lsm.c           |   7 ++
>>>>>  8 files changed, 223 insertions(+), 19 deletions(-)
>>>>>
>>>>> diff --git a/fs/proc/base.c b/fs/proc/base.c
>>>>> index 0f707003dda5..7432f24f0132 100644
>>>>> --- a/fs/proc/base.c
>>>>> +++ b/fs/proc/base.c
>>>>> @@ -2806,6 +2806,7 @@ static const struct pid_entry attr_dir_stuff[] = {
>>>>>  	ATTR(NULL, "fscreate",		0666),
>>>>>  	ATTR(NULL, "keycreate",		0666),
>>>>>  	ATTR(NULL, "sockcreate",	0666),
>>>>> +	ATTR(NULL, "display",		0666),
>>>> That's a vague name, any chance it can be more descriptive?
>>> Sure. How about lsm_display, or display_lsm? I wouldn't say that
>>> any of the files in /proc/*/attr have especially descriptive names,
>>> but that's hardly an excuse.
>> I still don't understand what "display" means in this context.  Perhaps
> its the LSM thats context is being displayed on the shared interface,
> ie. /proc/*/attr/*
>
> thinking about it more owner or even interface_owner might be a better
> name

I was hoping for a single word, but I see how something more descriptive
might be in order. How about "lsm_of_current"? Or "lsm_of_dot_slash_current",
if you want to be pedantic. "format_of_current" isn't quite accurate, but
might be easier for some people to understand. Maybe "interface_owning_lsm".

/proc/*/attr/display answers the question "Which LSM is providing the data
I see if I look in /proc/*/attr/current, prev or exec or if that process uses
SO_PEERSEC".


>> documentation will help clear it up?
>>
> yeah this needs documented.

Agreed. I've noticed that nothing in /proc/*/attr seems documented
in an orderly (documentation/ABI) fashion. I will have to fix some of
that for a description of /proc/*/attr/whatever_it_ends_up_getting_called
to make sense. Working on it.

>> thanks,
>>
>> greg k-h
>>