Re: [PATCH] userns: Replace netlink uses of cap_raised with capable.

[ebiederm@xmission.com (Eric W. Biederman)]

"Serge E. Hallyn" <serge@hallyn.com> writes:

> Quoting Eric W. Biederman (ebiederm@xmission.com):
>> 
>> In 2009 Philip Reiser notied that a few users of netlink connector
>> interface needed a capability check and added the idiom
>> cap_raised(nsp->eff_cap, CAP_SYS_ADMIN) to a few of them, on the premise
>> that netlink was asynchronous.
>> 
>> In 2011 Patrick McHardy noticed we were being silly because netlink is
>> synchronous and removed eff_cap from the netlink_skb_params and changed
>> the idiom to cap_raised(current_cap(), CAP_SYS_ADMIN).
>> 
>> Looking at those spots with a fresh eye we should be calling
>> capable(CAP_SYS_ADMIN).  The only reason I can see for not calling
>> capable is that it once appeared we were not in the same task as the
>> caller which would have made calling capable() impossible.
>
> And (just to make sure) that is now absolutely not the case?

Absolutely.  Netlink is synchronous.

I reread netlink_unicast_kernel() in net/netlink/af_netlink.c while
researching this patch.

Netlink has been synchronous since 2007.  In a world of perfect
knowledge transmission these calls could have been calls to capable
from the very beginning.

>> In the initial user_namespace the only difference between  between
>> cap_raised(current_cap(), CAP_SYS_ADMIN) and capable(CAP_SYS_ADMIN)
>> are a few sanity checks and the fact that capable(CAP_SYS_ADMIN)
>> sets PF_SUPERPRIV if we use the capability.
>> 
>> Since we are going to be using root privilege setting PF_SUPERPRIV
>> seems the right thing to do.
>> 
>> The motivation for this that patch is that in a child user namespace
>> cap_raised(current_cap(),...) tests your capabilities with respect to
>> that child user namespace not capabilities in the initial user namespace
>> and thus will allow processes that should be unprivielged to use the
>> kernel services that are only protected with
>> cap_raised(current_cap(),..).
>> 
>> To fix possible user_namespace issues and to just clean up the code
>> replace cap_raised(current_cap(), CAP_SYS_ADMIN) with
>> capable(CAP_SYS_ADMIN).
>> 
>> Cc: Patrick McHardy <kaber@trash.net>
>> Cc: Philipp Reisner <philipp.reisner@linbit.com>
>> Cc: Serge E. Hallyn <serge.hallyn@canonical.com>
>
> Acked-by: Serge E. Hallyn <serge.hallyn@canonical.com>
>
> thanks,
> -serge
>
>> Cc: Andrew Morgan <morgan@kernel.org>
>> Cc: Vasiliy Kulikov <segoon@openwall.com>
>> Cc: David Howells <dhowells@redhat.com>
>> Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
>> ---
>>  drivers/block/drbd/drbd_nl.c           |    2 +-
>>  drivers/md/dm-log-userspace-transfer.c |    2 +-
>>  drivers/video/uvesafb.c                |    2 +-
>>  3 files changed, 3 insertions(+), 3 deletions(-)
>> 
>> diff --git a/drivers/block/drbd/drbd_nl.c b/drivers/block/drbd/drbd_nl.c
>> index abfaaca..946166e 100644
>> --- a/drivers/block/drbd/drbd_nl.c
>> +++ b/drivers/block/drbd/drbd_nl.c
>> @@ -2297,7 +2297,7 @@ static void drbd_connector_callback(struct cn_msg *req, struct netlink_skb_parms
>>  		return;
>>  	}
>>  
>> -	if (!cap_raised(current_cap(), CAP_SYS_ADMIN)) {
>> +	if (!capable(CAP_SYS_ADMIN)) {
>>  		retcode = ERR_PERM;
>>  		goto fail;
>>  	}
>> diff --git a/drivers/md/dm-log-userspace-transfer.c b/drivers/md/dm-log-userspace-transfer.c
>> index 1f23e04..08d9a20 100644
>> --- a/drivers/md/dm-log-userspace-transfer.c
>> +++ b/drivers/md/dm-log-userspace-transfer.c
>> @@ -134,7 +134,7 @@ static void cn_ulog_callback(struct cn_msg *msg, struct netlink_skb_parms *nsp)
>>  {
>>  	struct dm_ulog_request *tfr = (struct dm_ulog_request *)(msg + 1);
>>  
>> -	if (!cap_raised(current_cap(), CAP_SYS_ADMIN))
>> +	if (!capable(CAP_SYS_ADMIN))
>>  		return;
>>  
>>  	spin_lock(&receiving_list_lock);
>> diff --git a/drivers/video/uvesafb.c b/drivers/video/uvesafb.c
>> index 260cca7..9f7d27a 100644
>> --- a/drivers/video/uvesafb.c
>> +++ b/drivers/video/uvesafb.c
>> @@ -73,7 +73,7 @@ static void uvesafb_cn_callback(struct cn_msg *msg, struct netlink_skb_parms *ns
>>  	struct uvesafb_task *utask;
>>  	struct uvesafb_ktask *task;
>>  
>> -	if (!cap_raised(current_cap(), CAP_SYS_ADMIN))
>> +	if (!capable(CAP_SYS_ADMIN))
>>  		return;
>>  
>>  	if (msg->seq >= UVESAFB_TASKS_MAX)
>> -- 
>> 1.7.2.5