linux-fsdevel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: "J. Bruce Fields" <bfields@fieldses.org>
To: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: linux-fsdevel@vger.kernel.org,
	Linux Containers <containers@lists.linux-foundation.org>,
	linux-kernel@vger.kernel.org,
	"Serge E. Hallyn" <serge@hallyn.com>,
	Trond Myklebust <Trond.Myklebust@netapp.com>
Subject: Re: [PATCH review 52/85] sunrpc: Properly encode kuids and kgids in auth.unix.gid rpc pipe upcalls.
Date: Wed, 13 Feb 2013 16:05:45 -0500	[thread overview]
Message-ID: <20130213210545.GO14195@fieldses.org> (raw)
In-Reply-To: <1360777934-5663-52-git-send-email-ebiederm@xmission.com>

On Wed, Feb 13, 2013 at 09:51:41AM -0800, Eric W. Biederman wrote:
> From: "Eric W. Biederman" <ebiederm@xmission.com>
> 
> When a new rpc connection is established with an in-kernel server, the
> traffic passes through svc_process_common, and svc_set_client and down
> into svcauth_unix_set_client if it is of type RPC_AUTH_NULL or
> RPC_AUTH_UNIX.
> 
> svcauth_unix_set_client then looks at the uid of the credential we
> have assigned to the incomming client and if we don't have the groups
> already cached makes an upcall to get a list of groups that the client
> can use.
> 
> The upcall encodes send a rpc message to user space encoding the uid
> of the user whose groups we want to know.  Encode the kuid of the user
> in the initial user namespace as nfs mounts can only happen today in
> the initial user namespace.

OK, I didn't know that.

(Though I'm unclear how it should matter to the server what user
namespace the client is in?)

> When a reply to an upcall comes in convert interpret the uid and gid values
> from the rpc pipe as uids and gids in the initial user namespace and convert
> them into kuids and kgids before processing them further.
> 
> When reading proc files listing the uid to gid list cache convert the
> kuids and kgids from into uids and gids the initial user namespace.  As we are
> displaying server internal details it makes sense to display these values
> from the servers perspective.

All of these caches are already per-network-namespace.  Ideally wouldn't
we also like to associate a user namespace with each cache somehow?

--b.

> 
> Cc: "J. Bruce Fields" <bfields@fieldses.org>
> Cc: Trond Myklebust <Trond.Myklebust@netapp.com>
> Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
> ---
>  net/sunrpc/svcauth_unix.c |   14 +++++++++-----
>  1 files changed, 9 insertions(+), 5 deletions(-)
> 
> diff --git a/net/sunrpc/svcauth_unix.c b/net/sunrpc/svcauth_unix.c
> index faf1719..bdea0a1 100644
> --- a/net/sunrpc/svcauth_unix.c
> +++ b/net/sunrpc/svcauth_unix.c
> @@ -470,7 +470,7 @@ static void unix_gid_request(struct cache_detail *cd,
>  	char tuid[20];
>  	struct unix_gid *ug = container_of(h, struct unix_gid, h);
>  
> -	snprintf(tuid, 20, "%u", ug->uid);
> +	snprintf(tuid, 20, "%u", from_kuid(&init_user_ns, ug->uid));
>  	qword_add(bpp, blen, tuid);
>  	(*bpp)[-1] = '\n';
>  }
> @@ -486,7 +486,8 @@ static int unix_gid_parse(struct cache_detail *cd,
>  			char *mesg, int mlen)
>  {
>  	/* uid expiry Ngid gid0 gid1 ... gidN-1 */
> -	int uid;
> +	int id;
> +	kuid_t uid;
>  	int gids;
>  	int rv;
>  	int i;
> @@ -498,9 +499,12 @@ static int unix_gid_parse(struct cache_detail *cd,
>  		return -EINVAL;
>  	mesg[mlen-1] = 0;
>  
> -	rv = get_int(&mesg, &uid);
> +	rv = get_int(&mesg, &id);
>  	if (rv)
>  		return -EINVAL;
> +	uid = make_kuid(&init_user_ns, id);
> +	if (!uid_valid(uid))
> +		return -EINVAL;
>  	ug.uid = uid;
>  
>  	expiry = get_expiry(&mesg);
> @@ -554,7 +558,7 @@ static int unix_gid_show(struct seq_file *m,
>  			 struct cache_detail *cd,
>  			 struct cache_head *h)
>  {
> -	struct user_namespace *user_ns = current_user_ns();
> +	struct user_namespace *user_ns = &init_user_ns;
>  	struct unix_gid *ug;
>  	int i;
>  	int glen;
> @@ -570,7 +574,7 @@ static int unix_gid_show(struct seq_file *m,
>  	else
>  		glen = 0;
>  
> -	seq_printf(m, "%u %d:", ug->uid, glen);
> +	seq_printf(m, "%u %d:", from_kuid_munged(user_ns, ug->uid), glen);
>  	for (i = 0; i < glen; i++)
>  		seq_printf(m, " %d", from_kgid_munged(user_ns, GROUP_AT(ug->gi, i)));
>  	seq_printf(m, "\n");
> -- 
> 1.7.5.4
> 

  reply	other threads:[~2013-02-13 21:05 UTC|newest]

Thread overview: 110+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-02-13 17:48 [PATCH review 00/85] userns changes for 9p, afs, ceph, cifs, coda, gfs2, ncpfs, nfs, nfsd, and ocfs2 Eric W. Biederman
2013-02-13 17:50 ` [PATCH review 01/85] ceph: Only allow mounts in the initial network namespace Eric W. Biederman
2013-02-13 17:50   ` [PATCH review 08/85] 9p: Transmit kuid and kgid values Eric W. Biederman
2013-02-13 17:51   ` [PATCH review 12/85] 9p: Modify v9fs_get_fsgid_for_create to return a kgid Eric W. Biederman
2013-02-13 17:51   ` [PATCH review 13/85] 9p: Allow building 9p with user namespaces enabled Eric W. Biederman
2013-02-13 17:51   ` [PATCH review 16/85] afs: Support interacting with multiple user namespaces Eric W. Biederman
2013-02-13 17:51   ` [PATCH review 17/85] coda: Restrict coda messages to the initial pid namespace Eric W. Biederman
2013-02-13 17:51   ` [PATCH review 22/85] ocfs2: convert between kuids and kgids and DLM locks Eric W. Biederman
     [not found]     ` <1360777934-5663-22-git-send-email-ebiederm-aS9lmoZGLiVWk0Htik3J/w@public.gmane.org>
2013-02-14  8:34       ` Joel Becker
2013-02-13 17:51   ` [PATCH review 23/85] ocfs2: Convert uid and gids between in core and on disk inodes Eric W. Biederman
     [not found]     ` <1360777934-5663-23-git-send-email-ebiederm-aS9lmoZGLiVWk0Htik3J/w@public.gmane.org>
2013-02-14  8:35       ` Joel Becker
2013-02-14  9:10         ` Eric W. Biederman
     [not found]   ` <1360777934-5663-1-git-send-email-ebiederm-aS9lmoZGLiVWk0Htik3J/w@public.gmane.org>
2013-02-13 17:50     ` [PATCH review 02/85] ceph: Translate between uid and gids in cap messages and kuids and kgids Eric W. Biederman
2013-02-13 17:50     ` [PATCH review 03/85] ceph: Translate inode uid and gid attributes to/from " Eric W. Biederman
2013-02-13 17:50     ` [PATCH review 04/85] ceph: Convert struct ceph_mds_request to use kuid_t and kgid_t Eric W. Biederman
2013-02-13 17:50     ` [PATCH review 05/85] ceph: Convert kuids and kgids before printing them Eric W. Biederman
2013-02-13 17:50     ` [PATCH review 06/85] ceph: Enable building when user namespaces are enabled Eric W. Biederman
2013-02-13 17:50     ` [PATCH review 07/85] 9p: Add 'u' and 'g' format specifies for kuids and kgids Eric W. Biederman
2013-02-13 17:50     ` [PATCH review 09/85] 9p: Modify the stat structures to use kuid_t and kgid_t Eric W. Biederman
2013-02-13 17:50     ` [PATCH review 10/85] 9p: Modify struct 9p_fid to use a kuid_t not a uid_t Eric W. Biederman
2013-02-13 17:51     ` [PATCH review 11/85] 9p: Modify struct v9fs_session_info to use a kuids and kgids Eric W. Biederman
2013-02-13 17:51     ` [PATCH review 14/85] afs: Remove unused structure afs_store_status Eric W. Biederman
2013-02-13 17:51     ` [PATCH review 15/85] afs: Only allow mounting afs in the intial network namespace Eric W. Biederman
2013-02-13 17:51     ` [PATCH review 18/85] coda: Restrict coda messages to the initial user namespace Eric W. Biederman
     [not found]       ` <1360777934-5663-18-git-send-email-ebiederm-aS9lmoZGLiVWk0Htik3J/w@public.gmane.org>
2013-02-13 18:14         ` Jan Harkes
     [not found]           ` <20130213181450.GJ30800-ETDLCGt7PQU3uPMLIKxrzw@public.gmane.org>
2013-02-13 18:44             ` Eric W. Biederman
2013-02-13 17:51     ` [PATCH review 19/85] coda: Cache permisions in struct coda_inode_info in a kuid_t Eric W. Biederman
2013-02-13 17:51     ` [PATCH review 20/85] coda: Allow coda to be built when user namespace support is enabled Eric W. Biederman
2013-02-13 17:51     ` [PATCH review 21/85] ocfs2: Handle kuids and kgids in acl/xattr conversions Eric W. Biederman
2013-02-13 17:51     ` [PATCH review 24/85] ocfs2: For tracing report the uid and gid values in the initial user namespace Eric W. Biederman
2013-02-13 17:51     ` [PATCH review 25/85] ocfs2: Compare kuids and kgids using uid_eq and gid_eq Eric W. Biederman
     [not found]       ` <1360777934-5663-25-git-send-email-ebiederm-aS9lmoZGLiVWk0Htik3J/w@public.gmane.org>
2013-02-14  8:37         ` Joel Becker
2013-02-13 17:51     ` [PATCH review 26/85] ocfs2: Enable building with user namespaces enabled Eric W. Biederman
     [not found]       ` <1360777934-5663-26-git-send-email-ebiederm-aS9lmoZGLiVWk0Htik3J/w@public.gmane.org>
2013-02-14  8:38         ` Joel Becker
2013-02-13 17:51     ` [PATCH review 35/85] gfs2: Remove the QUOTA_USER and QUOTA_GROUP defines Eric W. Biederman
2013-02-13 17:51     ` [PATCH review 36/85] gfs2: Use kuid_t and kgid_t types where appropriate Eric W. Biederman
2013-02-13 17:51     ` [PATCH review 37/85] gfs2: Use uid_eq and gid_eq " Eric W. Biederman
2013-02-13 17:51     ` [PATCH review 38/85] gfs2: Convert uids and gids between dinodes and vfs inodes Eric W. Biederman
2013-02-13 17:51     ` [PATCH review 39/85] gfs2: Enable building with user namespaces enabled Eric W. Biederman
2013-02-13 17:51     ` [PATCH review 42/85] sunrpc: Use userns friendly constants Eric W. Biederman
2013-02-13 17:51     ` [PATCH review 43/85] sunrpc: Use kuid_t and kgid_t where appropriate Eric W. Biederman
2013-02-13 17:51     ` [PATCH review 44/85] sunrpc: Use uid_eq and gid_eq " Eric W. Biederman
2013-02-13 17:51     ` [PATCH review 46/85] sunrpc: Convert kuids and kgids to uids and gids for printing Eric W. Biederman
2013-02-13 17:51     ` [PATCH review 47/85] sunrpc: Use gid_valid to test for gid != INVALID_GID Eric W. Biederman
2013-02-13 17:51     ` [PATCH review 48/85] sunrpc: Update gss uid to security context mapping Eric W. Biederman
     [not found]       ` <1360777934-5663-48-git-send-email-ebiederm-aS9lmoZGLiVWk0Htik3J/w@public.gmane.org>
2013-02-13 20:35         ` J. Bruce Fields
     [not found]           ` <20130213203528.GN14195-uC3wQj2KruNg9hUCZPvPmw@public.gmane.org>
2013-02-13 21:17             ` Eric W. Biederman
     [not found]               ` <87fw0zrk0e.fsf-aS9lmoZGLiVWk0Htik3J/w@public.gmane.org>
2013-02-13 21:24                 ` J. Bruce Fields
2013-02-13 17:51     ` [PATCH review 50/85] sunrpc: Hash uids by first computing their value in the initial userns Eric W. Biederman
2013-02-13 17:51     ` [PATCH review 51/85] sunrpc: Properly encode kuids and kgids in RPC_AUTH_UNIX credentials Eric W. Biederman
2013-02-13 17:51     ` [PATCH review 52/85] sunrpc: Properly encode kuids and kgids in auth.unix.gid rpc pipe upcalls Eric W. Biederman
2013-02-13 21:05       ` J. Bruce Fields [this message]
     [not found]         ` <20130213210545.GO14195-uC3wQj2KruNg9hUCZPvPmw@public.gmane.org>
2013-02-13 21:29           ` Eric W. Biederman
     [not found]             ` <874nhfrjgg.fsf-aS9lmoZGLiVWk0Htik3J/w@public.gmane.org>
2013-02-13 21:50               ` J. Bruce Fields
     [not found]                 ` <20130213215047.GR14195-uC3wQj2KruNg9hUCZPvPmw@public.gmane.org>
2013-02-13 22:32                   ` Eric W. Biederman
2013-02-13 22:58                     ` J. Bruce Fields
     [not found]                       ` <20130213225840.GV14195-uC3wQj2KruNg9hUCZPvPmw@public.gmane.org>
2013-02-13 23:22                         ` Eric W. Biederman
2013-02-14  7:12                           ` Stanislav Kinsbursky
     [not found]                             ` <511C8E50.8080007-bzQdu9zFT3WakBO8gow8eQ@public.gmane.org>
2013-02-14  8:42                               ` Eric W. Biederman
     [not found]                           ` <87ip5vn6iv.fsf-aS9lmoZGLiVWk0Htik3J/w@public.gmane.org>
2013-02-15 17:00                             ` J. Bruce Fields
     [not found]                               ` <20130215170052.GN8343-uC3wQj2KruNg9hUCZPvPmw@public.gmane.org>
2013-02-15 18:47                                 ` Eric W. Biederman
2013-02-13 17:51     ` [PATCH review 54/85] nfs: Pass GLOBAL_ROOT_UID and GLOBAL_ROOT_GID to keyring alloc Eric W. Biederman
2013-02-13 17:51     ` [PATCH review 55/85] nfs: Convert struct nfs_fattr to Use kuid_t and kgid_t Eric W. Biederman
2013-02-13 17:51     ` [PATCH review 57/85] nfs: Convert nfs2xdr to use kuids and kgids Eric W. Biederman
2013-02-13 17:51     ` [PATCH review 59/85] nfs: Convert nfs4xdr " Eric W. Biederman
2013-02-13 17:51     ` [PATCH review 60/85] nfs: kuid and kgid conversions for nfs/inode.c Eric W. Biederman
2013-02-13 17:51     ` [PATCH review 63/85] nfsd: idmap use u32 not uid_t as the intermediate type Eric W. Biederman
2013-02-13 17:51     ` [PATCH review 65/85] nfsd: Remove nfsd_luid, nfsd_lgid, nfsd_ruid and nfsd_rgid Eric W. Biederman
2013-02-13 17:51     ` [PATCH review 66/85] nfsd: Convert nfs3xdr to use kuids and kgids Eric W. Biederman
2013-02-13 17:51     ` [PATCH review 67/85] nfsd: Convert nfsxdr " Eric W. Biederman
2013-02-13 17:52     ` [PATCH review 71/85] nfsd: Properly compare and initialize " Eric W. Biederman
2013-02-13 17:52     ` [PATCH review 75/85] cifs: Pass GLOBAL_ROOT_UID and GLOBAL_ROOT_GID to keyring_alloc Eric W. Biederman
2013-02-13 17:52     ` [PATCH review 77/85] cifs: Convert from a kuid before printing current_fsuid Eric W. Biederman
2013-02-13 17:52     ` [PATCH review 80/85] cifs: Convert struct cifs_fattr to use kuid and kgids Eric W. Biederman
2013-02-13 17:52     ` [PATCH review 81/85] cifs: Convert struct cifsFileInfo to use a kuid Eric W. Biederman
2013-02-13 17:51   ` [PATCH review 27/85] gfs2: Remove improper checks in gfs2_set_dqblk Eric W. Biederman
2013-02-13 17:51   ` [PATCH review 28/85] gfs2: Split NO_QUOTA_CHANGE inot NO_UID_QUTOA_CHANGE and NO_GID_QUTOA_CHANGE Eric W. Biederman
2013-02-13 17:51   ` [PATCH review 29/85] gfs2: Report quotas in the caller's user namespace Eric W. Biederman
2013-02-13 17:51   ` [PATCH review 30/85] gfs2: Introduce qd2index Eric W. Biederman
2013-02-13 17:51   ` [PATCH review 31/85] gfs2: Modify struct gfs2_quota_change_host to use struct kqid Eric W. Biederman
2013-02-13 17:51   ` [PATCH review 32/85] gfs2: Modify qdsb_get to take a " Eric W. Biederman
2013-02-13 17:51   ` [PATCH review 33/85] gfs2: Convert gfs2_quota_refresh to take a kqid Eric W. Biederman
2013-02-13 17:51   ` [PATCH review 34/85] gfs2: Store qd_id in struct gfs2_quota_data as a struct kqid Eric W. Biederman
2013-02-13 17:51   ` [PATCH review 40/85] ncpfs: Support interacting with multiple user namespaces Eric W. Biederman
2013-02-13 17:51   ` [PATCH review 41/85] nfs_common: Update the translation between nfsv3 acls linux posix acls Eric W. Biederman
2013-02-13 17:51   ` [PATCH review 45/85] sunrpc: Simplify auth_unix now that everything is a kgid_t Eric W. Biederman
2013-02-13 17:51   ` [PATCH review 49/85] sunrpc: Update svcgss xdr handle to rpsec_contect cache Eric W. Biederman
     [not found]     ` <1360777934-5663-49-git-send-email-ebiederm-aS9lmoZGLiVWk0Htik3J/w@public.gmane.org>
2013-03-04 14:12       ` J. Bruce Fields
2013-03-04 17:11         ` Eric W. Biederman
     [not found]           ` <87txorjdjf.fsf-aS9lmoZGLiVWk0Htik3J/w@public.gmane.org>
2013-03-05 23:10             ` J. Bruce Fields
     [not found]               ` <20130305231001.GJ15816-uC3wQj2KruNg9hUCZPvPmw@public.gmane.org>
2013-03-05 23:43                 ` Eric W. Biederman
2013-02-13 17:51   ` [PATCH review 53/85] sunrpc: Properly decode kuids and kgids in RPC_AUTH_UNIX credentials Eric W. Biederman
2013-02-13 17:51   ` [PATCH review 56/85] nfs: Convert idmap to use kuids and kgids Eric W. Biederman
2013-02-13 17:51   ` [PATCH review 58/85] nfs: Convert nfs3xdr " Eric W. Biederman
2013-02-13 17:51   ` [PATCH review 61/85] nfs: Enable building with user namespaces enabled Eric W. Biederman
2013-02-13 17:51   ` [PATCH review 62/85] nfsd: Remove declaration of nonexistent nfs4_acl_permisison Eric W. Biederman
2013-02-13 17:51   ` [PATCH review 64/85] nfsd: Convert idmap to use kuids and kgids Eric W. Biederman
2013-02-13 17:51   ` [PATCH review 68/85] nfsd: Handle kuids and kgids in the nfs4acl to posix_acl conversion Eric W. Biederman
2013-02-13 17:51   ` [PATCH review 69/85] nfsd: Modify nfsd4_cb_sec to use kuids and kgids Eric W. Biederman
2013-02-13 17:51   ` [PATCH review 70/85] nfsd: Store ex_anon_uid and ex_anon_gid as " Eric W. Biederman
2013-02-13 17:52   ` [PATCH review 72/85] nfsd: Enable building with user namespaces enabled Eric W. Biederman
2013-02-13 17:52   ` [PATCH review 73/85] cifs: Override unmappable incoming uids and gids Eric W. Biederman
2013-02-13 17:52   ` [PATCH review 74/85] cifs: Use BUILD_BUG_ON to validate uids and gids are the same size Eric W. Biederman
2013-02-13 17:52   ` [PATCH review 76/85] cifs: Use kuids and kgids SID to uid/gid mapping Eric W. Biederman
2013-02-13 17:52   ` [PATCH review 78/85] cifs: Modify struct cifs_unix_set_info_args to hold a kuid_t and a kgid_t Eric W. Biederman
2013-02-13 17:52   ` [PATCH review 79/85] cifs: Convert struct tcon_link to use a kuid Eric W. Biederman
2013-02-13 17:52   ` [PATCH review 82/85] cifs: Modify struct smb_vol to use kuids and kgids Eric W. Biederman
2013-02-13 17:52   ` [PATCH review 83/85] cifs: Convert struct cifs_sb_info " Eric W. Biederman
2013-02-13 17:52   ` [PATCH review 84/85] cifs: Convert struct cifs_ses to use a kuid_t and a kgid_t Eric W. Biederman
2013-02-13 17:52   ` [PATCH review 85/85] cifs: Enable building with user namespaces enabled Eric W. Biederman

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20130213210545.GO14195@fieldses.org \
    --to=bfields@fieldses.org \
    --cc=Trond.Myklebust@netapp.com \
    --cc=containers@lists.linux-foundation.org \
    --cc=ebiederm@xmission.com \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=serge@hallyn.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).