linux-security-module.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Casey Schaufler <casey@schaufler-ca.com>
To: Stephen Smalley <sds@tycho.nsa.gov>,
	casey.schaufler@intel.com, jmorris@namei.org,
	linux-security-module@vger.kernel.org, selinux@vger.kernel.org
Cc: keescook@chromium.org, john.johansen@canonical.com,
	penguin-kernel@i-love.sakura.ne.jp, paul@paul-moore.com,
	casey@schaufler-ca.com
Subject: Re: [PATCH v4 15/23] LSM: Specify which LSM to display
Date: Fri, 28 Jun 2019 09:15:00 -0700	[thread overview]
Message-ID: <7944672e-a590-44a3-743a-48c1785a5464@schaufler-ca.com> (raw)
In-Reply-To: <89561452-86f9-fd9a-1390-0cd4bde1fd62@tycho.nsa.gov>

On 6/28/2019 7:45 AM, Stephen Smalley wrote:
> On 6/26/19 3:22 PM, Casey Schaufler wrote:
>> Create a new entry "display" in /proc/.../attr for controlling
>> which LSM security information is displayed for a process.
>> 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.
>
> IIUC, presently, at the end of the entire series,
>
> 1) Any process can change its display attribute to any enabled security module, and no security module can veto that change.

That is correct. If a security module could hoard the display it
could prevent user space from functioning in a multiple module
environment.

> 2) The display attribute is inherited across fork and exec, even execs that change credentials, and again no security module has control over the inheritance of this attribute.

Also correct. Scripts don't work otherwise.

>
> 3) Setting the display attribute affects more than just the contexts read or written by the process itself:
> - Contexts reported in audit logs,
> - Contexts passed across binder (generated in sender context, delivered to receiver),
> - Contexts passed to NFS servers for new files,
> - Contexts returned by NFS servers for existing files,
> - Netlink-related contexts (?possibly generated in sender context rather than receiver context?),
> - This list may not be complete.

Any of which can be changed should a more rational behavior be proposed.
One possibility is to use lsm='value',lsm='value' encoding for internal
communications, but there's been considerable resistance to anything
like that.

> 4) A security_secid_to_secctx() in one process' context (e.g. sender) or with one display value followed later by a security_secctx_to_secid() call in a different process' context (e.g. receiver) or with a different display value may ask a different security module to perform the reverse translation of the context than the forward translation.

Do you have an example of where this might happen?
Contexts are rarely used within the kernel. The usual
behavior is to generate them, send them out to user space,
and delete them. They get cached in some networking code,
but not in cases where more than one (existing) security
module will ever use them. Binder may be an exception, but
only SELinux (currently) supports binder.


> Is that correct?  If so, it seems problematic.

Balancing backward compatibility with new behavior is hard!
What would you suggest for audit logs? Should we put all LSM
data in every record? Is NFS a concern for anyone not using
SELinux?

There is no user space that uses display, and it's going
to take some time to work out all the kinks before we even
think about teaching systemd about it.

>     
>>
>> Signed-off-by: Casey Schaufler <casey@schaufler-ca.com>
>> ---
>>   fs/proc/base.c      |   1 +
>>   security/security.c | 129 ++++++++++++++++++++++++++++++++++++++------
>>   2 files changed, 113 insertions(+), 17 deletions(-)
>>
>> diff --git a/fs/proc/base.c b/fs/proc/base.c
>> index ddef482f1334..7bf70e041315 100644
>> --- a/fs/proc/base.c
>> +++ b/fs/proc/base.c
>> @@ -2618,6 +2618,7 @@ static const struct pid_entry attr_dir_stuff[] = {
>>       ATTR(NULL, "fscreate",        0666),
>>       ATTR(NULL, "keycreate",        0666),
>>       ATTR(NULL, "sockcreate",    0666),
>> +    ATTR(NULL, "display",        0666),
>>   #ifdef CONFIG_SECURITY_SMACK
>>       DIR("smack",            0555,
>>           proc_smack_attr_dir_inode_ops, proc_smack_attr_dir_ops),
>> diff --git a/security/security.c b/security/security.c
>> index 3180a6f30625..82e29c477fa4 100644
>> --- a/security/security.c
>> +++ b/security/security.c
>> @@ -46,7 +46,9 @@ static struct kmem_cache *lsm_file_cache;
>>   static struct kmem_cache *lsm_inode_cache;
>>     char *lsm_names;
>> -static struct lsm_blob_sizes blob_sizes __lsm_ro_after_init;
>> +static struct lsm_blob_sizes blob_sizes __lsm_ro_after_init = {
>> +    .lbs_task = sizeof(int),    /* slot number for the "display" LSM */
>> +};
>>     /* Boot-time LSM user choice */
>>   static __initdata const char *chosen_lsm_order;
>> @@ -423,8 +425,10 @@ static int lsm_append(const char *new, char **result)
>>     /*
>>    * Current index to use while initializing the lsmblob secid list.
>> + * Pointers to the LSM id structures for local use.
>>    */
>>   static int lsm_slot;
>> +static struct lsm_id *lsm_slotlist[LSMBLOB_ENTRIES];
>>     /**
>>    * security_add_hooks - Add a modules hooks to the hook lists.
>> @@ -444,6 +448,7 @@ void __init security_add_hooks(struct security_hook_list *hooks, int count,
>>       if (lsmid->slot == LSMBLOB_NEEDED) {
>>           if (lsm_slot >= LSMBLOB_ENTRIES)
>>               panic("%s Too many LSMs registered.\n", __func__);
>> +        lsm_slotlist[lsm_slot] = lsmid;
>>           lsmid->slot = lsm_slot++;
>>           init_debug("%s assigned lsmblob slot %d\n", lsmid->lsm,
>>                  lsmid->slot);
>> @@ -564,6 +569,8 @@ int lsm_inode_alloc(struct inode *inode)
>>    */
>>   static int lsm_task_alloc(struct task_struct *task)
>>   {
>> +    int *display;
>> +
>>       if (blob_sizes.lbs_task == 0) {
>>           task->security = NULL;
>>           return 0;
>> @@ -572,6 +579,15 @@ static int lsm_task_alloc(struct task_struct *task)
>>       task->security = kzalloc(blob_sizes.lbs_task, GFP_KERNEL);
>>       if (task->security == NULL)
>>           return -ENOMEM;
>> +
>> +    /*
>> +     * The start of the task blob contains the "display" LSM slot number.
>> +     * Start with it set to the invalid slot number, indicating that the
>> +     * default first registered LSM be displayed.
>> +     */
>> +    display = task->security;
>> +    *display = LSMBLOB_INVALID;
>> +
>>       return 0;
>>   }
>>   @@ -1563,14 +1579,24 @@ int security_file_open(struct file *file)
>>     int security_task_alloc(struct task_struct *task, unsigned long clone_flags)
>>   {
>> +    int *odisplay = current->security;
>> +    int *ndisplay;
>>       int rc = lsm_task_alloc(task);
>>   -    if (rc)
>> +    if (unlikely(rc))
>>           return rc;
>> +
>>       rc = call_int_hook(task_alloc, 0, task, clone_flags);
>> -    if (unlikely(rc))
>> +    if (unlikely(rc)) {
>>           security_task_free(task);
>> -    return rc;
>> +        return rc;
>> +    }
>> +
>> +    ndisplay = task->security;
>> +    if (ndisplay && odisplay)
>> +        *ndisplay = *odisplay;
>> +
>> +    return 0;
>>   }
>>     void security_task_free(struct task_struct *task)
>> @@ -1967,10 +1993,29 @@ int security_getprocattr(struct task_struct *p, const char *lsm, char *name,
>>                   char **value)
>>   {
>>       struct security_hook_list *hp;
>> +    int *display = current->security;
>> +    int slot = 0;
>> +
>> +    if (!strcmp(name, "display")) {
>> +        /*
>> +         * lsm_slot will be 0 if there are no displaying modules.
>> +         */
>> +        if (lsm_slot == 0)
>> +            return -EINVAL;
>> +        if (*display != LSMBLOB_INVALID)
>> +            slot = *display;
>> +        *value = kstrdup(lsm_slotlist[slot]->lsm, GFP_KERNEL);
>> +        if (*value)
>> +            return strlen(*value);
>> +        return -ENOMEM;
>> +    }
>>         hlist_for_each_entry(hp, &security_hook_heads.getprocattr, list) {
>>           if (lsm != NULL && strcmp(lsm, hp->lsmid->lsm))
>>               continue;
>> +        if (lsm == NULL && *display != LSMBLOB_INVALID &&
>> +            *display != hp->lsmid->slot)
>> +            continue;
>>           return hp->hook.getprocattr(p, name, value);
>>       }
>>       return -EINVAL;
>> @@ -1980,10 +2025,46 @@ int security_setprocattr(const char *lsm, const char *name, void *value,
>>                size_t size)
>>   {
>>       struct security_hook_list *hp;
>> +    char *term;
>> +    char *cp;
>> +    int *display = current->security;
>> +    int rc = -EINVAL;
>> +    int slot = 0;
>> +
>> +    if (!strcmp(name, "display")) {
>> +        /*
>> +         * lsm_slot will be 0 if there are no displaying modules.
>> +         */
>> +        if (lsm_slot == 0 || size == 0)
>> +            return -EINVAL;
>> +        cp = kzalloc(size + 1, GFP_KERNEL);
>> +        if (cp == NULL)
>> +            return -ENOMEM;
>> +        memcpy(cp, value, size);
>> +
>> +        term = strchr(cp, ' ');
>> +        if (term == NULL)
>> +            term = strchr(cp, '\n');
>> +        if (term != NULL)
>> +            *term = '\0';
>> +
>> +        for (slot = 0; slot < lsm_slot; slot++)
>> +            if (!strcmp(cp, lsm_slotlist[slot]->lsm)) {
>> +                *display = lsm_slotlist[slot]->slot;
>> +                rc = size;
>> +                break;
>> +            }
>> +
>> +        kfree(cp);
>> +        return rc;
>> +    }
>>         hlist_for_each_entry(hp, &security_hook_heads.setprocattr, list) {
>>           if (lsm != NULL && strcmp(lsm, hp->lsmid->lsm))
>>               continue;
>> +        if (lsm == NULL && *display != LSMBLOB_INVALID &&
>> +            *display != hp->lsmid->slot)
>> +            continue;
>>           return hp->hook.setprocattr(name, value, size);
>>       }
>>       return -EINVAL;
>> @@ -2003,15 +2084,15 @@ EXPORT_SYMBOL(security_ismaclabel);
>>   int security_secid_to_secctx(struct lsmblob *blob, char **secdata, u32 *seclen)
>>   {
>>       struct security_hook_list *hp;
>> -    int rc;
>> +    int *display = current->security;
>>         hlist_for_each_entry(hp, &security_hook_heads.secid_to_secctx, list) {
>>           if (WARN_ON(hp->lsmid->slot < 0 || hp->lsmid->slot >= lsm_slot))
>>               continue;
>> -        rc = hp->hook.secid_to_secctx(blob->secid[hp->lsmid->slot],
>> -                          secdata, seclen);
>> -        if (rc != 0)
>> -            return rc;
>> +        if (*display == LSMBLOB_INVALID || *display == hp->lsmid->slot)
>> +            return hp->hook.secid_to_secctx(
>> +                    blob->secid[hp->lsmid->slot],
>> +                    secdata, seclen);
>>       }
>>       return 0;
>>   }
>> @@ -2021,16 +2102,15 @@ int security_secctx_to_secid(const char *secdata, u32 seclen,
>>                    struct lsmblob *blob)
>>   {
>>       struct security_hook_list *hp;
>> -    int rc;
>> +    int *display = current->security;
>>         lsmblob_init(blob, 0);
>>       hlist_for_each_entry(hp, &security_hook_heads.secctx_to_secid, list) {
>>           if (WARN_ON(hp->lsmid->slot < 0 || hp->lsmid->slot >= lsm_slot))
>>               continue;
>> -        rc = hp->hook.secctx_to_secid(secdata, seclen,
>> -                          &blob->secid[hp->lsmid->slot]);
>> -        if (rc != 0)
>> -            return rc;
>> +        if (*display == LSMBLOB_INVALID || *display == hp->lsmid->slot)
>> +            return hp->hook.secctx_to_secid(secdata, seclen,
>> +                        &blob->secid[hp->lsmid->slot]);
>>       }
>>       return 0;
>>   }
>> @@ -2038,7 +2118,15 @@ EXPORT_SYMBOL(security_secctx_to_secid);
>>     void security_release_secctx(char *secdata, u32 seclen)
>>   {
>> -    call_void_hook(release_secctx, secdata, seclen);
>> +    struct security_hook_list *hp;
>> +    int *display = current->security;
>> +
>> +    hlist_for_each_entry(hp, &security_hook_heads.release_secctx, list)
>> +        if (*display == LSMBLOB_INVALID ||
>> +            *display == hp->lsmid->slot) {
>> +            hp->hook.release_secctx(secdata, seclen);
>> +            return;
>> +        }
>>   }
>>   EXPORT_SYMBOL(security_release_secctx);
>>   @@ -2163,8 +2251,15 @@ EXPORT_SYMBOL(security_sock_rcv_skb);
>>   int security_socket_getpeersec_stream(struct socket *sock, char __user *optval,
>>                         int __user *optlen, unsigned len)
>>   {
>> -    return call_int_hook(socket_getpeersec_stream, -ENOPROTOOPT, sock,
>> -                optval, optlen, len);
>> +    int *display = current->security;
>> +    struct security_hook_list *hp;
>> +
>> +    hlist_for_each_entry(hp, &security_hook_heads.socket_getpeersec_stream,
>> +                 list)
>> +        if (*display == LSMBLOB_INVALID || *display == hp->lsmid->slot)
>> +            return hp->hook.socket_getpeersec_stream(sock, optval,
>> +                                 optlen, len);
>> +    return -ENOPROTOOPT;
>>   }
>>     int security_socket_getpeersec_dgram(struct socket *sock, struct sk_buff *skb,
>>
>

  reply	other threads:[~2019-06-28 16:15 UTC|newest]

Thread overview: 96+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-06-26 19:22 [PATCH v4 00/23] LSM: Module stacking for AppArmor Casey Schaufler
2019-06-26 19:22 ` [PATCH v4 01/23] LSM: Infrastructure management of the superblock Casey Schaufler
2019-06-26 19:22 ` [PATCH v4 02/23] LSM: Infrastructure management of the sock security Casey Schaufler
2019-06-26 19:22 ` [PATCH v4 03/23] LSM: Infrastructure management of the key blob Casey Schaufler
2019-06-26 19:22 ` [PATCH v4 04/23] LSM: Create and manage the lsmblob data structure Casey Schaufler
2019-06-26 23:39   ` John Johansen
2019-06-27 17:43     ` Casey Schaufler
2019-06-26 19:22 ` [PATCH v4 05/23] LSM: Use lsmblob in security_audit_rule_match Casey Schaufler
2019-06-26 23:07   ` Kees Cook
2019-06-26 23:45   ` John Johansen
2019-06-26 19:22 ` [PATCH v4 06/23] LSM: Use lsmblob in security_kernel_act_as Casey Schaufler
2019-06-26 23:07   ` Kees Cook
2019-06-26 23:47   ` John Johansen
2019-06-26 19:22 ` [PATCH v4 07/23] net: Prepare UDS for secuirty module stacking Casey Schaufler
2019-06-26 23:07   ` Kees Cook
2019-06-26 23:48   ` John Johansen
2019-06-26 19:22 ` [PATCH v4 08/23] LSM: Use lsmblob in security_secctx_to_secid Casey Schaufler
2019-06-26 23:07   ` Kees Cook
2019-06-26 23:50   ` John Johansen
2019-06-26 19:22 ` [PATCH v4 09/23] LSM: Use lsmblob in security_secid_to_secctx Casey Schaufler
2019-06-26 23:07   ` Kees Cook
2019-06-26 23:51   ` John Johansen
2019-06-26 19:22 ` [PATCH v4 10/23] LSM: Use lsmblob in security_ipc_getsecid Casey Schaufler
2019-06-26 23:08   ` Kees Cook
2019-06-26 23:53   ` John Johansen
2019-06-26 19:22 ` [PATCH v4 11/23] LSM: Use lsmblob in security_task_getsecid Casey Schaufler
2019-06-26 23:08   ` Kees Cook
2019-06-26 23:55   ` John Johansen
2019-06-26 19:22 ` [PATCH v4 12/23] LSM: Use lsmblob in security_inode_getsecid Casey Schaufler
2019-06-26 23:08   ` Kees Cook
2019-06-26 23:56   ` John Johansen
2019-06-26 19:22 ` [PATCH v4 13/23] LSM: Use lsmblob in security_cred_getsecid Casey Schaufler
2019-06-26 23:09   ` Kees Cook
2019-06-26 23:57   ` John Johansen
2019-06-26 19:22 ` [PATCH v4 14/23] IMA: Change internal interfaces to use lsmblobs Casey Schaufler
2019-06-26 23:09   ` Kees Cook
2019-06-26 23:58   ` John Johansen
2019-06-26 19:22 ` [PATCH v4 15/23] LSM: Specify which LSM to display Casey Schaufler
2019-06-26 23:12   ` Kees Cook
2019-06-27 21:33   ` John Johansen
2019-06-28 14:45   ` Stephen Smalley
2019-06-28 16:15     ` Casey Schaufler [this message]
2019-06-28 18:08       ` John Johansen
     [not found]       ` <CAB9W1A1nwE7WBZqTe-GV8xNb83_B2ybV7cco++nfMjtDz9NJrg@mail.gmail.com>
2019-06-29  1:01         ` Fwd: " Stephen Smalley
2019-06-29 19:45           ` Casey Schaufler
2019-07-02  0:49           ` James Morris
2019-07-02  1:20             ` Casey Schaufler
2019-06-26 19:22 ` [PATCH v4 16/23] LSM: Use lsmcontext in security_secid_to_secctx Casey Schaufler
2019-06-26 23:14   ` Kees Cook
2019-06-27 21:34   ` John Johansen
2019-06-26 19:22 ` [PATCH v4 17/23] " Casey Schaufler
2019-06-26 23:15   ` Kees Cook
2019-06-27  3:53   ` Kees Cook
2019-06-27 16:29     ` Casey Schaufler
2019-06-27 17:17       ` Kees Cook
2019-06-27 17:36         ` Casey Schaufler
2019-06-27 21:34   ` John Johansen
2019-06-27 21:36     ` John Johansen
2019-06-26 19:22 ` [PATCH v4 18/23] LSM: Use lsmcontext in security_dentry_init_security Casey Schaufler
2019-06-26 23:15   ` Kees Cook
2019-06-27 21:34   ` John Johansen
2019-06-28 10:07   ` Ondrej Mosnacek
2019-06-26 19:22 ` [PATCH v4 19/23] LSM: Use lsmcontext in security_inode_getsecctx Casey Schaufler
2019-06-26 23:16   ` Kees Cook
2019-06-27 21:36   ` John Johansen
2019-06-26 19:22 ` [PATCH v4 20/23] LSM: security_secid_to_secctx in netlink netfilter Casey Schaufler
2019-06-26 23:16   ` Kees Cook
2019-06-27 21:37   ` John Johansen
2019-06-26 19:22 ` [PATCH v4 21/23] Audit: Store LSM audit information in an lsmblob Casey Schaufler
2019-06-26 23:16   ` Kees Cook
2019-06-27 21:37   ` John Johansen
2019-06-26 19:22 ` [PATCH v4 22/23] NET: Store LSM netlabel data in a lsmblob Casey Schaufler
2019-06-26 23:18   ` Kees Cook
2019-06-27 21:38   ` John Johansen
2019-06-26 19:22 ` [PATCH v4 23/23] AppArmor: Remove the exclusive flag Casey Schaufler
2019-06-26 23:18   ` Kees Cook
2019-06-27  2:22   ` James Morris
2019-06-27  3:28     ` Kees Cook
2019-06-27  3:44     ` John Johansen
2019-06-27  3:49       ` James Morris
2019-06-27 21:38   ` John Johansen
2019-06-26 21:04 ` [PATCH v4 00/23] LSM: Module stacking for AppArmor Kees Cook
2019-06-26 21:11   ` John Johansen
2019-06-26 23:04     ` Kees Cook
2019-06-26 21:25   ` Casey Schaufler
2019-06-27  2:41 ` James Morris
2019-06-27  2:46   ` James Morris
2019-06-27  3:45     ` James Morris
2019-06-27  4:10       ` James Morris
2019-06-27 17:07         ` Kees Cook
2019-06-27 18:10           ` James Morris
2019-06-27  3:51     ` Kees Cook
2019-06-27 21:52 ` John Johansen
2019-06-27 22:33   ` Casey Schaufler
2019-06-27 23:16   ` James Morris
2019-06-27 23:44     ` John Johansen

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=7944672e-a590-44a3-743a-48c1785a5464@schaufler-ca.com \
    --to=casey@schaufler-ca.com \
    --cc=casey.schaufler@intel.com \
    --cc=jmorris@namei.org \
    --cc=john.johansen@canonical.com \
    --cc=keescook@chromium.org \
    --cc=linux-security-module@vger.kernel.org \
    --cc=paul@paul-moore.com \
    --cc=penguin-kernel@i-love.sakura.ne.jp \
    --cc=sds@tycho.nsa.gov \
    --cc=selinux@vger.kernel.org \
    /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).