* Re: [Qemu-devel] [PATCH kernel v4 7/7] virtio-balloon: tell host vm's unused page info
From: Michael S. Tsirkin @ 2016-11-08 21:07 UTC (permalink / raw)
To: Dave Hansen
Cc: Li, Liang Z, pbonzini@redhat.com, amit.shah@redhat.com,
quintela@redhat.com, dgilbert@redhat.com, qemu-devel@nongnu.org,
kvm@vger.kernel.org, virtio-dev@lists.oasis-open.org,
linux-mm@kvack.org, linux-kernel@vger.kernel.org,
virtualization@lists.linux-foundation.org,
mgorman@techsingularity.net, cornelia.huck@de.ibm.com
In-Reply-To: <281acd8d-fd94-6318-35e5-9eb130303dc6@intel.com>
On Mon, Nov 07, 2016 at 09:23:38AM -0800, Dave Hansen wrote:
> On 11/06/2016 07:37 PM, Li, Liang Z wrote:
> >> Let's say we do a 32k bitmap that can hold ~1M pages. That's 4GB of RAM.
> >> On a 1TB system, that's 256 passes through the top-level loop.
> >> The bottom-level lists have tens of thousands of pages in them, even on my
> >> laptop. Only 1/256 of these pages will get consumed in a given pass.
> >>
> > Your description is not exactly.
> > A 32k bitmap is used only when there is few free memory left in the system and when
> > the extend_page_bitmap() failed to allocate more memory for the bitmap. Or dozens of
> > 32k split bitmap will be used, this version limit the bitmap count to 32, it means we can use
> > at most 32*32 kB for the bitmap, which can cover 128GB for RAM. We can increase the bitmap
> > count limit to a larger value if 32 is not big enough.
>
> OK, so it tries to allocate a large bitmap. But, if it fails, it will
> try to work with a smaller bitmap. Correct?
>
> So, what's the _worst_ case? It sounds like it is even worse than I was
> positing.
>
> >> That's an awfully inefficient way of doing it. This patch essentially changed
> >> the data structure without changing the algorithm to populate it.
> >>
> >> Please change the *algorithm* to use the new data structure efficiently.
> >> Such a change would only do a single pass through each freelist, and would
> >> choose whether to use the extent-based (pfn -> range) or bitmap-based
> >> approach based on the contents of the free lists.
> >
> > Save the free page info to a raw bitmap first and then process the raw bitmap to
> > get the proper ' extent-based ' and 'bitmap-based' is the most efficient way I can
> > come up with to save the virtio data transmission. Do you have some better idea?
>
> That's kinda my point. This patch *does* processing to try to pack the
> bitmaps full of pages from the various pfn ranges. It's a form of
> processing that gets *REALLY*, *REALLY* bad in some (admittedly obscure)
> cases.
>
> Let's not pretend that making an essentially unlimited number of passes
> over the free lists is not processing.
>
> 1. Allocate as large of a bitmap as you can. (what you already do)
> 2. Iterate from the largest freelist order. Store those pages in the
> bitmap.
> 3. If you can no longer fit pages in the bitmap, return the list that
> you have.
> 4. Make an approximation about where the bitmap does not make any more,
> and fall back to listing individual PFNs. This would make sens, for
> instance in a large zone with very few free order-0 pages left.
In practice, a single PFN using the bitmap format
only takes up twice the size: I think it's 128 instead of 64 bit
per entry.
So it's not a a given that point 4 is worth it at any point,
just packing multiple bitmaps might be good enough.
>
> > It seems the benefit we get for this feature is not as big as that in fast balloon inflating/deflating.
> >>
> >> You should not be using get_max_pfn(). Any patch set that continues to use
> >> it is not likely to be using a proper algorithm.
> >
> > Do you have any suggestion about how to avoid it?
>
> Yes: get the pfns from the page free lists alone. Don't derive them
> from the pfn limits of the system or zones.
^ permalink raw reply
* Re: [PATCH lttng-tools v2] Fix: test cases now rely on explicit workloads
From: Mathieu Desnoyers @ 2016-11-08 21:07 UTC (permalink / raw)
To: Francis Deslauriers; +Cc: lttng-dev, Jeremie Galarneau, Nathan Lynch
In-Reply-To: <CADcCL0gSJ68OFsOVjxwcUJdJxTh5DZTJ2KRBR_bkgR-o9PJVhw@mail.gmail.com>
[-- Attachment #1.1: Type: text/plain, Size: 2389 bytes --]
----- On Nov 8, 2016, at 3:46 PM, Francis Deslauriers <francis.deslauriers@efficios.com> wrote:
> 2016-11-08 15:06 GMT-05:00 Nathan Lynch < [ mailto:Nathan_Lynch@mentor.com |
> Nathan_Lynch@mentor.com ] > :
>> On 11/08/2016 01:49 PM, Francis Deslauriers wrote:
>> > Run a process explicitly in the tracing session to generate the enabled events
>> > rather than relying on the events generated by the lttng CLI.
>> [...]
>>> diff --git a/tests/regression/kernel/test_all_events
>> > b/tests/regression/kernel/test_all_events
>> > index 50002f4..80a8020 100755
>> > --- a/tests/regression/kernel/test_all_events
>> > +++ b/tests/regression/kernel/test_all_events
>> > @@ -21,6 +21,8 @@ CURDIR=$(dirname $0)/
>> > TESTDIR=$CURDIR/../..
>> > NUM_TESTS=8
>> > +TESTCMD="grep -V -q > /dev/null"
>> > +
>> What was wrong with /bin/true, as in v1 of this patch?
> I used grep because the configure script looks for it already
> At the moment, we don't check if /bin/true is installed on the machine using the
> configure script.
>> FYI, Busybox grep doesn't know -V:
>> # grep -V -q >/dev/null
>> grep: invalid option -- 'V'
>> BusyBox v1.24.1 (2016-11-08 12:59:13 CST) multi-call binary.
>> Usage: grep [-HhnlLoqvsriwFE] [-m N] [-A/B/C N] PATTERN/-e PATTERN.../-f
>> FILE [FILE]...
>> # echo $?
>> 1
>> It looks like the testcases don't care about the exit status of TESTCMD,
>> but you may want to redirect stderr to /dev/null as well.
> Good point.
> I could do: grep "foo" /dev/null
utils/utils.sh:45:# We set the default lttng-sessiond path to /bin/true to prevent the spawning
utils/utils.sh:49:export LTTNG_SESSIOND_PATH="/bin/true"
already used.
So unless anyone complain about it, let's use it.
If someone complains, then we need to fix the configure detection.
Thanks,
Mathieu
> Thank you,
> --
> Francis Deslauriers
> Software developer
> EfficiOS inc.
>> _______________________________________________
>> lttng-dev mailing list
>> [ mailto:lttng-dev@lists.lttng.org | lttng-dev@lists.lttng.org ]
>> [ https://lists.lttng.org/cgi-bin/mailman/listinfo/lttng-dev |
>> https://lists.lttng.org/cgi-bin/mailman/listinfo/lttng-dev ]
> _______________________________________________
> lttng-dev mailing list
> lttng-dev@lists.lttng.org
> https://lists.lttng.org/cgi-bin/mailman/listinfo/lttng-dev
--
Mathieu Desnoyers
EfficiOS Inc.
http://www.efficios.com
[-- Attachment #1.2: Type: text/html, Size: 4602 bytes --]
[-- Attachment #2: Type: text/plain, Size: 156 bytes --]
_______________________________________________
lttng-dev mailing list
lttng-dev@lists.lttng.org
https://lists.lttng.org/cgi-bin/mailman/listinfo/lttng-dev
^ permalink raw reply
* [PATCH v4 8/9] selinux: Add IB Port SMP access vector
From: Dan Jurgens @ 2016-11-08 21:06 UTC (permalink / raw)
To: chrisw, paul, sds, eparis, dledford, sean.hefty, hal.rosenstock
Cc: selinux, linux-security-module, linux-rdma, yevgenyp, liranl,
leonro, Daniel Jurgens
In-Reply-To: <1478639185-47521-1-git-send-email-danielj@mellanox.com>
From: Daniel Jurgens <danielj@mellanox.com>
Add a type for Infiniband ports and an access vector for subnet
management packets. Implement the ib_port_smp hook to check that the
caller has permission to send and receive SMPs on the end port specified
by the device name and port. Add interface to query the SID for a IB
port, which walks the IB_PORT ocontexts to find an entry for the
given name and port.
issue: 736423
Change-Id: If8b365f3cf32e77a2060073f1a53e27ea846804d
Signed-off-by: Daniel Jurgens <danielj@mellanox.com>
---
v2:
- Shorted ib_end_port. Paul Moore
- Pass void blobs to security hooks. Paul Moore
- Log specific IB port info in audit log. Paul Moore
- Don't create a new intial sid, use unlabeled. Stephen Smalley
- Changed "smp" to "manage_subnet". Paul Moore
v3:
- ib_port -> ib_endport. Paul Moore
- Don't log device name as untrusted string. Paul Moore
- Reorder parameters of LSM hook. Paul Moore
Signed-off-by: Daniel Jurgens <danielj@mellanox.com>
---
include/linux/lsm_audit.h | 8 +++++++
security/lsm_audit.c | 5 +++++
security/selinux/hooks.c | 25 ++++++++++++++++++++++
security/selinux/include/classmap.h | 2 ++
security/selinux/include/security.h | 2 ++
security/selinux/ss/services.c | 42 +++++++++++++++++++++++++++++++++++++
6 files changed, 84 insertions(+)
diff --git a/include/linux/lsm_audit.h b/include/linux/lsm_audit.h
index 402b770..7047b4c 100644
--- a/include/linux/lsm_audit.h
+++ b/include/linux/lsm_audit.h
@@ -21,6 +21,7 @@
#include <linux/path.h>
#include <linux/key.h>
#include <linux/skbuff.h>
+#include <rdma/ib_verbs.h>
struct lsm_network_audit {
int netif;
@@ -50,6 +51,11 @@ struct lsm_pkey_audit {
u16 pkey;
};
+struct lsm_ib_endport_audit {
+ char dev_name[IB_DEVICE_NAME_MAX];
+ u8 port_num;
+};
+
/* Auxiliary data to use in generating the audit record. */
struct common_audit_data {
char type;
@@ -66,6 +72,7 @@ struct common_audit_data {
#define LSM_AUDIT_DATA_IOCTL_OP 11
#define LSM_AUDIT_DATA_FILE 12
#define LSM_AUDIT_DATA_PKEY 13
+#define LSM_AUDIT_DATA_IB_ENDPORT 14
union {
struct path path;
struct dentry *dentry;
@@ -84,6 +91,7 @@ struct common_audit_data {
struct lsm_ioctlop_audit *op;
struct file *file;
struct lsm_pkey_audit *pkey;
+ struct lsm_ib_endport_audit *ib_endport;
} u;
/* this union contains LSM specific data */
union {
diff --git a/security/lsm_audit.c b/security/lsm_audit.c
index b18d277..549fe9d 100644
--- a/security/lsm_audit.c
+++ b/security/lsm_audit.c
@@ -423,6 +423,11 @@ static void dump_common_audit_data(struct audit_buffer *ab,
a->u.pkey->pkey, &sbn_pfx);
break;
}
+ case LSM_AUDIT_DATA_IB_ENDPORT:
+ audit_log_format(ab, " device=%s port_num=%u",
+ a->u.ib_endport->dev_name,
+ a->u.ib_endport->port_num);
+ break;
} /* switch (a->type) */
}
diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c
index 20fb292..ea3f6d0 100644
--- a/security/selinux/hooks.c
+++ b/security/selinux/hooks.c
@@ -6108,6 +6108,29 @@ static int selinux_ib_pkey_access(void *ib_sec, u64 subnet_prefix, u16 pkey_val)
INFINIBAND_PKEY__ACCESS, &ad);
}
+static int selinux_ib_endport_manage_subnet(void *ib_sec, const char *dev_name,
+ u8 port_num)
+{
+ struct common_audit_data ad;
+ int err;
+ u32 sid = 0;
+ struct ib_security_struct *sec = ib_sec;
+ struct lsm_ib_endport_audit ib_endport;
+
+ err = security_ib_endport_sid(dev_name, port_num, &sid);
+
+ if (err)
+ return err;
+
+ ad.type = LSM_AUDIT_DATA_IB_ENDPORT;
+ strncpy(ib_endport.dev_name, dev_name, sizeof(ib_endport.dev_name));
+ ib_endport.port_num = port_num;
+ ad.u.ib_endport = &ib_endport;
+ return avc_has_perm(sec->sid, sid,
+ SECCLASS_INFINIBAND_ENDPORT,
+ INFINIBAND_ENDPORT__MANAGE_SUBNET, &ad);
+}
+
static int selinux_ib_alloc_security(void **ib_sec)
{
struct ib_security_struct *sec;
@@ -6313,6 +6336,8 @@ static void selinux_ib_free_security(void *ib_sec)
LSM_HOOK_INIT(tun_dev_open, selinux_tun_dev_open),
#ifdef CONFIG_SECURITY_INFINIBAND
LSM_HOOK_INIT(ib_pkey_access, selinux_ib_pkey_access),
+ LSM_HOOK_INIT(ib_endport_manage_subnet,
+ selinux_ib_endport_manage_subnet),
LSM_HOOK_INIT(ib_alloc_security, selinux_ib_alloc_security),
LSM_HOOK_INIT(ib_free_security, selinux_ib_free_security),
#endif
diff --git a/security/selinux/include/classmap.h b/security/selinux/include/classmap.h
index d42dd4d..f93b64b 100644
--- a/security/selinux/include/classmap.h
+++ b/security/selinux/include/classmap.h
@@ -167,5 +167,7 @@ struct security_class_mapping secclass_map[] = {
{ COMMON_CAP2_PERMS, NULL } },
{ "infiniband_pkey",
{ "access", NULL } },
+ { "infiniband_endport",
+ { "manage_subnet", NULL } },
{ NULL }
};
diff --git a/security/selinux/include/security.h b/security/selinux/include/security.h
index 17afb7c..8a6e5e7 100644
--- a/security/selinux/include/security.h
+++ b/security/selinux/include/security.h
@@ -178,6 +178,8 @@ int security_get_user_sids(u32 callsid, char *username,
int security_pkey_sid(u64 subnet_prefix, u16 pkey_num, u32 *out_sid);
+int security_ib_endport_sid(const char *dev_name, u8 port_num, u32 *out_sid);
+
int security_netif_sid(char *name, u32 *if_sid);
int security_node_sid(u16 domain, void *addr, u32 addrlen,
diff --git a/security/selinux/ss/services.c b/security/selinux/ss/services.c
index 085c54b..db40ce7 100644
--- a/security/selinux/ss/services.c
+++ b/security/selinux/ss/services.c
@@ -2244,6 +2244,48 @@ int security_pkey_sid(u64 subnet_prefix, u16 pkey_num, u32 *out_sid)
}
/**
+ * security_ib_endport_sid - Obtain the SID for a subnet management interface.
+ * @dev_name: device name
+ * @port: port number
+ * @out_sid: security identifier
+ */
+int security_ib_endport_sid(const char *dev_name, u8 port_num, u32 *out_sid)
+{
+ struct ocontext *c;
+ int rc = 0;
+
+ read_lock(&policy_rwlock);
+
+ c = policydb.ocontexts[OCON_IB_ENDPORT];
+ while (c) {
+ if (c->u.ib_endport.port_num == port_num &&
+ !strncmp(c->u.ib_endport.dev_name,
+ dev_name,
+ IB_DEVICE_NAME_MAX))
+ break;
+
+ c = c->next;
+ }
+
+ if (c) {
+ if (!c->sid[0]) {
+ rc = sidtab_context_to_sid(&sidtab,
+ &c->context[0],
+ &c->sid[0]);
+ if (rc)
+ goto out;
+ }
+ *out_sid = c->sid[0];
+ } else {
+ *out_sid = SECINITSID_UNLABELED;
+ }
+
+out:
+ read_unlock(&policy_rwlock);
+ return rc;
+}
+
+/**
* security_netif_sid - Obtain the SID for a network interface.
* @name: interface name
* @if_sid: interface SID
--
1.8.3.1
^ permalink raw reply related
* [PATCH v4 7/9] selinux: Implement Infiniband PKey "Access" access vector
From: Dan Jurgens @ 2016-11-08 21:06 UTC (permalink / raw)
To: chrisw, paul, sds, eparis, dledford, sean.hefty, hal.rosenstock
Cc: selinux, linux-security-module, linux-rdma, yevgenyp, liranl,
leonro, Daniel Jurgens
In-Reply-To: <1478639185-47521-1-git-send-email-danielj@mellanox.com>
From: Daniel Jurgens <danielj@mellanox.com>
Add a type and access vector for PKeys. Implement the ib_pkey_access
hook to check that the caller has permission to access the PKey on the
given subnet prefix. Add an interface to get the PKey SID. Walk the PKey
ocontexts to find an entry for the given subnet prefix and pkey.
issue: 736423
Change-Id: Ib3f7df9b5f98da77a6048855d146f9dcaceba336
Signed-off-by: Daniel Jurgens <danielj@mellanox.com>
---
v2:
- Use void* blobs for security structs. Paul Moore
- Add pkey specific data to the audit log. Paul Moore
- Don't introduce a new initial sid, use unlabeled. Stephen Smalley
v3:
- Reorder parameters to pkey_access hook. Paul Moore
Signed-off-by: Daniel Jurgens <danielj@mellanox.com>
---
include/linux/lsm_audit.h | 7 +++++++
security/lsm_audit.c | 13 ++++++++++++
security/selinux/hooks.c | 23 +++++++++++++++++++++
security/selinux/include/classmap.h | 2 ++
security/selinux/include/security.h | 2 ++
security/selinux/ss/services.c | 41 +++++++++++++++++++++++++++++++++++++
6 files changed, 88 insertions(+)
diff --git a/include/linux/lsm_audit.h b/include/linux/lsm_audit.h
index e58e577..402b770 100644
--- a/include/linux/lsm_audit.h
+++ b/include/linux/lsm_audit.h
@@ -45,6 +45,11 @@ struct lsm_ioctlop_audit {
u16 cmd;
};
+struct lsm_pkey_audit {
+ u64 subnet_prefix;
+ u16 pkey;
+};
+
/* Auxiliary data to use in generating the audit record. */
struct common_audit_data {
char type;
@@ -60,6 +65,7 @@ struct common_audit_data {
#define LSM_AUDIT_DATA_DENTRY 10
#define LSM_AUDIT_DATA_IOCTL_OP 11
#define LSM_AUDIT_DATA_FILE 12
+#define LSM_AUDIT_DATA_PKEY 13
union {
struct path path;
struct dentry *dentry;
@@ -77,6 +83,7 @@ struct common_audit_data {
char *kmod_name;
struct lsm_ioctlop_audit *op;
struct file *file;
+ struct lsm_pkey_audit *pkey;
} u;
/* this union contains LSM specific data */
union {
diff --git a/security/lsm_audit.c b/security/lsm_audit.c
index 37f04da..b18d277 100644
--- a/security/lsm_audit.c
+++ b/security/lsm_audit.c
@@ -410,6 +410,19 @@ static void dump_common_audit_data(struct audit_buffer *ab,
audit_log_format(ab, " kmod=");
audit_log_untrustedstring(ab, a->u.kmod_name);
break;
+ case LSM_AUDIT_DATA_PKEY: {
+ struct in6_addr sbn_pfx;
+
+ memset(&sbn_pfx.s6_addr, 0,
+ sizeof(sbn_pfx.s6_addr));
+
+ memcpy(&sbn_pfx.s6_addr, &a->u.pkey->subnet_prefix,
+ sizeof(a->u.pkey->subnet_prefix));
+
+ audit_log_format(ab, " pkey=0x%x subnet_prefix=%pI6c",
+ a->u.pkey->pkey, &sbn_pfx);
+ break;
+ }
} /* switch (a->type) */
}
diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c
index f48759d..20fb292 100644
--- a/security/selinux/hooks.c
+++ b/security/selinux/hooks.c
@@ -6086,6 +6086,28 @@ static int selinux_key_getsecurity(struct key *key, char **_buffer)
#endif
#ifdef CONFIG_SECURITY_INFINIBAND
+static int selinux_ib_pkey_access(void *ib_sec, u64 subnet_prefix, u16 pkey_val)
+{
+ struct common_audit_data ad;
+ int err;
+ u32 sid = 0;
+ struct ib_security_struct *sec = ib_sec;
+ struct lsm_pkey_audit pkey;
+
+ err = security_pkey_sid(subnet_prefix, pkey_val, &sid);
+
+ if (err)
+ return err;
+
+ ad.type = LSM_AUDIT_DATA_PKEY;
+ pkey.subnet_prefix = subnet_prefix;
+ pkey.pkey = pkey_val;
+ ad.u.pkey = &pkey;
+ return avc_has_perm(sec->sid, sid,
+ SECCLASS_INFINIBAND_PKEY,
+ INFINIBAND_PKEY__ACCESS, &ad);
+}
+
static int selinux_ib_alloc_security(void **ib_sec)
{
struct ib_security_struct *sec;
@@ -6290,6 +6312,7 @@ static void selinux_ib_free_security(void *ib_sec)
LSM_HOOK_INIT(tun_dev_attach, selinux_tun_dev_attach),
LSM_HOOK_INIT(tun_dev_open, selinux_tun_dev_open),
#ifdef CONFIG_SECURITY_INFINIBAND
+ LSM_HOOK_INIT(ib_pkey_access, selinux_ib_pkey_access),
LSM_HOOK_INIT(ib_alloc_security, selinux_ib_alloc_security),
LSM_HOOK_INIT(ib_free_security, selinux_ib_free_security),
#endif
diff --git a/security/selinux/include/classmap.h b/security/selinux/include/classmap.h
index 1f1f4b2..d42dd4d 100644
--- a/security/selinux/include/classmap.h
+++ b/security/selinux/include/classmap.h
@@ -165,5 +165,7 @@ struct security_class_mapping secclass_map[] = {
{ COMMON_CAP_PERMS, NULL } },
{ "cap2_userns",
{ COMMON_CAP2_PERMS, NULL } },
+ { "infiniband_pkey",
+ { "access", NULL } },
{ NULL }
};
diff --git a/security/selinux/include/security.h b/security/selinux/include/security.h
index 6bb9b0a..17afb7c 100644
--- a/security/selinux/include/security.h
+++ b/security/selinux/include/security.h
@@ -176,6 +176,8 @@ int security_get_user_sids(u32 callsid, char *username,
int security_port_sid(u8 protocol, u16 port, u32 *out_sid);
+int security_pkey_sid(u64 subnet_prefix, u16 pkey_num, u32 *out_sid);
+
int security_netif_sid(char *name, u32 *if_sid);
int security_node_sid(u16 domain, void *addr, u32 addrlen,
diff --git a/security/selinux/ss/services.c b/security/selinux/ss/services.c
index 082b20c..085c54b 100644
--- a/security/selinux/ss/services.c
+++ b/security/selinux/ss/services.c
@@ -2203,6 +2203,47 @@ int security_port_sid(u8 protocol, u16 port, u32 *out_sid)
}
/**
+ * security_pkey_sid - Obtain the SID for a pkey.
+ * @subnet_prefix: Subnet Prefix
+ * @pkey_num: pkey number
+ * @out_sid: security identifier
+ */
+int security_pkey_sid(u64 subnet_prefix, u16 pkey_num, u32 *out_sid)
+{
+ struct ocontext *c;
+ int rc = 0;
+
+ read_lock(&policy_rwlock);
+
+ c = policydb.ocontexts[OCON_PKEY];
+ while (c) {
+ if (c->u.pkey.low_pkey <= pkey_num &&
+ c->u.pkey.high_pkey >= pkey_num &&
+ c->u.pkey.subnet_prefix == subnet_prefix)
+ break;
+
+ c = c->next;
+ }
+
+ if (c) {
+ if (!c->sid[0]) {
+ rc = sidtab_context_to_sid(&sidtab,
+ &c->context[0],
+ &c->sid[0]);
+ if (rc)
+ goto out;
+ }
+ *out_sid = c->sid[0];
+ } else {
+ *out_sid = SECINITSID_UNLABELED;
+ }
+
+out:
+ read_unlock(&policy_rwlock);
+ return rc;
+}
+
+/**
* security_netif_sid - Obtain the SID for a network interface.
* @name: interface name
* @if_sid: interface SID
--
1.8.3.1
^ permalink raw reply related
* [PATCH v4 9/9] selinux: Add a cache for quicker retreival of PKey SIDs
From: Dan Jurgens @ 2016-11-08 21:06 UTC (permalink / raw)
To: chrisw, paul, sds, eparis, dledford, sean.hefty, hal.rosenstock
Cc: selinux, linux-security-module, linux-rdma, yevgenyp, liranl,
leonro, Daniel Jurgens
In-Reply-To: <1478639185-47521-1-git-send-email-danielj@mellanox.com>
From: Daniel Jurgens <danielj@mellanox.com>
It is likely that the SID for the same PKey will be requested many
times. To reduce the time to modify QPs and process MADs use a cache to
store PKey SIDs.
This code is heavily based on the "netif" and "netport" concept
originally developed by James Morris <jmorris@redhat.com> and Paul Moore
<paul@paul-moore.com> (see security/selinux/netif.c and
security/selinux/netport.c for more information)
issue: 736423
Change-Id: I176c3079d5d84d06839b4f750100ac47a6081e94
Signed-off-by: Daniel Jurgens <danielj@mellanox.com>
---
v2:
- Renamed the files to ibpkey. Paul Moore
- Fixed a braket indentation mismatch in sel_pkey_find. Yuval Shaia
- Change spin_lock_bh to spin_lock_irqsave to resolve HARDIRQ lockdep
warning. Dan Jurgens
Signed-off-by: Daniel Jurgens <danielj@mellanox.com>
---
security/selinux/Makefile | 2 +-
security/selinux/hooks.c | 4 +-
security/selinux/ibpkey.c | 245 ++++++++++++++++++++++++++++++++++++++
security/selinux/include/ibpkey.h | 31 +++++
security/selinux/include/objsec.h | 6 +
5 files changed, 286 insertions(+), 2 deletions(-)
create mode 100644 security/selinux/ibpkey.c
create mode 100644 security/selinux/include/ibpkey.h
diff --git a/security/selinux/Makefile b/security/selinux/Makefile
index 3411c33..ff5895e 100644
--- a/security/selinux/Makefile
+++ b/security/selinux/Makefile
@@ -5,7 +5,7 @@
obj-$(CONFIG_SECURITY_SELINUX) := selinux.o
selinux-y := avc.o hooks.o selinuxfs.o netlink.o nlmsgtab.o netif.o \
- netnode.o netport.o exports.o \
+ netnode.o netport.o ibpkey.o exports.o \
ss/ebitmap.o ss/hashtab.o ss/symtab.o ss/sidtab.o ss/avtab.o \
ss/policydb.o ss/services.o ss/conditional.o ss/mls.o ss/status.o
diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c
index ea3f6d0..619d2ea 100644
--- a/security/selinux/hooks.c
+++ b/security/selinux/hooks.c
@@ -90,6 +90,7 @@
#include "netif.h"
#include "netnode.h"
#include "netport.h"
+#include "ibpkey.h"
#include "xfrm.h"
#include "netlabel.h"
#include "audit.h"
@@ -174,6 +175,7 @@ static int selinux_netcache_avc_callback(u32 event)
static int selinux_lsm_notifier_avc_callback(u32 event)
{
if (event == AVC_CALLBACK_RESET) {
+ sel_pkey_flush();
call_lsm_notifier(LSM_POLICY_CHANGE, NULL);
}
return 0;
@@ -6094,7 +6096,7 @@ static int selinux_ib_pkey_access(void *ib_sec, u64 subnet_prefix, u16 pkey_val)
struct ib_security_struct *sec = ib_sec;
struct lsm_pkey_audit pkey;
- err = security_pkey_sid(subnet_prefix, pkey_val, &sid);
+ err = sel_pkey_sid(subnet_prefix, pkey_val, &sid);
if (err)
return err;
diff --git a/security/selinux/ibpkey.c b/security/selinux/ibpkey.c
new file mode 100644
index 0000000..6e52c54
--- /dev/null
+++ b/security/selinux/ibpkey.c
@@ -0,0 +1,245 @@
+/*
+ * Pkey table
+ *
+ * SELinux must keep a mapping of Infinband PKEYs to labels/SIDs. This
+ * mapping is maintained as part of the normal policy but a fast cache is
+ * needed to reduce the lookup overhead.
+ *
+ * This code is heavily based on the "netif" and "netport" concept originally
+ * developed by
+ * James Morris <jmorris@redhat.com> and
+ * Paul Moore <paul@paul-moore.com>
+ * (see security/selinux/netif.c and security/selinux/netport.c for more
+ * information)
+ *
+ */
+
+/*
+ * (c) Mellanox Technologies, 2016
+ *
+ * This program is free software: you can redistribute it and/or modify
+ * it under the terms of version 2 of the GNU General Public License as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ */
+
+#include <linux/types.h>
+#include <linux/rcupdate.h>
+#include <linux/list.h>
+#include <linux/spinlock.h>
+
+#include "ibpkey.h"
+#include "objsec.h"
+
+#define SEL_PKEY_HASH_SIZE 256
+#define SEL_PKEY_HASH_BKT_LIMIT 16
+
+struct sel_pkey_bkt {
+ int size;
+ struct list_head list;
+};
+
+struct sel_pkey {
+ struct pkey_security_struct psec;
+ struct list_head list;
+ struct rcu_head rcu;
+};
+
+static LIST_HEAD(sel_pkey_list);
+static DEFINE_SPINLOCK(sel_pkey_lock);
+static struct sel_pkey_bkt sel_pkey_hash[SEL_PKEY_HASH_SIZE];
+
+/**
+ * sel_pkey_hashfn - Hashing function for the pkey table
+ * @pkey: pkey number
+ *
+ * Description:
+ * This is the hashing function for the pkey table, it returns the bucket
+ * number for the given pkey.
+ *
+ */
+static unsigned int sel_pkey_hashfn(u16 pkey)
+{
+ return (pkey & (SEL_PKEY_HASH_SIZE - 1));
+}
+
+/**
+ * sel_pkey_find - Search for a pkey record
+ * @subnet_prefix: subnet_prefix
+ * @pkey_num: pkey_num
+ *
+ * Description:
+ * Search the pkey table and return the matching record. If an entry
+ * can not be found in the table return NULL.
+ *
+ */
+static struct sel_pkey *sel_pkey_find(u64 subnet_prefix, u16 pkey_num)
+{
+ unsigned int idx;
+ struct sel_pkey *pkey;
+
+ idx = sel_pkey_hashfn(pkey_num);
+ list_for_each_entry_rcu(pkey, &sel_pkey_hash[idx].list, list) {
+ if (pkey->psec.pkey == pkey_num &&
+ pkey->psec.subnet_prefix == subnet_prefix)
+ return pkey;
+ }
+
+ return NULL;
+}
+
+/**
+ * sel_pkey_insert - Insert a new pkey into the table
+ * @pkey: the new pkey record
+ *
+ * Description:
+ * Add a new pkey record to the hash table.
+ *
+ */
+static void sel_pkey_insert(struct sel_pkey *pkey)
+{
+ unsigned int idx;
+
+ /* we need to impose a limit on the growth of the hash table so check
+ * this bucket to make sure it is within the specified bounds
+ */
+ idx = sel_pkey_hashfn(pkey->psec.pkey);
+ list_add_rcu(&pkey->list, &sel_pkey_hash[idx].list);
+ if (sel_pkey_hash[idx].size == SEL_PKEY_HASH_BKT_LIMIT) {
+ struct sel_pkey *tail;
+
+ tail = list_entry(
+ rcu_dereference_protected(
+ sel_pkey_hash[idx].list.prev,
+ lockdep_is_held(&sel_pkey_lock)),
+ struct sel_pkey, list);
+ list_del_rcu(&tail->list);
+ kfree_rcu(tail, rcu);
+ } else {
+ sel_pkey_hash[idx].size++;
+ }
+}
+
+/**
+ * sel_pkey_sid_slow - Lookup the SID of a pkey using the policy
+ * @subnet_prefix: subnet prefix
+ * @pkey_num: pkey number
+ * @sid: pkey SID
+ *
+ * Description:
+ * This function determines the SID of a pkey by querying the security
+ * policy. The result is added to the pkey table to speedup future
+ * queries. Returns zero on success, negative values on failure.
+ *
+ */
+static int sel_pkey_sid_slow(u64 subnet_prefix, u16 pkey_num, u32 *sid)
+{
+ int ret = -ENOMEM;
+ struct sel_pkey *pkey;
+ struct sel_pkey *new = NULL;
+ unsigned long flags;
+
+ spin_lock_irqsave(&sel_pkey_lock, flags);
+ pkey = sel_pkey_find(subnet_prefix, pkey_num);
+ if (pkey) {
+ *sid = pkey->psec.sid;
+ spin_unlock_irqrestore(&sel_pkey_lock, flags);
+ return 0;
+ }
+
+ ret = security_pkey_sid(subnet_prefix, pkey_num, sid);
+ if (ret != 0)
+ goto out;
+
+ new = kzalloc(sizeof(*new), GFP_ATOMIC);
+ if (!new)
+ goto out;
+
+ new->psec.subnet_prefix = subnet_prefix;
+ new->psec.pkey = pkey_num;
+ new->psec.sid = *sid;
+ sel_pkey_insert(new);
+
+out:
+ spin_unlock_irqrestore(&sel_pkey_lock, flags);
+ if (unlikely(ret))
+ kfree(new);
+
+ return ret;
+}
+
+/**
+ * sel_pkey_sid - Lookup the SID of a PKEY
+ * @subnet_prefix: subnet_prefix
+ * @pkey_num: pkey number
+ * @sid: pkey SID
+ *
+ * Description:
+ * This function determines the SID of a PKEY using the fastest method
+ * possible. First the pkey table is queried, but if an entry can't be found
+ * then the policy is queried and the result is added to the table to speedup
+ * future queries. Returns zero on success, negative values on failure.
+ *
+ */
+int sel_pkey_sid(u64 subnet_prefix, u16 pkey_num, u32 *sid)
+{
+ struct sel_pkey *pkey;
+
+ rcu_read_lock();
+ pkey = sel_pkey_find(subnet_prefix, pkey_num);
+ if (pkey) {
+ *sid = pkey->psec.sid;
+ rcu_read_unlock();
+ return 0;
+ }
+ rcu_read_unlock();
+
+ return sel_pkey_sid_slow(subnet_prefix, pkey_num, sid);
+}
+
+/**
+ * sel_pkey_flush - Flush the entire pkey table
+ *
+ * Description:
+ * Remove all entries from the pkey table
+ *
+ */
+void sel_pkey_flush(void)
+{
+ unsigned int idx;
+ struct sel_pkey *pkey, *pkey_tmp;
+ unsigned long flags;
+
+ spin_lock_irqsave(&sel_pkey_lock, flags);
+ for (idx = 0; idx < SEL_PKEY_HASH_SIZE; idx++) {
+ list_for_each_entry_safe(pkey, pkey_tmp,
+ &sel_pkey_hash[idx].list, list) {
+ list_del_rcu(&pkey->list);
+ kfree_rcu(pkey, rcu);
+ }
+ sel_pkey_hash[idx].size = 0;
+ }
+ spin_unlock_irqrestore(&sel_pkey_lock, flags);
+}
+
+static __init int sel_pkey_init(void)
+{
+ int iter;
+
+ if (!selinux_enabled)
+ return 0;
+
+ for (iter = 0; iter < SEL_PKEY_HASH_SIZE; iter++) {
+ INIT_LIST_HEAD(&sel_pkey_hash[iter].list);
+ sel_pkey_hash[iter].size = 0;
+ }
+
+ return 0;
+}
+
+subsys_initcall(sel_pkey_init);
diff --git a/security/selinux/include/ibpkey.h b/security/selinux/include/ibpkey.h
new file mode 100644
index 0000000..387885a
--- /dev/null
+++ b/security/selinux/include/ibpkey.h
@@ -0,0 +1,31 @@
+/*
+ * pkey table
+ *
+ * SELinux must keep a mapping of pkeys to labels/SIDs. This
+ * mapping is maintained as part of the normal policy but a fast cache is
+ * needed to reduce the lookup overhead.
+ *
+ */
+
+/*
+ * (c) Mellanox Technologies, 2016
+ *
+ * This program is free software: you can redistribute it and/or modify
+ * it under the terms of version 2 of the GNU General Public License as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ */
+
+#ifndef _SELINUX_IB_PKEY_H
+#define _SELINUX_IB_PKEY_H
+
+void sel_pkey_flush(void);
+
+int sel_pkey_sid(u64 subnet_prefix, u16 pkey, u32 *sid);
+
+#endif
diff --git a/security/selinux/include/objsec.h b/security/selinux/include/objsec.h
index 8e7db43..4139f28 100644
--- a/security/selinux/include/objsec.h
+++ b/security/selinux/include/objsec.h
@@ -133,6 +133,12 @@ struct ib_security_struct {
u32 sid; /* SID of the queue pair or MAD agent */
};
+struct pkey_security_struct {
+ u64 subnet_prefix; /* Port subnet prefix */
+ u16 pkey; /* PKey number */
+ u32 sid; /* SID of pkey */
+};
+
extern unsigned int selinux_checkreqprot;
#endif /* _SELINUX_OBJSEC_H_ */
--
1.8.3.1
^ permalink raw reply related
* Re: [PATCH kernel v4 7/7] virtio-balloon: tell host vm's unused page info
From: Michael S. Tsirkin @ 2016-11-08 21:07 UTC (permalink / raw)
To: Dave Hansen
Cc: Li, Liang Z, pbonzini@redhat.com, amit.shah@redhat.com,
quintela@redhat.com, dgilbert@redhat.com, qemu-devel@nongnu.org,
kvm@vger.kernel.org, virtio-dev@lists.oasis-open.org,
linux-mm@kvack.org, linux-kernel@vger.kernel.org,
virtualization@lists.linux-foundation.org,
mgorman@techsingularity.net, cornelia.huck@de.ibm.com
In-Reply-To: <281acd8d-fd94-6318-35e5-9eb130303dc6@intel.com>
On Mon, Nov 07, 2016 at 09:23:38AM -0800, Dave Hansen wrote:
> On 11/06/2016 07:37 PM, Li, Liang Z wrote:
> >> Let's say we do a 32k bitmap that can hold ~1M pages. That's 4GB of RAM.
> >> On a 1TB system, that's 256 passes through the top-level loop.
> >> The bottom-level lists have tens of thousands of pages in them, even on my
> >> laptop. Only 1/256 of these pages will get consumed in a given pass.
> >>
> > Your description is not exactly.
> > A 32k bitmap is used only when there is few free memory left in the system and when
> > the extend_page_bitmap() failed to allocate more memory for the bitmap. Or dozens of
> > 32k split bitmap will be used, this version limit the bitmap count to 32, it means we can use
> > at most 32*32 kB for the bitmap, which can cover 128GB for RAM. We can increase the bitmap
> > count limit to a larger value if 32 is not big enough.
>
> OK, so it tries to allocate a large bitmap. But, if it fails, it will
> try to work with a smaller bitmap. Correct?
>
> So, what's the _worst_ case? It sounds like it is even worse than I was
> positing.
>
> >> That's an awfully inefficient way of doing it. This patch essentially changed
> >> the data structure without changing the algorithm to populate it.
> >>
> >> Please change the *algorithm* to use the new data structure efficiently.
> >> Such a change would only do a single pass through each freelist, and would
> >> choose whether to use the extent-based (pfn -> range) or bitmap-based
> >> approach based on the contents of the free lists.
> >
> > Save the free page info to a raw bitmap first and then process the raw bitmap to
> > get the proper ' extent-based ' and 'bitmap-based' is the most efficient way I can
> > come up with to save the virtio data transmission. Do you have some better idea?
>
> That's kinda my point. This patch *does* processing to try to pack the
> bitmaps full of pages from the various pfn ranges. It's a form of
> processing that gets *REALLY*, *REALLY* bad in some (admittedly obscure)
> cases.
>
> Let's not pretend that making an essentially unlimited number of passes
> over the free lists is not processing.
>
> 1. Allocate as large of a bitmap as you can. (what you already do)
> 2. Iterate from the largest freelist order. Store those pages in the
> bitmap.
> 3. If you can no longer fit pages in the bitmap, return the list that
> you have.
> 4. Make an approximation about where the bitmap does not make any more,
> and fall back to listing individual PFNs. This would make sens, for
> instance in a large zone with very few free order-0 pages left.
In practice, a single PFN using the bitmap format
only takes up twice the size: I think it's 128 instead of 64 bit
per entry.
So it's not a a given that point 4 is worth it at any point,
just packing multiple bitmaps might be good enough.
>
> > It seems the benefit we get for this feature is not as big as that in fast balloon inflating/deflating.
> >>
> >> You should not be using get_max_pfn(). Any patch set that continues to use
> >> it is not likely to be using a proper algorithm.
> >
> > Do you have any suggestion about how to avoid it?
>
> Yes: get the pfns from the page free lists alone. Don't derive them
> from the pfn limits of the system or zones.
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
^ permalink raw reply
* Re: [PATCH kernel v4 7/7] virtio-balloon: tell host vm's unused page info
From: Michael S. Tsirkin @ 2016-11-08 21:07 UTC (permalink / raw)
To: Dave Hansen
Cc: Li, Liang Z, pbonzini@redhat.com, amit.shah@redhat.com,
quintela@redhat.com, dgilbert@redhat.com, qemu-devel@nongnu.org,
kvm@vger.kernel.org, virtio-dev@lists.oasis-open.org,
linux-mm@kvack.org, linux-kernel@vger.kernel.org,
virtualization@lists.linux-foundation.org,
mgorman@techsingularity.net, cornelia.huck@de.ibm.com
In-Reply-To: <281acd8d-fd94-6318-35e5-9eb130303dc6@intel.com>
On Mon, Nov 07, 2016 at 09:23:38AM -0800, Dave Hansen wrote:
> On 11/06/2016 07:37 PM, Li, Liang Z wrote:
> >> Let's say we do a 32k bitmap that can hold ~1M pages. That's 4GB of RAM.
> >> On a 1TB system, that's 256 passes through the top-level loop.
> >> The bottom-level lists have tens of thousands of pages in them, even on my
> >> laptop. Only 1/256 of these pages will get consumed in a given pass.
> >>
> > Your description is not exactly.
> > A 32k bitmap is used only when there is few free memory left in the system and when
> > the extend_page_bitmap() failed to allocate more memory for the bitmap. Or dozens of
> > 32k split bitmap will be used, this version limit the bitmap count to 32, it means we can use
> > at most 32*32 kB for the bitmap, which can cover 128GB for RAM. We can increase the bitmap
> > count limit to a larger value if 32 is not big enough.
>
> OK, so it tries to allocate a large bitmap. But, if it fails, it will
> try to work with a smaller bitmap. Correct?
>
> So, what's the _worst_ case? It sounds like it is even worse than I was
> positing.
>
> >> That's an awfully inefficient way of doing it. This patch essentially changed
> >> the data structure without changing the algorithm to populate it.
> >>
> >> Please change the *algorithm* to use the new data structure efficiently.
> >> Such a change would only do a single pass through each freelist, and would
> >> choose whether to use the extent-based (pfn -> range) or bitmap-based
> >> approach based on the contents of the free lists.
> >
> > Save the free page info to a raw bitmap first and then process the raw bitmap to
> > get the proper ' extent-based ' and 'bitmap-based' is the most efficient way I can
> > come up with to save the virtio data transmission. Do you have some better idea?
>
> That's kinda my point. This patch *does* processing to try to pack the
> bitmaps full of pages from the various pfn ranges. It's a form of
> processing that gets *REALLY*, *REALLY* bad in some (admittedly obscure)
> cases.
>
> Let's not pretend that making an essentially unlimited number of passes
> over the free lists is not processing.
>
> 1. Allocate as large of a bitmap as you can. (what you already do)
> 2. Iterate from the largest freelist order. Store those pages in the
> bitmap.
> 3. If you can no longer fit pages in the bitmap, return the list that
> you have.
> 4. Make an approximation about where the bitmap does not make any more,
> and fall back to listing individual PFNs. This would make sens, for
> instance in a large zone with very few free order-0 pages left.
In practice, a single PFN using the bitmap format
only takes up twice the size: I think it's 128 instead of 64 bit
per entry.
So it's not a a given that point 4 is worth it at any point,
just packing multiple bitmaps might be good enough.
>
> > It seems the benefit we get for this feature is not as big as that in fast balloon inflating/deflating.
> >>
> >> You should not be using get_max_pfn(). Any patch set that continues to use
> >> it is not likely to be using a proper algorithm.
> >
> > Do you have any suggestion about how to avoid it?
>
> Yes: get the pfns from the page free lists alone. Don't derive them
> from the pfn limits of the system or zones.
^ permalink raw reply
* [PATCH v4 4/9] IB/core: Enforce security on management datagrams
From: Dan Jurgens @ 2016-11-08 21:06 UTC (permalink / raw)
To: chrisw, paul, sds, eparis, dledford, sean.hefty, hal.rosenstock
Cc: selinux, linux-security-module, linux-rdma, yevgenyp, liranl,
leonro, Daniel Jurgens
In-Reply-To: <1478639185-47521-1-git-send-email-danielj@mellanox.com>
From: Daniel Jurgens <danielj@mellanox.com>
Allocate and free a security context when creating and destroying a MAD
agent. This context is used for controlling access to PKeys and sending
and receiving SMPs.
When sending or receiving a MAD check that the agent has permission to
access the PKey for the Subnet Prefix of the port.
During MAD and snoop agent registration for SMI QPs check that the
calling process has permission to access the manage the subnet and
register a callback with the LSM to be notified of policy changes. When
notificaiton of a policy change occurs recheck permission and set a flag
indicating sending and receiving SMPs is allowed.
When sending and receiving MADs check that the agent has access to the
SMI if it's on an SMI QP. Because security policy can change it's
possible permission was allowed when creating the agent, but no longer
is.
issue: 736423
Change-Id: I17343224da7ffffe7ee0cc67dadd44c473e453d0
Signed-off-by: Daniel Jurgens <danielj@mellanox.com>
---
v2:
- Squashed LSM hook additions. Paul Moore
- Changed security blobs to void*. Paul Moore
- Shorten end_port to port. Paul Moore
- Change "smp" to "manage_subnet". Paul Moore
- Use the LSM policy change notification and a flag to track permission
instead of calling the LSM hook for every SMP. Dan Jurgens
- Squashed PKey and SMP enforcement into the same patch and moved the
logic into security.c. Dan Jurgens
v3:
- ib_port -> ib_endport. Paul Moore
- Use notifier chains for LSM notification. Paul Moore
- Reorder LSM hook parameters to put sec first. Paul Moore
Signed-off-by: Daniel Jurgens <danielj@mellanox.com>
---
drivers/infiniband/core/core_priv.h | 35 ++++++++++++++
drivers/infiniband/core/mad.c | 52 +++++++++++++++++----
drivers/infiniband/core/security.c | 92 +++++++++++++++++++++++++++++++++++++
include/linux/lsm_hooks.h | 8 ++++
include/linux/security.h | 6 +++
include/rdma/ib_mad.h | 4 ++
security/security.c | 8 ++++
7 files changed, 197 insertions(+), 8 deletions(-)
diff --git a/drivers/infiniband/core/core_priv.h b/drivers/infiniband/core/core_priv.h
index 68e3de0..2c35162 100644
--- a/drivers/infiniband/core/core_priv.h
+++ b/drivers/infiniband/core/core_priv.h
@@ -37,6 +37,8 @@
#include <linux/spinlock.h>
#include <rdma/ib_verbs.h>
+#include <rdma/ib_mad.h>
+#include "mad_priv.h"
struct pkey_index_qp_list {
struct list_head pkey_index_list;
@@ -166,6 +168,11 @@ int ib_get_cached_subnet_prefix(struct ib_device *device,
u64 *sn_pfx);
#ifdef CONFIG_SECURITY_INFINIBAND
+int ib_security_pkey_access(struct ib_device *dev,
+ u8 port_num,
+ u16 pkey_index,
+ void *sec);
+
void ib_security_destroy_port_pkey_list(struct ib_device *device);
void ib_security_cache_change(struct ib_device *device,
@@ -183,7 +190,19 @@ int ib_security_modify_qp(struct ib_qp *qp,
void ib_destroy_qp_security_end(struct ib_qp_security *sec);
int ib_open_shared_qp_security(struct ib_qp *qp, struct ib_device *dev);
void ib_close_shared_qp_security(struct ib_qp_security *sec);
+int ib_mad_agent_security_setup(struct ib_mad_agent *agent,
+ enum ib_qp_type qp_type);
+void ib_mad_agent_security_cleanup(struct ib_mad_agent *agent);
+int ib_mad_enforce_security(struct ib_mad_agent_private *map, u16 pkey_index);
#else
+static inline int ib_security_pkey_access(struct ib_device *dev,
+ u8 port_num,
+ u16 pkey_index,
+ void *sec)
+{
+ return 0;
+}
+
static inline void ib_security_destroy_port_pkey_list(struct ib_device *device)
{
}
@@ -232,5 +251,21 @@ static inline int ib_open_shared_qp_security(struct ib_qp *qp,
static inline void ib_close_shared_qp_security(struct ib_qp_security *sec)
{
}
+
+static inline int ib_mad_agent_security_setup(struct ib_mad_agent *agent,
+ enum ib_qp_type qp_type)
+{
+ return 0;
+}
+
+static inline void ib_mad_agent_security_cleanup(struct ib_mad_agent *agent)
+{
+}
+
+static inline int ib_mad_enforce_security(struct ib_mad_agent_private *map,
+ u16 pkey_index)
+{
+ return 0;
+}
#endif
#endif /* _CORE_PRIV_H */
diff --git a/drivers/infiniband/core/mad.c b/drivers/infiniband/core/mad.c
index 40cbd6b..b6041ec 100644
--- a/drivers/infiniband/core/mad.c
+++ b/drivers/infiniband/core/mad.c
@@ -40,9 +40,11 @@
#include <linux/dma-mapping.h>
#include <linux/slab.h>
#include <linux/module.h>
+#include <linux/security.h>
#include <rdma/ib_cache.h>
#include "mad_priv.h"
+#include "core_priv.h"
#include "mad_rmpp.h"
#include "smi.h"
#include "opa_smi.h"
@@ -367,6 +369,12 @@ struct ib_mad_agent *ib_register_mad_agent(struct ib_device *device,
atomic_set(&mad_agent_priv->refcount, 1);
init_completion(&mad_agent_priv->comp);
+ ret2 = ib_mad_agent_security_setup(&mad_agent_priv->agent, qp_type);
+ if (ret2) {
+ ret = ERR_PTR(ret2);
+ goto error4;
+ }
+
spin_lock_irqsave(&port_priv->reg_lock, flags);
mad_agent_priv->agent.hi_tid = ++ib_mad_client_id;
@@ -384,7 +392,7 @@ struct ib_mad_agent *ib_register_mad_agent(struct ib_device *device,
if (method) {
if (method_in_use(&method,
mad_reg_req))
- goto error4;
+ goto error5;
}
}
ret2 = add_nonoui_reg_req(mad_reg_req, mad_agent_priv,
@@ -400,14 +408,14 @@ struct ib_mad_agent *ib_register_mad_agent(struct ib_device *device,
if (is_vendor_method_in_use(
vendor_class,
mad_reg_req))
- goto error4;
+ goto error5;
}
}
ret2 = add_oui_reg_req(mad_reg_req, mad_agent_priv);
}
if (ret2) {
ret = ERR_PTR(ret2);
- goto error4;
+ goto error5;
}
}
@@ -416,9 +424,10 @@ struct ib_mad_agent *ib_register_mad_agent(struct ib_device *device,
spin_unlock_irqrestore(&port_priv->reg_lock, flags);
return &mad_agent_priv->agent;
-
-error4:
+error5:
spin_unlock_irqrestore(&port_priv->reg_lock, flags);
+ ib_mad_agent_security_cleanup(&mad_agent_priv->agent);
+error4:
kfree(reg_req);
error3:
kfree(mad_agent_priv);
@@ -489,6 +498,7 @@ struct ib_mad_agent *ib_register_mad_snoop(struct ib_device *device,
struct ib_mad_agent *ret;
struct ib_mad_snoop_private *mad_snoop_priv;
int qpn;
+ int err;
/* Validate parameters */
if ((is_snooping_sends(mad_snoop_flags) && !snoop_handler) ||
@@ -523,17 +533,25 @@ struct ib_mad_agent *ib_register_mad_snoop(struct ib_device *device,
mad_snoop_priv->agent.port_num = port_num;
mad_snoop_priv->mad_snoop_flags = mad_snoop_flags;
init_completion(&mad_snoop_priv->comp);
+
+ err = ib_mad_agent_security_setup(&mad_snoop_priv->agent, qp_type);
+ if (err) {
+ ret = ERR_PTR(err);
+ goto error2;
+ }
+
mad_snoop_priv->snoop_index = register_snoop_agent(
&port_priv->qp_info[qpn],
mad_snoop_priv);
if (mad_snoop_priv->snoop_index < 0) {
ret = ERR_PTR(mad_snoop_priv->snoop_index);
- goto error2;
+ goto error3;
}
atomic_set(&mad_snoop_priv->refcount, 1);
return &mad_snoop_priv->agent;
-
+error3:
+ ib_mad_agent_security_cleanup(&mad_snoop_priv->agent);
error2:
kfree(mad_snoop_priv);
error1:
@@ -579,6 +597,8 @@ static void unregister_mad_agent(struct ib_mad_agent_private *mad_agent_priv)
deref_mad_agent(mad_agent_priv);
wait_for_completion(&mad_agent_priv->comp);
+ ib_mad_agent_security_cleanup(&mad_agent_priv->agent);
+
kfree(mad_agent_priv->reg_req);
kfree(mad_agent_priv);
}
@@ -597,6 +617,8 @@ static void unregister_mad_snoop(struct ib_mad_snoop_private *mad_snoop_priv)
deref_snoop_agent(mad_snoop_priv);
wait_for_completion(&mad_snoop_priv->comp);
+ ib_mad_agent_security_cleanup(&mad_snoop_priv->agent);
+
kfree(mad_snoop_priv);
}
@@ -1219,12 +1241,16 @@ int ib_post_send_mad(struct ib_mad_send_buf *send_buf,
/* Walk list of send WRs and post each on send list */
for (; send_buf; send_buf = next_send_buf) {
-
mad_send_wr = container_of(send_buf,
struct ib_mad_send_wr_private,
send_buf);
mad_agent_priv = mad_send_wr->mad_agent_priv;
+ ret = ib_mad_enforce_security(mad_agent_priv,
+ mad_send_wr->send_wr.pkey_index);
+ if (ret)
+ goto error;
+
if (!send_buf->mad_agent->send_handler ||
(send_buf->timeout_ms &&
!send_buf->mad_agent->recv_handler)) {
@@ -1958,6 +1984,14 @@ static void ib_mad_complete_recv(struct ib_mad_agent_private *mad_agent_priv,
struct ib_mad_send_wr_private *mad_send_wr;
struct ib_mad_send_wc mad_send_wc;
unsigned long flags;
+ int ret;
+
+ ret = ib_mad_enforce_security(mad_agent_priv,
+ mad_recv_wc->wc->pkey_index);
+ if (ret) {
+ ib_free_recv_mad(mad_recv_wc);
+ deref_mad_agent(mad_agent_priv);
+ }
INIT_LIST_HEAD(&mad_recv_wc->rmpp_list);
list_add(&mad_recv_wc->recv_buf.list, &mad_recv_wc->rmpp_list);
@@ -2015,6 +2049,8 @@ static void ib_mad_complete_recv(struct ib_mad_agent_private *mad_agent_priv,
mad_recv_wc);
deref_mad_agent(mad_agent_priv);
}
+
+ return;
}
static enum smi_action handle_ib_smi(const struct ib_mad_port_private *port_priv,
diff --git a/drivers/infiniband/core/security.c b/drivers/infiniband/core/security.c
index 45800dd..bc09af5 100644
--- a/drivers/infiniband/core/security.c
+++ b/drivers/infiniband/core/security.c
@@ -39,6 +39,7 @@
#include <rdma/ib_verbs.h>
#include <rdma/ib_cache.h>
#include "core_priv.h"
+#include "mad_priv.h"
static struct pkey_index_qp_list *get_pkey_idx_qp_list(struct ib_port_pkey *pp)
{
@@ -614,4 +615,95 @@ int ib_security_modify_qp(struct ib_qp *qp,
}
EXPORT_SYMBOL(ib_security_modify_qp);
+int ib_security_pkey_access(struct ib_device *dev,
+ u8 port_num,
+ u16 pkey_index,
+ void *sec)
+{
+ u64 subnet_prefix;
+ u16 pkey;
+ int ret;
+
+ ret = ib_get_cached_pkey(dev, port_num, pkey_index, &pkey);
+ if (ret)
+ return ret;
+
+ ret = ib_get_cached_subnet_prefix(dev, port_num, &subnet_prefix);
+
+ if (ret)
+ return ret;
+
+ return security_ib_pkey_access(sec, subnet_prefix, pkey);
+}
+EXPORT_SYMBOL(ib_security_pkey_access);
+
+static int ib_mad_agent_security_change(struct notifier_block *nb,
+ unsigned long event,
+ void *data)
+{
+ struct ib_mad_agent *ag = container_of(nb, struct ib_mad_agent, lsm_nb);
+
+ if (event != LSM_POLICY_CHANGE)
+ return NOTIFY_DONE;
+
+ ag->smp_allowed = !security_ib_endport_manage_subnet(ag->security,
+ ag->device->name,
+ ag->port_num);
+
+ return NOTIFY_OK;
+}
+
+int ib_mad_agent_security_setup(struct ib_mad_agent *agent,
+ enum ib_qp_type qp_type)
+{
+ int ret;
+
+ ret = security_ib_alloc_security(&agent->security);
+ if (ret)
+ return ret;
+
+ if (qp_type != IB_QPT_SMI)
+ return 0;
+
+ ret = security_ib_endport_manage_subnet(agent->security,
+ agent->device->name,
+ agent->port_num);
+ if (ret)
+ return ret;
+
+ agent->lsm_nb.notifier_call = ib_mad_agent_security_change;
+ ret = register_lsm_notifier(&agent->lsm_nb);
+ if (ret)
+ return ret;
+
+ agent->smp_allowed = true;
+ agent->lsm_nb_reg = true;
+ return 0;
+}
+
+void ib_mad_agent_security_cleanup(struct ib_mad_agent *agent)
+{
+ security_ib_free_security(agent->security);
+ if (agent->lsm_nb_reg)
+ unregister_lsm_notifier(&agent->lsm_nb);
+}
+
+int ib_mad_enforce_security(struct ib_mad_agent_private *map, u16 pkey_index)
+{
+ int ret;
+
+ if (map->agent.qp->qp_type == IB_QPT_SMI && !map->agent.smp_allowed)
+ return -EACCES;
+
+ ret = ib_security_pkey_access(map->agent.device,
+ map->agent.port_num,
+ pkey_index,
+ map->agent.security);
+
+ if (ret)
+ return ret;
+
+ return 0;
+}
+
#endif /* CONFIG_SECURITY_INFINIBAND */
diff --git a/include/linux/lsm_hooks.h b/include/linux/lsm_hooks.h
index d576efc..3936216 100644
--- a/include/linux/lsm_hooks.h
+++ b/include/linux/lsm_hooks.h
@@ -911,6 +911,11 @@
* @subnet_prefix the subnet prefix of the port being used.
* @pkey the pkey to be accessed.
* @sec pointer to a security structure.
+ * @ib_endport_manage_subnet:
+ * Check permissions to send and receive SMPs on a end port.
+ * @dev_name the IB device name (i.e. mlx4_0).
+ * @port_num the port number.
+ * @sec pointer to a security structure.
* @ib_alloc_security:
* Allocate a security structure for Infiniband objects.
* @sec pointer to a security structure pointer.
@@ -1629,6 +1634,8 @@
#ifdef CONFIG_SECURITY_INFINIBAND
int (*ib_pkey_access)(void *sec, u64 subnet_prefix, u16 pkey);
+ int (*ib_endport_manage_subnet)(void *sec, const char *dev_name,
+ u8 port_num);
int (*ib_alloc_security)(void **sec);
void (*ib_free_security)(void *sec);
#endif /* CONFIG_SECURITY_INFINIBAND */
@@ -1865,6 +1872,7 @@ struct security_hook_heads {
#endif /* CONFIG_SECURITY_NETWORK */
#ifdef CONFIG_SECURITY_INFINIBAND
struct list_head ib_pkey_access;
+ struct list_head ib_endport_manage_subnet;
struct list_head ib_alloc_security;
struct list_head ib_free_security;
#endif /* CONFIG_SECURITY_INFINIBAND */
diff --git a/include/linux/security.h b/include/linux/security.h
index d8c29cd..9241f86 100644
--- a/include/linux/security.h
+++ b/include/linux/security.h
@@ -1404,6 +1404,7 @@ static inline int security_tun_dev_open(void *security)
#ifdef CONFIG_SECURITY_INFINIBAND
int security_ib_pkey_access(void *sec, u64 subnet_prefix, u16 pkey);
+int security_ib_endport_manage_subnet(void *sec, const char *name, u8 port_num);
int security_ib_alloc_security(void **sec);
void security_ib_free_security(void *sec);
#else /* CONFIG_SECURITY_INFINIBAND */
@@ -1412,6 +1413,11 @@ static inline int security_ib_pkey_access(void *sec, u64 subnet_prefix, u16 pkey
return 0;
}
+static inline int security_ib_endport_manage_subnet(void *sec, const char *dev_name, u8 port_num)
+{
+ return 0;
+}
+
static inline int security_ib_alloc_security(void **sec)
{
return 0;
diff --git a/include/rdma/ib_mad.h b/include/rdma/ib_mad.h
index c8a773f..23c4a4e 100644
--- a/include/rdma/ib_mad.h
+++ b/include/rdma/ib_mad.h
@@ -537,6 +537,10 @@ struct ib_mad_agent {
u32 flags;
u8 port_num;
u8 rmpp_version;
+ void *security;
+ bool smp_allowed;
+ bool lsm_nb_reg;
+ struct notifier_block lsm_nb;
};
/**
diff --git a/security/security.c b/security/security.c
index 40326d4..e4fd2b6 100644
--- a/security/security.c
+++ b/security/security.c
@@ -1470,6 +1470,12 @@ int security_ib_pkey_access(void *sec, u64 subnet_prefix, u16 pkey)
}
EXPORT_SYMBOL(security_ib_pkey_access);
+int security_ib_endport_manage_subnet(void *sec, const char *dev_name, u8 port_num)
+{
+ return call_int_hook(ib_endport_manage_subnet, 0, sec, dev_name, port_num);
+}
+EXPORT_SYMBOL(security_ib_endport_manage_subnet);
+
int security_ib_alloc_security(void **sec)
{
return call_int_hook(ib_alloc_security, 0, sec);
@@ -1943,6 +1949,8 @@ struct security_hook_heads security_hook_heads = {
#ifdef CONFIG_SECURITY_INFINIBAND
.ib_pkey_access = LIST_HEAD_INIT(security_hook_heads.ib_pkey_access),
+ .ib_endport_manage_subnet =
+ LIST_HEAD_INIT(security_hook_heads.ib_endport_manage_subnet),
.ib_alloc_security =
LIST_HEAD_INIT(security_hook_heads.ib_alloc_security),
.ib_free_security =
--
1.8.3.1
^ permalink raw reply related
* Re: [PATCH kernel v4 7/7] virtio-balloon: tell host vm's unused page info
From: Michael S. Tsirkin @ 2016-11-08 21:07 UTC (permalink / raw)
To: Dave Hansen
Cc: virtio-dev@lists.oasis-open.org, kvm@vger.kernel.org,
linux-kernel@vger.kernel.org, Li, Liang Z, qemu-devel@nongnu.org,
dgilbert@redhat.com, linux-mm@kvack.org, amit.shah@redhat.com,
pbonzini@redhat.com, virtualization@lists.linux-foundation.org,
mgorman@techsingularity.net
In-Reply-To: <281acd8d-fd94-6318-35e5-9eb130303dc6@intel.com>
On Mon, Nov 07, 2016 at 09:23:38AM -0800, Dave Hansen wrote:
> On 11/06/2016 07:37 PM, Li, Liang Z wrote:
> >> Let's say we do a 32k bitmap that can hold ~1M pages. That's 4GB of RAM.
> >> On a 1TB system, that's 256 passes through the top-level loop.
> >> The bottom-level lists have tens of thousands of pages in them, even on my
> >> laptop. Only 1/256 of these pages will get consumed in a given pass.
> >>
> > Your description is not exactly.
> > A 32k bitmap is used only when there is few free memory left in the system and when
> > the extend_page_bitmap() failed to allocate more memory for the bitmap. Or dozens of
> > 32k split bitmap will be used, this version limit the bitmap count to 32, it means we can use
> > at most 32*32 kB for the bitmap, which can cover 128GB for RAM. We can increase the bitmap
> > count limit to a larger value if 32 is not big enough.
>
> OK, so it tries to allocate a large bitmap. But, if it fails, it will
> try to work with a smaller bitmap. Correct?
>
> So, what's the _worst_ case? It sounds like it is even worse than I was
> positing.
>
> >> That's an awfully inefficient way of doing it. This patch essentially changed
> >> the data structure without changing the algorithm to populate it.
> >>
> >> Please change the *algorithm* to use the new data structure efficiently.
> >> Such a change would only do a single pass through each freelist, and would
> >> choose whether to use the extent-based (pfn -> range) or bitmap-based
> >> approach based on the contents of the free lists.
> >
> > Save the free page info to a raw bitmap first and then process the raw bitmap to
> > get the proper ' extent-based ' and 'bitmap-based' is the most efficient way I can
> > come up with to save the virtio data transmission. Do you have some better idea?
>
> That's kinda my point. This patch *does* processing to try to pack the
> bitmaps full of pages from the various pfn ranges. It's a form of
> processing that gets *REALLY*, *REALLY* bad in some (admittedly obscure)
> cases.
>
> Let's not pretend that making an essentially unlimited number of passes
> over the free lists is not processing.
>
> 1. Allocate as large of a bitmap as you can. (what you already do)
> 2. Iterate from the largest freelist order. Store those pages in the
> bitmap.
> 3. If you can no longer fit pages in the bitmap, return the list that
> you have.
> 4. Make an approximation about where the bitmap does not make any more,
> and fall back to listing individual PFNs. This would make sens, for
> instance in a large zone with very few free order-0 pages left.
In practice, a single PFN using the bitmap format
only takes up twice the size: I think it's 128 instead of 64 bit
per entry.
So it's not a a given that point 4 is worth it at any point,
just packing multiple bitmaps might be good enough.
>
> > It seems the benefit we get for this feature is not as big as that in fast balloon inflating/deflating.
> >>
> >> You should not be using get_max_pfn(). Any patch set that continues to use
> >> it is not likely to be using a proper algorithm.
> >
> > Do you have any suggestion about how to avoid it?
>
> Yes: get the pfns from the page free lists alone. Don't derive them
> from the pfn limits of the system or zones.
^ permalink raw reply
* Re: [RFC][ATCH 1/3] ptrace: Remove maxargs from task_current_syscall()
From: Andy Lutomirski @ 2016-11-08 21:06 UTC (permalink / raw)
To: Steven Rostedt
Cc: Andrew Morton, linux-arch, Oleg Nesterov, Ingo Molnar,
Linux Kernel Mailing List, Peter Zijlstra, Roland McGrath,
Linus Torvalds
In-Reply-To: <20161108144851.7a35c37d@gandalf.local.home>
On Nov 8, 2016 11:48 AM, "Steven Rostedt" <rostedt@goodmis.org> wrote:
>
> On Tue, 8 Nov 2016 08:20:48 -0800
> Andy Lutomirski <luto@amacapital.net> wrote:
>
> > On Tue, Nov 8, 2016 at 8:16 AM, Linus Torvalds
> > <torvalds@linux-foundation.org> wrote:
> > > So I definitely approve of the change, but I wonder if we should go
> > > one step further:
> > >
> > > On Mon, Nov 7, 2016 at 1:26 PM, Steven Rostedt <rostedt@goodmis.org> wrote:
> > >>
> > >> extern int task_current_syscall(struct task_struct *target, long *callno,
> > >> - unsigned long args[6], unsigned int maxargs,
> > >> - unsigned long *sp, unsigned long *pc);
> > >> + unsigned long args[6], unsigned long *sp,
> > >> + unsigned long *pc);
> > >
> > > The thing is, in C, having an array in a function declaration is
> > > pretty much exactly the same as just having a pointer, so from a type
> > > checking standpoint it doesn't really help all that much (but from a
> > > "human documentation" side the "args[6]" is much better than "*args").
> > >
> > > However, what would really help type checking is making it a
> > > structure. And maybe that structure could just contain "callno", "sp"
> > > and "pc" too? That would not only fix the type checking, it would make
> > > the calling convention even cleaner. Just have one single structure
> > > that contains all the relevant data.
> >
> > I would propose calling this 'struct seccomp_data'.
>
> I'm assuming you mean to use the existing seccomp_data? But isn't that
> already defined as a user structure? Thus, we can't add sp and pc to it.
pc is there. sp isn't, but that could be separate. Or you could
embed seccomp_data in a bigger structure.
^ permalink raw reply
* [PATCH v4 9/9] selinux: Add a cache for quicker retreival of PKey SIDs
From: Dan Jurgens @ 2016-11-08 21:06 UTC (permalink / raw)
To: chrisw-69jw2NvuJkxg9hUCZPvPmw, paul-r2n+y4ga6xFZroRs9YW3xA,
sds-+05T5uksL2qpZYMLLGbcSA, eparis-FjpueFixGhCM4zKIHC2jIg,
dledford-H+wXaHxf7aLQT0dZR+AlfA,
sean.hefty-ral2JQCrhuEAvxtiuMwx3w,
hal.rosenstock-Re5JQEeQqe8AvxtiuMwx3w
Cc: selinux-+05T5uksL2qpZYMLLGbcSA,
linux-security-module-u79uwXL29TY76Z2rM5mHXA,
linux-rdma-u79uwXL29TY76Z2rM5mHXA,
yevgenyp-VPRAkNaXOzVWk0Htik3J/w, liranl-VPRAkNaXOzVWk0Htik3J/w,
leonro-VPRAkNaXOzVWk0Htik3J/w, Daniel Jurgens
In-Reply-To: <1478639185-47521-1-git-send-email-danielj-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
From: Daniel Jurgens <danielj-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
It is likely that the SID for the same PKey will be requested many
times. To reduce the time to modify QPs and process MADs use a cache to
store PKey SIDs.
This code is heavily based on the "netif" and "netport" concept
originally developed by James Morris <jmorris-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org> and Paul Moore
<paul-r2n+y4ga6xFZroRs9YW3xA@public.gmane.org> (see security/selinux/netif.c and
security/selinux/netport.c for more information)
issue: 736423
Change-Id: I176c3079d5d84d06839b4f750100ac47a6081e94
Signed-off-by: Daniel Jurgens <danielj-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
---
v2:
- Renamed the files to ibpkey. Paul Moore
- Fixed a braket indentation mismatch in sel_pkey_find. Yuval Shaia
- Change spin_lock_bh to spin_lock_irqsave to resolve HARDIRQ lockdep
warning. Dan Jurgens
Signed-off-by: Daniel Jurgens <danielj-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
---
security/selinux/Makefile | 2 +-
security/selinux/hooks.c | 4 +-
security/selinux/ibpkey.c | 245 ++++++++++++++++++++++++++++++++++++++
security/selinux/include/ibpkey.h | 31 +++++
security/selinux/include/objsec.h | 6 +
5 files changed, 286 insertions(+), 2 deletions(-)
create mode 100644 security/selinux/ibpkey.c
create mode 100644 security/selinux/include/ibpkey.h
diff --git a/security/selinux/Makefile b/security/selinux/Makefile
index 3411c33..ff5895e 100644
--- a/security/selinux/Makefile
+++ b/security/selinux/Makefile
@@ -5,7 +5,7 @@
obj-$(CONFIG_SECURITY_SELINUX) := selinux.o
selinux-y := avc.o hooks.o selinuxfs.o netlink.o nlmsgtab.o netif.o \
- netnode.o netport.o exports.o \
+ netnode.o netport.o ibpkey.o exports.o \
ss/ebitmap.o ss/hashtab.o ss/symtab.o ss/sidtab.o ss/avtab.o \
ss/policydb.o ss/services.o ss/conditional.o ss/mls.o ss/status.o
diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c
index ea3f6d0..619d2ea 100644
--- a/security/selinux/hooks.c
+++ b/security/selinux/hooks.c
@@ -90,6 +90,7 @@
#include "netif.h"
#include "netnode.h"
#include "netport.h"
+#include "ibpkey.h"
#include "xfrm.h"
#include "netlabel.h"
#include "audit.h"
@@ -174,6 +175,7 @@ static int selinux_netcache_avc_callback(u32 event)
static int selinux_lsm_notifier_avc_callback(u32 event)
{
if (event == AVC_CALLBACK_RESET) {
+ sel_pkey_flush();
call_lsm_notifier(LSM_POLICY_CHANGE, NULL);
}
return 0;
@@ -6094,7 +6096,7 @@ static int selinux_ib_pkey_access(void *ib_sec, u64 subnet_prefix, u16 pkey_val)
struct ib_security_struct *sec = ib_sec;
struct lsm_pkey_audit pkey;
- err = security_pkey_sid(subnet_prefix, pkey_val, &sid);
+ err = sel_pkey_sid(subnet_prefix, pkey_val, &sid);
if (err)
return err;
diff --git a/security/selinux/ibpkey.c b/security/selinux/ibpkey.c
new file mode 100644
index 0000000..6e52c54
--- /dev/null
+++ b/security/selinux/ibpkey.c
@@ -0,0 +1,245 @@
+/*
+ * Pkey table
+ *
+ * SELinux must keep a mapping of Infinband PKEYs to labels/SIDs. This
+ * mapping is maintained as part of the normal policy but a fast cache is
+ * needed to reduce the lookup overhead.
+ *
+ * This code is heavily based on the "netif" and "netport" concept originally
+ * developed by
+ * James Morris <jmorris-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org> and
+ * Paul Moore <paul-r2n+y4ga6xFZroRs9YW3xA@public.gmane.org>
+ * (see security/selinux/netif.c and security/selinux/netport.c for more
+ * information)
+ *
+ */
+
+/*
+ * (c) Mellanox Technologies, 2016
+ *
+ * This program is free software: you can redistribute it and/or modify
+ * it under the terms of version 2 of the GNU General Public License as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ */
+
+#include <linux/types.h>
+#include <linux/rcupdate.h>
+#include <linux/list.h>
+#include <linux/spinlock.h>
+
+#include "ibpkey.h"
+#include "objsec.h"
+
+#define SEL_PKEY_HASH_SIZE 256
+#define SEL_PKEY_HASH_BKT_LIMIT 16
+
+struct sel_pkey_bkt {
+ int size;
+ struct list_head list;
+};
+
+struct sel_pkey {
+ struct pkey_security_struct psec;
+ struct list_head list;
+ struct rcu_head rcu;
+};
+
+static LIST_HEAD(sel_pkey_list);
+static DEFINE_SPINLOCK(sel_pkey_lock);
+static struct sel_pkey_bkt sel_pkey_hash[SEL_PKEY_HASH_SIZE];
+
+/**
+ * sel_pkey_hashfn - Hashing function for the pkey table
+ * @pkey: pkey number
+ *
+ * Description:
+ * This is the hashing function for the pkey table, it returns the bucket
+ * number for the given pkey.
+ *
+ */
+static unsigned int sel_pkey_hashfn(u16 pkey)
+{
+ return (pkey & (SEL_PKEY_HASH_SIZE - 1));
+}
+
+/**
+ * sel_pkey_find - Search for a pkey record
+ * @subnet_prefix: subnet_prefix
+ * @pkey_num: pkey_num
+ *
+ * Description:
+ * Search the pkey table and return the matching record. If an entry
+ * can not be found in the table return NULL.
+ *
+ */
+static struct sel_pkey *sel_pkey_find(u64 subnet_prefix, u16 pkey_num)
+{
+ unsigned int idx;
+ struct sel_pkey *pkey;
+
+ idx = sel_pkey_hashfn(pkey_num);
+ list_for_each_entry_rcu(pkey, &sel_pkey_hash[idx].list, list) {
+ if (pkey->psec.pkey == pkey_num &&
+ pkey->psec.subnet_prefix == subnet_prefix)
+ return pkey;
+ }
+
+ return NULL;
+}
+
+/**
+ * sel_pkey_insert - Insert a new pkey into the table
+ * @pkey: the new pkey record
+ *
+ * Description:
+ * Add a new pkey record to the hash table.
+ *
+ */
+static void sel_pkey_insert(struct sel_pkey *pkey)
+{
+ unsigned int idx;
+
+ /* we need to impose a limit on the growth of the hash table so check
+ * this bucket to make sure it is within the specified bounds
+ */
+ idx = sel_pkey_hashfn(pkey->psec.pkey);
+ list_add_rcu(&pkey->list, &sel_pkey_hash[idx].list);
+ if (sel_pkey_hash[idx].size == SEL_PKEY_HASH_BKT_LIMIT) {
+ struct sel_pkey *tail;
+
+ tail = list_entry(
+ rcu_dereference_protected(
+ sel_pkey_hash[idx].list.prev,
+ lockdep_is_held(&sel_pkey_lock)),
+ struct sel_pkey, list);
+ list_del_rcu(&tail->list);
+ kfree_rcu(tail, rcu);
+ } else {
+ sel_pkey_hash[idx].size++;
+ }
+}
+
+/**
+ * sel_pkey_sid_slow - Lookup the SID of a pkey using the policy
+ * @subnet_prefix: subnet prefix
+ * @pkey_num: pkey number
+ * @sid: pkey SID
+ *
+ * Description:
+ * This function determines the SID of a pkey by querying the security
+ * policy. The result is added to the pkey table to speedup future
+ * queries. Returns zero on success, negative values on failure.
+ *
+ */
+static int sel_pkey_sid_slow(u64 subnet_prefix, u16 pkey_num, u32 *sid)
+{
+ int ret = -ENOMEM;
+ struct sel_pkey *pkey;
+ struct sel_pkey *new = NULL;
+ unsigned long flags;
+
+ spin_lock_irqsave(&sel_pkey_lock, flags);
+ pkey = sel_pkey_find(subnet_prefix, pkey_num);
+ if (pkey) {
+ *sid = pkey->psec.sid;
+ spin_unlock_irqrestore(&sel_pkey_lock, flags);
+ return 0;
+ }
+
+ ret = security_pkey_sid(subnet_prefix, pkey_num, sid);
+ if (ret != 0)
+ goto out;
+
+ new = kzalloc(sizeof(*new), GFP_ATOMIC);
+ if (!new)
+ goto out;
+
+ new->psec.subnet_prefix = subnet_prefix;
+ new->psec.pkey = pkey_num;
+ new->psec.sid = *sid;
+ sel_pkey_insert(new);
+
+out:
+ spin_unlock_irqrestore(&sel_pkey_lock, flags);
+ if (unlikely(ret))
+ kfree(new);
+
+ return ret;
+}
+
+/**
+ * sel_pkey_sid - Lookup the SID of a PKEY
+ * @subnet_prefix: subnet_prefix
+ * @pkey_num: pkey number
+ * @sid: pkey SID
+ *
+ * Description:
+ * This function determines the SID of a PKEY using the fastest method
+ * possible. First the pkey table is queried, but if an entry can't be found
+ * then the policy is queried and the result is added to the table to speedup
+ * future queries. Returns zero on success, negative values on failure.
+ *
+ */
+int sel_pkey_sid(u64 subnet_prefix, u16 pkey_num, u32 *sid)
+{
+ struct sel_pkey *pkey;
+
+ rcu_read_lock();
+ pkey = sel_pkey_find(subnet_prefix, pkey_num);
+ if (pkey) {
+ *sid = pkey->psec.sid;
+ rcu_read_unlock();
+ return 0;
+ }
+ rcu_read_unlock();
+
+ return sel_pkey_sid_slow(subnet_prefix, pkey_num, sid);
+}
+
+/**
+ * sel_pkey_flush - Flush the entire pkey table
+ *
+ * Description:
+ * Remove all entries from the pkey table
+ *
+ */
+void sel_pkey_flush(void)
+{
+ unsigned int idx;
+ struct sel_pkey *pkey, *pkey_tmp;
+ unsigned long flags;
+
+ spin_lock_irqsave(&sel_pkey_lock, flags);
+ for (idx = 0; idx < SEL_PKEY_HASH_SIZE; idx++) {
+ list_for_each_entry_safe(pkey, pkey_tmp,
+ &sel_pkey_hash[idx].list, list) {
+ list_del_rcu(&pkey->list);
+ kfree_rcu(pkey, rcu);
+ }
+ sel_pkey_hash[idx].size = 0;
+ }
+ spin_unlock_irqrestore(&sel_pkey_lock, flags);
+}
+
+static __init int sel_pkey_init(void)
+{
+ int iter;
+
+ if (!selinux_enabled)
+ return 0;
+
+ for (iter = 0; iter < SEL_PKEY_HASH_SIZE; iter++) {
+ INIT_LIST_HEAD(&sel_pkey_hash[iter].list);
+ sel_pkey_hash[iter].size = 0;
+ }
+
+ return 0;
+}
+
+subsys_initcall(sel_pkey_init);
diff --git a/security/selinux/include/ibpkey.h b/security/selinux/include/ibpkey.h
new file mode 100644
index 0000000..387885a
--- /dev/null
+++ b/security/selinux/include/ibpkey.h
@@ -0,0 +1,31 @@
+/*
+ * pkey table
+ *
+ * SELinux must keep a mapping of pkeys to labels/SIDs. This
+ * mapping is maintained as part of the normal policy but a fast cache is
+ * needed to reduce the lookup overhead.
+ *
+ */
+
+/*
+ * (c) Mellanox Technologies, 2016
+ *
+ * This program is free software: you can redistribute it and/or modify
+ * it under the terms of version 2 of the GNU General Public License as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ */
+
+#ifndef _SELINUX_IB_PKEY_H
+#define _SELINUX_IB_PKEY_H
+
+void sel_pkey_flush(void);
+
+int sel_pkey_sid(u64 subnet_prefix, u16 pkey, u32 *sid);
+
+#endif
diff --git a/security/selinux/include/objsec.h b/security/selinux/include/objsec.h
index 8e7db43..4139f28 100644
--- a/security/selinux/include/objsec.h
+++ b/security/selinux/include/objsec.h
@@ -133,6 +133,12 @@ struct ib_security_struct {
u32 sid; /* SID of the queue pair or MAD agent */
};
+struct pkey_security_struct {
+ u64 subnet_prefix; /* Port subnet prefix */
+ u16 pkey; /* PKey number */
+ u32 sid; /* SID of pkey */
+};
+
extern unsigned int selinux_checkreqprot;
#endif /* _SELINUX_OBJSEC_H_ */
--
1.8.3.1
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply related
* [PATCH v4 8/9] selinux: Add IB Port SMP access vector
From: Dan Jurgens @ 2016-11-08 21:06 UTC (permalink / raw)
To: chrisw-69jw2NvuJkxg9hUCZPvPmw, paul-r2n+y4ga6xFZroRs9YW3xA,
sds-+05T5uksL2qpZYMLLGbcSA, eparis-FjpueFixGhCM4zKIHC2jIg,
dledford-H+wXaHxf7aLQT0dZR+AlfA,
sean.hefty-ral2JQCrhuEAvxtiuMwx3w,
hal.rosenstock-Re5JQEeQqe8AvxtiuMwx3w
Cc: selinux-+05T5uksL2qpZYMLLGbcSA,
linux-security-module-u79uwXL29TY76Z2rM5mHXA,
linux-rdma-u79uwXL29TY76Z2rM5mHXA,
yevgenyp-VPRAkNaXOzVWk0Htik3J/w, liranl-VPRAkNaXOzVWk0Htik3J/w,
leonro-VPRAkNaXOzVWk0Htik3J/w, Daniel Jurgens
In-Reply-To: <1478639185-47521-1-git-send-email-danielj-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
From: Daniel Jurgens <danielj-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
Add a type for Infiniband ports and an access vector for subnet
management packets. Implement the ib_port_smp hook to check that the
caller has permission to send and receive SMPs on the end port specified
by the device name and port. Add interface to query the SID for a IB
port, which walks the IB_PORT ocontexts to find an entry for the
given name and port.
issue: 736423
Change-Id: If8b365f3cf32e77a2060073f1a53e27ea846804d
Signed-off-by: Daniel Jurgens <danielj-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
---
v2:
- Shorted ib_end_port. Paul Moore
- Pass void blobs to security hooks. Paul Moore
- Log specific IB port info in audit log. Paul Moore
- Don't create a new intial sid, use unlabeled. Stephen Smalley
- Changed "smp" to "manage_subnet". Paul Moore
v3:
- ib_port -> ib_endport. Paul Moore
- Don't log device name as untrusted string. Paul Moore
- Reorder parameters of LSM hook. Paul Moore
Signed-off-by: Daniel Jurgens <danielj-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
---
include/linux/lsm_audit.h | 8 +++++++
security/lsm_audit.c | 5 +++++
security/selinux/hooks.c | 25 ++++++++++++++++++++++
security/selinux/include/classmap.h | 2 ++
security/selinux/include/security.h | 2 ++
security/selinux/ss/services.c | 42 +++++++++++++++++++++++++++++++++++++
6 files changed, 84 insertions(+)
diff --git a/include/linux/lsm_audit.h b/include/linux/lsm_audit.h
index 402b770..7047b4c 100644
--- a/include/linux/lsm_audit.h
+++ b/include/linux/lsm_audit.h
@@ -21,6 +21,7 @@
#include <linux/path.h>
#include <linux/key.h>
#include <linux/skbuff.h>
+#include <rdma/ib_verbs.h>
struct lsm_network_audit {
int netif;
@@ -50,6 +51,11 @@ struct lsm_pkey_audit {
u16 pkey;
};
+struct lsm_ib_endport_audit {
+ char dev_name[IB_DEVICE_NAME_MAX];
+ u8 port_num;
+};
+
/* Auxiliary data to use in generating the audit record. */
struct common_audit_data {
char type;
@@ -66,6 +72,7 @@ struct common_audit_data {
#define LSM_AUDIT_DATA_IOCTL_OP 11
#define LSM_AUDIT_DATA_FILE 12
#define LSM_AUDIT_DATA_PKEY 13
+#define LSM_AUDIT_DATA_IB_ENDPORT 14
union {
struct path path;
struct dentry *dentry;
@@ -84,6 +91,7 @@ struct common_audit_data {
struct lsm_ioctlop_audit *op;
struct file *file;
struct lsm_pkey_audit *pkey;
+ struct lsm_ib_endport_audit *ib_endport;
} u;
/* this union contains LSM specific data */
union {
diff --git a/security/lsm_audit.c b/security/lsm_audit.c
index b18d277..549fe9d 100644
--- a/security/lsm_audit.c
+++ b/security/lsm_audit.c
@@ -423,6 +423,11 @@ static void dump_common_audit_data(struct audit_buffer *ab,
a->u.pkey->pkey, &sbn_pfx);
break;
}
+ case LSM_AUDIT_DATA_IB_ENDPORT:
+ audit_log_format(ab, " device=%s port_num=%u",
+ a->u.ib_endport->dev_name,
+ a->u.ib_endport->port_num);
+ break;
} /* switch (a->type) */
}
diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c
index 20fb292..ea3f6d0 100644
--- a/security/selinux/hooks.c
+++ b/security/selinux/hooks.c
@@ -6108,6 +6108,29 @@ static int selinux_ib_pkey_access(void *ib_sec, u64 subnet_prefix, u16 pkey_val)
INFINIBAND_PKEY__ACCESS, &ad);
}
+static int selinux_ib_endport_manage_subnet(void *ib_sec, const char *dev_name,
+ u8 port_num)
+{
+ struct common_audit_data ad;
+ int err;
+ u32 sid = 0;
+ struct ib_security_struct *sec = ib_sec;
+ struct lsm_ib_endport_audit ib_endport;
+
+ err = security_ib_endport_sid(dev_name, port_num, &sid);
+
+ if (err)
+ return err;
+
+ ad.type = LSM_AUDIT_DATA_IB_ENDPORT;
+ strncpy(ib_endport.dev_name, dev_name, sizeof(ib_endport.dev_name));
+ ib_endport.port_num = port_num;
+ ad.u.ib_endport = &ib_endport;
+ return avc_has_perm(sec->sid, sid,
+ SECCLASS_INFINIBAND_ENDPORT,
+ INFINIBAND_ENDPORT__MANAGE_SUBNET, &ad);
+}
+
static int selinux_ib_alloc_security(void **ib_sec)
{
struct ib_security_struct *sec;
@@ -6313,6 +6336,8 @@ static void selinux_ib_free_security(void *ib_sec)
LSM_HOOK_INIT(tun_dev_open, selinux_tun_dev_open),
#ifdef CONFIG_SECURITY_INFINIBAND
LSM_HOOK_INIT(ib_pkey_access, selinux_ib_pkey_access),
+ LSM_HOOK_INIT(ib_endport_manage_subnet,
+ selinux_ib_endport_manage_subnet),
LSM_HOOK_INIT(ib_alloc_security, selinux_ib_alloc_security),
LSM_HOOK_INIT(ib_free_security, selinux_ib_free_security),
#endif
diff --git a/security/selinux/include/classmap.h b/security/selinux/include/classmap.h
index d42dd4d..f93b64b 100644
--- a/security/selinux/include/classmap.h
+++ b/security/selinux/include/classmap.h
@@ -167,5 +167,7 @@ struct security_class_mapping secclass_map[] = {
{ COMMON_CAP2_PERMS, NULL } },
{ "infiniband_pkey",
{ "access", NULL } },
+ { "infiniband_endport",
+ { "manage_subnet", NULL } },
{ NULL }
};
diff --git a/security/selinux/include/security.h b/security/selinux/include/security.h
index 17afb7c..8a6e5e7 100644
--- a/security/selinux/include/security.h
+++ b/security/selinux/include/security.h
@@ -178,6 +178,8 @@ int security_get_user_sids(u32 callsid, char *username,
int security_pkey_sid(u64 subnet_prefix, u16 pkey_num, u32 *out_sid);
+int security_ib_endport_sid(const char *dev_name, u8 port_num, u32 *out_sid);
+
int security_netif_sid(char *name, u32 *if_sid);
int security_node_sid(u16 domain, void *addr, u32 addrlen,
diff --git a/security/selinux/ss/services.c b/security/selinux/ss/services.c
index 085c54b..db40ce7 100644
--- a/security/selinux/ss/services.c
+++ b/security/selinux/ss/services.c
@@ -2244,6 +2244,48 @@ int security_pkey_sid(u64 subnet_prefix, u16 pkey_num, u32 *out_sid)
}
/**
+ * security_ib_endport_sid - Obtain the SID for a subnet management interface.
+ * @dev_name: device name
+ * @port: port number
+ * @out_sid: security identifier
+ */
+int security_ib_endport_sid(const char *dev_name, u8 port_num, u32 *out_sid)
+{
+ struct ocontext *c;
+ int rc = 0;
+
+ read_lock(&policy_rwlock);
+
+ c = policydb.ocontexts[OCON_IB_ENDPORT];
+ while (c) {
+ if (c->u.ib_endport.port_num == port_num &&
+ !strncmp(c->u.ib_endport.dev_name,
+ dev_name,
+ IB_DEVICE_NAME_MAX))
+ break;
+
+ c = c->next;
+ }
+
+ if (c) {
+ if (!c->sid[0]) {
+ rc = sidtab_context_to_sid(&sidtab,
+ &c->context[0],
+ &c->sid[0]);
+ if (rc)
+ goto out;
+ }
+ *out_sid = c->sid[0];
+ } else {
+ *out_sid = SECINITSID_UNLABELED;
+ }
+
+out:
+ read_unlock(&policy_rwlock);
+ return rc;
+}
+
+/**
* security_netif_sid - Obtain the SID for a network interface.
* @name: interface name
* @if_sid: interface SID
--
1.8.3.1
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply related
* Re: [PATCH 15/33] userfaultfd: hugetlbfs: add __mcopy_atomic_hugetlb for huge page UFFDIO_COPY
From: Mike Kravetz @ 2016-11-08 21:06 UTC (permalink / raw)
To: Andrea Arcangeli
Cc: Hillf Danton, 'Andrew Morton', linux-mm,
'Dr. David Alan Gilbert', 'Shaohua Li',
'Pavel Emelyanov', 'Mike Rapoport'
In-Reply-To: <20161104193626.GU4611@redhat.com>
On 11/04/2016 12:36 PM, Andrea Arcangeli wrote:
> This is the current status, I'm sending a full diff against the
> previous submit for review of the latest updates. It's easier to
> review incrementally I think.
>
> Please test it, I updated the aa.git tree userfault branch in sync
> with this.
>
Hello Andrea,
I found a couple more issues with hugetlbfs support. The below patch
describes and addresses the issues. It is against your aa tree. Do note
that there is a patch going into mm tree that is a pre-req for this
patch. The patch is "mm/hugetlb: fix huge page reservation leak in
private mapping error paths".
http://marc.info/?l=linux-mm&m=147693310409312&w=2
--
Mike Kravetz
From: Mike Kravetz <mike.kravetz@oracle.com>
userfaultfd: hugetlbfs: fix __mcopy_atomic_hugetlb retry/error processing
The new routine copy_huge_page_from_user() uses kmap_atomic() to map
PAGE_SIZE pages. However, this prevents page faults in the subsequent
call to copy_from_user(). This is OK in the case where the routine
is copied with mmap_sema held. However, in another case we want to
allow page faults. So, add a new argument allow_pagefault to indicate
if the routine should allow page faults.
A patch (mm/hugetlb: fix huge page reservation leak in private mapping
error paths) was recently submitted and is being added to -mm tree. It
addresses the issue huge page reservations when a huge page is allocated,
and free'ed before being instantiated in an address space. This would
typically happen in error paths. The routine __mcopy_atomic_hugetlb has
such an error path, so it will need to call restore_reserve_on_error()
before free'ing the huge page. restore_reserve_on_error is currently
only visible in mm/hugetlb.c. So, add it to a header file so that it
can be used in mm/userfaultfd.c. Another option would be to move
__mcopy_atomic_hugetlb into mm/hugetlb.c
Signed-off-by: Mike Kravetz <mike.kravetz@oracle.com>
---
include/linux/hugetlb.h | 2 ++
include/linux/mm.h | 3 ++-
mm/hugetlb.c | 7 +++----
mm/memory.c | 13 ++++++++++---
mm/userfaultfd.c | 6 ++++--
5 files changed, 21 insertions(+), 10 deletions(-)
diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h
index fc27b66..bf02b7e 100644
--- a/include/linux/hugetlb.h
+++ b/include/linux/hugetlb.h
@@ -101,6 +101,8 @@ u32 hugetlb_fault_mutex_hash(struct hstate *h,
struct mm_struct *mm,
struct vm_area_struct *vma,
struct address_space *mapping,
pgoff_t idx, unsigned long address);
+void restore_reserve_on_error(struct hstate *h, struct vm_area_struct *vma,
+ unsigned long address, struct page *page);
pte_t *huge_pmd_share(struct mm_struct *mm, unsigned long addr, pud_t
*pud);
diff --git a/include/linux/mm.h b/include/linux/mm.h
index 39157f5..7c73a05 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -2417,7 +2417,8 @@ extern void copy_user_huge_page(struct page *dst,
struct page *src,
unsigned int pages_per_huge_page);
extern long copy_huge_page_from_user(struct page *dst_page,
const void __user *usr_src,
- unsigned int pages_per_huge_page);
+ unsigned int pages_per_huge_page,
+ bool allow_pagefault);
#endif /* CONFIG_TRANSPARENT_HUGEPAGE || CONFIG_HUGETLBFS */
extern struct page_ext_operations debug_guardpage_ops;
diff --git a/mm/hugetlb.c b/mm/hugetlb.c
index 7bfeee3..9ce8ecb 100644
--- a/mm/hugetlb.c
+++ b/mm/hugetlb.c
@@ -1935,9 +1935,8 @@ static long vma_add_reservation(struct hstate *h,
* reserve map here to be consistent with global reserve count adjustments
* to be made by free_huge_page.
*/
-static void restore_reserve_on_error(struct hstate *h,
- struct vm_area_struct *vma, unsigned long address,
- struct page *page)
+void restore_reserve_on_error(struct hstate *h, struct vm_area_struct *vma,
+ unsigned long address, struct page *page)
{
if (unlikely(PagePrivate(page))) {
long rc = vma_needs_reservation(h, vma, address);
@@ -3981,7 +3980,7 @@ int hugetlb_mcopy_atomic_pte(struct mm_struct *dst_mm,
ret = copy_huge_page_from_user(page,
(const void __user *) src_addr,
- pages_per_huge_page(h));
+ pages_per_huge_page(h), false);
/* fallback to copy_from_user outside mmap_sem */
if (unlikely(ret)) {
diff --git a/mm/memory.c b/mm/memory.c
index b911110..0137c4a 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -4106,7 +4106,8 @@ void copy_user_huge_page(struct page *dst, struct
page *src,
long copy_huge_page_from_user(struct page *dst_page,
const void __user *usr_src,
- unsigned int pages_per_huge_page)
+ unsigned int pages_per_huge_page,
+ bool allow_pagefault)
{
void *src = (void *)usr_src;
void *page_kaddr;
@@ -4114,11 +4115,17 @@ long copy_huge_page_from_user(struct page *dst_page,
unsigned long ret_val = pages_per_huge_page * PAGE_SIZE;
for (i = 0; i < pages_per_huge_page; i++) {
- page_kaddr = kmap_atomic(dst_page + i);
+ if (allow_pagefault)
+ page_kaddr = kmap(dst_page + i);
+ else
+ page_kaddr = kmap_atomic(dst_page + i);
rc = copy_from_user(page_kaddr,
(const void __user *)(src + i * PAGE_SIZE),
PAGE_SIZE);
- kunmap_atomic(page_kaddr);
+ if (allow_pagefault)
+ kunmap(page_kaddr);
+ else
+ kunmap_atomic(page_kaddr);
ret_val -= (PAGE_SIZE - rc);
if (rc)
diff --git a/mm/userfaultfd.c b/mm/userfaultfd.c
index e8d7a89..c8588aa 100644
--- a/mm/userfaultfd.c
+++ b/mm/userfaultfd.c
@@ -275,7 +275,7 @@ static __always_inline ssize_t
__mcopy_atomic_hugetlb(struct mm_struct *dst_mm,
err = copy_huge_page_from_user(page,
(const void __user *)src_addr,
- pages_per_huge_page(h));
+ pages_per_huge_page(h), true);
if (unlikely(err)) {
err = -EFAULT;
goto out;
@@ -302,8 +302,10 @@ static __always_inline ssize_t
__mcopy_atomic_hugetlb(struct mm_struct *dst_mm,
out_unlock:
up_read(&dst_mm->mmap_sem);
out:
- if (page)
+ if (page) {
+ restore_reserve_on_error(h, dst_vma, dst_addr, page);
put_page(page);
+ }
BUG_ON(copied < 0);
BUG_ON(err > 0);
BUG_ON(!copied && !err);
--
2.7.4
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
^ permalink raw reply related
* [PATCH v4 7/9] selinux: Implement Infiniband PKey "Access" access vector
From: Dan Jurgens @ 2016-11-08 21:06 UTC (permalink / raw)
To: chrisw-69jw2NvuJkxg9hUCZPvPmw, paul-r2n+y4ga6xFZroRs9YW3xA,
sds-+05T5uksL2qpZYMLLGbcSA, eparis-FjpueFixGhCM4zKIHC2jIg,
dledford-H+wXaHxf7aLQT0dZR+AlfA,
sean.hefty-ral2JQCrhuEAvxtiuMwx3w,
hal.rosenstock-Re5JQEeQqe8AvxtiuMwx3w
Cc: selinux-+05T5uksL2qpZYMLLGbcSA,
linux-security-module-u79uwXL29TY76Z2rM5mHXA,
linux-rdma-u79uwXL29TY76Z2rM5mHXA,
yevgenyp-VPRAkNaXOzVWk0Htik3J/w, liranl-VPRAkNaXOzVWk0Htik3J/w,
leonro-VPRAkNaXOzVWk0Htik3J/w, Daniel Jurgens
In-Reply-To: <1478639185-47521-1-git-send-email-danielj-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
From: Daniel Jurgens <danielj-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
Add a type and access vector for PKeys. Implement the ib_pkey_access
hook to check that the caller has permission to access the PKey on the
given subnet prefix. Add an interface to get the PKey SID. Walk the PKey
ocontexts to find an entry for the given subnet prefix and pkey.
issue: 736423
Change-Id: Ib3f7df9b5f98da77a6048855d146f9dcaceba336
Signed-off-by: Daniel Jurgens <danielj-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
---
v2:
- Use void* blobs for security structs. Paul Moore
- Add pkey specific data to the audit log. Paul Moore
- Don't introduce a new initial sid, use unlabeled. Stephen Smalley
v3:
- Reorder parameters to pkey_access hook. Paul Moore
Signed-off-by: Daniel Jurgens <danielj-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
---
include/linux/lsm_audit.h | 7 +++++++
security/lsm_audit.c | 13 ++++++++++++
security/selinux/hooks.c | 23 +++++++++++++++++++++
security/selinux/include/classmap.h | 2 ++
security/selinux/include/security.h | 2 ++
security/selinux/ss/services.c | 41 +++++++++++++++++++++++++++++++++++++
6 files changed, 88 insertions(+)
diff --git a/include/linux/lsm_audit.h b/include/linux/lsm_audit.h
index e58e577..402b770 100644
--- a/include/linux/lsm_audit.h
+++ b/include/linux/lsm_audit.h
@@ -45,6 +45,11 @@ struct lsm_ioctlop_audit {
u16 cmd;
};
+struct lsm_pkey_audit {
+ u64 subnet_prefix;
+ u16 pkey;
+};
+
/* Auxiliary data to use in generating the audit record. */
struct common_audit_data {
char type;
@@ -60,6 +65,7 @@ struct common_audit_data {
#define LSM_AUDIT_DATA_DENTRY 10
#define LSM_AUDIT_DATA_IOCTL_OP 11
#define LSM_AUDIT_DATA_FILE 12
+#define LSM_AUDIT_DATA_PKEY 13
union {
struct path path;
struct dentry *dentry;
@@ -77,6 +83,7 @@ struct common_audit_data {
char *kmod_name;
struct lsm_ioctlop_audit *op;
struct file *file;
+ struct lsm_pkey_audit *pkey;
} u;
/* this union contains LSM specific data */
union {
diff --git a/security/lsm_audit.c b/security/lsm_audit.c
index 37f04da..b18d277 100644
--- a/security/lsm_audit.c
+++ b/security/lsm_audit.c
@@ -410,6 +410,19 @@ static void dump_common_audit_data(struct audit_buffer *ab,
audit_log_format(ab, " kmod=");
audit_log_untrustedstring(ab, a->u.kmod_name);
break;
+ case LSM_AUDIT_DATA_PKEY: {
+ struct in6_addr sbn_pfx;
+
+ memset(&sbn_pfx.s6_addr, 0,
+ sizeof(sbn_pfx.s6_addr));
+
+ memcpy(&sbn_pfx.s6_addr, &a->u.pkey->subnet_prefix,
+ sizeof(a->u.pkey->subnet_prefix));
+
+ audit_log_format(ab, " pkey=0x%x subnet_prefix=%pI6c",
+ a->u.pkey->pkey, &sbn_pfx);
+ break;
+ }
} /* switch (a->type) */
}
diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c
index f48759d..20fb292 100644
--- a/security/selinux/hooks.c
+++ b/security/selinux/hooks.c
@@ -6086,6 +6086,28 @@ static int selinux_key_getsecurity(struct key *key, char **_buffer)
#endif
#ifdef CONFIG_SECURITY_INFINIBAND
+static int selinux_ib_pkey_access(void *ib_sec, u64 subnet_prefix, u16 pkey_val)
+{
+ struct common_audit_data ad;
+ int err;
+ u32 sid = 0;
+ struct ib_security_struct *sec = ib_sec;
+ struct lsm_pkey_audit pkey;
+
+ err = security_pkey_sid(subnet_prefix, pkey_val, &sid);
+
+ if (err)
+ return err;
+
+ ad.type = LSM_AUDIT_DATA_PKEY;
+ pkey.subnet_prefix = subnet_prefix;
+ pkey.pkey = pkey_val;
+ ad.u.pkey = &pkey;
+ return avc_has_perm(sec->sid, sid,
+ SECCLASS_INFINIBAND_PKEY,
+ INFINIBAND_PKEY__ACCESS, &ad);
+}
+
static int selinux_ib_alloc_security(void **ib_sec)
{
struct ib_security_struct *sec;
@@ -6290,6 +6312,7 @@ static void selinux_ib_free_security(void *ib_sec)
LSM_HOOK_INIT(tun_dev_attach, selinux_tun_dev_attach),
LSM_HOOK_INIT(tun_dev_open, selinux_tun_dev_open),
#ifdef CONFIG_SECURITY_INFINIBAND
+ LSM_HOOK_INIT(ib_pkey_access, selinux_ib_pkey_access),
LSM_HOOK_INIT(ib_alloc_security, selinux_ib_alloc_security),
LSM_HOOK_INIT(ib_free_security, selinux_ib_free_security),
#endif
diff --git a/security/selinux/include/classmap.h b/security/selinux/include/classmap.h
index 1f1f4b2..d42dd4d 100644
--- a/security/selinux/include/classmap.h
+++ b/security/selinux/include/classmap.h
@@ -165,5 +165,7 @@ struct security_class_mapping secclass_map[] = {
{ COMMON_CAP_PERMS, NULL } },
{ "cap2_userns",
{ COMMON_CAP2_PERMS, NULL } },
+ { "infiniband_pkey",
+ { "access", NULL } },
{ NULL }
};
diff --git a/security/selinux/include/security.h b/security/selinux/include/security.h
index 6bb9b0a..17afb7c 100644
--- a/security/selinux/include/security.h
+++ b/security/selinux/include/security.h
@@ -176,6 +176,8 @@ int security_get_user_sids(u32 callsid, char *username,
int security_port_sid(u8 protocol, u16 port, u32 *out_sid);
+int security_pkey_sid(u64 subnet_prefix, u16 pkey_num, u32 *out_sid);
+
int security_netif_sid(char *name, u32 *if_sid);
int security_node_sid(u16 domain, void *addr, u32 addrlen,
diff --git a/security/selinux/ss/services.c b/security/selinux/ss/services.c
index 082b20c..085c54b 100644
--- a/security/selinux/ss/services.c
+++ b/security/selinux/ss/services.c
@@ -2203,6 +2203,47 @@ int security_port_sid(u8 protocol, u16 port, u32 *out_sid)
}
/**
+ * security_pkey_sid - Obtain the SID for a pkey.
+ * @subnet_prefix: Subnet Prefix
+ * @pkey_num: pkey number
+ * @out_sid: security identifier
+ */
+int security_pkey_sid(u64 subnet_prefix, u16 pkey_num, u32 *out_sid)
+{
+ struct ocontext *c;
+ int rc = 0;
+
+ read_lock(&policy_rwlock);
+
+ c = policydb.ocontexts[OCON_PKEY];
+ while (c) {
+ if (c->u.pkey.low_pkey <= pkey_num &&
+ c->u.pkey.high_pkey >= pkey_num &&
+ c->u.pkey.subnet_prefix == subnet_prefix)
+ break;
+
+ c = c->next;
+ }
+
+ if (c) {
+ if (!c->sid[0]) {
+ rc = sidtab_context_to_sid(&sidtab,
+ &c->context[0],
+ &c->sid[0]);
+ if (rc)
+ goto out;
+ }
+ *out_sid = c->sid[0];
+ } else {
+ *out_sid = SECINITSID_UNLABELED;
+ }
+
+out:
+ read_unlock(&policy_rwlock);
+ return rc;
+}
+
+/**
* security_netif_sid - Obtain the SID for a network interface.
* @name: interface name
* @if_sid: interface SID
--
1.8.3.1
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply related
* [PATCH v4 6/9] selinux: Allocate and free infiniband security hooks
From: Dan Jurgens @ 2016-11-08 21:06 UTC (permalink / raw)
To: chrisw, paul, sds, eparis, dledford, sean.hefty, hal.rosenstock
Cc: selinux, linux-security-module, linux-rdma, yevgenyp, liranl,
leonro, Daniel Jurgens
In-Reply-To: <1478639185-47521-1-git-send-email-danielj@mellanox.com>
From: Daniel Jurgens <danielj@mellanox.com>
Implement and attach hooks to allocate and free Infiniband object
security structures.
issue: 736423
Change-Id: I3bdbecee7aab6d7615a02967c39a5a8792a14d44
Signed-off-by: Daniel Jurgens <danielj@mellanox.com>
---
v2:
- Use void * blobs for security structs. Paul Moore
- Shorten ib_end_port to ib_port. Paul Moore
- Allocate memory for security struct with GFP_KERNEL. Yuval Shaia
Signed-off-by: Daniel Jurgens <danielj@mellanox.com>
---
security/selinux/hooks.c | 25 ++++++++++++++++++++++++-
security/selinux/include/objsec.h | 5 +++++
2 files changed, 29 insertions(+), 1 deletion(-)
diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c
index 7578198..f48759d 100644
--- a/security/selinux/hooks.c
+++ b/security/selinux/hooks.c
@@ -17,6 +17,7 @@
* Paul Moore <paul@paul-moore.com>
* Copyright (C) 2007 Hitachi Software Engineering Co., Ltd.
* Yuichi Nakamura <ynakam@hitachisoft.jp>
+ * Copyright (C) 2016 Mellanox Technologies
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License version 2,
@@ -6082,7 +6083,26 @@ static int selinux_key_getsecurity(struct key *key, char **_buffer)
*_buffer = context;
return rc;
}
+#endif
+
+#ifdef CONFIG_SECURITY_INFINIBAND
+static int selinux_ib_alloc_security(void **ib_sec)
+{
+ struct ib_security_struct *sec;
+
+ sec = kzalloc(sizeof(*sec), GFP_KERNEL);
+ if (!sec)
+ return -ENOMEM;
+ sec->sid = current_sid();
+
+ *ib_sec = sec;
+ return 0;
+}
+static void selinux_ib_free_security(void *ib_sec)
+{
+ kfree(ib_sec);
+}
#endif
static struct security_hook_list selinux_hooks[] = {
@@ -6269,7 +6289,10 @@ static int selinux_key_getsecurity(struct key *key, char **_buffer)
LSM_HOOK_INIT(tun_dev_attach_queue, selinux_tun_dev_attach_queue),
LSM_HOOK_INIT(tun_dev_attach, selinux_tun_dev_attach),
LSM_HOOK_INIT(tun_dev_open, selinux_tun_dev_open),
-
+#ifdef CONFIG_SECURITY_INFINIBAND
+ LSM_HOOK_INIT(ib_alloc_security, selinux_ib_alloc_security),
+ LSM_HOOK_INIT(ib_free_security, selinux_ib_free_security),
+#endif
#ifdef CONFIG_SECURITY_NETWORK_XFRM
LSM_HOOK_INIT(xfrm_policy_alloc_security, selinux_xfrm_policy_alloc),
LSM_HOOK_INIT(xfrm_policy_clone_security, selinux_xfrm_policy_clone),
diff --git a/security/selinux/include/objsec.h b/security/selinux/include/objsec.h
index c21e135..8e7db43 100644
--- a/security/selinux/include/objsec.h
+++ b/security/selinux/include/objsec.h
@@ -10,6 +10,7 @@
*
* Copyright (C) 2001,2002 Networks Associates Technology, Inc.
* Copyright (C) 2003 Red Hat, Inc., James Morris <jmorris@redhat.com>
+ * Copyright (C) 2016 Mellanox Technologies
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License version 2,
@@ -128,6 +129,10 @@ struct key_security_struct {
u32 sid; /* SID of key */
};
+struct ib_security_struct {
+ u32 sid; /* SID of the queue pair or MAD agent */
+};
+
extern unsigned int selinux_checkreqprot;
#endif /* _SELINUX_OBJSEC_H_ */
--
1.8.3.1
^ permalink raw reply related
* [PATCH v4 5/9] selinux: Create policydb version for Infiniband support
From: Dan Jurgens @ 2016-11-08 21:06 UTC (permalink / raw)
To: chrisw, paul, sds, eparis, dledford, sean.hefty, hal.rosenstock
Cc: selinux, linux-security-module, linux-rdma, yevgenyp, liranl,
leonro, Daniel Jurgens
In-Reply-To: <1478639185-47521-1-git-send-email-danielj@mellanox.com>
From: Daniel Jurgens <danielj@mellanox.com>
Support for Infiniband requires the addition of two new object contexts,
one for infiniband PKeys and another IB Ports. Added handlers to read
and write the new ocontext types when reading or writing a binary policy
representation.
issue: 736423
Change-Id: I906bf7a1e0bd59c3cccdce278e5fe98ba8c404b8
Signed-off-by: Daniel Jurgens <danielj@mellanox.com>
Reviewed-by: Eli Cohen <eli@mellanox.com>
---
v2:
- Shorten ib_end_port to ib_port. Paul Moore
- Added bounds checking to port number. Paul Moore
- Eliminated {} in OCON_PKEY case statement. Yuval Shaia
v3:
- ib_port -> ib_endport. Paul Moore
v4:
- removed unneeded brackets in ocontext_read. Paul Moore
Signed-off-by: Daniel Jurgens <danielj@mellanox.com>
---
security/selinux/include/security.h | 3 +-
security/selinux/ss/policydb.c | 129 +++++++++++++++++++++++++++++++-----
security/selinux/ss/policydb.h | 27 +++++---
3 files changed, 135 insertions(+), 24 deletions(-)
diff --git a/security/selinux/include/security.h b/security/selinux/include/security.h
index 308a286..6bb9b0a 100644
--- a/security/selinux/include/security.h
+++ b/security/selinux/include/security.h
@@ -36,10 +36,11 @@
#define POLICYDB_VERSION_DEFAULT_TYPE 28
#define POLICYDB_VERSION_CONSTRAINT_NAMES 29
#define POLICYDB_VERSION_XPERMS_IOCTL 30
+#define POLICYDB_VERSION_INFINIBAND 31
/* Range of policy versions we understand*/
#define POLICYDB_VERSION_MIN POLICYDB_VERSION_BASE
-#define POLICYDB_VERSION_MAX POLICYDB_VERSION_XPERMS_IOCTL
+#define POLICYDB_VERSION_MAX POLICYDB_VERSION_INFINIBAND
/* Mask for just the mount related flags */
#define SE_MNTMASK 0x0f
diff --git a/security/selinux/ss/policydb.c b/security/selinux/ss/policydb.c
index d719db4..24e16da 100644
--- a/security/selinux/ss/policydb.c
+++ b/security/selinux/ss/policydb.c
@@ -17,6 +17,11 @@
*
* Added support for the policy capability bitmap
*
+ * Update: Mellanox Techonologies
+ *
+ * Added Infiniband support
+ *
+ * Copyright (C) 2016 Mellanox Techonologies
* Copyright (C) 2007 Hewlett-Packard Development Company, L.P.
* Copyright (C) 2004-2005 Trusted Computer Solutions, Inc.
* Copyright (C) 2003 - 2004 Tresys Technology, LLC
@@ -76,81 +81,86 @@ struct policydb_compat_info {
{
.version = POLICYDB_VERSION_BASE,
.sym_num = SYM_NUM - 3,
- .ocon_num = OCON_NUM - 1,
+ .ocon_num = OCON_NUM - 3,
},
{
.version = POLICYDB_VERSION_BOOL,
.sym_num = SYM_NUM - 2,
- .ocon_num = OCON_NUM - 1,
+ .ocon_num = OCON_NUM - 3,
},
{
.version = POLICYDB_VERSION_IPV6,
.sym_num = SYM_NUM - 2,
- .ocon_num = OCON_NUM,
+ .ocon_num = OCON_NUM - 2,
},
{
.version = POLICYDB_VERSION_NLCLASS,
.sym_num = SYM_NUM - 2,
- .ocon_num = OCON_NUM,
+ .ocon_num = OCON_NUM - 2,
},
{
.version = POLICYDB_VERSION_MLS,
.sym_num = SYM_NUM,
- .ocon_num = OCON_NUM,
+ .ocon_num = OCON_NUM - 2,
},
{
.version = POLICYDB_VERSION_AVTAB,
.sym_num = SYM_NUM,
- .ocon_num = OCON_NUM,
+ .ocon_num = OCON_NUM - 2,
},
{
.version = POLICYDB_VERSION_RANGETRANS,
.sym_num = SYM_NUM,
- .ocon_num = OCON_NUM,
+ .ocon_num = OCON_NUM - 2,
},
{
.version = POLICYDB_VERSION_POLCAP,
.sym_num = SYM_NUM,
- .ocon_num = OCON_NUM,
+ .ocon_num = OCON_NUM - 2,
},
{
.version = POLICYDB_VERSION_PERMISSIVE,
.sym_num = SYM_NUM,
- .ocon_num = OCON_NUM,
+ .ocon_num = OCON_NUM - 2,
},
{
.version = POLICYDB_VERSION_BOUNDARY,
.sym_num = SYM_NUM,
- .ocon_num = OCON_NUM,
+ .ocon_num = OCON_NUM - 2,
},
{
.version = POLICYDB_VERSION_FILENAME_TRANS,
.sym_num = SYM_NUM,
- .ocon_num = OCON_NUM,
+ .ocon_num = OCON_NUM - 2,
},
{
.version = POLICYDB_VERSION_ROLETRANS,
.sym_num = SYM_NUM,
- .ocon_num = OCON_NUM,
+ .ocon_num = OCON_NUM - 2,
},
{
.version = POLICYDB_VERSION_NEW_OBJECT_DEFAULTS,
.sym_num = SYM_NUM,
- .ocon_num = OCON_NUM,
+ .ocon_num = OCON_NUM - 2,
},
{
.version = POLICYDB_VERSION_DEFAULT_TYPE,
.sym_num = SYM_NUM,
- .ocon_num = OCON_NUM,
+ .ocon_num = OCON_NUM - 2,
},
{
.version = POLICYDB_VERSION_CONSTRAINT_NAMES,
.sym_num = SYM_NUM,
- .ocon_num = OCON_NUM,
+ .ocon_num = OCON_NUM - 2,
},
{
.version = POLICYDB_VERSION_XPERMS_IOCTL,
.sym_num = SYM_NUM,
+ .ocon_num = OCON_NUM - 2,
+ },
+ {
+ .version = POLICYDB_VERSION_INFINIBAND,
+ .sym_num = SYM_NUM,
.ocon_num = OCON_NUM,
},
};
@@ -2222,6 +2232,60 @@ static int ocontext_read(struct policydb *p, struct policydb_compat_info *info,
goto out;
break;
}
+ case OCON_PKEY:
+ rc = next_entry(nodebuf, fp, sizeof(u32) * 6);
+ if (rc)
+ goto out;
+
+ c->u.pkey.subnet_prefix = be64_to_cpu(*((__be64 *)nodebuf));
+ /* The subnet prefix is stored as an IPv6
+ * address in the policy.
+ *
+ * Check that the lower 2 DWORDS are 0.
+ */
+ if (nodebuf[2] || nodebuf[3]) {
+ rc = -EINVAL;
+ goto out;
+ }
+
+ if (nodebuf[4] > 0xffff ||
+ nodebuf[5] > 0xffff) {
+ rc = -EINVAL;
+ goto out;
+ }
+
+ c->u.pkey.low_pkey = le32_to_cpu(nodebuf[4]);
+ c->u.pkey.high_pkey = le32_to_cpu(nodebuf[5]);
+
+ rc = context_read_and_validate(&c->context[0],
+ p,
+ fp);
+ if (rc)
+ goto out;
+ break;
+ case OCON_IB_ENDPORT:
+ rc = next_entry(buf, fp, sizeof(u32) * 2);
+ if (rc)
+ goto out;
+ len = le32_to_cpu(buf[0]);
+
+ rc = str_read(&c->u.ib_endport.dev_name, GFP_KERNEL, fp, len);
+ if (rc)
+ goto out;
+
+ if (buf[1] > 0xff || buf[1] == 0) {
+ rc = -EINVAL;
+ goto out;
+ }
+
+ c->u.ib_endport.port_num = le32_to_cpu(buf[1]);
+
+ rc = context_read_and_validate(&c->context[0],
+ p,
+ fp);
+ if (rc)
+ goto out;
+ break;
}
}
}
@@ -3151,6 +3215,41 @@ static int ocontext_write(struct policydb *p, struct policydb_compat_info *info,
if (rc)
return rc;
break;
+ case OCON_PKEY:
+ *((__be64 *)nodebuf) = cpu_to_be64(c->u.pkey.subnet_prefix);
+
+ /*
+ * The low order 2 bits were confirmed to be 0
+ * when the policy was loaded. Write them out
+ * as zero
+ */
+ nodebuf[2] = 0;
+ nodebuf[3] = 0;
+
+ nodebuf[4] = cpu_to_le32(c->u.pkey.low_pkey);
+ nodebuf[5] = cpu_to_le32(c->u.pkey.high_pkey);
+
+ rc = put_entry(nodebuf, sizeof(u32), 6, fp);
+ if (rc)
+ return rc;
+ rc = context_write(p, &c->context[0], fp);
+ if (rc)
+ return rc;
+ break;
+ case OCON_IB_ENDPORT:
+ len = strlen(c->u.ib_endport.dev_name);
+ buf[0] = cpu_to_le32(len);
+ buf[1] = cpu_to_le32(c->u.ib_endport.port_num);
+ rc = put_entry(buf, sizeof(u32), 2, fp);
+ if (rc)
+ return rc;
+ rc = put_entry(c->u.ib_endport.dev_name, 1, len, fp);
+ if (rc)
+ return rc;
+ rc = context_write(p, &c->context[0], fp);
+ if (rc)
+ return rc;
+ break;
}
}
}
diff --git a/security/selinux/ss/policydb.h b/security/selinux/ss/policydb.h
index 725d594..edb329d 100644
--- a/security/selinux/ss/policydb.h
+++ b/security/selinux/ss/policydb.h
@@ -187,6 +187,15 @@ struct ocontext {
u32 addr[4];
u32 mask[4];
} node6; /* IPv6 node information */
+ struct {
+ u64 subnet_prefix;
+ u16 low_pkey;
+ u16 high_pkey;
+ } pkey;
+ struct {
+ char *dev_name;
+ u8 port_num;
+ } ib_endport;
} u;
union {
u32 sclass; /* security class for genfs */
@@ -215,14 +224,16 @@ struct genfs {
#define SYM_NUM 8
/* object context array indices */
-#define OCON_ISID 0 /* initial SIDs */
-#define OCON_FS 1 /* unlabeled file systems */
-#define OCON_PORT 2 /* TCP and UDP port numbers */
-#define OCON_NETIF 3 /* network interfaces */
-#define OCON_NODE 4 /* nodes */
-#define OCON_FSUSE 5 /* fs_use */
-#define OCON_NODE6 6 /* IPv6 nodes */
-#define OCON_NUM 7
+#define OCON_ISID 0 /* initial SIDs */
+#define OCON_FS 1 /* unlabeled file systems */
+#define OCON_PORT 2 /* TCP and UDP port numbers */
+#define OCON_NETIF 3 /* network interfaces */
+#define OCON_NODE 4 /* nodes */
+#define OCON_FSUSE 5 /* fs_use */
+#define OCON_NODE6 6 /* IPv6 nodes */
+#define OCON_PKEY 7 /* Infiniband PKeys */
+#define OCON_IB_ENDPORT 8 /* Infiniband end ports */
+#define OCON_NUM 9
/* The policy database */
struct policydb {
--
1.8.3.1
^ permalink raw reply related
* [PATCH v4 4/9] IB/core: Enforce security on management datagrams
From: Dan Jurgens @ 2016-11-08 21:06 UTC (permalink / raw)
To: chrisw-69jw2NvuJkxg9hUCZPvPmw, paul-r2n+y4ga6xFZroRs9YW3xA,
sds-+05T5uksL2qpZYMLLGbcSA, eparis-FjpueFixGhCM4zKIHC2jIg,
dledford-H+wXaHxf7aLQT0dZR+AlfA,
sean.hefty-ral2JQCrhuEAvxtiuMwx3w,
hal.rosenstock-Re5JQEeQqe8AvxtiuMwx3w
Cc: selinux-+05T5uksL2qpZYMLLGbcSA,
linux-security-module-u79uwXL29TY76Z2rM5mHXA,
linux-rdma-u79uwXL29TY76Z2rM5mHXA,
yevgenyp-VPRAkNaXOzVWk0Htik3J/w, liranl-VPRAkNaXOzVWk0Htik3J/w,
leonro-VPRAkNaXOzVWk0Htik3J/w, Daniel Jurgens
In-Reply-To: <1478639185-47521-1-git-send-email-danielj-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
From: Daniel Jurgens <danielj-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
Allocate and free a security context when creating and destroying a MAD
agent. This context is used for controlling access to PKeys and sending
and receiving SMPs.
When sending or receiving a MAD check that the agent has permission to
access the PKey for the Subnet Prefix of the port.
During MAD and snoop agent registration for SMI QPs check that the
calling process has permission to access the manage the subnet and
register a callback with the LSM to be notified of policy changes. When
notificaiton of a policy change occurs recheck permission and set a flag
indicating sending and receiving SMPs is allowed.
When sending and receiving MADs check that the agent has access to the
SMI if it's on an SMI QP. Because security policy can change it's
possible permission was allowed when creating the agent, but no longer
is.
issue: 736423
Change-Id: I17343224da7ffffe7ee0cc67dadd44c473e453d0
Signed-off-by: Daniel Jurgens <danielj-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
---
v2:
- Squashed LSM hook additions. Paul Moore
- Changed security blobs to void*. Paul Moore
- Shorten end_port to port. Paul Moore
- Change "smp" to "manage_subnet". Paul Moore
- Use the LSM policy change notification and a flag to track permission
instead of calling the LSM hook for every SMP. Dan Jurgens
- Squashed PKey and SMP enforcement into the same patch and moved the
logic into security.c. Dan Jurgens
v3:
- ib_port -> ib_endport. Paul Moore
- Use notifier chains for LSM notification. Paul Moore
- Reorder LSM hook parameters to put sec first. Paul Moore
Signed-off-by: Daniel Jurgens <danielj-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
---
drivers/infiniband/core/core_priv.h | 35 ++++++++++++++
drivers/infiniband/core/mad.c | 52 +++++++++++++++++----
drivers/infiniband/core/security.c | 92 +++++++++++++++++++++++++++++++++++++
include/linux/lsm_hooks.h | 8 ++++
include/linux/security.h | 6 +++
include/rdma/ib_mad.h | 4 ++
security/security.c | 8 ++++
7 files changed, 197 insertions(+), 8 deletions(-)
diff --git a/drivers/infiniband/core/core_priv.h b/drivers/infiniband/core/core_priv.h
index 68e3de0..2c35162 100644
--- a/drivers/infiniband/core/core_priv.h
+++ b/drivers/infiniband/core/core_priv.h
@@ -37,6 +37,8 @@
#include <linux/spinlock.h>
#include <rdma/ib_verbs.h>
+#include <rdma/ib_mad.h>
+#include "mad_priv.h"
struct pkey_index_qp_list {
struct list_head pkey_index_list;
@@ -166,6 +168,11 @@ int ib_get_cached_subnet_prefix(struct ib_device *device,
u64 *sn_pfx);
#ifdef CONFIG_SECURITY_INFINIBAND
+int ib_security_pkey_access(struct ib_device *dev,
+ u8 port_num,
+ u16 pkey_index,
+ void *sec);
+
void ib_security_destroy_port_pkey_list(struct ib_device *device);
void ib_security_cache_change(struct ib_device *device,
@@ -183,7 +190,19 @@ int ib_security_modify_qp(struct ib_qp *qp,
void ib_destroy_qp_security_end(struct ib_qp_security *sec);
int ib_open_shared_qp_security(struct ib_qp *qp, struct ib_device *dev);
void ib_close_shared_qp_security(struct ib_qp_security *sec);
+int ib_mad_agent_security_setup(struct ib_mad_agent *agent,
+ enum ib_qp_type qp_type);
+void ib_mad_agent_security_cleanup(struct ib_mad_agent *agent);
+int ib_mad_enforce_security(struct ib_mad_agent_private *map, u16 pkey_index);
#else
+static inline int ib_security_pkey_access(struct ib_device *dev,
+ u8 port_num,
+ u16 pkey_index,
+ void *sec)
+{
+ return 0;
+}
+
static inline void ib_security_destroy_port_pkey_list(struct ib_device *device)
{
}
@@ -232,5 +251,21 @@ static inline int ib_open_shared_qp_security(struct ib_qp *qp,
static inline void ib_close_shared_qp_security(struct ib_qp_security *sec)
{
}
+
+static inline int ib_mad_agent_security_setup(struct ib_mad_agent *agent,
+ enum ib_qp_type qp_type)
+{
+ return 0;
+}
+
+static inline void ib_mad_agent_security_cleanup(struct ib_mad_agent *agent)
+{
+}
+
+static inline int ib_mad_enforce_security(struct ib_mad_agent_private *map,
+ u16 pkey_index)
+{
+ return 0;
+}
#endif
#endif /* _CORE_PRIV_H */
diff --git a/drivers/infiniband/core/mad.c b/drivers/infiniband/core/mad.c
index 40cbd6b..b6041ec 100644
--- a/drivers/infiniband/core/mad.c
+++ b/drivers/infiniband/core/mad.c
@@ -40,9 +40,11 @@
#include <linux/dma-mapping.h>
#include <linux/slab.h>
#include <linux/module.h>
+#include <linux/security.h>
#include <rdma/ib_cache.h>
#include "mad_priv.h"
+#include "core_priv.h"
#include "mad_rmpp.h"
#include "smi.h"
#include "opa_smi.h"
@@ -367,6 +369,12 @@ struct ib_mad_agent *ib_register_mad_agent(struct ib_device *device,
atomic_set(&mad_agent_priv->refcount, 1);
init_completion(&mad_agent_priv->comp);
+ ret2 = ib_mad_agent_security_setup(&mad_agent_priv->agent, qp_type);
+ if (ret2) {
+ ret = ERR_PTR(ret2);
+ goto error4;
+ }
+
spin_lock_irqsave(&port_priv->reg_lock, flags);
mad_agent_priv->agent.hi_tid = ++ib_mad_client_id;
@@ -384,7 +392,7 @@ struct ib_mad_agent *ib_register_mad_agent(struct ib_device *device,
if (method) {
if (method_in_use(&method,
mad_reg_req))
- goto error4;
+ goto error5;
}
}
ret2 = add_nonoui_reg_req(mad_reg_req, mad_agent_priv,
@@ -400,14 +408,14 @@ struct ib_mad_agent *ib_register_mad_agent(struct ib_device *device,
if (is_vendor_method_in_use(
vendor_class,
mad_reg_req))
- goto error4;
+ goto error5;
}
}
ret2 = add_oui_reg_req(mad_reg_req, mad_agent_priv);
}
if (ret2) {
ret = ERR_PTR(ret2);
- goto error4;
+ goto error5;
}
}
@@ -416,9 +424,10 @@ struct ib_mad_agent *ib_register_mad_agent(struct ib_device *device,
spin_unlock_irqrestore(&port_priv->reg_lock, flags);
return &mad_agent_priv->agent;
-
-error4:
+error5:
spin_unlock_irqrestore(&port_priv->reg_lock, flags);
+ ib_mad_agent_security_cleanup(&mad_agent_priv->agent);
+error4:
kfree(reg_req);
error3:
kfree(mad_agent_priv);
@@ -489,6 +498,7 @@ struct ib_mad_agent *ib_register_mad_snoop(struct ib_device *device,
struct ib_mad_agent *ret;
struct ib_mad_snoop_private *mad_snoop_priv;
int qpn;
+ int err;
/* Validate parameters */
if ((is_snooping_sends(mad_snoop_flags) && !snoop_handler) ||
@@ -523,17 +533,25 @@ struct ib_mad_agent *ib_register_mad_snoop(struct ib_device *device,
mad_snoop_priv->agent.port_num = port_num;
mad_snoop_priv->mad_snoop_flags = mad_snoop_flags;
init_completion(&mad_snoop_priv->comp);
+
+ err = ib_mad_agent_security_setup(&mad_snoop_priv->agent, qp_type);
+ if (err) {
+ ret = ERR_PTR(err);
+ goto error2;
+ }
+
mad_snoop_priv->snoop_index = register_snoop_agent(
&port_priv->qp_info[qpn],
mad_snoop_priv);
if (mad_snoop_priv->snoop_index < 0) {
ret = ERR_PTR(mad_snoop_priv->snoop_index);
- goto error2;
+ goto error3;
}
atomic_set(&mad_snoop_priv->refcount, 1);
return &mad_snoop_priv->agent;
-
+error3:
+ ib_mad_agent_security_cleanup(&mad_snoop_priv->agent);
error2:
kfree(mad_snoop_priv);
error1:
@@ -579,6 +597,8 @@ static void unregister_mad_agent(struct ib_mad_agent_private *mad_agent_priv)
deref_mad_agent(mad_agent_priv);
wait_for_completion(&mad_agent_priv->comp);
+ ib_mad_agent_security_cleanup(&mad_agent_priv->agent);
+
kfree(mad_agent_priv->reg_req);
kfree(mad_agent_priv);
}
@@ -597,6 +617,8 @@ static void unregister_mad_snoop(struct ib_mad_snoop_private *mad_snoop_priv)
deref_snoop_agent(mad_snoop_priv);
wait_for_completion(&mad_snoop_priv->comp);
+ ib_mad_agent_security_cleanup(&mad_snoop_priv->agent);
+
kfree(mad_snoop_priv);
}
@@ -1219,12 +1241,16 @@ int ib_post_send_mad(struct ib_mad_send_buf *send_buf,
/* Walk list of send WRs and post each on send list */
for (; send_buf; send_buf = next_send_buf) {
-
mad_send_wr = container_of(send_buf,
struct ib_mad_send_wr_private,
send_buf);
mad_agent_priv = mad_send_wr->mad_agent_priv;
+ ret = ib_mad_enforce_security(mad_agent_priv,
+ mad_send_wr->send_wr.pkey_index);
+ if (ret)
+ goto error;
+
if (!send_buf->mad_agent->send_handler ||
(send_buf->timeout_ms &&
!send_buf->mad_agent->recv_handler)) {
@@ -1958,6 +1984,14 @@ static void ib_mad_complete_recv(struct ib_mad_agent_private *mad_agent_priv,
struct ib_mad_send_wr_private *mad_send_wr;
struct ib_mad_send_wc mad_send_wc;
unsigned long flags;
+ int ret;
+
+ ret = ib_mad_enforce_security(mad_agent_priv,
+ mad_recv_wc->wc->pkey_index);
+ if (ret) {
+ ib_free_recv_mad(mad_recv_wc);
+ deref_mad_agent(mad_agent_priv);
+ }
INIT_LIST_HEAD(&mad_recv_wc->rmpp_list);
list_add(&mad_recv_wc->recv_buf.list, &mad_recv_wc->rmpp_list);
@@ -2015,6 +2049,8 @@ static void ib_mad_complete_recv(struct ib_mad_agent_private *mad_agent_priv,
mad_recv_wc);
deref_mad_agent(mad_agent_priv);
}
+
+ return;
}
static enum smi_action handle_ib_smi(const struct ib_mad_port_private *port_priv,
diff --git a/drivers/infiniband/core/security.c b/drivers/infiniband/core/security.c
index 45800dd..bc09af5 100644
--- a/drivers/infiniband/core/security.c
+++ b/drivers/infiniband/core/security.c
@@ -39,6 +39,7 @@
#include <rdma/ib_verbs.h>
#include <rdma/ib_cache.h>
#include "core_priv.h"
+#include "mad_priv.h"
static struct pkey_index_qp_list *get_pkey_idx_qp_list(struct ib_port_pkey *pp)
{
@@ -614,4 +615,95 @@ int ib_security_modify_qp(struct ib_qp *qp,
}
EXPORT_SYMBOL(ib_security_modify_qp);
+int ib_security_pkey_access(struct ib_device *dev,
+ u8 port_num,
+ u16 pkey_index,
+ void *sec)
+{
+ u64 subnet_prefix;
+ u16 pkey;
+ int ret;
+
+ ret = ib_get_cached_pkey(dev, port_num, pkey_index, &pkey);
+ if (ret)
+ return ret;
+
+ ret = ib_get_cached_subnet_prefix(dev, port_num, &subnet_prefix);
+
+ if (ret)
+ return ret;
+
+ return security_ib_pkey_access(sec, subnet_prefix, pkey);
+}
+EXPORT_SYMBOL(ib_security_pkey_access);
+
+static int ib_mad_agent_security_change(struct notifier_block *nb,
+ unsigned long event,
+ void *data)
+{
+ struct ib_mad_agent *ag = container_of(nb, struct ib_mad_agent, lsm_nb);
+
+ if (event != LSM_POLICY_CHANGE)
+ return NOTIFY_DONE;
+
+ ag->smp_allowed = !security_ib_endport_manage_subnet(ag->security,
+ ag->device->name,
+ ag->port_num);
+
+ return NOTIFY_OK;
+}
+
+int ib_mad_agent_security_setup(struct ib_mad_agent *agent,
+ enum ib_qp_type qp_type)
+{
+ int ret;
+
+ ret = security_ib_alloc_security(&agent->security);
+ if (ret)
+ return ret;
+
+ if (qp_type != IB_QPT_SMI)
+ return 0;
+
+ ret = security_ib_endport_manage_subnet(agent->security,
+ agent->device->name,
+ agent->port_num);
+ if (ret)
+ return ret;
+
+ agent->lsm_nb.notifier_call = ib_mad_agent_security_change;
+ ret = register_lsm_notifier(&agent->lsm_nb);
+ if (ret)
+ return ret;
+
+ agent->smp_allowed = true;
+ agent->lsm_nb_reg = true;
+ return 0;
+}
+
+void ib_mad_agent_security_cleanup(struct ib_mad_agent *agent)
+{
+ security_ib_free_security(agent->security);
+ if (agent->lsm_nb_reg)
+ unregister_lsm_notifier(&agent->lsm_nb);
+}
+
+int ib_mad_enforce_security(struct ib_mad_agent_private *map, u16 pkey_index)
+{
+ int ret;
+
+ if (map->agent.qp->qp_type == IB_QPT_SMI && !map->agent.smp_allowed)
+ return -EACCES;
+
+ ret = ib_security_pkey_access(map->agent.device,
+ map->agent.port_num,
+ pkey_index,
+ map->agent.security);
+
+ if (ret)
+ return ret;
+
+ return 0;
+}
+
#endif /* CONFIG_SECURITY_INFINIBAND */
diff --git a/include/linux/lsm_hooks.h b/include/linux/lsm_hooks.h
index d576efc..3936216 100644
--- a/include/linux/lsm_hooks.h
+++ b/include/linux/lsm_hooks.h
@@ -911,6 +911,11 @@
* @subnet_prefix the subnet prefix of the port being used.
* @pkey the pkey to be accessed.
* @sec pointer to a security structure.
+ * @ib_endport_manage_subnet:
+ * Check permissions to send and receive SMPs on a end port.
+ * @dev_name the IB device name (i.e. mlx4_0).
+ * @port_num the port number.
+ * @sec pointer to a security structure.
* @ib_alloc_security:
* Allocate a security structure for Infiniband objects.
* @sec pointer to a security structure pointer.
@@ -1629,6 +1634,8 @@
#ifdef CONFIG_SECURITY_INFINIBAND
int (*ib_pkey_access)(void *sec, u64 subnet_prefix, u16 pkey);
+ int (*ib_endport_manage_subnet)(void *sec, const char *dev_name,
+ u8 port_num);
int (*ib_alloc_security)(void **sec);
void (*ib_free_security)(void *sec);
#endif /* CONFIG_SECURITY_INFINIBAND */
@@ -1865,6 +1872,7 @@ struct security_hook_heads {
#endif /* CONFIG_SECURITY_NETWORK */
#ifdef CONFIG_SECURITY_INFINIBAND
struct list_head ib_pkey_access;
+ struct list_head ib_endport_manage_subnet;
struct list_head ib_alloc_security;
struct list_head ib_free_security;
#endif /* CONFIG_SECURITY_INFINIBAND */
diff --git a/include/linux/security.h b/include/linux/security.h
index d8c29cd..9241f86 100644
--- a/include/linux/security.h
+++ b/include/linux/security.h
@@ -1404,6 +1404,7 @@ static inline int security_tun_dev_open(void *security)
#ifdef CONFIG_SECURITY_INFINIBAND
int security_ib_pkey_access(void *sec, u64 subnet_prefix, u16 pkey);
+int security_ib_endport_manage_subnet(void *sec, const char *name, u8 port_num);
int security_ib_alloc_security(void **sec);
void security_ib_free_security(void *sec);
#else /* CONFIG_SECURITY_INFINIBAND */
@@ -1412,6 +1413,11 @@ static inline int security_ib_pkey_access(void *sec, u64 subnet_prefix, u16 pkey
return 0;
}
+static inline int security_ib_endport_manage_subnet(void *sec, const char *dev_name, u8 port_num)
+{
+ return 0;
+}
+
static inline int security_ib_alloc_security(void **sec)
{
return 0;
diff --git a/include/rdma/ib_mad.h b/include/rdma/ib_mad.h
index c8a773f..23c4a4e 100644
--- a/include/rdma/ib_mad.h
+++ b/include/rdma/ib_mad.h
@@ -537,6 +537,10 @@ struct ib_mad_agent {
u32 flags;
u8 port_num;
u8 rmpp_version;
+ void *security;
+ bool smp_allowed;
+ bool lsm_nb_reg;
+ struct notifier_block lsm_nb;
};
/**
diff --git a/security/security.c b/security/security.c
index 40326d4..e4fd2b6 100644
--- a/security/security.c
+++ b/security/security.c
@@ -1470,6 +1470,12 @@ int security_ib_pkey_access(void *sec, u64 subnet_prefix, u16 pkey)
}
EXPORT_SYMBOL(security_ib_pkey_access);
+int security_ib_endport_manage_subnet(void *sec, const char *dev_name, u8 port_num)
+{
+ return call_int_hook(ib_endport_manage_subnet, 0, sec, dev_name, port_num);
+}
+EXPORT_SYMBOL(security_ib_endport_manage_subnet);
+
int security_ib_alloc_security(void **sec)
{
return call_int_hook(ib_alloc_security, 0, sec);
@@ -1943,6 +1949,8 @@ struct security_hook_heads security_hook_heads = {
#ifdef CONFIG_SECURITY_INFINIBAND
.ib_pkey_access = LIST_HEAD_INIT(security_hook_heads.ib_pkey_access),
+ .ib_endport_manage_subnet =
+ LIST_HEAD_INIT(security_hook_heads.ib_endport_manage_subnet),
.ib_alloc_security =
LIST_HEAD_INIT(security_hook_heads.ib_alloc_security),
.ib_free_security =
--
1.8.3.1
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply related
* [PATCH v4 3/9] selinux lsm IB/core: Implement LSM notification system
From: Dan Jurgens @ 2016-11-08 21:06 UTC (permalink / raw)
To: chrisw-69jw2NvuJkxg9hUCZPvPmw, paul-r2n+y4ga6xFZroRs9YW3xA,
sds-+05T5uksL2qpZYMLLGbcSA, eparis-FjpueFixGhCM4zKIHC2jIg,
dledford-H+wXaHxf7aLQT0dZR+AlfA,
sean.hefty-ral2JQCrhuEAvxtiuMwx3w,
hal.rosenstock-Re5JQEeQqe8AvxtiuMwx3w
Cc: selinux-+05T5uksL2qpZYMLLGbcSA,
linux-security-module-u79uwXL29TY76Z2rM5mHXA,
linux-rdma-u79uwXL29TY76Z2rM5mHXA,
yevgenyp-VPRAkNaXOzVWk0Htik3J/w, liranl-VPRAkNaXOzVWk0Htik3J/w,
leonro-VPRAkNaXOzVWk0Htik3J/w, Daniel Jurgens
In-Reply-To: <1478639185-47521-1-git-send-email-danielj-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
From: Daniel Jurgens <danielj-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
Add a generic notificaiton mechanism in the LSM. Interested consumers
can register a callback with the LSM and security modules can produce
events.
Because access to Infiniband QPs are enforced in the setup phase of a
connection security should be enforced again if the policy changes.
Register infiniband devices for policy change notification and check all
QPs on that device when the notification is received.
Add a call to the notification mechanism from SELinux when the AVC
cache changes or setenforce is cleared.
issue: 736423
Change-Id: I9f5a13026c9219105ee5e0cd99edf8eabdafe946
Signed-off-by: Daniel Jurgens <danielj-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
---
v2:
- new patch that has the generic notification, replaces selinux and
IB/core patches related to the ib_flush callback. Yuval Shaia and Paul
Moore
v3:
- use notifier chains. Paul Moore
v4:
- Seperate avc callback for LSM notifier. Paul Moore
Signed-off-by: Daniel Jurgens <danielj-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
---
drivers/infiniband/core/device.c | 53 ++++++++++++++++++++++++++++++++++++++++
include/linux/security.h | 8 ++++++
security/security.c | 20 +++++++++++++++
security/selinux/hooks.c | 11 +++++++++
security/selinux/selinuxfs.c | 2 ++
5 files changed, 94 insertions(+)
diff --git a/drivers/infiniband/core/device.c b/drivers/infiniband/core/device.c
index 5b42e83..7b6fd06 100644
--- a/drivers/infiniband/core/device.c
+++ b/drivers/infiniband/core/device.c
@@ -39,6 +39,8 @@
#include <linux/init.h>
#include <linux/mutex.h>
#include <linux/netdevice.h>
+#include <linux/security.h>
+#include <linux/notifier.h>
#include <rdma/rdma_netlink.h>
#include <rdma/ib_addr.h>
#include <rdma/ib_cache.h>
@@ -82,6 +84,14 @@ struct ib_client_data {
static DEFINE_MUTEX(device_mutex);
static DECLARE_RWSEM(lists_rwsem);
+static int ib_security_change(struct notifier_block *nb, unsigned long event,
+ void *lsm_data);
+static void ib_policy_change_task(struct work_struct *work);
+static DECLARE_WORK(ib_policy_change_work, ib_policy_change_task);
+
+static struct notifier_block ibdev_lsm_nb = {
+ .notifier_call = ib_security_change,
+};
static int ib_device_check_mandatory(struct ib_device *device)
{
@@ -344,6 +354,40 @@ static int setup_port_pkey_list(struct ib_device *device)
return 0;
}
+static void ib_policy_change_task(struct work_struct *work)
+{
+ struct ib_device *dev;
+
+ down_read(&lists_rwsem);
+ list_for_each_entry(dev, &device_list, core_list) {
+ int i;
+
+ for (i = rdma_start_port(dev); i <= rdma_end_port(dev); i++) {
+ u64 sp;
+ int ret = ib_get_cached_subnet_prefix(dev,
+ i,
+ &sp);
+
+ WARN_ONCE(ret,
+ "ib_get_cached_subnet_prefix err: %d, this should never happen here\n",
+ ret);
+ ib_security_cache_change(dev, i, sp);
+ }
+ }
+ up_read(&lists_rwsem);
+}
+
+static int ib_security_change(struct notifier_block *nb, unsigned long event,
+ void *lsm_data)
+{
+ if (event != LSM_POLICY_CHANGE)
+ return NOTIFY_DONE;
+
+ schedule_work(&ib_policy_change_work);
+
+ return NOTIFY_OK;
+}
+
/**
* ib_register_device - Register an IB device with IB core
* @device:Device to register
@@ -1075,10 +1119,18 @@ static int __init ib_core_init(void)
goto err_sa;
}
+ ret = register_lsm_notifier(&ibdev_lsm_nb);
+ if (ret) {
+ pr_warn("Couldn't register LSM notifier. ret %d\n", ret);
+ goto err_ibnl_clients;
+ }
+
ib_cache_setup();
return 0;
+err_ibnl_clients:
+ ib_remove_ibnl_clients();
err_sa:
ib_sa_cleanup();
err_mad:
@@ -1098,6 +1150,7 @@ static int __init ib_core_init(void)
static void __exit ib_core_cleanup(void)
{
+ unregister_lsm_notifier(&ibdev_lsm_nb);
ib_cache_cleanup();
ib_remove_ibnl_clients();
ib_sa_cleanup();
diff --git a/include/linux/security.h b/include/linux/security.h
index 342ca4c..d8c29cd 100644
--- a/include/linux/security.h
+++ b/include/linux/security.h
@@ -69,6 +69,14 @@
struct user_namespace;
struct timezone;
+enum lsm_event {
+ LSM_POLICY_CHANGE,
+};
+
+int call_lsm_notifier(enum lsm_event event, void *data);
+int register_lsm_notifier(struct notifier_block *nb);
+int unregister_lsm_notifier(struct notifier_block *nb);
+
/* These functions are in security/commoncap.c */
extern int cap_capable(const struct cred *cred, struct user_namespace *ns,
int cap, int audit);
diff --git a/security/security.c b/security/security.c
index 7d3bf2f..40326d4 100644
--- a/security/security.c
+++ b/security/security.c
@@ -33,6 +33,8 @@
/* Maximum number of letters for an LSM name string */
#define SECURITY_NAME_MAX 10
+static ATOMIC_NOTIFIER_HEAD(lsm_notifier_chain);
+
/* Boot-time LSM user choice */
static __initdata char chosen_lsm[SECURITY_NAME_MAX + 1] =
CONFIG_DEFAULT_SECURITY;
@@ -98,6 +100,24 @@ int __init security_module_enable(const char *module)
return !strcmp(module, chosen_lsm);
}
+int call_lsm_notifier(enum lsm_event event, void *data)
+{
+ return atomic_notifier_call_chain(&lsm_notifier_chain, event, data);
+}
+EXPORT_SYMBOL(call_lsm_notifier);
+
+int register_lsm_notifier(struct notifier_block *nb)
+{
+ return atomic_notifier_chain_register(&lsm_notifier_chain, nb);
+}
+EXPORT_SYMBOL(register_lsm_notifier);
+
+int unregister_lsm_notifier(struct notifier_block *nb)
+{
+ return atomic_notifier_chain_unregister(&lsm_notifier_chain, nb);
+}
+EXPORT_SYMBOL(unregister_lsm_notifier);
+
/*
* Hook list operation macros.
*
diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c
index 09fd610..7578198 100644
--- a/security/selinux/hooks.c
+++ b/security/selinux/hooks.c
@@ -170,6 +170,14 @@ static int selinux_netcache_avc_callback(u32 event)
return 0;
}
+static int selinux_lsm_notifier_avc_callback(u32 event)
+{
+ if (event == AVC_CALLBACK_RESET) {
+ call_lsm_notifier(LSM_POLICY_CHANGE, NULL);
+ }
+ return 0;
+}
+
/*
* initialise the security for the init task
*/
@@ -6325,6 +6333,9 @@ static __init int selinux_init(void)
if (avc_add_callback(selinux_netcache_avc_callback, AVC_CALLBACK_RESET))
panic("SELinux: Unable to register AVC netcache callback\n");
+ if (avc_add_callback(selinux_lsm_notifier_avc_callback, AVC_CALLBACK_RESET))
+ panic("SELinux: Unable to register AVC LSM notifier callback\n");
+
if (selinux_enforcing)
printk(KERN_DEBUG "SELinux: Starting in enforcing mode\n");
else
diff --git a/security/selinux/selinuxfs.c b/security/selinux/selinuxfs.c
index 72c145d..d3f9192 100644
--- a/security/selinux/selinuxfs.c
+++ b/security/selinux/selinuxfs.c
@@ -177,6 +177,8 @@ static ssize_t sel_write_enforce(struct file *file, const char __user *buf,
avc_ss_reset(0);
selnl_notify_setenforce(selinux_enforcing);
selinux_status_update_setenforce(selinux_enforcing);
+ if (!selinux_enforcing)
+ call_lsm_notifier(LSM_POLICY_CHANGE, NULL);
}
length = count;
out:
--
1.8.3.1
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply related
* [PATCH v4 2/9] IB/core: Enforce PKey security on QPs
From: Dan Jurgens @ 2016-11-08 21:06 UTC (permalink / raw)
To: chrisw, paul, sds, eparis, dledford, sean.hefty, hal.rosenstock
Cc: selinux, linux-security-module, linux-rdma, yevgenyp, liranl,
leonro, Daniel Jurgens
In-Reply-To: <1478639185-47521-1-git-send-email-danielj@mellanox.com>
From: Daniel Jurgens <danielj@mellanox.com>
Add new LSM hooks to allocate and free security contexts and check for
permission to access a PKey.
Allocate and free a security context when creating and destroying a QP.
This context is used for controlling access to PKeys.
When a request is made to modify a QP that changes the port, PKey index,
or alternate path, check that the QP has permission for the PKey in the
PKey table index on the subnet prefix of the port. If the QP is shared
make sure all handles to the QP also have access.
Store which port and PKey index a QP is using. After the reset to init
transition the user can modify the port, PKey index and alternate path
independently. So port and PKey settings changes can be a merge of the
previous settings and the new ones.
In order to maintain access control if there are PKey table or subnet
prefix change keep a list of all QPs are using each PKey index on
each port. If a change occurs all QPs using that device and port must
have access enforced for the new cache settings.
These changes add a transaction to the QP modify process. Association
with the old port and PKey index must be maintained if the modify fails,
and must be removed if it succeeds. Association with the new port and
PKey index must be established prior to the modify and removed if the
modify fails.
1. When a QP is modified to a particular Port, PKey index or alternate
path insert that QP into the appropriate lists.
2. Check permission to access the new settings.
3. If step 2 grants access attempt to modify the QP.
4a. If steps 2 and 3 succeed remove any prior associations.
4b. If ether fails remove the new setting associations.
If a PKey table or subnet prefix changes walk the list of QPs and
check that they have permission. If not send the QP to the error state
and raise a fatal error event. If it's a shared QP make sure all the
QPs that share the real_qp have permission as well. If the QP that
owns a security structure is denied access the security structure is
marked as such and the QP is added to an error_list. Once the moving
the QP to error is complete the security structure mark is cleared.
Maintaining the lists correctly turns QP destroy into a transaction.
The hardware driver for the device frees the ib_qp structure, so while
the destroy is in progress the ib_qp pointer in the ib_qp_security
struct is undefined. When the destroy process begins the ib_qp_security
structure is marked as destroying. This prevents any action from being
taken on the QP pointer. After the QP is destroyed successfully it
could still listed on an error_list wait for it to be processed by that
flow before cleaning up the structure.
If the destroy fails the QPs port and PKey settings are reinserted into
the appropriate lists, the destroying flag is cleared, and access control
is enforced, in case there were any cache changes during the destroy
flow.
To keep the security changes isolated a new file is used to hold security
related functionality.
issue: 736423
Change-Id: I0509141a57a8e08c91a9e0c059800251949b18e6
Signed-off-by: Daniel Jurgens <danielj@mellanox.com>
---
v2:
- Squashed LSM hook additions. Paul Moore
- Changed security blobs to void*. Paul Moore
v3:
- Change parameter order of pkey_access hook. Paul Moore
Signed-off-by: Daniel Jurgens <danielj@mellanox.com>
---
drivers/infiniband/core/Makefile | 3 +-
drivers/infiniband/core/cache.c | 21 +-
drivers/infiniband/core/core_priv.h | 77 +++++
drivers/infiniband/core/device.c | 33 ++
drivers/infiniband/core/security.c | 617 +++++++++++++++++++++++++++++++++++
drivers/infiniband/core/uverbs_cmd.c | 20 +-
drivers/infiniband/core/verbs.c | 27 +-
include/linux/lsm_hooks.h | 27 ++
include/linux/security.h | 21 ++
include/rdma/ib_verbs.h | 48 +++
security/Kconfig | 9 +
security/security.c | 31 ++
12 files changed, 925 insertions(+), 9 deletions(-)
create mode 100644 drivers/infiniband/core/security.c
diff --git a/drivers/infiniband/core/Makefile b/drivers/infiniband/core/Makefile
index edaae9f..da4e2c1 100644
--- a/drivers/infiniband/core/Makefile
+++ b/drivers/infiniband/core/Makefile
@@ -10,7 +10,8 @@ obj-$(CONFIG_INFINIBAND_USER_ACCESS) += ib_uverbs.o ib_ucm.o \
ib_core-y := packer.o ud_header.o verbs.o cq.o rw.o sysfs.o \
device.o fmr_pool.o cache.o netlink.o \
roce_gid_mgmt.o mr_pool.o addr.o sa_query.o \
- multicast.o mad.o smi.o agent.o mad_rmpp.o
+ multicast.o mad.o smi.o agent.o mad_rmpp.o \
+ security.o
ib_core-$(CONFIG_INFINIBAND_USER_MEM) += umem.o
ib_core-$(CONFIG_INFINIBAND_ON_DEMAND_PAGING) += umem_odp.o umem_rbtree.o
diff --git a/drivers/infiniband/core/cache.c b/drivers/infiniband/core/cache.c
index affc8ef..48eaeca 100644
--- a/drivers/infiniband/core/cache.c
+++ b/drivers/infiniband/core/cache.c
@@ -53,6 +53,7 @@ struct ib_update_work {
struct work_struct work;
struct ib_device *device;
u8 port_num;
+ bool enforce_security;
};
union ib_gid zgid;
@@ -1046,7 +1047,8 @@ int ib_get_cached_lmc(struct ib_device *device,
EXPORT_SYMBOL(ib_get_cached_lmc);
static void ib_cache_update(struct ib_device *device,
- u8 port)
+ u8 port,
+ bool enforce_security)
{
struct ib_port_attr *tprops = NULL;
struct ib_pkey_cache *pkey_cache = NULL, *old_pkey_cache;
@@ -1134,6 +1136,11 @@ static void ib_cache_update(struct ib_device *device,
tprops->subnet_prefix;
write_unlock_irq(&device->cache.lock);
+ if (enforce_security)
+ ib_security_cache_change(device,
+ port,
+ tprops->subnet_prefix);
+
kfree(gid_cache);
kfree(old_pkey_cache);
kfree(tprops);
@@ -1150,7 +1157,9 @@ static void ib_cache_task(struct work_struct *_work)
struct ib_update_work *work =
container_of(_work, struct ib_update_work, work);
- ib_cache_update(work->device, work->port_num);
+ ib_cache_update(work->device,
+ work->port_num,
+ work->enforce_security);
kfree(work);
}
@@ -1171,6 +1180,12 @@ static void ib_cache_event(struct ib_event_handler *handler,
INIT_WORK(&work->work, ib_cache_task);
work->device = event->device;
work->port_num = event->element.port_num;
+ if (event->event == IB_EVENT_PKEY_CHANGE ||
+ event->event == IB_EVENT_GID_CHANGE)
+ work->enforce_security = true;
+ else
+ work->enforce_security = false;
+
queue_work(ib_wq, &work->work);
}
}
@@ -1211,7 +1226,7 @@ int ib_cache_setup_one(struct ib_device *device)
return err;
for (p = 0; p <= rdma_end_port(device) - rdma_start_port(device); ++p)
- ib_cache_update(device, p + rdma_start_port(device));
+ ib_cache_update(device, p + rdma_start_port(device), true);
INIT_IB_EVENT_HANDLER(&device->cache.event_handler,
device, ib_cache_event);
diff --git a/drivers/infiniband/core/core_priv.h b/drivers/infiniband/core/core_priv.h
index ce826e4..68e3de0 100644
--- a/drivers/infiniband/core/core_priv.h
+++ b/drivers/infiniband/core/core_priv.h
@@ -38,6 +38,14 @@
#include <rdma/ib_verbs.h>
+struct pkey_index_qp_list {
+ struct list_head pkey_index_list;
+ u16 pkey_index;
+ /* Lock to hold while iterating the qp_list. */
+ spinlock_t qp_list_lock;
+ struct list_head qp_list;
+};
+
#if IS_ENABLED(CONFIG_INFINIBAND_ADDR_TRANS_CONFIGFS)
int cma_configfs_init(void);
void cma_configfs_exit(void);
@@ -156,4 +164,73 @@ int ib_nl_handle_ip_res_resp(struct sk_buff *skb,
int ib_get_cached_subnet_prefix(struct ib_device *device,
u8 port_num,
u64 *sn_pfx);
+
+#ifdef CONFIG_SECURITY_INFINIBAND
+void ib_security_destroy_port_pkey_list(struct ib_device *device);
+
+void ib_security_cache_change(struct ib_device *device,
+ u8 port_num,
+ u64 subnet_prefix);
+
+int ib_security_modify_qp(struct ib_qp *qp,
+ struct ib_qp_attr *qp_attr,
+ int qp_attr_mask,
+ struct ib_udata *udata);
+
+int ib_create_qp_security(struct ib_qp *qp, struct ib_device *dev);
+void ib_destroy_qp_security_begin(struct ib_qp_security *sec);
+void ib_destroy_qp_security_abort(struct ib_qp_security *sec);
+void ib_destroy_qp_security_end(struct ib_qp_security *sec);
+int ib_open_shared_qp_security(struct ib_qp *qp, struct ib_device *dev);
+void ib_close_shared_qp_security(struct ib_qp_security *sec);
+#else
+static inline void ib_security_destroy_port_pkey_list(struct ib_device *device)
+{
+}
+
+static inline void ib_security_cache_change(struct ib_device *device,
+ u8 port_num,
+ u64 subnet_prefix)
+{
+}
+
+static inline int ib_security_modify_qp(struct ib_qp *qp,
+ struct ib_qp_attr *qp_attr,
+ int qp_attr_mask,
+ struct ib_udata *udata)
+{
+ return qp->device->modify_qp(qp->real_qp,
+ qp_attr,
+ qp_attr_mask,
+ udata);
+}
+
+static inline int ib_create_qp_security(struct ib_qp *qp,
+ struct ib_device *dev)
+{
+ return 0;
+}
+
+static inline void ib_destroy_qp_security_begin(struct ib_qp_security *sec)
+{
+}
+
+static inline void ib_destroy_qp_security_abort(struct ib_qp_security *sec)
+{
+}
+
+static inline void ib_destroy_qp_security_end(struct ib_qp_security *sec)
+{
+}
+
+static inline int ib_open_shared_qp_security(struct ib_qp *qp,
+ struct ib_device *dev)
+{
+ return 0;
+}
+
+static inline void ib_close_shared_qp_security(struct ib_qp_security *sec)
+{
+}
+#endif
#endif /* _CORE_PRIV_H */
diff --git a/drivers/infiniband/core/device.c b/drivers/infiniband/core/device.c
index 760ef60..5b42e83 100644
--- a/drivers/infiniband/core/device.c
+++ b/drivers/infiniband/core/device.c
@@ -320,6 +320,30 @@ void ib_get_device_fw_str(struct ib_device *dev, char *str, size_t str_len)
}
EXPORT_SYMBOL(ib_get_device_fw_str);
+static int setup_port_pkey_list(struct ib_device *device)
+{
+ int i;
+
+ /**
+ * device->port_pkey_list is indexed directly by the port number,
+ * Therefore it is declared as a 1 based array with potential empty
+ * slots at the beginning.
+ */
+ device->port_pkey_list = kzalloc(sizeof(*device->port_pkey_list)
+ * (rdma_end_port(device) + 1),
+ GFP_KERNEL);
+
+ if (!device->port_pkey_list)
+ return -ENOMEM;
+
+ for (i = 0; i < (rdma_end_port(device) + 1); i++) {
+ spin_lock_init(&device->port_pkey_list[i].list_lock);
+ INIT_LIST_HEAD(&device->port_pkey_list[i].pkey_list);
+ }
+
+ return 0;
+}
+
/**
* ib_register_device - Register an IB device with IB core
* @device:Device to register
@@ -357,6 +381,12 @@ int ib_register_device(struct ib_device *device,
goto out;
}
+ ret = setup_port_pkey_list(device);
+ if (ret) {
+ dev_warn(device->dma_device, "Couldn't create per port_pkey_list\n");
+ goto out;
+ }
+
ret = ib_cache_setup_one(device);
if (ret) {
pr_warn("Couldn't set up InfiniBand P_Key/GID cache\n");
@@ -427,6 +457,9 @@ void ib_unregister_device(struct ib_device *device)
ib_device_unregister_sysfs(device);
ib_cache_cleanup_one(device);
+ ib_security_destroy_port_pkey_list(device);
+ kfree(device->port_pkey_list);
+
down_write(&lists_rwsem);
spin_lock_irqsave(&device->client_data_lock, flags);
list_for_each_entry_safe(context, tmp, &device->client_data_list, list)
diff --git a/drivers/infiniband/core/security.c b/drivers/infiniband/core/security.c
new file mode 100644
index 0000000..45800dd
--- /dev/null
+++ b/drivers/infiniband/core/security.c
@@ -0,0 +1,617 @@
+/*
+ * Copyright (c) 2016 Mellanox Technologies Ltd. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses. You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ * Redistribution and use in source and binary forms, with or
+ * without modification, are permitted provided that the following
+ * conditions are met:
+ *
+ * - Redistributions of source code must retain the above
+ * copyright notice, this list of conditions and the following
+ * disclaimer.
+ *
+ * - Redistributions in binary form must reproduce the above
+ * copyright notice, this list of conditions and the following
+ * disclaimer in the documentation and/or other materials
+ * provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#ifdef CONFIG_SECURITY_INFINIBAND
+
+#include <linux/security.h>
+#include <linux/completion.h>
+#include <linux/list.h>
+
+#include <rdma/ib_verbs.h>
+#include <rdma/ib_cache.h>
+#include "core_priv.h"
+
+static struct pkey_index_qp_list *get_pkey_idx_qp_list(struct ib_port_pkey *pp)
+{
+ struct pkey_index_qp_list *pkey = NULL;
+ struct pkey_index_qp_list *tmp_pkey;
+ struct ib_device *dev = pp->sec->dev;
+
+ spin_lock(&dev->port_pkey_list[pp->port_num].list_lock);
+ list_for_each_entry(tmp_pkey,
+ &dev->port_pkey_list[pp->port_num].pkey_list,
+ pkey_index_list) {
+ if (tmp_pkey->pkey_index == pp->pkey_index) {
+ pkey = tmp_pkey;
+ break;
+ }
+ }
+ spin_unlock(&dev->port_pkey_list[pp->port_num].list_lock);
+ return pkey;
+}
+
+static int get_pkey_and_subnet_prefix(struct ib_port_pkey *pp,
+ u16 *pkey,
+ u64 *subnet_prefix)
+{
+ struct ib_device *dev = pp->sec->dev;
+ int ret;
+
+ ret = ib_get_cached_pkey(dev, pp->port_num, pp->pkey_index, pkey);
+ if (ret)
+ return ret;
+
+ ret = ib_get_cached_subnet_prefix(dev, pp->port_num, subnet_prefix);
+
+ return ret;
+}
+
+static int enforce_qp_pkey_security(u16 pkey,
+ u64 subnet_prefix,
+ struct ib_qp_security *qp_sec)
+{
+ struct ib_qp_security *shared_qp_sec;
+ int ret;
+
+ ret = security_ib_pkey_access(qp_sec->security, subnet_prefix, pkey);
+ if (ret)
+ return ret;
+
+ if (qp_sec->qp == qp_sec->qp->real_qp) {
+ list_for_each_entry(shared_qp_sec,
+ &qp_sec->shared_qp_list,
+ shared_qp_list) {
+ ret = security_ib_pkey_access(shared_qp_sec->security,
+ subnet_prefix,
+ pkey);
+ if (ret)
+ return ret;
+ }
+ }
+ return 0;
+}
+
+/* The caller of this function must hold the QP security
+ * mutex of the QP of the security structure in *pps.
+ *
+ * It takes separate ports_pkeys and security structure
+ * because in some cases the pps will be for a new settings
+ * or the pps will be for the real QP and security structure
+ * will be for a shared QP.
+ */
+static int check_qp_port_pkey_settings(struct ib_ports_pkeys *pps,
+ struct ib_qp_security *sec)
+{
+ u64 subnet_prefix;
+ u16 pkey;
+ int ret = 0;
+
+ if (!pps)
+ return 0;
+
+ if (pps->main.state != IB_PORT_PKEY_NOT_VALID) {
+ get_pkey_and_subnet_prefix(&pps->main,
+ &pkey,
+ &subnet_prefix);
+
+ ret = enforce_qp_pkey_security(pkey,
+ subnet_prefix,
+ sec);
+ }
+ if (ret)
+ goto out;
+
+ if (pps->alt.state != IB_PORT_PKEY_NOT_VALID) {
+ get_pkey_and_subnet_prefix(&pps->alt,
+ &pkey,
+ &subnet_prefix);
+
+ ret = enforce_qp_pkey_security(pkey,
+ subnet_prefix,
+ sec);
+ }
+
+ if (ret)
+ goto out;
+
+out:
+ return ret;
+}
+
+/* The caller of this function must hold the QP security
+ * mutex.
+ */
+static void qp_to_error(struct ib_qp_security *sec)
+{
+ struct ib_qp_security *shared_qp_sec;
+ struct ib_qp_attr attr = {
+ .qp_state = IB_QPS_ERR
+ };
+ struct ib_event event = {
+ .event = IB_EVENT_QP_FATAL
+ };
+
+ /* If the QP is in the process of being destroyed
+ * the qp pointer in the security structure is
+ * undefined. It cannot be modified now.
+ */
+ if (sec->destroying)
+ return;
+
+ ib_modify_qp(sec->qp,
+ &attr,
+ IB_QP_STATE);
+
+ if (sec->qp->event_handler && sec->qp->qp_context) {
+ event.element.qp = sec->qp;
+ sec->qp->event_handler(&event,
+ sec->qp->qp_context);
+ }
+
+ list_for_each_entry(shared_qp_sec,
+ &sec->shared_qp_list,
+ shared_qp_list) {
+ struct ib_qp *qp = shared_qp_sec->qp;
+
+ if (qp->event_handler && qp->qp_context) {
+ event.element.qp = qp;
+ event.device = qp->device;
+ qp->event_handler(&event,
+ qp->qp_context);
+ }
+ }
+}
+
+static inline void check_pkey_qps(struct pkey_index_qp_list *pkey,
+ struct ib_device *device,
+ u8 port_num,
+ u64 subnet_prefix)
+{
+ struct ib_port_pkey *pp, *tmp_pp;
+ bool comp;
+ LIST_HEAD(to_error_list);
+ u16 pkey_val;
+
+ if (!ib_get_cached_pkey(device,
+ port_num,
+ pkey->pkey_index,
+ &pkey_val)) {
+ spin_lock(&pkey->qp_list_lock);
+ list_for_each_entry(pp, &pkey->qp_list, qp_list) {
+ if (atomic_read(&pp->sec->error_list_count))
+ continue;
+
+ if (enforce_qp_pkey_security(pkey_val,
+ subnet_prefix,
+ pp->sec)) {
+ atomic_inc(&pp->sec->error_list_count);
+ list_add(&pp->to_error_list,
+ &to_error_list);
+ }
+ }
+ spin_unlock(&pkey->qp_list_lock);
+ }
+
+ list_for_each_entry_safe(pp,
+ tmp_pp,
+ &to_error_list,
+ to_error_list) {
+ mutex_lock(&pp->sec->mutex);
+ qp_to_error(pp->sec);
+ list_del(&pp->to_error_list);
+ atomic_dec(&pp->sec->error_list_count);
+ comp = pp->sec->destroying;
+ mutex_unlock(&pp->sec->mutex);
+
+ if (comp)
+ complete(&pp->sec->error_complete);
+ }
+}
+
+/* The caller of this function must hold the QP security
+ * mutex.
+ */
+static int port_pkey_list_insert(struct ib_port_pkey *pp)
+{
+ struct pkey_index_qp_list *tmp_pkey;
+ struct pkey_index_qp_list *pkey;
+ struct ib_device *dev;
+ u8 port_num = pp->port_num;
+ int ret = 0;
+
+ if (pp->state != IB_PORT_PKEY_VALID)
+ return 0;
+
+ dev = pp->sec->dev;
+
+ pkey = get_pkey_idx_qp_list(pp);
+
+ if (!pkey) {
+ bool found = false;
+
+ pkey = kzalloc(sizeof(*pkey), GFP_KERNEL);
+ if (!pkey)
+ return -ENOMEM;
+
+ spin_lock(&dev->port_pkey_list[port_num].list_lock);
+ /* Check for the PKey again. A racing process may
+ * have created it.
+ */
+ list_for_each_entry(tmp_pkey,
+ &dev->port_pkey_list[port_num].pkey_list,
+ pkey_index_list) {
+ if (tmp_pkey->pkey_index == pp->pkey_index) {
+ kfree(pkey);
+ pkey = tmp_pkey;
+ found = true;
+ break;
+ }
+ }
+
+ if (!found) {
+ pkey->pkey_index = pp->pkey_index;
+ spin_lock_init(&pkey->qp_list_lock);
+ INIT_LIST_HEAD(&pkey->qp_list);
+ list_add(&pkey->pkey_index_list,
+ &dev->port_pkey_list[port_num].pkey_list);
+ }
+ spin_unlock(&dev->port_pkey_list[port_num].list_lock);
+ }
+
+ spin_lock(&pkey->qp_list_lock);
+ list_add(&pp->qp_list, &pkey->qp_list);
+ spin_unlock(&pkey->qp_list_lock);
+
+ pp->state = IB_PORT_PKEY_LISTED;
+
+ return ret;
+}
+
+/* The caller of this function must hold the QP security
+ * mutex.
+ */
+static void port_pkey_list_remove(struct ib_port_pkey *pp)
+{
+ struct pkey_index_qp_list *pkey;
+
+ if (pp->state != IB_PORT_PKEY_LISTED)
+ return;
+
+ pkey = get_pkey_idx_qp_list(pp);
+
+ spin_lock(&pkey->qp_list_lock);
+ list_del(&pp->qp_list);
+ spin_unlock(&pkey->qp_list_lock);
+
+ /* The setting may still be valid, i.e. after
+ * a destroy has failed for example.
+ */
+ pp->state = IB_PORT_PKEY_VALID;
+}
+
+static void destroy_qp_security(struct ib_qp_security *sec)
+{
+ security_ib_free_security(sec->security);
+ kfree(sec->ports_pkeys);
+ kfree(sec);
+}
+
+/* The caller of this function must hold the QP security
+ * mutex.
+ */
+static struct ib_ports_pkeys *get_new_pps(const struct ib_qp *qp,
+ const struct ib_qp_attr *qp_attr,
+ int qp_attr_mask)
+{
+ struct ib_ports_pkeys *new_pps;
+ struct ib_ports_pkeys *qp_pps = qp->qp_sec->ports_pkeys;
+
+ new_pps = kzalloc(sizeof(*new_pps), GFP_KERNEL);
+ if (!new_pps)
+ return NULL;
+
+ if (qp_attr_mask & (IB_QP_PKEY_INDEX | IB_QP_PORT)) {
+ if (!qp_pps) {
+ new_pps->main.port_num = qp_attr->port_num;
+ new_pps->main.pkey_index = qp_attr->pkey_index;
+ } else {
+ new_pps->main.port_num = (qp_attr_mask & IB_QP_PORT) ?
+ qp_attr->port_num :
+ qp_pps->main.port_num;
+
+ new_pps->main.pkey_index =
+ (qp_attr_mask & IB_QP_PKEY_INDEX) ?
+ qp_attr->pkey_index :
+ qp_pps->main.pkey_index;
+ }
+ new_pps->main.state = IB_PORT_PKEY_VALID;
+ } else if (qp_pps) {
+ new_pps->main.port_num = qp_pps->main.port_num;
+ new_pps->main.pkey_index = qp_pps->main.pkey_index;
+ if (qp_pps->main.state != IB_PORT_PKEY_NOT_VALID)
+ new_pps->main.state = IB_PORT_PKEY_VALID;
+ }
+
+ if (qp_attr_mask & IB_QP_ALT_PATH) {
+ new_pps->alt.port_num = qp_attr->alt_port_num;
+ new_pps->alt.pkey_index = qp_attr->alt_pkey_index;
+ new_pps->alt.state = IB_PORT_PKEY_VALID;
+ } else if (qp_pps) {
+ new_pps->alt.port_num = qp_pps->alt.port_num;
+ new_pps->alt.pkey_index = qp_pps->alt.pkey_index;
+ if (qp_pps->alt.state != IB_PORT_PKEY_NOT_VALID)
+ new_pps->alt.state = IB_PORT_PKEY_VALID;
+ }
+
+ new_pps->main.sec = qp->qp_sec;
+ new_pps->alt.sec = qp->qp_sec;
+ return new_pps;
+}
+
+int ib_open_shared_qp_security(struct ib_qp *qp, struct ib_device *dev)
+{
+ struct ib_qp *real_qp = qp->real_qp;
+ int ret;
+
+ ret = ib_create_qp_security(qp, dev);
+
+ if (ret)
+ goto out;
+
+ mutex_lock(&real_qp->qp_sec->mutex);
+ ret = check_qp_port_pkey_settings(real_qp->qp_sec->ports_pkeys,
+ qp->qp_sec);
+
+ if (ret)
+ goto ret;
+
+ if (qp != real_qp)
+ list_add(&qp->qp_sec->shared_qp_list,
+ &real_qp->qp_sec->shared_qp_list);
+ret:
+ mutex_unlock(&real_qp->qp_sec->mutex);
+ if (ret)
+ destroy_qp_security(qp->qp_sec);
+
+out:
+ return ret;
+}
+
+void ib_close_shared_qp_security(struct ib_qp_security *sec)
+{
+ struct ib_qp *real_qp = sec->qp->real_qp;
+
+ mutex_lock(&real_qp->qp_sec->mutex);
+ list_del(&sec->shared_qp_list);
+ mutex_unlock(&real_qp->qp_sec->mutex);
+
+ destroy_qp_security(sec);
+}
+
+int ib_create_qp_security(struct ib_qp *qp, struct ib_device *dev)
+{
+ int ret;
+
+ qp->qp_sec = kzalloc(sizeof(*qp->qp_sec), GFP_KERNEL);
+ if (!qp->qp_sec)
+ return -ENOMEM;
+
+ qp->qp_sec->qp = qp;
+ qp->qp_sec->dev = dev;
+ mutex_init(&qp->qp_sec->mutex);
+ INIT_LIST_HEAD(&qp->qp_sec->shared_qp_list);
+ atomic_set(&qp->qp_sec->error_list_count, 0);
+ init_completion(&qp->qp_sec->error_complete);
+ ret = security_ib_alloc_security(&qp->qp_sec->security);
+ if (ret)
+ kfree(qp->qp_sec);
+
+ return ret;
+}
+EXPORT_SYMBOL(ib_create_qp_security);
+
+void ib_destroy_qp_security_begin(struct ib_qp_security *sec)
+{
+ mutex_lock(&sec->mutex);
+
+ /* Remove the QP from the lists so it won't get added to
+ * a to_error_list during the destroy process.
+ */
+ if (sec->ports_pkeys) {
+ port_pkey_list_remove(&sec->ports_pkeys->main);
+ port_pkey_list_remove(&sec->ports_pkeys->alt);
+ }
+
+ /* If the QP is already in one or more of those lists
+ * the destroying flag will ensure the to error flow
+ * doesn't operate on an undefined QP.
+ */
+ sec->destroying = true;
+
+ /* Record the error list count to know how many completions
+ * to wait for.
+ */
+ sec->error_comps_pending = atomic_read(&sec->error_list_count);
+
+ mutex_unlock(&sec->mutex);
+}
+
+void ib_destroy_qp_security_abort(struct ib_qp_security *sec)
+{
+ int ret;
+ int i;
+
+ /* If a concurrent cache update is in progress this
+ * QP security could be marked for an error state
+ * transition. Wait for this to complete.
+ */
+ for (i = 0; i < sec->error_comps_pending; i++)
+ wait_for_completion(&sec->error_complete);
+
+ mutex_lock(&sec->mutex);
+ sec->destroying = false;
+
+ /* Restore the position in the lists and verify
+ * access is still allowed in case a cache update
+ * occurred while attempting to destroy.
+ *
+ * Because these setting were listed already
+ * and removed during ib_destroy_qp_security_begin
+ * we know the pkey_index_qp_list for the PKey
+ * already exists so port_pkey_list_insert won't fail.
+ */
+ if (sec->ports_pkeys) {
+ port_pkey_list_insert(&sec->ports_pkeys->main);
+ port_pkey_list_insert(&sec->ports_pkeys->alt);
+ }
+
+ ret = check_qp_port_pkey_settings(sec->ports_pkeys, sec);
+ if (ret)
+ qp_to_error(sec);
+
+ mutex_unlock(&sec->mutex);
+}
+
+void ib_destroy_qp_security_end(struct ib_qp_security *sec)
+{
+ int i;
+
+ /* If a concurrent cache update is occurring we must
+ * wait until this QP security structure is processed
+ * in the QP to error flow before destroying it because
+ * the to_error_list is in use.
+ */
+ for (i = 0; i < sec->error_comps_pending; i++)
+ wait_for_completion(&sec->error_complete);
+
+ destroy_qp_security(sec);
+}
+
+void ib_security_cache_change(struct ib_device *device,
+ u8 port_num,
+ u64 subnet_prefix)
+{
+ struct pkey_index_qp_list *pkey;
+
+ list_for_each_entry(pkey,
+ &device->port_pkey_list[port_num].pkey_list,
+ pkey_index_list) {
+ check_pkey_qps(pkey,
+ device,
+ port_num,
+ subnet_prefix);
+ }
+}
+
+void ib_security_destroy_port_pkey_list(struct ib_device *device)
+{
+ struct pkey_index_qp_list *pkey, *tmp_pkey;
+ int i;
+
+ for (i = rdma_start_port(device); i <= rdma_end_port(device); i++) {
+ spin_lock(&device->port_pkey_list[i].list_lock);
+ list_for_each_entry_safe(pkey,
+ tmp_pkey,
+ &device->port_pkey_list[i].pkey_list,
+ pkey_index_list) {
+ list_del(&pkey->pkey_index_list);
+ kfree(pkey);
+ }
+ spin_unlock(&device->port_pkey_list[i].list_lock);
+ }
+}
+
+int ib_security_modify_qp(struct ib_qp *qp,
+ struct ib_qp_attr *qp_attr,
+ int qp_attr_mask,
+ struct ib_udata *udata)
+{
+ int ret = 0;
+ struct ib_ports_pkeys *tmp_pps;
+ struct ib_ports_pkeys *new_pps;
+ bool special_qp = (qp->qp_type == IB_QPT_SMI ||
+ qp->qp_type == IB_QPT_GSI);
+ bool pps_change = ((qp_attr_mask & (IB_QP_PKEY_INDEX | IB_QP_PORT)) ||
+ (qp_attr_mask & IB_QP_ALT_PATH));
+
+ if (pps_change && !special_qp) {
+ mutex_lock(&qp->qp_sec->mutex);
+ new_pps = get_new_pps(qp,
+ qp_attr,
+ qp_attr_mask);
+
+ /* Add this QP to the lists for the new port
+ * and pkey settings before checking for permission
+ * in case there is a concurrent cache update
+ * occurring. Walking the list for a cache change
+ * doesn't acquire the security mutex unless it's
+ * sending the QP to error.
+ */
+ ret = port_pkey_list_insert(&new_pps->main);
+
+ if (!ret)
+ ret = port_pkey_list_insert(&new_pps->alt);
+
+ if (!ret)
+ ret = check_qp_port_pkey_settings(new_pps,
+ qp->qp_sec);
+ }
+
+ if (!ret)
+ ret = qp->device->modify_qp(qp->real_qp,
+ qp_attr,
+ qp_attr_mask,
+ udata);
+
+ if (pps_change && !special_qp) {
+ /* Clean up the lists and free the appropriate
+ * ports_pkeys structure.
+ */
+ if (ret) {
+ tmp_pps = new_pps;
+ } else {
+ tmp_pps = qp->qp_sec->ports_pkeys;
+ qp->qp_sec->ports_pkeys = new_pps;
+ }
+
+ if (tmp_pps) {
+ port_pkey_list_remove(&tmp_pps->main);
+ port_pkey_list_remove(&tmp_pps->alt);
+ }
+ kfree(tmp_pps);
+ mutex_unlock(&qp->qp_sec->mutex);
+ }
+ return ret;
+}
+EXPORT_SYMBOL(ib_security_modify_qp);
+
+#endif /* CONFIG_SECURITY_INFINIBAND */
diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c
index cb3f515a..ceb9154 100644
--- a/drivers/infiniband/core/uverbs_cmd.c
+++ b/drivers/infiniband/core/uverbs_cmd.c
@@ -37,7 +37,7 @@
#include <linux/fs.h>
#include <linux/slab.h>
#include <linux/sched.h>
-
+#include <linux/security.h>
#include <asm/uaccess.h>
#include "uverbs.h"
@@ -1915,6 +1915,10 @@ static int create_qp(struct ib_uverbs_file *file,
}
if (cmd->qp_type != IB_QPT_XRC_TGT) {
+ ret = ib_create_qp_security(qp, device);
+ if (ret)
+ goto err_destroy;
+
qp->real_qp = qp;
qp->device = device;
qp->pd = pd;
@@ -2405,10 +2409,18 @@ ssize_t ib_uverbs_modify_qp(struct ib_uverbs_file *file,
ret = ib_resolve_eth_dmac(qp, attr, &cmd.attr_mask);
if (ret)
goto release_qp;
- ret = qp->device->modify_qp(qp, attr,
- modify_qp_mask(qp->qp_type, cmd.attr_mask), &udata);
+
+ ret = ib_security_modify_qp(qp,
+ attr,
+ modify_qp_mask(qp->qp_type,
+ cmd.attr_mask),
+ &udata);
} else {
- ret = ib_modify_qp(qp, attr, modify_qp_mask(qp->qp_type, cmd.attr_mask));
+ ret = ib_security_modify_qp(qp,
+ attr,
+ modify_qp_mask(qp->qp_type,
+ cmd.attr_mask),
+ NULL);
}
if (ret)
diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c
index 8368764..9845e37 100644
--- a/drivers/infiniband/core/verbs.c
+++ b/drivers/infiniband/core/verbs.c
@@ -44,6 +44,7 @@
#include <linux/in.h>
#include <linux/in6.h>
#include <net/addrconf.h>
+#include <linux/security.h>
#include <rdma/ib_verbs.h>
#include <rdma/ib_cache.h>
@@ -708,12 +709,20 @@ static struct ib_qp *__ib_open_qp(struct ib_qp *real_qp,
{
struct ib_qp *qp;
unsigned long flags;
+ int err;
qp = kzalloc(sizeof *qp, GFP_KERNEL);
if (!qp)
return ERR_PTR(-ENOMEM);
qp->real_qp = real_qp;
+ err = ib_open_shared_qp_security(qp, real_qp->device);
+ if (err) {
+ kfree(qp);
+ return ERR_PTR(err);
+ }
+
+ qp->real_qp = real_qp;
atomic_inc(&real_qp->usecnt);
qp->device = real_qp->device;
qp->event_handler = event_handler;
@@ -799,6 +808,12 @@ struct ib_qp *ib_create_qp(struct ib_pd *pd,
if (IS_ERR(qp))
return qp;
+ ret = ib_create_qp_security(qp, device);
+ if (ret) {
+ ib_destroy_qp(qp);
+ return ERR_PTR(ret);
+ }
+
qp->device = device;
qp->real_qp = qp;
qp->uobject = NULL;
@@ -1257,7 +1272,7 @@ int ib_modify_qp(struct ib_qp *qp,
if (ret)
return ret;
- return qp->device->modify_qp(qp->real_qp, qp_attr, qp_attr_mask, NULL);
+ return ib_security_modify_qp(qp->real_qp, qp_attr, qp_attr_mask, NULL);
}
EXPORT_SYMBOL(ib_modify_qp);
@@ -1286,6 +1301,7 @@ int ib_close_qp(struct ib_qp *qp)
spin_unlock_irqrestore(&real_qp->device->event_handler_lock, flags);
atomic_dec(&real_qp->usecnt);
+ ib_close_shared_qp_security(qp->qp_sec);
kfree(qp);
return 0;
@@ -1326,6 +1342,7 @@ int ib_destroy_qp(struct ib_qp *qp)
struct ib_cq *scq, *rcq;
struct ib_srq *srq;
struct ib_rwq_ind_table *ind_tbl;
+ struct ib_qp_security *sec;
int ret;
WARN_ON_ONCE(qp->mrs_used > 0);
@@ -1341,6 +1358,9 @@ int ib_destroy_qp(struct ib_qp *qp)
rcq = qp->recv_cq;
srq = qp->srq;
ind_tbl = qp->rwq_ind_tbl;
+ sec = qp->qp_sec;
+ if (sec)
+ ib_destroy_qp_security_begin(sec);
if (!qp->uobject)
rdma_rw_cleanup_mrs(qp);
@@ -1357,6 +1377,11 @@ int ib_destroy_qp(struct ib_qp *qp)
atomic_dec(&srq->usecnt);
if (ind_tbl)
atomic_dec(&ind_tbl->usecnt);
+ if (sec)
+ ib_destroy_qp_security_end(sec);
+ } else {
+ if (sec)
+ ib_destroy_qp_security_abort(sec);
}
return ret;
diff --git a/include/linux/lsm_hooks.h b/include/linux/lsm_hooks.h
index 558adfa..d576efc 100644
--- a/include/linux/lsm_hooks.h
+++ b/include/linux/lsm_hooks.h
@@ -8,6 +8,7 @@
* Copyright (C) 2001 Silicon Graphics, Inc. (Trust Technology Group)
* Copyright (C) 2015 Intel Corporation.
* Copyright (C) 2015 Casey Schaufler <casey@schaufler-ca.com>
+ * Copyright (C) 2016 Mellanox Techonologies
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
@@ -903,6 +904,21 @@
* associated with the TUN device's security structure.
* @security pointer to the TUN devices's security structure.
*
+ * Security hooks for Infiniband
+ *
+ * @ib_pkey_access:
+ * Check permission to access a pkey when modifing a QP.
+ * @subnet_prefix the subnet prefix of the port being used.
+ * @pkey the pkey to be accessed.
+ * @sec pointer to a security structure.
+ * @ib_alloc_security:
+ * Allocate a security structure for Infiniband objects.
+ * @sec pointer to a security structure pointer.
+ * Returns 0 on success, non-zero on failure
+ * @ib_free_security:
+ * Deallocate an Infiniband security structure.
+ * @sec contains the security structure to be freed.
+ *
* Security hooks for XFRM operations.
*
* @xfrm_policy_alloc_security:
@@ -1611,6 +1627,12 @@
int (*tun_dev_open)(void *security);
#endif /* CONFIG_SECURITY_NETWORK */
+#ifdef CONFIG_SECURITY_INFINIBAND
+ int (*ib_pkey_access)(void *sec, u64 subnet_prefix, u16 pkey);
+ int (*ib_alloc_security)(void **sec);
+ void (*ib_free_security)(void *sec);
+#endif /* CONFIG_SECURITY_INFINIBAND */
+
#ifdef CONFIG_SECURITY_NETWORK_XFRM
int (*xfrm_policy_alloc_security)(struct xfrm_sec_ctx **ctxp,
struct xfrm_user_sec_ctx *sec_ctx,
@@ -1841,6 +1863,11 @@ struct security_hook_heads {
struct list_head tun_dev_attach;
struct list_head tun_dev_open;
#endif /* CONFIG_SECURITY_NETWORK */
+#ifdef CONFIG_SECURITY_INFINIBAND
+ struct list_head ib_pkey_access;
+ struct list_head ib_alloc_security;
+ struct list_head ib_free_security;
+#endif /* CONFIG_SECURITY_INFINIBAND */
#ifdef CONFIG_SECURITY_NETWORK_XFRM
struct list_head xfrm_policy_alloc_security;
struct list_head xfrm_policy_clone_security;
diff --git a/include/linux/security.h b/include/linux/security.h
index c2125e9..342ca4c 100644
--- a/include/linux/security.h
+++ b/include/linux/security.h
@@ -6,6 +6,7 @@
* Copyright (C) 2001 Networks Associates Technology, Inc <ssmalley@nai.com>
* Copyright (C) 2001 James Morris <jmorris@intercode.com.au>
* Copyright (C) 2001 Silicon Graphics, Inc. (Trust Technology Group)
+ * Copyright (C) 2016 Mellanox Techonologies
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
@@ -1393,6 +1394,26 @@ static inline int security_tun_dev_open(void *security)
}
#endif /* CONFIG_SECURITY_NETWORK */
+#ifdef CONFIG_SECURITY_INFINIBAND
+int security_ib_pkey_access(void *sec, u64 subnet_prefix, u16 pkey);
+int security_ib_alloc_security(void **sec);
+void security_ib_free_security(void *sec);
+#else /* CONFIG_SECURITY_INFINIBAND */
+static inline int security_ib_pkey_access(void *sec, u64 subnet_prefix, u16 pkey)
+{
+ return 0;
+}
+
+static inline int security_ib_alloc_security(void **sec)
+{
+ return 0;
+}
+
+static inline void security_ib_free_security(void *sec)
+{
+}
+#endif /* CONFIG_SECURITY_INFINIBAND */
+
#ifdef CONFIG_SECURITY_NETWORK_XFRM
int security_xfrm_policy_alloc(struct xfrm_sec_ctx **ctxp,
diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h
index db178fd..1143586 100644
--- a/include/rdma/ib_verbs.h
+++ b/include/rdma/ib_verbs.h
@@ -1511,6 +1511,45 @@ struct ib_rwq_ind_table_init_attr {
struct ib_wq **ind_tbl;
};
+enum port_pkey_state {
+ IB_PORT_PKEY_NOT_VALID = 0,
+ IB_PORT_PKEY_VALID = 1,
+ IB_PORT_PKEY_LISTED = 2,
+};
+
+struct ib_qp_security;
+
+struct ib_port_pkey {
+ enum port_pkey_state state;
+ u16 pkey_index;
+ u8 port_num;
+ struct list_head qp_list;
+ struct list_head to_error_list;
+ struct ib_qp_security *sec;
+};
+
+struct ib_ports_pkeys {
+ struct ib_port_pkey main;
+ struct ib_port_pkey alt;
+};
+
+struct ib_qp_security {
+ struct ib_qp *qp;
+ struct ib_device *dev;
+ /* Hold this mutex when changing port and pkey settings. */
+ struct mutex mutex;
+ struct ib_ports_pkeys *ports_pkeys;
+ /* A list of all open shared QP handles. Required to enforce security
+ * properly for all users of a shared QP.
+ */
+ struct list_head shared_qp_list;
+ void *security;
+ bool destroying;
+ atomic_t error_list_count;
+ struct completion error_complete;
+ int error_comps_pending;
+};
+
/*
* @max_write_sge: Maximum SGE elements per RDMA WRITE request.
* @max_read_sge: Maximum SGE elements per RDMA READ request.
@@ -1540,6 +1579,7 @@ struct ib_qp {
u32 max_read_sge;
enum ib_qp_type qp_type;
struct ib_rwq_ind_table *rwq_ind_tbl;
+ struct ib_qp_security *qp_sec;
};
struct ib_mr {
@@ -1820,6 +1860,12 @@ struct ib_port_immutable {
u32 max_mad_size;
};
+struct ib_port_pkey_list {
+ /* Lock to hold while modifying the list. */
+ spinlock_t list_lock;
+ struct list_head pkey_list;
+};
+
struct ib_device {
struct device *dma_device;
@@ -1842,6 +1888,8 @@ struct ib_device {
int num_comp_vectors;
+ struct ib_port_pkey_list *port_pkey_list;
+
struct iw_cm_verbs *iwcm;
/**
diff --git a/security/Kconfig b/security/Kconfig
index 118f454..85cf893 100644
--- a/security/Kconfig
+++ b/security/Kconfig
@@ -49,6 +49,15 @@ config SECURITY_NETWORK
implement socket and networking access controls.
If you are unsure how to answer this question, answer N.
+config SECURITY_INFINIBAND
+ bool "Infiniband Security Hooks"
+ depends on SECURITY && INFINIBAND
+ help
+ This enables the Infiniband security hooks.
+ If enabled, a security module can use these hooks to
+ implement Infiniband access controls.
+ If you are unsure how to answer this question, answer N.
+
config SECURITY_NETWORK_XFRM
bool "XFRM (IPSec) Networking Security Hooks"
depends on XFRM && SECURITY_NETWORK
diff --git a/security/security.c b/security/security.c
index f825304..7d3bf2f 100644
--- a/security/security.c
+++ b/security/security.c
@@ -4,6 +4,7 @@
* Copyright (C) 2001 WireX Communications, Inc <chris@wirex.com>
* Copyright (C) 2001-2002 Greg Kroah-Hartman <greg@kroah.com>
* Copyright (C) 2001 Networks Associates Technology, Inc <ssmalley@nai.com>
+ * Copyright (C) 2016 Mellanox Technologies
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
@@ -1441,6 +1442,27 @@ int security_tun_dev_open(void *security)
#endif /* CONFIG_SECURITY_NETWORK */
+#ifdef CONFIG_SECURITY_INFINIBAND
+
+int security_ib_pkey_access(void *sec, u64 subnet_prefix, u16 pkey)
+{
+ return call_int_hook(ib_pkey_access, 0, sec, subnet_prefix, pkey);
+}
+EXPORT_SYMBOL(security_ib_pkey_access);
+
+int security_ib_alloc_security(void **sec)
+{
+ return call_int_hook(ib_alloc_security, 0, sec);
+}
+EXPORT_SYMBOL(security_ib_alloc_security);
+
+void security_ib_free_security(void *sec)
+{
+ call_void_hook(ib_free_security, sec);
+}
+EXPORT_SYMBOL(security_ib_free_security);
+#endif /* CONFIG_SECURITY_INFINIBAND */
+
#ifdef CONFIG_SECURITY_NETWORK_XFRM
int security_xfrm_policy_alloc(struct xfrm_sec_ctx **ctxp,
@@ -1898,6 +1920,15 @@ struct security_hook_heads security_hook_heads = {
LIST_HEAD_INIT(security_hook_heads.tun_dev_attach),
.tun_dev_open = LIST_HEAD_INIT(security_hook_heads.tun_dev_open),
#endif /* CONFIG_SECURITY_NETWORK */
+
+#ifdef CONFIG_SECURITY_INFINIBAND
+ .ib_pkey_access = LIST_HEAD_INIT(security_hook_heads.ib_pkey_access),
+ .ib_alloc_security =
+ LIST_HEAD_INIT(security_hook_heads.ib_alloc_security),
+ .ib_free_security =
+ LIST_HEAD_INIT(security_hook_heads.ib_free_security),
+#endif /* CONFIG_SECURITY_INFINIBAND */
+
#ifdef CONFIG_SECURITY_NETWORK_XFRM
.xfrm_policy_alloc_security =
LIST_HEAD_INIT(security_hook_heads.xfrm_policy_alloc_security),
--
1.8.3.1
^ permalink raw reply related
* [PATCH v4 1/9] IB/core: IB cache enhancements to support Infiniband security
From: Dan Jurgens @ 2016-11-08 21:06 UTC (permalink / raw)
To: chrisw, paul, sds, eparis, dledford, sean.hefty, hal.rosenstock
Cc: selinux, linux-security-module, linux-rdma, yevgenyp, liranl,
leonro, Daniel Jurgens
In-Reply-To: <1478639185-47521-1-git-send-email-danielj@mellanox.com>
From: Daniel Jurgens <danielj@mellanox.com>
Cache the subnet prefix and add a function to access it. Enforcing
security requires frequent queries of the subnet prefix and the pkeys in
the pkey table.
Also removed an unneded pr_warn about memory allocation failure.
issue: 736423
Change-Id: Ifdef64b097a8d1d55db65f08ce401d9d2e4b025e
Signed-off-by: Daniel Jurgens <danielj@mellanox.com>
Reviewed-by: Eli Cohen <eli@mellanox.com>
Reviewed-by: Leon Romanovsky <leonro@mellanox.com>
---
v2:
- In ib_get_cached_subnet_prefix wait to initialize p until after
validation. Yuval Shaia
Signed-off-by: Daniel Jurgens <danielj@mellanox.com>
---
drivers/infiniband/core/cache.c | 36 ++++++++++++++++++++++++++++++++++--
drivers/infiniband/core/core_priv.h | 3 +++
include/rdma/ib_verbs.h | 1 +
3 files changed, 38 insertions(+), 2 deletions(-)
diff --git a/drivers/infiniband/core/cache.c b/drivers/infiniband/core/cache.c
index 1a2984c..affc8ef 100644
--- a/drivers/infiniband/core/cache.c
+++ b/drivers/infiniband/core/cache.c
@@ -934,6 +934,26 @@ int ib_get_cached_pkey(struct ib_device *device,
}
EXPORT_SYMBOL(ib_get_cached_pkey);
+int ib_get_cached_subnet_prefix(struct ib_device *device,
+ u8 port_num,
+ u64 *sn_pfx)
+{
+ unsigned long flags;
+ int p;
+
+ if (port_num < rdma_start_port(device) ||
+ port_num > rdma_end_port(device))
+ return -EINVAL;
+
+ p = port_num - rdma_start_port(device);
+ read_lock_irqsave(&device->cache.lock, flags);
+ *sn_pfx = device->cache.subnet_prefix_cache[p];
+ read_unlock_irqrestore(&device->cache.lock, flags);
+
+ return 0;
+}
+EXPORT_SYMBOL(ib_get_cached_subnet_prefix);
+
int ib_find_cached_pkey(struct ib_device *device,
u8 port_num,
u16 pkey,
@@ -1110,6 +1130,8 @@ static void ib_cache_update(struct ib_device *device,
device->cache.lmc_cache[port - rdma_start_port(device)] = tprops->lmc;
+ device->cache.subnet_prefix_cache[port - rdma_start_port(device)] =
+ tprops->subnet_prefix;
write_unlock_irq(&device->cache.lock);
kfree(gid_cache);
@@ -1168,9 +1190,18 @@ int ib_cache_setup_one(struct ib_device *device)
(rdma_end_port(device) -
rdma_start_port(device) + 1),
GFP_KERNEL);
+
+ device->cache.subnet_prefix_cache =
+ kcalloc((rdma_end_port(device) - rdma_start_port(device) + 1),
+ sizeof(*device->cache.subnet_prefix_cache),
+ GFP_KERNEL);
+
if (!device->cache.pkey_cache ||
- !device->cache.lmc_cache) {
- pr_warn("Couldn't allocate cache for %s\n", device->name);
+ !device->cache.lmc_cache ||
+ !device->cache.subnet_prefix_cache) {
+ kfree(device->cache.pkey_cache);
+ kfree(device->cache.lmc_cache);
+ kfree(device->cache.subnet_prefix_cache);
return -ENOMEM;
}
@@ -1213,6 +1244,7 @@ void ib_cache_release_one(struct ib_device *device)
gid_table_release_one(device);
kfree(device->cache.pkey_cache);
kfree(device->cache.lmc_cache);
+ kfree(device->cache.subnet_prefix_cache);
}
void ib_cache_cleanup_one(struct ib_device *device)
diff --git a/drivers/infiniband/core/core_priv.h b/drivers/infiniband/core/core_priv.h
index 19d499d..ce826e4 100644
--- a/drivers/infiniband/core/core_priv.h
+++ b/drivers/infiniband/core/core_priv.h
@@ -153,4 +153,7 @@ int ib_nl_handle_set_timeout(struct sk_buff *skb,
int ib_nl_handle_ip_res_resp(struct sk_buff *skb,
struct netlink_callback *cb);
+int ib_get_cached_subnet_prefix(struct ib_device *device,
+ u8 port_num,
+ u64 *sn_pfx);
#endif /* _CORE_PRIV_H */
diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h
index 5ad43a4..db178fd 100644
--- a/include/rdma/ib_verbs.h
+++ b/include/rdma/ib_verbs.h
@@ -1761,6 +1761,7 @@ struct ib_cache {
struct ib_pkey_cache **pkey_cache;
struct ib_gid_table **gid_cache;
u8 *lmc_cache;
+ u64 *subnet_prefix_cache;
};
struct ib_dma_mapping_ops {
--
1.8.3.1
^ permalink raw reply related
* [PATCH v4 0/9] SELinux support for Infiniband RDMA
From: Dan Jurgens @ 2016-11-08 21:06 UTC (permalink / raw)
To: chrisw, paul, sds, eparis, dledford, sean.hefty, hal.rosenstock
Cc: selinux, linux-security-module, linux-rdma, yevgenyp, liranl,
leonro, Daniel Jurgens
From: Daniel Jurgens <danielj@mellanox.com>
Infiniband applications access HW from user-space -- traffic is generated
directly by HW, bypassing the kernel. Consequently, Infiniband Partitions,
which are associated directly with HW transport endpoints, are a natural
choice for enforcing granular mandatory access control for Infiniband. QPs may
only send or receives packets tagged with the corresponding partition key
(PKey). The PKey is not a cryptographic key; it's a 16 bit number identifying
the partition.
Every Infiniband fabric is controlled by a central Subnet Manager (SM). The SM
provisions the partitions by assigning each port with the partitions it can
access. In addition, the SM tags each port with a subnet prefix, which
identifies the subnet. Determining which users are allowed to access which
partition keys on a given subnet forms an effective policy for isolating users
on the fabric. Any application that attempts to send traffic on a given subnet
is automatically subject to the policy, regardless of which device and port it
uses. SM software configures the subnet through a privileged Subnet Management
Interface (SMI), which is presented by each Infiniband port. Thus, the SMI must
also be controlled to prevent unauthorized changes to fabric configuration and
partitioning.
To support access control for IB partitions and subnet management, security
contexts must be provided for two new types of objects - PKeys and IB ports.
A PKey label consists of a subnet prefix and a range of PKey values and is
similar to the labeling mechanism for netports. Each Infiniband port can reside
on a different subnet. So labeling the PKey values for specific subnet prefixes
provides the user maximum flexibility, as PKey values may be determined
independently for different subnets. There is a single access vector for PKeys
called "access".
An Infiniband port is labeled by device name and port number. There is a single
access vector for IB ports called "manage_subnet".
Because RDMA allows kernel bypass, enforcement must be done during connection
setup. Communication over RDMA requires a send and receive queue, collectively
known as a Queue Pair (QP). A QP must be initialized by privileged system calls
before it can be used to send or receive data. During initialization the user
must provide the PKey and port the QP will use; at this time access control can
be enforced.
Because there is a possibility that the enforcement settings or security
policy can change, a means of notifying the ib_core module of such changes
is required. To facilitate this a generic notification callback mechanism
is added to the LSM. One callback is registered for checking the QP PKey
associations when the policy changes. Mad agents also register a callback,
they cache the permission to send and receive SMPs to avoid another per
packet call to the LSM.
Because frequent accesses to the same PKey's SID is expected a cache is
implemented which is very similar to the netport cache.
In order to properly enforce security when changes to the PKey table or
security policy or enforcement occur ib_core must track which QPs are
using which port, pkey index, and alternate path for every IB device.
This makes operations that used to be atomic transactional.
When modifying a QP, ib_core must associate it with the PKey index, port,
and alternate path specified. If the QP was already associated with
different settings, the QP is added to the new list prior to the
modification. If the modify succeeds then the old listing is removed. If
the modify fails the new listing is removed and the old listing remains
unchanged.
When destroying a QP the ib_qp structure is freed by the decive specific
driver (i.e. mlx4_ib) if the 'destroy' is successful. This requires storing
security related information in a separate structure. When a 'destroy'
request is in process the ib_qp structure is in an undefined state so if
there are changes to the security policy or PKey table, the security checks
cannot reset the QP if it doesn't have permission for the new setting. If
the 'destroy' fails, security for that QP must be enforced again and its
status in the list is restored. If the 'destroy' succeeds the security info
can be cleaned up and freed.
There are a number of locks required to protect the QP security structure
and the QP to device/port/pkey index lists. If multiple locks are required,
the safe locking order is: QP security structure mutex first, followed by
any list locks needed, which are sorted first by port followed by pkey
index.
---
v2:
- Use void* blobs in the LSM hooks. Paul Moore
- Make the policy change callback generic. Yuval Shaia, Paul Moore
- Squash LSM changes into the patches where the calls are added. Paul Moore
- Don't add new initial SIDs. Stephen Smalley
- Squash MAD agent PKey and SMI patches and move logic to IB security. Dan Jurgens
- Changed ib_end_port to ib_port. Paul Moore
- Changed ib_port access vector from smp to manage_subnet. Paul Moore
- Added pkey and ib_port details to the audit log. Paul Moore
- See individual patches for more detail.
v3:
- ib_port -> ib_endport. Paul Moore
- use notifier chains for LSM notifications. Paul Moore
- reorder parameters in hooks to put security blob first. Paul Moore
- Don't treat device name as untrusted string in audit log. Paul Moore
v4:
- Added seperate AVC callback for LSM notifier. Paul Moore
- Removed unneeded braces in ocontext_read. Paul Moore
Daniel Jurgens (9):
IB/core: IB cache enhancements to support Infiniband security
IB/core: Enforce PKey security on QPs
selinux lsm IB/core: Implement LSM notification system
IB/core: Enforce security on management datagrams
selinux: Create policydb version for Infiniband support
selinux: Allocate and free infiniband security hooks
selinux: Implement Infiniband PKey "Access" access vector
selinux: Add IB Port SMP access vector
selinux: Add a cache for quicker retreival of PKey SIDs
drivers/infiniband/core/Makefile | 3 +-
drivers/infiniband/core/cache.c | 57 ++-
drivers/infiniband/core/core_priv.h | 115 ++++++
drivers/infiniband/core/device.c | 86 +++++
drivers/infiniband/core/mad.c | 52 ++-
drivers/infiniband/core/security.c | 709 +++++++++++++++++++++++++++++++++++
drivers/infiniband/core/uverbs_cmd.c | 20 +-
drivers/infiniband/core/verbs.c | 27 +-
include/linux/lsm_audit.h | 15 +
include/linux/lsm_hooks.h | 35 ++
include/linux/security.h | 35 ++
include/rdma/ib_mad.h | 4 +
include/rdma/ib_verbs.h | 49 +++
security/Kconfig | 9 +
security/lsm_audit.c | 18 +
security/security.c | 59 +++
security/selinux/Makefile | 2 +-
security/selinux/hooks.c | 86 ++++-
security/selinux/ibpkey.c | 245 ++++++++++++
security/selinux/include/classmap.h | 4 +
security/selinux/include/ibpkey.h | 31 ++
security/selinux/include/objsec.h | 11 +
security/selinux/include/security.h | 7 +-
security/selinux/selinuxfs.c | 2 +
security/selinux/ss/policydb.c | 129 ++++++-
security/selinux/ss/policydb.h | 27 +-
security/selinux/ss/services.c | 83 ++++
27 files changed, 1875 insertions(+), 45 deletions(-)
create mode 100644 drivers/infiniband/core/security.c
create mode 100644 security/selinux/ibpkey.c
create mode 100644 security/selinux/include/ibpkey.h
--
1.8.3.1
^ permalink raw reply
* Re: svc_xprt_put is no longer BH-safe
From: Chuck Lever @ 2016-11-08 21:05 UTC (permalink / raw)
To: J. Bruce Fields; +Cc: Linux NFS Mailing List
In-Reply-To: <20161108202035.GC26589@fieldses.org>
> On Nov 8, 2016, at 3:20 PM, J. Bruce Fields <bfields@fieldses.org> wrote:
>
> On Tue, Nov 08, 2016 at 03:13:13PM -0500, Chuck Lever wrote:
>>
>>> On Nov 8, 2016, at 3:03 PM, J. Bruce Fields <bfields@fieldses.org> wrote:
>>>
>>> On Sat, Oct 29, 2016 at 12:21:03PM -0400, Chuck Lever wrote:
>>>> Hi Bruce-
>>>
>>> Sorry for the slow response!
>>>
>>> ...
>>>> In commit 39a9beab5acb83176e8b9a4f0778749a09341f1f ('rpc: share one xps between
>>>> all backchannels') you added:
>>>>
>>>> diff --git a/net/sunrpc/svc_xprt.c b/net/sunrpc/svc_xprt.c
>>>> index f5572e3..4f01f63 100644
>>>> --- a/net/sunrpc/svc_xprt.c
>>>> +++ b/net/sunrpc/svc_xprt.c
>>>> @@ -136,6 +136,8 @@ static void svc_xprt_free(struct kref *kref)
>>>> /* See comment on corresponding get in xs_setup_bc_tcp(): */
>>>> if (xprt->xpt_bc_xprt)
>>>> xprt_put(xprt->xpt_bc_xprt);
>>>> + if (xprt->xpt_bc_xps)
>>>> + xprt_switch_put(xprt->xpt_bc_xps);
>>>> xprt->xpt_ops->xpo_free(xprt);
>>>> module_put(owner);
>>>> }
>>>>
>>>> svc_xprt_free() is invoked by svc_xprt_put(). svc_xprt_put() is called
>>>> from svc_rdma in soft IRQ context (eg. svc_rdma_wc_receive).
>>>
>>> Is that necessary? I wonder why the svcrdma code seems to be taking so
>>> many of its own references on svc_xprts.
>>
>> The idea is to keep the xprt around while Work Requests (I/O) are running,
>> so that the xprt is guaranteed to be there during work completions. The
>> completion handlers (where svc_xprt_put is often invoked) run in soft IRQ
>> context.
>>
>> It's simple to change completions to use a Work Queue instead, but testing
>> so far shows that will result in a performance loss. I'm still studying it.
>>
>> Is there another way to keep the xprt's ref count boosted while I/O is
>> going on?
>
> Why do you need the svc_xprt in the completion?
1. The svc_xprt contains the svcrdma_xprt, which contains the Send Queue
accounting mechanism. SQ accounting has to be updated for each completion:
the completion indicates that the SQ entry used by this Work Request is
now available, and that other callers waiting for enough SQEs can go ahead.
2. When handling a Receive completion, the incoming RPC message is enqueued
on the svc_xprt via svc_xprt_enqueue, unless RDMA Reads are needed.
3. When handling a Read completion, the incoming RPC message is enqueued on
the svc_xprt via svc_xprt_enqueue.
> Can the xpo_detach method wait for any pending completions?
So the completions would call wake_up on a waitqueue, or use a kernel
completion? That sounds more expensive than managing an atomic reference
count.
I suppose some other reference count could be used to trigger a work item
that would invoke xpo_detach.
But based on your comments, then, svc_xprt_put() was never intended to be
BH-safe.
> --b.
>
>>
>>
>>>> However, xprt_switch_put() takes a spin lock (xps_lock) which is locked
>>>> everywhere without disabling BHs.
>>>>
>>>> It looks to me like 39a9beab5acb makes svc_xprt_put() no longer BH-safe?
>>>> Not sure if svc_xprt_put() was intended to be BH-safe beforehand.
>>>>
>>>> Maybe xprt_switch_put() could be invoked in ->xpo_free, but that seems
>>>> like a temporary solution.
>>>
>>> Since xpo_free is also called from svc_xprt_put that doesn't sound like
>>> it would change anything. Or do we not trunk over RDMA for some reason?
>>
>> It's quite desirable to trunk NFS/RDMA on multiple connections, and it
>> should work just like it does for TCP, but so far it's not been tested.
--
Chuck Lever
^ permalink raw reply
* [Intel-wired-lan] [next PATCH S52-V2 15/15] Changed version from 1.6.21 to 1.6.25
From: Bimmy Pujari @ 2016-11-08 21:05 UTC (permalink / raw)
To: intel-wired-lan
In-Reply-To: <1478639119-14656-1-git-send-email-bimmy.pujari@intel.com>
Signed-off-by: Bimmy Pujari <bimmy.pujari@intel.com>
---
drivers/net/ethernet/intel/i40e/i40e_main.c | 2 +-
drivers/net/ethernet/intel/i40evf/i40evf_main.c | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c b/drivers/net/ethernet/intel/i40e/i40e_main.c
index 8443e8c..a032dfd 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_main.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_main.c
@@ -41,7 +41,7 @@ static const char i40e_driver_string[] =
#define DRV_VERSION_MAJOR 1
#define DRV_VERSION_MINOR 6
-#define DRV_VERSION_BUILD 21
+#define DRV_VERSION_BUILD 25
#define DRV_VERSION __stringify(DRV_VERSION_MAJOR) "." \
__stringify(DRV_VERSION_MINOR) "." \
__stringify(DRV_VERSION_BUILD) DRV_KERN
diff --git a/drivers/net/ethernet/intel/i40evf/i40evf_main.c b/drivers/net/ethernet/intel/i40evf/i40evf_main.c
index ca85021..c0fc533 100644
--- a/drivers/net/ethernet/intel/i40evf/i40evf_main.c
+++ b/drivers/net/ethernet/intel/i40evf/i40evf_main.c
@@ -38,7 +38,7 @@ static const char i40evf_driver_string[] =
#define DRV_VERSION_MAJOR 1
#define DRV_VERSION_MINOR 6
-#define DRV_VERSION_BUILD 21
+#define DRV_VERSION_BUILD 25
#define DRV_VERSION __stringify(DRV_VERSION_MAJOR) "." \
__stringify(DRV_VERSION_MINOR) "." \
__stringify(DRV_VERSION_BUILD) \
--
2.4.11
^ permalink raw reply related
* [Intel-wired-lan] [next PATCH S52-V2 14/15] i40e: Blink LED on 1G BaseT boards
From: Bimmy Pujari @ 2016-11-08 21:05 UTC (permalink / raw)
To: intel-wired-lan
In-Reply-To: <1478639119-14656-1-git-send-email-bimmy.pujari@intel.com>
From: Henry Tieman <henry.w.tieman@intel.com>
Before this patch "ethtool -p" was not blinking the LEDs on boards
with 1G BaseT PHYs.
This commit identifies 1G BaseT boards as having the LEDs connected
to the MAC. Also, renamed the flag to be more descriptive of usage.
The flag is now I40E_FLAG_PHY_CONTROLS_LEDS.
Signed-off-by: Harshitha Ramamurthy <harshitha.ramamurthy@intel.com>
Signed-off-by: Henry Tieman <henry.w.tieman@intel.com>
Change-ID: I4eb741da9780da7849ddf2dc4c0cb27ffa42a801
---
drivers/net/ethernet/intel/i40e/i40e.h | 2 +-
drivers/net/ethernet/intel/i40e/i40e_ethtool.c | 10 +++++-----
drivers/net/ethernet/intel/i40e/i40e_main.c | 4 ++--
3 files changed, 8 insertions(+), 8 deletions(-)
diff --git a/drivers/net/ethernet/intel/i40e/i40e.h b/drivers/net/ethernet/intel/i40e/i40e.h
index 06e3c23..b8f2978 100644
--- a/drivers/net/ethernet/intel/i40e/i40e.h
+++ b/drivers/net/ethernet/intel/i40e/i40e.h
@@ -356,7 +356,7 @@ struct i40e_pf {
#define I40E_FLAG_NO_DCB_SUPPORT BIT_ULL(45)
#define I40E_FLAG_USE_SET_LLDP_MIB BIT_ULL(46)
#define I40E_FLAG_STOP_FW_LLDP BIT_ULL(47)
-#define I40E_FLAG_HAVE_10GBASET_PHY BIT_ULL(48)
+#define I40E_FLAG_PHY_CONTROLS_LEDS BIT_ULL(48)
#define I40E_FLAG_PF_MAC BIT_ULL(50)
#define I40E_FLAG_TRUE_PROMISC_SUPPORT BIT_ULL(51)
#define I40E_FLAG_HAVE_CRT_RETIMER BIT_ULL(52)
diff --git a/drivers/net/ethernet/intel/i40e/i40e_ethtool.c b/drivers/net/ethernet/intel/i40e/i40e_ethtool.c
index 76753e1..6ba0035 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_ethtool.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_ethtool.c
@@ -1890,7 +1890,7 @@ static int i40e_set_phys_id(struct net_device *netdev,
switch (state) {
case ETHTOOL_ID_ACTIVE:
- if (!(pf->flags & I40E_FLAG_HAVE_10GBASET_PHY)) {
+ if (!(pf->flags & I40E_FLAG_PHY_CONTROLS_LEDS)) {
pf->led_status = i40e_led_get(hw);
} else {
i40e_aq_set_phy_debug(hw, I40E_PHY_DEBUG_ALL, NULL);
@@ -1900,20 +1900,20 @@ static int i40e_set_phys_id(struct net_device *netdev,
}
return blink_freq;
case ETHTOOL_ID_ON:
- if (!(pf->flags & I40E_FLAG_HAVE_10GBASET_PHY))
+ if (!(pf->flags & I40E_FLAG_PHY_CONTROLS_LEDS))
i40e_led_set(hw, 0xf, false);
else
ret = i40e_led_set_phy(hw, true, pf->led_status, 0);
break;
case ETHTOOL_ID_OFF:
- if (!(pf->flags & I40E_FLAG_HAVE_10GBASET_PHY))
+ if (!(pf->flags & I40E_FLAG_PHY_CONTROLS_LEDS))
i40e_led_set(hw, 0x0, false);
else
ret = i40e_led_set_phy(hw, false, pf->led_status, 0);
break;
case ETHTOOL_ID_INACTIVE:
- if (!(pf->flags & I40E_FLAG_HAVE_10GBASET_PHY)) {
- i40e_led_set(hw, false, pf->led_status);
+ if (!(pf->flags & I40E_FLAG_PHY_CONTROLS_LEDS)) {
+ i40e_led_set(hw, pf->led_status, false);
} else {
ret = i40e_led_set_phy(hw, false, pf->led_status,
(pf->phy_led_val |
diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c b/drivers/net/ethernet/intel/i40e/i40e_main.c
index eddb14c..8443e8c 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_main.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_main.c
@@ -11368,8 +11368,8 @@ static int i40e_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
pf->main_vsi_seid);
if ((pf->hw.device_id == I40E_DEV_ID_10G_BASE_T) ||
- (pf->hw.device_id == I40E_DEV_ID_10G_BASE_T4))
- pf->flags |= I40E_FLAG_HAVE_10GBASET_PHY;
+ (pf->hw.device_id == I40E_DEV_ID_10G_BASE_T4))
+ pf->flags |= I40E_FLAG_PHY_CONTROLS_LEDS;
if (pf->hw.device_id == I40E_DEV_ID_SFP_I_X722)
pf->flags |= I40E_FLAG_HAVE_CRT_RETIMER;
/* print a string summarizing features */
--
2.4.11
^ permalink raw reply related
* [Intel-wired-lan] [next PATCH S52-V2 13/15] i40e: remove code to handle dev_addr specially
From: Bimmy Pujari @ 2016-11-08 21:05 UTC (permalink / raw)
To: intel-wired-lan
In-Reply-To: <1478639119-14656-1-git-send-email-bimmy.pujari@intel.com>
From: Jacob Keller <jacob.e.keller@intel.com>
The netdev->dev_addr MAC filter already exists in the
MAC/VLAN hash table, as it is added when we configure
the netdev in i40e_configure_netdev. Because we already
know that this address will be updated in the
hash_for_each loops, we do not need to handle it
specially. This removes duplicate code and simplifies
the i40e_vsi_add_vlan and i40e_vsi_kill_vlan functions.
Because we know these filters must be part of the
MAC/VLAN hash table, this should not have any functional
impact on what filters are included and is merely a code
simplification.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Change-ID: I5e648302dbdd7cc29efc6d203b7019c11f0b5705
---
Testing-hints:
Ensure that adding and removing VLANs correctly updates the filters
for the MAC address of the device.
drivers/net/ethernet/intel/i40e/i40e_main.c | 43 +++++------------------------
1 file changed, 7 insertions(+), 36 deletions(-)
diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c b/drivers/net/ethernet/intel/i40e/i40e_main.c
index ef6eb19..eddb14c 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_main.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_main.c
@@ -2503,17 +2503,6 @@ int i40e_vsi_add_vlan(struct i40e_vsi *vsi, s16 vid)
/* Locked once because all functions invoked below iterates list*/
spin_lock_bh(&vsi->mac_filter_hash_lock);
- if (vsi->netdev) {
- add_f = i40e_add_filter(vsi, vsi->netdev->dev_addr, vid);
- if (!add_f) {
- dev_info(&vsi->back->pdev->dev,
- "Could not add vlan filter %d for %pM\n",
- vid, vsi->netdev->dev_addr);
- spin_unlock_bh(&vsi->mac_filter_hash_lock);
- return -ENOMEM;
- }
- }
-
hash_for_each_safe(vsi->mac_filter_hash, bkt, h, f, hlist) {
if (f->state == I40E_FILTER_REMOVE)
continue;
@@ -2527,28 +2516,14 @@ int i40e_vsi_add_vlan(struct i40e_vsi *vsi, s16 vid)
}
}
- /* Now if we add a vlan tag, make sure to check if it is the first
- * tag (i.e. a "tag" -1 does exist) and if so replace the -1 "tag"
- * with 0, so we now accept untagged and specified tagged traffic
- * (and not all tags along with untagged)
+ /* When we add a new VLAN filter, we need to make sure that all existing
+ * filters which are marked as vid=-1 (I40E_VLAN_ANY) are converted to
+ * vid=0. The simplest way is just search for all filters marked as
+ * vid=-1 and replace them with vid=0. This converts all filters that
+ * were marked to receive all traffic (tagged or untagged) into
+ * filters to receive only untagged traffic, so that we don't receive
+ * tagged traffic for VLANs which we have not configured.
*/
- if (vid > 0 && vsi->netdev) {
- del_f = i40e_find_filter(vsi, vsi->netdev->dev_addr,
- I40E_VLAN_ANY);
- if (del_f) {
- __i40e_del_filter(vsi, del_f);
- add_f = i40e_add_filter(vsi, vsi->netdev->dev_addr, 0);
- if (!add_f) {
- dev_info(&vsi->back->pdev->dev,
- "Could not add filter 0 for %pM\n",
- vsi->netdev->dev_addr);
- spin_unlock_bh(&vsi->mac_filter_hash_lock);
- return -ENOMEM;
- }
- }
- }
-
- /* Do not assume that I40E_VLAN_ANY should be reset to VLAN 0 */
if (vid > 0 && !vsi->info.pvid) {
hash_for_each_safe(vsi->mac_filter_hash, bkt, h, f, hlist) {
if (f->state == I40E_FILTER_REMOVE)
@@ -2585,7 +2560,6 @@ int i40e_vsi_add_vlan(struct i40e_vsi *vsi, s16 vid)
**/
void i40e_vsi_kill_vlan(struct i40e_vsi *vsi, s16 vid)
{
- struct net_device *netdev = vsi->netdev;
struct i40e_mac_filter *f;
struct hlist_node *h;
int bkt;
@@ -2593,9 +2567,6 @@ void i40e_vsi_kill_vlan(struct i40e_vsi *vsi, s16 vid)
/* Locked once because all functions invoked below iterates list */
spin_lock_bh(&vsi->mac_filter_hash_lock);
- if (vsi->netdev)
- i40e_del_filter(vsi, netdev->dev_addr, vid);
-
hash_for_each_safe(vsi->mac_filter_hash, bkt, h, f, hlist) {
if (f->vlan == vid)
__i40e_del_filter(vsi, f);
--
2.4.11
^ permalink raw reply related
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.