Changes in tcp_metric hash table are protected by tcp_metrics_lock
only, not by genl_mutex
While we are at it use deref_locked() instead of rcu_dereference()
in tcp_new() to avoid unnecessary barrier, as we hold tcp_metrics_lock
as well.
Reported-by: Andrew Vagin <avagin@parallels.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Fixes: 098a697b49 ("tcp_metrics: Use a single hash table for all network namespaces.")
Reviewed-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This would fix randconfig compile error:
net/built-in.o: In function `netdev_switch_fib_ipv4_abort':
(.text+0xf7811): undefined reference to `fib_flush_external'
Also it fixes following warnings:
warning: (NET_DSA) selects NET_SWITCHDEV which has unmet direct dependencies (NET && INET)
warning: (NET_DSA_MV88E6060 && NET_DSA_MV88E6131 && NET_DSA_MV88E6123_61_65 && NET_DSA_MV88E6171 && NET_DSA_MV88E6352 && NET_DSA_BCM_SF2) selects NET_DSA which has unmet direct dependencies (NET && HAVE_NET_DSA && NET_SWITCHDEV)
Reported-by: Randy Dunlap <rdunlap@infradead.org>
Suggested-by: Alexei Starovoitov <alexei.starovoitov@gmail.com>
Signed-off-by: Jiri Pirko <jiri@resnulli.us>
Signed-off-by: David S. Miller <davem@davemloft.net>
MDIO controller on little endian Socs, e.g. ls2085a is similar to the
controller on big endian Socs, but the MDIO access is little endian,
we use I/O accessor function to handle endianness, so the driver can
run on little endian Socs. A property "little-endian" is used
in DTS to indicate the MDIO is little endian, if driver probes the
property, driver will access MDIO in little endian, otherwise, driver
works in big endian by default.
Signed-off-by: Shaohui Xie <Shaohui.Xie@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
There is a bug in xgmac_wait_until_done() which mdio_stat should be used
instead of mdio_data when checking if busy bit is cleared.
Signed-off-by: Shaohui Xie <Shaohui.Xie@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Creating a kernel socket with sock_create_kern() happens in "init_net"
namespace, however, releasing it with sk_release_kernel() occurs in
the current namespace which may be different with "init_net" namespace.
Therefore, we should guarantee that the namespace in which a kernel
socket is created is same as the socket is created.
Signed-off-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fixes the following sparse warnings:
lib/rhashtable.c:767:5: warning: context imbalance in 'rhashtable_walk_start' - wrong count at exit
lib/rhashtable.c:849:6: warning: context imbalance in 'rhashtable_walk_stop' - unexpected unlock
Fixes: f2dba9c6ff ("rhashtable: Introduce rhashtable_walk_*")
Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
In hast to fix some sparse warning, I hard-coded a fix-sized array on the stack
which is probably too big for kernel standards. Fix this by converting array
to dynamic allocation.
Signed-off-by: Scott Feldman <sfeldma@gmail.com>
Acked-by: Jiri Pirko <jiri@resnulli.us>
Signed-off-by: David S. Miller <davem@davemloft.net>
Eric Dumazet says:
====================
inet: tcp listener refactoring, part 10
We are getting close to the point where request sockets will be hashed
into generic hash table. Some followups are needed for netfilter and
will be handled in next patch series.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
reqsk_put() is the generic function that should be used
to release a refcount (and automatically call reqsk_free())
reqsk_free() might be called if refcount is known to be 0
or undefined.
refcnt is set to one in inet_csk_reqsk_queue_add()
As request socks are not yet in global ehash table,
I added temporary debugging checks in reqsk_put() and reqsk_free()
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
sock_edemux() is not used in fast path, and should
really call sock_gen_put() to save some code.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
inet_diag_fill_req() is renamed to inet_req_diag_fill()
and moved up, so that it can be called fom sk_diag_fill()
inet_diag_bc_sk() is ready to handle request socks.
inet_twsk_diag_dump() is no longer needed.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When a request socket is created, we do not cache ip route
dst entry, like for timewait sockets.
Let's use sk_fullsock() helper.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We have many places where we want to check if a socket is
not a timewait or request socket. Use a helper to avoid
hard coding this.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In case LE Secure Connections is not enabled, then the command for
returning local out-of-band data should not include the confirmation
and random value for LE SC pairing. All other fields are still valid,
but these two need to be left out. In that case it is also no needed
to generate the public/private key pair for out-of-band pairing.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
This seems to be a simple typo in the debugfs entry for the remote
out-of-band data entries.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
The variable for the out-of-band random number was badly named and
with that confusing. Just rename it to local_rand so it is clear
what value it represents.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Just for pure debugging purposes print the remote out-of-band data
that has been received and is going to be used.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
It might be a bit counterintuitive to set a 'local' flag based on remote
data. This patch adds a clarifying comment to the pairing req/rsp
handlers when setting the LOCAL_OOB flag based on the PDU received from
the remote side.
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
When we receive the remote public key, if we have remote OOB data
there's no point in sending our public key to the remote if the OOB data
doesn't match. This patch moves the test for this higher up in the
smp_cmd_public_key() function.
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
If we haven't received remote OOB data we cannot perform any special
checks on the confirm value. This patch updates the check after having
received the public key to only perform the verification if we have
remote OOB data present.
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
If the SMP Pairing Request or Response PDU received from the remote
device indicates that it has received our OOB data we should set the
SMP_FLAG_LOCAL_OOB flag.
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
There are several decisions in the SMP logic that depend not only on
whether we're doing SMP or not, but also whether local and/or remote OOB
data is present. This patch splits the existing SMP_FLAG_OOB into two
new flags to track local and remote OOB data respectively.
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
We need to store the local ra/rb value in order to verify the Check
value received from the remote. This patch adds a new 'lr' for the local
ra/rb value and makes sure it gets used when verifying the DHKey Check
PDU received from the remote.
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
The OOB public and secret key pair is different from the non-OOB pairing
procedure. SO when OOB method is in use, then use this key pair instead
of generating a new one.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Then the local out-of-band data for LE SC pairing is requested via Read
Local OOB Extended Data command, then fill in the values generated by
the smp_generate_oob function. Every call of this command will overwrite
previously generated values.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
This patch adds a smp_generate_oob function that allows to create
local out-of-band data that can be used for pairing and also provides
the confirmation and random value.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
The security manager device will require the use of AES-CMAC hash for
out-of-band data generation. This patch makes sure it is correctly
set up and available.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
The LE Secure Connections Confirmation Value and LE Secure Connections
Random Value contants are required for the out-of-band data and so
just define them.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Every Bluetooth Low Energy controller requires a local crypto context
to handle the resolvable private addresses. At the moment this is just
a single crypto context, but for out-of-band data generation it will
require an additional. To facility this, create a struct smp_dev that
will hold all the extra information. This patch is just the refactoring
in preparation for future changes.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Within the security manager, it makes sense to use kzfree instead of
kfree for all data structures. This ensures that no key material leaks
by accident.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
The HCI_CONN_REMOTE_OOB connection flag is used to indicate if the
pairing initiator has provided out-of-band data. However since that
value is no longer used in any decision making, just remove it.
It is actually unclear what purpose the OOB data present field from
the HCI IO Capability Response event serves in the first place. If
either side provided out-of-band data, then that data will be used
for pairing.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
When only the pairing initiator is providing out-of-band data, then
the receiver side was ignoring the data. For some reason the code was
checking if the initiator has received out-of-band data and only then
also provide the required inidication that the acceptor actually has
the needed data available.
For BR/EDR out-of-band pairing it is enough if one side has received
out-of-band data. There are no extra checks needed here to make this
work smoothly. The only thing that is needed is to tell the controller
if data is present (and if it is P-192 or P-256 or both) and then let
the controller actually figure out the rest.
This means the check for outgoing connection or if the initiator has
indicated data are completely pointless and are in fact actually
causing harm. The check in question is this one:
if (conn->out || test_bit(HCI_CONN_REMOTE_OOB, &conn->flags)) {
After just taking the conditional check out and always executing the
code for determining the type of out-of-band data, the pairing works
flawlessly and prodcudes authenticated link keys.
The patch itself looks more complicated due to the reformatting of the
indentation, but it essentially just a two-line change.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Scott Feldman says:
====================
switchdev: add swdev ops
v3:
- Fix missing include for DSA build
v2:
- Per Simon's review, squash some of the dependent commits into one to
make series git bisect safe.
v1:
Per discussions at netconf, move switchdev ndo ops to a new swdev_ops to
keep ndo namespace clean and maintain switchdev-related ops into one place.
There are no functional changes here; just shuffling ops around for better
organization.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Move swdev wrappers over to new swdev ops (from previous ndo ops). No
functional changes to the implementation.
Signed-off-by: Scott Feldman <sfeldma@gmail.com>
rocker: move to new swdev ops
Signed-off-by: Scott Feldman <sfeldma@gmail.com>
dsa: move to new swdev ops
Signed-off-by: Scott Feldman <sfeldma@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
As discussed at netconf, introduce swdev_ops as first step to move switchdev
ops from ndo to swdev. This will keep switchdev from cluttering up ndo ops
space.
Signed-off-by: Scott Feldman <sfeldma@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Herbert Xu says:
====================
rhashtable: Fix two bugs caused by multiple rehash preparation
While testing some new patches over the weekend I discovered a
couple of bugs in the series that had just been merged. These
two patches fix them:
1) A use-after-free in the walker that can cause crashes when
walking during a rehash.
2) When a second rehash starts during a single rhashtable_remove
call the remove may fail when it shouldn't.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
The commit 9d901bc051 ("rhashtable:
Free bucket tables asynchronously after rehash") causes gratuitous
failures in rhashtable_remove.
The reason is that it inadvertently introduced multiple rehashing
from the perspective of readers. IOW it is now possible to see
more than two tables during a single RCU critical section.
Fortunately the other reader rhashtable_lookup already deals with
this correctly thanks to c4db8848af
("rhashtable: rhashtable: Move future_tbl into struct bucket_table")
so only rhashtable_remove is broken by this change.
This patch fixes this by looping over every table from the first
one to the last or until we find the element that we were trying
to delete.
Incidentally the simple test for detecting rehashing to prevent
starting another shrinking no longer works. Since it isn't needed
anyway (the work queue and the mutex serves as a natural barrier
to unnecessary rehashes) I've simply killed the test.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
The commit c4db8848af ("rhashtable:
Move future_tbl into struct bucket_table") introduced a use-after-
free bug in rhashtable_walk_stop because it dereferences tbl after
droping the RCU read lock.
This patch fixes it by moving the RCU read unlock down to the bottom
of rhashtable_walk_stop. In fact this was how I had it originally
but it got dropped while rearranging patches because this one
depended on the async freeing of bucket_table.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add support for Hardware Filter Block (HFB) so that incoming Rx traffic
can be matched and directed to desired Rx queues.
Signed-off-by: Petri Gynther <pgynther@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Alexei Starovoitov says:
====================
bpf: allow eBPF access skb fields
V1->V2:
- refactored field access converter into common helper convert_skb_access()
used in both classic and extended BPF
- added missing build_bug_on for field 'len'
- added comment to uapi/linux/bpf.h as suggested by Daniel
- dropped exposing 'ifindex' field for now
classic BPF has a way to access skb fields, whereas extended BPF didn't.
This patch introduces this ability.
Classic BPF can access fields via negative SKF_AD_OFF offset.
Positive bpf_ld_abs N is treated as load from packet, whereas
bpf_ld_abs -0x1000 + N is treated as skb fields access.
Many offsets were hard coded over years: SKF_AD_PROTOCOL, SKF_AD_PKTTYPE, etc.
The problem with this approach was that for every new field classic bpf
assembler had to be tweaked.
I've considered doing the same for extended, but for every new field LLVM
compiler would have to be modifed. Since it would need to add a new intrinsic.
It could be done with single intrinsic and magic offset or use of inline
assembler, but neither are clean from compiler backend point of view, since
they look like calls but shouldn't scratch caller-saved registers.
Another approach was to introduce a new helper functions like bpf_get_pkt_type()
for every field that we want to access, but that is equally ugly for kernel
and slow, since helpers are calls and they are slower then just loads.
In theory helper calls can be 'inlined' inside kernel into direct loads, but
since they were calls for user space, compiler would have to spill registers
around such calls anyway. Teaching compiler to treat such helpers differently
is even uglier.
They were few other ideas considered. At the end the best seems to be to
introduce a user accessible mirror of in-kernel sk_buff structure:
struct __sk_buff {
__u32 len;
__u32 pkt_type;
__u32 mark;
__u32 queue_mapping;
};
bpf programs will do:
int bpf_prog1(struct __sk_buff *skb)
{
__u32 var = skb->pkt_type;
which will be compiled to bpf assembler as:
dst_reg = *(u32 *)(src_reg + 4) // 4 == offsetof(struct __sk_buff, pkt_type)
bpf verifier will check validity of access and will convert it to:
dst_reg = *(u8 *)(src_reg + offsetof(struct sk_buff, __pkt_type_offset))
dst_reg &= 7
since 'pkt_type' is a bitfield.
No new instructions added. LLVM doesn't need to be modified.
JITs don't change and verifier already knows when it accesses 'ctx' pointer.
The only thing needed was to convert user visible offset within __sk_buff
to kernel internal offset within sk_buff.
For 'len' and other fields conversion is trivial.
Converting 'pkt_type' takes 2 or 3 instructions depending on endianness.
More fields can be exposed by adding to the end of the 'struct __sk_buff'.
Like vlan_tci and others can be added later.
When pkt_type field is moved around, goes into different structure, removed or
its size changes, the function convert_skb_access() would need to updated and
it will cover both classic and extended.
Patch 2 updates examples to demonstrates how fields are accessed and
adds new tests for verifier, since it needs to detect a corner case when
attacker is using single bpf instruction in two branches with different
register types.
The 4 fields of __sk_buff are already exposed to user space via classic bpf and
I believe they're useful in extended as well.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
- modify sockex1 example to count number of bytes in outgoing packets
- modify sockex2 example to count number of bytes and packets per flow
- add 4 stress tests that exercise 'skb->field' code path of verifier
Signed-off-by: Alexei Starovoitov <ast@plumgrid.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
introduce user accessible mirror of in-kernel 'struct sk_buff':
struct __sk_buff {
__u32 len;
__u32 pkt_type;
__u32 mark;
__u32 queue_mapping;
};
bpf programs can do:
int bpf_prog(struct __sk_buff *skb)
{
__u32 var = skb->pkt_type;
which will be compiled to bpf assembler as:
dst_reg = *(u32 *)(src_reg + 4) // 4 == offsetof(struct __sk_buff, pkt_type)
bpf verifier will check validity of access and will convert it to:
dst_reg = *(u8 *)(src_reg + offsetof(struct sk_buff, __pkt_type_offset))
dst_reg &= 7
since skb->pkt_type is a bitfield.
Signed-off-by: Alexei Starovoitov <ast@plumgrid.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Daniel Borkmann says:
====================
eBPF updates
Two small eBPF helper additions to better match up with ancillary
classic BPF functionality.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch adds the possibility to obtain raw_smp_processor_id() in
eBPF. Currently, this is only possible in classic BPF where commit
da2033c282 ("filter: add SKF_AD_RXHASH and SKF_AD_CPU") has added
facilities for this.
Perhaps most importantly, this would also allow us to track per CPU
statistics with eBPF maps, or to implement a poor-man's per CPU data
structure through eBPF maps.
Example function proto-type looks like:
u32 (*smp_processor_id)(void) = (void *)BPF_FUNC_get_smp_processor_id;
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
This work is similar to commit 4cd3675ebf ("filter: added BPF
random opcode") and adds a possibility for packet sampling in eBPF.
Currently, this is only possible in classic BPF and useful to
combine sampling with f.e. packet sockets, possible also with tc.
Example function proto-type looks like:
u32 (*prandom_u32)(void) = (void *)BPF_FUNC_get_prandom_u32;
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Claudiu Manoil says:
====================
gianfar: ARM port driver updates (2/2)
The 2nd round of driver updates to make gianfar portable on ARM,
for the ARM based SoC that integrates eTSEC - "ls1021a".
The patches address the bulk of remaining endianess issues -
handling DMA fields (BD and FCB), and device tree properties.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Use of_property_read*() to get arch endian consistent
property values. Do some refactoring in the process.
Signed-off-by: Jingchang Lu <jingchang.lu@freescale.com>
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Eric Dumazet