global mutex, accept_mtx, which serializes access to the following
fields across all sockets:
so_qlen so_incqlen so_qstate
so_comp so_incomp so_list
so_head
While providing only coarse granularity, this approach avoids lock
order issues between sockets by avoiding ownership of the fields
by a specific socket and its per-socket mutexes.
While here, rewrite soclose(), sofree(), soaccept(), and
sonewconn() to add assertions, close additional races and address
lock order concerns. In particular:
- Reorganize the optimistic concurrency behavior in accept1() to
always allocate a file descriptor with falloc() so that if we do
find a socket, we don't have to encounter the "Oh, there wasn't
a socket" race that can occur if falloc() sleeps in the current
code, which broke inbound accept() ordering, not to mention
requiring backing out socket state changes in a way that raced
with the protocol level. We may want to add a lockless read of
the queue state if polling of empty queues proves to be important
to optimize.
- In accept1(), soref() the socket while holding the accept lock
so that the socket cannot be free'd in a race with the protocol
layer. Likewise in netgraph equivilents of the accept1() code.
- In sonewconn(), loop waiting for the queue to be small enough to
insert our new socket once we've committed to inserting it, or
races can occur that cause the incomplete socket queue to
overfill. In the previously implementation, it was sufficient
to simply tested once since calling soabort() didn't release
synchronization permitting another thread to insert a socket as
we discard a previous one.
- In soclose()/sofree()/et al, it is the responsibility of the
caller to remove a socket from the incomplete connection queue
before calling soabort(), which prevents soabort() from having
to walk into the accept socket to release the socket from its
queue, and avoids races when releasing the accept mutex to enter
soabort(), permitting soabort() to avoid lock ordering issues
with the caller.
- Generally cluster accept queue related operations together
throughout these functions in order to facilitate locking.
Annotate new locking in socketvar.h.
the socket is on an accept queue of a listen socket. This change
renames the flags to SQ_COMP and SQ_INCOMP, and moves them to a new
state field on the socket, so_qstate, as the locking for these flags
is substantially different for the locking on the remainder of the
flags in so_state.
behaviour lost in the change from 4.x style netgraph tee nodes.
Alter the tee node to use the new method. Document the behaviour.
Step the ABI version number... old netgraph klds will refuse to load.
Better than just crashing.
Submitted by: Gleb Smirnoff <glebius@cell.sick.ru>
state. Apparently it happens when both devices try to disconnect RFCOMM
multiplexor channel at the same time.
The scenario is as follows:
- local device initiates RFCOMM connection to the remote device. This
creates both RFCOMM multiplexor channel and data channel;
- remote device terminates RFCOMM data channel (inactivity timeout);
- local device acknowledges RFCOMM data channel termination. Because
there is no more active data channels and local device has initiated
connection it terminates RFCOMM multiplexor channel;
- remote device does not acknowledges RFCOMM multiplexor channel
termination. Instead it sends its own request to terminate RFCOMM
multiplexor channel. Even though local device acknowledges RFCOMM
multiplexor channel termination the remote device still keeps
L2CAP connection open.
Because of hanging RFCOMM multiplexor channel subsequent RFCOMM
connections between local and remote devices will fail.
Reported by: Johann Hugo <jhugo@icomtek.csir.co.za>
there so there are no ABI changes);
+ replace 5 redefinitions of the IPF2AC macro with one in if_arp.h
Eventually (but before freezing the ABI) we need to get rid of
struct arpcom (initially with the help of some smart #defines
to avoid having to touch each and every driver, see below).
Apart from the struct ifnet, struct arpcom now only stores a copy
of the MAC address (ac_enaddr, but we already have another copy in
the struct ifnet -- if_addrhead), and a netgraph-specific field
which is _always_ accessed through the ifp, so it might well go
into the struct ifnet too (where, besides, there is already an entry
for AF_NETGRAPH data...)
Too bad ac_enaddr is widely referenced by all drivers. But
this can be fixed as follows:
#define ac_enaddr ac_if.the_original_ac_enaddr_in_struct_ifnet
(note that the right hand side would likely be a pointer rather than
the base address of an array.)
of an interface. No functional change.
On passing, comment an useless invocation of TAILQ_INIT(&ifp->if_addrhead)
which could probably be removed in the interest of clarity.
ethernet (tested) and FDDI (not tested). The main use for this is on ADSL (or
other ATM) connections where bridged ethernet is used, PPPoE being a prime
example.
There is no manual page as yet, I will write one shortly.
Reviewed by: harti
functions in kern_socket.c.
Rename the "canwait" field to "mflags" and pass M_WAITOK and M_NOWAIT
in from the caller context rather than "1" or "0".
Correct mflags pass into mac_init_socket() from previous commit to not
include M_ZERO.
Submitted by: sam
Introduce d_version field in struct cdevsw, this must always be
initialized to D_VERSION.
Flip sense of D_NOGIANT flag to D_NEEDGIANT, this involves removing
four D_NOGIANT flags and adding 145 D_NEEDGIANT flags.
Free approx 86 major numbers with a mostly automatically generated patch.
A number of strategic drivers have been left behind by caution, and a few
because they still (ab)use their major number.
layering violation. As pointed out, there is much better way to do this.
Sorry guys, I need to find a better way to force reviews.
Requested by: harti, julian, scottl (mentor)
Pointy hat to: pjd
It works as follows:
In every 'interval' seconds defined links are checked.
If they are non-active they will not be used by to data transfer.
No response from: julian, archie
Silent on: net@
Approved by: scottl (mentor)
problem here still to be solved: the sockaddr_hci has still a 16 byte
field for the node name. The code currently does not correctly use the
length field in the sockaddr to handle the address length, so
node names get truncated to 15 characters when put into a sockaddr_hci.
introducing a START_NOW command. This command does not send
and GET_IFINDEX message downstream (to wait for the response from
the ETHERNET node), but directly starts the sending process. This allows
one to generate traffic as input for any hook on any node.
a new bpf_mtap2 routine that does the right thing for an mbuf
and a variable-length chunk of data that should be prepended.
o while we're sweeping the drivers, use u_int32_t uniformly when
when prepending the address family (several places were assuming
sizeof(int) was 4)
o return M_ASSERTVALID to BPF_MTAP* now that all stack-allocated
mbufs have been eliminated; this may better be moved to the bpf
routines
Reviewed by: arch@ and several others
which means "always stay in the standard mode of PPPoE operation
regardless of any junk floating around."
As the referenced PR stated clearly, the old default setting of 0
was extremely dangerous because it opened a possibility for a
spurious frame not only to put down a single PPPoE node running
FreeBSD, but to plague *every* FreeBSD node in a PPPoE network in
such a way that those nodes would keep poisoning each other until
rebooted simultaneously.
PR: kern/47920
Reviewed by: Gleb Smirnoff <glebius <at> cell.sick.ru>
MFC after: 1 week
nonstandard. They differ in the values of certain fields in
the PPPoE frame. Previously, ng_pppoe would start in standard
mode, yet switch to nonstandard one upon reception of a single
nonstandard frame. After having done so, ng_pppoe would be unable
to interact with standard PPPoE peers. Thus, a DoS condition
existed that could be triggered by a buggy peer or malicious party.
Since few people have expressed their displeasure WRT this problem,
the default operation of ng_pppoe is left untouched for now. However,
a new value for the sysctl net.graph.nonstandard_pppoe is introduced,
-1, which will force ng_pppoe stay in standard mode regardless of any
bogus frames floating around.
PR: kern/47920
Submitted by: Gleb Smirnoff <glebius <at> cell.sick.ru>
MFC after: 1 week
In practice it seems that in situations of high packet loss the ACK
timeout seems to hit this maximum (perhaps inappropriately, but the
estimation algorithm is not perfect, so apparently it happens). In
any case, 10 seconds is way too high a value so lower to 1 second.
MFC after: 3 days
the MAC label referenced from 'struct socket' in the IPv4 and
IPv6-based protocols. This permits MAC labels to be checked during
network delivery operations without dereferencing inp->inp_socket
to get to so->so_label, which will eventually avoid our having to
grab the socket lock during delivery at the network layer.
This change introduces 'struct inpcb' as a labeled object to the
MAC Framework, along with the normal circus of entry points:
initialization, creation from socket, destruction, as well as a
delivery access control check.
For most policies, the inpcb label will simply be a cache of the
socket label, so a new protocol switch method is introduced,
pr_sosetlabel() to notify protocols that the socket layer label
has been updated so that the cache can be updated while holding
appropriate locks. Most protocols implement this using
pru_sosetlabel_null(), but IPv4/IPv6 protocols using inpcbs use
the the worker function in_pcbsosetlabel(), which calls into the
MAC Framework to perform a cache update.
Biba, LOMAC, and MLS implement these entry points, as do the stub
policy, and test policy.
Reviewed by: sam, bms
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
defines for these constants that include the trailing NUL byte. These
new constants have SIZ in their name instead of LEN. As soon as all
consumers in the tree are converted to use the new defines the old
defines will be put under BURN_BRIDGES.
Reviewed by: archie, julian, ru
Approved by: re (in principle)
whether or not the isr needs to hold Giant when running; Giant-less
operation is also controlled by the setting of debug_mpsafenet
o mark all netisr's except NETISR_IP as needing Giant
o add a GIANT_REQUIRED assertion to the top of netisr's that need Giant
o pickup Giant (when debug_mpsafenet is 1) inside ip_input before
calling up with a packet
o change netisr handling so swi_net runs w/o Giant; instead we grab
Giant before invoking handlers based on whether the handler needs Giant
o change netisr handling so that netisr's that are marked MPSAFE may
have multiple instances active at a time
o add netisr statistics for packets dropped because the isr is inactive
Supported by: FreeBSD Foundation
conservative lock. The problem with the lock-less algorithm is that
it suffers from the ABA problem. Running an application with funnels
a couple of 100kpkts/s through the netgraph system on a dual CPU system
with MPSAFE drivers will panic almost immediatly with the old algorithm.
It may be possible to eliminate the contention between threads that insert
free items into the list and those that get free items by using the
Michael/Scott queue algorithm that has two locks.
if_xname, if_dname, and if_dunit. if_xname is the name of the interface
and if_dname/unit are the driver name and instance.
This change paves the way for interface renaming and enhanced pseudo
device creation and configuration symantics.
Approved By: re (in principle)
Reviewed By: njl, imp
Tested On: i386, amd64, sparc64
Obtained From: NetBSD (if_xname)
even could call VOP_REVOKE() on vnodes associated with its dev_t's
has originated, but it stops right here.
If there are things people belive destroy_dev() needs to learn how to
do, please tell me about it, preferably with a reproducible test case.
Include <sys/uio.h> in bluetooth code rather than rely on <sys/vnode.h>
to do so.
The fact that some of the USB code needs to include <sys/vnode.h>
still disturbs me greatly, but I do not have time to chase that.
messages are forwarded as netgraph control messages to the node
that is connected to the manage hook. If that hook is not connected,
the event is lost. Flow control events are converted to netgraph
flow control messages and send along the hook that is connected to
the flow controlled VC. ACR change events are converted to control
messages and sent along the hook for the given VC.
