Commit Graph

2055 Commits

Author SHA1 Message Date
yar
78898ea5a8 Fix a couple of typos in a comment. 2007-05-02 11:13:48 +00:00
thompsa
49fe4717c7 Set the master flag on the right variable. 2007-05-02 08:58:28 +00:00
thompsa
e6a03215f5 Test for IFM_FDX rather than IFM_HDX as the half-duplex bit may not be set even
if the link is not full-duplex.
2007-05-02 07:52:55 +00:00
thompsa
5fc175b7b4 Rename the trunk(4) driver to lagg(4) as it is too similar to vlan trunking.
The name trunk is misused as the networking term trunk means carrying multiple
VLANs over a single connection. The IEEE standard for link aggregation (802.3
section 3) does not talk about 'trunk' at all while it is used throughout IEEE
802.1Q in describing vlans.

The lagg(4) driver provides link aggregation, failover and fault tolerance.

Discussed on:	current@
2007-04-17 00:35:11 +00:00
thompsa
feb8fc32f4 Fix a case where the multicast addresses were not removed from some ports. The
first port to be removed from the trunk would free the multicast list so
subsequent removed ports didnt have their multicast addresses removed.
2007-04-12 01:58:57 +00:00
thompsa
bcc8cf69fe Fix an uninitialized variable warning. 2007-04-10 08:02:33 +00:00
thompsa
0f00c64853 Add the trunk(4) driver for providing link aggregation, failover and fault
tolerance.  This driver allows aggregation of multiple network interfaces as
one virtual interface using a number of different protocols/algorithms.

failover    - Sends traffic through the secondary port if the master becomes
              inactive.
fec         - Supports Cisco Fast EtherChannel.
lacp        - Supports the IEEE 802.3ad Link Aggregation Control Protocol
              (LACP) and the Marker Protocol.
loadbalance - Static loadbalancing using an outgoing hash.
roundrobin  - Distributes outgoing traffic using a round-robin scheduler
              through all active ports.

This code was obtained from OpenBSD and this also includes 802.3ad LACP support
from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
glebius
2f20235351 Fix regression in rev. 1.140.
Reported by:	Yuriy Tsibizov <Yuriy.Tsibizov gfk.ru>, bsam
2007-03-27 19:36:12 +00:00
bms
441b1fffcf Fix a case where hardware removal of an interface caused an attempt to
announce an ll_ifma which has gone away. Add a KASSERT to catch regressions.

Bug found by:	Tom Uffner
2007-03-27 16:11:28 +00:00
yar
0c5dfb27b3 Fix some statements in disc(4) and about it:
- ifnet is no more embedded in softc;
- the interface name is `disc', not `ds'.
2007-03-26 09:10:28 +00:00
yar
1e010686da Give a hint that softc can contain many things besides ifp. 2007-03-26 09:05:10 +00:00
yar
9b2be94dbc We no longer embed ifnet in softc, and the pointer to ifnet
doesn't need to be first in softc now.  (It was the whole
ifnet structure itself that needed to be first in the good
old days.)  Fix the respective comment accordingly.

Add xrefs to ifnet(9) in some other comments while I'm here.

Pointed out by:		thompsa
2007-03-26 08:58:22 +00:00
yar
2e04280ece Introduce a new toy interface, edsc(4). It's a discard interface
imitating an Ethernet device, so vlan(4) and if_bridge(4) can be
attached to it for testing and benchmarking purposes.  Its source
can be an introduction to the anatomy of a network interface driver
due to its simplicity as well as to a bunch of comments in it.

(The rest of needed changes were in my previous commit, which got
interrupted in the middle.  Alas, CVS commits are not atomic.)
2007-03-26 04:50:06 +00:00
bms
317f57e707 Fix a typo, and update a comment.
Submitted by:	yar
2007-03-22 19:08:39 +00:00
glebius
060c94f5be When working on an RTM_CHANGE do the route editing in the following
sequence. First, if rt_ifa is going to be changed, then call
ifa_rtrequest(RTM_DELETE). Second, if gateway is going to be changed,
then call rt_setgate(). Third, change rt_ifa.

With this change we are able to change a link level route to a
gateway one, that wasn't possible before:

	# ifconfig em0 192.168.22.1/24
        # arp -s 192.168.22.99 00:11:22:33:44:55
        # route change 192.168.22.99 192.168.22.199
        # ping 192.168.22.99
	db>

Reported by:	avatar
2007-03-22 10:51:03 +00:00
bms
2ad7143fb2 Make the m_pullup() diagnostic message compile-time conditional on DIAGNOSTIC.
Requested by:	glebius
2007-03-20 14:29:54 +00:00
bms
a5925f917c Fix tinderbox; ng_ether needs to see if_findmulti(). 2007-03-20 03:15:43 +00:00
bms
4ffc004901 Implement reference counting for ifmultiaddr, in_multi, and in6_multi
structures. Detect when ifnet instances are detached from the network
stack and perform appropriate cleanup to prevent memory leaks.

This has been implemented in such a way as to be backwards ABI compatible.
Kernel consumers are changed to use if_delmulti_ifma(); in_delmulti()
is unable to detect interface removal by design, as it performs searches
on structures which are removed with the interface.

With this architectural change, the panics FreeBSD users have experienced
with carp and pfsync should be resolved.

Obtained from:	p4 branch bms_netdev
Reviewed by:	andre
Sponsored by:	Garance A Drosehn
Idea from:	NetBSD
MFC after:	1 month
2007-03-20 00:36:10 +00:00
thompsa
260775bb6f etherbroadcastaddr is now unused. 2007-03-19 19:20:35 +00:00
thompsa
62bb42c44c M_BCAST & M_MCAST are now set by ether_input before passing to the bridge. 2007-03-19 19:13:02 +00:00
bms
0ad9209c94 Clean up the ether_input() path by using the M_PROMISC flag.
Main points of this change:
 * Drop frames immediately if the interface is not marked IFF_UP.
 * Always trim off the frame checksum if present.
 * Always use M_VLANTAG in preference to passing 802.1Q frames
   to consumers.
 * Use __func__ consistently for KASSERT().
 * Use the M_PROMISC flag to detect situations where ether_input()
   may reenter itself on the same call graph with the same mbuf which
   was promiscuously received on behalf of subsystems such as
   netgraph, carp, and vlan.
 * 802.1P frames (that is, VLAN frames with an ID of 0) will now be
   passed to layer 3 input paths.
 * Deal with the special case for CARP in a sane way.

This is a significant rewrite of code on the critical path. Please report
any issues to me if they arise. Frames will now only pass through dummynet
if M_PROMISC is cleared, to avoid problems with re-entry.

The handling of CARP needs to be revisited architecturally. The M_PROMISC
flag may potentially be demoted to a link-layer flag only as it is in
NetBSD, where the idea originated.

Discussed on:	net
Idea from:	NetBSD
Reviewed by:	yar
MFC after:	1 month
2007-03-19 18:39:36 +00:00
bms
9d0975ba99 Add a sysctl net.link.tap.up_on_open which defaults to zero; when it
is non-zero, tap(4) instances will be marked IFF_UP on attach.

PR:		110383
Requested by:	Frank Behrens
MFC after:	2 weeks
2007-03-19 18:17:31 +00:00
yar
33d5501a32 Now <net/if_arp.h> is unused here. 2007-03-19 18:01:39 +00:00
yar
2396d74a2c Fix a nameless constant: 6 -> ETHER_ADDR_LEN
Tested with:	md5(1)
2007-03-19 16:58:07 +00:00
yar
3590987ee8 Now that this driver uses ether_ioctl(), it no longer needs
the INET related include files.
2007-03-19 16:31:28 +00:00
rik
4e46a492aa Give a chance for packet to appear with a correct input interfaces
in case of multiple interfaces with the same MAC in the same bridge.
This commit do not solve the entire problem. Only case where packet
arrived from such interface.

PR:     kern/109815
MFC after:      7 days
Submitted by:   Eygene Ryabinkin and rik@
Discussed with: bms@, thompsa@, yar@
2007-03-18 23:28:53 +00:00
yar
c9ba410fa3 Remove a spurious blank line at the start of vlan_growhash().
Add a diagnostic message to the function about resizing vlan
hash table.
2007-03-15 14:10:51 +00:00
thompsa
e1cdb72246 Properly move the setting of bstp_linkstate_p to the bridgestp module. 2007-03-14 20:55:51 +00:00
yar
f78c9d73a4 Let vlan_ioctl() pass some work on to ether_ioctl()
and so reduce code duplication a bit.
2007-03-14 14:38:04 +00:00
yar
9e5ccd324f Emit load and unload messages under bootverbose.
This can help to spot bugs (which it did for me,)
and let people know which mode the vlan module is
actually using if they suspect it isn't picking its
options from the main kernel config file.
2007-03-12 12:42:14 +00:00
yar
e732c73d0a Fix some minor issues in the internal vlan lists:
- ifv_list member of struct ifvlan is unneeded in array mode,
  it's used only in hash mode to resolve hash collisions.

- We don't need the list of trunks at all.  (The initial reason for
  having it was to be able to destroy all trunks in the MOD_UNLOAD
  handler, but a trunk is not to be destroyed forcibly -- it will
  go away when all vlan interfaces on it have been deleted.
  Note that if_clone_detach() called first of all under MOD_UNLOAD
  will delete all vlan interfaces and thus make all trunks go away
  quietly.)

- It's enough to use a single [S]LIST_FIRST() in a typical list
  destruction loop.
2007-03-12 12:27:30 +00:00
thompsa
8841e4c963 Change the passing of callbacks to a struct in case this needs to be extended in the future. 2007-03-09 19:34:55 +00:00
bms
aac8b31811 Add Ethertype for 802.3ad LACP. 2007-03-07 12:51:52 +00:00
bms
f9e43f8ad0 Prepare for 802.1p:
Add macro EVL_APPLY_VLID() which may be used to apply an 802.1q VLAN ID
 to the M_VLANTAG field in an mbuf packet header non-destructively.
 This will be used by net80211 to begin with.

 Add macro EVL_APPLY_PRI() which may be used to apply an 802.1p priority
 class to the M_VLANTAG field in an mbuf packet header non-destructively.

 Add other macros for manipulating tags and the CFI bit.

Submitted by:	Boris Kovalenko (EVL_CFIOFTAG(), EVL_MAKETAG())
2007-02-28 22:05:30 +00:00
jkim
2bd7382fdc Add three new ioctl(2) commands for bpf(4).
- BIOCGDIRECTION and BIOCSDIRECTION get or set the setting determining
whether incoming, outgoing, or all packets on the interface should be
returned by BPF.  Set to BPF_D_IN to see only incoming packets on the
interface.  Set to BPF_D_INOUT to see packets originating locally and
remotely on the interface.  Set to BPF_D_OUT to see only outgoing
packets on the interface.  This setting is initialized to BPF_D_INOUT
by default.  BIOCGSEESENT and BIOCSSEESENT are obsoleted by these but
kept for backward compatibility.

- BIOCFEEDBACK sets packet feedback mode.  This allows injected packets
to be fed back as input to the interface when output via the interface is
successful.  When BPF_D_INOUT direction is set, injected outgoing packet
is not returned by BPF to avoid duplication.  This flag is initialized to
zero by default.

Note that libpcap has been modified to support BPF_D_OUT direction for
pcap_setdirection(3) and PCAP_D_OUT direction is functional now.

Reviewed by:	rwatson
2007-02-26 22:24:14 +00:00
thompsa
82cd4163b4 Move the lock init until after if_alloc in case the allocation fails and we
free the softc and return.

MFC after:	3 days
2007-02-23 19:37:55 +00:00
csjp
6169e2ea40 Use ETHER_BPF_MTAP() instead of BPF_MTAP() here. It's possible
incoming packets have had their 802.1Q tags processed by the
hardware, resulting in them being stripped from the packets, and
placed on the mbuf.  This fixes the processing of 802.1Q tags when
hardware offload of 802.1Q tags is enabled.
2007-02-22 14:50:31 +00:00
bms
aaa1e7fb11 Fix a bug in if_findmulti(), whereby it would not find (and thus delete)
a link-layer multicast group membership.
Such memberships are needed in order to support protocols such as
IS-IS without putting the interface into PROMISC or ALLMULTI modes.

sa_equal() is not OK for comparing sockaddr_dl as it has deeper structure
than a simple byte array, so add sa_dl_equal() and use that instead.

Reviewed by:	rwatson
Verified with:	/usr/sbin/mtest
Bug found by:	Jouke Witteveen
MFC after:	2 weeks
2007-02-22 00:14:02 +00:00
rwatson
800b6fbc65 Replace a suser() check with an explicit check for PRIV_NET_SETIFMTU. 2007-02-20 15:20:36 +00:00
cognet
638a7ba075 Use __NO_STRICT_ALIGNMENT, instead of special casing ia64 and sparc64.
This fixes panics I got on arm, with struct ip aligned on 4 bytes.

MFC After:	1 week
2007-02-09 00:09:35 +00:00
bms
94de0f0fd0 Fix devfs cloning for non-superusers when net.link.tap.user_open is non-zero.
Note: 'ifconfig tapX create' still requires PRIV_NET_IFCREATE privilege.

Reviewed by:	rwatson
2007-02-05 11:29:08 +00:00
bms
ece591cff4 Clean up after tun(4) properly; remove routes whose ifp is set to
that of the tun instance even for the !AF_INET case, and properly
remove configured addresses by calling if_purgeaddrs().

Maintain the TUN_DSTADDR behaviour for compatibility with the OS/390
emulator.

MFC after:	3 weeks
PR:		100080
Reviewed by:	bz
2007-02-05 11:15:52 +00:00
bms
77c2e11309 Implement ifnet cloning for tun(4)/tap(4).
Make devfs cloning a sysctl/tunable which defaults to on.

If devfs cloning is enabled, only the super-user may create
tun(4)/tap(4)/vmnet(4) instances. Devfs cloning is still enabled by
default; it may be disabled from the loader or via sysctl with
"net.link.tap.devfs_cloning" and "net.link.tun.devfs_cloning".

Disabling its use affects potentially all tun(4)/tap(4) consumers
including OpenSSH, OpenVPN and VMware.

PR:		105228 (potentially also 90413, 105570)
Submitted by:	Landon Fuller
Tested by:	Andrej Tobola
Approved by:	core (rwatson)
MFC after:	4 weeks
2007-02-04 16:32:46 +00:00
bms
cb84e5a9bd Drop unicast Ethernet frames not destined for the configured address
of a tap(4) instance, if IFF_PROMISC is not set.

In tap(4), we should emulate the effect IFF_PROMISC would have on
hardware, otherwise we risk introducing layer 2 loops if tap(4) is
used with bridges. This means not even bpf(4) gets to see them.

This patch has been tested in a variety of situations. Multicast and
broadcast frames are correctly allowed through. I have observed this
behaviour causing problems with multiple QEMU instances hosted on
the same FreeBSD machine.

The checks in in ether_demux() [if_ethersubr.c, rev 1.222, line 638]
are insufficient to prevent this bug from occurring, as ifp->if_vlantrunk
will always be NULL for the non-vlan case.

MFC after:	3 weeks
PR:		86429
Submitted by:	Pieter de Boer (with changes)
2007-02-03 02:57:45 +00:00
bms
a6c57fe6a9 Use int instead of u_int for the 'extra' argument to the
clone_create() KPI.
This fixes a signedness bug in unit number comparisons.

Submitted by:	imp, Landon Fuller
PR:		kern/105228
MFC after:	2 weeks
2007-02-02 22:27:45 +00:00
rwatson
c8cb2f0c11 Update comment for struct bpf_d: we now store buffered packets for BPF
in malloc'd storage, not in mbuf clusters.
2007-01-29 14:41:03 +00:00
rwatson
14613c4dc7 Remove slightly dubious comment; add descriptive strings for several
sysctls.

MFC after:	3 days
2007-01-28 16:38:44 +00:00
rwatson
708b428377 Remove BSD < 199103 compatibility entries in the bpf_d structure: they are
not used in any of our code.  Also remove explicit padding variable that
kept the bpf_d structure the same size before and after the change in
select implementation, since binary compatibility is not required for this
data structure on 7-CURRENT.
2007-01-27 18:12:50 +00:00
rwatson
ebd5cdbc2e Remove now unused bpf_compat.h. This compatibility file emulates malloc(9)
using the mbuf allocator.
2007-01-27 17:32:12 +00:00
thompsa
d02694d662 Set topology change propagation on all ports _except_ the caller. 2007-01-18 07:13:01 +00:00