Allow to enslave devices when the bonding device is not up. Over the discussion
held at the previous post this seemed to be the most clean way to go, where it
is not expected to cause instabilities.
Normally, the bonding driver is UP before any enslavement takes place.
Once a netdevice is UP, the network stack acts to have it join some multicast groups
(eg the all-hosts 224.0.0.1). Now, since ether_setup() have set the bonding device
type to be ARPHRD_ETHER and address len to be ETHER_ALEN, the net core code
computes a wrong multicast link address. This is b/c ip_eth_mc_map() is called
where for multicast joins taking place after the enslavement another ip_xxx_mc_map()
is called (eg ip_ib_mc_map() when the bond type is ARPHRD_INFINIBAND)
Signed-off-by: Moni Shoua <monis at voltaire.com>
Signed-off-by: Or Gerlitz <ogerlitz at voltaire.com>
Acked-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
This patch allows for enslaving netdevices which do not support
the set_mac_address() function. In that case the bond mac address is the one
of the active slave, where remote peers are notified on the mac address
(neighbour) change by Gratuitous ARP sent by bonding when fail-over occurs
(this is already done by the bonding code).
Signed-off-by: Moni Shoua <monis at voltaire.com>
Signed-off-by: Or Gerlitz <ogerlitz at voltaire.com>
Acked-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
This patch changes some of the bond netdevice attributes and functions
to be that of the active slave for the case of the enslaved device not being
of ARPHRD_ETHER type. Basically it overrides those setting done by ether_setup(),
which are netdevice **type** dependent and hence might be not appropriate for
devices of other types. It also enforces mutual exclusion on bonding slaves
from dissimilar ether types, as was concluded over the v1 discussion.
IPoIB (see Documentation/infiniband/ipoib.txt) MAC address is made of a 3 bytes
IB QP (Queue Pair) number and 16 bytes IB port GID (Global ID) of the port this
IPoIB device is bounded to. The QP is a resource created by the IB HW and the
GID is an identifier burned into the HCA (i have omitted here some details which
are not important for the bonding RFC).
Signed-off-by: Moni Shoua <monis at voltaire.com>
Signed-off-by: Or Gerlitz <ogerlitz at voltaire.com>
Acked-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
For the operations
get-tx-csum
get-sg
get-tso
get-ufo
the default ethtool_op_xxx behavior is fine for all drivers, so we
permit op==NULL to imply the default behavior.
This provides a more uniform behavior across all drivers, eliminating
ethtool(8) "ioctl not supported" errors on older drivers that had
not been updated for the latest sub-ioctls.
The ethtool_op_xxx() functions are left exported, in case anyone
wishes to call them directly from a driver-private implementation --
a not-uncommon case. Should an ethtool_op_xxx() helper remain unused
for a while, except by net/core/ethtool.c, we can un-export it at a
later date.
[ Resolved conflicts with set/get value ethtool patch... -DaveM ]
Signed-off-by: Jeff Garzik <jeff@garzik.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
It's been a useless no-op for long enough in 2.6 so I figured it's time to
remove it. The number of people that could object because they're
maintaining unified 2.4 and 2.6 drivers is probably rather small.
[ Handled drivers added by netdev tree and some missed IRDA cases... -DaveM ]
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch cleans up duplicate includes in
drivers/net/
Signed-off-by: Jesper Juhl <jesper.juhl@gmail.com>
Acked-by: "John W. Linville" <linville@tuxdriver.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
This patch makes most of the generic device layer network
namespace safe. This patch makes dev_base_head a
network namespace variable, and then it picks up
a few associated variables. The functions:
dev_getbyhwaddr
dev_getfirsthwbytype
dev_get_by_flags
dev_get_by_name
__dev_get_by_name
dev_get_by_index
__dev_get_by_index
dev_ioctl
dev_ethtool
dev_load
wireless_process_ioctl
were modified to take a network namespace argument, and
deal with it.
vlan_ioctl_set and brioctl_set were modified so their
hooks will receive a network namespace argument.
So basically anthing in the core of the network stack that was
affected to by the change of dev_base was modified to handle
multiple network namespaces. The rest of the network stack was
simply modified to explicitly use &init_net the initial network
namespace. This can be fixed when those components of the network
stack are modified to handle multiple network namespaces.
For now the ifindex generator is left global.
Fundametally ifindex numbers are per namespace, or else
we will have corner case problems with migration when
we get that far.
At the same time there are assumptions in the network stack
that the ifindex of a network device won't change. Making
the ifindex number global seems a good compromise until
the network stack can cope with ifindex changes when
you change namespaces, and the like.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Every user of the network device notifiers is either a protocol
stack or a pseudo device. If a protocol stack that does not have
support for multiple network namespaces receives an event for a
device that is not in the initial network namespace it quite possibly
can get confused and do the wrong thing.
To avoid problems until all of the protocol stacks are converted
this patch modifies all netdev event handlers to ignore events on
devices that are not in the initial network namespace.
As the rest of the code is made network namespace aware these
checks can be removed.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch modifies every packet receive function
registered with dev_add_pack() to drop packets if they
are not from the initial network namespace.
This should ensure that the various network stacks do
not receive packets in a anything but the initial network
namespace until the code has been converted and is ready
for them.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch makes /proc/net per network namespace. It modifies the global
variables proc_net and proc_net_stat to be per network namespace.
The proc_net file helpers are modified to take a network namespace argument,
and all of their callers are fixed to pass &init_net for that argument.
This ensures that all of the /proc/net files are only visible and
usable in the initial network namespace until the code behind them
has been updated to be handle multiple network namespaces.
Making /proc/net per namespace is necessary as at least some files
in /proc/net depend upon the set of network devices which is per
network namespace, and even more files in /proc/net have contents
that are relevant to a single network namespace.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
http://bugzilla.kernel.org/show_bug.cgi?id=8797 shows that the
bonding driver may produce bogus combinations of the checksum
flags and SG/TSO.
For example, if you bond devices with NETIF_F_HW_CSUM and
NETIF_F_IP_CSUM you'll end up with a bonding device that
has neither flag set. If both have TSO then this produces
an illegal combination.
The bridge device on the other hand has the correct code to
deal with this.
In fact, the same code can be used for both. So this patch
moves that logic into net/core/dev.c and uses it for both
bonding and bridging.
In the process I've made small adjustments such as only
setting GSO_ROBUST if at least one constituent device
supports it.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
During the transition to the ethtool_ops way of doing things, we supported
calling the device's ->do_ioctl method to allow unconverted drivers to
continue working. Those days are long behind us, all in-tree drivers
use the ethtool_ops way, and so we no longer need to support this.
The bonding driver is the biggest beneficiary of this; it no longer
needs to call ioctl() as a fallback if ethtool_ops aren't supported.
Also put a proper copyright statement on ethtool.c.
Signed-off-by: Matthew Wilcox <matthew@wil.cx>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Cc: Chad Tindel <ctindel@users.sourceforge.net>
Cc: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
At present, when a device is enslaved to bonding, if ipv6 is
active then addrconf will be initated on the slave (because it is closed
then opened during the enslavement processing). This causes DAD and RS
packets to be sent from the slave. These packets in turn can confuse
switches that perform ipv6 snooping, causing them to incorrectly update
their forwarding tables (if, e.g., the slave being added is an inactve
backup that won't be used right away) and direct traffic away from the
active slave to a backup slave (where the incoming packets will be
dropped).
This patch alters the behavior so that addrconf will only run on
the master device itself. I believe this is logically correct, as it
prevents slaves from having an IPv6 identity independent from the
master. This is consistent with the IPv4 behavior for bonding.
This is accomplished by (a) having bonding set IFF_SLAVE sooner
in the enslavement processing than currently occurs (before open, not
after), and (b) having ipv6 addrconf ignore UP and CHANGE events on
slave devices.
The eql driver also uses the IFF_SLAVE flag. I inspected eql,
and I believe this change is reasonable for its usage of IFF_SLAVE, but
I did not test it.
Signed-off-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
Modify carrier state determination for 802.3ad mode to comply
with section 43.3.9 of IEEE 802.3, which requires that "Links that are
not successful candidates for aggregation (e.g., links that are attached
to other devices that cannot perform aggregation or links that have been
manually configured to be non-aggregatable) are enabled to operate as
individual IEEE 802.3 links."
Bug reported by Laurent Chavey <chavey@google.com>. This patch
is an updated version of his patch that changes the wording of
commentary and adds an update to the driver version.
Signed-off-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: Laurent Chavey <chavey@google.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
The following patch (based on a patch from Stephen Hemminger
<shemminger@linux-foundation.org>) removes use after free conditions in
the unregister path for the bonding master. Without this patch, an
operation of the form "echo -bond0 > /sys/class/net/bonding_masters"
would trigger a NULL pointer dereference in sysfs. I was not able to
induce the failure with the non-sysfs code path, but for consistency I
updated that code as well.
I also did some testing of the bonding /proc file being open
while the bond is being deleted, and didn't see any problems there.
Signed-off-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
Herbert Xu conviced me that a new flag was overkill; every driver
currently overrides get_stats, so we might as well make the internal
one the default. If someone did fail to set get_stats, they would now
get all 0 stats instead of "No statistics available".
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Acked-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
Network drivers which keep stats allocate their own stats structure
then write a get_stats() function to return them. It would be nice if
this were done by default.
1) Add a new "stats" field to "struct net_device".
2) Add a new feature field to say "this driver uses the internal one"
3) Have a default "get_stats" which returns NULL if that feature not set.
4) Change callers to check result of get_stats call for NULL, not if
->get_stats is set.
This should not break backwards compatibility with older drivers, yet
allow modern drivers to shed some boilerplate code.
Lightly tested: works for a modified lguest network driver.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
Renaming skb->h to skb->transport_header, skb->nh to skb->network_header and
skb->mac to skb->mac_header, to match the names of the associated helpers
(skb[_[re]set]_{transport,network,mac}_header).
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Now the skb->nh union has just one member, .raw, i.e. it is just like the
skb->mac union, strange, no? I'm just leaving it like that till the transport
layer is done with, when we'll rename skb->mac.raw to skb->mac_header (or
->mac_header_offset?), ditto for ->{h,nh}.
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
For the places where we need a pointer to the network header, it is still legal
to touch skb->nh.raw directly if just adding to, subtracting from or setting it
to another layer header.
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
For consistency with all the other skb->nh.raw accessors.
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
For the common, open coded 'skb->mac.raw = skb->data' operation, so that we can
later turn skb->mac.raw into a offset, reducing the size of struct sk_buff in
64bit land while possibly keeping it as a pointer on 32bit.
This one touches just the most simple case, next will handle the slightly more
"complex" cases.
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In active-backup mode, the current bonding code duplicates IGMP
traffic to all slaves, so that switches are up to date in case of a
failover from an active to a backup interface. If bonding then fails
back to the original active interface, it is likely that the "active
slave" switch's IGMP forwarding for the port will be out of date until
some event occurs to refresh the switch (e.g., a membership query).
This patch alters the behavior of bonding to no longer flood
IGMP to all ports, and to issue IGMP JOINs to the newly active port at
the time of a failover. This insures that switches are kept up to date
for all cases.
"GOELLESCH Niels" <niels.goellesch@eurocontrol.int> originally
reported this problem, and included a patch. His original patch was
modified by Jay Vosburgh to additionally remove the existing IGMP flood
behavior, use RCU, streamline code paths, fix trailing white space, and
adjust for style.
Signed-off-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
The ARP validation code only needs ARPs for the bonding device.
Signed-off-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
Bonding can erroneously register the same packet_type to receive
ARPs (for use by ARP validation): once at device open time, and once via
sysfs. Since sysfs can change the validate setting (and thus register
or unregister) at any time, a flag is needed to synchronize with device
open in order to avoid double registrations, and the simplest place is
within the packet_type structure itself. Double unregister is not an
issue.
Bug reported by Ulrich Oelmann <ulrich.oelmann@web.de>.
Signed-off-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
This patch splits the vlan_group struct into a multi-allocated struct. On
x86_64, the size of the original struct is a little more than 32KB, causing
a 4-order allocation, which is prune to problems caused by buddy-system
external fragmentation conditions.
I couldn't just use vmalloc() because vfree() cannot be called in the
softirq context of the RCU callback.
Signed-off-by: Dan Aloni <da-x@monatomic.org>
Acked-by: Jeff Garzik <jeff@garzik.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
After Al Viro (finally) succeeded in removing the sched.h #include in module.h
recently, it makes sense again to remove other superfluous sched.h includes.
There are quite a lot of files which include it but don't actually need
anything defined in there. Presumably these includes were once needed for
macros that used to live in sched.h, but moved to other header files in the
course of cleaning it up.
To ease the pain, this time I did not fiddle with any header files and only
removed #includes from .c-files, which tend to cause less trouble.
Compile tested against 2.6.20-rc2 and 2.6.20-rc2-mm2 (with offsets) on alpha,
arm, i386, ia64, mips, powerpc, and x86_64 with allnoconfig, defconfig,
allmodconfig, and allyesconfig as well as a few randconfigs on x86_64 and all
configs in arch/arm/configs on arm. I also checked that no new warnings were
introduced by the patch (actually, some warnings are removed that were emitted
by unnecessarily included header files).
Signed-off-by: Tim Schmielau <tim@physik3.uni-rostock.de>
Acked-by: Russell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Many struct file_operations in the kernel can be "const". Marking them const
moves these to the .rodata section, which avoids false sharing with potential
dirty data. In addition it'll catch accidental writes at compile time to
these shared resources.
Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Replace kmalloc() + memset() pairs with the appropriate kzalloc() calls in
the bonding driver.
Signed-off-by: Joe Jin <lkmaillist@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
* master.kernel.org:/pub/scm/linux/kernel/git/gregkh/driver-2.6: (28 commits)
sysfs: Shadow directory support
Driver Core: Increase the default timeout value of the firmware subsystem
Driver core: allow to delay the uevent at device creation time
Driver core: add device_type to struct device
Driver core: add uevent vars for devices of a class
SYSFS: Fix missing include of list.h in sysfs.h
HOWTO: Add a reference to Harbison and Steele
sysfs: error handling in sysfs, fill_read_buffer()
kobject: kobject_put cleanup
sysfs: kobject_put cleanup
sysfs: suppress lockdep warnings
Driver core: fix race in sysfs between sysfs_remove_file() and read()/write()
driver core: Change function call order in device_bind_driver().
driver core: Don't stop probing on ->probe errors.
driver core fixes: device_register() retval check in platform.c
driver core fixes: make_class_name() retval checks
/sys/modules/*/holders
USB: add the sysfs driver name to all modules
SERIO: add the sysfs driver name to all modules
PCI: add the sysfs driver name to all modules
...
This lets the network core have the ability to handle suspend/resume
issues, if it wants to.
Thanks to Frederik Deweerdt <frederik.deweerdt@gmail.com> for the arm
driver fixes.
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The existing code would blindly attempt to create the
bonding_masters file (in /sys/class/net) every time the module was
loaded. When the module is loaded multiple times (which is the
historical method used by initscripts and sysconfig to create multiple
bonding interfaces), this caused load failure of the second module load
attempt, as the creation request would fail.
This changes the code to note the failure, arrange to not remove
the bonding_masters file upon module exit, and then return success.
Bonding interfaces created by the second or subsequent loads of
the module will not exist in bonding_masters. This is not a significant
change, as previously only the interfaces from the most recent load of
the module would be listed. Both situations are less than optimal, but
this case permits compatibility with existing distro configuration
scripts, and is consistent.
Note that previously, the sysfs create request would overwrite
the exsting bonding_masters file and succeed, allowing multiple loads of
the module. The sysfs code has recently changed to return an error if
the file being created already exists.
Patrick McHardy <kaber@trash.net>, who reported this problem,
observed crashes on the old kernel (before sysfs checked for
duplicates). I did not experience such crashes, but this change should
resolve them.
Signed-off-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
The existing code did not correctly handle failures to create
the per-interface sysfs group for bonding.
Modified code to notice errors, and correctly unwind.
Signed-off-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
The code to select names for the bonding interfaces was, for the
non-sysfs creation case, always using a hard-coded set of bond0, bond1,
etc, up to max_bonds. This caused conflicts for the second or
subsequent loads of the module.
Changed the code to obtain device names from dev_alloc_name().
Signed-off-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
"extern inline" generates a warning with -Wmissing-prototypes and I'm
currently working on getting the kernel cleaned up for adding this to
the CFLAGS since it will help us to avoid a nasty class of runtime
errors.
If there are places that really need a forced inline, __always_inline
would be the correct solution.
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
While working with the latest bonding code I noticed a nasty problem that
will prevent arp monitoring from always functioning correctly on x86_64
systems. Comparing ints to longs and expecting reliable results on x86_64
is a bad idea. With this patch, arp monitoring works correctly again.
Signed-off-by: Andy Gospodarek <andy@greyhouse.net>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Stephen Hemminger <shemminger@osdl.org>
Cc: Jeff Garzik <jeff@garzik.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
This is a small fix-up to finish out the work done by Jay Vosburgh to add
carrier-state support for bonding devices. The output in
/proc/net/bonding/bondX was correct, but when collecting the same info via
an iotcl it could still be incorrect.
Signed-off-by: Andy Gospodarek <andy@greyhouse.net>
Cc: Jeff Garzik <jeff@garzik.org>
Cc: Stephen Hemminger <shemminger@osdl.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
Bonding driver unconditionnaly dereference get_stats function pointer
for each of its slave device. This patch
- adds a check for NULL dev->get_stats pointer in bond_get_stats
- prints a notice when the bonding device enslave a device without
get_stats function.
Signed-off-by: Laurent Riffard <laurent.riffard@free.fr>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
In bond_alb_monitor the bond->curr_slave_lock write lock is taken
and then dev_set_promiscuity maybe called which can take some time,
depending on the network HW. If a network IRQ for this card come in
the softirq handler maybe try to deliver more packets which end up in
a request to the read lock of bond->curr_slave_lock -> deadlock.
This issue was found by a test lab during network stress tests, this patch
disable the softirq handler for this case and solved the issue.
Signed-off-by: Karsten Keil <kkeil@suse.de>
Acked-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
ifa_local, ifa_address, ifa_mask, ifa_broadcast and ifa_anycast are
net-endian. Annotated them and variables that are inferred to be
net-endian.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
I neglected to properly update the version number in the recent
patch series; this sets it to something reasonable.
Signed-off-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
Moni Shoua