Nexthops for MPLS routes have a via address field sized for the
largest via address that is expected, which is 32 bytes. This means
that in the most common case of having ipv4 via addresses, 28 bytes of
memory more than required are used per nexthop. In the other common
case of an ipv6 nexthop then 16 bytes more than required are
used. With large numbers of MPLS routes this extra memory usage could
start to become significant.
To avoid allocating memory for a maximum length via address when not
all of it is required and to allow for ease of iterating over
nexthops, then the via addresses are changed to be stored in the same
memory block as the route and nexthops, but in an array after the end
of the array of nexthops. New accessors are provided to retrieve a
pointer to the via address.
To allow for O(1) access without having to store a pointer or offset
per nh, the via address for each nexthop is sized according to the
maximum via address for any nexthop in the route, which is stored in a
new route field, rt_max_alen, but this is in an existing hole in
struct mpls_route so it doesn't increase the size of the
structure. Each via address is ensured to be aligned to VIA_ALEN_ALIGN
to account for architectures that don't allow unaligned accesses.
Signed-off-by: Robert Shearman <rshearma@brocade.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fill in the via address length for the predefined IPv4 and IPv6
explicit-null label routes.
Fixes: f8efb73c97 ("mpls: multipath route support")
Signed-off-by: Robert Shearman <rshearma@brocade.com>
Acked-by: Roopa Prabhu <roopa@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch reduces the overhead of locking for busy poll.
Previously the state was protected by a lock, whereas now
it's manipulated solely with atomic operations.
Signed-off-by: Shradha Shah <sshah@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
netstamp_needed is toggled for all socket families if they request
timestamping. But some protocols don't need the lower-layer timestamping
code at all. This patch starts disabling it for af-unix.
E.g. systemd enables timestamping during boot-up on the journald af-unix
sockets, thus causing the system to globally enable timestamping in the
lower networking stack. Still, it is very probable that timestamping
gets activated, by e.g. dhclient or various NTP implementations.
Reported-by: Jesper Dangaard Brouer <brouer@redhat.com>
Signed-off-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Ariel Elior says:
====================
Add new drivers: qed & qede
This series implements the driver set for Qlogic's new QL4xxx series.
These are 10/20/25/40/50/100 Gig capable converged nics, supporting
ethernet (obviously), iscsi, fcoe, roce and iwarp protocols.
The overall driver design includes a common module ('qed') and protocol
specific dependent modules for ethernet ('qede'), fcoe ('qedf'),
iscsi ('qedi') and roce ('qedr').
The common module contains all of the common logic, e.g. initialization,
cleanup, infrastructure for interrupt handling, link management, slowpath
etc. as well as protocol agnostic features, and supplying an abstraction
layer for other modules.
The protocol specific modules can be compiled and operated independently
of each other, with the exception of the rdma modules which are dependent
on the ethernet module, in accordance with the kernel rdma stack design.
This series only adds the core and ethernet modules, with basic L2
capabilities. Future series will add the rest of the modules and enhance
the L2 functionality.
Ths patch series is constructed of the following patches:
qed: Add module with basic common support
qed: Add basic L2 interface
qede: Add basic Network driver
qed: Add slowpath L2 support
qede: Add basic network device support
qede: Add classification configuration
qed: Add link support
qede: Add support for link
qed: Add statistics support
qede: Add basic ethtool support
This project is a team effort, thanks go to Yuval Mintz, Dmitry Kravkov,
Michal Kalderon, Tomer Tayar, Manish Chopra, Sudarsana Kalluru,
Rajesh Borundia, Sony Chacko, Artum Zolotushko, Harish Patil, Rasesh Mody,
Sergey Ukhterov and Elad Manela, as well as former team members,
Eilon Greenstein and Shmulik Ravid.
Changes from previos version:
-----------------------------
From Version 7:
- Various small fixes according to Dave's suggestions; Largest change
[code-wise] - don't use tabs for indenting function arguments.
From Version 6:
- Reduced the number of arguments for functions with exceptionally
high number of parameters.
From Version 5:
- Style change and fixes [mostly in 1, 4 and 7].
Thanks go to Francois Romieu, a mere mortal. ;-)
From Version 4:
- Drop dependency for x86_64.
From Version 3:
- Limit support of initial submission to x86_64.
- Fix endian problems appearing via sparse [although no BE support yet].
- Fix small issues suggested by the kbuild test robot.
From Version 2:
- Removed U64_{HI,LO}; Using {upper,lower}_32_bits instead.
- Use regular napi weight definition.
- [We still use the __le variants for variables, since we didn't get
a reply regarding the change into non-user API types].
From Version 1:
- Removed private license file; Instead revised comments at source headers.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
This adds basic ethtool operations to the qed driver, allowing support in:
- Statistics gathering [ethtool -S]
- Setting of debug level [ethtool -s <interface> msglvl]
- Getting basic information [ethtool, ethtool -i]
In addition it adds the ability to change the MTU.
Signed-off-by: Sudarsana Kalluru <Sudarsana.Kalluru@qlogic.com>
Signed-off-by: Yuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: Ariel Elior <Ariel.Elior@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Device statistics can be gathered on-demand. This adds the qed support for
reading the statistics [both function and port] from the device, and adds
to the public API a method for requesting the current statistics.
Signed-off-by: Manish Chopra <Manish.Chopra@qlogic.com>
Signed-off-by: Yuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: Ariel Elior <Ariel.Elior@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This adds basic link functionality to qede - driver still doesn't provide
users with an API to change any link property, but it does request qed to
initialize the link using default configuration, and registers a callback
that allows it to get link notifications.
This patch adds the ability of the driver to set the carrier as active and
to enable traffic as a result of async. link notifications.
Following this patch, driver should be capable of running traffic.
Signed-off-by: Sudarsana Kalluru <Sudarsana.Kalluru@qlogic.com>
Signed-off-by: Yuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: Ariel Elior <Ariel.Elior@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Physical link is handled by the management Firmware.
This patch lays the infrastructure for attention handling in the driver,
as link change notifications arrive via async. attentions,
as well the handling of such notifications.
This patch also extends the API with the protocol drivers by adding
registered callbacks which the protocol driver passes to qed in order
to be notified of async. events originating from the FW/HW.
Signed-off-by: Yuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: Ariel Elior <Ariel.Elior@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add the ability to configure basic classification in driver by
implementing ndo_set_mac_address() and ndo_set_rx_mode().
Signed-off-by: Sudarsana Kalluru <Sudarsana.Kalluru@qlogic.com>
Signed-off-by: Yuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: Ariel Elior <Ariel.Elior@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch includes the basic Rx/Tx support for the driver [although
carrier will still never be turned on].
Following this patch the driver registers a network device, initializes
it and prepares it for traffic.
Signed-off-by: Sudarsana Kalluru <Sudarsana.Kalluru@qlogic.com>
Signed-off-by: Yuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: Ariel Elior <Ariel.Elior@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch adds to the qed the support to configure various L2 elements,
such as channels and basic filtering conditions.
It also enhances its public API to allow qede to later utilize this
functionality.
Signed-off-by: Manish Chopra <Manish.Chopra@qlogic.com>
Signed-off-by: Yuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: Ariel Elior <Ariel.Elior@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The Qlogic Everest Driver for Ethernet is the Ethernet specific module for
QL4xxx ethernet products by Qlogic.
This patch adds a very minimal PCI driver, one that doesn't yet register
a network device, but one that does interact with qed and does a basic
initialization of the HW.
Signed-off-by: Yuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: Ariel Elior <Ariel.Elior@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch adds a public API for a network driver to work on top of QED.
The interface itself is very minimal - it's mostly infrastructure, as the
only content it has after this patch is a query for HW-based information
required for the creation of a network interface [I.e., no actual
protocol-specific configurations are supported].
Signed-off-by: Manish Chopra <Manish.Chopra@qlogic.com>
Signed-off-by: Yuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: Ariel Elior <Ariel.Elior@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The Qlogic Everest Driver is the backend module for the QL4xxx ethernet
products by Qlogic.
This module serves two main purposes:
1. It's responsible to contain all the common code that will be shared
between the various drivers that would be used with said line of
products. Flows such as chip initialization and de-initialization
fall under this category.
2. It would abstract the protocol-specific HW & FW components, allowing
the protocol drivers to have a clean APIs which is detached in its
slowpath configuration from the actual HSI.
This adds a very basic module without any protocol-specific bits.
I.e., this adds a basic implementation that almost entirely falls under
the first category.
Signed-off-by: Yuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: Ariel Elior <Ariel.Elior@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Adding IPv6 for the TSO helper API is trivial:
* Don't play with the id (which doesn't exist in IPv6)
* Correctly update the payload_len (don't include the
length of the IP header itself)
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
xceiver handling (enable/disable) of the com20020 cards. The driver now handles
link status change detection. The EAE PCI-ARCNET cards now make use of the
rotary encoded subdevice indexing and got support for led triggers on transmit
and reconnection events.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1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=NNwD
-----END PGP SIGNATURE-----
Merge tag 'arcnet-for-4.4-rc1' of git://git.pengutronix.de/git/mgr/linux
Michael Grzeschik says:
====================
This series includes code simplifaction. The main changes are the correct
xceiver handling (enable/disable) of the com20020 cards. The driver now handles
link status change detection. The EAE PCI-ARCNET cards now make use of the
rotary encoded subdevice indexing and got support for led triggers on transmit
and reconnection events.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Messages like "icmp6_send: no reply to icmp error" are close
to useless. Adding source and destination addresses to provide
some more clue.
Signed-off-by: Bjørn Mork <bjorn@mork.no>
Signed-off-by: David S. Miller <davem@davemloft.net>
exported perf symbols are GPL only, mark eBPF helper functions
used in tracing as GPL only as well.
Suggested-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fix safety checks for bpf_perf_event_read():
- only non-inherited events can be added to perf_event_array map
(do this check statically at map insertion time)
- dynamically check that event is local and !pmu->count
Otherwise buggy bpf program can cause kernel splat.
Also fix error path after perf_event_attrs()
and remove redundant 'extern'.
Fixes: 35578d7984 ("bpf: Implement function bpf_perf_event_read() that get the selected hardware PMU conuter")
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Tested-by: Wang Nan <wangnan0@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While the current driver mostly supports BCM7445 which has a hardcoded
location for its MoCA port on port 7 and port 0 for its internal PHY,
this is not necessarily true for all other chips out there such as
BCM3390 for instance.
Walk the list of ports from Device Tree, get their port number ("reg"
property), and then parse the "phy-mode" property and initialize two
internal variables: moca_port and a bitmask of internal PHYs. Since we
use interrupts for the MoCA port, we introduce two helper functions to
enable/disable interrupts and do this at the appropriate bank (INTRL2_0
or INTRL2_1).
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add support for the FDB add, delete, and dump operations. The add and
delete operations are implemented using directed ARL operations using
the specified MAC address and consist in a read operation, write and
readback operation.
The dump operation consists in using the ARL search and software
filtering entries which are not for the desired port.
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Reviewed-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The arcnet device has no interrupt to detect if the link has changed
from disconnected to connected. This patch adds an timer to toggle the
link detection. The timer will get retriggered as long as the
reconnection interrupts accure. If the recon interrupts hold off
for >1s we define the connection stable again.
Signed-off-by: Michael Grzeschik <m.grzeschik@pengutronix.de>
The EAE PLX-PCI card has special leds on the the main io pci resource
bar. This patch adds support to trigger the conflict and data leds with
the packages.
Signed-off-by: Michael Grzeschik <m.grzeschik@pengutronix.de>
The EAE PLX-PCI card has a special rotary encoder
to configure the address of every card individually.
We take this information for the initial setup of
the cards dev_id.
Signed-off-by: Michael Grzeschik <m.grzeschik@pengutronix.de>
This patch sets the dev_port according to the index of
the card. This can be used by udev to name the ports
in userspace.
Signed-off-by: Michael Grzeschik <m.grzeschik@pengutronix.de>
This patch changes the driver to properly work with the linux netif
interface. The controller gets enabled on open and disabled on close.
Therefor it removes every bogus start of the xceiver. It only gets
enabled on com20020_open and disabled on com20020_close.
Signed-off-by: Michael Grzeschik <m.grzeschik@pengutronix.de>
The call for dev_free_skb is done only once. This patch
moves its call to its only user and removes the obsolete
condition variable.
Signed-off-by: Michael Grzeschik <m.grzeschik@pengutronix.de>
bnxt_gro_skb() has unused variables when CONFIG_INET is not set. We
really cannot support hardware GRO if CONFIG_INET is not set, so
compile out bnxt_gro_skb() completely and define BNXT_FLAG_GRO to be 0
if CONFIG_INET is not set. This will effectively always disable
hardware GRO if CONFIG_INET is not set.
Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This fixes the mask used to update the LED configuration so that it clears
the necessary bits as well as setting the bits according to the mask.
Also reverse the LED configuration to show the Link state + collisions in
LEDA and the Link state + TX/RX events in LEDB.
Signed-off-by: Jon Ringle <jringle@gridpoint.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Russell King says:
====================
mvneta ethtool statistics
Sorry for v3 - I forgot to update the commit message on patch 1 as
requested by Marcin.
This short series adds ethtool statistics reporting to mvneta. Having
discussed with Andrew on IRC, we decided I'd pick up his patch into my
series.
My change for patch 1 compared to the previous RFC splits out the
reading of the statistics from the hardware into a separate function,
in order to facilitate work going on elsewhere to arrange for the
statistics to be preserved across a suspend/resume cycle.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
The existing function to clear the MIB statatistics was using the
wrong address for the registers. Also, the counters would of been
cleared when the interface was brought up, not during the
probe. Fix both of these.
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add support for the ethtool statistic interface, returning the full set
of statistics which both Armada 370, 38x and Armada XP can support.
Tested-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
poll(POLLOUT) on a listener should not report fd is ready for
a write().
This would break some applications using poll() and pfd.events = -1,
as they would not block in poll()
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Alan Burlison <Alan.Burlison@oracle.com>
Tested-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
TO: "David S. Miller" <davem@davemloft.net>
CC: netdev@vger.kernel.org
CC: Jon Maloy <jon.maloy@ericsson.com>
CC: Ying Xue <ying.xue@windriver.com>
CC: tipc-discussion@lists.sourceforge.net
CC: linux-kernel@vger.kernel.org
Signed-off-by: Fengguang Wu <fengguang.wu@intel.com>
Acked-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Jon Maloy says:
====================
tipc: improve broadcast implementation
The TIPC broadcast link implementation is currently complex and hard to
follow. It also incurs some amount of code and structure duplication,
something that can be reduced significantly with a little effort.
This commit series introduces a number of improvements which address
both the locking structure, the code/structure duplication issue, and
the overall readbility of the code.
The series consists of three main parts:
1-7: Adaptation to the new link structure, and preparation for the next
step. In particular, we want the broadcast transmission link to
have a life cycle that is longer than any of its potential (unicast
and broadcast receive links) users. This eliminates the need to
always test for the presence of this link before accessing it.
8-10: This is what is really new in this series. Commit #9 is by far
the largest and most important one, because it moves most of
the broadcast functionality into link.c, partially reusing the
fields and functionality of the unicast link. The removal of
the "node_map" infrastructure in commit #10 is also an important
achievement.
11-16: Some improvements leveraging the changes made in the previous
commits.
The series needs commit 53387c4e22 ("tipc: extend broadcast link window size")
and commit e53567948f ("tipc: conditionally expand buffer headroom over udp tunnel")
which are both present in 'net' but not yet in 'net-next', to apply cleanly.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
After the previous changes in this series, we can now remove some
unused code and structures, both in the broadcast, link aggregation
and link code.
There are no functional changes in this commit.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Correct synchronization of the broadcast link at first contact between
two nodes is dependent on the assumption that the binding table "bulk"
update passes via the same link as the initial broadcast syncronization
message, i.e., via the first link that is established.
This is not guaranteed in the current implementation. If two link
come up very close to each other in time, the "bulk" may quite well
pass via the second link, and hence void the guarantee of a correct
initial synchronization before the broadcast link is opened.
This commit makes two small changes to strengthen this guarantee.
1) We let the second established link occupy slot 1 of the
"active_links" array, while the first link will retain slot 0.
(This is in reality a cosmetic change, we could just as well keep
the current, opposite order)
2) We let the name distributor always use link selector/slot 0 when
it sends it binding table updates.
The extra traffic bias on the first link caused by this change should
be negligible, since binding table updates constitutes a very small
fraction of the total traffic.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
With the recent commit series, we have established a one-way dependency
between the link aggregation (struct tipc_node) instances and their
pertaining tipc_link instances. This has enabled quite significant code
and structure simplifications.
In this commit, we eliminate the field 'owner', which points to an
instance of struct tipc_node, from struct tipc_link, and replace it with
a pointer to struct net, which is the only external reference now needed
by a link instance.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since all packet transmitters (link, bcast, discovery) are now sending
consumable buffer clones to the bearer layer, we can remove the
redundant buffer cloning that is perfomed in the lower level functions
tipc_l2_send_msg() and tipc_udp_send_msg().
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The neighbor discovery function currently uses the function
tipc_bearer_send() for transmitting packets, assuming that the
sent buffers are not consumed by the called function.
We want to change this, in order to avoid unnecessary buffer cloning
elswhere in the code.
This commit introduces a new function tipc_bearer_skb() which consumes
the sent buffers, and let the discoverer functions use this new call
instead. The discoverer does now itself perform the cloning when
that is necessary.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Until now, we have only been supporting a fix MTU size of 1500 bytes
for all broadcast media, irrespective of their actual capability.
We now make the broadcast MTU adaptable to the carrying media, i.e.,
we use the smallest MTU supported by any of the interfaces attached
to TIPC.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Until now, we have been keeping track of the exact set of broadcast
destinations though the help structure tipc_node_map. This leads us to
have to maintain a whole infrastructure for supporting this, including
a pseudo-bearer and a number of functions to manipulate both the bearers
and the node map correctly. Apart from the complexity, this approach is
also limiting, as struct tipc_node_map only can support cluster local
broadcast if we want to avoid it becoming excessively large. We want to
eliminate this limitation, in order to enable introduction of scoped
multicast in the future.
A closer analysis reveals that it is unnecessary maintaining this "full
set" overview; it is sufficient to keep a counter per bearer, indicating
how many nodes can be reached via this bearer at the moment. The protocol
is now robust enough to handle transitional discrepancies between the
nominal number of reachable destinations, as expected by the broadcast
protocol itself, and the number which is actually reachable at the
moment. The initial broadcast synchronization, in conjunction with the
retransmission mechanism, ensures that all packets will eventually be
acknowledged by the correct set of destinations.
This commit introduces these changes.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The code path for receiving broadcast packets is currently distinct
from the unicast path. This leads to unnecessary code and data
duplication, something that can be avoided with some effort.
We now introduce separate per-peer tipc_link instances for handling
broadcast packet reception. Each receive link keeps a pointer to the
common, single, broadcast link instance, and can hence handle release
and retransmission of send buffers as if they belonged to the own
instance.
Furthermore, we let each unicast link instance keep a reference to both
the pertaining broadcast receive link, and to the common send link.
This makes it possible for the unicast links to easily access data for
broadcast link synchronization, as well as for carrying acknowledges for
received broadcast packets.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Until now, we have tried to support both the newer, dedicated broadcast
synchronization mechanism along with the older, less safe, RESET_MSG/
ACTIVATE_MSG based one. The latter method has turned out to be a hazard
in a highly dynamic cluster, so we find it safer to disable it completely
when we find that the former mechanism is supported by the peer node.
For this purpose, we now introduce a new capabability bit,
TIPC_BCAST_SYNCH, to inform any peer nodes that dedicated broadcast
syncronization is supported by the present node. The new bit is conveyed
between peers in the 'capabilities' field of neighbor discovery messages.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This commit simplifies the broadcast link transmission function, by
leveraging previous changes to the link transmission function and the
broadcast transmission link life cycle.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Realizing that unicast is just a special case of broadcast, we also see
that we can go in the other direction, i.e., that modest changes to the
current unicast link can make it generic enough to support broadcast.
The following changes are introduced here:
- A new counter ("ackers") in struct tipc_link, to indicate how many
peers need to ack a packet before it can be released.
- A corresponding counter in the skb user area, to keep track of how
many peers a are left to ack before a buffer can be released.
- A new counter ("acked"), to keep persistent track of how far a peer
has acked at the moment, i.e., where in the transmission queue to
start updating buffers when the next ack arrives. This is to avoid
double acknowledgements from a peer, with inadvertent relase of
packets as a result.
- A more generic tipc_link_retrans() function, where retransmit starts
from a given sequence number, instead of the first packet in the
transmision queue. This is to minimize the number of retransmitted
packets on the broadcast media.
When the new functionality is taken into use in the next commits,
we expect it to have minimal effect on unicast mode performance.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The broadcast link instance (struct tipc_link) used for sending is
currently aggregated into struct tipc_bclink. This means that we cannot
use the regular tipc_link_create() function for initiating the link, but
do instead have to initiate numerous fields directly from the
bcast_init() function.
We want to reduce dependencies between the broadcast functionality
and the inner workings of tipc_link. In this commit, we introduce
a new function tipc_bclink_create() to link.c, and allocate the
instance of the link separately using this function.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In reality, the link implementation is already independent from
struct tipc_bearer, in that it doesn't store any reference to it.
However, we still pass on a pointer to a bearer instance in the
function tipc_link_create(), just to have it extract some
initialization information from it.
I later commits, we need to create instances of tipc_link without
having any associated struct tipc_bearer. To facilitate this, we
want to extract the initialization data already in the creator
function in node.c, before calling tipc_link_create(), and pass
this info on as individual parameters in the call.
This commit introduces this change.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The broadcast transmission link is currently instantiated when the
network subsystem is started, i.e., on order from user space via netlink.
This forces the broadcast transmission code to do unnecessary tests for
the existence of the transmission link, as well in single mode node as
in network mode.
In this commit, we do instead create the link during initialization of
the name space, and remove it when it is stopped. The fact that the
transmission link now has a guaranteed longer life cycle than any of its
potential clients paves the way for further code simplifcations
and optimizations.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
