Tom Herbert says:

====================
This series implements so_reuseport (SO_REUSEPORT socket option) for
TCP and UDP.  For TCP, so_reuseport allows multiple listener sockets
to be bound to the same port.  In the case of UDP, so_reuseport allows
multiple sockets to bind to the same port.  To prevent port hijacking
all sockets bound to the same port using so_reuseport must have the
same uid.  Received packets are distributed to multiple sockets bound
to the same port using a 4-tuple hash.

The motivating case for so_resuseport in TCP would be something like
a web server binding to port 80 running with multiple threads, where
each thread might have it's own listener socket.  This could be done
as an alternative to other models: 1) have one listener thread which
dispatches completed connections to workers. 2) accept on a single
listener socket from multiple threads.  In case #1 the listener thread
can easily become the bottleneck with high connection turn-over rate.
In case #2, the proportion of connections accepted per thread tends
to be uneven under high connection load (assuming simple event loop:
while (1) { accept(); process() }, wakeup does not promote fairness
among the sockets.  We have seen the  disproportion to be as high
as 3:1 ratio between thread accepting most connections and the one
accepting the fewest.  With so_reusport the distribution is
uniform.

The TCP implementation has a problem in that the request sockets for a
listener are attached to a listener socket.  If a SYN is received, a
listener socket is chosen and request structure is created (SYN-RECV
state).  If the subsequent ack in 3WHS does not match the same port
by so_reusport, the connection state is not found (reset) and the
request structure is orphaned.  This scenario would occur when the
number of listener sockets bound to a port changes (new ones are
added, or old ones closed).  We are looking for a solution to this,
maybe allow multiple sockets to share the same request table...

The motivating case for so_reuseport in UDP would be something like a
DNS server.  An alternative would be to recv on the same socket from
multiple threads.  As in the case of TCP, the load across these threads
tends to be disproportionate and we also see a lot of contection on
the socket lock.  Note that SO_REUSEADDR already allows multiple UDP
sockets to bind to the same port, however there is no provision to
prevent hijacking and nothing to distribute packets across all the
sockets sharing the same bound port.  This patch does not change the
semantics of SO_REUSEADDR, but provides usable functionality of it
for unicast.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Motivation for soreuseport would be something like a DNS server.  An
alternative would be to recv on the same socket from multiple threads.
As in the case of TCP, the load across these threads tends to be
disproportionate and we also see a lot of contection on the socket lock.
Note that SO_REUSEADDR already allows multiple UDP sockets to bind to
the same port, however there is no provision to prevent hijacking and
nothing to distribute packets across all the sockets sharing the same
bound port.  This patch does not change the semantics of SO_REUSEADDR,
but provides usable functionality of it for unicast.
Signed-off-by: NTom Herbert <therbert@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Motivation for soreuseport would be something like a web server
binding to port 80 running with multiple threads, where each thread
might have it's own listener socket.  This could be done as an
alternative to other models: 1) have one listener thread which
dispatches completed connections to workers. 2) accept on a single
listener socket from multiple threads.  In case #1 the listener thread
can easily become the bottleneck with high connection turn-over rate.
In case #2, the proportion of connections accepted per thread tends
to be uneven under high connection load (assuming simple event loop:
while (1) { accept(); process() }, wakeup does not promote fairness
among the sockets.  We have seen the  disproportion to be as high
as 3:1 ratio between thread accepting most connections and the one
accepting the fewest.  With so_reusport the distribution is
uniform.
Signed-off-by: NTom Herbert <therbert@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Allow multiple UDP sockets to bind to the same port.

Motivation soreuseport would be something like a DNS server.  An
alternative would be to recv on the same socket from multiple threads.
As in the case of TCP, the load across these threads tends to be
disproportionate and we also see a lot of contection on the socketlock.
Note that SO_REUSEADDR already allows multiple UDP sockets to bind to
the same port, however there is no provision to prevent hijacking and
nothing to distribute packets across all the sockets sharing the same
bound port.  This patch does not change the semantics of SO_REUSEADDR,
but provides usable functionality of it for unicast.
Signed-off-by: NTom Herbert <therbert@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Allow multiple listener sockets to bind to the same port.

Motivation for soresuseport would be something like a web server
binding to port 80 running with multiple threads, where each thread
might have it's own listener socket.  This could be done as an
alternative to other models: 1) have one listener thread which
dispatches completed connections to workers. 2) accept on a single
listener socket from multiple threads.  In case #1 the listener thread
can easily become the bottleneck with high connection turn-over rate.
In case #2, the proportion of connections accepted per thread tends
to be uneven under high connection load (assuming simple event loop:
while (1) { accept(); process() }, wakeup does not promote fairness
among the sockets.  We have seen the  disproportion to be as high
as 3:1 ratio between thread accepting most connections and the one
accepting the fewest.  With so_reusport the distribution is
uniform.
Signed-off-by: NTom Herbert <therbert@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Definitions and macros for implementing soreusport.
Signed-off-by: NTom Herbert <therbert@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Sometimes it is useful to be able to change the MAC address of the
interface for netback devices. For example, when using ebtables it may
be useful to be able to distinguish traffic from different interfaces
without depending on the interface name.
Reported-by: NNikita Borzykh <sample.n@gmail.com>
Reported-by: NPaul Harvey <stockingpaul@hotmail.com>
Cc: netdev@vger.kernel.org
Cc: xen-devel@lists.xen.org
Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Acked-by: NIan Campbell <ian.campbell@citrix.com>
Signed-off-by: NMatt Wilson <msw@amazon.com>
Reviewed-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Fengguang reported:

   net/core/netpoll.c: In function 'netpoll_setup':
   net/core/netpoll.c:1049:6: warning: 'err' may be used uninitialized in this function [-Wmaybe-uninitialized]

in !CONFIG_IPV6 case, we may error out without initializing
'err'.
Reported-by: NFengguang Wu <fengguang.wu@intel.com>
Cc: David S. Miller <davem@davemloft.net>
Signed-off-by: NCong Wang <amwang@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

It is declared in:
include/net/ip6_route.h:187:int ip6_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *));

and net/ip6_route.h is already included.

Cc: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Cc: David S. Miller <davem@davemloft.net>
Signed-off-by: NCong Wang <amwang@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Paul Gortmaker says:

====================
The Ethernet-HowTo was maintained for roughly 10 years, from 1993 to 2003.
Fortunately sane hardware probing and auto detection (via PCI and ISA/PnP)
largely made the document a relic of the past, hence it being abandoned
a decade ago.

However, there is one last useful thing that we can extract from the
effort made in maintaining that document.  We can use it to guide us
with respect to what rare, experimental and/or super ancient 10Mbit
ISA drivers don't make sense to maintain in-tree anymore.

Nobody will argue that ISA is obsolete.  Availability went away at about
the time Pentium3 motherboards moved from 500MHz Slot1/SECC processors
to the green 500MHz Socket 370 Pentium3 chips, at the turn of the century.

In theory, it is possible that someone could still be running one of these
12+ year old P3 machines and want 3.9+ bleeding edge kernels (but unlikely).
In light of the above (remote) possibility, we can defer the removal of some
ISA network drivers that were highly popular and well tested.  Typically
that means the stuff more from the mid to late '90s, some with ISA PnP
support, like the 3c509, the wd/SMC 8390 based stuff, PCnet/lance etc.

But a lot of other drivers, typically from the early 1990s were for rare
hardware, and experimental (to the point of requiring a cron job that would
do a test ping, and then ifconfig down/up and/or a rmmod/insmod!).  And
some of these drivers (znet, and lp486e to name two) are physically tied
to platforms with on motherboard ethernet -- of 486 machines that date
from the early 1990s and can only have single digit amounts of memory.

What I'd like to achieve here with this series, is to get rid of those old
drivers that are no longer being used.  In an earlier discussion where
I'd proposed deleting a single driver, Alan suggested we instead dump
all the historical stuff in one go, to make it "...immediately obvious
where the break point is..."[1] and that it was "perfectly reasonable it
(and a pile of other ISA cards) ought to be shown the door"[2].  So that
is the goal here - make a clear line in the sand where the really ancient
stuff finally gets kicked to the curb.

Two old parallel port drivers are considered for removal here as well,
since in early 386/486 ISA machines, the parallel port was typically found
with the UARTS on the multi-I/O ISA controller card.  These drivers also date
from the early 1990's; parallel ports are no longer found on modern boards,
and their performance was not even capable of 10% of 10Mbit bandwidth.

Allow me a preemptive justification against the inevitable comments from
well meaning bystanders who suggest "why not just leave all this alone?".
Dead drivers cost us all if they are left in tree.  If you think that
is false, then please first consider:

-every time you type "git status", you are checking to see if modifications
 have been made by you to all that dead code.

-every time you type "git grep <regex>" you are searching through files
 which contain that dead code that simply does not interest you.

-every time you build a "allyesconfig" and an "allmodconfig" (don't tell
 me you skip this step before submitting your changes to a maintainer),
 you waste CPU cycles building this dead code.

-every time there is a tree wide API change, or cleanup, or file relocation,
 we pay the cost of updating dead code, or moving dead code.

-daily regression tests (take linux-next as the most transparent
 example) spend time building (and possibly running) this dead code.

-hard working people who regularly run auditing tools looking for lurking
 bugs (sparse/coverity/smatch/coccinelle) are wasting time checking for,
 and fixing bugs in this dead code.

This last one is key.  Please take a look at the git history for the
files that are proposed for removal here.  Look at the git history for
any one of them ("git whatchanged --follow drivers/net/.../driver.c")
Mentally sort the changes into two bins -- (1) the robotic tree-wide
changes, and (2) the "look I found a real run-time bug while using this"
category.  You will see that category #2 is essentially empty.

Further to that, realize that drivers don't simply disappear.  We are
not operating in the binary-only distribution space like other OS.  All
these drivers remain in the git history forever.  If a person is an
enthusiast for extreme legacy hardware, they are probably already
customizing their kernel source and building it themselves to support
such systems.  Also keep in mind that they could still build the 3.8
kernel exactly as-is, and run it (or a 3.8.x stable variant of it) for
several more years if they were really determined to cling to these old
experimental ISA drivers for some reason.

In summary, I hope that folks can be pragmatic about this, and not
get swept up in nostalgia.  Ask yourself whether it is realistic to
expect a person would have a genuine use case where they would
need to build a 3.9+ modern kernel and install it on some legacy hardware
that has no option but to absolutely _require_ one of the drivers
that are deleted here.

The following series was created with --irreversible-delete for
ease of review (it skips showing the content of files that are
deleted); however the complete patches can be pulled as per below.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NYOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NYOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NYOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NYOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Since we have removed NCE (Neighbour Cache Entry) reference from
routing entries, the only refcnt holders of an NCE are its timer
(if running) and its owner table, in usual cases.  As a result,
neigh_periodic_work() purges NCEs over and over again even for
gateways.

It does not make sense to purge entries, if number of them is
very small, so keep them.  The minimum number of entries to keep
is specified by gc_thresh1.
Signed-off-by: NYOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

mfc_mcastgrp and mfc_origin are __be32, thus we need to convert INADDR_ANY.
Because INADDR_ANY is 0, this patch just fix sparse warnings.
Reported-by: NFengguang Wu <fengguang.wu@intel.com>
Signed-off-by: NNicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The last update to the Ethernet HowTo (over 10 years ago) listed this:

 ------------------------
   SEEQ 8005

   Status: Obsolete, Driver Name: seeq8005

   There is little information about the card included in the driver,
   and hence little information to be put here. If you have a question,
   you are probably best trying to e-mail the driver author as listed
   in the source.

   It was marked obsolete as of the 2.4 series kernels.
 ------------------------

If it was obsolete over a decade ago, the situation can not have
improved with the passage of time, so let us act on that.  Even with
today's improved search engines, I was unable to locate any real
meaningful information on the ISA implementation of this rare chip.

There are ARM and SGI variants of the driver in tree, but they do
not depend on the original x86 driver source or header file.  We
leave those non-x86 drivers to be deleted by the arch maintainers
when they decide to expire those legacy platforms as a whole.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

This is another one that makes sense to target for obsolescence, since
it (a)appeared pre-1995, and (b)was rather rare, and (c)did not
really have any statistically significant active linux user base.

Removing this ISA 10Mbit driver support is unlikely to be even noticed
by the user base of 3.9+ linux kernels, especially when the documentation
clearly indicates the vintage with this text:

	 "...designed to  work with all kernels > 1.1.33"
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

These are old ISA 10Mbit cards from the 1st 1/2 of the 1990s and
required manual jumper settings in order to configure them.  Here
we remove them on the premise that they are no longer used in any
modern 3.9+ kernels.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

This is an area I know all too well, after being author of several 8390
drivers, and maintainer of all 8390 drivers during a large part of their
active lifecycle.

To that end, I can say this with a reasonable degree of confidence.
The drivers deleted here represent the earliest (as in early 1990)
hardware and/or rare hardware.  The remaining hardware not deleted
here is the more modern/sane of the lot, with ISA-PnP and jumperless
"soft configuration" like the wd and smc cards had.

The original ne2000 driver (ne.c) gets a pass at this time since
AT/LANTIC based cards that could be both ne2000 or wd-like (with
shared memory) and with jumperless configuration were made in the
mid to late 1990's, and performed reasonably well for their era.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

This is another driver for relatively rare 10Mbit hardware that
originated in the early 1990's.  So we select it for removal at
this point in time as well.

Cc: Mika Kuoppala <miku@iki.fi>
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

These Fujitsu MB86965 based ISA 10Mbit cards were another of the
relatively rare cards dating from the early 1990s that for one reason
or another didn't seem to get a lot of use in linux.  So we retire it
now with a reasonable degree of confidence that it won't impact anyone.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

These cards were only available in 8bit format, and in addition
they only had AUI and BNC(10-Base2) interfaces (i.e. no RJ-45).

In fact, they are so rare, that an internet search on these old
cards almost comes up empty, unless the "Micom interlan" name
is used.

This puts them in the equivalent domain as the 3c501, so there
should be no strong opposition to the driver removal, as nobody
is seriously using 3.9+ with 8 bit ISA hardware.

In doing so, the whole "ethernet/racal" category becomes empty,
so we clean up the Makefile/Kconfig and subdir appropriately.

Cc: Andreas Mohr <andi@lisas.de>
Cc: Jan-Pascal van Best <janpascal@vanbest.org>
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

Like the other drivers that were in the ISA i825xx family, the ni52
was rather rare, not widely used, and hence perhaps not as reliable
as the more mainstream ISA drivers that were heavily used.  Given
that, it is chosen for retirement at this time as well.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

This driver supported early to mid 1990's Zenith laptops, of the
2" thick variety.  The driver was already dead 10+ years ago, but
we see this in the source:

 ----------------
 /* 10/2002

 [...]

   Tested on a vintage Zenith Z-Note 433Lnp+. Probably broken on
   anything else. Testers (and detailed bug reports) are welcome :-).
 ----------------

To clarify, a 433 translates into a 486 at 33MHz, and a system with
a default of 4MB RAM.  I can't fault the noble effort to keep things
working a decade ago, but at this point in time, there is no valid
justification to continue carrying this driver along.

Note that there is no associated Space.c cleanup here since this
driver was using module_init to hook itself in.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

These old drivers should not be confused with the very common PCI
cards that are supported by e100.c -- these older 10Mbit ISA only
drivers were not as commonly used as some of the other ISA drivers,
simply due to hardware availability and pricing.

Given the rarity of the hardware, and the subsequent less extensive
use of the drivers, it makes sense to obsolete them at this point
in time, along with other rare/experimental ISA drivers.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

For those of us who were around in the early to mid 1990's, we
will remember that the i825xx ethernet support was not something
that was considered sufficiently vetted for 24/7 use.

Folks might be inclined to use *functional* ISA hardware on some
near expired P3 ISA machines for dedicated workhorse applications,
but the odds of using (and relying on) one of these old/experimental
drivers is essentially nil.  So lets remove them.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

The parallel port is largely replaced by USB, and even in the
day where these drivers were current, the documented speed was
less than 100kB/s.  Let us not pretend that anyone cares about
these drivers anymore, or worse - pretend that anyone is using
them on a modern kernel.

As a side bonus, this is the end of legacy parallel port ethernet,
so we get to drop the whole chunk relating to that in the legacy
Space.c file containing the non-PCI unified probe dispatch.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

It was amusing that linux was able to make use of this 1980's
technology on machines long past its intended lifespan, but
it probably should go now.

To set some context, the 3c501 was designed in the 1980's to be
used on 8088 PC-XT 8bit ISA machines.  It was built using a large
number of discrete TTL components and truly looks like a relic
of the ancient past before large scale integration was common.

But from a functional point of view, the real issue, as stated
in the (also obsolete) Ethernet-HowTo, is that "...the 3c501 can
only do one thing at a time -- while you are removing one packet
from the single-packet buffer it cannot receive another packet,
nor can it receive a packet while loading a transmit packet."

You know things are not good when the Kconfig help text suggests
you make a cron job doing a ping every minute.

Hardware that old and crippled is simply not going to be used by
anyone in a time where 10 year old 100Mbit PCI cards (that are
still functional) are largely give-away items.

Cc: Alan Cox <alan@linux.intel.com>
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

This driver was specific to a "professional workstation" line
of products from around 1993 that used the i82596 ethernet chip
as an on-board ethernet solution.

With a 486 processor, and the premium top of the line model maxing
out at a clock speed of 50MHz, we can safely retire this support.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

The Apricot was a 486 PC with 4MB RAM, and an on-board ethernet
via an intel i82596 hard-wired to i/o 0x300.

Those who were using linux in the 1990's will recall that the
i82596 driver was not one of the more stable or widely used
drivers of its day.  Combine that with the extremely limited
resources of the platform, and it is truly time to expire the
support for this thing.

There are some old m68k targets who were also using this chip,
so rather than poll the m68k user base, we simply cut out the
x86/Apricot support here in this commit.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

Cosmetic changes to drivers/isdn/gigaset/ev-layer.c and
drivers/isdn/gigaset/gigaset.h to improve readability.
Signed-off-by: NTilman Schmidt <tilman@imap.cc>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Rearrange the gigaset_freecs() function to make it more readable,
and adapt gigaset_initcs() accordingly.
Signed-off-by: NTilman Schmidt <tilman@imap.cc>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Avoid forward declarations and remove a needless initialization.
Signed-off-by: NTilman Schmidt <tilman@imap.cc>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Some firmware releases of Gigaset M105 do not accept AT+VLS=0 command
in DLE mode, so always leave DLE mode before sending the command.
Signed-off-by: NTilman Schmidt <tilman@imap.cc>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Fix up some of the readibility deterioration caused by last year's
ISDN whitespace coding style cleanup.
Note that the checkpatch complaints all apply to the state of the
source before this patch as well, and in many cases even more so.
Signed-off-by: NTilman Schmidt <tilman@imap.cc>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Delete successive tests to the same location.  rc was previously tested and
not subsequently updated.  efx_phc_adjtime can return an error code, so the
call is updated so that is tested instead.

A simplified version of the semantic match that finds this problem is as
follows: (http://coccinelle.lip6.fr/)

// <smpl>
@s exists@
local idexpression y;
expression x,e;
@@

*if ( \(x == NULL\|IS_ERR(x)\|y != 0\) )
 { ... when forall
   return ...; }
... when != \(y = e\|y += e\|y -= e\|y |= e\|y &= e\|y++\|y--\|&y\)
    when != \(XT_GETPAGE(...,y)\|WMI_CMD_BUF(...)\)
*if ( \(x == NULL\|IS_ERR(x)\|y != 0\) )
 { ... when forall
   return ...; }
// </smpl>
Signed-off-by: NJulia Lawall <Julia.Lawall@lip6.fr>
Acked-by: NBen Hutchings <bhutchings@solarflare.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Commit 2152caea ("ipv6: Do not depend on rt->n in rt6_probe().")
introduce a bug to try to update "updated" time in neighbour
structure.
Update the "updated" time only if neighbour is available.

Bug was found by Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NYOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Florian Fainelli says:

====================
These two patches are non-critical bugfixes based on net-next which I
stumbled upon while working on Device Tree bindings for DSA (will comme
as a separate patch later).

====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch changes dsa_switch_setup() to ensure that at least one valid
valid port name is specified and will bail out with an error in case we
walked the maximum number of port with a valid port name found.
Signed-off-by: NFlorian Fainelli <florian@openwrt.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>