Get rid of the old __dev_put macro that is just a hold over from pre 2.6
kernel.  And turn dev_hold into an inline instead of a macro.
Signed-off-by: NStephen Hemminger <shemminger@osdl.org>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

And not the silly LIMIT_NETDEBUG and silently return without inserting
the option requested.

Also drop some old debugging messages associated to option insertion.
Signed-off-by: NArnaldo Carvalho de Melo <acme@mandriva.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NArnaldo Carvalho de Melo <acme@mandriva.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Merging it with its only user: dccp_v[46]_reqsk_send_ack.
Signed-off-by: NArnaldo Carvalho de Melo <acme@mandriva.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Using this also provides opportunities for introducing
inet_csk_alloc_skb that would call alloc_skb, account it to the sock
and skb_reserve(max_header), but I'll leave this for later, for now
using sk_prot->max_header consistently is enough.
Signed-off-by: NArnaldo Carvalho de Melo <acme@mandriva.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

I.e. they should be just ignored, but we have to use 'break', not 'continue',
as we have to possibly reset the mandatory flag.
Signed-off-by: NArnaldo Carvalho de Melo <acme@mandriva.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

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

Here's an updated copy of the patch to use fget_light in net/socket.c.
Rerunning the tests show a drop of ~80Mbit/s on average, which looks
bad until you see the drop in cpu usage from ~89% to ~82%.  That will
get fixed in another patch...

Before: max 8113.70, min 8026.32, avg 8072.34
 87380  16384  16384    10.01      8045.55   87.11    87.11    1.774   1.774
 87380  16384  16384    10.01      8065.14   90.86    90.86    1.846   1.846
 87380  16384  16384    10.00      8077.76   89.85    89.85    1.822   1.822
 87380  16384  16384    10.00      8026.32   89.80    89.80    1.833   1.833
 87380  16384  16384    10.01      8108.59   89.81    89.81    1.815   1.815
 87380  16384  16384    10.01      8034.53   89.01    89.01    1.815   1.815
 87380  16384  16384    10.00      8113.70   90.45    90.45    1.827   1.827
 87380  16384  16384    10.00      8111.37   89.90    89.90    1.816   1.816
 87380  16384  16384    10.01      8077.75   87.96    87.96    1.784   1.784
 87380  16384  16384    10.00      8062.70   90.25    90.25    1.834   1.834

After: max 8035.81, min 7963.69, avg 7998.14
 87380  16384  16384    10.01      8000.93   82.11    82.11    1.682   1.682
 87380  16384  16384    10.01      8016.17   83.67    83.67    1.710   1.710
 87380  16384  16384    10.01      7963.69   83.47    83.47    1.717   1.717
 87380  16384  16384    10.01      8014.35   81.71    81.71    1.671   1.671
 87380  16384  16384    10.00      7967.68   83.41    83.41    1.715   1.715
 87380  16384  16384    10.00      7995.22   81.00    81.00    1.660   1.660
 87380  16384  16384    10.00      8002.61   83.90    83.90    1.718   1.718
 87380  16384  16384    10.00      8035.81   81.71    81.71    1.666   1.666
 87380  16384  16384    10.01      8005.36   82.56    82.56    1.690   1.690
 87380  16384  16384    10.00      7979.61   82.50    82.50    1.694   1.694
Signed-off-by: NBenjamin LaHaise <bcrl@linux.intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The functions list_del followed by list_add_tail is equivalent to the
existing inline list_move_tail. list_move_tail avoids unnecessary
_LIST_POISON.
Signed-off-by: NStephen Hemminger <shemminger@osdl.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

struct neigh_ops currently has a destructor field, which no in-kernel
drivers outside of infiniband use.  The infiniband/ulp/ipoib in-tree
driver stashes some info in the neighbour structure (the results of
the second-stage lookup from ARP results to real link-level path), and
it uses neigh->ops->destructor to get a callback so it can clean up
this extra info when a neighbour is freed.  We've run into problems
with this: since the destructor is in an ops field that is shared
between neighbours that may belong to different net devices, there's
no way to set/clear it safely.

The following patch moves this field to neigh_parms where it can be
safely set, together with its twin neigh_setup.  Two additional
patches in the patch series update ipoib to use this new interface.
Signed-off-by: NMichael S. Tsirkin <mst@mellanox.co.il>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Due to the thread's lock changes, we're at a new version now.
Signed-off-by: NLuiz Capitulino <lcapitulino@mandriva.com.br>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

As suggested by Arnaldo, this patch replaces the
thread_lock()/thread_unlock() by directly calls to
mutex_lock()/mutex_unlock().

This change makes the code a bit more readable, and the direct calls
are used everywhere in the kernel.
Signed-off-by: NLuiz Capitulino <lcapitulino@mandriva.com.br>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

pktgen's thread semaphores are strict mutexes, convert them to the
mutex implementation.
Signed-off-by: NLuiz Capitulino <lcapitulino@mandriva.com.br>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch turns the RTNL from a semaphore to a new 2.6.16 mutex and
gets rid of some of the leftover legacy.
Signed-off-by: NStephen Hemminger <shemminger@osdl.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

First, it warns when PAGE_SIZE >= 64K because the ctx_len
field is 16-bits.

Secondly, if there are any real length limitations it can
be verified by the security layer security_xfrm_state_alloc()
call.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Instead of estimating the time since the last congestion event, count
it directly.
Signed-off-by: NBaruch Even <baruch@ev-en.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Account for delayed-ACKs in H-TCP.

Delayed-ACKs cause H-TCP to be less aggressive than its design calls
for. It is especially true when the receiver is a Linux machine where
the average delayed ack is over 3 packets with values of 7 not unheard
of.
Signed-off-By: NBaruch Even <baruch@ev-en.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Use functions to calculate jiffies from milliseconds and not the old,
crude method of dividing HZ by a value. Ensures more accurate values
even in the face of strange HZ values.
Signed-off-By: NBaruch Even <baruch@ev-en.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

With all the previous changes, we're at a new version now.
Signed-off-by: NLuiz Capitulino <lcapitulino@mandriva.com.br>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch ports the per-thread interface list list to the in-kernel
linked list implementation. In the general, the resulting code is a
bit simpler.
Signed-off-by: NLuiz Capitulino <lcapitulino@mandriva.com.br>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Even if pktgen's thread initialization fails for all CPUs, the module
will be successfully loaded.

This patch changes that behaivor, by returning an error on module load time,
and also freeing all the resources allocated. It also prints a warning if a
thread initialization has failed.
Signed-off-by: NLuiz Capitulino <lcapitulino@mandriva.com.br>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Free all the alocated resources if kernel_thread() call fails.
Signed-off-by: NLuiz Capitulino <lcapitulino@mandriva.com.br>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The final result is a simpler and smaller code.

Note that I'm adding a new member in the struct pktgen_thread called
'removed'. The reason is that I didn't find a better wait condition to
be used in the place of the replaced one.
Signed-off-by: NLuiz Capitulino <lcapitulino@mandriva.com.br>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Lindet run, with some fixes made by hand.
Signed-off-by: NLuiz Capitulino <lcapitulino@mandriva.com.br>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

According to dccp draft (draft-ietf-dccp-spec-13.txt) section 5.8.2
(Mandatory Option) the following patch correct the handling of the
following cases:

1) "... and any Mandatory options received on DCCP-Data packets MUST be
  ignored."

2) "The connection is in error and should be reset with Reset Code 5, ...
  if option O is absent (Mandatory was the last byte of the option list), or
  if option O equals Mandatory."
Signed-off-by: NArnaldo Carvalho de Melo <acme@mandriva.com>
Signed-off-by: NHagen Paul Pfeifer <hagen@jauu.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

No changes in the logic where made.
Signed-off-by: NArnaldo Carvalho de Melo <acme@mandriva.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

No changes in the logic were made, just removing trailing whitespaces,
etc.
Signed-off-by: NArnaldo Carvalho de Melo <acme@mandriva.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Consolidating open coded sequences in tcp and dccp, v4 and v6.
Signed-off-by: NArnaldo Carvalho de Melo <acme@mandriva.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

I guess I forgot to add it, nah, now it just works:

18:04:33.274066 IP6 ::1.1476 > ::1.5001: request (service=0)
18:04:33.334482 IP6 ::1.5001 > ::1.1476: reset (code=bad_service_code)

Ditched IP_DCCP_UNLOAD_HACK, as now we would have to do it for both
IPv6 and IPv4, so I'll come up with another way for freeing the
control sockets in upcoming changesets.
Signed-off-by: NArnaldo Carvalho de Melo <acme@mandriva.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NArnaldo Carvalho de Melo <acme@mandriva.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

There's no reason for struct dccp_v4_prot being global.
Signed-off-by: NAdrian Bunk <bunk@stusta.de>
Signed-off-by: NArnaldo Carvalho de Melo <acme@mandriva.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

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

fib_triestats has been buggy and caused oopses some platforms as
openwrt.  The patch below should cure those problems.
Signed-off-by: NRobert Olsson <robert.olsson@its.uu.se>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In some kernel configs /proc functions seems to be accessed before the
trie is initialized. The patch below checks for this.
Signed-off-by: NRobert Olsson <robert.olsson@its.uu.se>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This moves some TCP-specific MTU probing state out of
inet_connection_sock back to tcp_sock.
Signed-off-by: NJohn Heffner <jheffner@psc.edu>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The patch below replaces a divide by 2 with a shift -- sk_sndbuf is an
integer, so gcc emits an idiv, which takes 10x longer than a shift by 1.
This improves af_unix bandwidth by ~6-10K/s.  Also, tidy up the comment
to fit in 80 columns while we're at it.
Signed-off-by: NBenjamin LaHaise <benjamin.c.lahaise@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

o Uninline kfree_skb, which saves some 15k of object code on my notebook.

o Allow kfree_skb to be called with a NULL argument.

  Subsequent patches can remove conditional from drivers and further
  reduce source and object size.
Signed-off-by: NJrn Engel <joern@wohnheim.fh-wedel.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Thanks to Leslie Harlley Watter <leslie@watter.org> for reporting the
problem an testing this patch.
Signed-off-by: NArnaldo Carvalho de Melo <acme@mandriva.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NArnaldo Carvalho de Melo <acme@mandriva.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Cut'n'paste error from ddp_proto.
Signed-off-by: NArnaldo Carvalho de Melo <acme@mandriva.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>