commit 8a27f7c90f
changed the output style of %pi4 to use fixed
width leading zero IP addresses "001.002.003.004".
It's useful when printing multiple lines of
addresses, but was a change in output style for
some existing uses.
Using %pI4 restores the previous output style.
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Use the same format string as net/ipv4/netfilter/nf_nat_ftp.c
to encode an ipv4 address and port.
Both uses should be a single common function.
Signed-off-by: Joe Perches <joe@perches.com>
Acked-by: Simon Horman <horms@verge.net.au>
Signed-off-by: Patrick McHardy <kaber@trash.net>
These functions merely exist to format a buffer and call
nf_nat_mangle_tcp_packet.
Format the buffer and perform the call in nf_nat_ftp instead.
Use %pI4 for the IP address.
Saves ~600 bytes of text
old:
$ size net/ipv4/netfilter/nf_nat_ftp.o
text data bss dec hex filename
2187 160 408 2755 ac3 net/ipv4/netfilter/nf_nat_ftp.o
new:
$ size net/ipv4/netfilter/nf_nat_ftp.o
text data bss dec hex filename
1532 112 288 1932 78c net/ipv4/netfilter/nf_nat_ftp.o
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Patrick McHardy <kaber@trash.net>
I was very frustrated about the fact that I have to recompile the kernel
to change the hash size. So, I created this patch.
If IPVS is built-in you can append ip_vs.conn_tab_bits=?? to kernel
command line, or, if you built IPVS as modules, you can add
options ip_vs conn_tab_bits=??.
To keep everything backward compatible, you still can select the size at
compile time, and that will be used as default.
It has been about a year since this patch was originally posted
and subsequently dropped on the basis of insufficient test data.
Mark Bergsma has provided the following test results which seem
to strongly support the need for larger hash table sizes:
We do however run into the same problem with the default setting (212 =
4096 entries), as most of our LVS balancers handle around a million
connections/SLAB entries at any point in time (around 100-150 kpps
load). With only 4096 hash table entries this implies that each entry
consists of a linked list of 256 connections *on average*.
To provide some statistics, I did an oprofile run on an 2.6.31 kernel,
with both the default 4096 table size, and the same kernel recompiled
with IP_VS_CONN_TAB_BITS set to 18 (218 = 262144 entries). I built a
quick test setup with a part of Wikimedia/Wikipedia's live traffic
mirrored by the switch to the test host.
With the default setting, at ~ 120 kpps packet load we saw a typical %si
CPU usage of around 30-35%, and oprofile reported a hot spot in
ip_vs_conn_in_get:
samples % image name app name
symbol name
1719761 42.3741 ip_vs.ko ip_vs.ko ip_vs_conn_in_get
302577 7.4554 bnx2 bnx2 /bnx2
181984 4.4840 vmlinux vmlinux __ticket_spin_lock
128636 3.1695 vmlinux vmlinux ip_route_input
74345 1.8318 ip_vs.ko ip_vs.ko ip_vs_conn_out_get
68482 1.6874 vmlinux vmlinux mwait_idle
After loading the recompiled kernel with 218 entries, %si CPU usage
dropped in half to around 12-18%, and oprofile looks much healthier,
with only 7% spent in ip_vs_conn_in_get:
samples % image name app name
symbol name
265641 14.4616 bnx2 bnx2 /bnx2
143251 7.7986 vmlinux vmlinux __ticket_spin_lock
140661 7.6576 ip_vs.ko ip_vs.ko ip_vs_conn_in_get
94364 5.1372 vmlinux vmlinux mwait_idle
86267 4.6964 vmlinux vmlinux ip_route_input
[ horms@verge.net.au: trivial up-port and minor style fixes ]
Signed-off-by: Catalin(ux) M. BOIE <catab@embedromix.ro>
Cc: Mark Bergsma <mark@wikimedia.org>
Signed-off-by: Simon Horman <horms@verge.net.au>
Signed-off-by: Patrick McHardy <kaber@trash.net>
We can initialize the random hash bytes on checkentry. This is
preferable since it is outside the hot path.
Reference: http://bugzilla.netfilter.org/show_bug.cgi?id=621
Signed-off-by: Jan Engelhardt <jengelh@medozas.de>
Signed-off-by: Patrick McHardy <kaber@trash.net>
"It is deliberately not done in the init function, since we might not
have sufficient random while booting."
Signed-off-by: Jan Engelhardt <jengelh@medozas.de>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Moving rnd_inited into the hole after the uint8 lets go of the uint32
rnd_inited was using, plus the padding that would follow the int group.
Signed-off-by: Jan Engelhardt <jengelh@medozas.de>
Signed-off-by: Patrick McHardy <kaber@trash.net>
obj has type struct snmp_object **, not struct snmp_object *. But indeed
it is not even clear why kmalloc is needed. The memory is freed by the end
of the function, so the local variable of pointer type should be sufficient.
The semantic patch that makes this change is as follows:
(http://coccinelle.lip6.fr/)
// <smpl>
@disable sizeof_type_expr@
type T;
T **x;
@@
x =
<+...sizeof(
- T
+ *x
)...+>
// </smpl>
Signed-off-by: Julia Lawall <julia@diku.dk>
Signed-off-by: Patrick McHardy <kaber@trash.net>
We can rely on kconfig to limit these numbers,
no need to limit them at compile time/run time.
Users who modify these numbers manually should
be responsible for themself. :)
Signed-off-by: WANG Cong <amwang@redhat.com>
Cc: Per Liden <per.liden@ericsson.com>
Cc: Jon Maloy <jon.maloy@ericsson.com>
Cc: Allan Stephens <allan.stephens@windriver.com>
Cc: David S. Miller <davem@davemloft.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
this patch removes the hlist that contains the CAN receiver filter lists.
It uses the 'midlayer private' pointer ml_priv and links the filters directly
to the CAN netdevice, which allows to omit the walk through the complete CAN
devices hlist for each received CAN frame.
This patch is tested and does not remove any locking.
Signed-off-by: Oliver Hartkopp <oliver@hartkopp.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
This allows a bond device to specify an arp_ip_target as a host that is
not on the same vlan as the base bond device and still use arp
validation. A configuration like this, now works:
BONDING_OPTS="mode=active-backup arp_interval=1000 arp_ip_target=10.0.100.1 arp_validate=3"
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
2: eth1: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master bond0 qlen 1000
link/ether 00:13:21:be:33:e9 brd ff:ff:ff:ff:ff:ff
3: eth0: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master bond0 qlen 1000
link/ether 00:13:21:be:33:e9 brd ff:ff:ff:ff:ff:ff
8: bond0: <BROADCAST,MULTICAST,MASTER,UP,LOWER_UP> mtu 1500 qdisc noqueue
link/ether 00:13:21:be:33:e9 brd ff:ff:ff:ff:ff:ff
inet6 fe80::213:21ff:febe:33e9/64 scope link
valid_lft forever preferred_lft forever
9: bond0.100@bond0: <BROADCAST,MULTICAST,MASTER,UP,LOWER_UP> mtu 1500 qdisc noqueue
link/ether 00:13:21:be:33:e9 brd ff:ff:ff:ff:ff:ff
inet 10.0.100.2/24 brd 10.0.100.255 scope global bond0.100
inet6 fe80::213:21ff:febe:33e9/64 scope link
valid_lft forever preferred_lft forever
Ethernet Channel Bonding Driver: v3.6.0 (September 26, 2009)
Bonding Mode: fault-tolerance (active-backup)
Primary Slave: None
Currently Active Slave: eth1
MII Status: up
MII Polling Interval (ms): 0
Up Delay (ms): 0
Down Delay (ms): 0
ARP Polling Interval (ms): 1000
ARP IP target/s (n.n.n.n form): 10.0.100.1
Slave Interface: eth1
MII Status: up
Link Failure Count: 1
Permanent HW addr: 00:40:05:30:ff:30
Slave Interface: eth0
MII Status: up
Link Failure Count: 0
Permanent HW addr: 00:13:21:be:33:e9
Signed-off-by: Andy Gospodarek <andy@greyhouse.net>
Signed-off-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
To make it easier to notice cases of calling sleeping ops in atomic context,
annotate driver-ops.h with appropiate might_sleep() calls. At the same time,
also document in mac80211.h the op functions with missing contexts.
mac80211 doesn't seem to use get_tx_stats anywhere currently. Just to be on
the safe side, I documented it to be atomic, but hopefully the op can be
removed in the future.
Compile-tested only.
Signed-off-by: Kalle Valo <kalle.valo@iki.fi>
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
There's no need to be requeueing the work struct
since we check for the scan after removing items
due to possible timeouts.
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
All its members (vif, mac_addr, type) are now available
in the vif struct directly, so we can pass that instead
of the conf struct. I generated this patch (except the
mac80211 and header file changes) with this semantic
patch:
@@
identifier conf, fn, hw;
type tp;
@@
tp fn(struct ieee80211_hw *hw,
-struct ieee80211_if_init_conf *conf)
+struct ieee80211_vif *vif)
{
<...
(
-conf->type
+vif->type
|
-conf->mac_addr
+vif->addr
|
-conf->vif
+vif
)
...>
}
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
When, for instance, a new IBSS peer is found, userspace
wants to be notified. Add events for all new stations
that mac80211 learns about.
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
This changes mac80211 to allow being off-channel for
any type of work, not just the 'remain-on-channel'
work. This also helps fast transition to a BSS on a
different channel.
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
This implements the new remain-on-channel cfg80211
command in mac80211, extending the work interface.
Also change the work purge code to be able to clean
up events properly (pretending they timed out.)
Signed-off-by: Jouni Malinen <jouni.malinen@atheros.com>
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Add new commands for requesting the driver to remain awake
on a specified channel for the specified amount of time
(and another command to cancel such an operation). This
can be used to implement userspace-controlled off-channel
operations, like Public Action frame exchange on another
channel than the operation channel.
The off-channel operation should behave similarly to scan,
i.e. the local station (if associated) moves into power
save mode to request the AP to buffer frames for it and
then moves to the other channel to allow the off-channel
operation to be completed. The duration parameter can be
used to request enough time to receive a response from
the target station.
Signed-off-by: Jouni Malinen <jouni.malinen@atheros.com>
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
The off-channel operations for going into power save mode (station
mode) or stop beaconing (AP/IBSS) are not limited to scanning. Move
these into a separate file and allow them to be used for other
purposes, too.
Signed-off-by: Jouni Malinen <jouni.malinen@atheros.com>
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
cfg80211 offers private data for each BSS struct,
which mac80211 uses. However, mac80211 uses internal
and external (cfg80211) BSS pointers interchangeably
and has a hack to put the cfg80211 bss struct into
the private struct.
Remove this hack, properly converting between the
pointers wherever necessary.
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Currently, we insert all user-specified IEs before the HT
IE for association, and after the HT IE for probe requests.
For association, that's correct only if the user-specified
IEs are RSN only, incorrect in all other cases including
WPA. Change this to split apart the user-specified IEs in
two places for association: before the HT IE (e.g. RSN),
after the HT IE (generally empty right now I think?) and
after WMM (all other vendor-specific IEs). For probes,
split the IEs in different places to be correct according
to the spec.
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Refactor the code to reserve an skb of the right size
(instead of hoping 200 bytes are enough forever), and
also put HT IE generation into an own function.
Additionally, put the HT IE before the vendor-specific
WMM IE. This still leaves things not quite ordered
correctly, due to user-specified IEs, add a note about
that for now.
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
The station we're authenticating/associating with
may not always be an AP in the sense that word is
mostly understood, so print only the MAC address
of the peer instead.
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
In order to use auth/assoc for different purposes
other than MLME, it needs to be split up. For other
purposes, a generic work handling (potentially on
another channel) will be useful.
To achieve that, this patch moves much of the MLME
work handling out of mlme into a new work API. The
API can currently handle probing a specific AP,
authentication and association. The MLME previously
handled probe/authentication as one step and will
continue to do so, but they are separate in the new
work handling.
Work items are RCU-managed to be able to check for
existence of an item for a specific frame in the RX
path, but they can be re-used which the MLME right
now will do for its combined probe/auth step.
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
As a first step of generalising management work,
this renames a few things and puts more information
directly into the struct so that auth/assoc need
not access the BSS pointer as often -- in fact it
can be removed from auth completely. Also since the
previous patch made sure a new work item is used
for association, we can make the different data a
union.
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
mac80211 currently hangs on to the auth state by
keeping it on the work list. That can lead to
confusing behaviour like rejecting scans while
authenticated to any AP (but not yet associated.)
It also means that it needs to keep track of the
work struct while associated for when it gets
disassociated (or disassociates.)
Change this to free the work struct after the
authentication completed successfully and
allocate a new one for associating, thereby
letting cfg80211 manage the auth state. Another
change necessary for this is to tell cfg80211
about all unicast deauth frames sent to mac80211
since now it can no longer check the auth state,
but that check was racy anyway.
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
We've long lacked a good confirmation that frames
have really gone out, e.g. before going off-channel
for a scan. Add a flush() operation that drivers
can implement to provide that confirmation, and use
it in a few places:
* before scanning sends the nullfunc frames
* after scanning sends the nullfunc frames, if any
* when going idle, to send any pending frames
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Instead of always using netif_running(sdata->dev)
use ieee80211_sdata_running(sdata) now which is
just an inline containing netif_running() for now.
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
mac80211 does not propagate failed hardware reconfiguration
requests. For suspend and resume this is important due to all
the possible issues that can come out of the suspend <-> resume
cycle. Not propagating the error means cfg80211 will assume
the resume for the device went through fine and mac80211 will
continue on trying to poke at the hardware, enable timers,
queue work, and so on for a device which is completley
unfunctional.
The least we can do is to propagate device start issues and
warn when this occurs upon resume. A side effect of this patch
is we also now propagate the start errors upon harware
reconfigurations (non-suspend), but this should also be desirable
anyway, there is not point in continuing to reconfigure a
device if mac80211 was unable to start the device.
For further details refer to the thread:
http://marc.info/?t=126151038700001&r=1&w=2
Cc: stable@kernel.org
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
When mac80211 suspends it calls a driver's suspend callback
as a last step and after that the driver assumes no calls will
be made to it until we resume and its start callback is kicked.
If such calls are made, however, suspend can end up throwing
hardware in an unexpected state and making the device unusable
upon resume.
Fix this by preventing mac80211 to schedule dynamic_ps_disable_work
by checking for when mac80211 starts to suspend and starts
quiescing. Frames should be allowed to go through though as
that is part of the quiescing steps and we do not flush the
mac80211 workqueue since it was already done towards the
beginning of suspend cycle.
The other mac80211 issue will be hanled in the next patch.
For further details see refer to the thread:
http://marc.info/?t=126144866100001&r=1&w=2
Cc: stable@kernel.org
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
If there's an invalid channel or SSID, the code leaks
the scan request. Always free the scan request, unless
it was successfully given to the driver.
Reported-by: Dan Carpenter <error27@gmail.com>
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Acked-by: Dan Carpenter <error27@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Joseph Nahmias reported, in http://bugs.debian.org/562016,
that he was getting the following warning (with some log
around the issue):
ath0: direct probe to AP 00:11:95:77:e0:b0 (try 1)
ath0: direct probe responded
ath0: authenticate with AP 00:11:95:77:e0:b0 (try 1)
ath0: authenticated
ath0: associate with AP 00:11:95:77:e0:b0 (try 1)
ath0: deauthenticating from 00:11:95:77:e0:b0 by local choice (reason=3)
ath0: direct probe to AP 00:11:95:77:e0:b0 (try 1)
ath0: RX AssocResp from 00:11:95:77:e0:b0 (capab=0x421 status=0 aid=2)
ath0: associated
------------[ cut here ]------------
WARNING: at net/wireless/mlme.c:97 cfg80211_send_rx_assoc+0x14d/0x152 [cfg80211]()
Hardware name: 7658CTO
...
Pid: 761, comm: phy0 Not tainted 2.6.32-trunk-686 #1
Call Trace:
[<c1030a5d>] ? warn_slowpath_common+0x5e/0x8a
[<c1030a93>] ? warn_slowpath_null+0xa/0xc
[<f86cafc7>] ? cfg80211_send_rx_assoc+0x14d/0x152
...
ath0: link becomes ready
ath0: deauthenticating from 00:11:95:77:e0:b0 by local choice (reason=3)
ath0: no IPv6 routers present
ath0: link is not ready
ath0: direct probe to AP 00:11:95:77:e0:b0 (try 1)
ath0: direct probe responded
ath0: authenticate with AP 00:11:95:77:e0:b0 (try 1)
ath0: authenticated
ath0: associate with AP 00:11:95:77:e0:b0 (try 1)
ath0: RX ReassocResp from 00:11:95:77:e0:b0 (capab=0x421 status=0 aid=2)
ath0: associated
It is not clear to me how the first "direct probe" here
happens, but this seems to be a race condition, if the
user requests to deauth after requesting assoc, but before
the assoc response is received. In that case, it may
happen that mac80211 tries to report the assoc success to
cfg80211, but gets blocked on the wdev lock that is held
because the user is requesting the deauth.
The result is that we run into a warning. This is mostly
harmless, but maybe cause an unexpected event to be sent
to userspace; we'd send an assoc success event although
userspace was no longer expecting that.
To fix this, remove the warning and check whether the
race happened and in that case abort processing.
Reported-by: Joseph Nahmias <joe@nahmias.net>
Cc: stable@kernel.org
Cc: 562016-quiet@bugs.debian.org
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
When fixed bssid is requested when joining an ibss network, incoming
beacons that match the configured bssid cause mac80211 to create new
sta entries, even before the ibss interface is in joined state.
When that happens, it fails to bring up the interface entirely, because
it checks for existing sta entries before joining.
This patch fixes this bug by refusing to create sta info entries before
the interface is fully operational.
Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
The SAP ref counter gets decremented twice when deleting a socket,
although for all but the first socket of a SAP the SAP ref counter was
incremented only once.
Signed-off-by: Octavian Purdila <opurdila@ixiacom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
For the cases where a lot of interfaces are used in conjunction with a
lot of LLC sockets bound to the same SAP, the iteration of the socket
list becomes prohibitively expensive.
Replacing the list with a a local address based hash significantly
improves the bind and listener lookup operations as well as the
datagram delivery.
Connected sockets delivery is also improved, but this patch does not
address the case where we have lots of sockets with the same local
address connected to different remote addresses.
In order to keep the socket sanity checks alive and fast a socket
counter was added to the SAP structure.
Signed-off-by: Octavian Purdila <opurdila@ixiacom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch adds a per SAP device based hash table to solve the
multicast delivery scalability issue when we have large number of
interfaces and a large number of sockets bound to the same SAP.
Signed-off-by: Octavian Purdila <opurdila@ixiacom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Optimize multicast delivery by doing the actual delivery without
holding the lock. Based on the same approach used in UDP code.
Signed-off-by: Octavian Purdila <opurdila@ixiacom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
For the reclamation phase we use the SLAB_DESTROY_BY_RCU mechanism,
which require some extra checks in the lookup code:
a) If the current socket was released, reallocated & inserted in
another list it will short circuit the iteration for the current list,
thus we need to restart the lookup.
b) If the current socket was released, reallocated & inserted in the
same list we just need to recheck it matches the look-up criteria and
if not we can skip to the next element.
In this case there is no need to restart the lookup, since sockets are
inserted at the start of the list and the worst that will happen is
that we will iterate throught some of the list elements more then
once.
Note that the /proc and multicast delivery was not yet converted to
RCU, it still uses spinlocks for protection.
Signed-off-by: Octavian Purdila <opurdila@ixiacom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Using bind(MAC address) with LLC sockets has O(n) complexity, where n
is the number of interfaces. To overcome this, we add support for
SO_BINDTODEVICE which drops the complexity to O(1).
Signed-off-by: Octavian Purdila <opurdila@ixiacom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Using dev_hard_header allows us to use LLC with VLANs and potentially
other Ethernet/TokernRing specific encapsulations. It also removes code
duplication between LLC and Ethernet/TokenRing core code.
Signed-off-by: Octavian Purdila <opurdila@ixiacom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Apparently some awk versions choke on C-style comments -- who knew? :-)
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add rtnetlink init_rcvwnd to set the TCP initial receive window size
advertised by passive and active TCP connections.
The current Linux TCP implementation limits the advertised TCP initial
receive window to the one prescribed by slow start. For short lived
TCP connections used for transaction type of traffic (i.e. http
requests), bounding the advertised TCP initial receive window results
in increased latency to complete the transaction.
Support for setting initial congestion window is already supported
using rtnetlink init_cwnd, but the feature is useless without the
ability to set a larger TCP initial receive window.
The rtnetlink init_rcvwnd allows increasing the TCP initial receive
window, allowing TCP connection to advertise larger TCP receive window
than the ones bounded by slow start.
Signed-off-by: Laurent Chavey <chavey@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>