This fixes a problem caused by the overlap of the connection-setup and
established-state phases of DCCP connections.
During connection setup, the client retransmits Confirm Feature-Negotiation
options until a response from the server signals that it can move from the
half-established PARTOPEN into the OPEN state, whereupon the connection is
fully established on both ends (RFC 4340, 8.1.5).
However, since the client may already send data while it is in the PARTOPEN
state, consequences arise for the Maximum Packet Size: the problem is that the
initial option overhead is much higher than for the subsequent established
phase, as it involves potentially many variable-length list-type options
(server-priority options, RFC 4340, 6.4).
Applying the standard MPS is insufficient here: especially with larger
payloads this can lead to annoying, counter-intuitive EMSGSIZE errors.
On the other hand, reducing the MPS available for the established phase by
the added initial overhead is highly wasteful and inefficient.
The solution chosen therefore is a two-phase strategy:
If the payload length of the DataAck in PARTOPEN is too large, an Ack is sent
to carry the options, and the feature-negotiation list is then flushed.
This means that the server gets two Acks for one Response. If both Acks get
lost, it is probably better to restart the connection anyway and devising yet
another special-case does not seem worth the extra complexity.
The result is a higher utilisation of the available packet space for the data
transmission phase (established state) of a connection.
The patch (over-)estimates the initial overhead to be 32*4 bytes -- commonly
seen values were around 90 bytes for initial feature-negotiation options.
It uses sizeof(u32) to mean "aligned units of 4 bytes".
For consistency, another use of 4-byte alignment is adapted.
Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch resolves a long-standing FIXME to dynamically update the Maximum
Packet Size depending on actual options usage.
It uses the flags set by the feature-negotiation infrastructure to compute
the required header option size.
Most options are fixed-size, a notable exception are Ack Vectors (required
currently only by CCID-2). These can have any length between 3 and 1020
bytes. As a result of testing, 16 bytes (2 bytes for type/length plus 14 Ack
Vector cells) have been found to be sufficient for loss-free situations.
There are currently no CCID-specific header options which may appear on data
packets, thus it is not necessary to define a corresponding CCID field as
suggested in the old comment.
Further changes:
----------------
Adjusted the type of 'cur_mps' to match the unsigned return type of the
function.
Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
Acked-by: Ian McDonald <ian.mcdonald@jandi.co.nz>
Signed-off-by: David S. Miller <davem@davemloft.net>
I guess these fields were one day 16-bit in the struct but
nowadays they're just using 8 bits anyway.
This is just a precaution, didn't result any change in my
case but who knows what all those varying gcc versions &
options do. I've been told that 16-bit is not so nice with
some cpus.
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
copied was assigned zero right before the goto, so if (copied)
cannot ever be true.
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
Also fixes insignificant bug that would cause sending of stale
SACK block (would occur in some corner cases).
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
It seems that implementation in yeah was inconsistent to what
other did as it would increase cwnd one ack earlier than the
others do.
Size benefits:
bictcp_cong_avoid | -36
tcp_cong_avoid_ai | +52
bictcp_cong_avoid | -34
tcp_scalable_cong_avoid | -36
tcp_veno_cong_avoid | -12
tcp_yeah_cong_avoid | -38
= -104 bytes total
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
Similar to what is done elsewhere in TCP code when double
state checks are being done.
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
Redundant checks made indentation impossible to follow.
However, it might be useful to make this ca_state+is_sack
indexed array.
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
Some comment about its current state added. So far I have
seen very few cases where the thing is actually useful,
usually just marginally (though admittedly I don't usually
see top of window losses where it seems possible that there
could be some gain), instead, more often the cases suffer
from L-marking spike which is certainly not desirable
(I'll bury improving it to my todo list, but on a low
prio position).
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
Arnd Hannemann <hannemann@nets.rwth-aachen.de> noticed and was
puzzled by the fact that !tcp_is_fack(tp) leads to early return
near the beginning and the later on tcp_is_fack(tp) was still
used in an if condition. The later check was a left-over from
RFC3517 SACK stuff (== !tcp_is_fack(tp) behavior nowadays) as
there wasn't clear way how to handle this particular check
cheaply in the spirit of RFC3517 (using only SACK blocks, not
holes + SACK blocks as with FACK). I sort of left it there as
a reminder but since it's confusing other people just remove
it and comment the missing-feature stuff instead.
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Cc: Arnd Hannemann <hannemann@nets.rwth-aachen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
If cur_mss grew very recently so that the previously G/TSOed skb
now fits well into a single segment it would get send up in
parts unless we calculate # of segments again. This corner-case
could happen eg. after mtu probe completes or less than
previously sack blocks are required for the opposite direction.
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
1) We didn't remove any skbs, so no need to handle stale refs.
2) scoreboard_skb_hint is trivial, no timestamps were changed
so no need to clear that one
3) lost_skb_hint needs tweaking similar to that of
tcp_sacktag_one().
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
If skb can be sent right away, we certainly should do that
if it's in the middle of the queue because it won't get
more data into it.
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
It is possible that lost_cnt_hint gets underflow in
tcp_clean_rtx_queue because the cumulative ACK can cover
the segment where lost_skb_hint points to only partially,
which means that the hint is not cleared, opposite to what
my (earlier) comment claimed.
Also I don't agree what I ended up writing about non-trivial
case there to be what I intented to say. It was not supposed
to happen that the hint won't get cleared and we underflow
in any scenario.
In general, this is quite hard to trigger in practice.
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
Backtracking to sacked skbs is a horrible performance killer
since the hint cannot be advanced successfully past them...
...And it's totally unnecessary too.
In theory this is 2.6.27..28 regression but I doubt anybody
can make .28 to have worse performance because of other TCP
improvements.
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
As reported by Stephen Rothwell.
> Today's linux-next build (powerpc allyesconfig) failed like this:
>
> net/rds/cong.c: In function 'rds_cong_set_bit':
> net/rds/cong.c:284: error: implicit declaration of function 'generic___set_le_bit'
> net/rds/cong.c: In function 'rds_cong_clear_bit':
> net/rds/cong.c:298: error: implicit declaration of function 'generic___clear_le_bit'
> net/rds/cong.c: In function 'rds_cong_test_bit':
> net/rds/cong.c:309: error: implicit declaration of function 'generic_test_le_bit'
Signed-off-by: David S. Miller <davem@davemloft.net>
There's conflicting assumptions in shifting, the caller assumes
that dupsack results in S'ed skbs (or a part of it) for sure but
never gave a hint to tcp_sacktag_one when dsack is actually in
use. Thus DSACK retrans_out -= pcount was not taken and the
counter became out of sync. Remove obstacle from that information
flow to get DSACKs accounted in tcp_sacktag_one as expected.
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Tested-by: Denys Fedoryshchenko <denys@visp.net.lb>
Signed-off-by: David S. Miller <davem@davemloft.net>
The netpoll entry checks are required to ensure that we don't
receive normal packets when invoked via netpoll. Unfortunately
it only ever worked for the netif_receive_skb/netif_rx entry
points. The VLAN (and subsequently GRO) entry point didn't
have the check and therefore can trigger all sorts of weird
problems.
This patch adds the netpoll check to all entry points.
I'm still uneasy with receiving at all under netpoll (which
apparently is only used by the out-of-tree kdump code). The
reason is it is perfectly legal to receive all data including
headers into highmem if netpoll is off, but if you try to do
that with netpoll on and someone gets a printk in an IRQ handler
you're going to get a nice BUG_ON.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
"cfg80211: add assert_cfg80211_lock() to ensure proper protection"
added assert_cfg80211_lock() calls into various places. At least
one of them, nl80211_send_wiphy(), should not have been there. That
triggers the BUG_ON in assert_cfg80211_lock() and pretty much kills
the kernel whenever someone runs hostapd.. Remove that call and make
assert_cfg80211_lock() use WARN_ON instead of BUG_ON to be a bit more
friendly to users.
Signed-off-by: Jouni Malinen <jouni.malinen@atheros.com>
Acked-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Give slow hardware some time to do the TSF sync, to not run into an
IBSS merging endless loop in some rarely situations.
Signed-off-by: Alina Friedrichsen <x-alina@gmx.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
It was not a good idea to do a TSF reset on strange IBSS merges to the same BSSID. For example it will break the TSF sync of ath9k completely and it is unnecessary as all hardware I have tested do a TSF sync to a higher value automatically and IBSS merges are only done to higher TSF values. It only need a TSF reset to accept a lower value, when the IBSS network is changed manually.
Signed-off-by: Alina Friedrichsen <x-alina@gmx.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Since we already have a regulatory request from the workqueue use that
and avoid a new kzalloc()
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
We were passing value by value, lets just pass the struct.
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
When devices are world roaming they cannot beacon or do active scan
on 5 GHz or on channels 12, 13 and 14 on the 2 GHz band. Although
we have a good regulatory API some cards may _always_ world roam, this
is also true when a system does not have CRDA present. Devices doing world
roaming can still passive scan, if they find a beacon from an AP on
one of the world roaming frequencies we make the assumption we can do
the same and we also remove the passive scan requirement.
This adds support for providing beacon regulatory hints based on scans.
This works for devices that do either hardware or software scanning.
If a channel has not yet been marked as having had a beacon present
on it we queue the beacon hint processing into the workqueue.
All wireless devices will benefit from beacon regulatory hints from
any wireless device on a system including new devices connected to
the system at a later time.
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
The current static world regulatory domain is too restrictive,
we can use some 5 GHz channels world wide so long as they do not
touch frequencies which require DFS. The compromise is we must
also enforce passive scanning and disallow usage of a mode of
operation that beacons: (AP | IBSS | Mesh)
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
This enables active scan and beaconing on Channels 1 through 11
on the static world regulatory domain.
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
This allows drivers that agree on regulatory to share their
regulatory domain.
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
All regulatory hints (core, driver, userspace and 11d) are now processed in
a workqueue.
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
This was never happening but it was still wrong, so correct it.
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Truth of the matter this was confusing people so mark it as
unlikely as that is the case now.
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
We were not protecting last_request there is a small possible race
between an 11d hint and another routine which calls reset_regdomains()
which can prevent a valid country IE from being processed. This is
not critical as it will still be procesed soon after but locking prior
to it is correct.
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
We do this so later on we can move the pending requests onto a
workqueue. By using the wiphy_idx instead of the wiphy we can
later easily check if the wiphy has disappeared or not.
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Calling kobject_uevent_env() can fail mainly due to out of
memory conditions. We do not want to continue during such
conditions so propagate that as well instead of letting
cfg80211 load as if everything is peachy.
Additionally lets clarify that when CRDA is not called during
cfg80211's initialization _and_ if the error is not an -ENOMEM
its because kobject_uevent_env() failed to call CRDA, not because
CRDA failed. For those who want to find out why we also let you
do so by enabling the kernel config CONFIG_CFG80211_REG_DEBUG --
you'll get an actual stack trace.
So for now we'll treat non -ENOMEM kobject_uevent_env() failures as
non fatal during cfg80211's initialization.
CC: Greg KH <greg@kroah.com>
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
This makes the core hint path more readable and allows for us to
later make it obvious under what circumstances we need locking or not.
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
If cfg80211 is built into the kernel there is perhaps a small
time window betwen nl80211_init() and regulatory_init() where
cfg80211_regdomain hasn't yet been initialized to let the
wireless core do its work. During that rare case and time
frame (if its even possible) we don't allow user regulatory
changes as cfg80211 is working on enabling its first regulatory
domain.
To check for cfg80211_regdomain we now contend the entire operation
using the cfg80211_mutex.
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
cfg80211_drv_mutex is protecting more than the driver list,
this renames it and documents what its currently supposed to
protect.
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
This will later be used by others, for now make use of it in
cfg80211_drv_by_wiphy_idx() to return early if an invalid
wiphy_idx has been provided.
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Makes it clearer to read when comparing to ifidx
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
This patch introduces a generic commit() function which initiate a
new network joining process. It should be called after some interface
config changes, so that the changes get applied more cleanly. Currently
set_ssid() and set_bssid() call it. Others can be added in future
patches.
In version 1 the header files was forgotten, sorry.
Signed-off-by: Alina Friedrichsen <x-alina@gmx.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
This adds optional notifier functions for software scan.
Signed-off-by: Michael Buesch <mb@bu3sch.de>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
The previous patch made cfg80211 generally aware of the signal
type a given hardware will give, so now it can implement
SIOCGIWRANGE itself, removing more wext stuff from mac80211.
Might need to be a little more parametrized once we have
more hardware using cfg80211 and new hardware capabilities.
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
It wasn't a good idea to make the signal type a per-BSS option,
although then it is closer to the actual value. Move it to be
a per-wiphy setting, update mac80211 to match.
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
