pandora-kernel.git
13 years agocnic: Add support for 57712 device
Michael Chan [Wed, 13 Oct 2010 14:06:51 +0000 (14:06 +0000)]
cnic: Add support for 57712 device

Add new interrupt ack functions and other hardware interface logic to
support the new device.

Update version to 2.2.6.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agocnic: Decouple uio close from cnic shutdown
Michael Chan [Wed, 13 Oct 2010 14:06:50 +0000 (14:06 +0000)]
cnic: Decouple uio close from cnic shutdown

During cnic shutdown, the original driver code requires userspace to
close the uio device within a few seconds.  This doesn't always happen
as the userapp may be hung or otherwise take a long time to close.  The
system may crash when this happens.

We fix the problem by decoupling the uio structures from the cnic
structures during cnic shutdown.  We do not unregister the uio device
until the cnic driver is unloaded.  This eliminates the unreliable wait
loop for uio to close.

All uio structures are kept in a linked list.  If the device is shutdown
and later brought back up again, the uio strcture will be found in the
linked list and coupled back to the cnic structures.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agocnic: Add cnic_uio_dev struct
Michael Chan [Wed, 13 Oct 2010 14:06:49 +0000 (14:06 +0000)]
cnic: Add cnic_uio_dev struct

and put all uio related structures and ring buffers in it.  This allows
uio operations to be done independent of the cnic device structures.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agocnic: Add cnic_free_uio()
Michael Chan [Wed, 13 Oct 2010 14:06:48 +0000 (14:06 +0000)]
cnic: Add cnic_free_uio()

to free all UIO related structures.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agocnic: Defer iscsi connection cleanup
Michael Chan [Wed, 13 Oct 2010 14:06:47 +0000 (14:06 +0000)]
cnic: Defer iscsi connection cleanup

The bnx2x devices require a 2 second quiet time before sending the last
RAMROD command to destroy a connection.  This sleep wait adds up to a
long delay when iscsid is serially destroying maultiple connections.

Create a workqueue to perform the final connection cleanup in the
background to speed up the process.  This significantly speeds up the
process as the wait time can be done in parallel for multiple connections.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agocnic: Add cnic_bnx2x_destroy_ramrod()
Michael Chan [Wed, 13 Oct 2010 14:06:46 +0000 (14:06 +0000)]
cnic: Add cnic_bnx2x_destroy_ramrod()

Refactoring code for the next patch to defer connection clean up.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agocnic: Convert ctx_flags to bit fields
Michael Chan [Wed, 13 Oct 2010 14:06:45 +0000 (14:06 +0000)]
cnic: Convert ctx_flags to bit fields

so that we can additional bit definitions without requiring a spinlock.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agocnic: Add common cnic_request_irq()
Michael Chan [Wed, 13 Oct 2010 14:06:44 +0000 (14:06 +0000)]
cnic: Add common cnic_request_irq()

to reduce some duplicate code.  Also, use tasklet_kill() in
cnic_free_irq() to wait for the cnic_irq_task to complete.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agoiwlwifi: move agn only eeprom functions to separate file
Wey-Yi Guy [Fri, 8 Oct 2010 23:05:19 +0000 (16:05 -0700)]
iwlwifi: move agn only eeprom functions to separate file

Some of the functions in iwl-eeprom.c file are for agn devices only,
Those functions do not have to be part of iwlcore.ko, so move those
to iwl-agn-eeprom.c file.

Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
13 years agoiwlagn: check beacon frame size
Johannes Berg [Thu, 7 Oct 2010 11:00:33 +0000 (04:00 -0700)]
iwlagn: check beacon frame size

When the beacon_skb is NULL, we might still
attempt to use it in this code path (if we
ever get here) -- make the code a bit more
defensive and check the return value of
iwl_fill_beacon_frame() against zero.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
13 years agoiwlwifi: blink LED in IBSS mode
Johannes Berg [Thu, 7 Oct 2010 10:52:35 +0000 (03:52 -0700)]
iwlwifi: blink LED in IBSS mode

We recently found that contrary to expectations,
the LED is not blinking in IBSS mode. Fix this.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
13 years agoiwlwifi: rewrite RXON checks
Johannes Berg [Thu, 7 Oct 2010 11:02:03 +0000 (04:02 -0700)]
iwlwifi: rewrite RXON checks

The RXON checking is a bit magical, and prints
out too much information if something goes wrong.
Make it less magical and print out only the items
that were actually wrong.

Also remove the comment about removing it -- the
driver is constantly changing so these checks are
useful.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
13 years agoiwlagn: 6050 ops should be used;
Wey-Yi Guy [Wed, 6 Oct 2010 20:46:11 +0000 (13:46 -0700)]
iwlagn: 6050 ops should be used;

For 6050 series of devices, 6050 ops should be used;
One of the 6050 config still use 6000 ops, fix it.

Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
13 years agoiwlwifi: clean up some beacon handling
Johannes Berg [Mon, 4 Oct 2010 12:50:36 +0000 (05:50 -0700)]
iwlwifi: clean up some beacon handling

There's no need to check for NULL before
calling dev_kfree_skb() since it is valid
to call it on NULL -- it becomes a no-op.

There's also no need to initialise the
beacon_skb variable to NULL just after
the memory it is in has been kzalloc'ed.

Some minor whitespace cleanups, and a
lock assertion in a function that needs
the mutex (to access the beacon_skb var)
complete the patch.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
13 years agoiwlwifi: rename ibss_beacon variable
Johannes Berg [Mon, 4 Oct 2010 12:50:06 +0000 (05:50 -0700)]
iwlwifi: rename ibss_beacon variable

Since we're also going to support AP (GO) mode,
the variable isn't used for just IBSS beacons
any more -- rename it to not mislead readers.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
13 years agoiwlwifi: allow probe-after-rx on 2.4 GHz
Johannes Berg [Mon, 4 Oct 2010 12:47:23 +0000 (05:47 -0700)]
iwlwifi: allow probe-after-rx on 2.4 GHz

There are two passive 2.4 GHz channels: 12 and 13.
If you have a hidden SSID on those, you will not
be able to connect to it because we don't send out
probe requests there. We can allow this by using
the firmware's probe-after-rx functionality on
those channels as well.

This fixes
https://bugzilla.kernel.org/show_bug.cgi?id=16462

Reported-by: Daniel J Blueman <daniel.blueman@gmail.com>
Tested-by: Daniel J Blueman <daniel.blueman@gmail.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
13 years agoiwlgn: need longer tx queue stuck timer for coex devices
Wey-Yi Guy [Thu, 23 Sep 2010 22:24:22 +0000 (15:24 -0700)]
iwlgn: need longer tx queue stuck timer for coex devices

For BT/WiFi combo devices, need longer tx stuck queue
timer, so those devices won't reload firmware too often.

Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
13 years agoDocumentation: Update Phonet doc for Pipe controller changes
Kumar Sanghvi [Tue, 12 Oct 2010 20:17:25 +0000 (20:17 +0000)]
Documentation: Update Phonet doc for Pipe controller changes

Updates to Phonet doc for Pipe controller 'connect' socket
implementation and changes related to socket options.

Signed-off-by: Kumar Sanghvi <kumar.sanghvi@stericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agoPhonet: 'connect' socket implementation for Pipe controller
Kumar Sanghvi [Tue, 12 Oct 2010 20:14:43 +0000 (20:14 +0000)]
Phonet: 'connect' socket implementation for Pipe controller

Based on suggestion by Rémi Denis-Courmont to implement 'connect'
for Pipe controller logic,  this patch implements 'connect' socket
call for the Pipe controller logic.
The patch does following:-
- Removes setsockopts for PNPIPE_CREATE and PNPIPE_DESTROY
- Adds setsockopt for setting the Pipe handle value
- Implements connect socket call
- Updates the Pipe controller logic

User-space should now follow below sequence with Pipe controller:-
-socket
-bind
-setsockopt for PNPIPE_PIPE_HANDLE
-connect
-setsockopt for PNPIPE_ENCAP_IP
-setsockopt for PNPIPE_ENABLE

GPRS/3G data has been tested working fine with this.

Signed-off-by: Kumar Sanghvi <kumar.sanghvi@stericsson.com>
Acked-by: Rémi Denis-Courmont <remi.denis-courmont@nokia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agotipc: clean out all instances of #if 0'd unused code
Paul Gortmaker [Tue, 12 Oct 2010 14:25:58 +0000 (14:25 +0000)]
tipc: clean out all instances of #if 0'd unused code

Remove all instances of legacy, or as yet to be implemented code
that is currently living within an #if 0 ... #endif block.
In the rare instance that some of it be needed in the future,
it can still be dragged out of history, but there is no need
for it to sit in mainline.

Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agoath9k: Fix potential use-after-free.
Ben Greear [Wed, 13 Oct 2010 19:01:23 +0000 (12:01 -0700)]
ath9k: Fix potential use-after-free.

The ath_debug_stat_tx references bf->bf_mpdu, which
is the skb consumed by  ath_tx_complete.  So, call
the ath_debug_stat_tx method first.

Signed-off-by: Ben Greear <greearb@candelatech.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agomac80211: fix SMPS request
Johannes Berg [Wed, 13 Oct 2010 17:23:21 +0000 (19:23 +0200)]
mac80211: fix SMPS request

It looks like I submitted a different patch
than I tested, because clearly the code in
mac80211 is missing actually propagating the
requested SMPS mode. Fix that!

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agoath9k: Set RX filter for Probe Request based on filter flag
Jouni Malinen [Wed, 13 Oct 2010 14:29:31 +0000 (17:29 +0300)]
ath9k: Set RX filter for Probe Request based on filter flag

This allows mac80211 to enable receiving of Probe Request frames in
station mode which is needed for P2P.

Signed-off-by: Jouni Malinen <jouni.malinen@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agomac80211: add probe request filter flag
Johannes Berg [Wed, 13 Oct 2010 10:06:24 +0000 (12:06 +0200)]
mac80211: add probe request filter flag

Using the frame registration notification, we
can see when probe requests are requested and
notify the low-level driver via filtering. The
flag is also set in AP and IBSS modes.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agocfg80211: notify drivers about frame registrations
Johannes Berg [Wed, 13 Oct 2010 10:06:23 +0000 (12:06 +0200)]
cfg80211: notify drivers about frame registrations

Drivers may need to adjust their filters according
to frame registrations, so notify them about them.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agoath9k: Fix documentation in rate control
Mohammed Shafi Shajakhan [Wed, 13 Oct 2010 06:17:09 +0000 (11:47 +0530)]
ath9k: Fix documentation in rate control

This fix updates the documenation in Rate Control Table structure

Signed-off-by: Mohammed Shafi Shajakhan <mshajakhan@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agoath5k: Move debugfs under ieee80211/[wiphy-name]
Ben Greear [Tue, 12 Oct 2010 17:55:38 +0000 (10:55 -0700)]
ath5k: Move debugfs under ieee80211/[wiphy-name]

This automatically keeps things proper when wiphy
is renamed.

Based on patch by Johannes Berg <johannes@sipsolutions.net>

Signed-off-by: Ben Greear <greearb@candelatech.com>
Acked-by: Bruno Randolf <br1@einfach.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agowext: fix alignment problem in serializing 'struct iw_point'
Gerrit Renker [Tue, 12 Oct 2010 05:07:42 +0000 (07:07 +0200)]
wext: fix alignment problem in serializing 'struct iw_point'

wext: fix alignment problem in serializing 'struct iw_point'

This fixes a typo in the definition of the serialized length of struct iw_point:
 a) wireless.h is exported to userspace, the typo causes IW_EV_POINT_PK_LEN
    to be 12 on 64-bit, and 8 on 32-bit systems (causing misalignment);
 b) in compat-64 mode iwe_stream_add_point() memcpys overlap (see below).

The second case in  in compat-64 mode looks like (variable names are as in
include/net/iw_handler.h:iwe_stream_add_point()):

 point_len = IW_EV_COMPAT_POINT_LEN = 8
 lcp_len   = IW_EV_COMPAT_LCP_LEN   = 4
 2nd memcpy: IW_EV_POINT_PK_LEN - IW_EV_LCP_PK_LEN = 12 - 4 = 8

 IW_EV_LCP_PK_LEN
 <-------------->                *---> 'extra' data area
 +-------+-------+-------+-------+---------------+------- ...-+
 | len   | cmd   |length | flags |  (empty) -> extra      ... |
 +-------+-------+-------+-------+---------------+------- ...-+
    2       2       2       2          4

     lcp_len
 <-------------->                <-!! OVERLAP !!>
 <--1st memcpy--><------- 2nd memcpy ----------->
                                 <---- 3rd memcpy ------- ... >
 <--------- point_len ---------->

This case could cause overrun whenever iw_point.length < 4.
The other two cases are -
 * 32-bit systems: IW_EV_POINT_PK_LEN - IW_EV_LCP_PK_LEN =  8 - 4 = 4,
   the second memcpy copies exactly the 4 required bytes;
 * 64-bit systems: IW_EV_POINT_PK_LEN - IW_EV_LCP_PK_LEN = 12 - 4 = 8,
   the second memcpy copies a superfluous (but non overlapping) 4 bytes.

The patch changes IW_EV_POINT_PK_LEN to be 8, so that in all 3 cases always only
the requested iw_point.{length,flags} (both __u16) are copied, avoiding overrrun
(compat-64) and superfluous copy (64-bit). In addition, the userspace header is
sanitized (in agreement with version 30 of the wireless tools).

Many thanks to Johannes Berg for help and review with this patch.

Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agos390: ctcm_mpc: Fix build after netdev refcount changes.
David S. Miller [Wed, 13 Oct 2010 16:11:26 +0000 (09:11 -0700)]
s390: ctcm_mpc: Fix build after netdev refcount changes.

Reported-by: Sachin Sant <sachinp@in.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agoath5k: fix build break from "ath5k: Print out opmode in debugfs"
Joe Perches [Tue, 12 Oct 2010 18:07:44 +0000 (11:07 -0700)]
ath5k: fix build break from "ath5k: Print out opmode in debugfs"

Also improve ath_opmode_to_string usage by having it return UNKNOWN
rather than NULL in the event of failure to map the opmode value to a
representative string.

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agowireless: Print wiphy name in sysfs.
Ben Greear [Mon, 11 Oct 2010 17:28:59 +0000 (10:28 -0700)]
wireless: Print wiphy name in sysfs.

The index cannot be used to reliably reconstruct a phy
name, so explicitly add the phy name to sysfs so that scripts
can figure out the parent phy device for a particular
wireless interface.

Signed-off-by: Ben Greear <greearb@candelatech.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agoMerge branch 'wireless-next-2.6' of git://git.kernel.org/pub/scm/linux/kernel/git...
John W. Linville [Tue, 12 Oct 2010 19:52:25 +0000 (15:52 -0400)]
Merge branch 'wireless-next-2.6' of git://git./linux/kernel/git/iwlwifi/iwlwifi-2.6

13 years agonet: percpu net_device refcount
Eric Dumazet [Mon, 11 Oct 2010 10:22:12 +0000 (10:22 +0000)]
net: percpu net_device refcount

We tried very hard to remove all possible dev_hold()/dev_put() pairs in
network stack, using RCU conversions.

There is still an unavoidable device refcount change for every dst we
create/destroy, and this can slow down some workloads (routers or some
app servers, mmap af_packet)

We can switch to a percpu refcount implementation, now dynamic per_cpu
infrastructure is mature. On a 64 cpus machine, this consumes 256 bytes
per device.

On x86, dev_hold(dev) code :

before
        lock    incl 0x280(%ebx)
after:
        movl    0x260(%ebx),%eax
        incl    fs:(%eax)

Stress bench :

(Sending 160.000.000 UDP frames,
IP route cache disabled, dual E5540 @2.53GHz,
32bit kernel, FIB_TRIE)

Before:

real    1m1.662s
user    0m14.373s
sys     12m55.960s

After:

real    0m51.179s
user    0m15.329s
sys     10m15.942s

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agobnx2x: Fixing a typo: added a missing RSS enablement
Dmitry Kravkov [Tue, 12 Oct 2010 09:02:21 +0000 (09:02 +0000)]
bnx2x: Fixing a typo: added a missing RSS enablement

Reported-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Dmitry Kravkov <dmitry@broadcom.com>
Signed-off-by: Vladislav Zolotarov <vladz@broadcom.com>
Tested-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agoMerge branch 'dccp' of git://eden-feed.erg.abdn.ac.uk/net-next-2.6
David S. Miller [Tue, 12 Oct 2010 18:43:42 +0000 (11:43 -0700)]
Merge branch 'dccp' of git://eden-feed.erg.abdn.ac.uk/net-next-2.6

13 years agoBluetooth: update MAINTAINERS for Bluetooth subsys
Gustavo F. Padovan [Fri, 8 Oct 2010 12:13:28 +0000 (09:13 -0300)]
Bluetooth: update MAINTAINERS for Bluetooth subsys

Add myself to MAINTAINERS and update the git trees.

Acked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
13 years agoBluetooth: hci_uart: Fix typo in stats for sco tx
Karl Beldan [Thu, 7 Oct 2010 19:57:10 +0000 (21:57 +0200)]
Bluetooth: hci_uart: Fix typo in stats for sco tx

s/stat.cmd_tx++/stat.sco_tx++ for HCI_SCODATA_PKT

Signed-off-by: Karl Beldan <karl.beldan@gmail.com>
Acked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
13 years agoBluetooth: clean up rfcomm code
Andrei Emeltchenko [Fri, 1 Oct 2010 09:05:11 +0000 (12:05 +0300)]
Bluetooth: clean up rfcomm code

Remove dead code and unused rfcomm thread events

Signed-off-by: Andrei Emeltchenko <andrei.emeltchenko@nokia.com>
Acked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
13 years agoBluetooth: Update conf_state before send config_req out
Haijun Liu [Thu, 30 Sep 2010 08:52:40 +0000 (16:52 +0800)]
Bluetooth: Update conf_state before send config_req out

Update conf_state with L2CAP_CONF_REQ_SENT before send config_req out in
l2cap_config_req().

Signed-off-by: Haijun Liu <haijun.liu@atheros.com>
Acked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
13 years agoBluetooth: Use the proper error value from bt_skb_send_alloc()
Gustavo F. Padovan [Fri, 24 Sep 2010 23:30:57 +0000 (20:30 -0300)]
Bluetooth: Use the proper error value from bt_skb_send_alloc()

&err points to the proper error set by bt_skb_send_alloc() when it
fails.

Acked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
13 years agoBluetooth: make batostr() print in the right order
Gustavo F. Padovan [Fri, 3 Sep 2010 21:29:46 +0000 (18:29 -0300)]
Bluetooth: make batostr() print in the right order

The Bluetooth core uses the the BD_ADDR in the opposite order from the
human readable order. So we are changing batostr() to print in the
correct order and then removing some baswap(), as they are not needed
anymore.

Acked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
13 years agoBluetooth: remove unused variable from cmtp
Gustavo F. Padovan [Fri, 3 Sep 2010 20:51:57 +0000 (17:51 -0300)]
Bluetooth: remove unused variable from cmtp

A value was attributed to 'src', but no one was using.

Acked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
13 years agoBluetooth: check for l2cap header in start fragment
Andrei Emeltchenko [Wed, 15 Sep 2010 11:28:43 +0000 (14:28 +0300)]
Bluetooth: check for l2cap header in start fragment

BLUETOOTH SPECIFICATION Version 4.0 [Vol 3] page 36 mentioned
"Note: Start Fragments always begin with the Basic L2CAP header
of a PDU."

Signed-off-by: Andrei Emeltchenko <andrei.emeltchenko@nokia.com>
Acked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
13 years agoBluetooth: check L2CAP length in first ACL fragment
Andrei Emeltchenko [Wed, 15 Sep 2010 11:28:44 +0000 (14:28 +0300)]
Bluetooth: check L2CAP length in first ACL fragment

Current Bluetooth code assembles fragments of big L2CAP packets
in l2cap_recv_acldata and then checks allowed L2CAP size in
assemled L2CAP packet (pi->imtu < skb->len).

The patch moves allowed L2CAP size check to the early stage when
we receive the first fragment of L2CAP packet. We do not need to
reserve and keep L2CAP fragments for bad packets.

Updated version after comments from Mat Martineau <mathewm@codeaurora.org>
and Gustavo Padovan <padovan@profusion.mobi>.

Trace below is received when using stress tools sending big
fragmented L2CAP packets.
...
[ 1712.798492] swapper: page allocation failure. order:4, mode:0x4020
[ 1712.804809] [<c0031870>] (unwind_backtrace+0x0/0xdc) from [<c00a1f70>]
(__alloc_pages_nodemask+0x4)
[ 1712.814666] [<c00a1f70>] (__alloc_pages_nodemask+0x47c/0x4d4) from
[<c00a1fd8>] (__get_free_pages+)
[ 1712.824645] [<c00a1fd8>] (__get_free_pages+0x10/0x3c) from [<c026eb5c>]
(__alloc_skb+0x4c/0xfc)
[ 1712.833465] [<c026eb5c>] (__alloc_skb+0x4c/0xfc) from [<bf28c738>]
(l2cap_recv_acldata+0xf0/0x1f8 )
[ 1712.843322] [<bf28c738>] (l2cap_recv_acldata+0xf0/0x1f8 [l2cap]) from
[<bf0094ac>] (hci_rx_task+0x)
...

Signed-off-by: Andrei Emeltchenko <andrei.emeltchenko@nokia.com>
Acked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
13 years agoBluetooth: Add support Bluetooth controller of MacbookPro 7,1
Nobuhiro Iwamatsu [Fri, 20 Aug 2010 07:24:07 +0000 (16:24 +0900)]
Bluetooth: Add support Bluetooth controller of MacbookPro 7,1

Bluetooth controller of MacbookPro 7,1 does not work.
Because Device Class of these controllers was set 255 (Vendor Sepecific Class).

T:  Bus=04 Lev=02 Prnt=04 Port=00 Cnt=01 Dev#=  5 Spd=12  MxCh= 0
D:  Ver= 2.00 Cls=ff(vend.) Sub=01 Prot=01 MxPS=64 #Cfgs=  1
P:  Vendor=05ac ProdID=8213 Rev=01.86
S:  Manufacturer=Apple Inc.
S:  Product=Bluetooth USB Host Controller
S:  SerialNumber=5C5948C81B99
C:  #Ifs= 4 Cfg#= 1 Atr=e0 MxPwr=0mA
I:  If#= 0 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=01 Prot=01 Driver=btusb
I:  If#= 1 Alt= 0 #EPs= 2 Cls=e0(wlcon) Sub=01 Prot=01 Driver=btusb
I:  If#= 2 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=ff Prot=ff Driver=(none)
I:  If#= 3 Alt= 0 #EPs= 0 Cls=fe(app. ) Sub=01 Prot=00 Driver=(none)

Signed-off-by: Nobuhiro Iwamatsu <iwamatsu@nigauri.org>
Acked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
13 years agoBluetooth: Add support Bluetooth controller of MacbookPro 6,2
Nobuhiro Iwamatsu [Fri, 20 Aug 2010 07:24:06 +0000 (16:24 +0900)]
Bluetooth: Add support Bluetooth controller of MacbookPro 6,2

Bluetooth controller of MacbookPro 6,2 does not work.
Because Device Class of these controllers was set 255 (Vendor Sepecific Class).

T:  Bus=01 Lev=03 Prnt=03 Port=02 Cnt=03 Dev#=  8 Spd=12  MxCh= 0
D:  Ver= 2.00 Cls=ff(vend.) Sub=01 Prot=01 MxPS=64 #Cfgs=  1
P:  Vendor=05ac ProdID=8218 Rev=00.22
S:  Manufacturer=Apple Inc.
S:  Product=Bluetooth USB Host Controller
C:  #Ifs= 4 Cfg#= 1 Atr=e0 MxPwr=0mA
I:  If#= 0 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=01 Prot=01 Driver=(none)
I:  If#= 1 Alt= 0 #EPs= 2 Cls=e0(wlcon) Sub=01 Prot=01 Driver=(none)
I:  If#= 2 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=ff Prot=ff Driver=(none)
I:  If#= 3 Alt= 0 #EPs= 0 Cls=fe(app. ) Sub=01 Prot=01 Driver=(none)

Signed-off-by: Nobuhiro Iwamatsu <iwamatsu@nigauri.org>
Acked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
13 years agoBluetooth: Don't clear the blacklist when closing the HCI device
Johan Hedberg [Thu, 9 Sep 2010 07:32:39 +0000 (10:32 +0300)]
Bluetooth: Don't clear the blacklist when closing the HCI device

Clearing the blacklist in hci_dev_do_close() would mean that user space
needs to do extra work to re-block devices after a DEVDOWN-DEVUP cycle.
This patch removes the clearing of the blacklist in this case and
thereby saves user space from the extra work.

Signed-off-by: Johan Hedberg <johan.hedberg@nokia.com>
Acked-by: Ville Tervo <ville.tervo@nokia.com>
Acked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
13 years agoBluetooth: remove extra newline from debug output
Andrei Emeltchenko [Wed, 8 Sep 2010 13:26:53 +0000 (16:26 +0300)]
Bluetooth: remove extra newline from debug output

Signed-off-by: Andrei Emeltchenko <andrei.emeltchenko@nokia.com>
Acked-by: Ville Tervo <ville.tervo@nokia.com>
Acked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
13 years agoBluetooth: Use a stream-oriented recvmsg with SOCK_STREAM L2CAP sockets.
Mat Martineau [Wed, 8 Sep 2010 17:05:29 +0000 (10:05 -0700)]
Bluetooth: Use a stream-oriented recvmsg with SOCK_STREAM L2CAP sockets.

L2CAP ERTM sockets can be opened with the SOCK_STREAM socket type,
which is a mandatory request for ERTM mode.

However, these sockets still have SOCK_SEQPACKET read semantics when
bt_sock_recvmsg() is used to pull data from the receive queue.  If the
application is only reading part of a frame, then the unread portion
of the frame is discarded.  If the application requests more bytes
than are in the current frame, only the current frame's data is
returned.

This patch utilizes common code derived from RFCOMM's recvmsg()
function to make L2CAP SOCK_STREAM reads behave like RFCOMM reads (and
other SOCK_STREAM sockets in general).  The application may read one
byte at a time from the input stream and not lose any data, and may
also read across L2CAP frame boundaries.

Signed-off-by: Mat Martineau <mathewm@codeaurora.org>
Acked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
13 years agoBluetooth: Use common SOCK_STREAM receive code in RFCOMM
Mat Martineau [Wed, 8 Sep 2010 17:05:28 +0000 (10:05 -0700)]
Bluetooth: Use common SOCK_STREAM receive code in RFCOMM

To reduce code duplication, have rfcomm_sock_recvmsg() call
bt_sock_stream_recvmsg().  The common bt_sock_stream_recvmsg()
code is nearly identical, with the RFCOMM-specific functionality
for deferred setup and connection unthrottling left in
rfcomm_sock_recvmsg().

Signed-off-by: Mat Martineau <mathewm@codeaurora.org>
Acked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
13 years agoBluetooth: Add common code for stream-oriented recvmsg()
Mat Martineau [Wed, 8 Sep 2010 17:05:27 +0000 (10:05 -0700)]
Bluetooth: Add common code for stream-oriented recvmsg()

This commit adds a bt_sock_stream_recvmsg() function for use by any
Bluetooth code that uses SOCK_STREAM sockets.  This code is copied
from rfcomm_sock_recvmsg() with minimal modifications to remove
RFCOMM-specific functionality and improve readability.

L2CAP (with the SOCK_STREAM socket type) and RFCOMM have common needs
when it comes to reading data.  Proper stream read semantics require
that applications can read from a stream one byte at a time and not
lose any data.  The RFCOMM code already operated on and pulled data
from the underlying L2CAP socket, so very few changes were required to
make the code more generic for use with non-RFCOMM data over L2CAP.

Applications that need more awareness of L2CAP frame boundaries are
still free to use SOCK_SEQPACKET sockets, and may verify that they
connection did not fall back to basic mode by calling getsockopt().

Signed-off-by: Mat Martineau <mathewm@codeaurora.org>
Acked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
13 years agoBluetooth: Validate PSM values in calls to connect() and bind()
Mat Martineau [Wed, 8 Sep 2010 17:05:26 +0000 (10:05 -0700)]
Bluetooth: Validate PSM values in calls to connect() and bind()

Valid L2CAP PSMs are odd numbers, and the least significant bit of the
most significant byte must be 0.

Signed-off-by: Mat Martineau <mathewm@codeaurora.org>
Acked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
13 years agoBluetooth: Fix RFCOMM RPN negotiation
Yuri Kululin [Fri, 23 Jul 2010 09:57:12 +0000 (13:57 +0400)]
Bluetooth: Fix RFCOMM RPN negotiation

According to the ETSI 3GPP TS 07.10 the default bit rate value for RFCOMM
is 9600 bit/s. Return this bit rate in case of RPN request and accept other
sane bit rates proposed by the sender in RPM command.

Signed-off-by: Yuri Kululin <ext-yuri.kululin@nokia.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
13 years agoBluetooth: Support SDIO devices that are AMP controllers
David Vrabel [Mon, 9 Aug 2010 21:42:21 +0000 (17:42 -0400)]
Bluetooth: Support SDIO devices that are AMP controllers

Signed-off-by: David Vrabel <david.vrabel@csr.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
13 years agoBluetooth: HCI devices are either BR/EDR or AMP radios
David Vrabel [Mon, 9 Aug 2010 21:38:10 +0000 (17:38 -0400)]
Bluetooth: HCI devices are either BR/EDR or AMP radios

HCI transport drivers may not know what type of radio an AMP device has
so only say whether they're BR/EDR or AMP devices.

Signed-off-by: David Vrabel <david.vrabel@csr.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
13 years agoiwl3945: fix queue allocation
Johannes Berg [Tue, 12 Oct 2010 09:26:13 +0000 (02:26 -0700)]
iwl3945: fix queue allocation

commit 6f98613258b966ffe0e6def18129b386514d10e0
Author: Jay Sternberg <jay.e.sternberg@intel.com>
Date:   Sat Sep 18 09:07:04 2010 -0700

    iwlagn: reduce redundant parameter definitions

broke 3945 because Jay accidentally removed the
num_of_queues parameter for 3945, so that we now
attempt to allocate a zero-sized queue array,
which leads to SLUB returning ZERO_SIZE_PTR (0x10)
which we then try to dereference thus crashing the
system. Restore the necessary num_of_queues param.

This fixes
http://bugzilla.intellinuxwireless.org/show_bug.cgi?id=2254

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
13 years agoiwlagn: prio_tbl need to download before calibration
Wey-Yi Guy [Mon, 11 Oct 2010 21:24:05 +0000 (14:24 -0700)]
iwlagn: prio_tbl need to download before calibration

For WiFi/BT combo devices, priority table always need to download
before perform any calibration operation.

Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
13 years agoiiwlagn: always download priority table
Wey-Yi Guy [Thu, 23 Sep 2010 16:56:51 +0000 (09:56 -0700)]
iiwlagn: always download priority table

For advance BT/WiFi co-exist, always download bt priority table
before sending bt_config command

Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
13 years agodccp: cosmetics - warning format
Gerrit Renker [Mon, 11 Oct 2010 18:44:42 +0000 (20:44 +0200)]
dccp: cosmetics - warning format

This  omits the redundant "DCCP:" in warning messages, since DCCP_WARN() already
echoes the function name, avoiding messages like

   kernel: [10988.766503] dccp_close: DCCP: ABORT -- 209 bytes unread

Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
13 years agodccp: schedule an Ack when receiving timestamps
Gerrit Renker [Mon, 11 Oct 2010 18:41:13 +0000 (20:41 +0200)]
dccp: schedule an Ack when receiving timestamps

This schedules an Ack when receiving a timestamp, exploiting the
existing inet_csk_schedule_ack() function, saving one case in the
`dccp_ack_pending()' function.

Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
13 years agodccp: generalise data-loss condition
Ivo Calado [Mon, 11 Oct 2010 18:40:04 +0000 (20:40 +0200)]
dccp: generalise data-loss condition

This patch generalises the task of determining data loss from RFC 4340, 7.7.1.

Let S_A, S_B be sequence numbers such that S_B is "after" S_A, and let
N_B be the NDP count of packet S_B. Then, using modulo-2^48 arithmetic,
 D = S_B - S_A - 1  is an upper bound of the number of lost data packets,
 D - N_B            is an approximation of the number of lost data packets
                    (there are cases where this is not exact).

The patch implements this as
 dccp_loss_count(S_A, S_B, N_B) := max(S_B - S_A - 1 - N_B, 0)

Signed-off-by: Ivo Calado <ivocalado@embedded.ufcg.edu.br>
Signed-off-by: Erivaldo Xavier <desadoc@gmail.com>
Signed-off-by: Leandro Sales <leandroal@gmail.com>
Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
13 years agodccp: remove unused argument in CCID tx function
Gerrit Renker [Mon, 11 Oct 2010 18:37:38 +0000 (20:37 +0200)]
dccp: remove unused argument in CCID tx function

This removes the argument `more' from ccid_hc_tx_packet_sent, since it was
nowhere used in the entire code.

(Btw, this argument was not even used in the original KAME code where the
 function initially came from; compare the variable moreToSend in the
 freebsd61-dccp-kame-28.08.2006.patch kept by Emmanuel Lochin.)

Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
13 years agodccp: merge now-reduced connect_init() function
Gerrit Renker [Mon, 11 Oct 2010 18:36:33 +0000 (20:36 +0200)]
dccp: merge now-reduced connect_init() function

After moving the assignment of GAR/ISS from dccp_connect_init() to
dccp_transmit_skb(), the former function becomes very small, so that
a merger with dccp_connect() suggests itself.

Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
13 years agodccp: fix the adjustments to AWL and SWL
Gerrit Renker [Mon, 11 Oct 2010 18:35:40 +0000 (20:35 +0200)]
dccp: fix the adjustments to AWL and SWL

This fixes a problem and a potential loophole with regard to seqno/ackno
validity: currently the initial adjustments to AWL/SWL are only performed
once at the begin of the connection, during the handshake.

Since the Sequence Window feature is always greater than Wmin=32 (7.5.2),
it is however necessary to perform these adjustments at least for the first
W/W' (variables as per 7.5.1) packets in the lifetime of a connection.

This requirement is complicated by the fact that W/W' can change at any time
during the lifetime of a connection.

Therefore it is better to perform that safety check each time SWL/AWL are
updated, as implemented by the patch.

A second problem solved by this patch is that the remote/local Sequence Window
feature values (which set the bounds for AWL/SWL/SWH) are undefined until the
feature negotiation has completed.

During the initial handshake we have more stringent sequence number protection;
the changes added by this patch effect that {A,S}W{L,H} are within the correct
bounds at the instant that feature negotiation completes (since the SeqWin
feature activation handlers call dccp_update_gsr/gss()).

Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
13 years agobnx2: Enable AER on PCIE devices only
Michael Chan [Mon, 11 Oct 2010 23:12:28 +0000 (16:12 -0700)]
bnx2: Enable AER on PCIE devices only

To prevent unnecessary error message.  pci_save_state() is also moved to
the end of ->probe() so that all PCI config, including AER state, will be
saved.

Update version to 2.0.18.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Reviewed-by: Benjamin Li <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agobnx2: Update firmware to 6.0.x.
Michael Chan [Mon, 11 Oct 2010 23:12:00 +0000 (16:12 -0700)]
bnx2: Update firmware to 6.0.x.

- Improved flow control and simplified interface
- Use hardware RSS indirection table instead of the slower firmware-
  based table
- Lower latency interrupt on 5709

Signed-off-by: Michael Chan <mchan@broadcom.com>
Reviewed-by: Benjamin Li <benli@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agoneigh: reorder struct neighbour fields
Eric Dumazet [Mon, 11 Oct 2010 12:20:54 +0000 (12:20 +0000)]
neigh: reorder struct neighbour fields

Le mardi 12 octobre 2010 à 00:02 +0200, Eric Dumazet a écrit :
> Here is the followup patch.
>
> Thanks !
>

Oops, this was an old version, the up2date ones also took care of "used"
field.

I guess its time for a sleep, sorry again.

[PATCH net-next V2] neigh: reorder struct neighbour fields

(refcnt) and (ha_lock, ha, used, dev, output, ops, primary_key) should
be placed on a separate cache lines.

refcnt can be often written, while other fields are mostly read.

This gave me good result on stress test :

before:

real    0m45.570s
user    0m15.525s
sys     9m56.669s

After:

real    0m41.841s
user    0m15.261s
sys     8m45.949s

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agonet dst: use a percpu_counter to track entries
Eric Dumazet [Fri, 8 Oct 2010 06:37:34 +0000 (06:37 +0000)]
net dst: use a percpu_counter to track entries

struct dst_ops tracks number of allocated dst in an atomic_t field,
subject to high cache line contention in stress workload.

Switch to a percpu_counter, to reduce number of time we need to dirty a
central location. Place it on a separate cache line to avoid dirtying
read only fields.

Stress test :

(Sending 160.000.000 UDP frames,
IP route cache disabled, dual E5540 @2.53GHz,
32bit kernel, FIB_TRIE, SLUB/NUMA)

Before:

real    0m51.179s
user    0m15.329s
sys     10m15.942s

After:

real 0m45.570s
user 0m15.525s
sys 9m56.669s

With a small reordering of struct neighbour fields, subject of a
following patch, (to separate refcnt from other read mostly fields)

real 0m41.841s
user 0m15.261s
sys 8m45.949s

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agoneigh: Protect neigh->ha[] with a seqlock
Eric Dumazet [Thu, 7 Oct 2010 10:44:07 +0000 (10:44 +0000)]
neigh: Protect neigh->ha[] with a seqlock

Add a seqlock in struct neighbour to protect neigh->ha[], and avoid
dirtying neighbour in stress situation (many different flows / dsts)

Dirtying takes place because of read_lock(&n->lock) and n->used writes.

Switching to a seqlock, and writing n->used only on jiffies changes
permits less dirtying.

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agoMerge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6
David S. Miller [Mon, 11 Oct 2010 19:30:34 +0000 (12:30 -0700)]
Merge branch 'master' of /linux/kernel/git/davem/net-2.6

Conflicts:
net/core/ethtool.c

13 years agonet: clear heap allocations for privileged ethtool actions
Kees Cook [Mon, 11 Oct 2010 19:23:25 +0000 (12:23 -0700)]
net: clear heap allocations for privileged ethtool actions

Several other ethtool functions leave heap uncleared (potentially) by
drivers. Some interfaces appear safe (eeprom, etc), in that the sizes
are well controlled. In some situations (e.g. unchecked error conditions),
the heap will remain unchanged in areas before copying back to userspace.
Note that these are less of an issue since these all require CAP_NET_ADMIN.

Cc: stable@kernel.org
Signed-off-by: Kees Cook <kees.cook@canonical.com>
Acked-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agort2x00: Fix URB error handling
Ivo van Doorn [Mon, 11 Oct 2010 13:39:48 +0000 (15:39 +0200)]
rt2x00: Fix URB error handling

kill_urb guarentees that when the function returns, the URB has
been fully killed. This means we don't need the extra sleeping
after the call to kill_urb.

kill_urb can however also guarentee the submit_urb to fail, as
a result, we must catch the return value from submit_urb an
correctly mark the entry as owned by the driver, and the
status as broken.

Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agort2x00: Don't perform watchdog checks on empty queue
Ivo van Doorn [Mon, 11 Oct 2010 13:39:04 +0000 (15:39 +0200)]
rt2x00: Don't perform watchdog checks on empty queue

The currently used watchdog functions cannot be applied
to empty queues.

Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Acked-by: Helmut Schaa <helmut.schaa@googlemail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agort2x00: Make queue_entry flags access atomic
Ivo van Doorn [Mon, 11 Oct 2010 13:38:45 +0000 (15:38 +0200)]
rt2x00: Make queue_entry flags access atomic

All access to the queue_entry->flags can be done concurrently,
so all flags must use the atomic operators. On most locations
this was already done, so just fix the last few non-atomic
versions.

Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Acked-by: Helmut Schaa <helmut.schaa@googlemail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agort2x00: Fix dead queue when skb allocation failed
Ivo van Doorn [Mon, 11 Oct 2010 13:38:26 +0000 (15:38 +0200)]
rt2x00: Fix dead queue when skb allocation failed

When the RX skb allocation failed, we should recycle
the previously allocated skbuffer. By calling return
we would kill the RX queue completely since the
entry would be invalidated.

Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Acked-by: Helmut Schaa <helmut.schaa@googlemail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agort2x00: Validate MCS on RX path
Ivo van Doorn [Mon, 11 Oct 2010 13:38:07 +0000 (15:38 +0200)]
rt2x00: Validate MCS on RX path

Similar to the PLCP signal and bitrates values,
we should validate the MCS value from the RX descriptor
before sending it to mac80211.

Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Acked-by: Helmut Schaa <helmut.schaa@googlemail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agort2x00: Move watchdog work to kernel work_queue
Ivo van Doorn [Mon, 11 Oct 2010 13:37:47 +0000 (15:37 +0200)]
rt2x00: Move watchdog work to kernel work_queue

The watchdog function must run on a work_queue
which is independent of any other work inside rt2x00.

The main reasons, being that a broken work on the mac80211
work_queue can otherwise prevent the watchdog to run (while
in fact the watchdog could fix the issue). And on the other
hand because the watchdog relies on the completion of the
completion handlers for RX/TX which for the USB case, occur
on the mac80211 workqueue.

This fixes some "Queue %d failed to flush" errors, which were
caused by the watchdog function waiting on the completion
handler which was scheduled to run right after the watchdog.

Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Acked-by: Helmut Schaa <helmut.schaa@googlemail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agort2x00: Simplify Queue function arguments
Ivo van Doorn [Mon, 11 Oct 2010 13:37:25 +0000 (15:37 +0200)]
rt2x00: Simplify Queue function arguments

A lot of functions accept a struct rt2x00_dev combined with
either a struct queue_entry or struct data_queue argument.
This can be simplified by only passing on the queue/entry
argument.

In cases where rt2x00_dev and a sk_buff are send together,
we can send the queue_entry instead.

rt2x00usb_alloc_urb and rt2x00usb_free_urb have a bit
of vague naming. Instead they allocate all the data which
belongs to a rt2x00 data queue entry.

Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Acked-by: Helmut Schaa <helmut.schaa@googlemail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agoWIRELESS: at76c50x, remove unneeded NULL check
Jiri Slaby [Mon, 11 Oct 2010 09:27:32 +0000 (11:27 +0200)]
WIRELESS: at76c50x, remove unneeded NULL check

Stanse found that urb cannot be NULL in at76_rx_tasklet because it is
dereferenced earlier, so remove the unneeded check.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agob43: N-PHY: prepare for rev3+ channel tables
Rafał Miłecki [Mon, 11 Oct 2010 01:19:22 +0000 (03:19 +0200)]
b43: N-PHY: prepare for rev3+ channel tables

Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agob43: N-PHY: fix logic in band switching
Rafał Miłecki [Mon, 11 Oct 2010 01:11:03 +0000 (03:11 +0200)]
b43: N-PHY: fix logic in band switching

Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agob43: N-PHY: store info about current channel's type
Rafał Miłecki [Mon, 11 Oct 2010 01:11:02 +0000 (03:11 +0200)]
b43: N-PHY: store info about current channel's type

Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agob43: N-PHY: grab more info about new channel
Rafał Miłecki [Mon, 11 Oct 2010 01:11:01 +0000 (03:11 +0200)]
b43: N-PHY: grab more info about new channel

Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agoath9k: built-in rate control A-MPDU fix
Björn Smedman [Sun, 10 Oct 2010 20:51:54 +0000 (22:51 +0200)]
ath9k: built-in rate control A-MPDU fix

This patch attempts to ensure that ath9k's built-in rate control algorithm
does not rely on the value of the ampdu_len and ampdu_ack_len tx status
fields unless the IEEE80211_TX_STAT_AMPDU flag is set.

This patch has not been tested.

Cc: <stable@kernel.org>
Signed-off-by: Björn Smedman <bjorn.smedman@venatech.se>
Acked-by: Felix Fietkau <nbd@openwrt.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agoath9k: A-MPDU rate control info fix
Björn Smedman [Sun, 10 Oct 2010 20:44:39 +0000 (22:44 +0200)]
ath9k: A-MPDU rate control info fix

This patch fixes the following problems with the rate control feedback
generated by ath9k for A-MPDU frames:

1. Rate control feedback is carried on the first frame of an aggregate
that is either ACKed, or has execeeded the software retry count and is
considered failed. However, ath9k would incorrectly assume the aggregate
had the length 1 if one of these conditions did not apply to the first
frame of the aggregate, but instead a later frame. This fix therefor
copies the bf_nframes field of the buffer in the same manner as the rates
field of the tx status.

2. Sometimes the ampdu_len and ampdu_ack_len fields of the tx status was
left uninitialized eventhough the IEEE80211_TX_STAT_AMPDU flag was set.
This is now avoid by setting flag and fields in the same place.

3. Even if a frame has been selected for aggregation by mac80211 and
marked with the IEEE80211_TX_CTL_AMPDU flag it can sometimes happen that
ath9k transmits the frame without aggregation. In these cases the
ampdu_ack_len field could be incorrectly computed because the nbad
parameter to ath_tx_rc_status was incorrect.

Cc: <stable@kernel.org>
Signed-off-by: Björn Smedman <bjorn.smedman@venatech.se>
Acked-by: Felix Fietkau <nbd@openwrt.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agomac80211: minstrel_ht A-MPDU fix
Björn Smedman [Sun, 10 Oct 2010 20:14:25 +0000 (22:14 +0200)]
mac80211: minstrel_ht A-MPDU fix

This patch fixes two problems with the minstrel_ht rate control
algorithms handling of A-MPDU frames:

1. The ampdu_len field of the tx status is not always initialized for
non-HT frames (and it would probably be unreasonable to require all
drivers to do so). This could cause rate control statistics to be
corrupted. We now trust the ampdu_len and ampdu_ack_len fields only when
the frame is marked with the IEEE80211_TX_STAT_AMPDU flag.

2. Successful transmission attempts where only recognized when the A-MPDU
subframe carrying the rate control status information was marked with the
IEEE80211_TX_STAT_ACK flag. If this information happed to be carried on a
frame that failed to be ACKed then the other subframes (which may have
succeeded) where not correctly registered. We now update rate control
statistics regardless of whether the subframe carrying the information was
ACKed or not.

Cc: <stable@kernel.org>
Signed-off-by: Björn Smedman <bjorn.smedman@venatech.se>
Acked-by: Felix Fietkau <nbd@openwrt.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agomac80211: don't kmalloc 16 bytes
Johannes Berg [Sun, 10 Oct 2010 16:52:10 +0000 (18:52 +0200)]
mac80211: don't kmalloc 16 bytes

Since this small buffer isn't used for DMA,
we can simply allocate it on the stack, it
just needs to be 16 bytes of which only 8
will be used for WEP40 keys.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agowl1251: update the new location to the maintainers file
Kalle Valo [Sun, 10 Oct 2010 08:28:33 +0000 (11:28 +0300)]
wl1251: update the new location to the maintainers file

wl1251 is grown up now and can have its own room^H^H^H^Hdirectory.

Signed-off-by: Kalle Valo <kvalo@adurom.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agowl1251: move to it's own directory
Kalle Valo [Sun, 10 Oct 2010 08:28:32 +0000 (11:28 +0300)]
wl1251: move to it's own directory

wl1271 driver is under heavy development but on the other hand the older
wl1251 driver is currently considered more as a legacy driver. To make it
easier to develop wl1271 features move wl1251 to it's own directory,
drivers/net/wireless/wl1251.

There are no functional changes, only moving of files. One regression
is that Kconfig won't be updated automatically and user needs to enable
wl1251 manually with an older config file.

Signed-off-by: Kalle Valo <kvalo@adurom.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agowl1251: create a copy of wl12xx_80211.h for wl1251
Kalle Valo [Sun, 10 Oct 2010 08:28:31 +0000 (11:28 +0300)]
wl1251: create a copy of wl12xx_80211.h for wl1251

In preparation of moving wl1251 out from drivers/net/wireless/wl12xx
create a separate copy of wl12xx_80211.h.

This file should not even exist, we should instead use generic ieee80211
definitions. This will be fixed in the future so that the file can
be removed.

Signed-off-by: Kalle Valo <kvalo@adurom.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agocarl9170: A-MPDU frame type filter
Christian Lamparter [Sat, 9 Oct 2010 23:15:07 +0000 (01:15 +0200)]
carl9170: A-MPDU frame type filter

In the past, carl9170 has been plagued by mysterious
ghosts.

e.g.:
 wlan4: deauthenticated from 02:04:d8:3c:ac:c1 (Reason: 0)

Apparently, the AP sent us a bogus deauthentication
notification. But upon closer inspection the
"management frame" turned out to be a corrupted
scrap of an unsuccessful A-MPDU.

Signed-off-by: Christian Lamparter <chunkeey@googlemail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agocarl9170: remove stale rx error path
Christian Lamparter [Sat, 9 Oct 2010 20:23:37 +0000 (22:23 +0200)]
carl9170: remove stale rx error path

The total/fatal error bit was erroneously prefixed
with AR9170_RX_ERROR instead of AR9170_RX_STATUS.
Luckily, the hardware specification confirmed that
the 0x80 flag will never be set for mac->error.
So, it was always just a dead branch.

This patch also imports the latest version of
shared wlan.h header from the firmware git.

Signed-off-by: Christian Lamparter <chunkeey@googlemail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agocarl9170: common error path for bad frames
Christian Lamparter [Sat, 9 Oct 2010 19:37:11 +0000 (21:37 +0200)]
carl9170: common error path for bad frames

This patch replaces several identical frame drop
paths with a single shared rx frame error handler.

Signed-off-by: Christian Lamparter <chunkeey@googlemail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agort2x00: Fix tx status handling in rt2800pci
Helmut Schaa [Sat, 9 Oct 2010 11:35:48 +0000 (13:35 +0200)]
rt2x00: Fix tx status handling in rt2800pci

The patches "rt2x00: Improve TX status entry validation" and "rt2x00: rework tx
status handling in rt2800pci" together were causing problems with tx status
processing in rt2800pci:

phy1 -> rt2800pci_txdone: Warning - Got TX status for an empty queue 3, dropping
phy1 -> rt2800pci_txdone: Warning - Got TX status for an unavailable queue 7, dropping

Fix this by using the correct field definition for getting the QID out of the
tx status report.

Reported-by: Luis Correia <luis.f.correia@gmail.com>
Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
Tested-by: Luis Correia <luis.f.correia@gmail.com>
Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agort2x00: Remove superfluous initialization of qidx
Helmut Schaa [Sat, 9 Oct 2010 11:35:13 +0000 (13:35 +0200)]
rt2x00: Remove superfluous initialization of qidx

There is no need to initialize qidx to zero as it will ever be
overwritten by the correct value.

Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agort2x00: Use unlikely for error case in rt2x00queue_write_tx_frame
Helmut Schaa [Sat, 9 Oct 2010 11:34:43 +0000 (13:34 +0200)]
rt2x00: Use unlikely for error case in rt2x00queue_write_tx_frame

This is an error condition that is not supposed to happen. Hence, it is
safe to add unlikely to this check.

Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agort2x00: Use proper type for rxwi_w2 in rt2800_agc_to_rssi
Helmut Schaa [Sat, 9 Oct 2010 11:34:11 +0000 (13:34 +0200)]
rt2x00: Use proper type for rxwi_w2 in rt2800_agc_to_rssi

Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agort2x00: Optimize unmapping of skbs
Helmut Schaa [Sat, 9 Oct 2010 11:33:43 +0000 (13:33 +0200)]
rt2x00: Optimize unmapping of skbs

Since no skb will be mapped for RX and TX at the same time we can
simply shortcut the check for SKBDESC_DMA_MAPPED_TX.

Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agort2x00: Shortcut link state updates when not operating as STA
Helmut Schaa [Sat, 9 Oct 2010 11:33:16 +0000 (13:33 +0200)]
rt2x00: Shortcut link state updates when not operating as STA

Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
13 years agomac80211: Update mesh constants to approved IEEE ANA values
Steve deRosier [Sun, 10 Oct 2010 00:23:28 +0000 (17:23 -0700)]
mac80211: Update mesh constants to approved IEEE ANA values

This patch updates IEEE802.11 mesh constants to be consistent with newly
approved values. It modifies some values, as well as adds many new constants
in preparation for updating mesh code to the current 802.11s drafts. ANA
numbers were taken from:
https://mentor.ieee.org/802.11/dcn/09/11-09-0031-12-0000-ana-database-assigned-number-authority.xls

A few notes are in order:
1. This will break backwards compatibility with existing Linux kernels as
over-the-air constants have changed.
2. Some old and obsolete constants have been retained for now as the mesh code
itself hasn't been updated yet to the new 802.11s draft. This was desired to
keep the existing mesh scheme working until it can be updated. Adding the
approved values is the first step in updating the mesh code.
3. Obsolete constants have been clearly marked.
4. All ANA approved 802.11s constants have been added.

Signed-off-by: Steve deRosier <steve@cozybit.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>