firefly-linux-kernel-4.4.55.git
15 years agomac80211: FIF_PSPOLL filter flag
Igor Perminov [Tue, 4 Aug 2009 12:48:51 +0000 (16:48 +0400)]
mac80211: FIF_PSPOLL filter flag

When an interface is configured in the AP mode, the mac80211
implementation doesn't inform the driver to receive PS Poll frames.
It leads to inability to communicate with power-saving stations
reliably.
The FIF_CONTROL flag isn't passed by mac80211 to
ieee80211_ops.configure_filter when an interface is in the AP mode.
And it's ok, because we don't want to receive ACK frames and other
control ones, but only PS Poll ones.

This patch introduces the FIF_PSPOLL filter flag in addition to
FIF_CONTROL, which means for the driver "pass PS Poll frames".

This flag is passed to the driver:
A) When an interface is configured in the AP mode.
B) In all cases, when the FIF_CONTROL flag was passed earlier (in
addition to it).

Signed-off-by: Igor Perminov <igor.perminov@inbox.ru>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: cancel xmit poll work at stop() callback
Luis R. Rodriguez [Sat, 25 Jul 2009 00:47:33 +0000 (20:47 -0400)]
ath9k: cancel xmit poll work at stop() callback

We forgot to cancel this work at the stop() callback.

------------[ cut here ]------------
WARNING: at net/mac80211/util.c:511 ieee80211_queue_delayed_work+0x3a/0x40 [mac80211]()
Hardware name: 6460DWU
queueing ieee80211 work while going to suspend
Modules linked in: <-- snip -->
Pid: 5124, comm: phy0 Tainted: G        W  2.6.31-rc3-wl #4
Call Trace:
 [<ffffffffa03adcda>] ? ieee80211_queue_delayed_work+0x3a/0x40 [mac80211]
 [<ffffffff810552b8>] warn_slowpath_common+0x78/0xd0
 [<ffffffff81055394>] warn_slowpath_fmt+0x64/0x70
 [<ffffffff814ed2c9>] ? thread_return+0x3e/0x635
 [<ffffffffa03adcda>] ieee80211_queue_delayed_work+0x3a/0x40 [mac80211]
 [<ffffffffa0297690>] ath_tx_complete_poll_work+0xc0/0x100 [ath9k]
 [<ffffffffa02975d0>] ? ath_tx_complete_poll_work+0x0/0x100 [ath9k]
 [<ffffffff81069b68>] worker_thread+0x178/0x260
 [<ffffffff8106ecc0>] ? autoremove_wake_function+0x0/0x40
 [<ffffffff810699f0>] ? worker_thread+0x0/0x260
 [<ffffffff8106e89e>] kthread+0x9e/0xb0
 [<ffffffff8101302a>] child_rip+0xa/0x20
 [<ffffffff8106e800>] ? kthread+0x0/0xb0
 [<ffffffff81013020>] ? child_rip+0x0/0x20

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: add initial hardware support for ar9271
Luis R. Rodriguez [Tue, 4 Aug 2009 03:14:12 +0000 (23:14 -0400)]
ath9k: add initial hardware support for ar9271

We will finalize this after some driver core changes, for now
we leave this unsupported.

Cc: Stephen Chen <stephen.chen@atheros.com>
Cc: Zhifeng Cai <zhifeng.cai@atheros.com>
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: add initvals and registry definitions for AR9271
Luis R. Rodriguez [Tue, 4 Aug 2009 03:14:11 +0000 (23:14 -0400)]
ath9k: add initvals and registry definitions for AR9271

Cc: Stephen Chen <stephen.chen@atheros.com>
Cc: Zhifeng Cai <zhifeng.cai@atheros.com>
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: add ar9271 revision and subrevision ID helpers
Luis R. Rodriguez [Tue, 4 Aug 2009 03:14:10 +0000 (23:14 -0400)]
ath9k: add ar9271 revision and subrevision ID helpers

These will be used later to add support for ar9271.

Cc: Stephen Chen <stephen.chen@atheros.com>
Cc: Zhifeng Cai <zhifeng.cai@atheros.com>
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: Remove _t postfix for ar9287_eeprom structure
Luis R. Rodriguez [Mon, 3 Aug 2009 21:31:25 +0000 (17:31 -0400)]
ath9k: Remove _t postfix for ar9287_eeprom structure

We don't use typdefs on ath9k, remove that _t.

Cc: Vivek Natarajan <vnatarajan@atheros.com>
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: remove spurious check for channel on keycache reset
Luis R. Rodriguez [Mon, 3 Aug 2009 19:24:56 +0000 (12:24 -0700)]
ath9k: remove spurious check for channel on keycache reset

ath9k_hw_keyreset() has a spurious check for ah->curchan..
remove it.

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: remove dangling error check on keycache reset on hw init
Luis R. Rodriguez [Mon, 3 Aug 2009 19:24:55 +0000 (12:24 -0700)]
ath9k: remove dangling error check on keycache reset on hw init

The keycache reset will not fail as right above we ensure
to set the sc->keymax to be <= ah->caps.keycache_size. Just remove
this dangling check.

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: call ath9k_hw_detach() once upon hw init failure
Luis R. Rodriguez [Mon, 3 Aug 2009 19:24:54 +0000 (12:24 -0700)]
ath9k: call ath9k_hw_detach() once upon hw init failure

If hw initialization fails (ath9k_hw_init()) on ath_init_softc()
we bail out and call ath9k_hw_detach(). The call ath9k_hw_detach()
is conditional though as ath9k_hw_init() could itself have called
ath9k_hw_detach(). Just describing this is itself a brain twister.
Avoid this nonsense by removing ath9k_hw_detach() from ath9k_hw_init().

Upon hw initialization failure we expect the callers to take care of
the cleanup.

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: set sc->sc_ah to NULL after freeing it
Luis R. Rodriguez [Mon, 3 Aug 2009 19:24:53 +0000 (12:24 -0700)]
ath9k: set sc->sc_ah to NULL after freeing it

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: set ah to null after freeing
Luis R. Rodriguez [Mon, 3 Aug 2009 19:24:52 +0000 (12:24 -0700)]
ath9k: set ah to null after freeing

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: rename ath9k_hw_ani_detach() to ath9k_hw_ani_disable()
Luis R. Rodriguez [Mon, 3 Aug 2009 19:24:51 +0000 (12:24 -0700)]
ath9k: rename ath9k_hw_ani_detach() to ath9k_hw_ani_disable()

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: rename ath9k_hw_rfdetach() to ath9k_hw_rf_free()
Luis R. Rodriguez [Mon, 3 Aug 2009 19:24:50 +0000 (12:24 -0700)]
ath9k: rename ath9k_hw_rfdetach() to ath9k_hw_rf_free()

This makes it clear what this does.

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: use helper macro to kfree and nullify on ath9k_hw_rfdetach()
Luis R. Rodriguez [Mon, 3 Aug 2009 19:24:49 +0000 (12:24 -0700)]
ath9k: use helper macro to kfree and nullify on ath9k_hw_rfdetach()

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: remove !NULL check before kfree()
Luis R. Rodriguez [Mon, 3 Aug 2009 19:24:48 +0000 (12:24 -0700)]
ath9k: remove !NULL check before kfree()

kfree(NULL) works so remove all those branches which check
for it before kfree()'ing on ath9k_hw_rfdetach().

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: distinguish between device initialization and ath_softc init
Luis R. Rodriguez [Mon, 3 Aug 2009 19:24:47 +0000 (12:24 -0700)]
ath9k: distinguish between device initialization and ath_softc init

We re-label the device driver initialization routines from the
ath_softc, the "Software Carrier" fillers. This should make it
clearer what each of these do.

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: describe hw initialization better
Luis R. Rodriguez [Mon, 3 Aug 2009 19:24:46 +0000 (12:24 -0700)]
ath9k: describe hw initialization better

During initialization ath9k tends to use "attach" to when we
initialize hardware due to the fact we used to attach a "HAL".
The notion of a HAL is long gone, so lets just be clear on what
we are doing.

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: break up hw initialization into a few more helpers
Luis R. Rodriguez [Mon, 3 Aug 2009 19:24:45 +0000 (12:24 -0700)]
ath9k: break up hw initialization into a few more helpers

This makes reading the hardware initialization process
easier to understand. The new helpers added are:

ath9k_hw_init_cal_settings()
ath9k_hw_init_mode_regs()
ath9k_hw_init_mode_gain_regs()
ath9k_hw_init_11a_eeprom_fix()

This patch has no functional changes.

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: remove debug message for no memoery on ath_init()
Luis R. Rodriguez [Mon, 3 Aug 2009 19:24:44 +0000 (12:24 -0700)]
ath9k: remove debug message for no memoery on ath_init()

We're now propagating the -ENOMEM error so there is no need to
keep a debug message there now.

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: rename ath9k_hw_set_defaults() to ath9k_hw_init_config()
Luis R. Rodriguez [Mon, 3 Aug 2009 19:24:43 +0000 (12:24 -0700)]
ath9k: rename ath9k_hw_set_defaults() to ath9k_hw_init_config()

This reflects better what we are actually doing there.

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: rename ath9k_hw_newstate() to ath9k_hw_init_defaults()
Luis R. Rodriguez [Mon, 3 Aug 2009 19:24:42 +0000 (12:24 -0700)]
ath9k: rename ath9k_hw_newstate() to ath9k_hw_init_defaults()

This reflects better what we are actually doing there.

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: move hw macrevision checker to helper
Luis R. Rodriguez [Mon, 3 Aug 2009 19:24:41 +0000 (12:24 -0700)]
ath9k: move hw macrevision checker to helper

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: pass only one argument to hw attach
Luis R. Rodriguez [Mon, 3 Aug 2009 19:24:40 +0000 (12:24 -0700)]
ath9k: pass only one argument to hw attach

The softc is cached and set within the ath_hw struct.

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: call hw initializer directly
Luis R. Rodriguez [Mon, 3 Aug 2009 19:24:39 +0000 (12:24 -0700)]
ath9k: call hw initializer directly

ath9k_hw_attach() was going first through some device id verifier,
and then calling some other helper which was doing the real hardware
initialization. Lets just do the devid checks within the real worker
by calling a helper ath9k_hw_devid_supported().

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: move cache setting of softc ah prior to attach
Luis R. Rodriguez [Mon, 3 Aug 2009 19:24:38 +0000 (12:24 -0700)]
ath9k: move cache setting of softc ah prior to attach

We do this in case attach and friends try to get back to
ah from the softc somehow.

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: move devid cache setting to ath_init()
Luis R. Rodriguez [Mon, 3 Aug 2009 19:24:37 +0000 (12:24 -0700)]
ath9k: move devid cache setting to ath_init()

This lets us trim one argument off of hw initializer routines.

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: move memory allocation of ath_hw to ath_init()
Luis R. Rodriguez [Mon, 3 Aug 2009 19:24:36 +0000 (12:24 -0700)]
ath9k: move memory allocation of ath_hw to ath_init()

This lets us simplify attach code and arguments passed.

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: propagate hw initialization errors
Luis R. Rodriguez [Mon, 3 Aug 2009 19:24:35 +0000 (12:24 -0700)]
ath9k: propagate hw initialization errors

We were never propagating hw initialization errors, lets
do that now and also use -EOPNOTSUPP when device revision is
not supported yet.

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: use a switch for revising supported hw mac revisions
Luis R. Rodriguez [Mon, 3 Aug 2009 19:24:34 +0000 (12:24 -0700)]
ath9k: use a switch for revising supported hw mac revisions

This makes adding new hw revisions a one line change here.

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: remove usage of AR_SREV_*() wrapper to detect supported hw
Luis R. Rodriguez [Mon, 3 Aug 2009 19:24:33 +0000 (12:24 -0700)]
ath9k: remove usage of AR_SREV_*() wrapper to detect supported hw

We will clean this up next to just use a switch.

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agocfg80211: clear SSID on disconnect/no connection
Johannes Berg [Tue, 4 Aug 2009 07:32:23 +0000 (09:32 +0200)]
cfg80211: clear SSID on disconnect/no connection

The SME state machine in cfg80211 uses the SSID stored
in struct wireless_dev internally, but fails to clear
it in multiple places (when giving up on a connection
attempt and when disconnecting). This doesn't matter to
the SME state machine, but does matter for IBSS. Thus,
in those cases, clear the SSID to avoid messing up the
IBSS state machine.

Reported-by: Joerg Albert <jal2@gmx.de>
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoiwmc3200wifi: avoid setting default key for 802.1X and RSNA
Zhu Yi [Mon, 3 Aug 2009 06:37:03 +0000 (14:37 +0800)]
iwmc3200wifi: avoid setting default key for 802.1X and RSNA

UMAC only allows us to set default key for WEP and auth type is
not 802.1X or RSNA. This patch fixes iwmc3200wifi for 802.1X with
WEP104.

Signed-off-by: Zhu Yi <yi.zhu@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoiwmc3200wifi: set WEP key static flag correctly
Zhu Yi [Mon, 3 Aug 2009 06:37:02 +0000 (14:37 +0800)]
iwmc3200wifi: set WEP key static flag correctly

We should only set the static_key flag for open and legacy
authentication types. It should not be set for 802.1X and TKIP.

Signed-off-by: Zhu Yi <yi.zhu@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoiwmc3200wifi: fix set_wpa_version and set_auth_type order
Zhu Yi [Mon, 3 Aug 2009 06:37:01 +0000 (14:37 +0800)]
iwmc3200wifi: fix set_wpa_version and set_auth_type order

iwm->umac_profile->sec.flags is set by iwm_set_wpa_version and
checked by iwm_set_auth_type. The patch changes the order to
make the flag used correctly.

Signed-off-by: Zhu Yi <yi.zhu@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agob43: implement baseband init for LP-PHY <= rev1
Gábor Stefanik [Sun, 2 Aug 2009 23:28:12 +0000 (01:28 +0200)]
b43: implement baseband init for LP-PHY <= rev1

Implement baseband init for rev.0 and rev.1 LP PHYs. Convert boardflags_hi values to defines.
Implement b43_phy_copy for easier copying between registers, as needed by LP-PHY init.

Signed-off-by: Gábor Stefanik<netrolller.3d@gmail.com>
Cc: Michael Buesch<mb@bu3sch.de>
Cc: Larry Finger<larry.finger@lwfinger.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agort2x00: cancel all work on disconnect
Pavel Roskin [Sun, 2 Aug 2009 18:30:15 +0000 (14:30 -0400)]
rt2x00: cancel all work on disconnect

Signed-off-by: Pavel Roskin <proski@gnu.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agort61pci: fix module reloading
Pavel Roskin [Sun, 2 Aug 2009 18:30:02 +0000 (14:30 -0400)]
rt61pci: fix module reloading

Unloading rt61pci can leave the device in such state that reloading
rt61pci would fail to reinitialize it.  Bogus data would be read from
the EEPROM and the RF version won't be recognized.

It appears that unloading rt61pci with power saving enabled would have
such effect.  To initialize the device properly, SOFT_RESET_CSR should
be set to the same value as rt61pci_config_ps() uses to wake up the
device.

Signed-off-by: Pavel Roskin <proski@gnu.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agob43legacy: Work around mac80211 race condition
Larry Finger [Sun, 2 Aug 2009 03:32:48 +0000 (22:32 -0500)]
b43legacy: Work around mac80211 race condition

As shown in http://thread.gmane.org/gmane.linux.kernel.wireless.general/36497,
mac80211 has a bug that allows a call to the TX routine after the queues have
been stopped. This situation will only occur under extreme stress. Although
b43legacy does not crash when this condition occurs, it does generate a WARN_ON
and also logs a queue overrun message. This patch recognizes b43legacy is not
at fault and logs a message only when the most verbose debugging mode is
enabled. In the unlikely event that the queue is not stopped when the DMA
queue becomes full, then a warning is issued.

This patch is based on the one used by b43.

Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agorfkill: add the GPS radio type
Tomas Winkler [Sat, 1 Aug 2009 23:36:49 +0000 (02:36 +0300)]
rfkill: add the GPS radio type

Althoug GPS is a technology w/o transmitting radio
and thus not a primary candidate for rfkill switch,
rfkill gives unified interface point for devices with
wireless technology.

The input key is not supplied as it is too be deprecated.

Cc: johannes@sipsolutions.net
Signed-off-by: Tomas Winkler <tomas.winkler@intel.com>
Acked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agodrivers/net/wireless: Use DIV_ROUND_CLOSEST
Julia Lawall [Sat, 1 Aug 2009 19:50:31 +0000 (21:50 +0200)]
drivers/net/wireless: Use DIV_ROUND_CLOSEST

The kernel.h macro DIV_ROUND_CLOSEST performs the computation (x + d/2)/d
but is perhaps more readable.

The semantic patch that makes this change is as follows:
(http://www.emn.fr/x-info/coccinelle/)

// <smpl>
@haskernel@
@@

#include <linux/kernel.h>

@depends on haskernel@
expression x,__divisor;
@@

- (((x) + ((__divisor) / 2)) / (__divisor))
+ DIV_ROUND_CLOSEST(x,__divisor)
// </smpl>

Signed-off-by: Julia Lawall <julia@diku.dk>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agolibertas: Fix WEP association failure with open source wpa_supplicant 0.5.10
Maithili Hinge [Sat, 1 Aug 2009 03:02:19 +0000 (20:02 -0700)]
libertas: Fix WEP association failure with open source wpa_supplicant 0.5.10

Add code to handle IW_AUTH_PRIVACY_INVOKED and IW_AUTH_RX_UNENCRYPTED_EAPOL
cases in lbs_set_auth() function in libertas code.

Signed-off-by: Maithili Hinge <maithili@marvell.com>
Signed-off-by: Bing Zhao <bzhao@marvell.com>
Acked-by: Dan Williams <dcbw@redhat.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoiwlwifi: remove duplicated version info from sysfs
Jay Sternberg [Fri, 31 Jul 2009 21:28:09 +0000 (14:28 -0700)]
iwlwifi: remove duplicated version info from sysfs

version info in sysfs had been determined to be unnecessary as it
is already provided in syslog info.  nvm version is added to syslog
version info as a debug level message to provide all info that was
in the version sysfs data.

Signed-off-by: Jay Sternberg <jay.e.sternberg@intel.com>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoiwlwifi: remove deprecated 6000 series adapters
Wey-Yi Guy [Fri, 31 Jul 2009 21:28:08 +0000 (14:28 -0700)]
iwlwifi: remove deprecated 6000 series adapters

Remove the support for deprecated devices. These devices are
engineering samples and no longer supported by the uCode.

Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoiwlwifi: Distinguish power amplifier for 6000 series
Wey-Yi Guy [Fri, 31 Jul 2009 21:28:07 +0000 (14:28 -0700)]
iwlwifi: Distinguish power amplifier for 6000 series

For 6x00 2x2 NIC, two types of Power Amplifier are available.
In order for uCode to apply correct tx power,
driver needs to program the CSR_GP_DRIVER_REG register and
let uCode know the type of PA.
If driver do not program CSR_GP_DRIVER_REG register (default to 0),
then it is uCode's decision for tx power

2x2 Hybrid card: use both internal and external PA
2x2 IPA(Internal Power Amplifier) card: internal PA only

Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoiwlwifi: clear iwl_cmd_meta structure before use
Daniel C Halperin [Fri, 31 Jul 2009 21:28:06 +0000 (14:28 -0700)]
iwlwifi: clear iwl_cmd_meta structure before use

Resolve an issue in which out-dated fields in iwl_cmd_meta
could be used for later hardware commands.

Signed-off-by: Daniel C Halperin <daniel.c.halperin@intel.com>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agob43: remove wrong probe_resp_plcp write
gregor kowski [Fri, 31 Jul 2009 20:35:49 +0000 (22:35 +0200)]
b43: remove wrong probe_resp_plcp write

The tkip hw support uncovered a bug in b43_write_probe_resp_template : it is
writing at the wrong shm offset, it is in the B43_SHM_SH_TKIPTSCTTAK zone. This
patch comments these writes.

Signed-off-by: Gregor Kowski <gregor.kowski@gmail.com>
Signed-off-by: Michael Buesch <mb@bu3sch.de>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agob43: Fix unaligned 32bit SHM-shared access
Michael Buesch [Fri, 31 Jul 2009 18:51:41 +0000 (20:51 +0200)]
b43: Fix unaligned 32bit SHM-shared access

This fixes unaligned 32bit SHM-shared read/write access.
The low and high 16 bits were swapped.
It also adds a testcase for this to the chipaccess validation.

(Thanks to Albert Herranz for tracking down this bug.)

Signed-off-by: Michael Buesch <mb@bu3sch.de>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agolibertas: check valid bits in SPI bus mode reg
Andrey Yurovsky [Fri, 31 Jul 2009 18:35:19 +0000 (11:35 -0700)]
libertas: check valid bits in SPI bus mode reg

The SPI driver writes to the bus mode register and performs a sanity
check by reading back what we wrote, however only the lower four bits of
that register are defined.  In some cases, the device side seems to set
the higher bits, causing us to fail the sanity check unnecessarily.
Check only the lower four bits instead.

Thanks to John Goyette from Schick Technologies for pointing out the
problem.

Signed-off-by: Andrey Yurovsky <andrey@cozybit.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agomac80211: Increase timeouts for station polling
Maxim Levitsky [Fri, 31 Jul 2009 15:54:23 +0000 (18:54 +0300)]
mac80211: Increase timeouts for station polling

Do a probe request every 30 seconds, and wait for probe response,
half a second This should lower the traffic that card sends, thus save
power Wainting longer for response makes probe more robust against
'slow' access points

Signed-off-by: Maxim Levitsky <maximlevitsky@gmail.com>
Acked-by: Johannes Berg <johannes@sipsolutions.net>
Tested-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agomac80211: Retry probe request few times
Maxim Levitsky [Fri, 31 Jul 2009 15:54:12 +0000 (18:54 +0300)]
mac80211: Retry probe request few times

Retry 5 times  (chosen arbitary ), before assuming
that station is out of range.

Fixes frequent disassociations while connected to weak,
and sometimes even strong access points.

Signed-off-by: Maxim Levitky <maximlevitsky@gmail.com>
Acked-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agocfg80211: lower dynamic PS timeout to 100ms
Johannes Berg [Fri, 31 Jul 2009 09:18:13 +0000 (11:18 +0200)]
cfg80211: lower dynamic PS timeout to 100ms

The default of 500ms is pretty high, and leads
to the device being awake at least 50% of the
time under such light traffic conditions as a
simple 1 second interval ping. Reduce to just
100ms -- it should have a similar effect while
providing a better sleep time.

Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Reviewed-by: Kalle Valo <kalle.valo@iki.fi>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoiwlwifi: remove usage of orig_flags
Luis R. Rodriguez [Fri, 31 Jul 2009 01:37:27 +0000 (21:37 -0400)]
iwlwifi: remove usage of orig_flags

This is a private flag, internal to cfg80211. cfg80211
will set orig_* stuff internally upon wiphy registration,
drivers do not need to muck with it.

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Acked-by: Reinette Chatre <reinette.chatre@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agocfg80211: enable country IE support to all cfg80211 drivers
Luis R. Rodriguez [Fri, 31 Jul 2009 00:38:09 +0000 (17:38 -0700)]
cfg80211: enable country IE support to all cfg80211 drivers

Since the bss is always set now once we are connected, if the
bss has its own information element we refer to it and pass that
instead of relying on mac80211's parsing.

Now all cfg80211 drivers get country IE support, automatically and
we reduce the call overhead that we had on mac80211 which called this
upon every beacon and instead now call this only upon a successfull
connection by a STA on cfg80211.

Acked-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agocfg80211: decouple regulatory variables from cfg80211_mutex
Luis R. Rodriguez [Fri, 31 Jul 2009 00:38:08 +0000 (17:38 -0700)]
cfg80211: decouple regulatory variables from cfg80211_mutex

We change regulatory code to be protected by its own regulatory
mutex and alleviate cfg80211_mutex to only be used to protect
cfg80211_rdev_list, the registered device list.

By doing this we will be able to work on regulatory core components
without having to have hog up the cfg80211_mutex. An example here is
we no longer need to use the cfg80211_mutex during driver specific
wiphy_apply_custom_regulatory(). We also no longer need it for the
the country IE regulatory hint; by doing so we end up curing this
new lockdep warning:

=======================================================
[ INFO: possible circular locking dependency detected ]
2.6.31-rc4-wl #12
-------------------------------------------------------
phy1/1709 is trying to acquire lock:
 (cfg80211_mutex){+.+.+.}, at: [<ffffffffa00af852>] regulatory_hint_11d+0x32/0x3f0 [cfg80211]

but task is already holding lock:
 (&ifmgd->mtx){+.+.+.}, at: [<ffffffffa0144228>] ieee80211_sta_work+0x108/0x10f0 [mac80211]

which lock already depends on the new lock.

the existing dependency chain (in reverse order) is:

-> #3 (&ifmgd->mtx){+.+.+.}:
       [<ffffffff810857b6>] __lock_acquire+0xd76/0x12b0
       [<ffffffff81085dd3>] lock_acquire+0xe3/0x120
       [<ffffffff814eeae4>] mutex_lock_nested+0x44/0x350
       [<ffffffffa0141bb8>] ieee80211_mgd_auth+0x108/0x1f0 [mac80211]
       [<ffffffffa0148563>] ieee80211_auth+0x13/0x20 [mac80211]
       [<ffffffffa00bc3a1>] __cfg80211_mlme_auth+0x1b1/0x2a0 [cfg80211]
       [<ffffffffa00bc516>] cfg80211_mlme_auth+0x86/0xc0 [cfg80211]
       [<ffffffffa00b368d>] nl80211_authenticate+0x21d/0x230 [cfg80211]
       [<ffffffff81416ba6>] genl_rcv_msg+0x1b6/0x1f0
       [<ffffffff81415c39>] netlink_rcv_skb+0x89/0xb0
       [<ffffffff814169d9>] genl_rcv+0x29/0x40
       [<ffffffff8141591d>] netlink_unicast+0x29d/0x2b0
       [<ffffffff81416514>] netlink_sendmsg+0x214/0x300
       [<ffffffff813e4407>] sock_sendmsg+0x107/0x130
       [<ffffffff813e45b9>] sys_sendmsg+0x189/0x320
       [<ffffffff81011f82>] system_call_fastpath+0x16/0x1b
       [<ffffffffffffffff>] 0xffffffffffffffff

-> #2 (&wdev->mtx){+.+.+.}:
       [<ffffffff810857b6>] __lock_acquire+0xd76/0x12b0
       [<ffffffff81085dd3>] lock_acquire+0xe3/0x120
       [<ffffffff814eeae4>] mutex_lock_nested+0x44/0x350
       [<ffffffffa00ab304>] cfg80211_netdev_notifier_call+0x1a4/0x390 [cfg80211]
       [<ffffffff814f3dff>] notifier_call_chain+0x3f/0x80
       [<ffffffff81075a91>] raw_notifier_call_chain+0x11/0x20
       [<ffffffff813f665a>] dev_open+0x10a/0x120
       [<ffffffff813f59bd>] dev_change_flags+0x9d/0x1e0
       [<ffffffff8144eb6e>] devinet_ioctl+0x6fe/0x760
       [<ffffffff81450204>] inet_ioctl+0x94/0xc0
       [<ffffffff813e25fa>] sock_ioctl+0x6a/0x290
       [<ffffffff8111e911>] vfs_ioctl+0x31/0xa0
       [<ffffffff8111ea9a>] do_vfs_ioctl+0x8a/0x5c0
       [<ffffffff8111f069>] sys_ioctl+0x99/0xa0
       [<ffffffff81011f82>] system_call_fastpath+0x16/0x1b
       [<ffffffffffffffff>] 0xffffffffffffffff

-> #1 (&rdev->mtx){+.+.+.}:
       [<ffffffff810857b6>] __lock_acquire+0xd76/0x12b0
       [<ffffffff81085dd3>] lock_acquire+0xe3/0x120
       [<ffffffff814eeae4>] mutex_lock_nested+0x44/0x350
       [<ffffffffa00ac4d0>] cfg80211_get_dev_from_ifindex+0x60/0x90 [cfg80211]
       [<ffffffffa00b21ff>] get_rdev_dev_by_info_ifindex+0x6f/0xa0 [cfg80211]
       [<ffffffffa00b51eb>] nl80211_set_interface+0x3b/0x260 [cfg80211]
       [<ffffffff81416ba6>] genl_rcv_msg+0x1b6/0x1f0
       [<ffffffff81415c39>] netlink_rcv_skb+0x89/0xb0
       [<ffffffff814169d9>] genl_rcv+0x29/0x40
       [<ffffffff8141591d>] netlink_unicast+0x29d/0x2b0
       [<ffffffff81416514>] netlink_sendmsg+0x214/0x300
       [<ffffffff813e4407>] sock_sendmsg+0x107/0x130
       [<ffffffff813e45b9>] sys_sendmsg+0x189/0x320
       [<ffffffff81011f82>] system_call_fastpath+0x16/0x1b
       [<ffffffffffffffff>] 0xffffffffffffffff

other info that might help us debug this:

3 locks held by phy1/1709:
 #0:  ((wiphy_name(local->hw.wiphy))){+.+.+.}, at: [<ffffffff8106b45d>] worker_thread+0x19d/0x340
 #1:  (&ifmgd->work){+.+.+.}, at: [<ffffffff8106b45d>] worker_thread+0x19d/0x340
 #2:  (&ifmgd->mtx){+.+.+.}, at: [<ffffffffa0144228>] ieee80211_sta_work+0x108/0x10f0 [mac80211]

Reported-by: Reinette Chatre <reinette.chatre@intel.com>
Acked-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agocfg80211: do not iterate over rdev list on country IE hint
Luis R. Rodriguez [Fri, 31 Jul 2009 00:38:07 +0000 (17:38 -0700)]
cfg80211: do not iterate over rdev list on country IE hint

Simplify the country IE hint code by just bailing out if
a previous country IE has been issued. We currently just trust
the first AP we connect to on any card. The idea was to perform
conflict resolution within this routine but since we can no longer
iterate over the registered device list here we leave conflict
resolution to be dealt with at a later time on the workqueue.

This code has no functional changes other than saving us an
interation over the registered device list when a second card
is connected, or you unplug and connect the same one, and a
country IE is received. This would have been done upon every
beacon received.

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agocfg80211: use goto out on country IE reg hint failure
Luis R. Rodriguez [Fri, 31 Jul 2009 00:38:06 +0000 (17:38 -0700)]
cfg80211: use goto out on country IE reg hint failure

This has no functional changes.

Acked-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agomac80211: fix powersave
Johannes Berg [Thu, 30 Jul 2009 19:43:55 +0000 (21:43 +0200)]
mac80211: fix powersave

Some of the recent MLME rework I did broke powersave
because the ps_sdata isn't assigned at the right time,
and the work item wasn't removed from the list before
calling ieee80211_recalc_ps(). To be more specific,
this broke the case where you'd enabled PS before
associating, either automatically or with iwconfig.

Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agorndis_wlan: rework key handling
Jussi Kivilinna [Thu, 30 Jul 2009 16:42:08 +0000 (19:42 +0300)]
rndis_wlan: rework key handling

Organize key data in private structure better and store WPA keys, so
they can be restored as WEP keys.

Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agorndis_wlan: add missing padding to struct rndis_80211_remove_key
Jussi Kivilinna [Thu, 30 Jul 2009 16:42:03 +0000 (19:42 +0300)]
rndis_wlan: add missing padding to struct rndis_80211_remove_key

OID_802_11_REMOVE_KEY failed with invalid length error, add missing padding to
structure fix this.

Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agorndis_wlan: handle 802.11 indications from device
Jussi Kivilinna [Thu, 30 Jul 2009 16:41:58 +0000 (19:41 +0300)]
rndis_wlan: handle 802.11 indications from device

Add handling for 802.11 specific rndis indications.

Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agorndis_host: allow rndis_wlan to see all indications
Jussi Kivilinna [Thu, 30 Jul 2009 16:41:52 +0000 (19:41 +0300)]
rndis_host: allow rndis_wlan to see all indications

Allow rndis_wlan to see all indications. Currently rndis_host lets rndis_wlan to
know about link state changes only, but there is whole set of other
802.11-specific indications that rndis_wlan should handle properly. So rename
link_change() to indication() and convert rndis_wlan to use it.

Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Cc: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agorndis_wlan: add rndis_set/query_oid debugging
Jussi Kivilinna [Thu, 30 Jul 2009 16:41:47 +0000 (19:41 +0300)]
rndis_wlan: add rndis_set/query_oid debugging

Add better debugging for failed OID queries.

Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agorndis_wlan: set current packet filter to zero on stop
Jussi Kivilinna [Thu, 30 Jul 2009 16:41:42 +0000 (19:41 +0300)]
rndis_wlan: set current packet filter to zero on stop

Set current packet filter to zero to block receiving data packets from
device.

Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agorndis_wlan: reset device and restore multicast list on rndis_wlan_reset()
Jussi Kivilinna [Thu, 30 Jul 2009 16:41:37 +0000 (19:41 +0300)]
rndis_wlan: reset device and restore multicast list on rndis_wlan_reset()

Reset device properly with RNDIS_MSG_RESET in rndis_wlan_reset() and restore
multicast list afterwards.

Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agorndis_wlan: clear cfg80211 scan on rndis_wlan_stop()
Jussi Kivilinna [Thu, 30 Jul 2009 16:41:31 +0000 (19:41 +0300)]
rndis_wlan: clear cfg80211 scan on rndis_wlan_stop()

Scanning gets stuck if device is stopped when scan is active. Fix by
clearing/aborting cfg80211 scan on rndis_wlan_stop().

Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agorndis_wlan: stop workers on rndis_wlan_stop() and restore on rndis_wlan_reset()
Jussi Kivilinna [Thu, 30 Jul 2009 16:41:26 +0000 (19:41 +0300)]
rndis_wlan: stop workers on rndis_wlan_stop() and restore on rndis_wlan_reset()

Driver doesn't need to poll statistics/link status when stopped.

Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agousbnet: allow "minidriver" to prevent urb unlinking on usbnet_stop
Jussi Kivilinna [Thu, 30 Jul 2009 16:41:20 +0000 (19:41 +0300)]
usbnet: allow "minidriver" to prevent urb unlinking on usbnet_stop

rndis_wlan devices freeze after running usbnet_stop several times. It appears
that firmware freezes in state where it does not respond to any RNDIS commands
and device have to be physically unplugged/replugged. This patch lets
minidrivers to disable unlink_urbs on usbnet_stop through new info flag.

Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Cc: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agocfg80211: fix NETDEV_UNREGISTER notifier
Johannes Berg [Thu, 30 Jul 2009 12:04:01 +0000 (14:04 +0200)]
cfg80211: fix NETDEV_UNREGISTER notifier

It's possible to get the NETDEV_UNREGISTER callback multiple
times (see net/core/dev.c:netdev_wait_allrefs) and this will
completely mess up our cleanup code. To avoid that, clean up
only when the interface is still on the wiphy interface list
from which it's removed on the first NETDEV_UNREGISTER call.

Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoiwlwifi: don't export symbols not needed in other modules
Johannes Berg [Thu, 30 Jul 2009 09:58:16 +0000 (11:58 +0200)]
iwlwifi: don't export symbols not needed in other modules

Even with the split into iwlcore/agn/3945 not all symbols
that cross file boundaries are needed in other modules, a
few are only used within iwlcore, for example.

Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agomac80211: redefine usage of the mac80211 workqueue
Luis R. Rodriguez [Thu, 30 Jul 2009 00:08:07 +0000 (20:08 -0400)]
mac80211: redefine usage of the mac80211 workqueue

The mac80211 workqueue exists to enable mac80211 and drivers
to queue their own work on a single threaded workqueue. mac80211
takes care to flush the workqueue during suspend but we never
really had requirements on drivers for how they should use
the workqueue in consideration for suspend.

We extend mac80211 to document how the mac80211 workqueue should
be used, how it should not be used and finally move raw access to
the workqueue to mac80211 only. Drivers and mac80211 use helpers
to queue work onto the mac80211 workqueue:

  * ieee80211_queue_work()
  * ieee80211_queue_delayed_work()

These helpers will now warn if mac80211 already completed its
suspend cycle and someone is trying to queue work. mac80211
flushes the mac80211 workqueue prior to suspend a few times,
but we haven't taken the care to ensure drivers won't add more
work after suspend. To help with this we add a warning when
someone tries to add work and mac80211 already completed the
suspend cycle.

Drivers should ensure they cancel any work or delayed work
in the mac80211 stop() callback.

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoat76c50x-usb: remove unneeded flush_workqueue() at usb disconnect
Luis R. Rodriguez [Wed, 29 Jul 2009 21:20:41 +0000 (14:20 -0700)]
at76c50x-usb: remove unneeded flush_workqueue() at usb disconnect

This driver only uses the mac80211 workqueue and mac80211 requires us to
cancel all work at driver stop. Since we now have the cancels in the right
places at stop() we really don't need to flush the mac80211 workqueue so
remove it.

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Acked-by: Kalle Valo <kalle.valo@iki.fi>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath5k: fix CAB queue operation
Bob Copeland [Wed, 29 Jul 2009 21:32:28 +0000 (17:32 -0400)]
ath5k: fix CAB queue operation

We need to process tx descriptors for all queues (currently main tx
queue and cabq) which may have triggered the TX completion interrupt.
Otherwise, the queues can get stuck after sending a few frames.

Signed-off-by: Bob Copeland <me@bobcopeland.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoorinoco: enable cfg80211 "set_channel" operation
Holger Schurig [Tue, 28 Jul 2009 14:34:26 +0000 (16:34 +0200)]
orinoco: enable cfg80211 "set_channel" operation

Signed-off-by: Holger Schurig <hs4233@mail.mn-solutions.de>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoar9170: remove EXPERIMENTAL marker
Luis R. Rodriguez [Mon, 27 Jul 2009 23:52:10 +0000 (16:52 -0700)]
ar9170: remove EXPERIMENTAL marker

Cc: Christian Lamparter <chunkeey@web.de>
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agob43: Work around mac80211 race condition
Larry Finger [Wed, 29 Jul 2009 15:54:06 +0000 (10:54 -0500)]
b43: Work around mac80211 race condition

As shown in http://thread.gmane.org/gmane.linux.kernel.wireless.general/36497,
mac80211 has a bug that allows a call to the TX routine after the queues have
been stopped. This situation will only occur under extreme stress. Although
b43 does not crash when this condition occurs, it does generate a WARN_ON and
also logs a queue overrun message. This patch recognizes b43 is not at fault
and logs a message only when the most verbose debugging mode is enabled. In
the unlikely event that the queue is not stopped when the DMA queue becomes
full, then a warning is issued.

During testing of this patch with one output stream running repeated tcpperf
writes and a second running a flood ping, this routine was entered with
the DMA ring stopped about once per hour. The condition where the DMA queue is
full but the ring has not been stopped has never been seen by me.

Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath5k: update PCU opmode whenever a new interface is added
Bob Copeland [Wed, 29 Jul 2009 14:29:03 +0000 (10:29 -0400)]
ath5k: update PCU opmode whenever a new interface is added

Previously, we would store the operating mode at interface up time,
but only update the PCU registers when the next reset happened.
The result is that if beacon configuration (ops->bss_info_changed)
happens before ops->config, we will program the wrong things into
the timer registers.  Consequently, beacons won't work in AP mode
until after a reset (channel change, scan etc.).

This is fragile anyway so just program the opmode as soon as
mac80211 gives it to us.

Signed-off-by: Bob Copeland <me@bobcopeland.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agomac80211: fix sparse warnings/errors
Johannes Berg [Wed, 29 Jul 2009 13:46:21 +0000 (15:46 +0200)]
mac80211: fix sparse warnings/errors

sparse complains about a shadowed variable, which
we can just rename, and lots of stuff if the API
tracer is enabled, so kick out the tracer code in
a sparse run -- the macros just confuse it.

Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: Remove unused ath9k_hw_intrget()
Vasanthakumar Thiagarajan [Wed, 29 Jul 2009 09:35:23 +0000 (15:05 +0530)]
ath9k: Remove unused ath9k_hw_intrget()

Signed-off-by: Vasanthakumar Thiagarajan <vasanth@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: Maintain monotonicity of PER while going across different phy
Vasanthakumar Thiagarajan [Wed, 29 Jul 2009 09:35:22 +0000 (15:05 +0530)]
ath9k: Maintain monotonicity of PER while going across different phy

Monotonicity of packet error rate should be kept when moving
from one phy to another (legacy to ht, ht single stream to dual,
etc). Current code skips updating per for other phys.

Signed-off-by: Vasanthakumar Thiagarajan <vasanth@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: Update rate control for 11NA HT40 mode
Vasanthakumar Thiagarajan [Wed, 29 Jul 2009 09:35:21 +0000 (15:05 +0530)]
ath9k: Update rate control for 11NA HT40 mode

Now the lowest rate in 11na ht40 mode is 13.5Mbps this shortens the
range when compared to 11na ht20 mode where the lowest rate is 6.5Mbps.
To improve the range, make 6.5Mbps as the lowest rate in 11na ht40 mode,
this improves the range by approximately 2dB. 11ng ht40 does not have
this issue as it also has basic rates (1, 2, 5.5 and 11).

Signed-off-by: Vasanthakumar Thiagarajan <vasanth@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agocfg80211: keep track of current_bss for userspace SME
Johannes Berg [Wed, 29 Jul 2009 09:23:49 +0000 (11:23 +0200)]
cfg80211: keep track of current_bss for userspace SME

When a userspace SME is active, we're currently not
keeping track of the BSS properly for reporting the
current link and for internal use. Additionally, it
looks like there is a possible BSS leak in that the
BSS never gets removed from auth_bsses[]. To fix it,
pass the BSS struct to __cfg80211_connect_result in
this case.

Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agomac80211: disable beacons before removing the associated interface
Bob Copeland [Wed, 29 Jul 2009 08:13:03 +0000 (10:13 +0200)]
mac80211: disable beacons before removing the associated interface

When downing interfaces, it's a good idea to tell the driver to
stop sending beacons; that way the driver doesn't need special
code in ops->remove_interface() when it should already handle the
case in bss_info_changed().

This fixes a potential crash with at least ath5k since the vif
pointer will be nullified while beacon interrupts are still active.

Signed-off-by: Bob Copeland <me@bobcopeland.com>
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agort2x00: Fix build error when crypto support is disabled
Ivo van Doorn [Tue, 28 Jul 2009 16:58:54 +0000 (18:58 +0200)]
rt2x00: Fix build error when crypto support is disabled

When only rt2400pci or rt2500pci is compiled without any of the other
rt2x00 modules, then CONFIG_RT2X00_LIB_CRYPTO will not be enabled.

However rt2x00mac_set_tim() implemented within #ifdef CONFIG_RT2X00_LIB_CRYPTO
statements while the declaration is placed outside the definition. This results in linking
errors as reporte by Ken.

rt2x00_set_tim() has nothing to do with crypto, and thus should be moved outside
of the #ifdef statements.

Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agort2x00: move experimental on kconfig only to rt2800usb
Luis R. Rodriguez [Tue, 28 Jul 2009 15:03:42 +0000 (08:03 -0700)]
rt2x00: move experimental on kconfig only to rt2800usb

These drivers have been around for a while, if there are issues
they should be reported. rt2800usb is still a bit flaky though.

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Acked-by: Ivo van Doorn <IvDoorn@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agowireless: use menuconfig for WLAN_PRE80211 and WLAN_80211
Luis R. Rodriguez [Mon, 27 Jul 2009 23:56:15 +0000 (16:56 -0700)]
wireless: use menuconfig for WLAN_PRE80211 and WLAN_80211

This should make it very clear which are pre-802.11 or not

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath5k: remove EXPERIMENTAL marker
Luis R. Rodriguez [Mon, 27 Jul 2009 23:52:09 +0000 (16:52 -0700)]
ath5k: remove EXPERIMENTAL marker

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath: use menuconfig to put ath stuff in its own page
Luis R. Rodriguez [Mon, 27 Jul 2009 23:52:08 +0000 (16:52 -0700)]
ath: use menuconfig to put ath stuff in its own page

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath: simplify kconfig dependency and add documentation
Luis R. Rodriguez [Mon, 27 Jul 2009 23:52:07 +0000 (16:52 -0700)]
ath: simplify kconfig dependency and add documentation

Make atheros wireless drivers visible when you select
"Atheros wirless drivers". Adds links to ath.ko page,
and Atheros drivers page on the wiki.

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath: depend on cfg80211
Luis R. Rodriguez [Mon, 27 Jul 2009 23:52:06 +0000 (16:52 -0700)]
ath: depend on cfg80211

The ath.ko module itself depends on cfg80211

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agowext: remove extra return on wireless_nlevent_init()
Luis R. Rodriguez [Mon, 27 Jul 2009 23:10:22 +0000 (16:10 -0700)]
wext: remove extra return on wireless_nlevent_init()

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoar9170: remove unneeded flush_workqueue()
Luis R. Rodriguez [Mon, 27 Jul 2009 19:51:37 +0000 (12:51 -0700)]
ar9170: remove unneeded flush_workqueue()

cancel_delayed_work_sync() and cancel_work_sync() are already being
used therefore already waiting for all pending work by the driver
to have been completed, no need to flush the mac80211 workqueue.

Cc: Christian Lamparter <chunkeey@web.de>
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoat76c50x-usb: cancel scan work at stop callback
Luis R. Rodriguez [Mon, 27 Jul 2009 18:56:24 +0000 (11:56 -0700)]
at76c50x-usb: cancel scan work at stop callback

This should fix suspend as mac80211 expects all work queued
to the mac80211 workqueue to be canceled at driver stop().

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Acked-by: Kalle Valo <kalle.valo@iki.fi>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoar76c50x-usb: cancel promisc work during mac80211 stop
Luis R. Rodriguez [Mon, 27 Jul 2009 18:56:23 +0000 (11:56 -0700)]
ar76c50x-usb: cancel promisc work during mac80211 stop

We weren't ever cancelling this.

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Acked-by: Kalle Valo <kalle.valo@iki.fi>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: move workqueue cancels to stop callback
Luis R. Rodriguez [Mon, 27 Jul 2009 18:53:04 +0000 (11:53 -0700)]
ath9k: move workqueue cancels to stop callback

We should be cancelling our work at the stop callback since
we are borrowing the mac80211 workqueue for our work. As it
stands mac80211 expects this for suspend purposes.

The ath9k specific virtual wiphy stuff need only be
cancelled only when the we have no secondary virtual wiphys.

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: move cancel_delayed_work_sync() out of ath_deinit_leds()
Luis R. Rodriguez [Mon, 27 Jul 2009 18:53:03 +0000 (11:53 -0700)]
ath9k: move cancel_delayed_work_sync() out of ath_deinit_leds()

We do this as we'll be moving the cancel elsewhere later.

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoath9k: re-order cancelling of work on mac80211 workqueue
Luis R. Rodriguez [Mon, 27 Jul 2009 18:53:02 +0000 (11:53 -0700)]
ath9k: re-order cancelling of work on mac80211 workqueue

ath9k uses the mac80211 workqueue for 4 different types of work:

 * Led blink work
 * TX hang monitoring work
 * internal wiphy schedular work
 * channel change work done for internal wiphy schedular

Since the internal wiphy schedular can end up kicking off some
channel channel change work we should first cancel the wiphy
schedular work and then the channel change work.

The TX hang work can be cancelled second since we're going down
anyway.

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agomac80211: disable software retry for now
Johannes Berg [Mon, 27 Jul 2009 18:28:40 +0000 (20:28 +0200)]
mac80211: disable software retry for now

Pavel Roskin reported a problem that seems to be due to
software retry of already transmitted frames. It turns
out that we've never done that correctly, but due to
some recent changes it now crashes in the TX code. I've
added a comment in the patch that explains the problem
better and also points to possible solutions -- which
I can't implement right now.

Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
15 years agoneigh: Convert garbage collection from softirq to workqueue
Eric Dumazet [Thu, 30 Jul 2009 03:15:07 +0000 (03:15 +0000)]
neigh: Convert garbage collection from softirq to workqueue

Current neigh_periodic_timer() function is fired by timer IRQ, and
scans one hash bucket each round (very litle work in fact)

As we are supposed to scan whole hash table in 15 seconds, this means
neigh_periodic_timer() can be fired very often. (depending on the number
of concurrent hash entries we stored in this table)

Converting this to a workqueue permits scanning whole table, minimizing
icache pollution, and firing this work every 15 seconds, independantly
of hash table size.

This 15 seconds delay is not a hard number, as work is a deferrable one.

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
15 years agoIPVS: use pr_err and friends instead of IP_VS_ERR and friends
Hannes Eder [Sun, 2 Aug 2009 11:05:41 +0000 (11:05 +0000)]
IPVS: use pr_err and friends instead of IP_VS_ERR and friends

Since pr_err and friends are used instead of printk there is no point
in keeping IP_VS_ERR and friends.  Furthermore make use of '__func__'
instead of hard coded function names.

Signed-off-by: Hannes Eder <heder@google.com>
Acked-by: Simon Horman <horms@verge.net.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
15 years agosmsc95xx: remove EEPROM loaded check
Steve Glendinning [Tue, 28 Jul 2009 02:37:58 +0000 (02:37 +0000)]
smsc95xx: remove EEPROM loaded check

The eeprom read & write commands currently check the E2P_CMD_LOADED_ bit is
set before allowing any operations.  This prevents any reading or writing
unless a correctly programmed EEPROM is installed.

This patch removes the check, so it is possible to program blank EEPROMS
via ethtool.

Signed-off-by: Steve Glendinning <steve.glendinning@smsc.com>
Signed-off-by: David S. Miller <davem@davemloft.net>