Signed-off-by: NOleksij Rempel <linux@rempel-privat.de>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

Signed-off-by: NOleksij Rempel <linux@rempel-privat.de>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

One of the benefits of platform_driver_probe() is that you can make
the probe function __init.
Signed-off-by: NJean Delvare <jdelvare@suse.de>
Cc: Hante Meuleman <meuleman@broadcom.com>
Cc: Arend van Spriel <arend@broadcom.com>
Cc: John W. Linville <linville@tuxdriver.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

The use of __constant_<foo> has been unnecessary for quite awhile now.

Make these uses consistent with the rest of the kernel.
Signed-off-by: NJoe Perches <joe@perches.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

Trivial patch to address this trace. Now calls dma_mapping_error and
return -ENOSPC if a problem found.

WARNING: at lib/dma-debug.c:933 check_unmap+0x47b/0x960()
Hardware name: Aspire 5515
ath5k 0000:02:00.0: DMA-API: device driver failed to check map
error[device address=0x00000000874fcd42] [size=45 bytes] [mapped as
single]
Modules linked in: bnep bluetooth ebtable_filter ebtables
ip6table_filter ip6_tables be2iscsi iscsi_boot_sysfs bnx2i cnic uio
cxgb4i cxgb4 cxgb3i cxgb3 mdio libcxgbi ib_iser rdma_cm ib_addr iw_cm
ib_cm ib_sa ib_mad ib_core iscsi_tcp libiscsi_tcp libiscsi
scsi_transport_iscsi arc4 snd_hda_codec_realtek snd_hda_intel
snd_hda_codec snd_hwdep snd_seq snd_seq_device ath5k ath snd_pcm
sparse_keymap snd_page_alloc mac80211 snd_timer sp5100_tco snd edac_core
k8temp soundcore edac_mce_amd i2c_piix4 cfg80211 rfkill shpchp vhost_net
tun macvtap macvlan kvm_amd kvm uinput dm_crypt ata_generic pata_acpi
radeon i2c_algo_bit pata_atiixp drm_kms_helper ttm drm r8169 mii
i2c_core wmi video sunrpc
Pid: 820, comm: firewalld Not tainted 3.9.0-0.rc3.git1.4.fc19.x86_64 #1
Call Trace:
 <IRQ>  [<ffffffff81068df0>] warn_slowpath_common+0x70/0xa0
 [<ffffffff81068e6c>] warn_slowpath_fmt+0x4c/0x50
 [<ffffffff8137ebeb>] check_unmap+0x47b/0x960
 [<ffffffff81021cf5>] ? native_sched_clock+0x15/0x80
 [<ffffffff81021d69>] ? sched_clock+0x9/0x10
 [<ffffffff8137f12f>] debug_dma_unmap_page+0x5f/0x70
 [<ffffffffa048fa07>] ath5k_tasklet_tx+0x157/0x3f0 [ath5k]
 [<ffffffff810acc0d>] ? sched_clock_local+0x1d/0x80
 [<ffffffff81072aa6>] ? tasklet_action+0x56/0x210
 [<ffffffff81072ae7>] tasklet_action+0x97/0x210
 [<ffffffff8107343f>] __do_softirq+0xff/0x400
 [<ffffffff81073905>] irq_exit+0xb5/0xc0
 [<ffffffff81728916>] do_IRQ+0x56/0xc0
 [<ffffffff8171dc32>] common_interrupt+0x72/0x72
 <EOI>  [<ffffffff811f1f81>] ? dput+0x111/0x310
 [<ffffffff811f1ea7>] ? dput+0x37/0x310
 [<ffffffff811e75e8>] link_path_walk+0x528/0x910
 [<ffffffff811ea624>] path_openat+0x94/0x530
 [<ffffffff811eb148>] do_filp_open+0x38/0x80
 [<ffffffff811e1dda>] open_exec+0x4a/0x130
 [<ffffffff81238be3>] load_elf_binary+0x7f3/0x18e0
 [<ffffffff81021d69>] ? sched_clock+0x9/0x10
 [<ffffffff810acc0d>] ? sched_clock_local+0x1d/0x80
 [<ffffffff810acd98>] ? sched_clock_cpu+0xa8/0x100
 [<ffffffff810d515d>] ? trace_hardirqs_off+0xd/0x10
 [<ffffffff810acedf>] ? local_clock+0x5f/0x70
 [<ffffffff810d5dbf>] ? lock_release_holdtime.part.28+0xf/0x190
 [<ffffffff812383f0>] ? elf_core_dump+0x1980/0x1980
 [<ffffffff811e1151>] search_binary_handler+0x1a1/0x4f0
 [<ffffffff811e1017>] ? search_binary_handler+0x67/0x4f0
 [<ffffffff811e28fc>] do_execve_common.isra.26+0x64c/0x710
 [<ffffffff811e23c2>] ? do_execve_common.isra.26+0x112/0x710
 [<ffffffff811e2cc6>] sys_execve+0x36/0x50
 [<ffffffff81727249>] stub_execve+0x69/0xa0
Signed-off-by: NJohn Greene <jogreene@redhat.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

On older chips, the INI value differ in similar ways as cycpwr_thr1, so
convert it to absolute values as well.

Since the ANI algorithm is different here compared to the old
implementation (fewer steps, controlled at a different point in time),
it makes sense to use values similar to what would be applied for newer
chips, just without relying on INI defaults.
Signed-off-by: NFelix Fietkau <nbd@openwrt.org>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

The table was copied from the ANI implementation of AR9300. It assumes
that the INI values contain a baseline value that is usable as reference
from which to increase/decrease based on the noise immunity value.

On older chips, the differences are bigger and especially AR5008/AR9001
are configured to much more sensitive values than what is useful.

Improve ANI behavior by reverting to the absolute values used in the
previous implementation (expressed as a simple formula instead of the
old table).
Signed-off-by: NFelix Fietkau <nbd@openwrt.org>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

The primary purpose of this piece of code was to selectively disable
OFDM weak signal detection. The checks for this are elsewhere, and an
earlier commit relaxed the restrictions for older chips, which are more
sensitive to interference.
Signed-off-by: NFelix Fietkau <nbd@openwrt.org>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

commit 4c59ff22 "wireless: Kconfig: add missing dependency" added a number
of 'depends on CFG80211' statements, but missed the AIRO_CS driver that
also causes the airo.c file to be built. This adds the (hopefully) last
such missing statement

Cc: "Zhao, Gang" <gamerh2o@gmail.com>
Cc: "John W. Linville" <linville@tuxdriver.com>
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

The brcms_attach function is defined as static but the comment is
saying that it should not be static or gcc will issue a warning.
I believe we can remove the comment as I don't se a problem with
this function being defined as static.
Signed-off-by: NEmil Goode <emilgoode@gmail.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

When brcm80211 firmware is not installed networking hangs.
A deadlock happens because we call ieee80211_unregister_hw()
from the .start callback of struct ieee80211_ops. When .start
is called we are under rtnl lock and ieee80211_unregister_hw()
tries to take it again.

Function call stack:

dev_change_flags()
	__dev_change_flags()
		__dev_open()
			ASSERT_RTNL() <-- Assert rtnl lock
			ops->ndo_open()

.ndo_open = ieee80211_open,

ieee80211_open()
	ieee80211_do_open()
		drv_start()
			local->ops->start()

.start = brcms_ops_start,

brcms_ops_start()
	brcms_remove()
		ieee80211_unregister_hw()
			rtnl_lock() <-- Here we deadlock

Introduced by:
commit 25b5632f
("brcmsmac: request firmware in .start() callback")

This patch fixes the bug by removing the call to brcms_remove()
and moves the brcms_request_fw() call to the top of the .start
callback to not initiate anything unless firmware is installed.
Signed-off-by: NEmil Goode <emilgoode@gmail.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

Unify scnprintf calls and include the current OFDM/CCK immunity level.
Signed-off-by: NFelix Fietkau <nbd@openwrt.org>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

qi->tqi_readyTime is written directly to registers that expect
microseconds as unit instead of TU.
When setting the CABQ ready time, cur_conf->beacon_interval is in TU, so
convert it to microseconds before passing it to ath9k_hw.

This should hopefully fix some Tx DMA issues with buffered multicast
frames in AP mode.

Cc: stable@vger.kernel.org
Signed-off-by: NFelix Fietkau <nbd@openwrt.org>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

Currently a "r8185" integer variable is used as a boolean flag to
indicate whether the card is a rtl8185 or not.
Since now the driver supports only rtl8185 and rtl8180 cards, if
"r8185" variable is zero then the card is implicitly assumed to
be a rtl8180.

Now I'm preparing to add support for a third card type (rtl8187se).

This patch changes the "r8185" flag with an enum variable to
explicitly indicate which card type we have.

I'm submitting this this patch now, even if I still have to submit
other patches that not pertain with rtl8187se support, because
IMHO it's not worth rebasing them on the current code, using r8185
flag, and then changing them back again nearly immediately.
BTW if someone feels I really should do this, please tell me..
Signed-off-by: NAndrea Merello <andrea.merello@gmail.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

Currently AMPDU aggregation is preferred over AMSDU. AMSDU
aggregation is performed only if AMPDU streams in firmware
are full.
This patch adds simultaneous AMSDU and AMPDU aggregation
support. This mechanism helps to improve throughput.
AMSDU is enabled only for 8897 chipsets which supports 4K
transmit buffer. User can disable AMSDU using
'disable_tx_amsdu' module parameter.
Signed-off-by: NAmitkumar Karwar <akarwar@marvell.com>
Signed-off-by: NBing Zhao <bzhao@marvell.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

Existing mwifiex_11n_dispatch_pkt() function is renamed as
mwifiex_11n_dispatch_pkt_until_start_win() and a new function
mwifiex_11n_dispatch_pkt() is created for a common code which
dispatches single packet based on interface type.
Signed-off-by: NAmitkumar Karwar <akarwar@marvell.com>
Signed-off-by: NBing Zhao <bzhao@marvell.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

Use negative check for 'status' and return from the function.
This improves readability by avoiding line splits. Also, local
variable is used for start window calculations.
Signed-off-by: NAmitkumar Karwar <akarwar@marvell.com>
Signed-off-by: NBing Zhao <bzhao@marvell.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

As V15 firmware supports VHT rate configuration, we can use this
information received in set bitrate mask handler.
Signed-off-by: NAmitkumar Karwar <akarwar@marvell.com>
Signed-off-by: NBing Zhao <bzhao@marvell.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

During Tx rate configuration, newer firmware V15 expects bitmap
for VHT MCS rates as well.
Signed-off-by: NAmitkumar Karwar <akarwar@marvell.com>
Signed-off-by: NBing Zhao <bzhao@marvell.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

The firmware API version number will be used for future patches
to support different firmware API specs.
Signed-off-by: NAmitkumar Karwar <akarwar@marvell.com>
Signed-off-by: NBing Zhao <bzhao@marvell.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

If p54u_load_firmware() fails, p54u_probe() does not deallocate
already allocated resources. The patch adds proper failure handling.

Found by Linux Driver Verification project (linuxtesting.org).
Signed-off-by: NAlexey Khoroshilov <khoroshilov@ispras.ru>
Acked-by: NChristian Lamparter <chunkeey@googlemail.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

Commit a619d1ab leads to the following static checker warning:

drivers/net/wireless/rtlwifi/rtl8723be/phy.c:667 _rtl8723be_store_tx_power_by_rate()
error: buffer overflow 'rtlphy->tx_power_by_rate_offset[band]' 4 <= 5

This warning arises because the code is testing the indices for the wrong maximum
values. In addition, the tests merely putput a warning, and then procedes to
corrupt memory. With this change, any such invalid memory access is avoided.
Signed-off-by: NLarry Finger <Larry.Finger@lwfinger.net>
Reported-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

There is a flow in which we send the host command in SYNC
mode, but we don't take priv->mutex.

Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1046495

Cc: <stable@vger.kernel.org>
Reviewed-by: NJohannes Berg <johannes.berg@intel.com>
Signed-off-by: NEmmanuel Grumbach <emmanuel.grumbach@intel.com>

Seems like we have an old bug, where we incidently overwrites
the max_antenna_gain we pass to firmware, with zero value.

End of all we are artifically reducing the output power.

This patch removes the excessive assignment on max_antenna_gain,
which is being provided by regulatory domain, and consequently
improves the tx power.
Signed-off-by: NMarek Puzyniak <marek.puzyniak@tieto.com>
Signed-off-by: NBartosz Markowski <bartosz.markowski@tieto.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

Consolidate the list of msdu skbs into the msdu-head skb, delete the
rest of the skbs, pass the msdu-head skb on up the stack as normal.

Tested with high-speed TCP and UDP traffic on modified firmware that
supports raw-rx.
Signed-off-by: NBen Greear <greearb@candelatech.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

It fixes following sparse check warning
>#make C=1 CF=-D__CHECK_ENDIAN__ drivers/net/wireless/mwl8k.o
>drivers/net/wireless/mwl8k.c:3089:6: warning: symbol 'mwl8k_update_survey' was not declared. Should it be static?
Reported-by: NFengguang Wu <fengguang.wu@intel.com>
Signed-off-by: NNishant Sarmukadam <nishants@marvell.com>
Signed-off-by: NYogesh Ashok Powar <yogeshp@marvell.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

It fixes couple of sparse check
>#make C=1 CF=-D__CHECK_ENDIAN__ drivers/net/wireless/mwl8k.o
>drivers/net/wireless/mwl8k.c:3104:19: warning: cast to restricted __le32
>drivers/net/wireless/mwl8k.c:3108:18: warning: cast to restricted __le32
Reported-by: NFengguang Wu <fengguang.wu@intel.com>
Signed-off-by: NNishant Sarmukadam <nishants@marvell.com>
Signed-off-by: NYogesh Ashok Powar <yogeshp@marvell.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

The intstatus in sdio code can be updated from different
threads. To protect intstatus access, atomic functions are
used. One of them is set_bit, but this function is not
guaranteed atomic on all platforms. The loop was replaced
with local created OR function.
Reviewed-by: NArend Van Spriel <arend@broadcom.com>
Reviewed-by: NFranky (Zhenhui) Lin <frankyl@broadcom.com>
Reviewed-by: NPieter-Paul Giesberts <pieterpg@broadcom.com>
Reviewed-by: NDaniel (Deognyoun) Kim <dekim@broadcom.com>
Signed-off-by: NHante Meuleman <meuleman@broadcom.com>
Signed-off-by: NArend van Spriel <arend@broadcom.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

This reverts commit c98db0be.

The function atomic_set_mask() is not architecture independent
so it can not be used in the driver as is.
Signed-off-by: NArend van Spriel <arend@broadcom.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

Smatch reports the following:

drivers/net/wireless/rtlwifi/rtl8723be/fw.c:208 _rtl8723be_fill_h2c_command() warn: variable dereferenced before check 'rtlhal' (see line 69)
drivers/net/wireless/rtlwifi/rtl8723be/hw.c:1732 _rtl8723be_read_adapter_info() error: __builtin_memcpy() '&rtlefuse->efuse_map[0][0]' too small (256 vs 512)

The first one is fixed by removing two pointless tests for NULL pointers.
Reported-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NLarry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

Sparse reports the following:

drivers/net/wireless/rtlwifi/rtl8723be/sw.c:374:14: sparse: duplicate const

drivers/net/wireless/rtlwifi/rtl8723be/hw.c:2214:30: sparse: cast to restricted __le32
Reported-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NLarry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

This patch prevents log spamming by adding a case for a previously
unhandled case.
Signed-off-by: NLarry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

In commit f78bccd7 entitled "rtlwifi:
rtl8192ce: Fix too long disable of IRQs", Olivier Langlois
<olivier@trillion01.com> fixed a problem caused by an extra long disabling
of interrupts. This patch makes the same fix for rtl8723ae.
Signed-off-by: NLarry Finger <Larry.Finger@lwfinger.net>
Cc: Stable <stable@vger.kernel.org>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

In commit f78bccd7 entitled "rtlwifi:
rtl8192ce: Fix too long disable of IRQs", Olivier Langlois
<olivier@trillion01.com> fixed a problem caused by an extra long disabling
of interrupts. This patch makes the same fix for rtl8188ee.
Signed-off-by: NLarry Finger <Larry.Finger@lwfinger.net>
Cc: Stable <stable@vger.kernel.org>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

In commit f78bccd7 entitled "rtlwifi:
rtl8192ce: Fix too long disable of IRQs", Olivier Langlois
<olivier@trillion01.com> fixed a problem caused by an extra long disabling
of interrupts. This patch makes the same fix for rtl8192se.
Signed-off-by: NLarry Finger <Larry.Finger@lwfinger.net>
Cc: Stable <stable@vger.kernel.org>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

In commit f78bccd7 entitled "rtlwifi:
rtl8192ce: Fix too long disable of IRQs", Olivier Langlois
<olivier@trillion01.com> fixed a problem caused by an extra long disabling
of interrupts. This patch makes the same fix for rtl8192cu.
Signed-off-by: NLarry Finger <Larry.Finger@lwfinger.net>
Cc: Stable <stable@vger.kernel.org>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

In a previous commit, Peter Wu removed this call as configure_filter takes care
of setting/clearing RCR_AAP. This patch makes the same change for rtl8723be.

In addition, a change is made in the logging level for one debug printout.
Signed-off-by: NLarry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

Signed-off-by: NLarry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

rtlwifi: rtl8192ce: rtl8192cu: rtl8192de: rtl8192se: rtl8723ae: rtl8723be: rtl8188eu: Modify for new API

The addition of a driver for the RTL8821AE requires a new API for the
fill_tx_desc() and set_desc() callback routines. This commit makes the
appropriate modifications in all the other drivers.
Signed-off-by: NLarry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

iwlegacy version of this iwlwifi patch:

commit aed7d9ac
Author: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Date:   Wed Feb 20 11:33:00 2013 +0200

    iwlwifi: disable 8K A-MSDU by default

    Supporting 8K A-MSDU means that we need to allocate order 1
    pages for every Rx packet. Even when there is no traffic.
    This adds stress on the memory manager. The handling of
    compound pages is also less trivial for the memory manager
    and not using them will make the allocation code run faster
    although I didn't really measure.
    Eric also pointed out that having huge buffers with little
    data in them is not very nice towards the TCP stack since
    the truesize of the skb is huge. This doesn't allow TCP
    to have a big Rx window.
    See https://patchwork.kernel.org/patch/2167711/ for details.

    Note that very few vendors will actually send A-MSDU.
    Disable this feature by default.
Signed-off-by: NEmmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>
Signed-off-by: NStanislaw Gruszka <sgruszka@redhat.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>