When bluetooth indicated high load, we should use
only antenna A in 2.4 GHz for management frames.
Add this condition to iwl_toggle_tx_ant() to make
sure it'll always be met. Note that scanning has
a separate way of forcing the antenna, because we
should scan on antenna A only regardless of BT
traffic load.
Signed-off-by: Johannes Berg <[email protected]>
Signed-off-by: Wey-Yi Guy <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
Disable plcp error checking for 6000g2b devices, with wifi/bt coex;
this got trigger too often, disable for now until finding better
trigger point.
Also extend the firmware reload timer much longer for BT coex to make sure
there are will be no mistake to reload firmware too fast and too often.
Signed-off-by: Wey-Yi Guy <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
The BT SCO needs to be re-applied to the device,
while the traffic load just needs to be correct
in software.
Signed-off-by: Johannes Berg <[email protected]>
Signed-off-by: Wey-Yi Guy <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
IBSS doesn't allow for coexistence, so it
should be disabled.
Additionally, disable reacting to the BT
profile notification when in IBSS mode,
it likely won't be sent by the device to
start with though.
Also, in IBSS mode, BT coexistence isn't as fully-featured
and we must use a single antenna only. So instead of
peppering the code with new checks, simply pretend
we are in high BT traffic load, which has the needed
effect of disabling antenna B use.
Signed-off-by: Johannes Berg <[email protected]>
Signed-off-by: Wey-Yi Guy <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
When we turn off the device, reset BT
data so that we don't have outdated
information when we come up again.
Signed-off-by: Johannes Berg <[email protected]>
Signed-off-by: Wey-Yi Guy <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
Depending on the amount of bluetooth traffic,
using the shared antenna (antenna B) will have
adverse impact on both bluetooth and wireless
traffic. Add controls to improve the situation
by making rate scaling depend on the BT load.
When there's high bluetooth traffic load, there's
little point in trying to aggregate as BT traffic
would disrupt the aggregated frames all the time,
so simply don't start sessions then.
When BT traffic returns to lower levels, the rate
scaling will come here again automatically when
wifi traffic is high enough, and then it will be
able to successfully enable aggregation.
Signed-off-by: Johannes Berg <[email protected]>
Signed-off-by: Wey-Yi Guy <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
When advanced bt coex enabled, uCode will send bt status
notification to driver, here add support for it.
Signed-off-by: Johannes Berg <[email protected]>
Signed-off-by: Wey-Yi Guy <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
The BT ignore bit should be set when transmitting
auth, assoc response and eap frames.
Also, scanning should set the BT ignore bit for the
probe request transmission; Note that we only use
the non-shared antenna.
Signed-off-by: Johannes Berg <[email protected]>
Signed-off-by: Wey-Yi Guy <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
6000g2b hardware implements advance bluetooth coexist command,
implement base on the new API command strucutre.
Also increment the API 5 to support the advance BT/WIfi coex.
Signed-off-by: Johannes Berg <[email protected]>
Signed-off-by: Wey-Yi Guy <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
The new 6000g2b hardware requires a different
bluetooth coexist implementation on the host,
this adds the command/notification definitions
for it.
Signed-off-by: Johannes Berg <[email protected]>
Signed-off-by: Wey-Yi Guy <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
No code or function changes, just some cleanup work for
out-of-date comments
Signed-off-by: Wey-Yi Guy <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
No functional changes, update comments to match current file structure
Signed-off-by: Wey-Yi Guy <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
Unlike most other workqueue-tasks, the restart_work is
not scheduled onto mac80211's private per-interface
workqueue, but onto one of the system-wide workqueues.
Therefore the mac80211-stack has to cancel any pending
restarts, before destroying the shared device context
and handing back the memory. Otherwise - under very
unlucky circumstances - there could be a stale work-
item left, because some other kernel component might
have delayed the execution of ieee80211_restart_work
for too long.
Signed-off-by: Christian Lamparter <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
The ISL3887 chip needs a USB reset, whenever the
usb-frontend module "p54usb" is reloaded.
This patch fixes an off-by-one bug, if the user
is running a kernel without the CONFIG_PM option
set and for some reason (e.g.: compat-wireless)
wants to switch between different p54usb modules.
Cc: <[email protected]>
Signed-off-by: Christian Lamparter <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
As my nokia address doesn't work anymore, it's better just to remove the
copyright contact altogether. We have MODULE_AUTHOR() scripts for these
anyway.
Signed-off-by: Kalle Valo <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
Luciano should be the contact person for the include/linux/spi/wl12xx.h file.
Signed-off-by: Kalle Valo <[email protected]>
Acked-by: Luciano Coelho <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
This patch adds an Kconfig option, which allows the user
to select, whenever he/she wants to include a 4k blob for
generic calibration and interface values into the driver,
or cut the module size by about 15 to 20%.
Signed-off-by: Christian Lamparter <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
Descriptors are currently logged with ATH5K_DEBUG_RESET,
which isn't really apt, and also means we can't see just
the descriptor setup or just the resets. Add a new
debug level just for that.
Acked-by: Bruno Randolf <[email protected]>
Signed-off-by: Bob Copeland <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
Fix some comments:
s/transmition/transmission/
s/puting/putting/
Signed-off-by: Bob Copeland <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
Fix spelling and readability of a few lines of kernel doc:
s/issueing/issuing/g
s/approriate/appropriate/g
s/supported by simply/supported simply by/
s/IEEE80211_HW_BEACON_FILTERING/IEEE80211_HW_BEACON_FILTER/g
Signed-off-by: Bob Copeland <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
mesh_hdr only used when CONFIG_MAC80211_MESH is defined
Signed-off-by: Wey-Yi Guy <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
Fix walking past the end of the bitrate_table array
in the case when the loop counter == BITRATE_TABLE_SIZE.
Reported by: Denis Kirjanov <[email protected]>
Signed-off-by: David Kilroy <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
To make del_timer_sync() works we have to assure that timer function
does not rearm the timer. To achieve that we cancel timer with
STATUS_EXIT_PENDING bit set in __iwl{3945,}_down function.
Patch also fix priv->txq memory usage after free for iwl3945, because
it move timer cancellation before iwl3945_hw_txq_ctx_free(priv) is called.
Signed-off-by: Stanislaw Gruszka <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
drivers/net/mlx4/en_rx.c: In function ‘mlx4_en_create_rx_ring’:
drivers/net/mlx4/en_rx.c:305: warning: label ‘err_map’ defined but not used
Signed-off-by: David S. Miller <[email protected]>
Adding capability for RSS for UDP traffic, hashing is done based on
IP addresses and UDP port number.
The support depends on HW/FW capabilities.
Signed-off-by: Yevgeny Petrilin <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
When Mac address is removed from one port of the CX2 device, the other device
should reconfigure its Mac.
This fixes an issue with failover, when both ports have the same Mac address,
and one of the ports, is closed, the second one stops receiving traffic.
(bugzilla #1965 at bugs.openfabrics.org)
Signed-off-by: Yevgeny Petrilin <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
In case that the rings were not fully filled, the report in Ethtool
should be according to the actual size that was filled.
Signed-off-by: Yevgeny Petrilin <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
The report now based on query from FW, giving the correct tranciever type
and link speed.
Signed-off-by: Yevgeny Petrilin <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
The selftest includes 5 features:
1. Interrupt test: Executing commands and receiving command completion
on all our interrupt vectors.
2. Link test: Verifying we are connected to valid link partner.
3. Speed test: Check that we negotiated link speed correctly.
4. Registers test: Activate HW health check command.
5. Loopback test: Send a packet on loopback interface and catch it on RX side.
Signed-off-by: Yevgeny Petrilin <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
We might try to use resources which were not activated.
Signed-off-by: Yevgeny Petrilin <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
Defined as sum of default number of TX rings and number of
rings used for priority flow control mode.
Signed-off-by: Yevgeny Petrilin <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
All port's resources should be closed before we give the
CLOSE_PORT command to the firmware.
Signed-off-by: Yevgeny Petrilin <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
When configuring HW resources, the RX ring size that is passed should be the actual
size that depends on number of buffers that we succeeded to allocate.
The mask for ring size should also be configured accordingly
Signed-off-by: Yevgeny Petrilin <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
When allocating new fragments to replace the ones that would be passed to the stack,
The fragments that should be replaced, are the ones that were already used.
Signed-off-by: Yevgeny Petrilin <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
Trivial extension to existing meta data match rules to allow
matching on skb receive hash value.
Signed-off-by: Stephen Hemminger <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
Compiler is not smart enough to avoid a conditional branch.
Signed-off-by: Eric Dumazet <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
Add some documentation for cfg80211. I'm hoping some of
the regulatory documentation will be filled by somebody
more familiar with it, hint hint! :)
Signed-off-by: Johannes Berg <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
When a scan is aborted because the corresponding
virtual interface is removed, we may still later
attempt to tell mac80211 that the scan completed.
This is obviously wrong, since we already told it
that it was aborted, so don't do that.
Signed-off-by: Johannes Berg <[email protected]>
Signed-off-by: Wey-Yi Guy <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
