Replace SLEEP_OVER_DTIM by SKIP_OVER_DTIM.
Add iwl_mvm prefix to a function name.
Signed-off-by: Alexander Bondar <[email protected]>
Reviewed-by: Emmanuel Grumbach <[email protected]>
Signed-off-by: Johannes Berg <[email protected]>
Current available FW still doesn't support new PM API.
Therefore, to enable basic power management with the
existing firmware, change the API in the driver back
to the API used in the current firmware.
Signed-off-by: Alexander Bondar <[email protected]>
Reviewed-by: Emmanuel Grumbach <[email protected]>
Signed-off-by: Johannes Berg <[email protected]>
When we roam to A band, we don't need to constraint WiFi
any more since it is operating on a different band.
Signed-off-by: Emmanuel Grumbach <[email protected]>
Signed-off-by: Johannes Berg <[email protected]>
This allows to have better wifi TPT when BT is active under
good RSSI conditions.
Wifi will have better chance to send Acks and Cts even if BT
is active.
Signed-off-by: Emmanuel Grumbach <[email protected]>
Signed-off-by: Johannes Berg <[email protected]>
Use the per-interface debugfs infrastructure to create a
directory and symlink, and add a file containing debug
data related to each virtual interface.
Signed-off-by: Johannes Berg <[email protected]>
Firmware provides the driver with credits used to transmit packets
to the firmware. When credits run out the packets should be queued
and dequeued when receiving creditback signals from the firmware.
Reviewed-by: Pieter-Paul Giesberts <[email protected]>
Reviewed-by: Hante Meuleman <[email protected]>
Signed-off-by: Arend van Spriel <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
The fcmode provided by module parameter defaults to NONE, which
means no flow-control is required. In this case flow-control
signals should not be enabled.
Reviewed-by: Pieter-Paul Giesberts <[email protected]>
Reviewed-by: Piotr Haber <[email protected]>
Reviewed-by: Hante Meuleman <[email protected]>
Signed-off-by: Arend van Spriel <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
Bail out of brcmf_fws_init() when no firmware-signalling is asked
for. Need to take this into account in brcmf_fws_deinit() as well.
Reviewed-by: Pieter-Paul Giesberts <[email protected]>
Reviewed-by: Piotr Haber <[email protected]>
Signed-off-by: Arend van Spriel <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
If iovar to the firmware fails the firmware-signalling module
does a cleanup for which it needs pointer to struct brcmf_pub, which
it gets from struct brcmf_fws_info::drvr. Assign this field before
doing the tlv iovar.
Reviewed-by: Pieter-Paul Giesberts <[email protected]>
Reviewed-by: Piotr Haber <[email protected]>
Reviewed-by: Hante Meuleman <[email protected]>
Signed-off-by: Arend van Spriel <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
The low-level sdio code has a large number of trace and info messages
that are mostly useful looking into bus specific issues. For tracing
higher-level driver functions it is better to have a dedicated level
for low-level sdio debugging.
Reviewed-by: Hante Meuleman <[email protected]>
Reviewed-by: Piotr Haber <[email protected]>
Signed-off-by: Arend van Spriel <[email protected]>
Change-Id: Ia424ff18d9033b97aeffc248358e50c51805e815
Reviewed-on: http://lb-bun-88.bun.broadcom.com:8080/74
Signed-off-by: John W. Linville <[email protected]>
During SDIO layer flow control signalling firmware can issue
invalid packets. Prevent printing of parsing errors in such case.
Reviewed-by: Hante Meuleman <[email protected]>
Reviewed-by: Arend van Spriel <[email protected]>
Signed-off-by: Piotr Haber <[email protected]>
Signed-off-by: Arend van Spriel <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
Parsing the tlv upon receiving frames can fail. Instead of printing
an error message, just count the parse failure. On some devices we
receive a lot of invalid tlv signals.
this commit will be squashed.
Signed-off-by: Arend van Spriel <[email protected]>
Change-Id: I08e0f62c55e5028f9aa70c396d291679abd273c9
Reviewed-on: http://lb-bun-88.bun.broadcom.com:8080/72
Reviewed-by: Hante Meuleman <[email protected]>
Reviewed-by: Piotr Haber <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
When firmware-signalling is active the brcmf_txcomplete() does
a free of the sk_buff when transfer to firmware fails in the
bus-specific driver code. However, it should also cleanup the
packet from the hanger. This patch fixes that.
Reviewed-by: Hante Meuleman <[email protected]>
Reviewed-by: Pieter-Paul Giesberts <[email protected]>
Reviewed-by: Piotr Haber <[email protected]>
Signed-off-by: Arend van Spriel <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
Enabling the tx status signalling, which requires packet tagging
before sending to the firmware and handling the tx status signal.
Reviewed-by: Pieter-Paul Giesberts <[email protected]>
Reviewed-by: Hante Meuleman <[email protected]>
Reviewed-by: Piotr Haber <[email protected]>
Signed-off-by: Arend van Spriel <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
Upcoming patches will add firmware signalled flow control. Prepare
by adding the mode, which defaults to disable it. The mode can be
queried by brcmf_fws_fc_active() and set by a module parameter.
Reviewed-by: Hante Meuleman <[email protected]>
Reviewed-by: Piotr Haber <[email protected]>
Reviewed-by: Pieter-Paul Giesberts <[email protected]>
Signed-off-by: Arend van Spriel <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
Add a callback to obtain packet queue from the bus-specific code
used to cleanup packet buffers from firmware-signalling code.
Reviewed-by: Pieter-Paul Giesberts <[email protected]>
Reviewed-by: Hante Meuleman <[email protected]>
Reviewed-by: Piotr Haber <[email protected]>
Reviewed-by: Franky (Zhenhui) Lin <[email protected]>
Signed-off-by: Arend van Spriel <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
The hanger for firmware-signalling is used to retain information for
outstanding transmit packets that await tx status.
Reviewed-by: Pieter-Paul Giesberts <[email protected]>
Reviewed-by: Hante Meuleman <[email protected]>
Reviewed-by: Piotr Haber <[email protected]>
Signed-off-by: Arend van Spriel <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
When an interface is deleted make sure to cleanup all packet
buffers related to that interface.
Reviewed-by: Pieter-Paul Giesberts <[email protected]>
Reviewed-by: Hante Meuleman <[email protected]>
Reviewed-by: Piotr Haber <[email protected]>
Signed-off-by: Arend van Spriel <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
The sk_buff structure contains a control buffer that can be used
by different layers in the networking stack for holding packet
associated information. In brcmfmac it is used to hold firmware
signalling related information.
Reviewed-by: Pieter-Paul Giesberts <[email protected]>
Reviewed-by: Hante Meuleman <[email protected]>
Reviewed-by: Piotr Haber <[email protected]>
Signed-off-by: Arend van Spriel <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
Currently, the netif queue is only stopped when the bus interface is
giving a push back. This will change soon so prepare the driver by
adding a stop reason and stop/resume the queue accordingly.
Reviewed-by: Pieter-Paul Giesberts <[email protected]>
Reviewed-by: Hante Meuleman <[email protected]>
Reviewed-by: Piotr Haber <[email protected]>
Signed-off-by: Arend van Spriel <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
Add a cleanup function releasing any queued packet buffers in
the mac descriptor entries.
Reviewed-by: Pieter-Paul Giesberts <[email protected]>
Reviewed-by: Hante Meuleman <[email protected]>
Reviewed-by: Piotr Haber <[email protected]>
Signed-off-by: Arend van Spriel <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
The firmware signalling functionality needs the credit map firmware
events. This patch adds registration of a handler for this event.
Reviewed-by: Pieter-Paul Giesberts <[email protected]>
Reviewed-by: Hante Meuleman <[email protected]>
Reviewed-by: Piotr Haber <[email protected]>
Signed-off-by: Arend van Spriel <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
Firmware can signal the driver to allocate descriptor info for a given
mac address, which will be used for flow control and host queueing.
Reviewed-by: Pieter-Paul Giesberts <[email protected]>
Reviewed-by: Hante Meuleman <[email protected]>
Reviewed-by: Piotr Haber <[email protected]>
Signed-off-by: Arend van Spriel <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
Firmware signalling needs to handle resources upon interface
events. This patch add calls in the interface event handling
routine.
Reviewed-by: Pieter-Paul Giesberts <[email protected]>
Reviewed-by: Hante Meuleman <[email protected]>
Reviewed-by: Piotr Haber <[email protected]>
Signed-off-by: Arend van Spriel <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
Use information from the device to determine the bands property
of the wiphy object. After this change the support of 80211n is
correctly presented in the bands property.
Reviewed-by: Arend Van Spriel <[email protected]>
Reviewed-by: Pieter-Paul Giesberts <[email protected]>
Reviewed-by: Piotr Haber <[email protected]>
Signed-off-by: Hante Meuleman <[email protected]>
Signed-off-by: Arend van Spriel <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
In brcmf_sdbrcm_txpkt() a new packet is allocated and used to transmit
to firmware freeing up the original packet. However, that packet is
still referenced in firmware-signalling so this would result in a
double free. Using skb_cow() avoids this as the packet reference is
unchanged.
Reviewed-by: Hante Meuleman <[email protected]>
Reviewed-by: Piotr Haber <[email protected]>
Reviewed-by: Pieter-Paul Giesberts <[email protected]>
Signed-off-by: Arend van Spriel <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
When taking care of packet alignment to 64-byte boundary padding may
be added between SDPCM header and CDC data. It clear both SDPCM header
space and padding space. Changed it to only clear padding space. In
filling the SDPCM header it uses unaligned access to set SDPCM software
header, but preceding code assures it is properly aligned.
Signed-off-by: Arend van Spriel <[email protected]>
Change-Id: Iad22f277f3496440ba4d2db771205714774570ac
Reviewed-on: http://lb-bun-88.bun.broadcom.com:8080/76
Reviewed-by: Franky Lin <[email protected]>
Reviewed-by: Piotr Haber <[email protected]>
Reviewed-by: Pieter-Paul Giesberts <[email protected]>
Reviewed-by: Hante Meuleman <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
The function brcmf_sdbrcm_txpkt() calls brcmf_txcomplete() with
a parameter success. For this parameter it passes ret != 0, but
that condition is true upon failure.
Reviewed-by: Franky (Zhenhui) Lin <[email protected]>
Reviewed-by: Piotr Haber <[email protected]>
Reviewed-by: Pieter-Paul Giesberts <[email protected]>
Reviewed-by: Hante Meuleman <[email protected]>
Signed-off-by: Arend van Spriel <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
This reverts commit 63cd353c34.
We no longer need this fix as the MEI bus APIs are now merged into
char-misc-next.
Signed-off-by: Samuel Ortiz <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
The rt2800pci driver supports the built-in wireless
MAC of the Ralink RT3x5x SoCs. However building the
driver for these SoCs leads to the following error:
LD init/built-in.o
drivers/built-in.o: In function `rt2800pci_rxdone_tasklet':
<...>/drivers/net/wireless/rt2x00/rt2800pci.c:1012: undefined reference to `rt2x00pci_rxdone'
drivers/built-in.o:(.rodata+0x4780): undefined reference to `rt2x00pci_initialize'
drivers/built-in.o:(.rodata+0x4784): undefined reference to `rt2x00pci_uninitialize'
drivers/built-in.o:(.rodata+0x47bc): undefined reference to `rt2x00pci_flush_queue'
drivers/built-in.o:(.rodata+0x4818): undefined reference to `rt2x00pci_regbusy_read'
make[5]: *** [vmlinux] Error 1
The missing functions are provided by the rt2x00pci
module. This module is only selected by the rt2800pci
driver if PCI support is enabled in the kernel, because
some parts of the rt2x00pci code depends on PCI support.
PCI support is not available on the RT3x5x SoCs because
those have no PCI host controller at all.
Move the non PCI specific code from rt2x00pci into a
separate module. This makes it possible to use that
code even if PCI support is disabled. The affected
functions are used by all of the rt2x00 PCI drivers
so select the new module for those drivers.
Signed-off-by: Gabor Juhos <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
ath_complete_reset will not be called if ath9k_hw_reset
is unsuccessful, so we need to re-enable intertupts to
balence the previous ath_prepare_reset call. Also schedule a
reset as a best effort method to recover the chip from
whatever state caused the channel change failure.
Fixes https://bugzilla.kernel.org/show_bug.cgi?id=55771
Signed-off-by: Robert Shade <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
Nvram contains critical initialization parameter for firmware to run. Host
driver should not proceed if nvram fails to be downloaded to dongle.
Reviewed-by: Piotr Haber <[email protected]>
Reviewed-by: Arend van Spriel <[email protected]>
Reviewed-by: Pieter-Paul Giesberts <[email protected]>
Signed-off-by: Franky Lin <[email protected]>
Signed-off-by: Arend van Spriel <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
When multiple cipher suites have been programmed then the lowest
suite is to be retured. This fixes issue when AP mode is using
CCMP and TKIP WPA combination where rekeying will fail.
Reviewed-by: Arend Van Spriel <[email protected]>
Reviewed-by: Pieter-Paul Giesberts <[email protected]>
Reviewed-by: Piotr Haber <[email protected]>
Signed-off-by: Hante Meuleman <[email protected]>
Signed-off-by: Arend van Spriel <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
on stop_ap the dongle was not properly shutdown. As a result it was
not possible to restart AP or STA after AP operation without
restarting the device. This patch will fix that.
Reviewed-by: Arend Van Spriel <[email protected]>
Reviewed-by: Pieter-Paul Giesberts <[email protected]>
Reviewed-by: Piotr Haber <[email protected]>
Signed-off-by: Hante Meuleman <[email protected]>
Signed-off-by: Arend van Spriel <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
tx and rx michael tkip keys are always swapped in case being
configured per mac. This is wrong for AP. The swap should only
be done for STA mode.
Reviewed-by: Arend Van Spriel <[email protected]>
Reviewed-by: Pieter-Paul Giesberts <[email protected]>
Reviewed-by: Piotr Haber <[email protected]>
Signed-off-by: Hante Meuleman <[email protected]>
Signed-off-by: Arend van Spriel <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
We are waiting on first scan command of internal scan request
before association, so we should complete on last internal scan
command response.
Cc: <[email protected]> # 3.8
Tested-by: Daniel Drake <[email protected]>
Tested-by: Marco Cesarano <[email protected]>
Signed-off-by: Bing Zhao <[email protected]>
Signed-off-by: Amitkumar Karwar <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
This patch combines the remaining changes in the rtlwifi family to handle
the addition of rtl8188ee. A number of these changes eliminate some CamelCase
variable names, and other shorten common variable names so that long lines
in the new driver could be shortened.
Signed-off-by: Larry Finger <[email protected]>
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]
Signed-off-by: John W. Linville <[email protected]>
In commit a5ffbe0, some of the calls to rtl_lps_leave() were switched
to be called from a work queue to avoid a scheduling while atomic bug.
This patch converts the remaining calls to use the work queue. In
addition, the call to rtl_lps_enter() is also switched to the work
queue. None of these newly converted calls had triggered the bug (yet),
but this change make all of them fit a single pattern.
Signed-off-by: Larry Finger <[email protected]>
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]
Signed-off-by: John W. Linville <[email protected]>
These changes add the new variables for P2P and modify the various struct
definitions for other new features.
This patch updates files base.{c,h} for the changes in the newest vendor
driver.
This patch updates files ps.{c,h} for the changes in the newest vendor
driver.
This patch updates files debug.{c,h}, efuse.c, pci.{c,h}, and wifi.h for
the changes in the newest vendor driver.
This patch updates files core.c, ps.c, rc.c, and wifi.h for
the changes in the newest vendor driver.
Signed-off-by: Larry Finger <[email protected]>
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]
Signed-off-by: John W. Linville <[email protected]>
