staging: wilc1000: remove line over 80 char in cfg_connect_result()

Fix 'line over 80 characters' issues reported by checkpatch.pl script in
cfg_connect_result().

Signed-off-by: Ajay Singh <ajay.kathat@microchip.com>
Reviewed-by: Claudiu Beznea <claudiu.beznea@microchip.com>
Reviewed-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: wilc1000: fix line over 80 char issue in clear_shadow_scan()

Remove 'line over 80 char' issue found by checkpatch.pl script.

Signed-off-by: Ajay Singh <ajay.kathat@microchip.com>
Reviewed-by: Claudiu Beznea <claudiu.beznea@microchip.com>
Reviewed-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: wilc1000: rename hAgingTimer to avoid camelCase issue

Fix 'Avoid camelCase' issue found by checkpatch.pl script.

Signed-off-by: Ajay Singh <ajay.kathat@microchip.com>
Reviewed-by: Claudiu Beznea <claudiu.beznea@microchip.com>
Reviewed-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: wilc1000: refactor mgmt_tx to fix line over 80 chars

Refactor mgmt_tx() to fix line over 80 characters issue. Split the
function to avoid the checkpatch.pl warning. Returning the same error
code in case of memory allocation failure.

Signed-off-by: Ajay Singh <ajay.kathat@microchip.com>
Reviewed-by: Claudiu Beznea <claudiu.beznea@microchip.com>
Reviewed-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: wilc1000: rename WILC_WFI_p2p_rx & s32Freq to avoid camelCase

Fix 'Avoid camelCase' issue found by checkpatch.pl script.

Signed-off-by: Ajay Singh <ajay.kathat@microchip.com>
Reviewed-by: Claudiu Beznea <claudiu.beznea@microchip.com>
Reviewed-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: wilc1000: refactor WILC_WFI_p2p_rx() to avoid line over 80 char

Fix 'line over 80 characters' issue found by checkpatch.pl script.
Refactor and split the function to avoid the checkpatch reported issues.

Signed-off-by: Ajay Singh <ajay.kathat@microchip.com>
Reviewed-by: Claudiu Beznea <claudiu.beznea@microchip.com>
Reviewed-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: wilc1000: remove line over 80 char warnings in set_wiphy_params()

Fix 'line over 80 character' issue reported by checkpatch.pl script in
set_wiphy_params(). Directly used the 'wiphy' pointer received as
function argument instead of using 'priv->dev->ieee80211_ptr->wiphy'.

Signed-off-by: Ajay Singh <ajay.kathat@microchip.com>
Reviewed-by: Claudiu Beznea <claudiu.beznea@microchip.com>
Reviewed-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: wilc1000: removed unused static variables for gtk and ptk information

Removed the unnecessary static variables used to store gtk and ptk
information. Key data stored in the params was never access using these
variables.

Variables given below are removed

g_add_gtk_key_params;
g_key_gtk_params;
g_add_ptk_key_params;
g_key_ptk_params;
g_key_wep_params;
g_ptk_keys_saved;
g_gtk_keys_saved;
g_wep_keys_saved;

Signed-off-by: Ajay Singh <ajay.kathat@microchip.com>
Reviewed-by: Claudiu Beznea <claudiu.beznea@microchip.com>
Reviewed-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: wilc1000: refactor scan() to free kmalloc memory on failure cases

Added changes to free the allocated memory in scan() for error condition.
Also added 'NULL' check validation before accessing allocated memory.
Copied the SSID information in consecutive slots to avoid inbetween
holes while filling into array.

Signed-off-by: Ajay Singh <ajay.kathat@microchip.com>
Reviewed-by: Claudiu Beznea <claudiu.beznea@microchip.com>
Reviewed-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: refactor hostif_sme_set_pmksa function

This commits refactor a bit hostif_sme_set_pmksa function:
 - avoid one level indentation changing if condition.
 - use ether_addr_copy to copy bssid addresses.
 - move 'i' initialization to declaration place.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: use ether_addr_copy in hostif_sme_multicast_set

Use ether_addr_copy to copy ethernet addresses in function
hostif_sme_multicast_set instead of memcpy.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: use ether_addr_equal in hostif_data_request

Use ether_addr_equal to compare addresses in ether_addr_equal
function instead of comparing using memcmp.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: use ether_addr_equal in hostif_data_indication

Instead of comparing ethernet address using memcmp in function
hostif_data_indication use ether_addr_equal function created
for this function.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: use ether_addr_copy in hostif_mib_get_confirm

This commit improves readability changing custom ethernet
addresses copies in favour of using ether_addr_copy() function.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: use ether_addr_copy() instead of custom copy

In order to achieve ethernet address copies, ether_addr_copy()
function exists. So just use it and avoid the byte by byte copy.
This increase readability.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: remove duplicated ks_wlan_handler_def declaration

This declaration is declared twice so just remove this one because
the other one is the one which contains static struct initializers.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: remove WPS definition conditional code

WPS definition was defined by default in ks_wlan.h header
file. So it makes no sense to have conditional preprocessor
stuff along the code about this.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: fix line exceding 80 characters in ks_wlan_get_range

Avoid very long if condition just changing its style. This makes
checkpatch script not complains about this line.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: refactor ks_wlan_get_mode function

Avoid the use of switch-case block which is not necessary
at all and just use a ternary operator to achieve this.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: refactor ks_wlan_set_mode function

Most cases which are being handled in the switch-case of
ks_wlan_set_mode function are just returning EINVAL. Avoid
the use of switch-case stament and just use a simple if
to handle those. This decrease LOC as well as improves
readability.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: remove some duplicated definitions in ks_wlan_net.c

This definitions are in linux/wireless.h header so it is not
necessary at all to have this compatibility stuff duplicated here.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: avoid use of brackets in single line if

Singles if does not need at all to use brackets in its body
so just remove them which is the preferred style.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: refactor hostif_sme_power_mgmt_set function

Some minor changes have been done in this function to clean
it a bit:
    - POWER_MGMT_ACTIVE and default case are the same so
      just handle that with the same block of code,
    - POWER_MGMT_SAVE1 replaces if-else with a ternary operator.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: use definitions from kernel headers in hostif_data_indication function

Function hostif_data_indication checks some hardcoded values in a
switch-case block. This values are defined in uapi/linux/llc.h
header. Just use them and avoid a comment in the code improving
readability a bit.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: join some debug traces in get_current_ap function

There is a lot of netdev_dbg calls related with the access point
in this function and all of them can be joined in only one call.
This makes code a bit simplier.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: avoid some if-else code in get_current_ap function

This commits avoid some if-else code extracting common code before
its use and making use of a new variable 'size' which is assigned
using a ternary operator. This improves readability.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: remove KSC_OPNOTSUPP related code

This commit reviews KSC_OPNOTSUPP related code. The
preprocessor KSC_OPNOTSUPP is defined by default so
related wext functions are not being used. Just clean
code removing all of this stuff.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: use IW_HANDLER macro in ks_wlan_handler

This commit make use of IW_HANDLER to set wext operations
of the device. Using this, comments are not neccessary anymore
and also NULL entries so readability is clearly increased.
In order to avoid casting because of the use of a different
prototype in all related functions, those which are affected
have been updated also to make use of the  union iwreq_data
as third parameter updating code accordly.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: review includes of ks_hostif file

This commit reviews includes of ks_hostif.c source file.
Those which are not being used at all have been removed.
Driver header includes have been moved after the kernel
header includes to make style consistent in different
files of the driver. The need of ks_wlan.h header include
is only because of some preprocessor conditional code
of WPS definition. This definition is in ks_wlan.h
so this one must be included before the ks_hostif.h header
file.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: removes data_buff field of ks_wlan_private struct

This commit removes data_buff array field of ks_wlan_private
which is not being used at all.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: remove two fields of ks_wlan_private struct

This commit removes l2_dev and l2_fd fields of ks_wlan_private
struct because they are not being used at all.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: avoid blank line between definitions in hostif_data_request

This commit removes a blank line between definition in
hostif_data_request function.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: fix warning aout long line in init_request

This commit fix length of the definition line of init_request
function. Warning from checkpatch script for this is fixed.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: remove superfluous comments in ks_hostif source file

This commit removes some comments which are not necessary at all
because code is clear enough to understand its intention.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: avoid one level indentation in devio_rec_ind function

This commit changes logic to handle with the status of the device
at first checking for close state to return directly instead
of just do the stuff when device is open. This improves readability
avoiding one level indentation.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: change name and type for device_open_status field

This commit changes type for device_open_status field of ks_wlan_private
structure from int to bool. This variable is only be set to 1
on ks_wlan_net_start and set to 0 on ks_wlan_net_stop. For this
purpose it is not necessary at all to use an integer because a bool
is enough. This also renames field name from device_open_status to
is_device_open.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: assign dev_alloc_name() result to variable before check it

This commit assigns dev_alloc_name() call to 'ret' variable to
check it after instead of check directly the call in the if
condition. This improves a bit readability. It also add an empty
line before the new assignment to separate it from the previous
check statement block.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: use u8 instead of unsigned char for firmware buffers

This commit replaces type unsigned char which is the one which
is being used for firmware buffers with u8 type which is preferred.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: fix error paths in ks7010_sdio_remove function

This commit reviews and fixes error paths in ks7010_sdio_remove
driver function. It change logic to handle error directly
after priv dereference to avoid one level indentation. It also
removes a temporal netdev variable which wasn't being used in all
of the function calls. Also if send_stop_request call fails it
was making a direct 'return' instead of doing a properly cleaning.
Because of this a new 'err_free_card' label has been added.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: check sdio_set_block_size return value

This commit checks sdio_set_block_size function return value.
If it fails abort driver initialization.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: replace create_workqueue with alloc_workqueue

This commit replaces deprecated create_workqueue call with the
alloc_workqueue one which is the one to be used now for this
purpose.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: move MODULE_DEVICE_TABLE related code

This commit moves MODULE_DEVICE_TABLE related code to the end of
the file. This is not necessary at all but moving it just before
its use improves readability.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: avoid one extra level indentation in ks_wlan_hw_rx function

This commit use an and operator in a if condition to avoid one
indentation level which is not needed at all.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: review debug and error messages in ks7010_sdio source

This commit reviews debug and error messages in code located
in ks7010_sdio source file avoiding to use 'error' or 'ks7010'
because this file is using netdev_* functions and has non
sense to repeat information in log messages.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: review comment style in ks7010_sdio source file

This commit reviews comment style used in ks7010_sdio source file
in order to be coherent with the rest of the code. Most comments
in this source are before definitions but only two of them have
been written at the right. So, be coherent moving this two to the
top of definitions. Also fix one multiline comment style to use
the normal preferred kernel style.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: add REG suffix to sdio register definitions

This commit adds REG suffix to register definitions related
with SDIO in order to improve readability.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: delete not used definitions in ks7010_sdio source

This commit removes two definitions inside ks7010_sdio source file
because they are not being used at all.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: move sdio specific register definitions into source file

This commit moves SDIO related register definitions from header
to source file. There is no need to have those into the header
because they are only being used in specific SDIO code.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: move ROM_FILE definition into source file

This commit moves ROM_FILE from header to source file because
there is not being used outside this.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: ks7010: avoid casts in michael_mic_function calls

This commit removes casts in calls to michael_mic_function.
Most of them are nosense because types match perfectly function
parameters. To avoid also int casting for len parameter just
pass unsigned len to function which makes sense because is only
being called with unsigned int len parameters.

Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: mt7621-mmc: Fix calculation typo in msdc_init_gpd_bd

When refactoring the function msdc_init_gpd_bd, an error in the
calculation of the bd.next address was introduced. The offset has to
be added to the base address, not multiplied.

Fixes: 8f2395586cf0 ("staging: mt7621-mmc: Refactor msdc_init_gpd_bd")
Reported-by: NeilBrown <neil@brown.name>
Signed-off-by: Christian Lütke-Stetzkamp <christian@lkamp.de>
Tested-by: NeilBrown <neil@brown.name>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: mt7621-mmc: Remove unused card_workqueue from msdc_host

The card_workqueue field of msdc_host is already if 0'd out and there
are no references to it in the code (not even in unused code), so
remove it.

Signed-off-by: Christian Lütke-Stetzkamp <christian@lkamp.de>
Reviewed-by: NeilBrown <neil@brown.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: mt7621-mmc: Remove old references to tasklet

In the current code there are '#if 0' out references on using a
tasklet instead of delayed_work. Removing these improves readability.

Signed-off-by: Christian Lütke-Stetzkamp <christian@lkamp.de>
Reviewed-by: NeilBrown <neil@brown.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: mt7621-mmc: Add annotations about held locks

The functions msdc_command_resp and msdc_do_request are always called
with the host->lock lock held. By adding annotations, sparse is
informed about that.

Signed-off-by: Christian Lütke-Stetzkamp <christian@lkamp.de>
Reviewed-by: NeilBrown <neil@brown.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: mt7621-mmc: Correct datatypes for io and sanitize io access

Current code discard the address space information on the base address
of the mmc controller, that causes sparse warnings. It uses the raw
read write function, that is correct for the mips architecture (little
endian), but for portability the non-raw function should be used. Also
the clear/set bit macros do direct memory access, that is also correct
for mips, but not portable.

So the type of the base address is changed to void __iomem *, that is
the type returned by the ioremap function. The set/clear bit macros
are changed to functions, that use the portable read and write
functions. The use of the raw access functions is changed to use the
non-raw ones.

Signed-off-by: Christian Lütke-Stetzkamp <christian@lkamp.de>
Reviewed-by: NeilBrown <neil@brown.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: mt7621-mmc: Replace dma dir with mmc_get_dma_dir

Currently the dma direction is manually determined by the read status,
there is a more portable function for it, mmc_get_dma_dir, use it.

Signed-off-by: Christian Lütke-Stetzkamp <christian@lkamp.de>
Reviewed-by: NeilBrown <neil@brown.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: mt7621-mmc: Fix dma_map_sg may map to fewer entries

The dma_map_sg function may merge several sglist entries into one, the
return value has to be saved to consider that. The data->sg_count
field is the position, where it should be saved, like it is done in
other mmc host drivers. Also the count of mapped entries is needed for
the dma setup.

Signed-off-by: Christian Lütke-Stetzkamp <christian@lkamp.de>
Reviewed-by: NeilBrown <neil@brown.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: mt7621-mmc: Remove multiple assignments

Fix checkpatch: multiple assignments should be avoided, to improve
readability.
It aslo moves the second assignment out of the if/else block, that is
valid, because drv_mode is of type msdc_mode, an enum with only those
three elements, so one of the if/ else if statements is always
taken. And the second assignment can happen after the conditions.

Signed-off-by: Christian Lütke-Stetzkamp <christian@lkamp.de>
Reviewed-by: NeilBrown <neil@brown.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: mt7621-mmc: Fix dereference before check in msdc_drv_pm

In the msdc_drv_pm function the variable mmc is dereferenced before
checked. Reordering fixes that.

Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Christian Lütke-Stetzkamp <christian@lkamp.de>
Reviewed-by: NeilBrown <neil@brown.name>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Revert "staging: nvec: Augment TODO file with GPIO work item"

This reverts commit 5a3e59ce3acf2f52c8b4a6fc656c3b1e4a329f06, because the
necessary changes were applied in the "staging: nvec: convert to use GPIO
descriptors" patch.

Signed-off-by: Marc Dietrich <marvin24@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

MAINTAINERS: add maintainer for the DPAA2 PTP clock driver

This patch is to add maintainer for the DPAA2 PTP clock driver.

Signed-off-by: Yangbo Lu <yangbo.lu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: fsl-dpaa2/rtc: add rtc driver

This patch is to add driver for the DPAA2 1588 timer module (RTC)
which interfaces to up to an unlimited number of 10/100/1000 or
10G ethernet MACs, providing current time, alarm, and fiper support.
The 1588 IP control block includes these distinctive features.

- External GPIO trigger for time-stamping
- 2 Time-stamp alarms
- 3 FIPER pulse generators
- Phase adjusted output timer clock

Currently this driver only supports basic functions like
settime/gettime/adjtime/adjfreq.

Signed-off-by: Yangbo Lu <yangbo.lu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

drivers/staging/fsl-mc: Fix DPIO error path issue

Remove unneeded call to dev_set_drvdata(dev, NULL) since the device
will be deallocated in any case.

Signed-off-by: Roy Pledge <roy.pledge@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

drivers/staging/fsl-mc: Use devm_memremap/devm_ioremap for DPIO

Change the mapping of the QBMan cache enabled area from using ioremap_wc()
to devm_memremap(). This allows the __iomem attribute to be removed from
the pointer (which makes sense as accesses treat this as cacheable memory
not IO memory).  These changes allow sparse checks to pass.

Also use devm_ioremap() for the cache inhibited area so unmap occurs
automatically when the device is released.

Signed-off-by: Roy Pledge <roy.pledge@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: rtl8192u: Replace mdelay with usleep_range in dm_TXPowerTrackingCallback_TSSI

dm_TXPowerTrackingCallback_TSSI() is never called in atomic context.

dm_TXPowerTrackingCallback_TSSI() is only called by
dm_txpower_trackingcallback(), which is set a parameter of
INIT_DELAYED_WORK() in rtl8192_init_priv_task().

Despite never getting called from atomic context,
dm_TXPowerTrackingCallback_TSSI() calls mdelay() to busily wait.
This is not necessary and can be replaced with usleep_range() to
avoid busy waiting.

This is found by a static analysis tool named DCNS written by myself.
And I also manually check it

Signed-off-by: Jia-Ju Bai <baijiaju1990@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: rtl8192u: Replace mdelay with usleep_range in rtl8192_usb_disconnect

rtl8192_usb_disconnect() is never called in atomic context.

rtl8192_usb_disconnect() is only set as ".disconnect" in
struct usb_driver.

Despite never getting called from atomic context,
rtl8192_usb_disconnect() calls mdelay() to busily wait.
This is not necessary and can be replaced with usleep_range() to
avoid busy waiting.

This is found by a static analysis tool named DCNS written by myself.
And I also manually check it.

Signed-off-by: Jia-Ju Bai <baijiaju1990@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: rtl8192u: Replace GFP_ATOMIC with GFP_KERNEL in ieee80211_softmac_init

ieee80211_softmac_init() is never called in atomic context.

The call chains ending up at ieee80211_softmac_init() is:
[1] ieee80211_softmac_init() <- alloc_ieee80211_rsl() <-
    rtl8192_usb_probe()

rtl8192_usb_probe() is set as ".probe" in struct usb_driver.

Despite never getting called from atomic context,
ieee80211_softmac_init() calls kzalloc() with GFP_ATOMIC,
which does not sleep for allocation.
GFP_ATOMIC is not necessary and can be replaced with GFP_KERNEL,
which can sleep and improve the possibility of sucessful allocation.

This is found by a static analysis tool named DCNS written by myself.
And I also manually check it.

Signed-off-by: Jia-Ju Bai <baijiaju1990@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: rtl8192u: Replace GFP_ATOMIC with GFP_KERNEL in prism2_wep_init

prism2_wep_init() is never called in atomic context.

prism2_wep_init() is only set as ".init" in
struct ieee80211_crypto_ops.
The call chains ending up at "->init" function are:
[1] ->init() <- ieee80211_wpa_set_encryption() <-
    ieee80211_wpa_supplicant_ioctl()
[2] ->init() <- ieee80211_wx_set_encode_ext_rsl() <-
    r8192_wx_set_enc_ext()
[3] ->init() <- ieee80211_wx_set_encode_rsl() <-
    r8192_wx_set_enc()

ieee80211_wpa_supplicant_ioctl(), r8192_wx_set_enc_ext() and
r8192_wx_set_enc() call mutex_lock(), which indicates these functions
are not called in atomic context.

Despite never getting called from atomic context,
prism2_wep_init() calls kzalloc() with GFP_ATOMIC,
which does not sleep for allocation.
GFP_ATOMIC is not necessary and can be replaced with GFP_KERNEL,
which can sleep and improve the possibility of sucessful allocation.

This is found by a static analysis tool named DCNS written by myself.
And I also manually check it

Signed-off-by: Jia-Ju Bai <baijiaju1990@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: rtl8192u: Replace GFP_ATOMIC with GFP_KERNEL in ieee80211_tkip_init

ieee80211_tkip_init() is never called in atomic context.

ieee80211_tkip_init() is only set as ".init" in
struct ieee80211_crypto_ops.
The call chains ending up at "->init" function are:
[1] ->init() <- ieee80211_wpa_set_encryption() <-
    ieee80211_wpa_supplicant_ioctl()
[2] ->init() <- ieee80211_wx_set_encode_ext_rsl() <-
    r8192_wx_set_enc_ext()
[3] ->init() <- ieee80211_wx_set_encode_rsl() <-
    r8192_wx_set_enc()

ieee80211_wpa_supplicant_ioctl(), r8192_wx_set_enc_ext() and
r8192_wx_set_enc() call mutex_lock(), which indicates these functions
are not called in atomic context.

Despite never getting called from atomic context,
ieee80211_tkip_init() calls kzalloc() with GFP_ATOMIC,
which does not sleep for allocation.
GFP_ATOMIC is not necessary and can be replaced with GFP_KERNEL,
which can sleep and improve the possibility of sucessful allocation.

This is found by a static analysis tool named DCNS written by myself.
And I also manually check it.

Signed-off-by: Jia-Ju Bai <baijiaju1990@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: rtl8192u: Replace GFP_ATOMIC with GFP_KERNEL in ieee80211_ccmp_init

ieee80211_ccmp_init() is never called in atomic context.

ieee80211_ccmp_init() is only set as ".init" in
struct ieee80211_crypto_ops.
The call chains ending up at "->init" function are:
[1] ->init() <- ieee80211_wpa_set_encryption() <-
ieee80211_wpa_supplicant_ioctl()
[2] ->init() <- ieee80211_wx_set_encode_ext_rsl() <-
r8192_wx_set_enc_ext()
[3] ->init() <- ieee80211_wx_set_encode_rsl() <-
r8192_wx_set_enc()

ieee80211_wpa_supplicant_ioctl(), r8192_wx_set_enc_ext() and
r8192_wx_set_enc() call mutex_lock(), which indicates these functions
are not called in atomic context.

Despite never getting called from atomic context,
ieee80211_ccmp_init() calls kzalloc() with GFP_ATOMIC,
which does not sleep for allocation.
GFP_ATOMIC is not necessary and can be replaced with GFP_KERNEL,
which can sleep and improve the possibility of sucessful allocation.

This is found by a static analysis tool named DCNS written by myself.
And I also manually check it.

Signed-off-by: Jia-Ju Bai <baijiaju1990@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: rtl8192e: Fixed coding style around a plus sign

Added blank spaces between a plus sign in order to standardize the coding
style.

Signed-off-by: Joao Dalben <jhdalben@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: comedi: cb_pcidas64: fix alignment of function parameters

Fix most checkpatch.pl issues of type:

CHECK: Alignment should match open parenthesis

Signed-off-by: Gabriel Francisco Mandaji <gfmandaji@gmail.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: vc04_services: Remove depends on HAS_DMA in case of platform dependency

Remove dependencies on HAS_DMA where a Kconfig symbol depends on another
symbol that implies HAS_DMA, and, optionally, on "|| COMPILE_TEST".
In most cases this other symbol is an architecture or platform specific
symbol, or PCI.

Generic symbols and drivers without platform dependencies keep their
dependencies on HAS_DMA, to prevent compiling subsystems or drivers that
cannot work anyway.

This simplifies the dependencies, and allows to improve compile-testing.

Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
Reviewed-by: Mark Brown <broonie@kernel.org>
Acked-by: Robin Murphy <robin.murphy@arm.com>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: vchiq_core: Free kthreads in error case

We need to free the kthreads in error case before leaving
vchiq_init_state() otherwise we leak resources.

Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: vchiq_core: Move all wake-ups to one point

Move all calls of wake_up_process to one point, whichs makes the
following implementation of clean-up code easier.

Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: vchiq_arm: Avoid long udelay

vchiq_initialise() is used in non-interrupt context, so we can
replace udelay with usleep_range as suggested by timers-howto.txt.

Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: vchiq_arm: Fix multiline dereferences

Multiline dereferences aren't nice to review. So fix this checkpatch
warning.

Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: vc04_services: Use __func__ macro

It's better to use the __func__ macro instead of open-code the function
name. This fixes the following checkpatch warning:

WARNING: Prefer using '"%s...", __func__' to using 'x',
this function's name, in a string

Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: vchiq_core: remove BITSET_ZERO

Hiding memset behind a macro isn't the best, because it relies on that
the parameter is not a pointer. Luckily all user has been removed, so
we can remove BITSET_ZERO too.

Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: vchiq_core: Remove stackhog in process_free_queue

This removes the stackhog in process_free_queue by allocating the
necessary memory within the recycle thread main function instead
of the stack.

Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: vchiq_core: Fix missing semaphore release in error case

The bail out branch in case of a invalid tx_pos missed a semaphore
release. Dan Carpenter found this with a static checker.

Fixes: d1eab9dec610 ("staging: vchiq_core: Bail out in case of invalid tx_pos")
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: vboxvideo: Fix line size exceeding 80 columns.

This patch fixes the checkpatch.pl warning:

WARNING: line over 80 characters
+               vbva = (void __force *)vbox->vbva_buffers + i *
VBVA_MIN_BUFFER_SIZE;

Signed-off-by: Daniel Junho <djunho@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: wilc1000: Remove unnecessary braces {} around single statement block

Remove unnecessary braces {} around an 'if' statement block with a single
statement. Issue found by checkpatch.

Signed-off-by: Eyal Ilsar <edilsar@gmail.com>
Reviewed-by: Claudiu Beznea <claudiu.beznea@microchip.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Staging: most: Move comments to the end of line

Signed-off-by: Ian Liu Rodrigues <ian.liu88@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: pi433: Fixed typos and grammar in documentation

Some typos and grammar issues were found in the documentation. These
mistakes were fixed.

Signed-off-by: Shannon Booth <shannon.ml.booth@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: pi433: use help in Kconfig description

Fixes checkpatch warning:

  WARNING: prefer 'help' over '---help---' for new help texts

Signed-off-by: Valentin Vidic <Valentin.Vidic@CARNet.hr>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: pi433: break long lines in rf69.c

Breaks long lines in rf69.c, fixing checkpatch.pl warnings:
"WARNING: line over 80 characters"

Signed-off-by: Simon Sandström <simon@nikanor.nu>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: pi433: break long lines

Breaks long lines in rf69.h, fixing checkpatch.pl warnings:
"WARNING: line over 80 characters"

Signed-off-by: Simon Sandström <simon@nikanor.nu>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: pi433: cleanup tx_fifo locking

pi433_write requires locking due to multiple writers.  After acquiring
the lock check if enough free space is available in the kfifo to write
the whole message. This check should prevent partial writes to tx_fifo
so kfifo_reset is not needed anymore.

pi433_tx_thread is the only reader so it does not require locking
after kfifo_reset is removed.

Signed-off-by: Valentin Vidic <Valentin.Vidic@CARNet.hr>
Reviewed-by: Marcus Wolf <marcus.wolf@wolf-entwicklungen.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: greybus: simplify getting .drvdata

We should get drvdata from struct device directly. Going via
platform_device is an unneeded step back and forth.

Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Acked-by: Johan Hovold <johan@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: greybus: Fix warning to limit chars per line

Wrap comment to fix warning "prefer a maximum 75 chars per line"

Signed-off-by: Gaurav Dhingra <gauravdhingra.gxyd@gmail.com>
Acked-by: Vaibhav Agarwal <vaibhav.sr@gmail.com>
Reviewed-by: Mark Greer <mgreer@animalcreek.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: mt7621-mmc: Remove unused field dma_addr of msdc_host

The dma_addr field of msdc_host is never used, so remove it.

Signed-off-by: Christian Lütke-Stetzkamp <christian@lkamp.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: mt7621-mmc: Remove unused field cmd_rsp_done of msdc_host

The cmd_rsp_done field of msdc_host is never used, so remove it.

Signed-off-by: Christian Lütke-Stetzkamp <christian@lkamp.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: mt7621-mmc: Remove unused field cmd_r1b_done in msdc_host

The field cmd_r1b_done in msdc_host is unused, so remove it.

Signed-off-by: Christian Lütke-Stetzkamp <christian@lkamp.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: mt7621-mmc: Remove unused field dma_left_size of msdc_host

The dma_left_size field of msdc_host is never used, remove it.

Signed-off-by: Christian Lütke-Stetzkamp <christian@lkamp.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: mt7621:mmc: Remove unused field reserved from msdc_host

The reserved field of msdc_host is never used and msdc_host is also
never exposed to hardware, so remove that field.

Signed-off-by: Christian Lütke-Stetzkamp <christian@lkamp.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: mt7621-mmc: Remove unused field starttime from msdc_host

The starttime field of msdc_host is never used, remove it.

Signed-off-by: Christian Lütke-Stetzkamp <christian@lkamp.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: mt7621-mmc: Change default transfer mode to DMA

The current default transfer is to use DMA or not depending on the
size of the data. The upstream driver mtk-sd uses DMA all times,
change the standard mode here to DMA for testing, if there are any
performance problems with DMA for small data sizes. If not, the option
for transfer mode should be removed in the future,

Signed-off-by: Christian Lütke-Stetzkamp <christian@lkamp.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: mt7621-mmc: Refactor msdc_init_gpd_bd

The msdc_init_gpd_bd function is currently hard to read, because of
old, commented out code and a while loop, where a for loop is much
easier to read. Refactor it to make the code more readable.

Signed-off-by: Christian Lütke-Stetzkamp <christian@lkamp.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: mt7621-mmc: Remove unused field xfersz from msdc_dma

The xfersz field of msdc_dma is only set to host->xfer_size and this
is also available at the (few) positions, where dma.xfersz is used, so
it is removed.

Signed-off-by: Christian Lütke-Stetzkamp <christian@lkamp.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>