staging: rtl8723bs: fix rtw_cfg80211_monitor_if_xmit_entry()'s return type
The method ndo_start_xmit() is defined as returning an 'netdev_tx_t',
which is a typedef for an enum type, but the implementation in this
driver returns an 'int'.
Fix this by returning 'netdev_tx_t' in this driver too.
Signed-off-by: Luc Van Oostenryck <luc.vanoostenryck@gmail.com> Reviewed-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
staging: visornic: fix visornic_xmit()'s return type
The method ndo_start_xmit() is defined as returning an 'netdev_tx_t',
which is a typedef for an enum type, but the implementation in this
driver returns an 'int'.
Fix this by returning 'netdev_tx_t' in this driver too.
Signed-off-by: Luc Van Oostenryck <luc.vanoostenryck@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
staging: wlan-ng: fix p80211knetdev_hard_start_xmit()'s return type
The method ndo_start_xmit() is defined as returning an 'netdev_tx_t',
which is a typedef for an enum type, but the implementation in this
driver returns an 'int'.
Fix this by returning 'netdev_tx_t' in this driver too.
Signed-off-by: Luc Van Oostenryck <luc.vanoostenryck@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
staging/wilc1000: fix wilc_mac_xmit()'s return type
The method ndo_start_xmit() is defined as returning an 'netdev_tx_t',
which is a typedef for an enum type, but the implementation in this
driver returns an 'int'.
Fix this by returning 'netdev_tx_t' in this driver too.
Signed-off-by: Luc Van Oostenryck <luc.vanoostenryck@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Removed the items from WILC1000 TODO list, which are already addressed
to keep it updated. The removed items are already taken care by
previously submitted patches.
staging: wilc1000: remove registering of ndo_do_ioctl callback
Remove registering of ndo_do_ioctl in wilc_netdev_ops structure.
As RSSI information is already avaliable through cfg80211, so remove
the IOCTL call use to fetch the same information. Wext support is not
present in the driver. Its also done to address the TODO list item
mentioned below:
"use wext-core handling instead of private SIOCSIWPRIV implementation"
staging: wilc1000: remove multiple define for mac connect and disconnect
Cleanup patch to have commonly used macro in common header file to avoid
same defination in mulitple file. Removed MAC_CONNECTED &
MAC_DISCONNECTED macro from coreconfigurator.h header.
staging: wilc1000: remove multiple define used for MAX_SSID_LEN
Cleanup patch to have commonly used macro in common header file to avoid
same defination in mulitple file. Removed MAX_SSID_LEN macro from
coreconfigurator.h header as its already defined in wilc_wlan_if.h.
staging: wilc1000: remove unused macros in wilc module
Cleanup patch to remove the macros which are defined by not used.
Below mentioned macros are removed:
SCAN_DONE
SCAN_EVENT_DONE_ABORTED
WILC_WFI_RX_INTR
WILC_WFI_TX_INTR
WILC_WFI_TIMEOUT
WILC_WFI_DWELL_PASSIVE
WILC_WFI_DWELL_ACTIVE
MAX_SURVEY_RESULT_FRAG_SIZE
SURVEY_RESULT_LENGTH
NUM_BASIC_SWITCHES
NUM_FHSS_SWITCHES
NUM_11N_BASIC_SWITCHES
NUM_11N_HUT_SWITCHES
BA_SESSION_DEFAULT_BUFFER_SIZE
BA_SESSION_DEFAULT_TIMEOUT
BLOCK_ACK_REQ_SIZE
staging: wilc1000: handle error condition in add_key() and remove auth_type variable
Added the code to return correct error code in add_key() and also removed
'auth_type' variable. Now passing diretly to function instead of using
the 'auth_type' variable.
staging: wilc1000: split add_key() to avoid line over 80 chars
Cleanup changes to fix 'line over 80 chars' issue found by checkpatch.pl
script by spliting the function. Also make use of kzalloc() instead
kmalloc().
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.
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'.
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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
Marc Dietrich [Mon, 23 Apr 2018 15:17:51 +0000 (17:17 +0200)]
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>
Yangbo Lu [Mon, 23 Apr 2018 03:55:00 +0000 (11:55 +0800)]
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.
Roy Pledge [Tue, 27 Mar 2018 18:13:31 +0000 (14:13 -0400)]
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>
Jia-Ju Bai [Tue, 10 Apr 2018 12:25:34 +0000 (20:25 +0800)]
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>
Jia-Ju Bai [Tue, 10 Apr 2018 12:25:09 +0000 (20:25 +0800)]
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>
Jia-Ju Bai [Tue, 10 Apr 2018 12:24:23 +0000 (20:24 +0800)]
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>
Jia-Ju Bai [Tue, 10 Apr 2018 12:23:54 +0000 (20:23 +0800)]
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>
Jia-Ju Bai [Tue, 10 Apr 2018 12:23:29 +0000 (20:23 +0800)]
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>
Jia-Ju Bai [Tue, 10 Apr 2018 12:23:00 +0000 (20:23 +0800)]
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>
projects / mirror_ubuntu-kernels.git / log