We fixed this to use free_netdev() instead of kfree() but unfortunately
free_netdev() doesn't accept NULL pointers. Smatch complains about
this, it's not something I discovered through testing.
Fixes: 3030d40b50 ('staging: vt6655: use free_netdev instead of kfree')
Fixes: 0a438d5b38 ('staging: vt6656: use free_netdev instead of kfree')
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
space after '*' in pointer declaration is not preferred in linux
and results in checkpatch error. Hence removed the spaces present
after the '*' in pointer declarations to fix the checkpatch errors
Signed-off-by: Nandini Hanumanthagowda <nandu.hgowda@gmail.com>
Reviewed-by: Lisa Nguyen <lisa@xenapiadmin.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This patch fixes a code indentation error found by checkpatch.pl
where a line was indented with spaces instead of tabs
Signed-off-by: Johannes Löthberg <johannes@kyriasis.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This patch fixes the following type of sparse warnings:
drivers/staging/vt6656/hostap.c:688:42: warning: cast from restricted gfp_t
drivers/staging/vt6656/hostap.c:688:42: warning: incorrect type in argument 2 (different base types)
drivers/staging/vt6656/hostap.c:688:42: expected restricted gfp_t [usertype] flags
drivers/staging/vt6656/hostap.c:688:42: got int [signed] <noident>
Signed-off-by: Ebru Akagunduz <ebru.akagunduz@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This patch fixes a brace warning in power.c found by checkpatch.pl
Signed-off-by: Johannes Löthberg <johannes@kyriasis.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
removed braces from if-control flow structs whenever it
contained only one statement to comply with the coding style
Signed-off-by: Nandini Hanumanthagowda <nandu.hgowda@gmail.com>
Reviewed-by: Lisa Nguyen <lisa@xenapiadmin.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Fixed the indentation errors and removed unnecessary space
warnings to comply with coding style
Signed-off-by: Nandini Hanumanthagowda <nandu.hgowda@gmail.com>
Reviewed-by: Lisa Nguyen <lisa@xenapiadmin.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The USB reset occurs after netdev registration if network manager
calls device_open too fast causing USB fails in main_usb.c: device_init_registers.
Move the usb reset to before register_netdev.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Removed the checkpatch warning of line over 80 chars
by breaking the long line into sensible chunks of 2 lines
to comply with coding style
Signed-off-by: Nandini Hanumanthagowda <nandu.hgowda@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Linux style for comment is C89 style "/* */" and it
doesn't prefer C99-style comment "//...". Hence replaced
C99-style comments used in code by C89 style comment to
comply with linux coding style
Signed-off-by: Nandini Hanumanthagowda <nandu.hgowda@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
There was parentheses around return statement's value which
was not required since return statement is not a function.
Hence removed the parentheses to eliminate the checkpatch error
which states:
ERROR: return is not a function, parentheses are not required
Signed-off-by: Nandini Hanumanthagowda <nandu.hgowda@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Removed braces under if-else control flow statement whenever
there is only one statement under if-else control statement
to comply with linux coding style
Signed-off-by: Nandini Hanumanthagowda <nandu.hgowda@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Removed unnecessary white spaces at beginning of line
and added proper indentation to fix checkpatch warnings/errors
to improve the readability of code
Signed-off-by: Nandini Hanumanthagowda <nandu.hgowda@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Set the two TX fall back rates in s_bPacketToWirelessUsb and
pass to private area of driver in variables tx_rate_fb0
and tx_rate_fb1 from the wFB_Opt0/wFB_Opt1 array.
Apply these rates were needed in the TX structure and
remove byFBOption settings in s_uGetRTSCTSDuration.
This greatly simplifies s_uGetRTSCTSDuration and
more future flexibility of setting rates from
upper levels of driver.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Removed whitespace related errors and warnings by removing
unnecessary spaces before new line in the quoted string
and spaces at start of any line or before tab to comply
with the coding style.
Signed-off-by: Nandini Hanumanthagowda <nandu.hgowda@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Removed spaces around operators where it was not required
as per coding style
Added spaces around operators where it was necessary as
per coding style
Signed-off-by: Nandini Hanumanthagowda <nandu.hgowda@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Fixes occasional urb submitted while active.
Even thought the fMP_CONTROL_WRITES/fMP_CONTROL_READS flags
are cleared in the return context urb->hcpriv is not NULL.
check for hcpriv and return STATUS_FAILURE if not NULL.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Remove sTimerTxData 10 second timer which triggers PSbSendNullPacket
every 40 seconds when bLinkPass == true.
Move the 40 second timer to the existing BSSvSecondCallBack
one second delayed workqueue and trigger every 40 seconds when
bLinkPass == true.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Return from work queues on flag fMP_DISCONNECTED to prevent
any scheduling threads past closing of device.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Correct goto of patch
staging: vt6656: make pControlURB available life time of driver.
Which should free_netdev.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
There was one line which was added twice in the
definition of iwctl_handler and hence was redundant.
It was defined at line nos 1798 and 1818. Removed
the code at line 1818.
The redundant code was:
IW_HANDLER(SIOCSIWMLME, iwctl_siwmlme)
which was defined twice.
Signed-off-by: Nandini Hanumanthagowda <nandu.hgowda@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
There is no need remove and add the control urb in device open/close.
Move to the probe and disconnect. This make the USB control in/out
functions always available to driver.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Remove white space and commented out code.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
uFillDataHead in is now dead so remove and pvTxDataHd.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Add no rts/cts vnt_tx_datahead_ab to union vnt_tx_data_head.
The need_mic now applies to all modes of packet type PK_TYPE_11B.
and return vnt_rxtx_datahead_ab in s_vGenerateTxParameter
This removes pvTxDataHd pointer assignment and calls to s_uFillDataHead.
Since pvTxDataHd is now null a small vendor hack to vDMA0_tx_80211/csMgmt_xmit
to do with Disassociation-packet is also changed correct head position.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Parse struct vnt_tx_datahead_ab from s_uFillDataHead and
move to struct vnt_rts_ab structure returning duration down
functions.
Both are returned in s_uFillDataHead at this stage as pvTxDataHd
is only null in s_bPacketToWirelessUsb.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Add no rts/cts vnt_tx_datahead_a_fb to union vnt_tx_data_head.
The need_mic now applies to all modes of packet type PK_TYPE_11A.
and return vnt_rxtx_datahead_a_fb in s_vGenerateTxParameter
This removes pvTxDataHd pointer assignment.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Parse struct vnt_tx_datahead_g from s_uFillDataHead and
move to struct vnt_rts_g/struct vnt_cts structure.
This removes pvTxDataHd pointer assignment.
The duration_id(uDuration) field is now returned in s_vGenerateTxParameter.
The return of s_uFillDataHead temporary or'ed to retain functionality of
other functions and is removed at the end of this patch series.
Since pvTxDataHd is now null a small vendor hack to vDMA0_tx_80211/csMgmt_xmit
to do with Disassociation-packet is also changed correct head position.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
s_vGenerateTxParameter, s_vFillCTSHead, s_vFillRTSHead, vnt_rxtx_rts* functions
to return u16
Initially this is the duration_id field for ieee80211_hdr
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Tasklet to workqueue.
RxMngWorkItem -> rx_mng_work_item
Reduce atomic area of driver and dependency on system timer.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
timer to delay workqueue.
sTimerSecondCallback -> second_callback_work
The delayed work queue is declared in device.h
This timer is very heavy on the system.
Improves over performance of driver and reduce the atomic
area of driver.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
timer to delay workqueue.
sTimerCommand -> run_command_work
sTimerCommand is very heavy on the system timer.
Improves over performance of driver and reduce the atomic
area of driver.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Create double pointer to pass back vnt_mic_hdr address position
of relevant structure if need_mic is more than 0. Otherwise it
is NULL.
This replaces the pMICHDR void pointer with struct vnt_mic_hdr
and assignments in the calling functions.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Create struct vnt_tx_mic_hdr for with union vnt_tx_data_head
for mic mode.
Create union vnt_tx for mic and non mic mode.
Attach vnt_rrv_time_* to relevant tx_head.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Attach union vnt_tx_head to structure vnt_tx_buffer replacing pointer
pvRrvTime.
In s_vGenerateTxParameter the relevant struct vnt_rrv_time_*
structure is attached to the vnt_tx_head.
The NULL check is now pFifoHead.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Don't allow entry to iwctl_siwencodeext if device not open.
This fixes a race condition where wpa supplicant/network manager
enters the function when the device is already closed.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Cc: stable@vger.kernel.org # 3.8+
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The vt6656 is prone to resetting on the usb bus.
It seems there is a race condition and wpa supplicant is
trying to open the device via iw_handlers before its actually
closed at a stage that the buffers are being removed.
The device is longer considered open when the
buffers are being removed. So move ~DEVICE_FLAGS_OPENED
flag to before freeing the device buffers.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
There seems to be race condition that the device is ndo_start_xmit
at a point where the device is closing and apTD is NULL resulting
in dead lock.
Add a NULL check to apTD and return NULL to calling functions.
This is more likely on 64 bit systems.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The variables bMultiBand and bKeyTableFull are only assigned the
values true and false. Change its type to bool.
The simplified semantic patch that find this problem is as
follows (http://coccinelle.lip6.fr/):
@exists@
type T;
identifier b;
@@
- T
+ bool
b = ...;
... when any
b = \(true\|false\)
Signed-off-by: Peter Senna Tschudin <peter.senna@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Malcolm Priestley
Dan Carpenter
Nandini Hanumanthagowda
Johannes Löthberg
Ebru Akagunduz
Ashvini Varatharaj
Greg Kroah-Hartman
Peter Senna Tschudin