While testing Fabio's patch I hit lockdep warning about possible
deadlock:
[ 252.063305] WARNING: possible recursive locking detected
[ 252.063642] 5.14.0+ #9 Tainted: G C
[ 252.063946] --------------------------------------------
[ 252.064282] ip/335 is trying to acquire lock:
[ 252.064560] ffff888009ebad28 (pmutex){+.+.}-{4:4}, at: usbctrl_vendorreq+0xc5/0x4a0 [r8188eu]
[ 252.065168]
[ 252.065168] but task is already holding lock:
[ 252.065536] ffffffffc021b3b8 (pmutex){+.+.}-{4:4}, at: netdev_open+0x3a/0x5f [r8188eu]
[ 252.066085]
[ 252.066085] other info that might help us debug this:
[ 252.066494] Possible unsafe locking scenario:
[ 252.066494]
[ 252.066866] CPU0
[ 252.067025] ----
[ 252.067184] lock(pmutex);
[ 252.067367] lock(pmutex);
There is one problem with this warning: there is no pmutex in this
driver, *BUT* all mutexes are initialized via private _rtw_mutex_init
API, which had struct mutex *pmutex argument.
So, all mutexes in this driver had same name in lockdep map. Of course,
lockdep will complain about any nested locking.
Fix it by open-coding _rtw_mutex_{init,free} wrappers, because we do not
need them at all.
Cc: Fabio M. De Francesco <fmdefrancesco@gmail.com>
Signed-off-by: Pavel Skripkin <paskripkin@gmail.com>
Link: https://lore.kernel.org/r/20210904124747.30050-1-paskripkin@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
smatch says:
usb_intf.c:326 rtw_hw_suspend() warn: variable dereferenced before check 'padapter' (see line 323)
usb_intf.c:387 rtw_hw_resume() warn: variable dereferenced before check 'padapter' (see line 385)
There is only one caller of rtw_hw_suspend() and it does not check
padapter pointer, so let's just omit this check to make smatch happy.
Signed-off-by: Pavel Skripkin <paskripkin@gmail.com>
Link: https://lore.kernel.org/r/20210905204218.19317-1-paskripkin@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Remove rtw_hal_c2h_id_filter_ccx from hal/hal_intf.c and its one caller
from core/rtw_cmd.c. This function is a wrapper function which returns
the c2h_id_filter_ccx function pointer of struct hal_ops unconditionally.
As this function pointer is never set, and the function call's return
value is subsequently called inside an if condition, this could lead to
an attempt to deference a NULL pointer, which would crash the driver.
Signed-off-by: Phillip Potter <phil@philpotter.co.uk>
Link: https://lore.kernel.org/r/20210906010106.898-14-phil@philpotter.co.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Remove Efuse_PgPacketWrite_BT function pointer from struct hal_ops, as it
is unused.
Signed-off-by: Phillip Potter <phil@philpotter.co.uk>
Link: https://lore.kernel.org/r/20210906010106.898-13-phil@philpotter.co.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Remove static Efuse_PgPacketWrite_BT and its one caller from
core/rtw_efuse.c. This function is a wrapper function which calls
the Efuse_PgPacketWrite_BT function pointer of struct hal_ops
unconditionally. As this function pointer is never set, and this
function call is possible to reach, this could lead to an attempt to
deference a NULL pointer, which would crash the driver.
Signed-off-by: Phillip Potter <phil@philpotter.co.uk>
Link: https://lore.kernel.org/r/20210906010106.898-12-phil@philpotter.co.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Remove hal_xmitframe_enqueue function pointer from struct hal_ops, as it
is unused.
Signed-off-by: Phillip Potter <phil@philpotter.co.uk>
Link: https://lore.kernel.org/r/20210906010106.898-11-phil@philpotter.co.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Remove rtw_hal_xmitframe_enqueue from hal/hal_intf.c and its one
caller from core/rtw_recv.c, and remove its declaration from
include/hal_intf.h as well. This is just a wrapper function that calls
the function pointer hal_xmitframe_enqueue in struct hal_ops if it
is set, which it never is.
Signed-off-by: Phillip Potter <phil@philpotter.co.uk>
Link: https://lore.kernel.org/r/20210906010106.898-10-phil@philpotter.co.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Remove rtw_hal_interrupt_handler from hal/hal_intf.c, and remove its
declaration from include/hal_intf.h as well. This is just a wrapper
function that calls the function pointer interrupt_handler in struct
hal_ops if it is set, which it never is. In addition, this wrapper
function is unused anyway.
Signed-off-by: Phillip Potter <phil@philpotter.co.uk>
Link: https://lore.kernel.org/r/20210906010106.898-8-phil@philpotter.co.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Remove rtw_hal_disable_interrupt from hal/hal_intf.c, and remove its
declaration from include/hal_intf.h as well. This is just a wrapper
function that calls the function pointer disable_interrupt in struct
hal_ops if it is set, which it never is. In addition, this wrapper
function is unused anyway.
Signed-off-by: Phillip Potter <phil@philpotter.co.uk>
Link: https://lore.kernel.org/r/20210906010106.898-6-phil@philpotter.co.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Remove rtw_hal_enable_interrupt from hal/hal_intf.c, and remove its
declaration from include/hal_intf.h as well. This is just a wrapper
function that calls the function pointer enable_interrupt in struct
hal_ops if it is set, which it never is. In addition, this wrapper
function is unused anyway.
Signed-off-by: Phillip Potter <phil@philpotter.co.uk>
Link: https://lore.kernel.org/r/20210906010106.898-4-phil@philpotter.co.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Remove hal_reset_security_engine function pointer from struct hal_ops,
as it is unused.
Reviewed-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Phillip Potter <phil@philpotter.co.uk>
Link: https://lore.kernel.org/r/20210906010106.898-3-phil@philpotter.co.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Remove rtw_hal_reset_security_engine and its one caller from
hal/hal_intf.c, and remove its declaration from include/hal_intf.h as
well. This is just a wrapper function that calls the function pointer
hal_reset_security_engine in struct hal_ops if it is set, which it never
is.
Signed-off-by: Phillip Potter <phil@philpotter.co.uk>
Link: https://lore.kernel.org/r/20210906010106.898-2-phil@philpotter.co.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Remove condition which never get execute.
as pattrib->mdata is always zero before if condition check.
Acked-by: Phillip Potter <phil@philpotter.co.uk>
Acked-by: Michael Straube <straube.linux@gmail.com>
Signed-off-by: Saurav Girepunje <saurav.girepunje@gmail.com>
Link: https://lore.kernel.org/r/YTNrjG1y05ByN54+@user
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Remove GetHalODMVarHandler from hal_ops ant remove its wrapper
rtw_hal_get_odm_var(), both are unused.
Acked-by: Phillip Potter <phil@philpotter.co.uk>
Signed-off-by: Michael Straube <straube.linux@gmail.com>
Link: https://lore.kernel.org/r/20210904124033.14244-6-straube.linux@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Function ReadChipVersion8188E() is only called by the wrapper
rtl8188e_read_chip_version(). Rename ReadChipVersion8188E()
to rtl8188e_read_chip_version() and change the return type to
void since the return value is never used.
Acked-by: Phillip Potter <phil@philpotter.co.uk>
Signed-off-by: Michael Straube <straube.linux@gmail.com>
Link: https://lore.kernel.org/r/20210904124033.14244-5-straube.linux@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
From enum hardware_type only HARDWARE_TYPE_RTL8188EU is used in the
driver. Remove the enum and define HARDWARE_TYPE_RTL8188EU.
Signed-off-by: Michael Straube <straube.linux@gmail.com>
Link: https://lore.kernel.org/r/20210903093354.12645-7-straube.linux@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The IS_HARDWARE_TYPE_8188* macros are not needed because the driver
is for RTL8188EU chips, remove them.
Signed-off-by: Michael Straube <straube.linux@gmail.com>
Link: https://lore.kernel.org/r/20210903093354.12645-6-straube.linux@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
All functions declared in rtw_ioctl_rtl.h are not implemented.
Remove rtw_ioctl_rtl.h and its includes.
Signed-off-by: Michael Straube <straube.linux@gmail.com>
Link: https://lore.kernel.org/r/20210903093354.12645-2-straube.linux@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Add a missing blank line after declarations in rtw_macaddr_cfg()
to clear a checkpatch warning.
WARNING: Missing a blank line after declarations
Signed-off-by: Michael Straube <straube.linux@gmail.com>
Link: https://lore.kernel.org/r/20210830121508.6411-8-straube.linux@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Use ether_addr_copy() in rtw_macaddr_cfg() to copy the mac address.
The buffers are properly aligned.
Signed-off-by: Michael Straube <straube.linux@gmail.com>
Link: https://lore.kernel.org/r/20210830121508.6411-7-straube.linux@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Use a random default mac address instead of a fixed one to reduce the
likelihood of mac address collision.
Signed-off-by: Michael Straube <straube.linux@gmail.com>
Link: https://lore.kernel.org/r/20210830121508.6411-6-straube.linux@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Use is_broadcast_ether_addr() and is_zero_ether_addr() in
rtw_macaddr_cfg(). The buffer is properly aligned.
Signed-off-by: Michael Straube <straube.linux@gmail.com>
Link: https://lore.kernel.org/r/20210830121508.6411-5-straube.linux@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Use ETH_ALEN instead of hard-coding the value for the mac_addr buffer
in struct eeprom_priv.
Signed-off-by: Michael Straube <straube.linux@gmail.com>
Link: https://lore.kernel.org/r/20210830121508.6411-4-straube.linux@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Ensure the mac address buffer in struct eeprom_priv is properly
aligned for use with functions from <linux/etherdevice.h>.
Signed-off-by: Michael Straube <straube.linux@gmail.com>
Link: https://lore.kernel.org/r/20210830121508.6411-3-straube.linux@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Instead of a custom approach use mac_pton() to convert the mac
address string. With mac_pton() the driver also validates the
input now.
Signed-off-by: Michael Straube <straube.linux@gmail.com>
Link: https://lore.kernel.org/r/20210830121508.6411-2-straube.linux@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
IC type is always 8188E. Remove ICType from struct HAL_VERSION,
all macros that check it and related code.
Signed-off-by: Michael Straube <straube.linux@gmail.com>
Link: https://lore.kernel.org/r/20210830092153.9283-4-straube.linux@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The local variable rfPath in Hal_SetAntennaPathPower() and
Hal_SetTxPower() is set but not used. Remove the variable
and related code.
Signed-off-by: Michael Straube <straube.linux@gmail.com>
Link: https://lore.kernel.org/r/20210830092153.9283-3-straube.linux@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
