According to the documentation for Laird SD40 radio modules (which use
the BCM4329 chipset), the order of the Enable_BREAK_To_Host and
Pulsed_HOST_WAKE parameters in the sleep mode struct is reversed
vis-à-vis our struct declaration.  See page 46 of this PDF:

http://cdn.lairdtech.com/home/brandworld/files/Application%20Note%20-%2040%20Series%20Bluetooth.pdf

The documentation is dated Oct 2015, so fairly recent, making it appear
more likely that the documentation is correct and our code is wrong.
Amend our code to be in congruence with the documentation.

Cc: Sue White <sue.white@lairdtech.com>
Reviewed-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NLukas Wunner <lukas@wunner.de>
Signed-off-by: NMarcel Holtmann <marcel@holtmann.org>

The driver calls mdelay(15) in the ->suspend, ->resume, ->runtime_suspend
and ->runtime_resume hook, however spinning for such a long period of
time is discouraged as per Documentation/timers/timers-howto.txt.

The use of mdelay() seems unnecessary, it is allowed to sleep in the
system sleep and runtime PM hooks (with the exception of ->suspend_noirq
and ->resume_noirq) and the driver itself also does not rely on a
non-sleeping ->runtime_resume as the only place where a synchronous
resume is performed, in bcm_dequeue(), is called from a work item in
hci_ldisc.c and hci_serdev.c.

So replace the mdelay(15) with msleep(15).

Note that the delay is inserted after asserting or deasserting the
device wake pin, but in bcm_gpio_set_power() that pin is asserted or
deasserted *without* observing a delay.  It is thus unclear if the delay
is necessary at all.  It is likewise unclear why it is exactly 15 ms,
the commit introducing it, 118612fb ("Bluetooth: hci_bcm: Add
suspend/resume PM functions"), does not provide a rationale.

Cc: Frédéric Danis <frederic.danis.oss@gmail.com>
Suggested-and-reviewed-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NLukas Wunner <lukas@wunner.de>
Signed-off-by: NMarcel Holtmann <marcel@holtmann.org>

The host wake IRQ is optional, but if none is found, "BCM irq: -22" is
logged which may irritate users.  This is really a debug message, so use
dev_dbg() instead of dev_info().  If users are interested in the IRQ,
they can always consult /proc/interrupts.

Cc: Frédéric Danis <frederic.danis.oss@gmail.com>
Reviewed-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NLukas Wunner <lukas@wunner.de>
Signed-off-by: NMarcel Holtmann <marcel@holtmann.org>

Enable Bluetooth on the following Macs which provide custom ACPI methods
to toggle the GPIOs for device wake and shutdown instead of accessing
the pins directly:

    MacBook8,1     2015  12"
    MacBook9,1     2016  12"
    MacBook10,1    2017  12"
    MacBookPro13,1 2016  13"
    MacBookPro13,2 2016  13" with Touch Bar
    MacBookPro13,3 2016  15" with Touch Bar
    MacBookPro14,1 2017  13"
    MacBookPro14,2 2017  13" with Touch Bar
    MacBookPro14,3 2017  15" with Touch Bar

On the MacBook8,1 Bluetooth is muxed with a second device (a debug port
on the SSD) under the control of PCH GPIO 36.  Because serdev cannot
deal with multiple slaves yet, it is currently necessary to patch the
DSDT and remove the SSDC device.

The custom ACPI methods are called:

    BTLP (Low Power) takes one argument, toggles device wake GPIO
    BTPU (Power Up) tells SMC to drive shutdown GPIO high
    BTPD (Power Down) tells SMC to drive shutdown GPIO low
    BTRS (Reset) calls BTPD followed by BTPU
    BTRB unknown, not present on all MacBooks

Search for the BTLP, BTPU and BTPD methods on ->probe and cache them in
struct bcm_device if the machine is a Mac.

Additionally, set the init_speed based on a custom device property
provided by Apple in lieu of _CRS resources.  The Broadcom UART's speed
is fixed on Apple Macs:  Any attempt to change it results in Bluetooth
status code 0x0c and bcm_set_baudrate() thus always returns -EBUSY.
By setting only the init_speed and leaving oper_speed at zero, we can
achieve that the host UART's speed is adjusted but the Broadcom UART's
speed is left as is.

The host wake pin goes into the SMC which handles it independently
of the OS, so there's no IRQ for it.

Thanks to Ronald Tschalär who did extensive debugging and testing of
this patch and contributed fixes.

ACPI snippet containing the custom methods and device properties
(taken from a MacBook8,1):

    Method (BTLP, 1, Serialized)
    {
        If (LEqual (Arg0, 0x00))
        {
            Store (0x01, GD54) /* set PCH GPIO 54 direction to input */
        }

        If (LEqual (Arg0, 0x01))
        {
            Store (0x00, GD54) /* set PCH GPIO 54 direction to output */
            Store (0x00, GP54) /* set PCH GPIO 54 value to low */
        }
    }

    Method (BTPU, 0, Serialized)
    {
        Store (0x01, \_SB.PCI0.LPCB.EC.BTPC)
        Sleep (0x0A)
    }

    Method (BTPD, 0, Serialized)
    {
        Store (0x00, \_SB.PCI0.LPCB.EC.BTPC)
        Sleep (0x0A)
    }

    Method (BTRS, 0, Serialized)
    {
        BTPD ()
        BTPU ()
    }

    Method (_DSM, 4, NotSerialized)  // _DSM: Device-Specific Method
    {
        If (LEqual (Arg0, ToUUID ("a0b5b7c6-1318-441c-b0c9-fe695eaf949b")))
        {
            Store (Package (0x08)
                {
                    "baud",
                    Buffer (0x08)
                    { 0xC0, 0xC6, 0x2D, 0x00, 0x00, 0x00, 0x00, 0x00 },

                    "parity",
                    Buffer (0x08)
                    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },

                    "dataBits",
                    Buffer (0x08)
                    { 0x08, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },

                    "stopBits",
                    Buffer (0x08)
                    { 0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 }
                }, Local0)
            DTGP (Arg0, Arg1, Arg2, Arg3, RefOf (Local0))
            Return (Local0)
        }
        Return (0x00)
    }

Link: https://github.com/Dunedan/mbp-2016-linux/issues/29
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=110901Reported-by: NLeif Liddy <leif.liddy@gmail.com>
Cc: Mika Westerberg <mika.westerberg@linux.intel.com>
Cc: Frédéric Danis <frederic.danis.oss@gmail.com>
Cc: Loic Poulain <loic.poulain@linaro.org>
Cc: Hans de Goede <hdegoede@redhat.com>
Tested-by: Max Shavrick <mxms@me.com>                     [MacBook8,1]
Tested-by: Leif Liddy <leif.liddy@gmail.com>              [MacBook9,1]
Tested-by: Daniel Roschka <danielroschka@phoenitydawn.de> [MacBookPro13,2]
Tested-by: Ronald Tschalär <ronald@innovation.ch>         [MacBookPro13,3]
Tested-by: Peter Y. Chuang <peteryuchuang@gmail.com>      [MacBookPro14,1]
Reviewed-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NRonald Tschalär <ronald@innovation.ch>
Signed-off-by: NLukas Wunner <lukas@wunner.de>
Signed-off-by: NMarcel Holtmann <marcel@holtmann.org>

A significant portion of this driver lacks error handling.  As a first
step, add error paths to bcm_gpio_set_power(), bcm_open(), bcm_close(),
bcm_suspend_device(), bcm_resume_device(), bcm_resume(), bcm_probe() and
bcm_serdev_probe().  (I've also scrutinized bcm_suspend() but think it's
fine as is.)

Those are all the functions accessing the device wake and shutdown GPIO.
On Apple Macs the pins are accessed through ACPI methods, which may fail
for various reasons, hence proper error handling is necessary.  Non-Macs
access the pins directly, which may fail as well but the GPIO core does
not yet pass back errors to consumers.

Cc: Frédéric Danis <frederic.danis.oss@gmail.com>
Cc: Hans de Goede <hdegoede@redhat.com>
Reviewed-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NLukas Wunner <lukas@wunner.de>
Signed-off-by: NMarcel Holtmann <marcel@holtmann.org>

MacBooks provides custom ACPI methods to toggle the GPIOs for device
wake and shutdown instead of accessing the pins directly.  Prepare for
their support by adding callbacks to toggle the GPIOs, which on non-Macs
do nothing more but call gpiod_set_value().

No functional change intended.
Suggested-and-reviewed-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NLukas Wunner <lukas@wunner.de>
Signed-off-by: NMarcel Holtmann <marcel@holtmann.org>

Reviewed-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NLukas Wunner <lukas@wunner.de>
Signed-off-by: NMarcel Holtmann <marcel@holtmann.org>

If devm_request_irq() fails, the driver bails out of bcm_request_irq()
but continues to ->setup the device (because the IRQ is optional).

The driver subsequently calls devm_free_irq(), enable_irq_wake() and
disable_irq_wake() on the IRQ even though requesting it failed.

Avoid by invalidating the IRQ on request failure.

Cc: Frédéric Danis <frederic.danis.oss@gmail.com>
Reviewed-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NLukas Wunner <lukas@wunner.de>
Signed-off-by: NMarcel Holtmann <marcel@holtmann.org>

On ->setup, pm_runtime_enable() is only called if a valid IRQ was found,
but on ->close(), pm_runtime_disable() is called unconditionally.
Disablement of runtime PM is recorded in a counter, so every
pm_runtime_disable() needs to be balanced.  Fix it.

Cc: Frédéric Danis <frederic.danis.oss@gmail.com>
Reported-and-reviewed-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NLukas Wunner <lukas@wunner.de>
Signed-off-by: NMarcel Holtmann <marcel@holtmann.org>

Upon ->close, the driver powers the Bluetooth controller down, deasserts
the device wake pin, updates the runtime PM status to "suspended" and
finally frees the IRQ.

Because the IRQ is freed last, a runtime resume can take place after
the controller was powered down.  The impact is not grave, the worst
thing that can happen is that the device wake pin is reasserted (should
have no effect while the regulator is off) and that setting the runtime
PM status to "suspended" does not reflect reality.

Still, it's wrong, so free the IRQ first.

Cc: Frédéric Danis <frederic.danis.oss@gmail.com>
Reviewed-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NLukas Wunner <lukas@wunner.de>
Signed-off-by: NMarcel Holtmann <marcel@holtmann.org>

pm_runtime_disable() and pm_runtime_set_suspended() are replaced with
empty inlines if CONFIG_PM is disabled, so there's no need to #ifdef
them.

device_init_wakeup() is likewise replaced with an inline, though it's
not empty, but it and devm_free_irq() can be made conditional on
IS_ENABLED(CONFIG_PM), which is preferable to #ifdef as per section 20
of Documentation/process/coding-style.rst.

Cc: Frédéric Danis <frederic.danis.oss@gmail.com>
Reviewed-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NLukas Wunner <lukas@wunner.de>
Signed-off-by: NMarcel Holtmann <marcel@holtmann.org>

The ->close, ->suspend and ->resume hooks assume presence of a valid IRQ
if the device is wakeup capable.  However it's entirely possible that
wakeup was enabled by some other entity besides this driver and in this
case the user will get a WARN splat if no valid IRQ was found.  Avoid by
checking if the IRQ is valid, i.e. > 0.

Case in point:  On recent MacBook Pros, the Bluetooth device lacks an
IRQ (because host wakeup is handled by the SMC, independently of the
operating system), but it does possess a _PRW method (which specifies
the SMC's GPE as wake event).  The ACPI core therefore automatically
marks the physical Bluetooth device wakeup capable upon binding it to
its ACPI companion:

device_set_wakeup_capable+0x96/0xb0
acpi_bind_one+0x28a/0x310
acpi_platform_notify+0x20/0xa0
device_add+0x215/0x690
serdev_device_add+0x57/0xf0
acpi_serdev_add_device+0xc9/0x110
acpi_ns_walk_namespace+0x131/0x280
acpi_walk_namespace+0xf5/0x13d
serdev_controller_add+0x6f/0x110
serdev_tty_port_register+0x98/0xf0
tty_port_register_device_attr_serdev+0x3a/0x70
uart_add_one_port+0x268/0x500
serial8250_register_8250_port+0x32e/0x490
dw8250_probe+0x46c/0x720
platform_drv_probe+0x35/0x90
driver_probe_device+0x300/0x450
bus_for_each_drv+0x67/0xb0
__device_attach+0xde/0x160
bus_probe_device+0x9c/0xb0
device_add+0x448/0x690
platform_device_add+0x10e/0x260
mfd_add_device+0x392/0x4c0
mfd_add_devices+0xb1/0x110
intel_lpss_probe+0x2a9/0x610 [intel_lpss]
intel_lpss_pci_probe+0x7a/0xa8 [intel_lpss_pci]
Reviewed-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NRonald Tschalär <ronald@innovation.ch>
[lukas: fix up ->suspend and ->resume as well, add commit message]
Signed-off-by: NLukas Wunner <lukas@wunner.de>
Signed-off-by: NMarcel Holtmann <marcel@holtmann.org>

Commit 0395ffc1 ("Bluetooth: hci_bcm: Add PM for BCM devices")
amended this driver to request a shutdown and device wake GPIO on probe,
but mandated that only one of them need to be present:

	/* Make sure at-least one of the GPIO is defined and that
	 * a name is specified for this instance
	 */
	if ((!dev->device_wakeup && !dev->shutdown) || !dev->name) {
		dev_err(&pdev->dev, "invalid platform data\n");
		return -EINVAL;
	}

However the same commit added a call to bcm_gpio_set_power() to the
->probe hook, which unconditionally accesses *both* GPIOs.  Luckily,
the resulting NULL pointer deref was never reported, suggesting there's
no machine where either GPIO is missing.

Commit 8a920568 ("Bluetooth: hci_bcm: Add (runtime)pm support to the
serdev driver") removed the check whether at least one of the GPIOs is
present without specifying a reason.

Because commit 62aaefa7 ("Bluetooth: hci_bcm: improve use of gpios
API") refactored the driver to use devm_gpiod_get_optional() instead of
devm_gpiod_get(), one is now tempted to believe that the driver doesn't
require *any* of the two GPIOs.

Which is wrong, the driver still requires both GPIOs to avoid a NULL
pointer deref.  To this end, establish the status quo ante and request
the GPIOs with devm_gpiod_get() again.  Bail out of ->probe if either
of them is missing.

Oddly enough, whereas bcm_gpio_set_power() accesses the device wake pin
unconditionally, bcm_suspend_device() and bcm_resume_device() do check
for its presence before accessing it.  Those checks are superfluous,
so remove them.

Cc: Frédéric Danis <frederic.danis.oss@gmail.com>
Cc: Loic Poulain <loic.poulain@linaro.org>
Cc: Hans de Goede <hdegoede@redhat.com>
Cc: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Cc: Linus Walleij <linus.walleij@linaro.org>
Reviewed-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NLukas Wunner <lukas@wunner.de>
Signed-off-by: NMarcel Holtmann <marcel@holtmann.org>

Commit 7d06d589 ("Revert "Bluetooth: btusb: fix QCA...suspend/resume"")
removed the setting of the BTUSB_RESET_RESUME quirk for QCA Rome devices,
instead favoring adding USB_QUIRK_RESET_RESUME quirks in usb/core/quirks.c.

This was done because the DIY BTUSB_RESET_RESUME reset-resume handling
has several issues (see the original commit message). An added advantage
of moving over to the USB-core reset-resume handling is that it also
disables autosuspend for these devices, which is similarly broken on these.

But there are 2 issues with this approach:
1) It leaves the broken DIY BTUSB_RESET_RESUME code in place for Realtek
   devices.
2) Sofar only 2 of the 10 QCA devices known to the btusb code have been
   added to usb/core/quirks.c and if we fix the Realtek case the same way
   we need to add an additional 14 entries. So in essence we need to
   duplicate a large part of the usb_device_id table in btusb.c in
   usb/core/quirks.c and manually keep them in sync.

This commit instead restores setting a reset-resume quirk for QCA devices
in the btusb.c code, avoiding the duplicate usb_device_id table problem.

This commit avoids the problems with the original DIY BTUSB_RESET_RESUME
code by simply setting the USB_QUIRK_RESET_RESUME quirk directly on the
usb_device.

This commit also moves the BTUSB_REALTEK case over to directly setting the
USB_QUIRK_RESET_RESUME on the usb_device and removes the now unused
BTUSB_RESET_RESUME code.

BugLink: https://bugzilla.redhat.com/show_bug.cgi?id=1514836
Fixes: 7d06d589 ("Revert "Bluetooth: btusb: fix QCA...suspend/resume"")
Cc: stable@vger.kernel.org
Cc: Leif Liddy <leif.linux@gmail.com>
Cc: Matthias Kaehlcke <mka@chromium.org>
Cc: Brian Norris <briannorris@chromium.org>
Cc: Daniel Drake <drake@endlessm.com>
Cc: Kai-Heng Feng <kai.heng.feng@canonical.com>
Signed-off-by: NHans de Goede <hdegoede@redhat.com>
Signed-off-by: NMarcel Holtmann <marcel@holtmann.org>

Don't populate the const read-only array 'param' on the stack but instead
make it static. Makes the object code smaller by nearly 20 bytes:

Before:
   text	   data	    bss	    dec	    hex	filename
  11605	   2629	     64	  14298	   37da	linux/drivers/bluetooth/btintel.o

After:
   text	   data	    bss	    dec	    hex	filename
  11531	   2685	     64	  14280	   37c8	linux/drivers/bluetooth/btintel.o

(gcc version 7.2.0 x86_64)
Signed-off-by: NColin Ian King <colin.king@canonical.com>
Signed-off-by: NMarcel Holtmann <marcel@holtmann.org>

Don't populate the const read-only array 'req' on the stack but instead
make it static. Makes the object code smaller by over 40 bytes:

Before:
   text	   data	    bss	    dec	    hex	filename
   8497	   3408	    128	  12033	   2f01	linux/drivers/bluetooth/bpa10x.o

After:
   text	   data	    bss	    dec	    hex	filename
   8366	   3496	    128	  11990	   2ed6	linux/drivers/bluetooth/bpa10x.o

(gcc version 7.2.0 x86_64)
Signed-off-by: NColin Ian King <colin.king@canonical.com>
Signed-off-by: NMarcel Holtmann <marcel@holtmann.org>

Commit 27378f4c ("Bluetooth: Avoid WARN splat due to missing
GPIOLIB") amended Kconfig to select GPIOLIB if BT_HCIUART_NOKIA,
BT_HCIUART_INTEL or BT_HCIUART_BCM is enabled since all three drivers
require it to function.

The diagnosis was correct but the treatment was not.  As stated in
Documentation/gpio/consumer.txt:

    Guidelines for GPIOs consumers
    ==============================

    Drivers that can't work without standard GPIO calls should have
    Kconfig entries that depend on GPIOLIB.
                         ^^^^^^^^^
Fix it.
Reported-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NLukas Wunner <lukas@wunner.de>
Reviewed-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NMarcel Holtmann <marcel@holtmann.org>

When the hci support is built-in, but mvmem is a loadable module, we
get a link failure:

drivers/bluetooth/hci_ll.o: In function `hci_ti_probe':
hci_ll.c:(.text+0x226): undefined reference to `nvmem_cell_get'
hci_ll.c:(.text+0x238): undefined reference to `nvmem_cell_read'
hci_ll.c:(.text+0x244): undefined reference to `nvmem_cell_put'

This adds another Kconfig dependency to enforce valid configurations.

Fixes: 0e58d0cd ("Bluetooth: hci_ll: Add optional nvmem BD address source")
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Signed-off-by: NMarcel Holtmann <marcel@holtmann.org>

Device 0cf3:e010 is one of the QCA ROME family.

T:  Bus=01 Lev=01 Prnt=01 Port=13 Cnt=03 Dev#=  4 Spd=12  MxCh= 0
D:  Ver= 2.01 Cls=e0(wlcon) Sub=01 Prot=01 MxPS=64 #Cfgs=  1
P:  Vendor=0cf3 ProdID=e010 Rev=00.01
C:  #Ifs= 2 Cfg#= 1 Atr=e0 MxPwr=100mA
I:  If#= 0 Alt= 0 #EPs= 3 Cls=e0(wlcon) Sub=01 Prot=01 Driver=btusb
I:  If#= 1 Alt= 0 #EPs= 2 Cls=e0(wlcon) Sub=01 Prot=01 Driver=btusb
Signed-off-by: NAceLan Kao <acelan.kao@canonical.com>
Signed-off-by: NMarcel Holtmann <marcel@holtmann.org>

This reverts commit fd865802.

This commit causes a regression on some QCA ROME chips. The USB device
reset happens in btusb_open(), hence firmware loading gets interrupted.

Furthermore, this commit stops working after commit
("a0085f25 Bluetooth: btusb: driver to
enable the usb-wakeup feature"). Reset-resume quirk only gets enabled in
btusb_suspend() when it's not a wakeup source.

If we really want to reset the USB device, we need to do it before
btusb_open(). Let's handle it in drivers/usb/core/quirks.c.

Cc: stable@vger.kernel.org
Cc: Leif Liddy <leif.linux@gmail.com>
Cc: Matthias Kaehlcke <mka@chromium.org>
Cc: Brian Norris <briannorris@chromium.org>
Cc: Daniel Drake <drake@endlessm.com>
Signed-off-by: NKai-Heng Feng <kai.heng.feng@canonical.com>
Reviewed-by: NBrian Norris <briannorris@chromium.org>
Tested-by: NBrian Norris <briannorris@chromium.org>
Signed-off-by: NMarcel Holtmann <marcel@holtmann.org>

This driver seeks to force the Bluetooth device on for the duration of
5 seconds when the Bluetooth device has woken the host and after a
complete packet has been received.  It does that by calling:

    pm_runtime_get();
    pm_runtime_mark_last_busy();
    pm_runtime_put_autosuspend();

The same can be achieved more succinctly with:

    pm_request_resume();

That's because after runtime resuming the device, rpm_resume() invokes
pm_runtime_mark_last_busy() followed by rpm_idle(), which will cause
the device to be suspended after expiration of the autosuspend_delay.

No functional change intended.

Cc: Frédéric Danis <frederic.danis.oss@gmail.com>
Signed-off-by: NLukas Wunner <lukas@wunner.de>
Signed-off-by: NMarcel Holtmann <marcel@holtmann.org>

Loading hci_bcm with CONFIG_GPIOLIB=n results in the following splat
when calling gpiod_to_irq() from bcm_get_resources():

    WARNING: CPU: 0 PID: 1006 at ./include/linux/gpio/consumer.h:450 bcm_get_resources+0x50/0x80
    CPU: 0 PID: 1006 Comm: kworker/u8:4 Tainted: G       A         4.15.0-rc4custom+ #4
    Hardware name: Apple Inc. MacBook8,1/Mac-BE0E8AC46FE800CC, BIOS MB81.88Z.0168.B00.1708080033 08/08/2017
    Call Trace:
    bcm_serdev_probe+0x8b/0xc0
    driver_probe_device+0x202/0x310
    __driver_attach+0x85/0x90
    ? driver_probe_device+0x310/0x310
    bus_for_each_dev+0x57/0x80
    async_run_entry_fn+0x2c/0xd0
    process_one_work+0x1d2/0x3d0
    worker_thread+0x26/0x3c0
    ? process_one_work+0x3d0/0x3d0
    kthread+0x10c/0x130
    ? kthread_create_on_node+0x40/0x40
    ret_from_fork+0x1f/0x30

We could call gpiod_to_irq() only if IS_ENABLED(CONFIG_GPIOLIB) but
without GPIOLIB, the driver's power saving features can't be used,
so selecting GPIOLIB seems more appropriate.

The same issue is present in hci_intel.c and hci_nokia.c, fix those up
as well.
Reported-by: NMax Shavrick <mxms@me.com>
Signed-off-by: NLukas Wunner <lukas@wunner.de>
Signed-off-by: NMarcel Holtmann <marcel@holtmann.org>

This patch adds the 04ca:3015 (from a QCA9377 board) Bluetooth device
to the btusb blacklist and makes the kernel use the btqca module
instead of btusb. The patch is necessary because, without it the
04ca:3015 device defaults to using the btusb driver, which makes the
WIFI side of the QCA9377 board unusable (obtains 0 MBps in speedtest,
when the 04ca:3015 bluetooth is used with an audio headset).

/sys/kernel/debug/usb/devices:

    T:  Bus=01 Lev=01 Prnt=01 Port=04 Cnt=01 Dev#=  2 Spd=12   MxCh= 0
    D:  Ver= 2.01 Cls=e0(wlcon) Sub=01 Prot=01 MxPS=64 #Cfgs=  1
    P:  Vendor=04ca ProdID=3015 Rev= 0.01
    C:* #Ifs= 2 Cfg#= 1 Atr=e0 MxPwr=100mA
    I:* If#= 0 Alt= 0 #EPs= 3 Cls=e0(wlcon) Sub=01 Prot=01 Driver=btusb
    E:  Ad=81(I) Atr=03(Int.) MxPS=  16 Ivl=1ms
    E:  Ad=82(I) Atr=02(Bulk) MxPS=  64 Ivl=0ms
    E:  Ad=02(O) Atr=02(Bulk) MxPS=  64 Ivl=0ms
    I:* If#= 1 Alt= 0 #EPs= 2 Cls=e0(wlcon) Sub=01 Prot=01 Driver=btusb
    E:  Ad=83(I) Atr=01(Isoc) MxPS=   0 Ivl=1ms
    E:  Ad=03(O) Atr=01(Isoc) MxPS=   0 Ivl=1ms
    I:  If#= 1 Alt= 1 #EPs= 2 Cls=e0(wlcon) Sub=01 Prot=01 Driver=btusb
    E:  Ad=83(I) Atr=01(Isoc) MxPS=   9 Ivl=1ms
    E:  Ad=03(O) Atr=01(Isoc) MxPS=   9 Ivl=1ms
    I:  If#= 1 Alt= 2 #EPs= 2 Cls=e0(wlcon) Sub=01 Prot=01 Driver=btusb
    E:  Ad=83(I) Atr=01(Isoc) MxPS=  17 Ivl=1ms
    E:  Ad=03(O) Atr=01(Isoc) MxPS=  17 Ivl=1ms
    I:  If#= 1 Alt= 3 #EPs= 2 Cls=e0(wlcon) Sub=01 Prot=01 Driver=btusb
    E:  Ad=83(I) Atr=01(Isoc) MxPS=  25 Ivl=1ms
    E:  Ad=03(O) Atr=01(Isoc) MxPS=  25 Ivl=1ms
    I:  If#= 1 Alt= 4 #EPs= 2 Cls=e0(wlcon) Sub=01 Prot=01 Driver=btusb
    E:  Ad=83(I) Atr=01(Isoc) MxPS=  33 Ivl=1ms
    E:  Ad=03(O) Atr=01(Isoc) MxPS=  33 Ivl=1ms
    I:  If#= 1 Alt= 5 #EPs= 2 Cls=e0(wlcon) Sub=01 Prot=01 Driver=btusb
    E:  Ad=83(I) Atr=01(Isoc) MxPS=  49 Ivl=1ms
    E:  Ad=03(O) Atr=01(Isoc) MxPS=  49 Ivl=1ms
Signed-off-by: NIoan Moldovan <ioan.moldovan1999@gmail.com>
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>

In case of tx clean up, we set '-1' as budget. This means clean up until
wq is empty or till (1 << 32) pkts are cleaned. Under heavy load this
will run for long time and cause
"watchdog: BUG: soft lockup - CPU#25 stuck for 21s!" warning.

This patch sets wq clean up budget to 256.
Signed-off-by: NGovindarajulu Varadarajan <gvaradar@cisco.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

ASSERT_RTNL() macro is actual open-coded variant of WARN_ONCE() with
two exceptions. First, it prints stack for multiple hits and not only
once as WARN_ONCE() does. Second, the user can disable prints of
WARN_ONCE by setting CONFIG_BUG to N.

The multiple prints of dump stack are actually not needed, because calls
without rtnl lock are programming errors and user can't do anything
about them except to complain to the mailing list after first occurrence
of such failure.

The user who disabled BUG/WARN prints did it explicitly because by default
in upstream kernel and distributions this option is enabled. It means
that user doesn't want to see prints about missing locks too.

This patch replaces open-coded variant in favor of already existing
macro and change error prints to be once only.
Reviewed-by: NMark Bloch <markb@mellanox.com>
Signed-off-by: NLeon Romanovsky <leonro@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Remove superfluous pin setup to get out of accessing invalid I/O pin
registers because the way for pin configuring tends to be different from
various SoCs and thus it should be better being managed and controlled by
the pinctrl driver which MT7622 already can support.
Signed-off-by: NSean Wang <sean.wang@mediatek.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The property "mediatek,pctl" is only required for SoCs such as MT2701 and
MT7623, so adding a few words for stating the condition.
Signed-off-by: NSean Wang <sean.wang@mediatek.com>
Reviewed-by: NRob Herring <robh@kernel.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Lots of overlapping changes.  Also on the net-next side
the XDP state management is handled more in the generic
layers so undo the 'net' nfp fix which isn't applicable
in net-next.

Include a necessary change by Jakub Kicinski, with log message:

====================
cls_bpf no longer takes care of offload tracking.  Make sure
netdevsim performs necessary checks.  This fixes a warning
caused by TC trying to remove a filter it has not added.
Signed-off-by: NJakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: NQuentin Monnet <quentin.monnet@netronome.com>
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Pull networking fixes from David Miller"
 "What's a holiday weekend without some networking bug fixes? [1]

   1) Fix some eBPF JIT bugs wrt. SKB pointers across helper function
      calls, from Daniel Borkmann.

   2) Fix regression from errata limiting change to marvell PHY driver,
      from Zhao Qiang.

   3) Fix u16 overflow in SCTP, from Xin Long.

   4) Fix potential memory leak during bridge newlink, from Nikolay
      Aleksandrov.

   5) Fix BPF selftest build on s390, from Hendrik Brueckner.

   6) Don't append to cfg80211 automatically generated certs file,
      always write new ones from scratch. From Thierry Reding.

   7) Fix sleep in atomic in mac80211 hwsim, from Jia-Ju Bai.

   8) Fix hang on tg3 MTU change with certain chips, from Brian King.

   9) Add stall detection to arc emac driver and reset chip when this
      happens, from Alexander Kochetkov.

  10) Fix MTU limitng in GRE tunnel drivers, from Xin Long.

  11) Fix stmmac timestamping bug due to mis-shifting of field. From
      Fredrik Hallenberg.

  12) Fix metrics match when deleting an ipv4 route. The kernel sets
      some internal metrics bits which the user isn't going to set when
      it makes the delete request. From Phil Sutter.

  13) mvneta driver loop over RX queues limits on "txq_number" :-) Fix
      from Yelena Krivosheev.

  14) Fix double free and memory corruption in get_net_ns_by_id, from
      Eric W. Biederman.

  15) Flush ipv4 FIB tables in the reverse order. Some tables can share
      their actual backing data, in particular this happens for the MAIN
      and LOCAL tables. We have to kill the LOCAL table first, because
      it uses MAIN's backing memory. Fix from Ido Schimmel.

  16) Several eBPF verifier value tracking fixes, from Edward Cree, Jann
      Horn, and Alexei Starovoitov.

  17) Make changes to ipv6 autoflowlabel sysctl really propagate to
      sockets, unless the socket has set the per-socket value
      explicitly. From Shaohua Li.

  18) Fix leaks and double callback invocations of zerocopy SKBs, from
      Willem de Bruijn"

[1] Is this a trick question? "Relaxing"? "Quiet"? "Fine"? - Linus.

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (77 commits)
  skbuff: skb_copy_ubufs must release uarg even without user frags
  skbuff: orphan frags before zerocopy clone
  net: reevalulate autoflowlabel setting after sysctl setting
  openvswitch: Fix pop_vlan action for double tagged frames
  ipv6: Honor specified parameters in fibmatch lookup
  bpf: do not allow root to mangle valid pointers
  selftests/bpf: add tests for recent bugfixes
  bpf: fix integer overflows
  bpf: don't prune branches when a scalar is replaced with a pointer
  bpf: force strict alignment checks for stack pointers
  bpf: fix missing error return in check_stack_boundary()
  bpf: fix 32-bit ALU op verification
  bpf: fix incorrect tracking of register size truncation
  bpf: fix incorrect sign extension in check_alu_op()
  bpf/verifier: fix bounds calculation on BPF_RSH
  ipv4: Fix use-after-free when flushing FIB tables
  s390/qeth: fix error handling in checksum cmd callback
  tipc: remove joining group member from congested list
  selftests: net: Adding config fragment CONFIG_NUMA=y
  nfp: bpf: keep track of the offloaded program
  ...

Sven Eckelmann says:

====================
flow_dissector: Provide basic batman-adv unicast handling

we are currently starting to use batman-adv as mesh protocol on multicore
embedded devices. These usually don't have a lot of CPU power per core but
are reasonable fast when using multiple cores.

It was noticed that sending was working very well but receiving was
basically only using on CPU core per neighbor. The reason for that is
format of the (normal) incoming packet:

  +--------------------+
  | ip(v6)hdr          |
  +--------------------+
  | inner ethhdr       |
  +--------------------+
  | batadv unicast hdr |
  +--------------------+
  | outer ethhdr       |
  +--------------------+

The flow dissector will therefore stop after parsing the outer ethernet
header and will not parse the actual ipv(4|6)/... header of the packet. Our
assumption was now that it would help us to add minimal support to the flow
dissector to jump over the batman-adv unicast and inner ethernet header
(like in gre ETH_P_TEB). The patch was implemented in a slightly hacky
way [1] and the results looked quite promising.

I didn't get any feedback how the files should actually be named. So I am
now just using the names from RFC v3

The discussion of the RFC v3 can be found in the related patches of
https://patchwork.ozlabs.org/cover/849345/

The discussion of the RFC v2 can be found in the related patches of
https://patchwork.ozlabs.org/cover/844783/

Changes in v4:
==============

* added  patch to change the u8/u16 to __u8/__u16 in
  include/uapi/linux/batadv_packet.h
  - requested by Willem de Bruijn <willemdebruijn.kernel@gmail.com>

Changes in v3:
==============

* removed change of uapi/linux/batman_adv.h to uapi/linux/batadv_genl.h
  - requested by Willem de Bruijn <willemdebruijn.kernel@gmail.com>
* removed naming fixes for enums/defines in uapi/linux/batadv_genl.h
  - requested by Willem de Bruijn <willemdebruijn.kernel@gmail.com>
* renamed uapi/linux/batadv.h to uapi/linux/batadv_packet.h
* moved batadv dissector functionality in own function
  - requested by Tom Herbert <tom@herbertland.com>
* added support for flags FLOW_DISSECTOR_F_STOP_AT_ENCAP and
  FLOW_DIS_ENCAPSULATION
  - requested by Willem de Bruijn <willemdebruijn.kernel@gmail.com>

Changes in v2:
==============

* removed the batman-adv unicast packet header definition from flow_dissector.c
* moved the batman-adv packet.h/uapi headers around to provide the correct
  definitions to flow_dissector.c
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The batman-adv unicast packets contain a full layer 2 frame in encapsulated
form. The flow dissector must therefore be able to parse the batman-adv
unicast header to reach the layer 2+3 information.

  +--------------------+
  | ip(v6)hdr          |
  +--------------------+
  | inner ethhdr       |
  +--------------------+
  | batadv unicast hdr |
  +--------------------+
  | outer ethhdr       |
  +--------------------+

The obtained information from the upper layer can then be used by RPS to
schedule the processing on separate cores. This allows better distribution
of multiple flows from the same neighbor to different cores.
Signed-off-by: NSven Eckelmann <sven.eckelmann@openmesh.com>
Reviewed-by: NJiri Pirko <jiri@mellanox.com>
Acked-by: NWillem de Bruijn <willemb@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The header file is used by different userspace programs to inject packets
or to decode sniffed packets. It should therefore be available to them as
userspace header.

Also other components in the kernel (like the flow dissector) require
access to the packet definitions to be able to decode ETH_P_BATMAN ethernet
packets.
Signed-off-by: NSven Eckelmann <sven.eckelmann@openmesh.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The uapi headers use the __u8/__u16/... version of the fixed width types
instead of u8/u16/... The use of the latter must be avoided before
packet.h is copied to include/uapi/linux/.
Signed-off-by: NSven Eckelmann <sven.eckelmann@openmesh.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The BIT(x) macro is no longer available for uapi headers because it is
defined outside of it (linux/bitops.h). The use of it must therefore be
avoided and replaced by an appropriate other representation.
Signed-off-by: NSven Eckelmann <sven.eckelmann@openmesh.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The headers used by packet.h should also be included by it directly. main.h
is currently dealing with it in batman-adv, but this will no longer work
when this header is moved to include/uapi/linux/.
Signed-off-by: NSven Eckelmann <sven.eckelmann@openmesh.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Bert Kenward says:

====================
sfc: support extra stats on Medford2

X2000-series NICs add port stats for two new features: FEC (Forward Error
 Correction, used on 25G links) and CTPIO (cut-through programmed I/O).
This patch series adds support for reporting both of these sets of stats

v2: add additional Signed-off-by
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

While the Linux driver doesn't use CTPIO ('cut-through programmed I/O'),
 other drivers on the same port might, so if we're responsible for
 reporting per-port stats we need to include the CTPIO stats.
Signed-off-by: NBert Kenward <bkenward@solarflare.com>
Signed-off-by: NEdward Cree <ecree@solarflare.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

There's no explicit capability bit, so we just condition them on having
 efx->num_mac_stats >= MC_CMD_MAC_NSTATS_V2.
Signed-off-by: NEdward Cree <ecree@solarflare.com>
Signed-off-by: NBert Kenward <bkenward@solarflare.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Medford2 NICs support more than MC_CMD_MAC_NSTATS stats, and report the new
 count in a field of MC_CMD_GET_CAPABILITIES_V4.  This also means that the
 end generation count moves (it is, as before, the last 64 bits of the DMA
 buffer, but that is no longer MC_CMD_MAC_GENERATION_END).
So read num_mac_stats from the GET_CAPABILITIES response, if present;
 otherwise assume MC_CMD_MAC_NSTATS; and always use num_mac_stats - 1 rather
 than MC_CMD_MAC_GENERATION_END.
Signed-off-by: NEdward Cree <ecree@solarflare.com>
Signed-off-by: NBert Kenward <bkenward@solarflare.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NEdward Cree <ecree@solarflare.com>
Signed-off-by: NBert Kenward <bkenward@solarflare.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>