of_match_table was not filled which prevents device to be
instantiated from device tree node.
Signed-off-by: NLukasz Gemborowski <lukasz.gemborowski@nokia.com>
Reviewed-by: NAlexander Sverdlin <alexander.sverdlin@nokia.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

With a i2c topology like the following

                       GPIO ---|  ------ BAT1
                        |      v /
   I2C  -----+----------+---- MUX
             |                   \
           EEPROM                 ------ BAT2

there is a locking problem with the GPIO controller since it is a client
on the same i2c bus that it muxes. Transfers to the mux clients (e.g. BAT1)
will lock the whole i2c bus prior to attempting to switch the mux to the
correct i2c segment. In the above case, the GPIO device is an I/O expander
with an i2c interface, and since the GPIO subsystem knows nothing (and
rightfully so) about the lockless needs of the i2c mux code, this results
in a deadlock when the GPIO driver issues i2c transfers to modify the
mux.

So, observing that while it is needed to have the i2c bus locked during the
actual MUX update in order to avoid random garbage on the slave side, it
is not strictly a must to have it locked over the whole sequence of a full
select-transfer-deselect mux client operation. The mux itself needs to be
locked, so transfers to clients behind the mux are serialized, and the mux
needs to be stable during all i2c traffic (otherwise individual mux slave
segments might see garbage, or worse).

Introduce this new locking concept as "mux-locked" muxes, and call the
pre-existing mux locking scheme "parent-locked".

Modify the i2c mux locking so that muxes that are "mux-locked" locks only
the muxes on the parent adapter instead of the whole i2c bus when there is
a transfer to the slave side of the mux. This lock serializes transfers to
the slave side of the muxes on the parent adapter.

Add code to i2c-mux-gpio and i2c-mux-pinctrl that checks if all involved
gpio/pinctrl devices have a parent that is an i2c adapter in the same
adapter tree that is muxed, and request a "mux-locked mux" if that is the
case.

Modify the select-transfer-deselect code for "mux-locked" muxes so
that each of the select-transfer-deselect ops locks the mux parent
adapter individually.
Signed-off-by: NPeter Rosin <peda@axentia.se>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

smatch rightfully says:

drivers/i2c/muxes/i2c-mux-pinctrl.c:175 i2c_mux_pinctrl_probe() warn: inconsistent indenting
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>
Acked-by: NPeter Rosin <peda@axentia.se>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

Allocate an explicit i2c mux core to handle parent and child adapters
etc. Update the select/deselect ops to be in terms of the i2c mux core
instead of the child adapter.
Signed-off-by: NPeter Rosin <peda@axentia.se>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

Allocate an explicit i2c mux core to handle parent and child adapters
etc. Update the select/deselect ops to be in terms of the i2c mux core
instead of the child adapter.

Add a mask to handle the case where not all child adapters should
cause a mux deselect to happen, now that there is a common deselect op
for all child adapters.
Signed-off-by: NPeter Rosin <peda@axentia.se>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

Allocate an explicit i2c mux core to handle parent and child adapters
etc. Update the select/deselect ops to be in terms of the i2c mux core
instead of the child adapter.
Signed-off-by: NPeter Rosin <peda@axentia.se>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

Allocate an explicit i2c mux core to handle parent and child adapters
etc. Update the select/deselect ops to be in terms of the i2c mux core
instead of the child adapter.
Signed-off-by: NPeter Rosin <peda@axentia.se>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

Allocate an explicit i2c mux core to handle parent and child adapters
etc. Update the select/deselect ops to be in terms of the i2c mux core
instead of the child adapter.
Signed-off-by: NPeter Rosin <peda@axentia.se>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

Allocate an explicit i2c mux core to handle parent and child adapters
etc. Update the select/deselect ops to be in terms of the i2c mux core
instead of the child adapter.
Signed-off-by: NPeter Rosin <peda@axentia.se>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

sysfs attributes should use the same format for reads and writes,
rather than pretty-printing on read.

* Make the "cur_master" attribute read back as just the name of the
  master
* Expose the list of all masters as a read-only "available_masters"
  attribute, using space separators as in similar attributes of other
  devices

Also, spell out "cur_master" in full as "current_master".

Fixes: 50a5ba87 ("i2c: mux: demux-pinctrl: add driver")
Signed-off-by: NBen Hutchings <ben.hutchings@codethink.co.uk>
Tested-by: NWolfram Sang <wsa+renesas@sang-engineering.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

This driver allows an I2C bus to switch between multiple masters. This
is not hot-switching because connected I2C slaves will be
re-instantiated. It is meant to select the best I2C core at runtime once
the task is known. Example: Prefer i2c-gpio over another I2C core
because of HW errata affecting your use case.
Signed-off-by: NWolfram Sang <wsa+renesas@sang-engineering.com>
Acked-by: NRob Herring <robh@kernel.org>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

Reading the register (if allowed) after writing is to ensure writing
is completed on a posted bus. The endianness of reading doesn't matter.
Signed-off-by: NYork Sun <yorksun@freescale.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

With CONFIG_OF=n, the following warnings occur when compiling
drivers/i2c/muxes/i2c-mux-reg.c:

  CC      drivers/i2c/muxes/i2c-mux-reg.o
drivers/i2c/muxes/i2c-mux-reg.c:170:13: warning: ‘struct gpiomux’ declared inside parameter list [enabled by default]
      struct platform_device *pdev)
             ^
drivers/i2c/muxes/i2c-mux-reg.c:170:13: warning: its scope
is only this definition or declaration, which is probably not what you want [enabled by default]
drivers/i2c/muxes/i2c-mux-reg.c: In function ‘i2c_mux_reg_probe’:
drivers/i2c/muxes/i2c-mux-reg.c:201:3: warning: passing argument 1 of ‘i2c_mux_reg_probe_dt’ from incompatible pointer type [enabled by default]
   ret = i2c_mux_reg_probe_dt(mux, pdev);
   ^
Signed-off-by: NMike Rapoport <mike.rapoport@gmail.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

Checking was done at three different locations, just do it once and
properly at probing time.
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>
Tested-by: NYork Sun <yorksun@freescale.com>

Every call of of_parse_phandle() increments user count of found device
node, if OF_DYNAMIC is enabled.

The change fixes all similar addressed cases in drivers/i2c.
Signed-off-by: NVladimir Zapolskiy <vz@mleia.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

Based on i2c-mux-gpio driver, similarly the register-based mux
switch from one bus to another by setting a single register.
The register can be on PCIe bus, local bus, or any memory-mapped
address. The endianness of such register can be specified in device
tree if used, or in platform data.
Signed-off-by: NYork Sun <yorksun@freescale.com>
Acked-by: NAlexander Sverdlin <alexander.sverdlin@nokia.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

This change is needed to properly lock I2C parent bus driver.

Prior to this change i2c_put_adapter() is misused, which may lead
to an overflow over zero of I2C bus driver user counter.
Signed-off-by: NVladimir Zapolskiy <vladimir_zapolskiy@mentor.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

i2c_driver does not need to set an owner because i2c_register_driver()
will set it.
Signed-off-by: NKrzysztof Kozlowski <k.kozlowski@samsung.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

pca9541 and pca954x are calling master_xfer() of the parent adapter directly
thus bypassing the quirks checks of the adapter. Use __i2c_transfer() instead.
Signed-off-by: NAlexander Sverdlin <alexander.sverdlin@nokia.com>
Tested-by: NŁukasz Gemborowski <lukasz.gemborowski@nokia.com>
Acked-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: NJisheng Zhang <jszhang@marvell.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>
Cc: stable@kernel.org
Fixes: b7f62584 ("i2c: add quirk checks to core")

The GPIO subsystem provides dummy GPIO consumer functions if GPIOLIB is
not enabled. Hence drivers that depend on GPIOLIB, but use GPIO consumer
functionality only, can still be compiled if GPIOLIB is not enabled.

Relax the dependency on GPIOLIB if COMPILE_TEST is enabled, where
appropriate.
Signed-off-by: NGeert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

Probe deferral is not an error case. It happens only when
the necessary dependencies are not there yet.

The driver core is already printing a message when a driver
requests probe deferral, so this can be traced in the logs
without these error prints.

This patch removes the error messages for these deferral cases.
Signed-off-by: NIonut Nicu <ioan.nicu.ext@nokia.com>
Acked-by: NAlexander Sverdlin <alexander.sverdlin@nokia.com>
Acked-by: NPeter Korsgaard <peter.korsgaard@barco.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

Since 39b2bbe3 (gpio: add flags argument to gpiod_get*() functions)
which appeared in v3.17-rc1, the gpiod_get* functions take an additional
parameter that allows to specify direction and initial value for
outputs.

Also there is an *_optional variant that serves well here.  The sematics
is slightly changed here by using it. Now if a reset gpio is specified
and getting hold on it fails, pca954x_probe fails, too.
Signed-off-by: NUwe Kleine-König <u.kleine-koenig@pengutronix.de>
Acked-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

Add i2c-mux-idle-disconnect device tree property to PCA954x mux driver. The new
property forces the multiplexer to disconnect child buses in idle state. This is
used, for example, when there are several multiplexers on the same bus and the
devices on the underlying buses might have same I2C addresses.
Signed-off-by: NAlexander Sverdlin <alexander.sverdlin@nsn.com>
[wsa: added a newline]
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

A platform_driver does not need to set an owner, it will be populated by the
driver core.
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

pca954x may be power lost during suspend, so after resume we also suffer
the issue fixed by commit cd823db8,

 "pca954x power-on default is channel 0 connected. If multiple pca954x
 muxes are connected to the same physical I2C bus, the parent bus will
 see channel 0 devices behind both muxes by default."

What's more, when resume bootloader may also operate the mux, so the
the channel connected after that may not be the one driver thought.

We fix this problem by putting the mux to disconnected state and
clearing last_chan in the resume hook.
Signed-off-by: NJisheng Zhang <jszhang@marvell.com>
Reviewed-by: NJean Delvare <jdelvare@suse.de>
Acked-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

This driver causes the following randconfig build error:

drivers/i2c/muxes/i2c-mux-pca954x.c: In function ‘pca954x_probe’:
drivers/i2c/muxes/i2c-mux-pca954x.c:204:2: error: implicit declaration
of function ‘devm_gpiod_get’ [-Werror=implicit-function-declaration]
  gpio = devm_gpiod_get(&client->dev, "reset");
  ^
drivers/i2c/muxes/i2c-mux-pca954x.c:204:7: warning: assignment makes
pointer from integer without a cast [enabled by default]
  gpio = devm_gpiod_get(&client->dev, "reset");
       ^
drivers/i2c/muxes/i2c-mux-pca954x.c:206:3: error: implicit declaration
of function ‘gpiod_direction_output’
[-Werror=implicit-function-declaration]
   gpiod_direction_output(gpio, 0);
   ^
cc1: some warnings being treated as errors
make[3]: *** [drivers/i2c/muxes/i2c-mux-pca954x.o] Error 1

This is because it is getting compiled without gpiolib, so
introduce an explicit dependency.
Reported-by: NJim Davis <jim.epost@gmail.com>
Signed-off-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

The pca954x driver recently switched to the GPIO descriptor API without
including the correct <linux/gpio/consumer.h> header. This breaks
compilation without CONFIG_GPIOLIB.

drivers/i2c/muxes/i2c-mux-pca954x.c: In function ‘pca954x_probe’:
drivers/i2c/muxes/i2c-mux-pca954x.c:204:2: error: implicit declaration
of function ‘devm_gpiod_get’ [-Werror=implicit-function-declaration]
  gpio = devm_gpiod_get(&client->dev, "reset");
  ^
drivers/i2c/muxes/i2c-mux-pca954x.c:204:7: warning: assignment makes
pointer from integer without a cast [enabled by default]
  gpio = devm_gpiod_get(&client->dev, "reset");
       ^
drivers/i2c/muxes/i2c-mux-pca954x.c:206:3: error: implicit declaration
of function ‘gpiod_direction_output’
[-Werror=implicit-function-declaration]
   gpiod_direction_output(gpio, 0);
   ^
cc1: some warnings being treated as errors
make[3]: *** [drivers/i2c/muxes/i2c-mux-pca954x.o] Error 1

Fix it by including the right header.
Reported-by: NJim Davis <jim.epost@gmail.com>
Signed-off-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

The ID-based GPIO API pushes handling of GPIO polarity to drivers.
Simplify the driver by switching to the descriptor-based GPIO API.

This also fixes a mismatch between the pca954x DT bindings that document
a "reset-gpios" property and the driver that requests a "reset-gpio"
property.
Signed-off-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

Signed-off-by: NJean Delvare <khali@linux-fr.org>

None of these files are actually using any __init type directives
and hence don't need to include <linux/init.h>.  Most are just a
left over from __devinit and __cpuinit removal, or simply due to
code getting copied from one driver to the next.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>
Acked-by: NJean Delvare <khali@linux-fr.org>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

If a reset GPIO support is specified, request the GPIO and get the chip
out of reset at probe time.
Signed-off-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

This simplifies error and removal paths.
Signed-off-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

This helps avoiding duplicate includes.
Signed-off-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

The i2c-mux driver requires that the chan_id parameter
passed to the i2c_add_mux_adapter() function is equal
to the reg value for that adapter:

for_each_child_of_node(mux_dev->of_node, child) {
	ret = of_property_read_u32(child, "reg", &reg);
	if (ret)
		continue;
	if (chan_id == reg) {
		priv->adap.dev.of_node = child;
		break;
	}
}

The i2c-mux-gpio driver uses an internal logical index
for chan_id when calling i2c_add_mux_adapter() instead
of using the reg value.

Because of this, there will problems in selecting the
right adapter when the i2c-mux-gpio's index into
mux->data.values doesn't match the reg value.

An example of such a case:

mux->data.values = { 1, 0 }

For chan_id = 0, i2c-mux will bind the adapter to the
of_node with reg = <0>, but when it will call the
select() callback with chan_id set to 0, the i2c-mux-gpio
will use it as an index into mux->data.values and it will
actually select the bus with reg = <1>.
Signed-off-by: NIonut Nicu <ioan.nicu.ext@nsn.com>
Acked-by: NAlexander Sverdlin <alexander.sverdlin@nsn.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>
Cc: stable@kernel.org

Some gpio chips may have get/set operations that
can sleep. gpio_set_value() only works for chips
which do not sleep, for the others we will get a
kernel warning. Using gpio_set_value_cansleep()
will work for both chips that do sleep and those
who don't.
Signed-off-by: NIonut Nicu <ioan.nicu.ext@nsn.com>
Acked-by: NPeter Korsgaard <peter.korsgaard@barco.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>
Cc: stable@kernel.org

'of_match_ptr' is defined in linux/of.h. Include it explicitly to
avoid build breakage in the future.
Signed-off-by: NSachin Kamat <sachin.kamat@linaro.org>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

If it is not there yet, it might appear later.
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

If it is not there yet, it might appear later.
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

If the i2c-parent bus driver is not loaded, returning
-ENODEV will force people to unload and then reload the
module again to get it working.
Signed-off-by: NIonut Nicu <ioan.nicu.ext@nsn.com>
Acked-by: NPeter Korsgaard <peter.korsgaard@barco.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

of_get_named_gpio could return -E_PROBE_DEFER or another
error code. This error should be passed further instead
of being ignored.
Signed-off-by: NIonut Nicu <ioan.nicu.ext@nsn.com>
Acked-by: NPeter Korsgaard <peter.korsgaard@barco.com>
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>