Check the class flags before allocating the temporary i2c_client
structure, to avoid allocating it when we don't need it.

Also optimize the inner loop a bit.
Signed-off-by: NJean Delvare <khali@linux-fr.org>
Cc: Jacob Pan <jacob.jun.pan@linux.intel.com>

The "new_device" sysfs interface has been there for quite some time
now, nobody complained about it so it must be good enough. Time to
remove the warning and call it stable.
Signed-off-by: NJean Delvare <khali@linux-fr.org>
Acked-by: NMichael Lawnick <ml.lawnick@gmx.de>

This makes the calling site's code clearer IMHO.
Signed-off-by: NJean Delvare <khali@linux-fr.org>
Acked-by: NMichael Lawnick <ml.lawnick@gmx.de>

Commit 959e85f7, "i2c: add OF-style registration and binding" caused a
module dependency loop where of_i2c.c calls functions in i2c-core, and
i2c-core calls of_i2c_register_devices() in of_i2c.  This means that
when i2c support is built as a module when CONFIG_OF is set, then
neither i2c_core nor of_i2c are able to be loaded.

This patch fixes the problem by moving the of_i2c_register_devices()
calls back into the device drivers.  Device drivers already
specifically request the core code to parse the device tree for
devices anyway by setting the of_node pointer, so it isn't a big
deal to also call the registration function.  The drivers just become
slightly more verbose.
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>
Signed-off-by: NJean Delvare <khali@linux-fr.org>

For devices which are not adapted to runtime PM a call to
pm_runtime_suspended always returns true.

Hence the pm_runtime_suspended checks below prevent legacy
suspend from getting called.

So do a pm_runtime_suspended check only for devices with a
dev_pm_ops populated (which hence do not rely on the legacy
suspend.)
Signed-off-by: NRajendra Nayak <rnayak@ti.com>
Acked-by: NRafael J. Wysocki <rjw@sisk.pl>
Cc: Ben Dooks <ben-linux@fluff.org>
Cc: Mark Brown <broonie@opensource.wolfsonmicro.com>
Cc: Kevin Hilman <khilman@deeprootsystems.com>
Signed-off-by: NJean Delvare <khali@linux-fr.org>

In current i2c core driver, call to pm_runtime_set_active from
i2c_device_pm_resume will unconditionally enable i2c module and
increment child count of the parent. Because of this, in CPU Idle
path, i2c does not idle, preventing Core to enter retention. Also i2c
module will not be suspended upon system suspend as
pm_runtime_set_suspended is not called from i2c_device_pm_suspend.

This issue is fixed by removing pm_runtime_set_active call from resume
path which is not necessary.
This fix has been tested on OMAP4430.
Signed-off-by: NPartha Basak <p-basak2@ti.com>
Signed-off-by: NVishwanath BS <vishwanath.bs@ti.com>
Acked-by: NRafael J. Wysocki <rjw@sisk.pl>
Cc: Kevin Hilman <khilman@deeprootsystems.com>
Cc: Ben Dooks <ben-linux@fluff.org>
Signed-off-by: NJean Delvare <khali@linux-fr.org>

Add multiplexed bus core support. I2C multiplexer and switches
like pca954x get instantiated as new adapters per port.
Signed-off-by: NMichael Lawnick <ml.lawnick@gmx.de>
Acked-by: NRodolfo Giometti <giometti@linux.it>
Signed-off-by: NJean Delvare <khali@linux-fr.org>

Moving userspace-instantiated clients to separate lists wasn't nearly
enough to avoid deadlocks in multiplexed bus cases. We also want to
have a dedicated mutex to protect each list.
Signed-off-by: NJean Delvare <khali@linux-fr.org>
Cc: Michael Lawnick <ml.lawnick@gmx.de>

Make i2c_default_probe self-sufficient, so that callers don't have to
do functionality checks themselves. This ensures everything is and
will stay consistent.
Signed-off-by: NJean Delvare <khali@linux-fr.org>

Now that bus_for_each_drv() is no longer __must_check, we can drop the
dummy variable that was used to store the returned value.
Signed-off-by: NJean Delvare <khali@linux-fr.org>
Cc: Justin P. Mattock <justinmattock@gmail.com>

Uninline i2c adapter locking helper functions, move them to i2c-core,
and use them in i2c-core itself. The functions are still exported for
external users. This makes future updates to the locking model (which
will be needed for multiplexing support) possible and transparent.
Signed-off-by: NJean Delvare <khali@linux-fr.org>
Cc: Michael Lawnick <ml.lawnick@gmx.de>

Now that i2c-core offers the possibility to provide custom probing
function for I2C devices, let's make use of it.
Signed-off-by: NJean Delvare <khali@linux-fr.org>
Acked-by: NMauro Carvalho Chehab <mchehab@redhat.com>

The probe method used by i2c_new_probed_device() may not be suitable
for all cases. Let the caller provide its own, optional probe
function.
Signed-off-by: NJean Delvare <khali@linux-fr.org>
Acked-by: NMauro Carvalho Chehab <mchehab@redhat.com>

The new unified probing function differs from the original code, and
the preliminary test whether probing is possible must be updated
accordingly.
Signed-off-by: NJean Delvare <khali@linux-fr.org>

This patch adds OF hooks to the i2c core so that devices can automatically
be registered based on device tree data.
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

Otherwise it's not clear what it is checking.

Also move the function to save a forward declaration.
Signed-off-by: NJean Delvare <khali@linux-fr.org>

Move strict I2C address validity check to a single function, and
document the reserved I2C addresses there.
Signed-off-by: NJean Delvare <khali@linux-fr.org>
Cc: Hans Verkuil <hverkuil@xs4all.nl>

Do basic address validity checks when a client is being registered. We
already had checks in place for devices which are being detected, but
not for devices which are simply instantiated.

This is a very basic check. We don't want to do strict checking here
because some devices are known to infringe the I2C address constraints
(e.g. IR receivers at 7-bit address 0x7a while this value is
supposedly reserved for 10-bit addresses.) So we assume the caller
knows what it is doing.
Signed-off-by: NJean Delvare <khali@linux-fr.org>
Cc: Matthieu Castet <castet.matthieu@free.fr>

Use the same I2C device presence detection code for legacy and new
device detection functions. This is more consistent and makes the code
smaller.
Signed-off-by: NJean Delvare <khali@linux-fr.org>
Cc: Matthieu Castet <castet.matthieu@free.fr>

Fix up coding style issues found by the checkpatch.pl tool.
Signed-off-by: NFarid Hammane <farid.hammane@gmail.com>
Signed-off-by: NJean Delvare <khali@linux-fr.org>

There are three issues with the i2c bus type's power management
callbacks at the moment.  First, they don't include any hibernate
callbacks, although they should at least include the .restore()
callback (there's no guarantee that the driver will be present in
memory before loading the image kernel and we must restore the
pre-hibernation state of the device).  Second, the "legacy"
callbacks are not going to be invoked by the PM core since the bus
type's pm object is not NULL.  Finally, the system sleep PM
(ie. suspend/resume) callbacks don't check if the device has been
already suspended at run time, in which case they should skip
suspending it.  Also, it looks like the i2c bus type can use the
generic subsystem-level runtime PM callbacks.

For these reasons, rework the system sleep PM callbacks provided by
the i2c bus type to handle hibernation correctly and to invoke the
"legacy" callbacks for drivers that provide them.  In addition to
that make the i2c bus type use the generic subsystem-level runtime
PM callbacks.
Signed-off-by: NRafael J. Wysocki <rjw@sisk.pl>
Acked-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NJean Delvare <khali@linux-fr.org>

Using a single list for all userspace devices leads to a dead lock
on multiplexed buses in some circumstances (mux chip instantiated
from userspace). This is solved by using a separate list for each
bus segment.
Signed-off-by: NJean Delvare <khali@linux-fr.org>
Acked-by: NMichael Lawnick <ml.lawnick@gmx.de>

Some FSC hardware monitoring chips (Syleus at least) doesn't like
quick writes we typically use to probe for I2C chips. Use a regular
byte read instead for the address they live at (0x73). These are the
only known chips living at this address on PC systems.

For clarity, this fix should not be needed for kernels 2.6.30 and
later, as we started instantiating the hwmon devices explicitly based
on DMI data. Still, this fix is valuable in the following two cases:
* Support for recent FSC chips on older kernels. The DMI-based device
  instantiation is more difficult to backport than the device support
  itself.
* Case where the DMI-based device instantiation fails, whatever the
  reason. We fall back to probing in that case, so it should work.

This fixes kernel bug #15634:
https://bugzilla.kernel.org/show_bug.cgi?id=15634Signed-off-by: NJean Delvare <khali@linux-fr.org>
Acked-by: NHans de Goede <hdegoede@redhat.com>
Cc: stable@kernel.org

After discovering that a lot of i2c-drivers leave the pointer to their
clientdata dangling, it was decided to let the core handle this issue.
It is assumed that the core may access the private data after remove()
as there are no guarantees for the lifetime of such pointers anyhow (see
thread starting at http://lkml.org/lkml/2010/3/21/68)
Signed-off-by: NWolfram Sang <w.sang@pengutronix.de>
Signed-off-by: NJean Delvare <khali@linux-fr.org>

The struct device_node *of_node pointer is moving out of dev->archdata
and into the struct device proper.  of_i2c.c needs to set the of_node
pointer before the device is registered.  Since the i2c subsystem
doesn't allow 2 stage allocation and registration of i2c devices, the
of_node pointer needs to be passed via the i2c_board_info structure
so that it is set prior to registration.

This patch adds of_node to struct i2c_board_info (conditional on
CONFIG_OF), sets of_node in i2c_new_device(), and modifies of_i2c.c
to use the new parameter.  The calling of dev_archdata_set_node()
from of_i2c will be removed in a subsequent patch when of_node is
removed from archdata and all users are converted over.
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

i2c_master_send & i2c_master_recv do not support more than 64 kb
transfer, since msg.len is u16.
Signed-off-by: NZhangfei Gao <zgao6@marvell.com>
Signed-off-by: NJean Delvare <khali@linux-fr.org>

Allow I2C drivers to make use of the runtime PM framework by adding
bus implementations of the runtime PM operations. These simply
immediately suspend when the device is idle. The runtime PM framework
provides drivers with off the shelf refcounts for enables and sysfs
control for managing runtime suspend from userspace so is useful even
without meaningful input from the bus.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NJean Delvare <khali@linux-fr.org>

dev_dbg outputs dev_name, which is released with device_unregister. This bug
resulted in output like this:

i2c Xy2�0: adapter [SMBus I801 adapter at 1880] unregistered

The right output would be:
i2c i2c-0: adapter [SMBus I801 adapter at 1880] unregistered
Signed-off-by: NThadeu Lima de Souza Cascardo <cascardo@holoscopio.com>
Signed-off-by: NJean Delvare <khali@linux-fr.org>

The C99 specification states in section 6.11.5:

The placement of a storage-class specifier other than at the beginning
of the declaration specifiers in a declaration is an obsolescent
feature.
Signed-off-by: NTobias Klauser <tklauser@distanz.ch>
Signed-off-by: NJean Delvare <khali@linux-fr.org>

Struct dev_pm_ops is not configured in current i2c bus type. i2c drivers
only depends on suspend/resume entries in struct dev_pm_ops are not
informed of PM suspend and resume events by i2c framework.
Signed-off-by: NSonic Zhang <sonic.zhang@analog.com>
Signed-off-by: NJean Delvare <khali@linux-fr.org>

Struct i2c_client_address_data only contains one field at this point,
which makes its usefulness questionable. Get rid of it and pass simple
address lists around instead.
Signed-off-by: NJean Delvare <khali@linux-fr.org>
Tested-by: NWolfram Sang <w.sang@pengutronix.de>

The "kind" parameter always has value -1, and nobody is using it any
longer, so we can remove it.
Signed-off-by: NJean Delvare <khali@linux-fr.org>
Tested-by: NWolfram Sang <w.sang@pengutronix.de>

Functions i2c_do_add_adapter() and __attach_adapter() do essentially
the same thing, differing only in how the parameters are passed. Same
for i2c_do_add_adapter() and __detach_adapter(). Introduce wrappers to
normalize the parameters, so that we do not have to duplicate the
code.
Signed-off-by: NJean Delvare <khali@linux-fr.org>
Cc: David Brownell <dbrownell@users.sourceforge.net>

The kind parameter of i2c_detect_address() always has value -1, so we
can get rid of it.

Next step is to update all i2c detect callback functions to get rid of
this now useless parameter.
Signed-off-by: NJean Delvare <khali@linux-fr.org>

The legacy probe and force module parameters are obsolete now, the
same can be achieved using the new_device sysfs interface, which is
both more flexible and cheaper (it is implemented by i2c-core rather
than replicated in every driver module.)

The legacy ignore module parameters can be dropped as well. Ignoring
can be done by instantiating a "dummy" device at the problematic
address.

This is the first step of a huge cleanup to i2c-core's i2c_detect
function, i2c.h's I2C_CLIENT_INSMOD* macros, and all drivers that made
use of them.
Signed-off-by: NJean Delvare <khali@linux-fr.org>

Low priority thread holding the i2c bus mutex could block higher
priority threads to access the bus resulting in unacceptable
latencies. Change the mutex type to rt_mutex preventing priority
inversion.
Tested-by: NPeter Ujfalusi <peter.ujfalusi@nokia.com>
Signed-off-by: NMika Kuoppala <mika.kuoppala@nokia.com>
Signed-off-by: NJean Delvare <khali@linux-fr.org>

Fix userspace_device list corruption. The corruption was caused by
clients not being removed when adapters with such clients were
themselves removed. Something like the following would trigger it
(assuming i2c-stub gets adapter number 3):

# modprobe i2c-stub chip_addr=0x50
# echo 24c08 0x50 > /sys/bus/i2c/devices/i2c-3/new_device 
# rmmod i2c-stub
# modprobe i2c-stub chip_addr=0x50
# echo 24c08 0x50 > /sys/bus/i2c/devices/i2c-3/new_device 

For the records, the stack trace in the kernel logs look like this:

kernel: WARNING: at lib/list_debug.c:30 __list_add+0x8b/0x90()
kernel: Hardware name: (...)
kernel: list_add corruption. prev->next should be next (c137fc84), but was (null). (prev=f57111b8).
kernel: Modules linked in: (...)
kernel: Pid: 4669, comm: bash Not tainted 2.6.32-rc8 #259
kernel: Call Trace:
kernel:  [<c111eb8b>] ? __list_add+0x8b/0x90
kernel:  [<c111eb8b>] ? __list_add+0x8b/0x90
kernel:  [<c103265c>] warn_slowpath_common+0x6c/0xc0
kernel:  [<c111eb8b>] ? __list_add+0x8b/0x90
kernel:  [<c10326f6>] warn_slowpath_fmt+0x26/0x30
kernel:  [<c111eb8b>] __list_add+0x8b/0x90
kernel:  [<c11ba165>] i2c_sysfs_new_device+0x1c5/0x250
kernel:  [<c10861be>] ? might_fault+0x2e/0x80
kernel:  [<c11b9fa0>] ? i2c_sysfs_new_device+0x0/0x250
kernel:  [<c118c625>] dev_attr_store+0x25/0x30
kernel:  [<c10e305c>] sysfs_write_file+0x9c/0xf0
kernel:  [<c109d35c>] vfs_write+0x9c/0x160
kernel:  [<c10e2fc0>] ? sysfs_write_file+0x0/0xf0
kernel:  [<c109d4dd>] sys_write+0x3d/0x70
kernel:  [<c1002ed8>] sysenter_do_call+0x12/0x36
Signed-off-by: NJean Delvare <khali@linux-fr.org>

Some user-space applications may be relying on i2c adapters showing up
as class devices in sysfs. Provide compatibility links for them for
the time being. We will remove them after a long transition period.
Signed-off-by: NJean Delvare <khali@linux-fr.org>
Cc: Kay Sievers <kay.sievers@vrfy.org>

Kay says i2c adapters shouldn't be class devices but bus devices.
Convert them that way, using a device type.
Signed-off-by: NJean Delvare <khali@linux-fr.org>
Cc: Kay Sievers <kay.sievers@vrfy.org>

This is required for upcoming changes.
Signed-off-by: NJean Delvare <khali@linux-fr.org>
Cc: Kay Sievers <kay.sievers@vrfy.org>