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: Peter Rosin <peda@axentia.se>
Signed-off-by: Wolfram 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: Peter Rosin <peda@axentia.se>
Signed-off-by: Wolfram 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: Peter Rosin <peda@axentia.se>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

All i2c-muxes have a parent adapter and one or many child
adapters. A mux also has some means of selection. Previously,
this was stored per child adapter, but it is only needed
to keep track of this per mux.

Add an i2c mux core, that keeps track of this consistently.

Also add some glue for users of the old interface, which will
create one implicit mux core per child adapter.
Signed-off-by: Peter Rosin <peda@axentia.se>
Tested-by: Antti Palosaari <crope@iki.fi>
Tested-by: Crestez Dan Leonard <leonard.crestez@intel.com>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

We cannot expect msleep(1) to actually sleep for a period shorter than
20 ms. Replace all calls to msleep() with usleep_range().
Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

The second check for I2C_FUNC_I2C is reduntant, so remove it.
Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
[wsa: reworded commit message]
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

Switch to the new generic functions: i2c_parse_fw_timings().
Signed-off-by: David Wu <david.wu@rock-chips.com>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

Remove superfluous check and stray newline.
Signed-off-by: Jan Glauber <jglauber@cavium.com>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

Add helper functions for control, data and status register access.
This simplifies the code and makes the purpose of the register
access clearer.
Signed-off-by: Jan Glauber <jglauber@cavium.com>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

Rename the [read|write]_sw functions to make it clearer they access
the TWSI registers.
Signed-off-by: Jan Glauber <jglauber@cavium.com>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

No functional change, just moving the functions upward in
preparation of improving the recovery.
Signed-off-by: Jan Glauber <jglauber@cavium.com>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

Convert the adapter timeout to 2 ms independently of depending on CONFIG_HZ.
CONFIG_HZ is 100 for MIPS Cavium-Octeon so the timeout value is not changed.

Also set retries to 5 to improve robustness.
Signed-off-by: Jan Glauber <jglauber@cavium.com>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

Enable multi-master mode in I2C_CNFG reg based on hw features.
Using single/multi-master mode bit introduced for Tegra210,
whereas multi-master mode is enabled by default in HW for T124 and
earlier Tegra SOC. Enabling this bit doesn't explicitly start
treating the bus has having multiple masters, but will start
checking for arbitration lost and reporting when it occurs.

The Tegra210 I2C controller supports single/multi master mode.
Add chipdata for Tegra210 and its compatibility string so that
Tegra210 will select data that enables multi master mode correctly.

Do below prerequisites for multi-master bus if "multi-master"
dt property entry is added.
 1. Enable 1st level clock always set.
 2. Disable 2nd level clock gating (slcg which
    is supported from T124 SOC and later chips)
Signed-off-by: Shardar Shariff Md <smohammed@nvidia.com>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

Allow runtime PM so that PM and PCI core can put the device into low-power
state when idle and resume it back when needed in those platforms that
support PM for i801 device.

Enable also autosuspend with 1 second delay in order to not needlessly
toggle power state on and off if there are multiple transactions during
short time.

Device is resumed at the beginning of bus access and marked idle ready
for autosuspend at the end of it.
Signed-off-by: Jarkko Nikula <jarkko.nikula@linux.intel.com>
Tested-by: Reinette Chatre <reinette.chatre@intel.com>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

Stop using legacy PCI PM support and convert to standard dev_pm_ops.
This provides more straightforward path to add runtime PM.

While at it remove explicit PCI power state control and configuration space
save/restore as the PCI core does it.
Signed-off-by: Jarkko Nikula <jarkko.nikula@linux.intel.com>
Tested-by: Reinette Chatre <reinette.chatre@intel.com>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

When using a certain I2C device with runtime PM enabled on
a certain I2C bus adaper the following happens:

struct amba_device *foo
   \
   struct i2c_adapter *bar
      \
      struct i2c_client *baz

The AMBA device foo has its device PM struct set to ignore
children with pm_suspend_ignore_children(&foo->dev, true).
This makes runtime PM work just fine locally in the driver:
the fact that devices on the bus are suspended or resumed
individually does not affect its operation, and the hardware
does not power up unless transferring messages.

However this child ignorance property is not inherited into
the struct i2c_adapter *bar.

On system suspend things will work fine.

On system resume the following annoying phenomenon occurs:

- In the pm_runtime_force_resume() path of
  struct i2c_client *baz, pm_runtime_set_active(&baz->dev); is
  eventually called.

- This becomes __pm_runtime_set_status(&baz->dev, RPM_ACTIVE);

- __pm_runtime_set_status() detects that RPM state is changed,
  and checks whether the parent is:
  not active (RPM_ACTIVE) and not ignoring its children
  If this happens it concludes something is wrong, because
  a parent that is not ignoring its children must be active
  before any children activate.

- Since the struct i2c_adapter *bar does not ignore
  its children, the PM core thinks that it must indeed go
  online before its children, the check bails out with
  -EBUSY, i.e. the i2c_client *baz thinks it can't work
  because it's parent is not online, and it respects its
  parent.

- In the driver the .resume() callback returns -EBUSY from
  the runtime_force_resume() call as per above. This leaves
  the device in a suspended state, leading to bad behaviour
  later when the device is used. The following debug
  print is made with an extra printg patch but illustrates
  the problem:

[   17.040832] bh1780 2-0029: parent (i2c-2) is not active
               parent->power.ignore_children = 0
[   17.040832] bh1780 2-0029: pm_runtime_force_resume:
               pm_runtime_set_active() failed (-16)
[   17.040863] dpm_run_callback():
               pm_runtime_force_resume+0x0/0x88 returns -16
[   17.040863] PM: Device 2-0029 failed to resume: error -16

Fix this by letting all struct i2c_adapter:s ignore their
children: i2c children have no business doing keeping
their parents awake: they are completely autonomous
devices that just use their parent to talk, a usecase
which must be power managed in the host on a per-message
basis.
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

Drivers do this in various ways, let's use one standard way of doing it.
Note: I2C_M_RD is bit 0, so the code could be simplified. To be extremly
robust and to advertise good coding practices, I still use the ternary
operator and let the compilers do the optimizing job.
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

There is code out there in user space and kernel space which relies on
I2C_M_RD being bit 0 to simplify their bit operations. Add a comment to
make sure this will never break. Do proper sorting of the defines while
we are here.
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>

Pull ARM fixes from Russell King:
 "A couple of small fixes, and wiring up the new syscalls which appeared
  during the merge window"

* 'fixes' of git://ftp.arm.linux.org.uk/~rmk/linux-arm:
  ARM: 8550/1: protect idiv patching against undefined gcc behavior
  ARM: wire up preadv2 and pwritev2 syscalls
  ARM: SMP enable of cache maintanence broadcast

Pull MMC fixes from Ulf Hansson:
 "Here are a couple of mmc fixes intended for v4.6 rc3:

  MMC host:
   - sdhci: Fix regression setting power on Trats2 board
   - sdhci-pci: Add support and PCI IDs for more Broxton host controllers"

* tag 'mmc-v4.6-rc1' of git://git.linaro.org/people/ulf.hansson/mmc:
  mmc: sdhci-pci: Add support and PCI IDs for more Broxton host controllers
  mmc: sdhci: Fix regression setting power on Trats2 board

Pull i2c fixes from Wolfram Sang:
 "Some bugfixes from I2C:

   - fix a uevent triggered boot problem by removing a useless debug
     print

   - fix sysfs-attributes of the new i2c-demux-pinctrl driver to follow
     standard kernel behaviour

   - fix a potential division-by-zero error (needed two takes)"

* 'i2c/for-current' of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux:
  i2c: jz4780: really prevent potential division by zero
  Revert "i2c: jz4780: prevent potential division by zero"
  i2c: jz4780: prevent potential division by zero
  i2c: mux: demux-pinctrl: Update docs to new sysfs-attributes
  i2c: mux: demux-pinctrl: Clean up sysfs attributes
  i2c: prevent endless uevent loop with CONFIG_I2C_DEBUG_CORE

This reverts commit 1028b55b.

It's broken: it makes ext4 return an error at an invalid point, causing
the readdir wrappers to write the the position of the last successful
directory entry into the position field, which means that the next
readdir will now return that last successful entry _again_.

You can only return fatal errors (that terminate the readdir directory
walk) from within the filesystem readdir functions, the "normal" errors
(that happen when the readdir buffer fills up, for example) happen in
the iterorator where we know the position of the actual failing entry.

I do have a very different patch that does the "signal_pending()"
handling inside the iterator function where it is allowable, but while
that one passes all the sanity checks, I screwed up something like four
times while emailing it out, so I'm not going to commit it today.

So my track record is not good enough, and the stars will have to align
better before that one gets committed.  And it would be good to get some
review too, of course, since celestial alignments are always an iffy
debugging model.

IOW, let's just revert the commit that caused the problem for now.
Reported-by: Greg Thelen <gthelen@google.com>
Cc: Theodore Ts'o <tytso@mit.edu>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Pull parisc fixes from Helge Deller:
 "Since commit 0de79858 ("parisc: Use generic extable search and
  sort routines") module loading is boken on parisc, because the parisc
  module loader wasn't prepared for the new R_PARISC_PCREL32 relocations.

  In addition, due to that breakage, Mikulas Patocka noticed that
  handling exceptions from modules probably never worked on parisc.  It
  was just masked by the fact that exceptions from modules don't happen
  during normal use.

  This patch series fixes those issues and survives the tests of the
  lib/test_user_copy kernel module test.  Some patches are tagged for
  stable"

* 'parisc-4.6-3' of git://git.kernel.org/pub/scm/linux/kernel/git/deller/parisc-linux:
  parisc: Update comment regarding relative extable support
  parisc: Unbreak handling exceptions from kernel modules
  parisc: Fix kernel crash with reversed copy_from_user()
  parisc: Avoid function pointers for kernel exception routines
  parisc: Handle R_PARISC_PCREL32 relocations in kernel modules