Provide tracepoints for the lifecycle of a message from submission to
completion and for the active time for masters to help with performance
analysis of SPI I/O.
Signed-off-by: NMark Brown <broonie@linaro.org>

spi: quad: fix the name of DT property in patch

The previous property name spi-tx-nbits and spi-rx-nbits looks not
human-readable. To make it consistent with other devices, using property
name spi-tx-bus-width and spi-rx-bus-width instead of the previous one
specify the number of data wires that spi controller will work in.
Add the specification in spi-bus.txt.
Signed-off-by: Nwangyuhang <wangyuhang2014@gmail.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

In the case spi_master_initialize_queue() fails, current code calls
device_unregister() before return error from spi_register_master().
However, all the drivers call spi_master_put() in the error path if
spi_register_master() fails. Thus we should call device_del() rather than
device_unregister() before return error from spi_register_master().

This also makes all the spi_register_master() error handling consistent,
because all other error paths of spi_register_master() expect drivers to
call spi_master_put() if spi_register_master() fails.
Signed-off-by: NAxel Lin <axel.lin@ingics.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

The addition SPI quad support made the DT properties mandatory, breaking
compatibility with existing systems. Fix that by making them optional,
also improving the error messages while we're at it.
Signed-off-by: NMark Brown <broonie@linaro.org>
Tested-by: NStephen Warren <swarren@nvidia.com>

Delete a "return" when commit the patch to a new kernel version
 by mistake. So recover it.
Signed-off-by: Nwangyuhang <wangyuhang2014@gmail.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

We have tested master->running immediately after grab the master->queue_lock.
The status of master->running won't be changed until we release the lock.
Thus remove a redundant test for master->running.
Signed-off-by: NAxel Lin <axel.lin@ingics.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

There is a bug in the following patch:
http://comments.gmane.org/gmane.linux.kernel.spi.devel/14420

spi: DUAL and QUAD support

    fix the previous patch some mistake below:
    1. DT in slave node, use "spi-tx-nbits = <1/2/4>" in place of using
       "spi-tx-dual, spi-tx-quad" directly, same to rx. So correct the
       previous way to get the property in @of_register_spi_devices().
    2. Change the value of transfer bit macro(SPI_NBITS_SINGLE, SPI_NBITS_DUAL
       SPI_NBITS_QUAD) to 0x01, 0x02 and 0x04 to match the actual wires.
    3. Add the following check
       (1)keep the tx_nbits and rx_nbits in spi_transfer is not beyond the
          single, dual and quad.
       (2)keep tx_nbits and rx_nbits are contained by @spi_device->mode
          example: if @spi_device->mode = DUAL, then tx/rx_nbits can not be set
                   to QUAD(SPI_NBITS_QUAD)
       (3)if "@spi_device->mode & SPI_3WIRE", then tx/rx_nbits should be in
          single(SPI_NBITS_SINGLE)

Checking of the tx/rx transfer bits and mode bits should be done conditionally
based on type of buffer filled else EINVAL condition will
always get hit either for rx or tx.
Signed-off-by: NSourav Poddar <sourav.poddar@ti.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

fix the previous patch some mistake below:
1. DT in slave node, use "spi-tx-nbits = <1/2/4>" in place of using
   "spi-tx-dual, spi-tx-quad" directly, same to rx. So correct the
   previous way to get the property in @of_register_spi_devices().
2. Change the value of transfer bit macro(SPI_NBITS_SINGLE, SPI_NBITS_DUAL
   SPI_NBITS_QUAD) to 0x01, 0x02 and 0x04 to match the actual wires.
3. Add the following check
   (1)keep the tx_nbits and rx_nbits in spi_transfer is not beyond the
      single, dual and quad.
   (2)keep tx_nbits and rx_nbits are contained by @spi_device->mode
      example: if @spi_device->mode = DUAL, then tx/rx_nbits can not be set
               to QUAD(SPI_NBITS_QUAD)
   (3)if "@spi_device->mode & SPI_3WIRE", then tx/rx_nbits should be in
      single(SPI_NBITS_SINGLE)
Signed-off-by: Nwangyuhang <wangyuhang2014@gmail.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

Since the 'spi: Support transfer speed checking in the core'
change, the SPI core validates the desired speed of a given
transfer against the minimum and maximum speeds supported by
the controller.

If the speed of a transfer is not specified, the core uses
the maximum speed of the actual SPI device. However if the
maximum speed of the actual device is greater than the
maximum speed of the controller, the core will reject the
transfer due to the aforementioned change.

Change the code to use the maximum speed of the controller
by default if that is below the device's maximum speed.
Signed-off-by: NGabor Juhos <juhosg@openwrt.org>
Signed-off-by: NMark Brown <broonie@linaro.org>

Most SPI drivers that implement runtime PM support use identical code to
do so: they acquire a runtime PM lock in prepare_transfer_hardware() and
then they release it in unprepare_transfer_hardware(). The variations in
this are mostly missing error checking and the choice to use autosuspend.

Since these runtime PM calls are normally the only thing in the prepare
and unprepare callbacks and the autosuspend API transparently does the
right thing on devices with autosuspend disabled factor all of this out
into the core with a flag to enable the behaviour.
Signed-off-by: NMark Brown <broonie@linaro.org>
Reviewed-by: NStephen Warren <swarren@nvidia.com>
Acked-by: NLinus Walleij <linus.walleij@linaro.org>

Make spi core calculate the message length while
populating the other transfer parameters.

Usecase, driver can use it to populate framelength filed in their
controller.
Signed-off-by: NSourav Poddar <sourav.poddar@ti.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

Check that transfers are non-empty and that there is a completion for
them.
Signed-off-by: NMark Brown <broonie@linaro.org>

Allow drivers to avoid implementing their own checks for simple rates by
specifying the limits in the master structure.
Signed-off-by: NMark Brown <broonie@linaro.org>

Calling kthread_run with a single name parameter causes it to be handled
as a format string. Many callers are passing potentially dynamic string
content, so use "%s" in those cases to avoid any potential accidents.
Signed-off-by: NKees Cook <keescook@chromium.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

The "runtime idle" helper routine, rpm_idle(), currently ignores
return values from .runtime_idle() callbacks executed by it.
However, it turns out that many subsystems use
pm_generic_runtime_idle() which checks the return value of the
driver's callback and executes pm_runtime_suspend() for the device
unless that value is not 0.  If that logic is moved to rpm_idle()
instead, pm_generic_runtime_idle() can be dropped and its users
will not need any .runtime_idle() callbacks any more.

Moreover, the PCI, SCSI, and SATA subsystems' .runtime_idle()
routines, pci_pm_runtime_idle(), scsi_runtime_idle(), and
ata_port_runtime_idle(), respectively, as well as a few drivers'
ones may be simplified if rpm_idle() calls rpm_suspend() after 0 has
been returned by the .runtime_idle() callback executed by it.

To reduce overall code bloat, make the changes described above.
Tested-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Tested-by: NKevin Hilman <khilman@linaro.org>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: NKevin Hilman <khilman@linaro.org>
Reviewed-by: NUlf Hansson <ulf.hansson@linaro.org>
Acked-by: NAlan Stern <stern@rowland.harvard.edu>

This makes sure that an error is returned on an incorrectly formed
"cs-gpios" property, but reports success when the "cs-gpios" property is
well formed or missing.

When holes in the cs-gpios property phandle list is used to indicate
that some other form of chipselect is to be used it is important that
failure to read a broken "cs-gpios" property does not silently fail
leading to the spi controller to use an unintended chipselect.
Signed-off-by: NAndreas Larsson <andreas@gaisler.com>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

The return value from of_get_named_gpio is -ENOENT when the given index
matches a hole in the "cs-gpios" property phandle list. However, the
default value of cs_gpio in struct spi_device and entries of cs_gpios in
struct spi_master is -EINVAL, which is documented to indicate that a
GPIO line should not be used for the given spi_device.

This sets the default value of cs_gpio in struct spi_device and entries
of cs_gpios in struct spi_master to -ENOENT. Thus, -ENOENT is the only
value used to indicate that no GPIO line should be used.
Signed-off-by: NAndreas Larsson <andreas@gaisler.com>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

The ACPI handle of struct spi_master's dev member should not be
set, because this causes that struct spi_master to be associated
with the ACPI device node corresponding to its parent as the
second "physical_device", which is incorrect (this happens during
the registration of struct spi_master).  Consequently,
acpi_register_spi_devices() should use the ACPI handle of the
parent of the struct spi_master it is called for rather than that
struct spi_master's ACPI handle (which should be NULL).

Make that happen and modify the spi-pxa2xx driver, which currently is
the only driver for ACPI-enumerated SPI controller chips, not to set
the ACPI handle for the struct spi_master it creates.
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Acked-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

spi_pump_messages() calls into a controller driver with
unprepare_transfer_hardware() which is documented as "This may sleep".
As in the prepare_transfer_hardware() call below, we should release the
queue_lock spinlock before making the call.
Rework the logic a bit to hold queue_lock to protect the 'busy' flag,
then release it to call unprepare_transfer_hardware().
Signed-off-by: NBryan Freed <bfreed@chromium.org>
Reviewed-by: NDoug Anderson <dianders@chromium.org>
Signed-off-by: NDoug Anderson <dianders@chromium.org>
Acked-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

Allow SPI masters to define the set of bits_per_word values they support.
If they do this, then the SPI core will reject transfers that attempt to
use an unsupported bits_per_word value. This eliminates the need for each
SPI driver to implement this checking in most cases.
Signed-off-by: NStephen Warren <swarren@wwwdotorg.org>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

This patch replaces the horribly coded of_count_named_gpios() with a
call to of_count_phandle_with_args() which is far more efficient. This
also changes the return value of of_gpio_count() & of_gpio_named_count()
from 'unsigned int' to 'int' so that it can return an error code. All
the users of that function are fixed up to correctly handle a negative
return value.

v2: Split GPIO portion into a separate patch
Tested-by: NAndreas Larsson <andreas@gaisler.com>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>
Cc: Linus Walleij <linus.walleij@linaro.org>
Cc: Rob Herring <rob.herring@calxeda.com>

Using memset does not set an array of integers properly. Replace with a
loop to set each element properly.
Signed-off-by: NAndreas Larsson <andreas@gaisler.com>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

Move rt scheduler definitions out of include/linux/sched.h into
new file include/linux/sched/rt.h
Signed-off-by: NClark Williams <williams@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Link: http://lkml.kernel.org/r/20130207094707.7b9f825f@riff.lanSigned-off-by: NIngo Molnar <mingo@kernel.org>

All in-kernel users of class_find_device() don't really need mutable
data for match callback.

In two places (kernel/power/suspend_test.c, drivers/scsi/osd/osd_uld.c)
this patch changes match callbacks to use const search data.

The const is propagated to rtc_class_open() and power_supply_get_by_name()
parameters.

Note that there's a dev reference leak in suspend_test.c that's not
touched in this patch.
Signed-off-by: NMichał Mirosław <mirq-linux@rere.qmqm.pl>
Acked-by: NGrant Likely <grant.likely@secretlab.ca>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Use GFP_DMA in order to ensure that the memory we allocate for transfers
in spi_write_then_read() can be DMAed. On most platforms this will have
no effect.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

When spi client does the spi transfer and if it does not set
the speed for each transfer then set it as default
of spi device in spi core before calling low level transfer.

This will remove the extra check in low level driver for setting
speed.
Signed-off-by: NLaxman Dewangan <ldewangan@nvidia.com>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

Linux assigns a number to each spi_master in the system, but when the
platform used the device tree, the numbers are dynamically assigned and
are not predictable. In general this shouldn't matter since the kernel
doesn't use the bus number for anything other than matching a bus to
spi_boardinfo (not used for DT). However, sometimes userspace needs to
figure out which bus is which, so it makes sense to use the global
/aliases namespace to choose a specific bus number.

It is safe to derive the bus number from an alias because aliases will
never cause two buses to try and use the same bus number. (At one time
the cell-index property was used for this purpose, but cell-index has
the risk of an id collision).
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>
Cc: Anatolij Gustschin <agust@denx.de>
Cc: Mark Brown <broonie@opensource.wolfsonmicro.com>

To get modprobe to automatically load the proper modules, we need to
prefix things with "spi:".  Partially based on Grant Likely's suggestions.
Signed-off-by: NDavid Daney <david.daney@cavium.com>
[grant.likely: reworked because drivers/of/of_spi.c has been removed]
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

The spi support code works on SPARC too. No reason to exclude it from
the party.
Signed-off-by: NAndreas Larsson <andreas@gaisler.com>
Acked-by: NDavid S. Miller <davem@davemloft.net>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

CONFIG_HOTPLUG is going away as an option.  As result the __dev*
markings will be going away.

Remove use of __devinit, __devexit_p, __devinitdata, __devinitconst,
and __devexit.

Bill Pemberton has done most of the legwork on this series. I've used
his script to purge the attributes from the drivers/gpio tree.
Reported-by: NBill Pemberton <wfp5p@virginia.edu>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

This patch allows to specify that a SPI device is connected in 3-wire mode via
device tree.
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Acked-by: NRob Herring <rob.herring@calxeda.com>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

Fix problem discovered with sparse:
+ drivers/spi/spi.c:1554:37: sparse: incompatible types in comparison expression (different signedness)
drivers/spi/spi.c: In function 'spi_write_then_read':
drivers/spi/spi.c:1554:23: warning: comparison of distinct pointer types lacks a cast [enabled by default]

The change to SPI_BUFSIZ was introduced in commit b3a223ee, "spi:
Remove SPI_BUFSIZ restriction on spi_write_then_read()"

Cc: Mark Brown <broonie@opensource.wolfsonmicro.com>
Reported-by: Nkbuild test robot <fengguang.wu@intel.com>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

In order to avoid constantly allocating and deallocating there is a fixed
buffer which spi_write_then_read() uses for transfers, with an early error
check to ensure that the transfer fits within the buffer. This limits the
size of transfers to this size, currently max(32, SMP_CACHE_BYTES).

Since we can dynamically allocate and in fact already have a fallback
to do so when there is contention for the fixed buffer remove this
restriction and instead dynamically allocate a suitably sized buffer if
the transfer won't fit.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

ACPI 5 introduced SPISerialBus resource that allows us to enumerate and
configure the SPI slave devices behind the SPI controller. This patch adds
support for this to the SPI core.

In addition we bind ACPI nodes to SPI devices. This makes it possible for
the slave drivers to get the ACPI handle for further configuration.
Signed-off-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Acked-by: NGrant Likely <grant.likely@secretlab.ca>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

Currently of_register_spi_devices() function registers all SPI devices,
independetly from their status property in device tree. According to
"ePAPR 1.1" spec, device should only be registered if there is no
"status" property, or it has "ok" (or "okay") value (see
of_device_is_available()). In case of "platform devices",
of_platform_device_create_pdata() checks for "status" and ensures
that disabled devices are not pupulated. But such check for SPI buses
was missing until now. Fix it.
Signed-off-by: NAlexander Sverdlin <alexander.sverdlin@sysgo.com>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

This will allow to use gpio for chip select with no modification in the
driver binding

When use the cs-gpios, the gpio number will be passed via the cs_gpio field
and the number of chip select will automatically increased with max(hw cs, gpio cs).

So if for example the controller has 2 CS lines, and the cs-gpios
property looks like this:

cs-gpios = <&gpio1 0 0> <0> <&gpio1 1 0> <&gpio1 2 0>;

Then it should be configured so that num_chipselect = 4 with the
following mapping:

cs0 : &gpio1 0 0
cs1 : native
cs2 : &gpio1 1 0
cs3 : &gpio1 2 0
Signed-off-by: NJean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
Cc: devicetree-discuss@lists.ozlabs.org
Cc: spi-devel-general@lists.sourceforge.net
Signed-off-by: NRichard Genoud <richard.genoud@gmail.com>
[grant.likely: fixed up type of cs count so min() can do type checking]
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

Currently the master->setup() is called unconditionally.
The assumption is that every driver need to implement this
callback. This encourages drivers to populate empty functions
to prevent crashing.

This patch prevents the call of master->setup() if it is not
populated.
Signed-off-by: NLaxman Dewangan <ldewangan@nvidia.com>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

When spi client does the spi transfer and does not sets
the bits_per_word for each transfer then set it as default
of spi device in spi core before calling low level transfer.

Removing the similar code from spi-tegra20-slink driver as
it is not required.
Signed-off-by: NLaxman Dewangan <ldewangan@nvidia.com>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

The modalias attribute of spi devices doesn't have the "spi:" prefix
that is used in the UEVENT and in spi device drivers.  This patch adds
the prefix so the modprobe can correctly match modules to devices.
Reported-by: NDavid Daney <david.daney@cavium.com>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

Trivial simplification.  Instead of requiring spi master drivers to
always set the bus number (even when a dynamic number is desired),
this patch modifies spi_alloc_master() to initialize num_chipselect to
1 (because there will always be at least one CS) and bus_num to -1 for
dynamic allocation.  This simplifies the code needed to be written for
drivers.
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>