This patch adds device tree support to the spi-pl022 driver.

Based on the initial patch by Alexandre Pereira da Silva <aletes.xgr@gmail.com>
Signed-off-by: NRoland Stigge <stigge@antcom.de>
Acked-by: NAlexandre Pereira da Silva <aletes.xgr@gmail.com>
Reviewed-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

This patch adds the ability for the driver to control the chip select directly.
This enables independence from cs_control callbacks.  Configurable via
platform_data, to be extended as DT in the following patch.

Based on the initial patch by Alexandre Pereira da Silva <aletes.xgr@gmail.com>
Signed-off-by: NRoland Stigge <stigge@antcom.de>
Acked-by: NAlexandre Pereira da Silva <aletes.xgr@gmail.com>
Reviewed-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

The call sequence spi_alloc_master/spi_register_master/spi_unregister_master
is complete; it reduces the device reference count to zero, which results in
device memory being freed. An extra call to spi_master_put is unnecessary
and results in an access to free memory. Drop it.
Signed-off-by: NGuenter Roeck <linux@roeck-us.net>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

If we're gonna use runtime PM it's a pretty good idea to actually
enable it in probe() and disable it in remove() too, so it
gets used for real. Up until now we only fooled around with the
reference count.

Cc: Vinit Shenoy <vinit.shenoy@st.com>
Signed-off-by: NLinus Walleij <linus.walleij@linaro.org>

This ASIC variant has been deleted from the ARM tree, no need
to keep support around.

Cc: Vinit Shenoy <vinit.shenoy@st.com>
Signed-off-by: NLinus Walleij <linus.walleij@linaro.org>

Commit ffbbdd21
"spi: create a message queueing infrastructure"
Accidentally deleted the logic to disable the port
when unused leading to higher power consumption.
Fix this up.

Cc: stable@kernel.org
Cc: Vinit Shenoy <vinit.shenoy@st.com>
Signed-off-by: NVirupax Sadashivpetimath <virupax.sadashivpetimath@stericsson.com>
Signed-off-by: NLinus Walleij <linus.walleij@linaro.org>

Currently, if we request for frequency greater than maximum possible, spi driver
returns error.

For example, if the spi block src frequency is 333/4 MHz, i.e. 83.33.. MHz,
maximum frequency programmable would be src/2. Which would come around 41.6...

It is difficult to pass frequency in these figures. We normally try to program
in round figures, like 42 MHz and it should get programmed to <=
requested_frequency, i.e. 41.6...

For this to happen, we must not return error even if requested freq is higher
than max possible. But should program it to max possible.
Reported-by: NVinit Kamalaksha Shenoy <vinit.shenoy@st.com>
Signed-off-by: NViresh Kumar <viresh.kumar@st.com>
Acked-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

calculate_effective_freq() was still not optimized and there were cases when it
returned without error and with values of cpsr and scr as zero.

Also, the variable named found is not used well.

This patch targets to optimize and correct this routine. Tested for SPEAr.
Signed-off-by: NViresh Kumar <viresh.kumar@st.com>
Tested-by: NVinit Kamalaksha Shenoy <vinit.shenoy@st.com>
Acked-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

pl022 ssp controller supports word lengths from 4 to 16 (or 32) bits.
Currently implemented checks were incorrect. It has following check

if (pl022->vendor->max_bpw >= 32)

which must be checking for <=.

Also error print message is incorrect, that prints "range is from 1 to
16".

Fix both these issues.
Signed-off-by: NVinit Shenoy <vinit.shenoy@st.com>
Acked-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

The AMBA bus regulator support is being used to model on/off switches
for power domains which isn't terribly idiomatic for modern kernels with
the generic power domain code and creates integration problems on platforms
which don't use regulators for their power domains as it's hard to tell
the difference between a regulator that is needed but failed to be provided
and one that isn't supposed to be there (though DT does make that easier).

Platforms that wish to use the regulator API to manage their power domains
can indirect via the power domain interface.

This feature is only used with the vape supply of the db8500 PRCMU
driver which supplies the UARTs and MMC controllers, none of which have
support for managing vcore at runtime in mainline (only pl022 SPI
controller does).  Update that supply to have an always_on constraint
until the power domain support for the system is updated so that it is
enabled for these users, this is likely to have no impact on practical
systems as probably at least one of these devices will be active and
cause AMBA to hold the supply on anyway.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NLinus Walleij <linus.walleij@linaro.org>
Tested-by: NShawn Guo <shawn.guo@linaro.org>
Signed-off-by: NRussell King <rmk+kernel@arm.linux.org.uk>

Add inline wrappers for device_prep_slave_sg() and device_prep_dma_cyclic()
interfaces to hide new parameter from current users of affected interfaces.
Convert current users to use new wrappers instead of direct calls.
Suggested by Russell King [https://lkml.org/lkml/2012/2/3/269].
Signed-off-by: NAlexandre Bounine <alexandre.bounine@idt.com>
Signed-off-by: NVinod Koul <vinod.koul@linux.intel.com>

This rips the message queue in the PL022 driver out and pushes
it into (optional) common infrastructure. Drivers that want to
use the message pumping thread will need to define the new
per-messags transfer methods and leave the deprecated transfer()
method as NULL.

Most of the design is described in the documentation changes that
are included in this patch.

Since there is a queue that need to be stopped when the system
is suspending/resuming, two new calls are implemented for the
device drivers to call in their suspend()/resume() functions:
spi_master_suspend() and spi_master_resume().

ChangeLog v1->v2:
- Remove Kconfig entry and do not make the queue support optional
  at all, instead be more agressive and have it as part of the
  compulsory infrastructure.
- If the .transfer() method is implemented, delete print a small
  deprecation notice and do not start the transfer pump.
- Fix a bitrotted comment.
ChangeLog v2->v3:
- Fix up a problematic sequence courtesy of Chris Blair.
- Stop rather than destroy the queue on suspend() courtesy of
  Chris Blair.
Signed-off-by: NChris Blair <chris.blair@stericsson.com>
Signed-off-by: NLinus Walleij <linus.walleij@linaro.org>
Tested-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Reviewed-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

.device_fc is added in struct dma_slave_config recently. All user drivers, which
want DMA to be the flow controller must pass this field as false. As earlier
driver don't look to use this feature, mark it false for now.
Signed-off-by: NViresh Kumar <viresh.kumar@st.com>
Acked-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NVinod Koul <vinod.koul@linux.intel.com>

WARNING: drivers/spi/built-in.o(.devinit.text+0xdb8): Section mismatch in reference from the function pl022_probe() to the function .init.text:pl022_dma_probe()
The function __devinit pl022_probe() references
a function __init pl022_dma_probe().
If pl022_dma_probe is only used by pl022_probe then
annotate pl022_dma_probe with a matching annotation.
Signed-off-by: NRussell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

This switches the PL022 worker to a kthread in order to get
hold of a mechanism to control the message pump priority. On
low-latency systems elevating the message kthread to realtime
priority give a real sleek response curve. This has been
confirmed by measurements. Realtime priority elevation for
a certain PL022 port can be requested from platform data.

Cc: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NViresh Kumar <viresh.kumar@st.com>
Signed-off-by: NChris Blair <chris.blair@stericsson.com>
Signed-off-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

There is a possibility that the pump_message and giveback
run in parallel on a SMP system. Both the pump_message and
giveback threads work on the same SPI message queue.
Results will be in correct if the pump_message gets to
work on the queue first.

when the pump_message works with the queue, it reads the
head of the queue and removes it from the queue. pump_message
activates the chip select depending on this message read.

This leads to giveback working on the modified queue or a
emptied queue. If the queue is empty or if the next message on
the queue (which is not the actual next message, as the pump
message has removed the next message from the queue) is not for
the same SPI device as that Of the previous one, giveback will
de-activate the chip select activated by pump_message(),
which is wrong.

To solve this problem pump_message is made not to run and
access the queue until the giveback is done handling the queue.
I.e. by making the cur_msg NULL after the giveback has read the
queue.

Also a state variable has been introduced to keep track of when
the CS for next message is already activated and avoid to
double-activate it.
Reviewed-by: NViresh Kumar <viresh.kumar@st.com>
Signed-off-by: NVirupax Sadashivpetimath <virupax.sadashivpetimath@stericsson.com>
Signed-off-by: NLinus Walleij <linus.walleij@linaro.org>

Adds support for configuring the spi bus to use autosuspend for
runtime power management. This can reduce the latency in starting an
spi transfer by not suspending the device immediately following
completion of a transfer. If another transfer then takes place before
the autosuspend timeout, the call to resume the device can return
immediately rather than needing to risk sleeping in order to resume
the device.
Reviewed-by: NViresh Kumar <viresh.kumar@st.com>
Signed-off-by: NChris Blair <chris.blair@stericsson.com>
Signed-off-by: NLinus Walleij <linus.walleij@linaro.org>

Make sure we clear the enable bit when the block is not used.
This will save some energy in certain hardware versions.
Reviewed-by: NViresh Kumar <viresh.kumar@st.com>
Signed-off-by: NVirupax Sadashivpetimath <virupax.sadashivpetimath@stericsson.com>
Signed-off-by: NLinus Walleij <linus.walleij@linaro.org>

Moves the disabling of the device and clocks to the same thread in
which the device and clocks are enabled. This avoids SMP issues where
the device can be enabled for a transfer by one thread and then
disabled by the completion of the previous transfer in another thread.
Reviewed-by: NViresh Kumar <viresh.kumar@st.com>
Signed-off-by: NChris Blair <chris.blair@stericsson.com>
Signed-off-by: NLinus Walleij <linus.walleij@linaro.org>

The loading of the default configuration before suspending has
been in the driver since its inception, but it is not really
needed. Especially so since we take to all the trouble of
enabling and disabling power and clock just to do this. Let's
scrap this now.
Reviewed-by: NViresh Kumar <viresh.kumar@st.com>
Signed-off-by: NVirupax Sadashivpetimath <virupax.sadashivpetimath@stericsson.com>
Signed-off-by: NLinus Walleij <linus.walleij@linaro.org>

The driver build complains with newer compilers unless you
initialize this struct properly.
Reviewed-by: NViresh Kumar <viresh.kumar@st.com>
Signed-off-by: NJonas Aaberg <jonas.aberg@stericsson.com>
Signed-off-by: NLinus Walleij <linus.walleij@linaro.org>

For interrupt mode transfers, start with only TX interrupts enabled
to reduce the overall number of interrupts received. Once TX is
complete, enable RX interrupts to complete the transfer.
Reviewed-by: NViresh Kumar <viresh.kumar@st.com>
Signed-off-by: NChris Blair <chris.blair@stericsson.com>
Signed-off-by: NLinus Walleij <linus.walleij@linaro.org>

Signed-off-by: NDave Martin <dave.martin@linaro.org>
Acked-by: NGrant Likely <grant.likely@secretlab.ca>
Acked-by: NLinus Walleij <linus.walleij@linaro.org>

Make sure we enable the clock before leaving probe.
Signed-off-by: NUlf Hansson <ulf.hansson@stericsson.com>
Acked-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NRussell King <rmk+kernel@arm.linux.org.uk>

fixup usage of dma direction by introducing dma_transfer_direction,
this patch moves spi, serial drivers to use new enum
Signed-off-by: NVinod Koul <vinod.koul@linux.intel.com>
Cc: Grant Likely <grant.likely@secretlab.ca>
Cc: Alan Cox <alan@linux.intel.com>
Acked-by: NMika Westerberg <mika.westerberg@iki.fi>

Signed-off-by: NRussell King <rmk+kernel@arm.linux.org.uk>

This patch fixes a build error, introduced by commit (67fc8b9f, "PM: add
runtime PM support to core Primecell driver") which unfortunately was a little
bit incomplete and did contain a typo (11 instead of 22).
I'm not sure how this patch could have been tested back then, if it
doesn't even compile ;)

The build failure was:
drivers/spi/spi-pl022.c:2292: error: 'adev' undeclared (first use in
this function)
drivers/spi/spi-pl022.c:2344: error: 'pl022_suspend' undeclared here
(not in a function)

The build failure appears e.g. for the u8500 and realview defconfig.
Signed-off-by: NPeter Huewe <peterhuewe@gmx.de>
Signed-off-by: NRussell King <rmk+kernel@arm.linux.org.uk>

Add runtime PM support to the core Primecell driver, following the PCI
model of how this is done.

Rather than having every driver fiddle about with enabling runtime PM,
that's dealt with in the core and instead, drivers just do a put() in
their probe and a balancing get() in their remove function to activate
runtime PM for the device.

As we're dealing with enabling runtime PM in the core, fix up spi-pl022
as it must not enable and disable runtime PM itself anymore.
Tested-by: NLinus Walleij <linus.walleij@linaro.org>
Acked-by: NRafael J. Wysocki <rjw@sisk.pl>
Signed-off-by: NRussell King <rmk+kernel@arm.linux.org.uk>

pl022_dma_remove() should be called only if enable_dma is true. There is no
point calling it when pl022_dma_probe() is not called, which again depends on
enable_dma.
Signed-off-by: NViresh Kumar <viresh.kumar@st.com>
Tested-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

There were few issues with calculate_effective_freq() routine:
- It was returning first rate found >= requested rate. Now, if system have spi's
  rate as 83 MHz, with possible prescaled rates as 83, 41.5, 20.75, 13.83 (as we
  can prescale with multiples of 2). If user has given rate to be programmed as
  22 MHz, then driver programmes it to 41.5 MHz. This looks to be incorrect, as
  user might have given the upper limit of the device, and we are programming it
  above it.
- Driver finds the first satisfying rate and programmes it, but with other
  values of scr & cpsdvsr, it is possible to get more closer rate.

This patch fixes these two issues, with some reformatting inside the code.  This
also creates a inline routine to calculate prescaled rate based on spi's rate,
cpsdvsr and scr.
Signed-off-by: NViresh Kumar <viresh.kumar@st.com>
Tested-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

In routine configure_dma(), if transfer->len = PAGE_SIZE, then pages is one more
than required. While leads to one more sg getting allocated.

This is wrong. Correct this to allocate correct number of sg.
Signed-off-by: NViresh Kumar <viresh.kumar@st.com>
Tested-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

tasklets don't allow invocation to sleeping routines. In configure_dma()
routine, sg_alloc_table() was called with GFP_KERNEL flag and so this causes
crash when called from tasklet.

Replace GFP_KERNEL with GFP_ATOMIC to get this fixed.
Signed-off-by: NViresh Kumar <viresh.kumar@st.com>
Tested-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

There were few formatting related issues in code. This patch fixes them.
Fixes include:
- Remove extra blank lines
- align code to 80 cols
- combine several lines to one line
- Replace multiple spaces with tabs
- Remove spaces before labels
Signed-off-by: NViresh Kumar <viresh.kumar@st.com>
Tested-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

The remove function in the PL022 driver cannot abort the remove
function any way, so restructure the code so as not to make that
assumption. Remove will now proceed no matter whether it can
stop the transfer queue or not.
Reported-by: NRussell King <linux@arm.linux.org.uk>
Signed-off-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

The platform configuration can select custom FIFO watermarks, but
these may conflict the actual FIFO size of the PL022 variant if
set too high. So strengthen the sanity checks to deny any
conflicting settings.
Signed-off-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

Insert notifiers for the runtime PM API. With this the runtime
PM layer kicks in to action where used.
Signed-off-by: NRabin Vincent <rabin.vincent@stericsson.com>
Reviewed-by: NVirupax Sadashivpetimath <virupax.sadashivpetimath@stericsson.com>
Reviewed-by: NJonas Aberg <jonas.aberg@stericsson.com>
Reviewed-by: NSrinidhi Kasagar <srinidhi.kasagar@stericsson.com>
[Rebased to Linux 3.0-rc3, edit description]
Signed-off-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

Configure the DMA burstsize from the FIFO trigger level supplied
with the controller configuration data. This is based on a patch
from Virupax, but I rewrote it differently.
Reported-by: NVirupax Sadashivpetimath <virupax.sadashivpetimath@stericsson.com>
Signed-off-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

Sort the SPI makefile and enforce the naming convention spi_*.c for
spi drivers.

This change also rolls the contents of atmel_spi.h into the .c file
since there is only one user of that particular include file.

v2: - Use 'spi-' prefix instead of 'spi_' to match what seems to be
      be the predominant pattern for subsystem prefixes.
    - Clean up filenames in Kconfig and header comment blocks
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>
Acked-by: NWolfram Sang <w.sang@pengutronix.de>
Acked-by: NLinus Walleij <linus.walleij@linaro.org>

If pl022_dma_probe fails, we can try to transfer data in polling or interrupt
mode. Also, set platform_info->enable_dma to 0, so that no other code tries to
use dma.
Signed-off-by: NViresh Kumar <viresh.kumar@st.com>
Acked-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>

This relaxes the cpu in the polling busy-wait loop.
Reported-by: NVitaly Wool <vitalywool@gmail.com>
Signed-off-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>