No file should be explicitly referencing its own platform headers
by specifying an absolute include path.  Fix these paths to use
standard <asm/arch/...> includes.
Signed-off-by: NRussell King <rmk+kernel@arm.linux.org.uk>

Change to use linux/semaphore.h
Signed-off-by: NMatthew Wilcox <willy@linux.intel.com>

The actual warning is
	arch/arm/mach-ns9xxx/irq.c:65:6: warning: symbol 'handle_prio_irq' was not declared. Should it be static?
Signed-off-by: NUwe Kleine-König <Uwe.Kleine-Koenig@digi.com>

When I copy-adapted handle_level_irq I skipped note_interrupt because
I considered it unimportant.  If I had understand its importance I would
have saved myself some ours of debugging.
Signed-off-by: NUwe Kleine-König <Uwe.Kleine-Koenig@digi.com>

When an irq is reported all lower prio irqs are masked until the current
irq is acked.  So never leave handle_prio_irq without acking.

desc->status & IRQ_INPROGRESS should never become true because the current
irq is masked until it is acked, too.
Signed-off-by: NUwe Kleine-König <Uwe.Kleine-Koenig@digi.com>

Signed-off-by: NUwe Kleine-König <Uwe.Kleine-Koenig@digi.com>

prio2irq(0) is the highest prioritized interrupt.  For now there is no
binary change intended.

If you intend to change the priorisation, you have to assert:

	∀ x ∈ {0, 1, ⋯ 31}: (irq2prio ∘ prio2irq)(x) = x
Signed-off-by: NUwe Kleine-König <Uwe.Kleine-Koenig@digi.com>

The irq controller automatically masks the current and all lower prioritised
tasks until the current irq is acked.
Signed-off-by: NUwe Kleine-König <Uwe.Kleine-Koenig@digi.com>

The hardware team changed some things that were taken as being common to
all ns9xxx processors up to now.

This patch addresses:

- irqs: s/IRQ_/IRQ_NS9360_/

- system module registers: some registers are still general, their
  definition lives now in include/asm-arm/arch-ns9xxx/regs-sys-common.h.
  The ns9360 specific ones are in .../regs-sys-ns9360.h
  As a result ns9360_systemclock cannot be static inline any more as its
  definition needs regs-sys-ns9360.h.  This becomes a real problem when
  adding support for ns9215 as this will need regs-sys-ns9215.h and
  including both files will not work.  For the same reason
  ns9360_reset() is now non-inline and gpio functions live in their own
  file.

- register mapping: s/ns9xxx_map_io/ns9360_map_io/

- timer registers:  move time.c to time-ns9360.c;
  s/ns9xxx_timer/ns9360_timer/
Signed-off-by: NUwe Kleine-König <Uwe.Kleine-Koenig@digi.com>

Moreover now it's the module that selects the processor, and not the machine
anymore.
Signed-off-by: NUwe Kleine-König <Uwe.Kleine-Koenig@digi.com>

Signed-off-by: NUwe Kleine-König <Uwe.Kleine-Koenig@digi.com>

Now the needed structs are allocated dynamically from __init code reducing
memory usage if the kernel runs on a board different from a9m9750dev.
Signed-off-by: NUwe Kleine-König <Uwe.Kleine-Koenig@digi.com>

GPIO_MAX is the number of the last gpio, not the number of gpios.  So
the bitmap must provide GPIO_MAX + 1 bits.
Signed-off-by: NUwe Kleine-König <Uwe.Kleine-Koenig@digi.com>
Signed-off-by: NRussell King <rmk+kernel@arm.linux.org.uk>

As a consequence registers are now accessed with __raw_{read,write}[bl].
Signed-off-by: NUwe Kleine-König <ukleinek@informatik.uni-freiburg.de>
Signed-off-by: NRussell King <rmk+kernel@arm.linux.org.uk>

Up to now only board-a9m9750dev.c used GPIOs.  It just wrote directly into the
corresponding registers.  Now it properly reserves the gpio and uses the API
function to configure it.
Signed-off-by: NUwe Kleine-König <ukleinek@informatik.uni-freiburg.de>
Signed-off-by: NRussell King <rmk+kernel@arm.linux.org.uk>

Signed-off-by: NUwe Kleine-König <ukleinek@informatik.uni-freiburg.de>
Signed-off-by: NRussell King <rmk+kernel@arm.linux.org.uk>

Signed-off-by: NUwe Kleine-König <ukleinek@informatik.uni-freiburg.de>
Signed-off-by: NRussell King <rmk+kernel@arm.linux.org.uk>

This implementation conforms to the general GPIO API
introduced in 2.6.21.

This patch was signed-of by David Brownell before I exported the functions
using EXPORT_SYMBOL.
Signed-off-by: NUwe Kleine-König <ukleinek@informatik.uni-freiburg.de>
Signed-off-by: NRussell King <rmk+kernel@arm.linux.org.uk>

The FPGA on the a9m9750dev board interrupts the CPU via EXT2. So
to acknowledge any FPGA interrupt IRQ_EXT2 must be acknowledged.
Signed-off-by: NUwe Kleine-König <ukleinek@informatik.uni-freiburg.de>
Signed-off-by: NRussell King <rmk+kernel@arm.linux.org.uk>

Now the drivers are responsible to clear the irq in the respective
device, which seems to be the normal thing to do.

So the ack'ing of the timer irq moved to time.c.
Signed-off-by: NUwe Kleine-König <ukleinek@informatik.uni-freiburg.de>
Signed-off-by: NRussell King <rmk+kernel@arm.linux.org.uk>

The HRM states that is must be done this way ...
Signed-off-by: NUwe Kleine-König <ukleinek@informatik.uni-freiburg.de>
Signed-off-by: NRussell King <rmk+kernel@arm.linux.org.uk>

I changed the naming to be more obvious---unfortunately the HRM
doesn't specify these.

Moreover the numbering is changed to be zero indexed as this is more
natural.

Adjust all callers.
Signed-off-by: NUwe Kleine-König <ukleinek@informatik.uni-freiburg.de>
Signed-off-by: NRussell King <rmk+kernel@arm.linux.org.uk>

Fixed all users.
Signed-off-by: NUwe Kleine-König <ukleinek@informatik.uni-freiburg.de>
Signed-off-by: NRussell King <rmk+kernel@arm.linux.org.uk>

Signed-off-by: NUwe Kleine-König <ukleinek@informatik.uni-freiburg.de>
Signed-off-by: NRussell King <rmk+kernel@arm.linux.org.uk>

make ns9xxx_ack_irq_functions static and add one include to get declarations
for ns9xxx_map_io and ns9xxx_init_machine.
Signed-off-by: NUwe Kleine-König <ukleinek@informatik.uni-freiburg.de>
Signed-off-by: NRussell King <rmk+kernel@arm.linux.org.uk>

I must have written the patch introducing support for this machine
deep in the night...
Signed-off-by: NUwe Kleine-König <ukleinek@informatik.uni-freiburg.de>
Signed-off-by: NRussell King <rmk+kernel@arm.linux.org.uk>

Spelling fixes in arch/arm/.
Signed-off-by: NSimon Arlott <simon@fire.lp0.eu>
Signed-off-by: NRussell King <rmk+kernel@arm.linux.org.uk>

Add IRQF_IRQPOLL for each timer interrupt.
Signed-off-by: NBernhard Walle <bwalle@suse.de>
Cc: Russell King <rmk@arm.linux.org.uk>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

The support for that machine is not yet complete, but it's enough to
be useful as a test platform for the serial and ethernet driver.

Moreover a typo in the product name is fixed that I missed in the
last patch.
Signed-off-by: NUwe Kleine-König <ukleinek@informatik.uni-freiburg.de>
Signed-off-by: NRussell King <rmk+kernel@arm.linux.org.uk>

The Product Manager of the cc9p insist on using the correct product names.
Signed-off-by: NUwe Kleine-König <ukleinek@informatik.uni-freiburg.de>
Signed-off-by: NRussell King <rmk+kernel@arm.linux.org.uk>

Signed-off-by: NUwe Kleine-König <ukleinek@informatik.uni-freiburg.de>
Signed-off-by: NRussell King <rmk+kernel@arm.linux.org.uk>